Cleveland Clinic Journal of Medicine

This centennial year update focuses primarily on immunization, but also reviews epidemiology, transmission, and treatment.

EPIDEMIOLOGY

2017–2018 was a bad season

The 2017–2018 influenza epidemic was memorable, dominated by influenza A(H3N2) viruses with morbidity and mortality rates approaching pandemic numbers. It lasted 19 weeks, killed more people than any other epidemic since 2010, particularly children, and was associated with 30,453 hospitalizations—almost twice the previous season high in some parts of the United States.²

Regrettably, 171 unvaccinated children died during 2017–2018, accounting for almost 80% of deaths.² The mean age of the children who died was 7.1 years; 51% had at least 1 underlying medical condition placing them at risk for influenza-related complications, and 57% died after hospitalization.²

Recent estimates of the incidence of symptomatic influenza among all ages ranged from 3% to 11%, which is slightly lower than historical estimates. The rates were higher for children under age 18 than for adults.³ Interestingly, influenza A(H3N2) accounted for 50% of cases of non-mumps viral parotitis during the 2014–2015 influenza season in the United States.⁴

Influenza C exists but is rare

Influenza A and B account for almost all influenza-related outpatient visits and hospitalizations. Surveillance data from May 2013 through December 2016 showed that influenza C accounts for 0.5% of influenza-related outpatient visits and hospitalizations, particularly affecting children ages 6 to 24 months. Medical comorbidities and copathogens were seen in all patients requiring intensive care and in most hospitalizations.⁵ Diagnostic tests for influenza C are not widely available.

Dogs and cats: Factories for new flu strains?

While pigs and birds are the major reservoirs of influenza viral genetic diversity from which infection is transmitted to humans, dogs and cats have recently emerged as possible sources of novel reassortant influenza A.⁶ With their frequent close contact with humans, our pets may prove to pose a significant threat.

Obesity a risk factor for influenza

Obesity emerged as a risk factor for severe influenza in the 2009 pandemic. Recent data also showed that obesity increases the duration of influenza A virus shedding, thus increasing duration of contagiousness.⁷

Influenza a cardiovascular risk factor

Previous data showed that influenza was a risk factor for cardiovascular events. Two recent epi­demiologic studies from the United Kingdom showed that laboratory-confirmed influenza was associated with higher rates of myocardial infarction and stroke for up to 4 weeks.^8,9

Which strain is the biggest threat?

Predicting which emerging influenza serotype may cause the next pandemic is difficult, but influenza A(H7N9), which had not infected humans until 2013 but has since infected about 1,600 people in China and killed 37% of them, appears to have the greatest potential.¹⁰

National influenza surveillance programs and influenza-related social media applications have been developed and may get a boost from technology. A smartphone equipped with a temperature sensor can instantly detect one’s temperature with great precision. A 2018 study suggested that a smartphone-driven thermometry application correlated well with national influenza-like illness activity and improved its forecast in real time and up to 3 weeks in advance.¹¹

TRANSMISSION

Humidity may not block transmission

Animal studies have suggested that humidity in the air interferes with transmission of airborne influenza virus, partially from biologic inactivation. But when a recent study used humidity-controlled chambers to investigate the stability of the 2009 influenza A(H1N1) virus in suspended aerosols and stationary droplets, the virus remained infectious in aerosols across a wide range of relative humidities, challenging the common belief that humidity destabilizes respiratory viruses in aerosols.¹²

One sick passenger may not infect the whole plane

Transmission of respiratory viruses on airplane flights has long been considered a potential avenue for spreading influenza. However, a recent study that monitored movements of individuals on 10 transcontinental US flights and simulated inflight transmission based on these data showed a low probability of direct transmission, except for passengers seated in close proximity to an infectious passenger.¹³

WHAT’S IN THE NEW FLU SHOT?

The 2018–2019 quadrivalent vaccine for the Northern Hemisphere¹⁴ contains the following strains:

• A/Michigan/45/2015 A(H1N1)pdm09-like virus
• A/Singapore/INFIMH-16-0019/2016 (H3N2)-like virus
• B/Colorado/06/2017-like virus (Victoria lineage)
• B/Phuket/3073/2013-like virus (Yamagata lineage).

The A(H3N2) (Singapore) and B/Victoria lineage components are new this year. The A(H3N2) strain was the main cause of the 2018 influenza epidemic in the Southern Hemisphere.

The quadrivalent live-attenuated vaccine, which was not recommended during the 2016–2017 and 2017–2018 influenza seasons, has made a comeback and is recommended for the 2018–2019 season in people for whom it is appropriate based on age and comorbidities.¹⁵ Although it was effective against influenza B and A(H3N2) viruses, it was less effective against the influenza A(H1N1)pdm09-like viruses during the 2013–2014 and 2015–2016 seasons.

A/Slovenia/2903/2015, the new A(H1N1)pdm09-like virus included in the 2018–2019 quadrivalent live-attenuated vaccine, is significantly more immunogenic than its predecessor, A/Bolivia/559/2013, but its clinical effectiveness remains to be seen.

How effective is it?

Influenza vaccine effectiveness in the 2017–2018 influenza season was 36% overall, 67% against A(H1N1), 42% against influenza B, and 25% against A(H3N2).¹⁶ It is estimated that influenza vaccine prevents 300 to 4,000 deaths annually in the United States alone.¹⁷

A 2018 Cochrane review¹⁷ concluded that vaccination reduced the incidence of influenza by about half, with 2.3% of the population contracting the flu without vaccination compared with 0.9% with vaccination (risk ratio 0.41, 95% confidence interval 0.36–0.47). The same review found that 71 healthy adults need to be vaccinated to prevent 1 from experiencing influenza, and 29 to prevent 1 influenza-like illness.

Several recent studies showed that influenza vaccine effectiveness varied based on age and influenza serotype, with higher effectiveness in people ages 5 to 17 and ages 18 to 64 than in those age 65 and older.^18–20 A mathematical model of influenza transmission and vaccination in the United States determined that even relatively low-efficacy influenza vaccines can be very useful if optimally distributed across age groups.²¹

Vaccination rates are low, and ‘antivaxxers’ are on the rise

Although the influenza vaccine is recommended in the United States for all people age 6 months and older regardless of the state of their health, vaccination rates remain low. In 2016, only 37% of employed adults were vaccinated. The highest rate was for government employees (45%), followed by private employees (36%), followed by the self-employed (30%).²²

A national goal is to immunize 80% of all Americans and 90% of at-risk populations (which include children and the elderly).²³ The number of US hospitals that require their employees to be vaccinated increased from 37.1% in 2013 to 61.4% in 2017.²⁴ Regrettably, as of March 2018, 14 lawsuits addressing religious objections to hospital influenza vaccination mandates have been filed.²⁵

Despite hundreds of studies demonstrating the efficacy, safety, and cost savings of influenza vaccination, the antivaccine movement has been growing in the United States and worldwide.²⁶ All US states except West Virginia, Mississippi, and California allow nonmedical exemptions from vaccination based on religious or personal belief.²⁷ Several US metropolitan areas represent “hot spots” for these exemptions.²⁸ This may render such areas vulnerable to vaccine-preventable diseases, including influenza.

Herd immunity: We’re all in this together

Some argue that the potential adverse effects and the cost of vaccination outweigh the benefits, but the protective benefits of herd immunity are significant for those with comorbidities or compromised immunity.

Educating the public about herd immunity and local influenza vaccination uptake increases people’s willingness to be vaccinated.²⁹ A key educational point is that at least 70% of a community needs to be vaccinated to prevent community outbreaks; this protects everyone, including those who do not mount a protective antibody response to influenza vaccination and those who are not vaccinated.

DOES ANNUAL VACCINATION BLUNT ITS EFFECTIVENESS?

Some studies from the 1970s and 1980s raised concern over a possible negative effect of annual influenza vaccination on vaccine effectiveness. The “antigenic distance hypothesis” holds that vaccine effectiveness is influenced by antigenic similarity between the previous season’s vaccine serotypes and the epidemic serotypes, as well as the antigenic similarity between the serotypes of the current and previous seasons.

A meta-analysis of studies from 2010 through 2015 showed significant inconsistencies in repeat vaccination effects within and between seasons and serotypes. It also showed that vaccine effectiveness may be influenced by more than 1 previous season, particularly for influenza A(H3N2), in which repeated vaccination can blunt the hemagglutinin antibody response.³⁰

A study from Japan showed that people who needed medical attention for influenza in the previous season were at lower risk of a similar event in the current season.³¹ Prior-season influenza vaccination reduced current-season vaccine effectiveness only in those who did not have medically attended influenza in the prior season. This suggests that infection is more immunogenic than vaccination, but only against the serotype causing the infection and not the other serotypes included in the vaccine.

An Australian study showed that annual influenza vaccination did not decrease vaccine effectiveness against influenza-associated hospitalization. Rather, effectiveness increased by about 15% in those vaccinated in both current and previous seasons compared with those vaccinated in either season alone.³²

European investigators showed that repeated seasonal influenza vaccination in the elderly prevented the need for hospitalization due to influenza A(H3N2) and B, but not A(H1N1)pdm09.³³

High-dose vaccine for older adults

The high-dose influenza vaccine has been licensed since 2009 for use in the United States for people ages 65 and older.

Recent studies confirmed that high-dose vaccine is more effective than standard-dose vaccine in veterans³⁴ and US Medicare beneficiaries.³⁵

The high-dose vaccine is rapidly becoming the primary vaccine given to people ages 65 and older in retail pharmacies, where vaccination begins earlier in the season than in providers’ offices.³⁶ Some studies have shown that the standard-dose vaccine wanes in effectiveness toward the end of the influenza season (particularly if the season is long) if it is given very early. It remains to be seen whether the same applies to the high-dose influenza vaccine.

Some advocate twice-annual influenza vaccination, particularly for older adults living in tropical and subtropical areas, where influenza seasons are more prolonged. However, a recently published study observed reductions in influenza-specific hemagglutination inhibition and cell-mediated immunity after twice-annual vaccination.³⁷

Vaccination is beneficial during pregnancy

Many studies have shown the value of influenza vaccination during pregnancy for both mothers and their infants.

One recently published study showed that 18% of infants who developed influenza required hospitalization.³⁸ In that study, prenatal and postpartum maternal influenza vaccination decreased the odds of influenza in infants by 61% and 53%, respectively.

Another study showed that vaccine effectiveness did not vary by gestational age at vaccination.³⁹

Some studies have shown that influenza virus infection can increase susceptibility to certain bacterial infections. A post hoc analysis of an influenza vaccination study in pregnant women suggested that the vaccine was also associated with decreased rates of pertussis in these women.⁴⁰

Factors that make vaccination less effective

Several factors including age-related frailty and iatrogenic and disease-related immunosuppression can affect vaccine effectiveness.

Frailty. A recent study showed that vaccine effectiveness was 77.6% in nonfrail older adults but only 58.7% in frail older adults.⁴¹

Immunosuppression. Temporary discontinuation of methotrexate for 2 weeks after influenza vaccination in patients with rheumatoid arthritis improves vaccine immunogenicity without precipitating disease flare.⁴² Solid-organ and hematopoietic stem cell transplant recipients who received influenza vaccine were less likely to develop pneumonia and require intensive care unit admission.⁴³

The high-dose influenza vaccine is more immunogenic than the standard-dose vaccine in solid-organ transplant recipients.⁴⁴

Statins are widely prescribed and have recently been associated with reduced influenza vaccine effectiveness against medically attended acute respiratory illness, but their benefits in preventing cardiovascular events outweigh this risk.⁴⁵

FUTURE VACCINE CONSIDERATIONS

Moving away from eggs

During the annual egg-based production process, which takes several months, the influenza vaccine acquires antigenic changes that allow replication in eggs, particularly in the hemagglutinin protein, which mediates receptor binding. This process of egg adaptation may cause antigenic changes that decrease vaccine effectiveness against circulating viruses.

The cell-based baculovirus influenza vaccine grown in dog kidney cells has higher antigenic content and is not subject to the limitations of egg-based vaccine, although it still requires annual updates. A recombinant influenza vaccine reduces the probability of influenza-like illness by 30% compared with the egg-based influenza vaccine, but also still requires annual updates.⁴⁶ The market share of these non-egg-based vaccines is small, and thus their effectiveness has yet to be demonstrated.

The US Department of Defense administered the cell-based influenza vaccine to about one-third of Armed Forces personnel, their families, and retirees in the 2017–2018 influenza seasons, and data on its effectiveness are expected in the near future.⁴⁷

A universal vaccine would be ideal

The quest continues for a universal influenza vaccine, one that remains protective for several years and does not require annual updates.⁴⁸ Such a vaccine would protect against seasonal epidemic influenza drift variants and pandemic strains. More people could likely be persuaded to be vaccinated once rather than every year.

David Schumick, medical illustrator; adapted from reference 49; illustration of the influenza hemagglutinin created in part with data from reference 50.

An ideal universal vaccine would be suitable for all age groups, at least 75% effective against symptomatic influenza virus infection, protective against all influenza A viruses (influenza A, not B, causes pandemics and seasonal epidemics), and durable through multiple influenza seasons.⁵¹

Research and production of such a vaccine are expected to require funding of about $1 billion over the next 5 years.

Boosting effectiveness

Estimates of influenza vaccine effectiveness range from 40% to 60% in years when the vaccine viruses closely match the circulating viruses, and variably lower when they do not match. The efficacy of most other vaccines given to prevent other infections is much higher.

New technologies to improve influenza vaccine effectiveness are needed, particularly for influenza A(H3N2) viruses, which are rapidly evolving and are highly susceptible to egg-adaptive mutations in the manufacturing process.

In one study, a nanoparticle vaccine formulated with a saponin-based adjuvant induced hemagglutination inhibition responses that were even greater than those induced by the high-dose vaccine.⁵²

Immunoglobulin A (IgA) may be a more effective vaccine target than traditional influenza vaccines that target IgG, since different parts of IgA may engage the influenza virus simultaneously.⁵³

Vaccines can be developed more quickly than in the past. The timeline from viral sequencing to human studies with deoxyribonucleic acid plasmid vaccines decreased from 20 months in 2003 for the severe acquired respiratory syndrome coronavirus to 11 months in 2006 for influenza A/Indonesia/2006 (H5), to 4 months in 2009 for influenza A/California/2009 (H1), to 3.5 months in 2016 for Zika virus.⁵⁴ This is because it is possible today to sequence a virus and insert the genetic material into a vaccine platform without ever having to grow the virus.

Numerous studies have found anti-influenza medications to be effective. Nevertheless, in an analysis of the 2011–2016 influenza seasons, only 15% of high-risk patients were prescribed anti-influenza medications within 2 days of symptom onset, including 37% in those with laboratory-confirmed influenza.⁵⁵ Fever was associated with an increased rate of antiviral treatment, but 25% of high-risk outpatients were afebrile. Empiric treatment of 4 high-risk outpatients with acute respiratory illness was needed to treat 1 patient with influenza.⁵⁵

Treatment with a neuraminidase inhibitor within 2 days of illness has recently been shown to improve survival and shorten duration of viral shedding in patients with avian influenza A(H7N9) infection.⁵⁶ Antiviral treatment within 2 days of illness is associated with improved outcomes in transplant recipients⁵⁷ and with a lower risk of otitis media in children.⁵⁸

Appropriate anti-influenza treatment is as important as avoiding unnecessary antibiotics. Regrettably, as many as one-third of patients with laboratory-confirmed influenza are prescribed antibiotics.⁵⁹

The US Food and Drug Administration warns against fraudulent unapproved over-the-counter influenza products.⁶⁰

Baloxavir marboxil

Baloxavir marboxil is a new anti-influenza medication approved in Japan in February 2018 and anticipated to be available in the United States sometime in 2019.

This prodrug is hydrolyzed in vivo to the active metabolite, which selectively inhibits cap-dependent endonuclease enzyme, a key enzyme in initiation of messenger ribonucleic acid synthesis required for influenza viral replication.⁶¹

In a double-blind phase 3 trial, the median time to alleviation of influenza symptoms is 26.5 hours shorter with baloxavir marboxil than with placebo. One tablet was as effective as 5 days of the neuraminidase inhibitor oseltamivir and was associated with greater reduction in viral load 1 day after initiation, and similar side effects.⁶² Of concern is the emergence of nucleic acid substitutions conferring resistance to baloxavir; this occurred in 2.2% and 9.7% of baloxavir recipients in the phase 2 and 3 trials, respectively.

CLOSING THE GAPS

Several gaps in the management of influenza persist since the 1918 pandemic.¹ These include gaps in epidemiology, prevention, diagnosis, treatment, and prognosis.

• Global networks wider than current ones are needed to address this global disease and to prioritize coordination efforts.
• Establishing and strengthening clinical capacity is needed in limited resource settings. New technologies are needed to expedite vaccine development and to achieve progress toward a universal vaccine.
• Current diagnostic tests do not distinguish between seasonal and novel influenza A viruses of zoonotic origin, which are expected to cause the next pandemic.
• Current antivirals have been shown to shorten duration of illness in outpatients with uncomplicated influenza, but the benefit in hospitalized patients has been less well established.
• In 2007, resistance of seasonal influenza A(H1N1) to oseltamivir became widespread. In 2009, pandemic influenza A(H1N1), which is highly susceptible to oseltamivir, replaced the seasonal virus and remains the predominantly circulating A(H1N1) strain.
• A small-molecule fragment, N-cyclohexyaltaurine, binds to the conserved hemagglutinin receptor-binding site in a manner that mimics the binding mode of the natural receptor sialic acid. This can serve as a template to guide the development of novel broad-spectrum small-molecule anti-influenza drugs.⁶³
• Biomarkers that can accurately predict development of severe disease in patients with influenza are needed.

This centennial year update focuses primarily on immunization, but also reviews epidemiology, transmission, and treatment.

EPIDEMIOLOGY

2017–2018 was a bad season

The 2017–2018 influenza epidemic was memorable, dominated by influenza A(H3N2) viruses with morbidity and mortality rates approaching pandemic numbers. It lasted 19 weeks, killed more people than any other epidemic since 2010, particularly children, and was associated with 30,453 hospitalizations—almost twice the previous season high in some parts of the United States.²

Regrettably, 171 unvaccinated children died during 2017–2018, accounting for almost 80% of deaths.² The mean age of the children who died was 7.1 years; 51% had at least 1 underlying medical condition placing them at risk for influenza-related complications, and 57% died after hospitalization.²

Recent estimates of the incidence of symptomatic influenza among all ages ranged from 3% to 11%, which is slightly lower than historical estimates. The rates were higher for children under age 18 than for adults.³ Interestingly, influenza A(H3N2) accounted for 50% of cases of non-mumps viral parotitis during the 2014–2015 influenza season in the United States.⁴

Influenza C exists but is rare

Influenza A and B account for almost all influenza-related outpatient visits and hospitalizations. Surveillance data from May 2013 through December 2016 showed that influenza C accounts for 0.5% of influenza-related outpatient visits and hospitalizations, particularly affecting children ages 6 to 24 months. Medical comorbidities and copathogens were seen in all patients requiring intensive care and in most hospitalizations.⁵ Diagnostic tests for influenza C are not widely available.

Dogs and cats: Factories for new flu strains?

While pigs and birds are the major reservoirs of influenza viral genetic diversity from which infection is transmitted to humans, dogs and cats have recently emerged as possible sources of novel reassortant influenza A.⁶ With their frequent close contact with humans, our pets may prove to pose a significant threat.

Obesity a risk factor for influenza

Obesity emerged as a risk factor for severe influenza in the 2009 pandemic. Recent data also showed that obesity increases the duration of influenza A virus shedding, thus increasing duration of contagiousness.⁷

Influenza a cardiovascular risk factor

Previous data showed that influenza was a risk factor for cardiovascular events. Two recent epi­demiologic studies from the United Kingdom showed that laboratory-confirmed influenza was associated with higher rates of myocardial infarction and stroke for up to 4 weeks.^8,9

Which strain is the biggest threat?

Predicting which emerging influenza serotype may cause the next pandemic is difficult, but influenza A(H7N9), which had not infected humans until 2013 but has since infected about 1,600 people in China and killed 37% of them, appears to have the greatest potential.¹⁰

National influenza surveillance programs and influenza-related social media applications have been developed and may get a boost from technology. A smartphone equipped with a temperature sensor can instantly detect one’s temperature with great precision. A 2018 study suggested that a smartphone-driven thermometry application correlated well with national influenza-like illness activity and improved its forecast in real time and up to 3 weeks in advance.¹¹

TRANSMISSION

Humidity may not block transmission

Animal studies have suggested that humidity in the air interferes with transmission of airborne influenza virus, partially from biologic inactivation. But when a recent study used humidity-controlled chambers to investigate the stability of the 2009 influenza A(H1N1) virus in suspended aerosols and stationary droplets, the virus remained infectious in aerosols across a wide range of relative humidities, challenging the common belief that humidity destabilizes respiratory viruses in aerosols.¹²

One sick passenger may not infect the whole plane

Transmission of respiratory viruses on airplane flights has long been considered a potential avenue for spreading influenza. However, a recent study that monitored movements of individuals on 10 transcontinental US flights and simulated inflight transmission based on these data showed a low probability of direct transmission, except for passengers seated in close proximity to an infectious passenger.¹³

WHAT’S IN THE NEW FLU SHOT?

The 2018–2019 quadrivalent vaccine for the Northern Hemisphere¹⁴ contains the following strains:

• A/Michigan/45/2015 A(H1N1)pdm09-like virus
• A/Singapore/INFIMH-16-0019/2016 (H3N2)-like virus
• B/Colorado/06/2017-like virus (Victoria lineage)
• B/Phuket/3073/2013-like virus (Yamagata lineage).

The A(H3N2) (Singapore) and B/Victoria lineage components are new this year. The A(H3N2) strain was the main cause of the 2018 influenza epidemic in the Southern Hemisphere.

The quadrivalent live-attenuated vaccine, which was not recommended during the 2016–2017 and 2017–2018 influenza seasons, has made a comeback and is recommended for the 2018–2019 season in people for whom it is appropriate based on age and comorbidities.¹⁵ Although it was effective against influenza B and A(H3N2) viruses, it was less effective against the influenza A(H1N1)pdm09-like viruses during the 2013–2014 and 2015–2016 seasons.

A/Slovenia/2903/2015, the new A(H1N1)pdm09-like virus included in the 2018–2019 quadrivalent live-attenuated vaccine, is significantly more immunogenic than its predecessor, A/Bolivia/559/2013, but its clinical effectiveness remains to be seen.

How effective is it?

Influenza vaccine effectiveness in the 2017–2018 influenza season was 36% overall, 67% against A(H1N1), 42% against influenza B, and 25% against A(H3N2).¹⁶ It is estimated that influenza vaccine prevents 300 to 4,000 deaths annually in the United States alone.¹⁷

A 2018 Cochrane review¹⁷ concluded that vaccination reduced the incidence of influenza by about half, with 2.3% of the population contracting the flu without vaccination compared with 0.9% with vaccination (risk ratio 0.41, 95% confidence interval 0.36–0.47). The same review found that 71 healthy adults need to be vaccinated to prevent 1 from experiencing influenza, and 29 to prevent 1 influenza-like illness.

Several recent studies showed that influenza vaccine effectiveness varied based on age and influenza serotype, with higher effectiveness in people ages 5 to 17 and ages 18 to 64 than in those age 65 and older.^18–20 A mathematical model of influenza transmission and vaccination in the United States determined that even relatively low-efficacy influenza vaccines can be very useful if optimally distributed across age groups.²¹

Vaccination rates are low, and ‘antivaxxers’ are on the rise

Although the influenza vaccine is recommended in the United States for all people age 6 months and older regardless of the state of their health, vaccination rates remain low. In 2016, only 37% of employed adults were vaccinated. The highest rate was for government employees (45%), followed by private employees (36%), followed by the self-employed (30%).²²

A national goal is to immunize 80% of all Americans and 90% of at-risk populations (which include children and the elderly).²³ The number of US hospitals that require their employees to be vaccinated increased from 37.1% in 2013 to 61.4% in 2017.²⁴ Regrettably, as of March 2018, 14 lawsuits addressing religious objections to hospital influenza vaccination mandates have been filed.²⁵

Despite hundreds of studies demonstrating the efficacy, safety, and cost savings of influenza vaccination, the antivaccine movement has been growing in the United States and worldwide.²⁶ All US states except West Virginia, Mississippi, and California allow nonmedical exemptions from vaccination based on religious or personal belief.²⁷ Several US metropolitan areas represent “hot spots” for these exemptions.²⁸ This may render such areas vulnerable to vaccine-preventable diseases, including influenza.

Herd immunity: We’re all in this together

Some argue that the potential adverse effects and the cost of vaccination outweigh the benefits, but the protective benefits of herd immunity are significant for those with comorbidities or compromised immunity.

Educating the public about herd immunity and local influenza vaccination uptake increases people’s willingness to be vaccinated.²⁹ A key educational point is that at least 70% of a community needs to be vaccinated to prevent community outbreaks; this protects everyone, including those who do not mount a protective antibody response to influenza vaccination and those who are not vaccinated.

DOES ANNUAL VACCINATION BLUNT ITS EFFECTIVENESS?

Some studies from the 1970s and 1980s raised concern over a possible negative effect of annual influenza vaccination on vaccine effectiveness. The “antigenic distance hypothesis” holds that vaccine effectiveness is influenced by antigenic similarity between the previous season’s vaccine serotypes and the epidemic serotypes, as well as the antigenic similarity between the serotypes of the current and previous seasons.

A meta-analysis of studies from 2010 through 2015 showed significant inconsistencies in repeat vaccination effects within and between seasons and serotypes. It also showed that vaccine effectiveness may be influenced by more than 1 previous season, particularly for influenza A(H3N2), in which repeated vaccination can blunt the hemagglutinin antibody response.³⁰

A study from Japan showed that people who needed medical attention for influenza in the previous season were at lower risk of a similar event in the current season.³¹ Prior-season influenza vaccination reduced current-season vaccine effectiveness only in those who did not have medically attended influenza in the prior season. This suggests that infection is more immunogenic than vaccination, but only against the serotype causing the infection and not the other serotypes included in the vaccine.

An Australian study showed that annual influenza vaccination did not decrease vaccine effectiveness against influenza-associated hospitalization. Rather, effectiveness increased by about 15% in those vaccinated in both current and previous seasons compared with those vaccinated in either season alone.³²

European investigators showed that repeated seasonal influenza vaccination in the elderly prevented the need for hospitalization due to influenza A(H3N2) and B, but not A(H1N1)pdm09.³³

High-dose vaccine for older adults

The high-dose influenza vaccine has been licensed since 2009 for use in the United States for people ages 65 and older.

Recent studies confirmed that high-dose vaccine is more effective than standard-dose vaccine in veterans³⁴ and US Medicare beneficiaries.³⁵

The high-dose vaccine is rapidly becoming the primary vaccine given to people ages 65 and older in retail pharmacies, where vaccination begins earlier in the season than in providers’ offices.³⁶ Some studies have shown that the standard-dose vaccine wanes in effectiveness toward the end of the influenza season (particularly if the season is long) if it is given very early. It remains to be seen whether the same applies to the high-dose influenza vaccine.

Some advocate twice-annual influenza vaccination, particularly for older adults living in tropical and subtropical areas, where influenza seasons are more prolonged. However, a recently published study observed reductions in influenza-specific hemagglutination inhibition and cell-mediated immunity after twice-annual vaccination.³⁷

Vaccination is beneficial during pregnancy

Many studies have shown the value of influenza vaccination during pregnancy for both mothers and their infants.

One recently published study showed that 18% of infants who developed influenza required hospitalization.³⁸ In that study, prenatal and postpartum maternal influenza vaccination decreased the odds of influenza in infants by 61% and 53%, respectively.

Another study showed that vaccine effectiveness did not vary by gestational age at vaccination.³⁹

Some studies have shown that influenza virus infection can increase susceptibility to certain bacterial infections. A post hoc analysis of an influenza vaccination study in pregnant women suggested that the vaccine was also associated with decreased rates of pertussis in these women.⁴⁰

Factors that make vaccination less effective

Several factors including age-related frailty and iatrogenic and disease-related immunosuppression can affect vaccine effectiveness.

Frailty. A recent study showed that vaccine effectiveness was 77.6% in nonfrail older adults but only 58.7% in frail older adults.⁴¹

Immunosuppression. Temporary discontinuation of methotrexate for 2 weeks after influenza vaccination in patients with rheumatoid arthritis improves vaccine immunogenicity without precipitating disease flare.⁴² Solid-organ and hematopoietic stem cell transplant recipients who received influenza vaccine were less likely to develop pneumonia and require intensive care unit admission.⁴³

The high-dose influenza vaccine is more immunogenic than the standard-dose vaccine in solid-organ transplant recipients.⁴⁴

Statins are widely prescribed and have recently been associated with reduced influenza vaccine effectiveness against medically attended acute respiratory illness, but their benefits in preventing cardiovascular events outweigh this risk.⁴⁵

FUTURE VACCINE CONSIDERATIONS

Moving away from eggs

During the annual egg-based production process, which takes several months, the influenza vaccine acquires antigenic changes that allow replication in eggs, particularly in the hemagglutinin protein, which mediates receptor binding. This process of egg adaptation may cause antigenic changes that decrease vaccine effectiveness against circulating viruses.

The cell-based baculovirus influenza vaccine grown in dog kidney cells has higher antigenic content and is not subject to the limitations of egg-based vaccine, although it still requires annual updates. A recombinant influenza vaccine reduces the probability of influenza-like illness by 30% compared with the egg-based influenza vaccine, but also still requires annual updates.⁴⁶ The market share of these non-egg-based vaccines is small, and thus their effectiveness has yet to be demonstrated.

The US Department of Defense administered the cell-based influenza vaccine to about one-third of Armed Forces personnel, their families, and retirees in the 2017–2018 influenza seasons, and data on its effectiveness are expected in the near future.⁴⁷

A universal vaccine would be ideal

The quest continues for a universal influenza vaccine, one that remains protective for several years and does not require annual updates.⁴⁸ Such a vaccine would protect against seasonal epidemic influenza drift variants and pandemic strains. More people could likely be persuaded to be vaccinated once rather than every year.

David Schumick, medical illustrator; adapted from reference 49; illustration of the influenza hemagglutinin created in part with data from reference 50.

An ideal universal vaccine would be suitable for all age groups, at least 75% effective against symptomatic influenza virus infection, protective against all influenza A viruses (influenza A, not B, causes pandemics and seasonal epidemics), and durable through multiple influenza seasons.⁵¹

Research and production of such a vaccine are expected to require funding of about $1 billion over the next 5 years.

Boosting effectiveness

Estimates of influenza vaccine effectiveness range from 40% to 60% in years when the vaccine viruses closely match the circulating viruses, and variably lower when they do not match. The efficacy of most other vaccines given to prevent other infections is much higher.

New technologies to improve influenza vaccine effectiveness are needed, particularly for influenza A(H3N2) viruses, which are rapidly evolving and are highly susceptible to egg-adaptive mutations in the manufacturing process.

In one study, a nanoparticle vaccine formulated with a saponin-based adjuvant induced hemagglutination inhibition responses that were even greater than those induced by the high-dose vaccine.⁵²

Immunoglobulin A (IgA) may be a more effective vaccine target than traditional influenza vaccines that target IgG, since different parts of IgA may engage the influenza virus simultaneously.⁵³

Vaccines can be developed more quickly than in the past. The timeline from viral sequencing to human studies with deoxyribonucleic acid plasmid vaccines decreased from 20 months in 2003 for the severe acquired respiratory syndrome coronavirus to 11 months in 2006 for influenza A/Indonesia/2006 (H5), to 4 months in 2009 for influenza A/California/2009 (H1), to 3.5 months in 2016 for Zika virus.⁵⁴ This is because it is possible today to sequence a virus and insert the genetic material into a vaccine platform without ever having to grow the virus.

Numerous studies have found anti-influenza medications to be effective. Nevertheless, in an analysis of the 2011–2016 influenza seasons, only 15% of high-risk patients were prescribed anti-influenza medications within 2 days of symptom onset, including 37% in those with laboratory-confirmed influenza.⁵⁵ Fever was associated with an increased rate of antiviral treatment, but 25% of high-risk outpatients were afebrile. Empiric treatment of 4 high-risk outpatients with acute respiratory illness was needed to treat 1 patient with influenza.⁵⁵

Treatment with a neuraminidase inhibitor within 2 days of illness has recently been shown to improve survival and shorten duration of viral shedding in patients with avian influenza A(H7N9) infection.⁵⁶ Antiviral treatment within 2 days of illness is associated with improved outcomes in transplant recipients⁵⁷ and with a lower risk of otitis media in children.⁵⁸

Appropriate anti-influenza treatment is as important as avoiding unnecessary antibiotics. Regrettably, as many as one-third of patients with laboratory-confirmed influenza are prescribed antibiotics.⁵⁹

The US Food and Drug Administration warns against fraudulent unapproved over-the-counter influenza products.⁶⁰

Baloxavir marboxil

Baloxavir marboxil is a new anti-influenza medication approved in Japan in February 2018 and anticipated to be available in the United States sometime in 2019.

This prodrug is hydrolyzed in vivo to the active metabolite, which selectively inhibits cap-dependent endonuclease enzyme, a key enzyme in initiation of messenger ribonucleic acid synthesis required for influenza viral replication.⁶¹

In a double-blind phase 3 trial, the median time to alleviation of influenza symptoms is 26.5 hours shorter with baloxavir marboxil than with placebo. One tablet was as effective as 5 days of the neuraminidase inhibitor oseltamivir and was associated with greater reduction in viral load 1 day after initiation, and similar side effects.⁶² Of concern is the emergence of nucleic acid substitutions conferring resistance to baloxavir; this occurred in 2.2% and 9.7% of baloxavir recipients in the phase 2 and 3 trials, respectively.

CLOSING THE GAPS

Several gaps in the management of influenza persist since the 1918 pandemic.¹ These include gaps in epidemiology, prevention, diagnosis, treatment, and prognosis.

• Global networks wider than current ones are needed to address this global disease and to prioritize coordination efforts.
• Establishing and strengthening clinical capacity is needed in limited resource settings. New technologies are needed to expedite vaccine development and to achieve progress toward a universal vaccine.
• Current diagnostic tests do not distinguish between seasonal and novel influenza A viruses of zoonotic origin, which are expected to cause the next pandemic.
• Current antivirals have been shown to shorten duration of illness in outpatients with uncomplicated influenza, but the benefit in hospitalized patients has been less well established.
• In 2007, resistance of seasonal influenza A(H1N1) to oseltamivir became widespread. In 2009, pandemic influenza A(H1N1), which is highly susceptible to oseltamivir, replaced the seasonal virus and remains the predominantly circulating A(H1N1) strain.
• A small-molecule fragment, N-cyclohexyaltaurine, binds to the conserved hemagglutinin receptor-binding site in a manner that mimics the binding mode of the natural receptor sialic acid. This can serve as a template to guide the development of novel broad-spectrum small-molecule anti-influenza drugs.⁶³
• Biomarkers that can accurately predict development of severe disease in patients with influenza are needed.

This centennial year update focuses primarily on immunization, but also reviews epidemiology, transmission, and treatment.

EPIDEMIOLOGY

2017–2018 was a bad season

The 2017–2018 influenza epidemic was memorable, dominated by influenza A(H3N2) viruses with morbidity and mortality rates approaching pandemic numbers. It lasted 19 weeks, killed more people than any other epidemic since 2010, particularly children, and was associated with 30,453 hospitalizations—almost twice the previous season high in some parts of the United States.²

Regrettably, 171 unvaccinated children died during 2017–2018, accounting for almost 80% of deaths.² The mean age of the children who died was 7.1 years; 51% had at least 1 underlying medical condition placing them at risk for influenza-related complications, and 57% died after hospitalization.²

Recent estimates of the incidence of symptomatic influenza among all ages ranged from 3% to 11%, which is slightly lower than historical estimates. The rates were higher for children under age 18 than for adults.³ Interestingly, influenza A(H3N2) accounted for 50% of cases of non-mumps viral parotitis during the 2014–2015 influenza season in the United States.⁴

Influenza C exists but is rare

Influenza A and B account for almost all influenza-related outpatient visits and hospitalizations. Surveillance data from May 2013 through December 2016 showed that influenza C accounts for 0.5% of influenza-related outpatient visits and hospitalizations, particularly affecting children ages 6 to 24 months. Medical comorbidities and copathogens were seen in all patients requiring intensive care and in most hospitalizations.⁵ Diagnostic tests for influenza C are not widely available.

Dogs and cats: Factories for new flu strains?

While pigs and birds are the major reservoirs of influenza viral genetic diversity from which infection is transmitted to humans, dogs and cats have recently emerged as possible sources of novel reassortant influenza A.⁶ With their frequent close contact with humans, our pets may prove to pose a significant threat.

Obesity a risk factor for influenza

Obesity emerged as a risk factor for severe influenza in the 2009 pandemic. Recent data also showed that obesity increases the duration of influenza A virus shedding, thus increasing duration of contagiousness.⁷

Influenza a cardiovascular risk factor

Previous data showed that influenza was a risk factor for cardiovascular events. Two recent epi­demiologic studies from the United Kingdom showed that laboratory-confirmed influenza was associated with higher rates of myocardial infarction and stroke for up to 4 weeks.^8,9

Which strain is the biggest threat?

Predicting which emerging influenza serotype may cause the next pandemic is difficult, but influenza A(H7N9), which had not infected humans until 2013 but has since infected about 1,600 people in China and killed 37% of them, appears to have the greatest potential.¹⁰

National influenza surveillance programs and influenza-related social media applications have been developed and may get a boost from technology. A smartphone equipped with a temperature sensor can instantly detect one’s temperature with great precision. A 2018 study suggested that a smartphone-driven thermometry application correlated well with national influenza-like illness activity and improved its forecast in real time and up to 3 weeks in advance.¹¹

TRANSMISSION

Humidity may not block transmission

Animal studies have suggested that humidity in the air interferes with transmission of airborne influenza virus, partially from biologic inactivation. But when a recent study used humidity-controlled chambers to investigate the stability of the 2009 influenza A(H1N1) virus in suspended aerosols and stationary droplets, the virus remained infectious in aerosols across a wide range of relative humidities, challenging the common belief that humidity destabilizes respiratory viruses in aerosols.¹²

One sick passenger may not infect the whole plane

Transmission of respiratory viruses on airplane flights has long been considered a potential avenue for spreading influenza. However, a recent study that monitored movements of individuals on 10 transcontinental US flights and simulated inflight transmission based on these data showed a low probability of direct transmission, except for passengers seated in close proximity to an infectious passenger.¹³

WHAT’S IN THE NEW FLU SHOT?

The 2018–2019 quadrivalent vaccine for the Northern Hemisphere¹⁴ contains the following strains:

• A/Michigan/45/2015 A(H1N1)pdm09-like virus
• A/Singapore/INFIMH-16-0019/2016 (H3N2)-like virus
• B/Colorado/06/2017-like virus (Victoria lineage)
• B/Phuket/3073/2013-like virus (Yamagata lineage).

The A(H3N2) (Singapore) and B/Victoria lineage components are new this year. The A(H3N2) strain was the main cause of the 2018 influenza epidemic in the Southern Hemisphere.

The quadrivalent live-attenuated vaccine, which was not recommended during the 2016–2017 and 2017–2018 influenza seasons, has made a comeback and is recommended for the 2018–2019 season in people for whom it is appropriate based on age and comorbidities.¹⁵ Although it was effective against influenza B and A(H3N2) viruses, it was less effective against the influenza A(H1N1)pdm09-like viruses during the 2013–2014 and 2015–2016 seasons.

A/Slovenia/2903/2015, the new A(H1N1)pdm09-like virus included in the 2018–2019 quadrivalent live-attenuated vaccine, is significantly more immunogenic than its predecessor, A/Bolivia/559/2013, but its clinical effectiveness remains to be seen.

How effective is it?

Influenza vaccine effectiveness in the 2017–2018 influenza season was 36% overall, 67% against A(H1N1), 42% against influenza B, and 25% against A(H3N2).¹⁶ It is estimated that influenza vaccine prevents 300 to 4,000 deaths annually in the United States alone.¹⁷

A 2018 Cochrane review¹⁷ concluded that vaccination reduced the incidence of influenza by about half, with 2.3% of the population contracting the flu without vaccination compared with 0.9% with vaccination (risk ratio 0.41, 95% confidence interval 0.36–0.47). The same review found that 71 healthy adults need to be vaccinated to prevent 1 from experiencing influenza, and 29 to prevent 1 influenza-like illness.

Several recent studies showed that influenza vaccine effectiveness varied based on age and influenza serotype, with higher effectiveness in people ages 5 to 17 and ages 18 to 64 than in those age 65 and older.^18–20 A mathematical model of influenza transmission and vaccination in the United States determined that even relatively low-efficacy influenza vaccines can be very useful if optimally distributed across age groups.²¹

Vaccination rates are low, and ‘antivaxxers’ are on the rise

Although the influenza vaccine is recommended in the United States for all people age 6 months and older regardless of the state of their health, vaccination rates remain low. In 2016, only 37% of employed adults were vaccinated. The highest rate was for government employees (45%), followed by private employees (36%), followed by the self-employed (30%).²²

A national goal is to immunize 80% of all Americans and 90% of at-risk populations (which include children and the elderly).²³ The number of US hospitals that require their employees to be vaccinated increased from 37.1% in 2013 to 61.4% in 2017.²⁴ Regrettably, as of March 2018, 14 lawsuits addressing religious objections to hospital influenza vaccination mandates have been filed.²⁵

Despite hundreds of studies demonstrating the efficacy, safety, and cost savings of influenza vaccination, the antivaccine movement has been growing in the United States and worldwide.²⁶ All US states except West Virginia, Mississippi, and California allow nonmedical exemptions from vaccination based on religious or personal belief.²⁷ Several US metropolitan areas represent “hot spots” for these exemptions.²⁸ This may render such areas vulnerable to vaccine-preventable diseases, including influenza.

Herd immunity: We’re all in this together

Some argue that the potential adverse effects and the cost of vaccination outweigh the benefits, but the protective benefits of herd immunity are significant for those with comorbidities or compromised immunity.

Educating the public about herd immunity and local influenza vaccination uptake increases people’s willingness to be vaccinated.²⁹ A key educational point is that at least 70% of a community needs to be vaccinated to prevent community outbreaks; this protects everyone, including those who do not mount a protective antibody response to influenza vaccination and those who are not vaccinated.

DOES ANNUAL VACCINATION BLUNT ITS EFFECTIVENESS?

Some studies from the 1970s and 1980s raised concern over a possible negative effect of annual influenza vaccination on vaccine effectiveness. The “antigenic distance hypothesis” holds that vaccine effectiveness is influenced by antigenic similarity between the previous season’s vaccine serotypes and the epidemic serotypes, as well as the antigenic similarity between the serotypes of the current and previous seasons.

A meta-analysis of studies from 2010 through 2015 showed significant inconsistencies in repeat vaccination effects within and between seasons and serotypes. It also showed that vaccine effectiveness may be influenced by more than 1 previous season, particularly for influenza A(H3N2), in which repeated vaccination can blunt the hemagglutinin antibody response.³⁰

A study from Japan showed that people who needed medical attention for influenza in the previous season were at lower risk of a similar event in the current season.³¹ Prior-season influenza vaccination reduced current-season vaccine effectiveness only in those who did not have medically attended influenza in the prior season. This suggests that infection is more immunogenic than vaccination, but only against the serotype causing the infection and not the other serotypes included in the vaccine.

An Australian study showed that annual influenza vaccination did not decrease vaccine effectiveness against influenza-associated hospitalization. Rather, effectiveness increased by about 15% in those vaccinated in both current and previous seasons compared with those vaccinated in either season alone.³²

European investigators showed that repeated seasonal influenza vaccination in the elderly prevented the need for hospitalization due to influenza A(H3N2) and B, but not A(H1N1)pdm09.³³

High-dose vaccine for older adults

The high-dose influenza vaccine has been licensed since 2009 for use in the United States for people ages 65 and older.

Recent studies confirmed that high-dose vaccine is more effective than standard-dose vaccine in veterans³⁴ and US Medicare beneficiaries.³⁵

The high-dose vaccine is rapidly becoming the primary vaccine given to people ages 65 and older in retail pharmacies, where vaccination begins earlier in the season than in providers’ offices.³⁶ Some studies have shown that the standard-dose vaccine wanes in effectiveness toward the end of the influenza season (particularly if the season is long) if it is given very early. It remains to be seen whether the same applies to the high-dose influenza vaccine.

Some advocate twice-annual influenza vaccination, particularly for older adults living in tropical and subtropical areas, where influenza seasons are more prolonged. However, a recently published study observed reductions in influenza-specific hemagglutination inhibition and cell-mediated immunity after twice-annual vaccination.³⁷

Vaccination is beneficial during pregnancy

Many studies have shown the value of influenza vaccination during pregnancy for both mothers and their infants.

One recently published study showed that 18% of infants who developed influenza required hospitalization.³⁸ In that study, prenatal and postpartum maternal influenza vaccination decreased the odds of influenza in infants by 61% and 53%, respectively.

Another study showed that vaccine effectiveness did not vary by gestational age at vaccination.³⁹

Some studies have shown that influenza virus infection can increase susceptibility to certain bacterial infections. A post hoc analysis of an influenza vaccination study in pregnant women suggested that the vaccine was also associated with decreased rates of pertussis in these women.⁴⁰

Factors that make vaccination less effective

Several factors including age-related frailty and iatrogenic and disease-related immunosuppression can affect vaccine effectiveness.

Frailty. A recent study showed that vaccine effectiveness was 77.6% in nonfrail older adults but only 58.7% in frail older adults.⁴¹

Immunosuppression. Temporary discontinuation of methotrexate for 2 weeks after influenza vaccination in patients with rheumatoid arthritis improves vaccine immunogenicity without precipitating disease flare.⁴² Solid-organ and hematopoietic stem cell transplant recipients who received influenza vaccine were less likely to develop pneumonia and require intensive care unit admission.⁴³

The high-dose influenza vaccine is more immunogenic than the standard-dose vaccine in solid-organ transplant recipients.⁴⁴

Statins are widely prescribed and have recently been associated with reduced influenza vaccine effectiveness against medically attended acute respiratory illness, but their benefits in preventing cardiovascular events outweigh this risk.⁴⁵

FUTURE VACCINE CONSIDERATIONS

Moving away from eggs

During the annual egg-based production process, which takes several months, the influenza vaccine acquires antigenic changes that allow replication in eggs, particularly in the hemagglutinin protein, which mediates receptor binding. This process of egg adaptation may cause antigenic changes that decrease vaccine effectiveness against circulating viruses.

The cell-based baculovirus influenza vaccine grown in dog kidney cells has higher antigenic content and is not subject to the limitations of egg-based vaccine, although it still requires annual updates. A recombinant influenza vaccine reduces the probability of influenza-like illness by 30% compared with the egg-based influenza vaccine, but also still requires annual updates.⁴⁶ The market share of these non-egg-based vaccines is small, and thus their effectiveness has yet to be demonstrated.

The US Department of Defense administered the cell-based influenza vaccine to about one-third of Armed Forces personnel, their families, and retirees in the 2017–2018 influenza seasons, and data on its effectiveness are expected in the near future.⁴⁷

A universal vaccine would be ideal

The quest continues for a universal influenza vaccine, one that remains protective for several years and does not require annual updates.⁴⁸ Such a vaccine would protect against seasonal epidemic influenza drift variants and pandemic strains. More people could likely be persuaded to be vaccinated once rather than every year.

David Schumick, medical illustrator; adapted from reference 49; illustration of the influenza hemagglutinin created in part with data from reference 50.

An ideal universal vaccine would be suitable for all age groups, at least 75% effective against symptomatic influenza virus infection, protective against all influenza A viruses (influenza A, not B, causes pandemics and seasonal epidemics), and durable through multiple influenza seasons.⁵¹

Research and production of such a vaccine are expected to require funding of about $1 billion over the next 5 years.

Boosting effectiveness

Estimates of influenza vaccine effectiveness range from 40% to 60% in years when the vaccine viruses closely match the circulating viruses, and variably lower when they do not match. The efficacy of most other vaccines given to prevent other infections is much higher.

New technologies to improve influenza vaccine effectiveness are needed, particularly for influenza A(H3N2) viruses, which are rapidly evolving and are highly susceptible to egg-adaptive mutations in the manufacturing process.

In one study, a nanoparticle vaccine formulated with a saponin-based adjuvant induced hemagglutination inhibition responses that were even greater than those induced by the high-dose vaccine.⁵²

Immunoglobulin A (IgA) may be a more effective vaccine target than traditional influenza vaccines that target IgG, since different parts of IgA may engage the influenza virus simultaneously.⁵³

Vaccines can be developed more quickly than in the past. The timeline from viral sequencing to human studies with deoxyribonucleic acid plasmid vaccines decreased from 20 months in 2003 for the severe acquired respiratory syndrome coronavirus to 11 months in 2006 for influenza A/Indonesia/2006 (H5), to 4 months in 2009 for influenza A/California/2009 (H1), to 3.5 months in 2016 for Zika virus.⁵⁴ This is because it is possible today to sequence a virus and insert the genetic material into a vaccine platform without ever having to grow the virus.

Numerous studies have found anti-influenza medications to be effective. Nevertheless, in an analysis of the 2011–2016 influenza seasons, only 15% of high-risk patients were prescribed anti-influenza medications within 2 days of symptom onset, including 37% in those with laboratory-confirmed influenza.⁵⁵ Fever was associated with an increased rate of antiviral treatment, but 25% of high-risk outpatients were afebrile. Empiric treatment of 4 high-risk outpatients with acute respiratory illness was needed to treat 1 patient with influenza.⁵⁵

Treatment with a neuraminidase inhibitor within 2 days of illness has recently been shown to improve survival and shorten duration of viral shedding in patients with avian influenza A(H7N9) infection.⁵⁶ Antiviral treatment within 2 days of illness is associated with improved outcomes in transplant recipients⁵⁷ and with a lower risk of otitis media in children.⁵⁸

Appropriate anti-influenza treatment is as important as avoiding unnecessary antibiotics. Regrettably, as many as one-third of patients with laboratory-confirmed influenza are prescribed antibiotics.⁵⁹

The US Food and Drug Administration warns against fraudulent unapproved over-the-counter influenza products.⁶⁰

Baloxavir marboxil

Baloxavir marboxil is a new anti-influenza medication approved in Japan in February 2018 and anticipated to be available in the United States sometime in 2019.

This prodrug is hydrolyzed in vivo to the active metabolite, which selectively inhibits cap-dependent endonuclease enzyme, a key enzyme in initiation of messenger ribonucleic acid synthesis required for influenza viral replication.⁶¹

In a double-blind phase 3 trial, the median time to alleviation of influenza symptoms is 26.5 hours shorter with baloxavir marboxil than with placebo. One tablet was as effective as 5 days of the neuraminidase inhibitor oseltamivir and was associated with greater reduction in viral load 1 day after initiation, and similar side effects.⁶² Of concern is the emergence of nucleic acid substitutions conferring resistance to baloxavir; this occurred in 2.2% and 9.7% of baloxavir recipients in the phase 2 and 3 trials, respectively.

CLOSING THE GAPS

Several gaps in the management of influenza persist since the 1918 pandemic.¹ These include gaps in epidemiology, prevention, diagnosis, treatment, and prognosis.

• Global networks wider than current ones are needed to address this global disease and to prioritize coordination efforts.
• Establishing and strengthening clinical capacity is needed in limited resource settings. New technologies are needed to expedite vaccine development and to achieve progress toward a universal vaccine.
• Current diagnostic tests do not distinguish between seasonal and novel influenza A viruses of zoonotic origin, which are expected to cause the next pandemic.
• Current antivirals have been shown to shorten duration of illness in outpatients with uncomplicated influenza, but the benefit in hospitalized patients has been less well established.
• In 2007, resistance of seasonal influenza A(H1N1) to oseltamivir became widespread. In 2009, pandemic influenza A(H1N1), which is highly susceptible to oseltamivir, replaced the seasonal virus and remains the predominantly circulating A(H1N1) strain.
• A small-molecule fragment, N-cyclohexyaltaurine, binds to the conserved hemagglutinin receptor-binding site in a manner that mimics the binding mode of the natural receptor sialic acid. This can serve as a template to guide the development of novel broad-spectrum small-molecule anti-influenza drugs.⁶³
• Biomarkers that can accurately predict development of severe disease in patients with influenza are needed.

A plethora of studies are under way in the field of perioperative medicine. As a result, evidence-based care of surgical patients is evolving at an exponential rate.

We performed a literature search and, using consensus, identified recent articles we believe will have a great impact on perioperative cardiovascular medicine. These articles report studies that were presented at national meetings in 2018, including the Perioperative Medicine Summit, Society of General Internal Medicine, and Society of Hospital Medicine. These articles are grouped under 5 questions that will help guide clinical practice in perioperative cardiovascular medicine.

SHOULD ASPIRIN BE CONTINUED PERIOPERATIVELY IN PATIENTS WITH A CORONARY STENT?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial¹ found that giving aspirin before surgery and throughout the early postoperative period had no significant effect on the rate of a composite of death or nonfatal myocardial infarction; moreover, aspirin increased the risk of major bleeding. However, many experts felt uncomfortable stopping aspirin preoperatively in patients taking it for secondary prophylaxis, particularly patients with a coronary stent.

[Graham MM, Sessler DI, Parlow JL, et al. Aspirin in patients with previous percutaneous coronary intervention undergoing noncardiac surgery. Ann Intern Med 2018; 168(4):237–244.]

This post hoc subgroup analysis² of POISE-2 evaluated the benefit and harm of perioperative aspirin in patients who had previously undergone percutaneous coronary intervention, more than 90% of whom had received a stent. Patients were age 45 or older with atherosclerotic heart disease or risk factors for it who had previously undergone percutaneous coronary intervention and were now undergoing noncardiac surgery.

Patients who had received a bare-metal stent within the previous 6 weeks or a drug-eluting stent within 12 months before surgery were excluded because guidelines at that time said to continue dual antiplatelet therapy for that long. Recommendations have since changed; the optimal duration for dual antiplatelet therapy with drug-eluting stents is now 6 months. Second-generation drug-eluting stents pose a lower risk of stent thrombosis and require a shorter duration of dual antiplatelet therapy than first-generation drug-eluting stents. Approximately 25% of the percutaneous coronary intervention subgroup had a drug-eluting stent, but the authors did not specify the type of drug-eluting stent.

The post hoc analysis² included a subgroup of 234 of 4,998 patients receiving aspirin and 236 of 5,012 patients receiving placebo initiated within 4 hours before surgery and continued postoperatively. The primary outcome measured was the rate of death or nonfatal myocardial infarction within 30 days after surgery, and bleeding was a secondary outcome.

Findings. Although the overall POISE-2 study found no benefit from aspirin, in the subgroup who had previously undergone percutaneous coronary intervention, aspirin significantly reduced the risk of the primary outcome, which occurred in 6% vs 11.5% of the patients:

• Absolute risk reduction 5.5% (95% confidence interval 0.4%–10.5%)
• Hazard ratio 0.50 (0.26–0.95).

The reduction was primarily due to fewer myocardial infarctions:

• Absolute risk reduction 5.9% (1.0%–10.8%)
• Hazard ratio 0.44 (0.22–0.87).

The type of stent had no effect on the primary outcome, although this subgroup analysis had limited power. In the nonpercutaneous coronary intervention subgroup, there was no significant difference in outcomes between the aspirin and placebo groups. This subgroup analysis was underpowered to evaluate the effect of aspirin on the composite of major and life-threatening bleeding in patients with prior percutaneous coronary intervention, which was reported as “uncertain” due to wide confidence intervals (absolute risk increase 1.3%, 95% confidence interval –2.6% to 5.2%), but the increased risk of major or life-threatening bleeding with aspirin demonstrated in the overall POISE-2 study population likely applies:

• Absolute risk increase 0.8% (0.1%–1.6%)
• Hazard ratio 1.22 (1.01–1.48).

Limitations. This was a nonspecified subgroup analysis that was underpowered and had a relatively small sample size with few events.

Conclusion. In the absence of a very high bleeding risk, continuing aspirin perioperatively in patients with prior percutaneous coronary intervention undergoing noncardiac surgery is more likely to result in benefit than harm. This finding is in agreement with current recommendations from the American College Cardiology and American Heart Association (class I; level of evidence C).³

Despite advances in anesthesia and surgical techniques, about 1% of patients over age 45 die within 30 days of noncardiac surgery.⁴ Studies have demonstrated a high mortality rate in patients who experience myocardial injury after noncardiac surgery (MINS), defined as elevations of troponin T with or without ischemic symptoms or electrocardiographic changes.⁵ Most of these studies used earlier, “non-high-sensitivity” troponin T assays. Fifth-generation, highly sensitive troponin T assays are now available that can detect troponin T at lower concentrations, but their utility in predicting postoperative outcomes remains uncertain. Two recent studies provide further insight into these issues.

[Writing Committee for the VISION Study Investigators, Devereaux PJ, Biccard BM, Sigamani A, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317(16):1642–1651.]

The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study⁵ was an international, prospective cohort study that initially evaluated the association between MINS and the 30-day mortality rate using a non-high-sensitivity troponin T assay (Roche fourth-generation Elecsys TnT assay) in patients age 45 or older undergoing noncardiac surgery and requiring hospital admission for at least 1 night. After the first 15,000 patients, the study switched to the Roche fifth-generation assay, with measurements at 6 to 12 hours after surgery and on postoperative days 1, 2, and 3.

A 2017 analysis by Devereaux et al⁶ included only these later-enrolled patients and correlated their high-sensitivity troponin T levels with 30-day mortality rates. Patients with a level 14 ng/L or higher, the upper limit of normal in this study, were also assessed for ischemic symptoms and electrocardiographic changes. Although not required by the study, more than 7,800 patients had their troponin T levels measured before surgery, and the absolute change was also analyzed for an association with the 30-day mortality rate.

Findings. Of the 21,842 patients, about two-thirds underwent some form of major surgery; some of them had more than 1 type. A total of 1.2% of the patients died within 30 days of surgery.

Based on their analysis, the authors proposed that MINS be defined as:

• A postoperative troponin T level of 65 ng/L or higher, or
• A level in the range of 20 ng/L to less than 65 ng/L with an absolute increase from the preoperative level at least 5 ng/L, not attributable to a nonischemic cause.

Seventeen percent of the study patients met these criteria, and of these, 21.7% met the universal definition of myocardial infarction, although only 6.9% had symptoms of it.

Limitations. Only 40.4% of the patients had a preoperative high-sensitivity troponin T measurement for comparison, and in 13.8% of patients who had an elevated perioperative measurement, their preoperative value was the same or higher than their postoperative one. Thus, the incidence of MINS may have been overestimated if patients were otherwise not known to have troponin T elevations before surgery.

[Puelacher C, Lurati Buse G, Seeberger D, et al. Perioperative myocardial injury after noncardiac surgery: incidence, mortality, and characterization. Circulation 2018; 137(12):1221–1232.]

Puelacher et al⁷ investigated the prevalence of MINS in 2,018 patients at increased cardiovascular risk (age ≥ 65, or age ≥ 45 with a history of coronary artery disease, peripheral vascular disease, or stroke) who underwent major noncardiac surgery (planned overnight stay ≥ 24 hours) at a university hospital in Switzerland. Patients had their troponin T measured with a high-sensitivity assay within 30 days before surgery and on postoperative days 1 and 2.

Instead of MINS, the investigators used the term “perioperative myocardial injury” (PMI), defined as an absolute increase in troponin T of at least 14 ng/L from before surgery to the peak postoperative reading. Similar to MINS, PMI did not require ischemic features, but in this study, noncardiac triggers (sepsis, stroke, or pulmonary embolus) were not excluded.

Findings. PMI occurred in 16% of surgeries, and of the patients with PMI, 6% had typical chest pain and 18% had any ischemic symptoms. Unlike in the POISE-2 study discussed above, PMI triggered an automatic referral to a cardiologist.

The unadjusted 30-day mortality rate was 8.9% among patients with PMI and 1.5% in those without. Multivariable logistic regression analysis showed an adjusted hazard ratio for 30-day mortality of 2.7 (95% CI 1.5–4.8) for those with PMI vs without, and this difference persisted for at least 1 year.

In patients with PMI, the authors compared the 30-day mortality rate of those with no ischemic signs or symptoms (71% of the patients) with those who met the criteria for myocardial infarction and found no difference. Patients with PMI triggered by a noncardiac event had a worse prognosis than those with a presumed cardiac etiology.

Limitations. Despite the multivariate analysis that included adjustment for age, nonelective surgery, and Revised Cardiac Risk Index (RCRI), the increased risk associated with PMI could simply reflect higher risk at baseline. Although PMI resulted in automatic referral to a cardiologist, only 10% of patients eventually underwent coronary angiography; a similar percentage were discharged with additional medical therapy such as aspirin, a statin, or a beta-blocker. The effect of these interventions is not known.

Conclusions. MINS is common and has a strong association with mortality risk proportional to the degree of troponin T elevation using high-sensitivity assays, consistent with data from previous studies of earlier assays. Because the mechanism of MINS may differ from that of myocardial infarction, its prevention and treatment may differ, and it remains unclear how serial measurement in postoperative patients should change clinical practice.

The recently published Dabigatran in Patients With Myocardial Injury After Non-cardiac Surgery (MANAGE) trial⁸ suggests that dabigatran may reduce arterial and venous complications in patients with MINS, but the study had a number of limitations that may restrict the clinical applicability of this finding.

While awaiting further clinical outcomes data, pre- and postoperative troponin T measurement may be beneficial in higher-risk patients (such as those with cardiovascular disease or multiple RCRI risk factors) if the information will change perioperative management.

Intraoperative hypotension is associated with organ ischemia, which may cause postoperative myocardial infarction, myocardial injury, and acute kidney injury.⁹ Traditional anesthesia practice is to maintain intraoperative blood pressure within 20% of the preoperative baseline, based on the notion that hypertensive patients require higher perfusion pressures.

[Futier E, Lefrant J-Y, Guinot P-G, et al. Effect of individualized vs standard blood pressure management strategies on postoperative organ dysfunction among high-risk patients undergoing major surgery: a randomized clinical trial. JAMA 2017; 318(14):1346–1357.]

Futier et al¹⁰ sought to address uncertainty in intraoperative and immediate postoperative management of systolic blood pressure. In this multicenter, randomized, parallel-group trial, 298 patients at increased risk of postoperative renal complications were randomized to blood pressure management that was either “individualized” (within 10% of resting systolic pressure) or “standard” (≥ 80 mm Hg or ≥ 40% of resting systolic pressure) from induction to 4 hours postoperatively.

Blood pressure was monitored using radial arterial lines and maintained using a combination of intravenous fluids, norepinephrine (the first-line agent for the individualized group), and ephedrine (in the standard treatment group only). The primary outcome was a composite of systemic inflammatory response syndrome (SIRS) and organ dysfunction affecting at least 1 organ system (cardiovascular, respiratory, renal, hematologic, or neurologic).

Findings. Data on the primary outcome were available for 292 of 298 patients enrolled. The mean age was 70 years, 15% were women, and 82% had previously diagnosed hypertension. Despite the requirement for an elevated risk of acute kidney injury, only 13% of the patients had a baseline estimated glomerular filtration rate of less than 60 mL/min/1.73 m², and the median was 88 mL/min/1.73 m². Ninety-five percent of patients underwent abdominal surgery, and 50% of the surgeries were elective.

The mean systolic blood pressure was 123 mm Hg in the individualized treatment group compared with 116 mm Hg in the standard treatment group. Despite this small difference, 96% of individualized treatment patients received norepinephrine, compared with 26% in the standard treatment group.

The primary outcome of SIRS with organ dysfunction occurred in 38.1% of patients in the individualized treatment group and 51.7% of those in the standard treatment group. After adjusting for center, surgical urgency, surgical site, and acute kidney injury risk index, the relative risk of developing SIRS in those receiving individualized management was 0.73 (P = .02). Renal dysfunction (based on Acute Dialysis Quality Initiative criteria¹¹) occurred in 32.7% of individualized treatment patients and 49% of standardized treatment patients.  

Limitations of this study included differences in pharmacologic approach to maintain blood pressure in the 2 protocols (ephedrine and fluids vs norepinephrine) and a modest sample size.

Conclusions. Despite this, the difference in organ dysfunction was striking, with a number needed to treat of only 7 patients. This intervention extended 4 hours postoperatively, a time when many of these patients have left the postanesthesia care unit and have returned to hospitalist care on inpatient wards.

While optimal management of intraoperative and immediate postoperative blood pressure may not be settled, this study suggests that even mild relative hypotension may justify immediate action. Further studies may be useful to delineate high- and low-risk populations, the timing of greatest risk, and indications for intraarterial blood pressure monitoring.

[Salmasi V, Maheswari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology 2017; 126(1):47–65.]

This retrospective cohort study¹² assessed the association between myocardial or kidney injury and absolute or relative thresholds of intraoperative mean arterial pressure. It included 57,315 adults who underwent inpatient noncardiac surgery, had a preoperative and at least 1 postoperative serum creatinine measurement within 7 days, and had blood pressure recorded in preoperative appointments within 6 months. Patients with chronic kidney disease (glomerular filtration rate < 60 mL/min/1.73 m²) and those on dialysis were excluded. The outcomes were MINS⁵ and acute kidney injury as defined by the Acute Kidney Injury Network.⁹

Findings. A mean arterial pressure below an absolute threshold of 65 mm Hg or a relative threshold of 20% lower than baseline value was associated with myocardial and kidney injury. At each threshold, prolonged periods of hypotension were associated with progressively increased risk.

An important conclusion of the study was that relative thresholds of mean arterial pressure were not any more predictive than absolute thresholds. Absolute thresholds are easier to use intraoperatively, especially when baseline values are not available. The authors did not find a clinically significant interaction between baseline blood pressure and the association of hypotension and myocardial and kidney injury.

Limitations included use of cardiac enzymes postoperatively to define MINS. Since these were not routinely collected, clinically silent myocardial injury may have been missed. Baseline blood pressure may have important implications in other forms of organ injury (ie, cerebral ischemia) that were not studied.

Summary. The lowest absolute mean arterial pressure is as predictive of postoperative myocardial and kidney injury as the relative pressure reduction, at least in patients with normal renal function. Limiting exposure to intraoperative hypotension is important. Baseline blood pressure values may have limited utility for intraoperative management.

In combination, these studies confirm that intraoperative hypotension is a predictor of postoperative organ dysfunction, but the definition and management remain unclear. While aggressive intraoperative management is likely beneficial, how to manage the anti­hypertensive therapy the patient has been taking as an outpatient when he or she comes into the hospital for surgery remains uncertain.

Perioperative stroke is an uncommon, severe complication of noncardiac surgery. The pathophysiology has been better defined in cardiac than in noncardiac surgeries. In nonsurgical patients, patent foramen ovale (PFO) is associated with stroke, even in patients considered to be at low risk.¹³ Perioperative patients have additional risk for venous thromboembolism and may have periprocedural antithrombotic medications altered, increasing their risk of paradoxical embolism through the PFO.

[Ng PY, Ng AK, Subramaniam B, et al. Association of preoperatively diagnosed patent foramen ovale with perioperative ischemic stroke. JAMA 2018; 319(5):452–462.]

This retrospective cohort study of noncardiac surgery patients at 3 hospitals¹⁴ sought to determine the association of preoperatively diagnosed PFO with the risk of perioperative ischemic stroke identified by International Classification of Diseases diagnoses.

Of 150,198 patients, 1.0% had a preoperative diagnosis of PFO, and at baseline, those with PFO had significantly more comorbidities than those without PFO. Stroke occurred in 3.2% of patients with PFO vs 0.5% of those without. Patients known to have a PFO were much more likely to have cardiovascular and thromboembolic risk factors for stroke. In the adjusted analysis, the absolute risk difference between groups was 0.4% (95% CI 0.2–0.6%), with an estimated perioperative stroke risk of 5.9 per 1,000 in patients with known patent foramen ovale and 2.2 per 1,000 in those without. A diagnosis of PFO was also associated with increased risk of large-vessel-territory stroke and more severe neurologic deficit.

Further attempts to adjust for baseline risk factors and other potential bias, including a propensity score-matched cohort analysis and an analysis limited to patients who had echocardiography performed in the same healthcare system, still showed a higher risk of perioperative stroke among patients with preoperatively detected patent foramen ovale.

Limitations. The study was retrospective and observational, used administrative data, and had a low rate of PFO diagnosis (1%), compared with about 25% in population-based studies.¹⁵ Indications for preoperative echocardiography are unknown. In addition, the study specifically examined preoperatively diagnosed PFO, rather than including those diagnosed in the postoperative period.

Discussion. How does this study affect clinical practice? The absolute stroke risk was increased by 0.4% in patients with PFO compared with those without. Although this is a relatively small increase, millions of patients undergo noncardiac surgery annually. The risks of therapeutic anticoagulation or PFO closure are likely too high in this context; however, clinicians may approach the perioperative management of antiplatelet agents and venous thromboembolism prophylaxis in patients with known PFO with additional caution.

HOW DOES TIMING OF EMERGENCY SURGERY AFTER PRIOR STROKE AFFECT OUTCOMES?

A history of stroke or transient ischemic attack is a known risk factor for perioperative vascular complications. A recent large cohort study demonstrated that a history of stroke within 9 months of elective surgery was associated with increased adverse outcomes.¹⁶ Little is known, however, of the perioperative risk in patients with a history of stroke who undergo emergency surgery.

[Christiansen MN, Andersson C, Gislason GH, et al. Risks of cardiovascular adverse events and death in patients with previous stroke undergoing emergency noncardiac, nonintracranial surgery: the importance of operative timing. Anesthesiology 2017; 127(1):9–19.]

In this study,¹⁷ all emergency noncardiac and nonintracranial surgeries from 2005 to 2011 were analyzed using multiple national patient registries in Denmark according to time elapsed between previous stroke and surgery. Primary outcomes were 30-day all-cause mortality and 30-day major adverse cardiac events (MACE), defined as nonfatal ischemic stroke, nonfatal myocardial infarction, and cardiovascular death. Statistical analysis to assess the risk of adverse outcomes included logistic regression models, spline analyses, and propensity-score matching.

Findings. The authors identified 146,694 emergency surgeries, with 7,861 patients (5.4%) having had a previous stroke (transient ischemic attacks and hemorrhagic strokes were not included). Rates of postoperative stroke were as follows:

• 9.9% in patents with a history of ischemic stroke within 3 months of surgery
• 2.8% in patients with a history of stroke 3 to 9 months before surgery
• 0.3% in patients with no previous stroke.

The risk plateaued when the time between stroke and surgery exceeded 4 to 5 months.¹⁵

Interestingly, in patients who underwent emergency surgery within 14 days of stroke, the risk of MACE was significantly lower immediately after surgery (1–3 days after stroke) compared with surgery that took place 4 to 14 days after stroke. The authors hypothesized that because cerebral autoregulation does not become compromised until approximately 5 days after a stroke, the risk was lower 1 to 3 days after surgery and increased thereafter.

Limitations of this study included the possibility of residual confounding, given its retrospective design using administrative data, not accounting for preoperative antithrombotic and anticoagulation therapy, and lack of information regarding the etiology of recurrent stroke (eg, thromboembolic, atherothrombotic, hypoperfusion).

Conclusions. Although it would be impractical to postpone emergency surgery in a patient who recently had a stroke, this study shows that the incidence rates of postoperative recurrent stroke and MACE are high. Therefore, it is important that the patient and perioperative team be aware of the risk. Further research is needed to confirm these estimates of postoperative adverse events in more diverse patient populations.

A plethora of studies are under way in the field of perioperative medicine. As a result, evidence-based care of surgical patients is evolving at an exponential rate.

We performed a literature search and, using consensus, identified recent articles we believe will have a great impact on perioperative cardiovascular medicine. These articles report studies that were presented at national meetings in 2018, including the Perioperative Medicine Summit, Society of General Internal Medicine, and Society of Hospital Medicine. These articles are grouped under 5 questions that will help guide clinical practice in perioperative cardiovascular medicine.

SHOULD ASPIRIN BE CONTINUED PERIOPERATIVELY IN PATIENTS WITH A CORONARY STENT?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial¹ found that giving aspirin before surgery and throughout the early postoperative period had no significant effect on the rate of a composite of death or nonfatal myocardial infarction; moreover, aspirin increased the risk of major bleeding. However, many experts felt uncomfortable stopping aspirin preoperatively in patients taking it for secondary prophylaxis, particularly patients with a coronary stent.

[Graham MM, Sessler DI, Parlow JL, et al. Aspirin in patients with previous percutaneous coronary intervention undergoing noncardiac surgery. Ann Intern Med 2018; 168(4):237–244.]

This post hoc subgroup analysis² of POISE-2 evaluated the benefit and harm of perioperative aspirin in patients who had previously undergone percutaneous coronary intervention, more than 90% of whom had received a stent. Patients were age 45 or older with atherosclerotic heart disease or risk factors for it who had previously undergone percutaneous coronary intervention and were now undergoing noncardiac surgery.

Patients who had received a bare-metal stent within the previous 6 weeks or a drug-eluting stent within 12 months before surgery were excluded because guidelines at that time said to continue dual antiplatelet therapy for that long. Recommendations have since changed; the optimal duration for dual antiplatelet therapy with drug-eluting stents is now 6 months. Second-generation drug-eluting stents pose a lower risk of stent thrombosis and require a shorter duration of dual antiplatelet therapy than first-generation drug-eluting stents. Approximately 25% of the percutaneous coronary intervention subgroup had a drug-eluting stent, but the authors did not specify the type of drug-eluting stent.

The post hoc analysis² included a subgroup of 234 of 4,998 patients receiving aspirin and 236 of 5,012 patients receiving placebo initiated within 4 hours before surgery and continued postoperatively. The primary outcome measured was the rate of death or nonfatal myocardial infarction within 30 days after surgery, and bleeding was a secondary outcome.

Findings. Although the overall POISE-2 study found no benefit from aspirin, in the subgroup who had previously undergone percutaneous coronary intervention, aspirin significantly reduced the risk of the primary outcome, which occurred in 6% vs 11.5% of the patients:

• Absolute risk reduction 5.5% (95% confidence interval 0.4%–10.5%)
• Hazard ratio 0.50 (0.26–0.95).

The reduction was primarily due to fewer myocardial infarctions:

• Absolute risk reduction 5.9% (1.0%–10.8%)
• Hazard ratio 0.44 (0.22–0.87).

The type of stent had no effect on the primary outcome, although this subgroup analysis had limited power. In the nonpercutaneous coronary intervention subgroup, there was no significant difference in outcomes between the aspirin and placebo groups. This subgroup analysis was underpowered to evaluate the effect of aspirin on the composite of major and life-threatening bleeding in patients with prior percutaneous coronary intervention, which was reported as “uncertain” due to wide confidence intervals (absolute risk increase 1.3%, 95% confidence interval –2.6% to 5.2%), but the increased risk of major or life-threatening bleeding with aspirin demonstrated in the overall POISE-2 study population likely applies:

• Absolute risk increase 0.8% (0.1%–1.6%)
• Hazard ratio 1.22 (1.01–1.48).

Limitations. This was a nonspecified subgroup analysis that was underpowered and had a relatively small sample size with few events.

Conclusion. In the absence of a very high bleeding risk, continuing aspirin perioperatively in patients with prior percutaneous coronary intervention undergoing noncardiac surgery is more likely to result in benefit than harm. This finding is in agreement with current recommendations from the American College Cardiology and American Heart Association (class I; level of evidence C).³

Despite advances in anesthesia and surgical techniques, about 1% of patients over age 45 die within 30 days of noncardiac surgery.⁴ Studies have demonstrated a high mortality rate in patients who experience myocardial injury after noncardiac surgery (MINS), defined as elevations of troponin T with or without ischemic symptoms or electrocardiographic changes.⁵ Most of these studies used earlier, “non-high-sensitivity” troponin T assays. Fifth-generation, highly sensitive troponin T assays are now available that can detect troponin T at lower concentrations, but their utility in predicting postoperative outcomes remains uncertain. Two recent studies provide further insight into these issues.

[Writing Committee for the VISION Study Investigators, Devereaux PJ, Biccard BM, Sigamani A, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317(16):1642–1651.]

The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study⁵ was an international, prospective cohort study that initially evaluated the association between MINS and the 30-day mortality rate using a non-high-sensitivity troponin T assay (Roche fourth-generation Elecsys TnT assay) in patients age 45 or older undergoing noncardiac surgery and requiring hospital admission for at least 1 night. After the first 15,000 patients, the study switched to the Roche fifth-generation assay, with measurements at 6 to 12 hours after surgery and on postoperative days 1, 2, and 3.

A 2017 analysis by Devereaux et al⁶ included only these later-enrolled patients and correlated their high-sensitivity troponin T levels with 30-day mortality rates. Patients with a level 14 ng/L or higher, the upper limit of normal in this study, were also assessed for ischemic symptoms and electrocardiographic changes. Although not required by the study, more than 7,800 patients had their troponin T levels measured before surgery, and the absolute change was also analyzed for an association with the 30-day mortality rate.

Findings. Of the 21,842 patients, about two-thirds underwent some form of major surgery; some of them had more than 1 type. A total of 1.2% of the patients died within 30 days of surgery.

Based on their analysis, the authors proposed that MINS be defined as:

• A postoperative troponin T level of 65 ng/L or higher, or
• A level in the range of 20 ng/L to less than 65 ng/L with an absolute increase from the preoperative level at least 5 ng/L, not attributable to a nonischemic cause.

Seventeen percent of the study patients met these criteria, and of these, 21.7% met the universal definition of myocardial infarction, although only 6.9% had symptoms of it.

Limitations. Only 40.4% of the patients had a preoperative high-sensitivity troponin T measurement for comparison, and in 13.8% of patients who had an elevated perioperative measurement, their preoperative value was the same or higher than their postoperative one. Thus, the incidence of MINS may have been overestimated if patients were otherwise not known to have troponin T elevations before surgery.

[Puelacher C, Lurati Buse G, Seeberger D, et al. Perioperative myocardial injury after noncardiac surgery: incidence, mortality, and characterization. Circulation 2018; 137(12):1221–1232.]

Puelacher et al⁷ investigated the prevalence of MINS in 2,018 patients at increased cardiovascular risk (age ≥ 65, or age ≥ 45 with a history of coronary artery disease, peripheral vascular disease, or stroke) who underwent major noncardiac surgery (planned overnight stay ≥ 24 hours) at a university hospital in Switzerland. Patients had their troponin T measured with a high-sensitivity assay within 30 days before surgery and on postoperative days 1 and 2.

Instead of MINS, the investigators used the term “perioperative myocardial injury” (PMI), defined as an absolute increase in troponin T of at least 14 ng/L from before surgery to the peak postoperative reading. Similar to MINS, PMI did not require ischemic features, but in this study, noncardiac triggers (sepsis, stroke, or pulmonary embolus) were not excluded.

Findings. PMI occurred in 16% of surgeries, and of the patients with PMI, 6% had typical chest pain and 18% had any ischemic symptoms. Unlike in the POISE-2 study discussed above, PMI triggered an automatic referral to a cardiologist.

The unadjusted 30-day mortality rate was 8.9% among patients with PMI and 1.5% in those without. Multivariable logistic regression analysis showed an adjusted hazard ratio for 30-day mortality of 2.7 (95% CI 1.5–4.8) for those with PMI vs without, and this difference persisted for at least 1 year.

In patients with PMI, the authors compared the 30-day mortality rate of those with no ischemic signs or symptoms (71% of the patients) with those who met the criteria for myocardial infarction and found no difference. Patients with PMI triggered by a noncardiac event had a worse prognosis than those with a presumed cardiac etiology.

Limitations. Despite the multivariate analysis that included adjustment for age, nonelective surgery, and Revised Cardiac Risk Index (RCRI), the increased risk associated with PMI could simply reflect higher risk at baseline. Although PMI resulted in automatic referral to a cardiologist, only 10% of patients eventually underwent coronary angiography; a similar percentage were discharged with additional medical therapy such as aspirin, a statin, or a beta-blocker. The effect of these interventions is not known.

Conclusions. MINS is common and has a strong association with mortality risk proportional to the degree of troponin T elevation using high-sensitivity assays, consistent with data from previous studies of earlier assays. Because the mechanism of MINS may differ from that of myocardial infarction, its prevention and treatment may differ, and it remains unclear how serial measurement in postoperative patients should change clinical practice.

The recently published Dabigatran in Patients With Myocardial Injury After Non-cardiac Surgery (MANAGE) trial⁸ suggests that dabigatran may reduce arterial and venous complications in patients with MINS, but the study had a number of limitations that may restrict the clinical applicability of this finding.

While awaiting further clinical outcomes data, pre- and postoperative troponin T measurement may be beneficial in higher-risk patients (such as those with cardiovascular disease or multiple RCRI risk factors) if the information will change perioperative management.

Intraoperative hypotension is associated with organ ischemia, which may cause postoperative myocardial infarction, myocardial injury, and acute kidney injury.⁹ Traditional anesthesia practice is to maintain intraoperative blood pressure within 20% of the preoperative baseline, based on the notion that hypertensive patients require higher perfusion pressures.

[Futier E, Lefrant J-Y, Guinot P-G, et al. Effect of individualized vs standard blood pressure management strategies on postoperative organ dysfunction among high-risk patients undergoing major surgery: a randomized clinical trial. JAMA 2017; 318(14):1346–1357.]

Futier et al¹⁰ sought to address uncertainty in intraoperative and immediate postoperative management of systolic blood pressure. In this multicenter, randomized, parallel-group trial, 298 patients at increased risk of postoperative renal complications were randomized to blood pressure management that was either “individualized” (within 10% of resting systolic pressure) or “standard” (≥ 80 mm Hg or ≥ 40% of resting systolic pressure) from induction to 4 hours postoperatively.

Blood pressure was monitored using radial arterial lines and maintained using a combination of intravenous fluids, norepinephrine (the first-line agent for the individualized group), and ephedrine (in the standard treatment group only). The primary outcome was a composite of systemic inflammatory response syndrome (SIRS) and organ dysfunction affecting at least 1 organ system (cardiovascular, respiratory, renal, hematologic, or neurologic).

Findings. Data on the primary outcome were available for 292 of 298 patients enrolled. The mean age was 70 years, 15% were women, and 82% had previously diagnosed hypertension. Despite the requirement for an elevated risk of acute kidney injury, only 13% of the patients had a baseline estimated glomerular filtration rate of less than 60 mL/min/1.73 m², and the median was 88 mL/min/1.73 m². Ninety-five percent of patients underwent abdominal surgery, and 50% of the surgeries were elective.

The mean systolic blood pressure was 123 mm Hg in the individualized treatment group compared with 116 mm Hg in the standard treatment group. Despite this small difference, 96% of individualized treatment patients received norepinephrine, compared with 26% in the standard treatment group.

The primary outcome of SIRS with organ dysfunction occurred in 38.1% of patients in the individualized treatment group and 51.7% of those in the standard treatment group. After adjusting for center, surgical urgency, surgical site, and acute kidney injury risk index, the relative risk of developing SIRS in those receiving individualized management was 0.73 (P = .02). Renal dysfunction (based on Acute Dialysis Quality Initiative criteria¹¹) occurred in 32.7% of individualized treatment patients and 49% of standardized treatment patients.  

Limitations of this study included differences in pharmacologic approach to maintain blood pressure in the 2 protocols (ephedrine and fluids vs norepinephrine) and a modest sample size.

Conclusions. Despite this, the difference in organ dysfunction was striking, with a number needed to treat of only 7 patients. This intervention extended 4 hours postoperatively, a time when many of these patients have left the postanesthesia care unit and have returned to hospitalist care on inpatient wards.

While optimal management of intraoperative and immediate postoperative blood pressure may not be settled, this study suggests that even mild relative hypotension may justify immediate action. Further studies may be useful to delineate high- and low-risk populations, the timing of greatest risk, and indications for intraarterial blood pressure monitoring.

[Salmasi V, Maheswari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology 2017; 126(1):47–65.]

This retrospective cohort study¹² assessed the association between myocardial or kidney injury and absolute or relative thresholds of intraoperative mean arterial pressure. It included 57,315 adults who underwent inpatient noncardiac surgery, had a preoperative and at least 1 postoperative serum creatinine measurement within 7 days, and had blood pressure recorded in preoperative appointments within 6 months. Patients with chronic kidney disease (glomerular filtration rate < 60 mL/min/1.73 m²) and those on dialysis were excluded. The outcomes were MINS⁵ and acute kidney injury as defined by the Acute Kidney Injury Network.⁹

Findings. A mean arterial pressure below an absolute threshold of 65 mm Hg or a relative threshold of 20% lower than baseline value was associated with myocardial and kidney injury. At each threshold, prolonged periods of hypotension were associated with progressively increased risk.

An important conclusion of the study was that relative thresholds of mean arterial pressure were not any more predictive than absolute thresholds. Absolute thresholds are easier to use intraoperatively, especially when baseline values are not available. The authors did not find a clinically significant interaction between baseline blood pressure and the association of hypotension and myocardial and kidney injury.

Limitations included use of cardiac enzymes postoperatively to define MINS. Since these were not routinely collected, clinically silent myocardial injury may have been missed. Baseline blood pressure may have important implications in other forms of organ injury (ie, cerebral ischemia) that were not studied.

Summary. The lowest absolute mean arterial pressure is as predictive of postoperative myocardial and kidney injury as the relative pressure reduction, at least in patients with normal renal function. Limiting exposure to intraoperative hypotension is important. Baseline blood pressure values may have limited utility for intraoperative management.

In combination, these studies confirm that intraoperative hypotension is a predictor of postoperative organ dysfunction, but the definition and management remain unclear. While aggressive intraoperative management is likely beneficial, how to manage the anti­hypertensive therapy the patient has been taking as an outpatient when he or she comes into the hospital for surgery remains uncertain.

Perioperative stroke is an uncommon, severe complication of noncardiac surgery. The pathophysiology has been better defined in cardiac than in noncardiac surgeries. In nonsurgical patients, patent foramen ovale (PFO) is associated with stroke, even in patients considered to be at low risk.¹³ Perioperative patients have additional risk for venous thromboembolism and may have periprocedural antithrombotic medications altered, increasing their risk of paradoxical embolism through the PFO.

[Ng PY, Ng AK, Subramaniam B, et al. Association of preoperatively diagnosed patent foramen ovale with perioperative ischemic stroke. JAMA 2018; 319(5):452–462.]

This retrospective cohort study of noncardiac surgery patients at 3 hospitals¹⁴ sought to determine the association of preoperatively diagnosed PFO with the risk of perioperative ischemic stroke identified by International Classification of Diseases diagnoses.

Of 150,198 patients, 1.0% had a preoperative diagnosis of PFO, and at baseline, those with PFO had significantly more comorbidities than those without PFO. Stroke occurred in 3.2% of patients with PFO vs 0.5% of those without. Patients known to have a PFO were much more likely to have cardiovascular and thromboembolic risk factors for stroke. In the adjusted analysis, the absolute risk difference between groups was 0.4% (95% CI 0.2–0.6%), with an estimated perioperative stroke risk of 5.9 per 1,000 in patients with known patent foramen ovale and 2.2 per 1,000 in those without. A diagnosis of PFO was also associated with increased risk of large-vessel-territory stroke and more severe neurologic deficit.

Further attempts to adjust for baseline risk factors and other potential bias, including a propensity score-matched cohort analysis and an analysis limited to patients who had echocardiography performed in the same healthcare system, still showed a higher risk of perioperative stroke among patients with preoperatively detected patent foramen ovale.

Limitations. The study was retrospective and observational, used administrative data, and had a low rate of PFO diagnosis (1%), compared with about 25% in population-based studies.¹⁵ Indications for preoperative echocardiography are unknown. In addition, the study specifically examined preoperatively diagnosed PFO, rather than including those diagnosed in the postoperative period.

Discussion. How does this study affect clinical practice? The absolute stroke risk was increased by 0.4% in patients with PFO compared with those without. Although this is a relatively small increase, millions of patients undergo noncardiac surgery annually. The risks of therapeutic anticoagulation or PFO closure are likely too high in this context; however, clinicians may approach the perioperative management of antiplatelet agents and venous thromboembolism prophylaxis in patients with known PFO with additional caution.

HOW DOES TIMING OF EMERGENCY SURGERY AFTER PRIOR STROKE AFFECT OUTCOMES?

A history of stroke or transient ischemic attack is a known risk factor for perioperative vascular complications. A recent large cohort study demonstrated that a history of stroke within 9 months of elective surgery was associated with increased adverse outcomes.¹⁶ Little is known, however, of the perioperative risk in patients with a history of stroke who undergo emergency surgery.

[Christiansen MN, Andersson C, Gislason GH, et al. Risks of cardiovascular adverse events and death in patients with previous stroke undergoing emergency noncardiac, nonintracranial surgery: the importance of operative timing. Anesthesiology 2017; 127(1):9–19.]

In this study,¹⁷ all emergency noncardiac and nonintracranial surgeries from 2005 to 2011 were analyzed using multiple national patient registries in Denmark according to time elapsed between previous stroke and surgery. Primary outcomes were 30-day all-cause mortality and 30-day major adverse cardiac events (MACE), defined as nonfatal ischemic stroke, nonfatal myocardial infarction, and cardiovascular death. Statistical analysis to assess the risk of adverse outcomes included logistic regression models, spline analyses, and propensity-score matching.

Findings. The authors identified 146,694 emergency surgeries, with 7,861 patients (5.4%) having had a previous stroke (transient ischemic attacks and hemorrhagic strokes were not included). Rates of postoperative stroke were as follows:

• 9.9% in patents with a history of ischemic stroke within 3 months of surgery
• 2.8% in patients with a history of stroke 3 to 9 months before surgery
• 0.3% in patients with no previous stroke.

The risk plateaued when the time between stroke and surgery exceeded 4 to 5 months.¹⁵

Interestingly, in patients who underwent emergency surgery within 14 days of stroke, the risk of MACE was significantly lower immediately after surgery (1–3 days after stroke) compared with surgery that took place 4 to 14 days after stroke. The authors hypothesized that because cerebral autoregulation does not become compromised until approximately 5 days after a stroke, the risk was lower 1 to 3 days after surgery and increased thereafter.

Limitations of this study included the possibility of residual confounding, given its retrospective design using administrative data, not accounting for preoperative antithrombotic and anticoagulation therapy, and lack of information regarding the etiology of recurrent stroke (eg, thromboembolic, atherothrombotic, hypoperfusion).

Conclusions. Although it would be impractical to postpone emergency surgery in a patient who recently had a stroke, this study shows that the incidence rates of postoperative recurrent stroke and MACE are high. Therefore, it is important that the patient and perioperative team be aware of the risk. Further research is needed to confirm these estimates of postoperative adverse events in more diverse patient populations.

A plethora of studies are under way in the field of perioperative medicine. As a result, evidence-based care of surgical patients is evolving at an exponential rate.

We performed a literature search and, using consensus, identified recent articles we believe will have a great impact on perioperative cardiovascular medicine. These articles report studies that were presented at national meetings in 2018, including the Perioperative Medicine Summit, Society of General Internal Medicine, and Society of Hospital Medicine. These articles are grouped under 5 questions that will help guide clinical practice in perioperative cardiovascular medicine.

SHOULD ASPIRIN BE CONTINUED PERIOPERATIVELY IN PATIENTS WITH A CORONARY STENT?

The Perioperative Ischemic Evaluation 2 (POISE-2) trial¹ found that giving aspirin before surgery and throughout the early postoperative period had no significant effect on the rate of a composite of death or nonfatal myocardial infarction; moreover, aspirin increased the risk of major bleeding. However, many experts felt uncomfortable stopping aspirin preoperatively in patients taking it for secondary prophylaxis, particularly patients with a coronary stent.

[Graham MM, Sessler DI, Parlow JL, et al. Aspirin in patients with previous percutaneous coronary intervention undergoing noncardiac surgery. Ann Intern Med 2018; 168(4):237–244.]

This post hoc subgroup analysis² of POISE-2 evaluated the benefit and harm of perioperative aspirin in patients who had previously undergone percutaneous coronary intervention, more than 90% of whom had received a stent. Patients were age 45 or older with atherosclerotic heart disease or risk factors for it who had previously undergone percutaneous coronary intervention and were now undergoing noncardiac surgery.

Patients who had received a bare-metal stent within the previous 6 weeks or a drug-eluting stent within 12 months before surgery were excluded because guidelines at that time said to continue dual antiplatelet therapy for that long. Recommendations have since changed; the optimal duration for dual antiplatelet therapy with drug-eluting stents is now 6 months. Second-generation drug-eluting stents pose a lower risk of stent thrombosis and require a shorter duration of dual antiplatelet therapy than first-generation drug-eluting stents. Approximately 25% of the percutaneous coronary intervention subgroup had a drug-eluting stent, but the authors did not specify the type of drug-eluting stent.

The post hoc analysis² included a subgroup of 234 of 4,998 patients receiving aspirin and 236 of 5,012 patients receiving placebo initiated within 4 hours before surgery and continued postoperatively. The primary outcome measured was the rate of death or nonfatal myocardial infarction within 30 days after surgery, and bleeding was a secondary outcome.

Findings. Although the overall POISE-2 study found no benefit from aspirin, in the subgroup who had previously undergone percutaneous coronary intervention, aspirin significantly reduced the risk of the primary outcome, which occurred in 6% vs 11.5% of the patients:

• Absolute risk reduction 5.5% (95% confidence interval 0.4%–10.5%)
• Hazard ratio 0.50 (0.26–0.95).

The reduction was primarily due to fewer myocardial infarctions:

• Absolute risk reduction 5.9% (1.0%–10.8%)
• Hazard ratio 0.44 (0.22–0.87).

The type of stent had no effect on the primary outcome, although this subgroup analysis had limited power. In the nonpercutaneous coronary intervention subgroup, there was no significant difference in outcomes between the aspirin and placebo groups. This subgroup analysis was underpowered to evaluate the effect of aspirin on the composite of major and life-threatening bleeding in patients with prior percutaneous coronary intervention, which was reported as “uncertain” due to wide confidence intervals (absolute risk increase 1.3%, 95% confidence interval –2.6% to 5.2%), but the increased risk of major or life-threatening bleeding with aspirin demonstrated in the overall POISE-2 study population likely applies:

• Absolute risk increase 0.8% (0.1%–1.6%)
• Hazard ratio 1.22 (1.01–1.48).

Limitations. This was a nonspecified subgroup analysis that was underpowered and had a relatively small sample size with few events.

Conclusion. In the absence of a very high bleeding risk, continuing aspirin perioperatively in patients with prior percutaneous coronary intervention undergoing noncardiac surgery is more likely to result in benefit than harm. This finding is in agreement with current recommendations from the American College Cardiology and American Heart Association (class I; level of evidence C).³

Despite advances in anesthesia and surgical techniques, about 1% of patients over age 45 die within 30 days of noncardiac surgery.⁴ Studies have demonstrated a high mortality rate in patients who experience myocardial injury after noncardiac surgery (MINS), defined as elevations of troponin T with or without ischemic symptoms or electrocardiographic changes.⁵ Most of these studies used earlier, “non-high-sensitivity” troponin T assays. Fifth-generation, highly sensitive troponin T assays are now available that can detect troponin T at lower concentrations, but their utility in predicting postoperative outcomes remains uncertain. Two recent studies provide further insight into these issues.

[Writing Committee for the VISION Study Investigators, Devereaux PJ, Biccard BM, Sigamani A, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317(16):1642–1651.]

The Vascular Events in Noncardiac Surgery Patients Cohort Evaluation (VISION) study⁵ was an international, prospective cohort study that initially evaluated the association between MINS and the 30-day mortality rate using a non-high-sensitivity troponin T assay (Roche fourth-generation Elecsys TnT assay) in patients age 45 or older undergoing noncardiac surgery and requiring hospital admission for at least 1 night. After the first 15,000 patients, the study switched to the Roche fifth-generation assay, with measurements at 6 to 12 hours after surgery and on postoperative days 1, 2, and 3.

A 2017 analysis by Devereaux et al⁶ included only these later-enrolled patients and correlated their high-sensitivity troponin T levels with 30-day mortality rates. Patients with a level 14 ng/L or higher, the upper limit of normal in this study, were also assessed for ischemic symptoms and electrocardiographic changes. Although not required by the study, more than 7,800 patients had their troponin T levels measured before surgery, and the absolute change was also analyzed for an association with the 30-day mortality rate.

Findings. Of the 21,842 patients, about two-thirds underwent some form of major surgery; some of them had more than 1 type. A total of 1.2% of the patients died within 30 days of surgery.

Based on their analysis, the authors proposed that MINS be defined as:

• A postoperative troponin T level of 65 ng/L or higher, or
• A level in the range of 20 ng/L to less than 65 ng/L with an absolute increase from the preoperative level at least 5 ng/L, not attributable to a nonischemic cause.

Seventeen percent of the study patients met these criteria, and of these, 21.7% met the universal definition of myocardial infarction, although only 6.9% had symptoms of it.

Limitations. Only 40.4% of the patients had a preoperative high-sensitivity troponin T measurement for comparison, and in 13.8% of patients who had an elevated perioperative measurement, their preoperative value was the same or higher than their postoperative one. Thus, the incidence of MINS may have been overestimated if patients were otherwise not known to have troponin T elevations before surgery.

[Puelacher C, Lurati Buse G, Seeberger D, et al. Perioperative myocardial injury after noncardiac surgery: incidence, mortality, and characterization. Circulation 2018; 137(12):1221–1232.]

Puelacher et al⁷ investigated the prevalence of MINS in 2,018 patients at increased cardiovascular risk (age ≥ 65, or age ≥ 45 with a history of coronary artery disease, peripheral vascular disease, or stroke) who underwent major noncardiac surgery (planned overnight stay ≥ 24 hours) at a university hospital in Switzerland. Patients had their troponin T measured with a high-sensitivity assay within 30 days before surgery and on postoperative days 1 and 2.

Instead of MINS, the investigators used the term “perioperative myocardial injury” (PMI), defined as an absolute increase in troponin T of at least 14 ng/L from before surgery to the peak postoperative reading. Similar to MINS, PMI did not require ischemic features, but in this study, noncardiac triggers (sepsis, stroke, or pulmonary embolus) were not excluded.

Findings. PMI occurred in 16% of surgeries, and of the patients with PMI, 6% had typical chest pain and 18% had any ischemic symptoms. Unlike in the POISE-2 study discussed above, PMI triggered an automatic referral to a cardiologist.

The unadjusted 30-day mortality rate was 8.9% among patients with PMI and 1.5% in those without. Multivariable logistic regression analysis showed an adjusted hazard ratio for 30-day mortality of 2.7 (95% CI 1.5–4.8) for those with PMI vs without, and this difference persisted for at least 1 year.

In patients with PMI, the authors compared the 30-day mortality rate of those with no ischemic signs or symptoms (71% of the patients) with those who met the criteria for myocardial infarction and found no difference. Patients with PMI triggered by a noncardiac event had a worse prognosis than those with a presumed cardiac etiology.

Limitations. Despite the multivariate analysis that included adjustment for age, nonelective surgery, and Revised Cardiac Risk Index (RCRI), the increased risk associated with PMI could simply reflect higher risk at baseline. Although PMI resulted in automatic referral to a cardiologist, only 10% of patients eventually underwent coronary angiography; a similar percentage were discharged with additional medical therapy such as aspirin, a statin, or a beta-blocker. The effect of these interventions is not known.

Conclusions. MINS is common and has a strong association with mortality risk proportional to the degree of troponin T elevation using high-sensitivity assays, consistent with data from previous studies of earlier assays. Because the mechanism of MINS may differ from that of myocardial infarction, its prevention and treatment may differ, and it remains unclear how serial measurement in postoperative patients should change clinical practice.

The recently published Dabigatran in Patients With Myocardial Injury After Non-cardiac Surgery (MANAGE) trial⁸ suggests that dabigatran may reduce arterial and venous complications in patients with MINS, but the study had a number of limitations that may restrict the clinical applicability of this finding.

While awaiting further clinical outcomes data, pre- and postoperative troponin T measurement may be beneficial in higher-risk patients (such as those with cardiovascular disease or multiple RCRI risk factors) if the information will change perioperative management.

Intraoperative hypotension is associated with organ ischemia, which may cause postoperative myocardial infarction, myocardial injury, and acute kidney injury.⁹ Traditional anesthesia practice is to maintain intraoperative blood pressure within 20% of the preoperative baseline, based on the notion that hypertensive patients require higher perfusion pressures.

[Futier E, Lefrant J-Y, Guinot P-G, et al. Effect of individualized vs standard blood pressure management strategies on postoperative organ dysfunction among high-risk patients undergoing major surgery: a randomized clinical trial. JAMA 2017; 318(14):1346–1357.]

Futier et al¹⁰ sought to address uncertainty in intraoperative and immediate postoperative management of systolic blood pressure. In this multicenter, randomized, parallel-group trial, 298 patients at increased risk of postoperative renal complications were randomized to blood pressure management that was either “individualized” (within 10% of resting systolic pressure) or “standard” (≥ 80 mm Hg or ≥ 40% of resting systolic pressure) from induction to 4 hours postoperatively.

Blood pressure was monitored using radial arterial lines and maintained using a combination of intravenous fluids, norepinephrine (the first-line agent for the individualized group), and ephedrine (in the standard treatment group only). The primary outcome was a composite of systemic inflammatory response syndrome (SIRS) and organ dysfunction affecting at least 1 organ system (cardiovascular, respiratory, renal, hematologic, or neurologic).

Findings. Data on the primary outcome were available for 292 of 298 patients enrolled. The mean age was 70 years, 15% were women, and 82% had previously diagnosed hypertension. Despite the requirement for an elevated risk of acute kidney injury, only 13% of the patients had a baseline estimated glomerular filtration rate of less than 60 mL/min/1.73 m², and the median was 88 mL/min/1.73 m². Ninety-five percent of patients underwent abdominal surgery, and 50% of the surgeries were elective.

The mean systolic blood pressure was 123 mm Hg in the individualized treatment group compared with 116 mm Hg in the standard treatment group. Despite this small difference, 96% of individualized treatment patients received norepinephrine, compared with 26% in the standard treatment group.

The primary outcome of SIRS with organ dysfunction occurred in 38.1% of patients in the individualized treatment group and 51.7% of those in the standard treatment group. After adjusting for center, surgical urgency, surgical site, and acute kidney injury risk index, the relative risk of developing SIRS in those receiving individualized management was 0.73 (P = .02). Renal dysfunction (based on Acute Dialysis Quality Initiative criteria¹¹) occurred in 32.7% of individualized treatment patients and 49% of standardized treatment patients.  

Limitations of this study included differences in pharmacologic approach to maintain blood pressure in the 2 protocols (ephedrine and fluids vs norepinephrine) and a modest sample size.

Conclusions. Despite this, the difference in organ dysfunction was striking, with a number needed to treat of only 7 patients. This intervention extended 4 hours postoperatively, a time when many of these patients have left the postanesthesia care unit and have returned to hospitalist care on inpatient wards.

While optimal management of intraoperative and immediate postoperative blood pressure may not be settled, this study suggests that even mild relative hypotension may justify immediate action. Further studies may be useful to delineate high- and low-risk populations, the timing of greatest risk, and indications for intraarterial blood pressure monitoring.

[Salmasi V, Maheswari K, Yang D, et al. Relationship between intraoperative hypotension, defined by either reduction from baseline or absolute thresholds, and acute kidney and myocardial injury after noncardiac surgery: a retrospective cohort analysis. Anesthesiology 2017; 126(1):47–65.]

This retrospective cohort study¹² assessed the association between myocardial or kidney injury and absolute or relative thresholds of intraoperative mean arterial pressure. It included 57,315 adults who underwent inpatient noncardiac surgery, had a preoperative and at least 1 postoperative serum creatinine measurement within 7 days, and had blood pressure recorded in preoperative appointments within 6 months. Patients with chronic kidney disease (glomerular filtration rate < 60 mL/min/1.73 m²) and those on dialysis were excluded. The outcomes were MINS⁵ and acute kidney injury as defined by the Acute Kidney Injury Network.⁹

Findings. A mean arterial pressure below an absolute threshold of 65 mm Hg or a relative threshold of 20% lower than baseline value was associated with myocardial and kidney injury. At each threshold, prolonged periods of hypotension were associated with progressively increased risk.

An important conclusion of the study was that relative thresholds of mean arterial pressure were not any more predictive than absolute thresholds. Absolute thresholds are easier to use intraoperatively, especially when baseline values are not available. The authors did not find a clinically significant interaction between baseline blood pressure and the association of hypotension and myocardial and kidney injury.

Limitations included use of cardiac enzymes postoperatively to define MINS. Since these were not routinely collected, clinically silent myocardial injury may have been missed. Baseline blood pressure may have important implications in other forms of organ injury (ie, cerebral ischemia) that were not studied.

Summary. The lowest absolute mean arterial pressure is as predictive of postoperative myocardial and kidney injury as the relative pressure reduction, at least in patients with normal renal function. Limiting exposure to intraoperative hypotension is important. Baseline blood pressure values may have limited utility for intraoperative management.

In combination, these studies confirm that intraoperative hypotension is a predictor of postoperative organ dysfunction, but the definition and management remain unclear. While aggressive intraoperative management is likely beneficial, how to manage the anti­hypertensive therapy the patient has been taking as an outpatient when he or she comes into the hospital for surgery remains uncertain.

Perioperative stroke is an uncommon, severe complication of noncardiac surgery. The pathophysiology has been better defined in cardiac than in noncardiac surgeries. In nonsurgical patients, patent foramen ovale (PFO) is associated with stroke, even in patients considered to be at low risk.¹³ Perioperative patients have additional risk for venous thromboembolism and may have periprocedural antithrombotic medications altered, increasing their risk of paradoxical embolism through the PFO.

[Ng PY, Ng AK, Subramaniam B, et al. Association of preoperatively diagnosed patent foramen ovale with perioperative ischemic stroke. JAMA 2018; 319(5):452–462.]

This retrospective cohort study of noncardiac surgery patients at 3 hospitals¹⁴ sought to determine the association of preoperatively diagnosed PFO with the risk of perioperative ischemic stroke identified by International Classification of Diseases diagnoses.

Of 150,198 patients, 1.0% had a preoperative diagnosis of PFO, and at baseline, those with PFO had significantly more comorbidities than those without PFO. Stroke occurred in 3.2% of patients with PFO vs 0.5% of those without. Patients known to have a PFO were much more likely to have cardiovascular and thromboembolic risk factors for stroke. In the adjusted analysis, the absolute risk difference between groups was 0.4% (95% CI 0.2–0.6%), with an estimated perioperative stroke risk of 5.9 per 1,000 in patients with known patent foramen ovale and 2.2 per 1,000 in those without. A diagnosis of PFO was also associated with increased risk of large-vessel-territory stroke and more severe neurologic deficit.

Further attempts to adjust for baseline risk factors and other potential bias, including a propensity score-matched cohort analysis and an analysis limited to patients who had echocardiography performed in the same healthcare system, still showed a higher risk of perioperative stroke among patients with preoperatively detected patent foramen ovale.

Limitations. The study was retrospective and observational, used administrative data, and had a low rate of PFO diagnosis (1%), compared with about 25% in population-based studies.¹⁵ Indications for preoperative echocardiography are unknown. In addition, the study specifically examined preoperatively diagnosed PFO, rather than including those diagnosed in the postoperative period.

Discussion. How does this study affect clinical practice? The absolute stroke risk was increased by 0.4% in patients with PFO compared with those without. Although this is a relatively small increase, millions of patients undergo noncardiac surgery annually. The risks of therapeutic anticoagulation or PFO closure are likely too high in this context; however, clinicians may approach the perioperative management of antiplatelet agents and venous thromboembolism prophylaxis in patients with known PFO with additional caution.

HOW DOES TIMING OF EMERGENCY SURGERY AFTER PRIOR STROKE AFFECT OUTCOMES?

A history of stroke or transient ischemic attack is a known risk factor for perioperative vascular complications. A recent large cohort study demonstrated that a history of stroke within 9 months of elective surgery was associated with increased adverse outcomes.¹⁶ Little is known, however, of the perioperative risk in patients with a history of stroke who undergo emergency surgery.

[Christiansen MN, Andersson C, Gislason GH, et al. Risks of cardiovascular adverse events and death in patients with previous stroke undergoing emergency noncardiac, nonintracranial surgery: the importance of operative timing. Anesthesiology 2017; 127(1):9–19.]

In this study,¹⁷ all emergency noncardiac and nonintracranial surgeries from 2005 to 2011 were analyzed using multiple national patient registries in Denmark according to time elapsed between previous stroke and surgery. Primary outcomes were 30-day all-cause mortality and 30-day major adverse cardiac events (MACE), defined as nonfatal ischemic stroke, nonfatal myocardial infarction, and cardiovascular death. Statistical analysis to assess the risk of adverse outcomes included logistic regression models, spline analyses, and propensity-score matching.

Findings. The authors identified 146,694 emergency surgeries, with 7,861 patients (5.4%) having had a previous stroke (transient ischemic attacks and hemorrhagic strokes were not included). Rates of postoperative stroke were as follows:

• 9.9% in patents with a history of ischemic stroke within 3 months of surgery
• 2.8% in patients with a history of stroke 3 to 9 months before surgery
• 0.3% in patients with no previous stroke.

The risk plateaued when the time between stroke and surgery exceeded 4 to 5 months.¹⁵

Interestingly, in patients who underwent emergency surgery within 14 days of stroke, the risk of MACE was significantly lower immediately after surgery (1–3 days after stroke) compared with surgery that took place 4 to 14 days after stroke. The authors hypothesized that because cerebral autoregulation does not become compromised until approximately 5 days after a stroke, the risk was lower 1 to 3 days after surgery and increased thereafter.

Limitations of this study included the possibility of residual confounding, given its retrospective design using administrative data, not accounting for preoperative antithrombotic and anticoagulation therapy, and lack of information regarding the etiology of recurrent stroke (eg, thromboembolic, atherothrombotic, hypoperfusion).

Conclusions. Although it would be impractical to postpone emergency surgery in a patient who recently had a stroke, this study shows that the incidence rates of postoperative recurrent stroke and MACE are high. Therefore, it is important that the patient and perioperative team be aware of the risk. Further research is needed to confirm these estimates of postoperative adverse events in more diverse patient populations.

Primary care physicians are tasked with a wide variety of issues affecting men. This article reviews the latest research in 4 areas of men’s health commonly addressed in primary care:

• Medical management of benign prostatic hyperplasia (BPH)
• Prostate cancer screening and treatment
• Medical management of erectile dysfunction
• Use of supplements.

See related commentary

MEDICAL MANAGEMENT OF BPH

An 84-year-old man with a history of hypertension, type 2 diabetes, hyperlipidemia, BPH, mild cognitive impairment, and osteoarthritis presents for a 6-month follow-up, accompanied by his son.

Two years ago he was started on a 5-alpha reductase inhibitor and an alpha-blocker for worsening BPH symptoms. His BPH symptoms are currently under control, with an American Urological Association (AUA) symptom index score of 7 of a possible 35 (higher scores being worse).

However, both the patient and son are concerned about the number of medications he is on and wonder if some could be eliminated.

Assessment tools

BPH is a common cause of lower urinary tract symptoms in older men. Evidence-based tools to help the clinician and patient decide on when to consider treatment for symptoms are:

• The AUA symptom index¹
• The International Prostate Symptom Score (IPSS).²

An AUA symptom index score or IPSS score of 8 through 19 of a possible 35 is consistent with moderate symptoms, while a score of 20 or higher indicates severe symptoms.

Combination therapy or monotherapy?

Monotherapy with an alpha-blocker or a 5-alpha reductase inhibitor is often the first-line treatment for BPH-related lower urinary tract symptoms.³ However, combination therapy with both an alpha-blocker and a 5-alpha reductase inhibitor is another evidence-based option.

The Medical Therapy of Prostatic Symptoms study,⁴ a randomized controlled trial, reported that long-term combination therapy reduced the risk of BPH clinical progression better than monotherapy. The same trial also found that either combination therapy or finasteride alone (a 5-alpha reductase inhibitor) reduced the risk of acute urinary retention and the future need for invasive therapy.

Monotherapy after a period of combination therapy?

There is also evidence to support switching from combination to monotherapy after an initial treatment period.

Matsukawa et al⁵ examined the effects of withdrawing the alpha-blocker from BPH combination therapy in a study in 140 patients. For 12 months, all patients received the alpha-blocker silodosin and the 5-alpha reductase inhibitor dutasteride. At 12 months, the remaining 132 patients (8 patients had been lost to follow-up) were randomized to continue combination therapy or to take dutasteride alone for another 12 months. They were evaluated at 0, 12, and 24 months by questionnaires (the IPSS and Overactive Bladder Symptom Score) and urodynamic testing (uroflowmetry, cystometrography, and pressure-flow studies).

There were no significant differen­ces in subjective symptoms and bladder outlet obstruction between patients who continued combination therapy and those who switched to dutasteride monotherapy. In the monotherapy group, those whose symptoms worsened weighed more (68.8 kg vs 62.6 kg, P =.002) and had a higher body mass index (BMI) (26.2 kg/m² vs 22.8 kg/m², P < .001) than those whose symptoms stayed the same or got better.

These findings of successful alpha-blocker withdrawal were consistent with those of other studies.

The Symptom Management After Reducing Therapy study⁶ showed that 80% of men with an IPSS score less than 20 who changed to dutasteride monotherapy did not have a noticeable worsening of their symptoms.

Baldwin et al⁷ noted similar success after withdrawing the alpha-blocker doxazosin in patients on finasteride.

Review all medications

The National Health and Nutrition Examination Survey noted that the estimated prevalence of polypharmacy increased from 8% in 1999 to 15% in 2011.⁸ Many commonly used medications, such as decongestants, antihistamines, and anticholinergic agents, can worsen BPH symptoms,⁹ so it is reasonable to consistently review the patient’s medications to weigh the risks and benefits and determine which ones align with the patient’s personal care goals.

BPH: Take-home points

• Combination therapy with an alpha-blocker and a 5-alpha reductase inhibitor is an effective regimen for BPH.
• Polypharmacy is a significant problem in the elderly.
• Withdrawing the alpha-blocker component from BPH combination therapy can be considered after 1 year of combination therapy in patients whose symptoms have been well controlled.

A 60-year-old patient calls you after receiving his laboratory testing report from his insurance physical. His prostate-specific antigen (PSA) level is 5.1 ng/mL, and he has several questions:

• Should he have agreed to the screening?
• How effective is the screening?
• What are the next steps?

Is PSA screening useful?

Over the last few years, there has been great debate as to the utility of screening for prostate cancer.

The US Centers for Disease Control and Prevention¹⁰ reported that in 2014, an estimated 172,258 men in the United States were diagnosed with prostate cancer, but only 28,343 men died of it. These statistics support the notion that screening programs may be detecting what might otherwise be a silent disease.

The US Preventive Services Task Force (USPSTF)¹¹ recommends against blanket PSA screening, in view of the low probability that it reduces the risk of death from prostate cancer. For men ages 55 through 69, current guidelines give a grade C recommendation to PSA screening, meaning there is moderate agreement that the benefit is likely small, and screening should be selectively offered based on professional judgment and patient preference. In men ages 70 and older who are not at high risk, the guideline gives screening a grade D recommendation, meaning there is moderate evidence that there is no benefit from the practice. This is a change from the 2012 USPSTF guidelines,¹² which gave a grade D recommendation to PSA screening for all ages.

The American Urological Association¹³ recommends against PSA screening in men under age 40 or ages 70 and older. It does not recommend routine screening in those ages 40 to 54 at average risk, but it says the decision should be individualized in this age group in those at higher risk (eg, with a positive family history, African American). At ages 55 through 69, it recommends shared decision-making, taking into account cancer risk and life expectancy. In those who opt for screening, an interval of 2 years or more may be preferred over annual screening to reduce the risk of overdiagnosis.

The USPSTF recommendations rely heavily on data from 2 trials: the European Randomized Study of Screening for Prostate Cancer (ERSPC)¹⁴ and the Prostate, Lung, Colorectal, and Ovarian Screening (PLCO) trial.¹⁵

The ERSPC¹⁴ demonstrated that screening for prostate cancer reduced deaths from prostate cancer by 20%, with an absolute risk difference of 0.71 deaths per 1,000 men; 1,410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent 1 death from prostate cancer. Screening also decreased the risk of developing metastatic disease by 30%.¹⁶ On the negative side, screening increased the risk of overdiagnosis and other harms such as bleeding, sepsis, and incontinence.

The PLCO trial,¹⁵ in contrast, found no difference in death rates between men randomly assigned to annual screening and those assigned to usual care. Differences between the trial results were thought to be due to different practice settings as well as study implementation and compliance.

Tsodikov et al¹⁷ reanalyzed data from the ERSPC and the PLCO trial using 3 different mathematical models to estimate the effects of screening in both trials compared with no screening. The analysis found no evidence that the effects of screening vs not screening differed between the 2 trials, ultimately concluding that PSA screening reduced prostate cancer deaths by 25% to 32%, which the authors inferred was primarily a result of earlier detection of cancer.

The Cluster Randomized Trial of PSA Testing for Prostate Cancer,¹⁸ published in March 2018, explored the effect of single PSA screening vs no screening on prostate cancer mortality rates in 419,582 men ages 50 through 69. Although screening detected more cases of low-risk prostate cancer, there was no significant difference in prostate cancer mortality rates after a median follow-up of 10 years. However, 10% to 15% of the control group was estimated to have also been screened, and these results do not directly speak to the efficacy of serial PSA screening.

Extended follow-up of this trial is planned to report on long-term survival benefits and whether screening lowers the risk of metastasis.

Imaging-guided prostate biopsy

Once a patient is found to have an elevated PSA level, standard practice has been to perform transrectal ultrasonography to obtain 12 core biopsy samples. The results indicate whether the prostate contains cancer, how aggressive the cancer is (Gleason score), and whether there is extracapsular extension.

In the past, magnetic resonance imaging (MRI) of the prostate before biopsy was thought to be too costly, and many insurance plans do not currently cover it.

Pahwa et al,¹⁹ however, in a cost-effectiveness study using a decision-analysis model, found that using MRI to detect lesions and then guide biopsy by triaging patients into proper treatment pathways added health benefits in a cost-effective manner in 94.05% of simulations. These benefits were found across all age groups.

This study demonstrated that doctors could use MRI to better evaluate patients for potentially harmful lesions. If a focus of cancer is found, it can be biopsied; if no cancer is seen on MRI, the patient can avoid biopsy completely. Additionally, though MRI tended to miss low-risk cancers, these cancers are thought to disproportionately lead to higher healthcare costs through unnecessary treatment. Therefore, a negative MRI study was believed to be an excellent sign that the patient does not have aggressive prostate cancer. This approach led to a net gain of 0.251 additional quality-adjusted life years compared with the standard biopsy strategy.

The Prostate MRI Imaging Study²⁰ also found MRI to be effective in the prostate cancer workup. In this trial, 576 men who had never undergone biopsy underwent multiparametric MRI, transrectal ultrasonography-guided biopsy, and the reference standard, ie, transperineal template prostate mapping biopsy. Of those who underwent biopsy, 71% received a diagnosis of prostate cancer, and 40% had clinically significant disease. In patients with clinically significant disease, MRI was more sensitive than ultrasonography-guided biopsy (93% vs 48%, P < .0001) but less specific (41% vs 96%, P < .0001).

Based on these findings, if biopsy were performed only in those who had suspicious lesions on MRI, 27% of men with elevated PSA could avoid biopsy and its potential complications such as bleeding and sepsis, which occurred in 5.9% of the biopsy group.

The Prostate Evaluation for Clinically Important Disease: Sampling Using Image Guidance or Not? trial²¹ more recently studied MRI with or without targeted biopsy vs standard transrectal ultrasonography-guided biopsy in 500 men who had not undergone biopsy before, and reported similar results. MRI with or without biopsy led to fewer biopsies and less overdetection of clinically insignificant prostate cancers compared with the standard approach. Furthermore, those in the MRI-targeted biopsy group were 13% less likely to receive a diagnosis of clinically insignificant cancer than those who received the standard biopsy (adjusted difference −13 percentage points, 95% confidence interval [CI] −19 to −7, P < .001).

Together, these data provide another argument for adding multiparametric MRI to the workup of men with an elevated PSA level.

Surveillance vs treatment for prostate cancer

Once prostate cancer is diagnosed, surveillance is becoming an increasingly common management strategy.

The Prostate Cancer Intervention Versus Observation Trial (PIVOT),²² one of the largest and longest trials involving cancer patients, offered further evidence that active surveillance and less intervention for men with prostate cancer is a better approach in many cases. This trial compared prostatectomy and observation alone in a randomized fashion. Inclusion for the study required men to be medically fit for radical prostatectomy, along with having histologically confirmed localized prostate cancer (stage T1-T2NxM0 in the tumor-node-metastasis classification system) of any grade diagnosed within the last 12 months.

During 19.5 years of follow-up, 223 (61.3%) of the 364 men randomly assigned to radical prostatectomy died, compared with 245 (66.8%) of 367 men in the observation group; the difference was not statistically different (P = .06). Only 9.4% of the deaths were due to prostate cancer, 7.4% in the surgery group and 11.4% in the observation group (P = .06).

Surgery was associated with a lower all-cause mortality rate than observation in the subgroup of patients with intermediate-risk prostate cancer (defined as PSA 10–20 ng/mL and a Gleason score of 7). Surgery was also associated with less disease progression.²²

This finding is in line with previous data from the Scandinavian Prostate Cancer Group Study Number 4,²³ as well as the much larger Prostate Testing for Cancer and Treatment (ProtecT) trial,²⁴ both of which reported that metastasis was 1.5 and 2.6 times as common, respectively, in participants in the active surveillance groups. However, in the PIVOT trial, those in the surgery group were significantly more likely than those in the observation group to have erectile dysfunction and urinary incontinence at 10 years.

Therefore, in men with localized disease and in those with low-risk PSA levels, both the PIVOT and ProtecT trials suggest that death from prostate cancer is uncommon and that observation may be more appropriate.

Prostate cancer: Take-home points

• A new look at 2 large trials of PSA screening strengthened evidence that testing in the right patient population can reduce deaths from prostate cancer, but a third recently published trial that found no benefit from 1-time screening may reopen debate on the topic.
• MRI offers a better method than ultrasonography-guided biopsy to triage patients thought to be at high risk of prostate cancer and tends to limit costly overtreatment of disease that likely would not cause death.
• Surgery for prostate cancer may not prolong life but could reduce disease progression, at the risk of more adverse effects.
• Shared decision-making should be practiced when deciding whether to use active surveillance or active treatment of diagnosed prostate cancer.

A 62-year-old man with hypertension, hyperlipidemia, peripheral artery disease, and type 2 diabetes presents for a 6-month follow-up. His medications include aspirin, metformin, lisinopril, and atorvastatin, all of which he takes without problems. Over the past several months, he has noticed that his erections are not adequate for sexual intercourse. He recently heard that a generic version of sildenafil has just become available, and he wonders if it might benefit him.

Erectile dysfunction is common, associated with chronic diseases

Erectile dysfunction, ie, persistent inability to obtain and maintain an erection sufficient to permit satisfactory sexual intercourse,^25,26 is estimated to affect nearly 20% of men over the age of 20 and 75% of men over the age of 75.²⁷

In age-adjusted models, erectile dysfunction has been shown²⁸ to be associated with:

• History of cardiovascular disease (odds ratio [OR] 1.63, 95% CI 1.02–2.63)
• Diabetes (OR 3.90, 95% CI 2.16–7.04)
• Treated hypertension vs no hypertension (OR 2.22, 95% CI 1.30–3.80)
• Current smoking vs never smoking (OR 1.63, 95% CI 1.01–2.62)
• BMI greater than 30 kg/m² vs less than 25 kg/m² (OR 1.80, 95% CI 1.03–3.14).

Because of the strong association between cardiovascular disease and erectile dysfunction, the presence of one often suggests the need to screen for the other.²⁹ While tools such as the International Index of Erectile Function (IIEF-5) have been developed to evaluate erectile dysfunction, it is most often diagnosed on the basis of clinical impression, while validated assessment methods are reserved for clinical trials.²⁸

Multiple causes of erectile dysfunction

Erectile dysfunction arises from inadequate penile tissue response to a sexual signal. The response can be disrupted at several points. For example, damage to vascular smooth muscle cells (eg, from age or obesity) and endothelial cells (from smoking or diabetes) and narrowing of the vascular lumen (from atherosclerosis or hypertension) have all been shown to impair engorgement of the corpus cavernosum.³⁰ In addition, denervation from prostate surgery or spinal trauma and psychogenic causes should be recognized in discussions with patients.

Drugs for erectile dysfunction

Pharmacologic management of erectile dysfunction includes oral, sublingual, intracavernosal, and intraurethral therapies.³¹ Treatment in primary care settings usually includes addressing underlying chronic diseases³² and prescribing phosphodiesterase-5 inhibitors (sildenafil, tadalafil, vardenafil, and avanafil). These drugs work by increasing local concentrations of cyclic guanosine monophosphate in the corpus cavernosum to induce vasodilation.³³

While these 4 drugs are still patent-protected, a manufacturer has been allowed to introduce a generic version of sildenafil into US markets, and a generic version of tadalafil is expected to be available soon.

Sildenafil, tadalafil, and vardenafil have been studied and found to have some degree of effectiveness in erectile dysfunction caused by damage to the penile vasculature, denervation, and spinal cord injury.³⁴ All drugs of this class have adverse effects including headache, facial flushing, and nasal congestion, but the drugs are generally well tolerated.³⁵

Sildenafil and tadalafil improve IIEF-5 scores by a similar margin, raising scores on the erectile domain subsection from approximately 14 of a possible 30 to approximately 24 of 30 in a trial of both drugs.³⁶ However, multiple crossover studies comparing the 2 drugs have shown that nearly 75% of patients prefer tadalafil to sildenafil,^36,37 perhaps because of tadalafil’s longer duration of action.³⁴

There is little evidence to suggest that vardenafil is more effective or more often preferred by patients than tadalafil or sidenafil.^34,38 And though data on the newest drug on the market, avanafil, are limited, a meta-analysis concluded that it may be less effective than tadalafil and without significant differences in terms of safety.³⁹

Other treatments

Lifestyle modifications, especially smoking cessation and exercise, have been shown to reduce the risk of erectile dysfunction with varying effect sizes across studies.^40–42 Moreover, factors such as obesity, alcohol use, and smoking may cause irreversible harm, and thus a healthy lifestyle should be encouraged.⁴¹

While there is only weak evidence for the use of psychological interventions alone for treating most types of erectile dysfunction, one meta-analysis found that the combination of psychological intervention and a phosphodiesterase-5 inhibitor improved sexual satisfaction more than drug therapy alone.⁴³

Erectile dysfunction: Take-home points

• Erectile dysfunction is common, affecting nearly 20% of men over the age of 20 and over 75% of men over the age of 75.
• Erectile dysfunction is often associated with chronic disease and may suggest the need to screen for cardiovascular disease.
• Treating underlying chronic diseases may help, and phosphodiesterase-5 inhibitors are effective; tadalafil may be most often preferred.

A 68-year-old man with a history of hypertension, BPH, and erectile dysfunction presents for a 6-month follow-up. His medication use includes lisinopril, which he takes without problems. He denies any new physical symptoms. His physical examination is unremarkable. He says he has heard about supplements that might help with his sexual performance and hopes to discuss recommendations during the visit.

A burgeoning, unregulated industry

Since the passage of the Dietary Supplement and Health Education Act in 1994, a law that decreased oversight of the supplement industry, spending on supplements has skyrocketed to over $41.1 billion each year.⁴⁴ Advertisements for these products typically claim that they improve general mental and physical health, sexual and romantic performance, leanness, and muscularity.⁴⁵ A national survey of men ages 57 and older reported that the most popular products were aimed at nutrition (such as multivitamins), cardiovascular health (such as omega-3 fatty acids), and chronic conditions (such as saw palmetto for BPH).⁴⁶

Little evidence of efficacy

There is little evidence to support the use of most supplements to improve men’s health. For example, a study in 82,405 men found no association between mortality rates and multivitamin use (hazard ratio [HR] 1.07, 95% CI 0.96–1.19).⁴⁷ Even for specific uses, such as cognitive performance, randomized trials exploring the effects of multivitamins in men have been largely negative.⁴⁸

The positive trials that have been reported are often of low quality and are funded by supplement manufacturers. For example, one of the few trials that reported a positive association between multivitamin supplementation and cognition in men was underpowered (N = 51) and found improvement in only 1 of 19 cognitive domains.⁴⁹ Despite the poor design and results to the contrary, this industry-funded study nevertheless concluded that multivitamins may play a role in improving elements of memory.

Evidence of possible harm from antioxidants

While not always specific to men, many meta-analyses have explored the effects of antioxidant supplements on cardiovascular and mortality risk. Most of them concluded that antioxidant supplements have no benefit and that some may actually be harmful.

For example, multiple meta-analyses of vitamin E supplementation found no cardiovascular benefit but possible increases in all-cause mortality rates in those taking high doses (risk ratio 1.04, 95% CI 1.01–1.07).^50,51

Another meta-analysis of 180,938 partici­pants in high-quality studies found an increased risk of all-cause mortality associated with independent intake of several antioxidant vitamins, including beta-carotene (risk ratio 1.07, 95% CI 1.02–1.11) and vitamin A (risk ratio 1.16, 95% CI 1.10–1.24), while intake of vitamin C and selenium had no impact on mortality.⁵²

Similarly, although nearly 10% of US adults report taking omega-3 fatty acid supplements, a review of 24 randomized controlled trials and meta-analyses published between 2005 and 2012 concluded that only 2 supported the use of these supplements for any health benefit.⁵³

Can supplements improve sexual function, prostate health?

To improve sexual function. A 2015 narrative review of the ingredients in General Nutrition Center’s top 30 best-selling products targeted at improving men’s sexual performance (including improving libido and erectile dysfunction) found only poor evidence for any efficacy.⁵⁴ The few studies that did support the use of select supplements, including B vitamins in people with diabetes, L-arginine, and yohimbine, were deemed to be of poor quality or showed a smaller effect size compared with standard medical therapy.

To prevent prostate cancer. Studies of supplement use to improve prostate health have had mixed results. For example, multiple large case-control studies have suggested that taking vitamin D^55,56 or vitamin C⁵⁷ is not associated with prostate cancer risk, while increased vitamin A^58,59 and E^60,61 intake is associated with inconsistent increases in prostate cancer risk.

In the Selenium and Vitamin E Cancer Prevention Trial,⁶² a randomized controlled trial in 35,533 men, those assigned to receive vitamin E supplementation were 17% more likely to get prostate cancer than were those assigned to placebo (HR 1.17, 99% CI 1.004–1.36, P = .008).

However, there are plausible biologic links between nutraceuticals and prostate cancer. For example, studies have linked genetic polymorphisms in vitamin D receptors⁶³ as well as intake of natural androgen receptor modulators, such as the most active polyphenol in green tea,⁶⁴ to prostate cancer risk and aggressiveness in certain populations. This led a recent review to conclude that there is some biologic plausibility, but at present little epidemiologic evidence, to support any dietary supplement’s ability to broadly affect prostate cancer risk.⁶⁵

Interest continues in exploring the targeted use of nutraceuticals as adjuvant therapy in specific populations at risk of prostate cancer.^66,67

To treat BPH. There is a similar dearth of clinical or population-based evidence that supplements can broadly affect BPH symptoms. For example, in a 2012 Cochrane review of Serenoa repens (saw palmetto) utilizing only high-quality evidence, there was no evidence that supplement use significantly reduced lower urinary tract symptoms, nocturia, or peak urine flow in BPH patients, and this was true even when the supplement was taken at triple-strength doses.⁶⁸

For other diseases. There is also limited evidence that supplements can affect other chronic diseases. For example, a meta-analysis of 3,803 patients found that glucosamine, chondroitin, and their combination had no impact on joint pain or joint space narrowing in patients with osteoarthritis of the knee or hip.⁶⁹

Even when there is some evidence to suggest benefit from supplementation, study heterogeneity and varying evidence quality limit confidence in the conclusions. For example, meta-analyses suggest garlic may improve blood pressure control in those with hypertension⁷⁰ and improve lipid and blood glucose control in type 2 diabetes.⁷¹ However, most of the trials included in those systematic reviews were underpowered, with samples as low as 10 patients, and many suffered from improper design, such as inadequate blinding of researchers. In addition, these meta-analyses often do not report adverse events, suggesting that higher quality studies would be needed to adequately measure event rates. As such, there is need for caution and a case-by-case review before recommending even a seemingly benign supplement like garlic to patients.

In total, there is only limited evidence to support the efficacy of supplements across many diseases and concerns common to men in primary care. This includes improving general health, cardiovascular health, sexual functioning, or other chronic diseases. While a supplement’s placebo effect may at times provide some benefit, supplements are much less strictly regulated since the passing of the 1994 act, and even vitamin supplementation has been shown to be associated with negative health outcomes. As such, a patient’s use of supplements requires careful consideration and shared decision-making.

Supplements: Take-home points

• Supplements are only loosely regulated by the federal government.
• There is some biologic but limited epidemiologic evidence for the use of multivitamins to improve cognition or mortality rates; for the use of antioxidant vitamins or omega-3 fatty acids to improve cardiovascular health; for the use of any of the top-selling sexual enhancement supplements to improve libido or erectile function; and for the use of vitamins or other supplements for improving BPH or reducing prostate cancer risk. Using supplements may in some cases be harmful.
• Given the heterogeneity of studies of supplements to manage chronic diseases and a lack of reporting of adverse events, careful consideration is needed when recommending supplements to patients.

Primary care physicians are tasked with a wide variety of issues affecting men. This article reviews the latest research in 4 areas of men’s health commonly addressed in primary care:

• Medical management of benign prostatic hyperplasia (BPH)
• Prostate cancer screening and treatment
• Medical management of erectile dysfunction
• Use of supplements.

See related commentary

MEDICAL MANAGEMENT OF BPH

An 84-year-old man with a history of hypertension, type 2 diabetes, hyperlipidemia, BPH, mild cognitive impairment, and osteoarthritis presents for a 6-month follow-up, accompanied by his son.

Two years ago he was started on a 5-alpha reductase inhibitor and an alpha-blocker for worsening BPH symptoms. His BPH symptoms are currently under control, with an American Urological Association (AUA) symptom index score of 7 of a possible 35 (higher scores being worse).

However, both the patient and son are concerned about the number of medications he is on and wonder if some could be eliminated.

Assessment tools

BPH is a common cause of lower urinary tract symptoms in older men. Evidence-based tools to help the clinician and patient decide on when to consider treatment for symptoms are:

• The AUA symptom index¹
• The International Prostate Symptom Score (IPSS).²

An AUA symptom index score or IPSS score of 8 through 19 of a possible 35 is consistent with moderate symptoms, while a score of 20 or higher indicates severe symptoms.

Combination therapy or monotherapy?

Monotherapy with an alpha-blocker or a 5-alpha reductase inhibitor is often the first-line treatment for BPH-related lower urinary tract symptoms.³ However, combination therapy with both an alpha-blocker and a 5-alpha reductase inhibitor is another evidence-based option.

The Medical Therapy of Prostatic Symptoms study,⁴ a randomized controlled trial, reported that long-term combination therapy reduced the risk of BPH clinical progression better than monotherapy. The same trial also found that either combination therapy or finasteride alone (a 5-alpha reductase inhibitor) reduced the risk of acute urinary retention and the future need for invasive therapy.

Monotherapy after a period of combination therapy?

There is also evidence to support switching from combination to monotherapy after an initial treatment period.

Matsukawa et al⁵ examined the effects of withdrawing the alpha-blocker from BPH combination therapy in a study in 140 patients. For 12 months, all patients received the alpha-blocker silodosin and the 5-alpha reductase inhibitor dutasteride. At 12 months, the remaining 132 patients (8 patients had been lost to follow-up) were randomized to continue combination therapy or to take dutasteride alone for another 12 months. They were evaluated at 0, 12, and 24 months by questionnaires (the IPSS and Overactive Bladder Symptom Score) and urodynamic testing (uroflowmetry, cystometrography, and pressure-flow studies).

There were no significant differen­ces in subjective symptoms and bladder outlet obstruction between patients who continued combination therapy and those who switched to dutasteride monotherapy. In the monotherapy group, those whose symptoms worsened weighed more (68.8 kg vs 62.6 kg, P =.002) and had a higher body mass index (BMI) (26.2 kg/m² vs 22.8 kg/m², P < .001) than those whose symptoms stayed the same or got better.

These findings of successful alpha-blocker withdrawal were consistent with those of other studies.

The Symptom Management After Reducing Therapy study⁶ showed that 80% of men with an IPSS score less than 20 who changed to dutasteride monotherapy did not have a noticeable worsening of their symptoms.

Baldwin et al⁷ noted similar success after withdrawing the alpha-blocker doxazosin in patients on finasteride.

Review all medications

The National Health and Nutrition Examination Survey noted that the estimated prevalence of polypharmacy increased from 8% in 1999 to 15% in 2011.⁸ Many commonly used medications, such as decongestants, antihistamines, and anticholinergic agents, can worsen BPH symptoms,⁹ so it is reasonable to consistently review the patient’s medications to weigh the risks and benefits and determine which ones align with the patient’s personal care goals.

BPH: Take-home points

• Combination therapy with an alpha-blocker and a 5-alpha reductase inhibitor is an effective regimen for BPH.
• Polypharmacy is a significant problem in the elderly.
• Withdrawing the alpha-blocker component from BPH combination therapy can be considered after 1 year of combination therapy in patients whose symptoms have been well controlled.

A 60-year-old patient calls you after receiving his laboratory testing report from his insurance physical. His prostate-specific antigen (PSA) level is 5.1 ng/mL, and he has several questions:

• Should he have agreed to the screening?
• How effective is the screening?
• What are the next steps?

Is PSA screening useful?

Over the last few years, there has been great debate as to the utility of screening for prostate cancer.

The US Centers for Disease Control and Prevention¹⁰ reported that in 2014, an estimated 172,258 men in the United States were diagnosed with prostate cancer, but only 28,343 men died of it. These statistics support the notion that screening programs may be detecting what might otherwise be a silent disease.

The US Preventive Services Task Force (USPSTF)¹¹ recommends against blanket PSA screening, in view of the low probability that it reduces the risk of death from prostate cancer. For men ages 55 through 69, current guidelines give a grade C recommendation to PSA screening, meaning there is moderate agreement that the benefit is likely small, and screening should be selectively offered based on professional judgment and patient preference. In men ages 70 and older who are not at high risk, the guideline gives screening a grade D recommendation, meaning there is moderate evidence that there is no benefit from the practice. This is a change from the 2012 USPSTF guidelines,¹² which gave a grade D recommendation to PSA screening for all ages.

The American Urological Association¹³ recommends against PSA screening in men under age 40 or ages 70 and older. It does not recommend routine screening in those ages 40 to 54 at average risk, but it says the decision should be individualized in this age group in those at higher risk (eg, with a positive family history, African American). At ages 55 through 69, it recommends shared decision-making, taking into account cancer risk and life expectancy. In those who opt for screening, an interval of 2 years or more may be preferred over annual screening to reduce the risk of overdiagnosis.

The USPSTF recommendations rely heavily on data from 2 trials: the European Randomized Study of Screening for Prostate Cancer (ERSPC)¹⁴ and the Prostate, Lung, Colorectal, and Ovarian Screening (PLCO) trial.¹⁵

The ERSPC¹⁴ demonstrated that screening for prostate cancer reduced deaths from prostate cancer by 20%, with an absolute risk difference of 0.71 deaths per 1,000 men; 1,410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent 1 death from prostate cancer. Screening also decreased the risk of developing metastatic disease by 30%.¹⁶ On the negative side, screening increased the risk of overdiagnosis and other harms such as bleeding, sepsis, and incontinence.

The PLCO trial,¹⁵ in contrast, found no difference in death rates between men randomly assigned to annual screening and those assigned to usual care. Differences between the trial results were thought to be due to different practice settings as well as study implementation and compliance.

Tsodikov et al¹⁷ reanalyzed data from the ERSPC and the PLCO trial using 3 different mathematical models to estimate the effects of screening in both trials compared with no screening. The analysis found no evidence that the effects of screening vs not screening differed between the 2 trials, ultimately concluding that PSA screening reduced prostate cancer deaths by 25% to 32%, which the authors inferred was primarily a result of earlier detection of cancer.

The Cluster Randomized Trial of PSA Testing for Prostate Cancer,¹⁸ published in March 2018, explored the effect of single PSA screening vs no screening on prostate cancer mortality rates in 419,582 men ages 50 through 69. Although screening detected more cases of low-risk prostate cancer, there was no significant difference in prostate cancer mortality rates after a median follow-up of 10 years. However, 10% to 15% of the control group was estimated to have also been screened, and these results do not directly speak to the efficacy of serial PSA screening.

Extended follow-up of this trial is planned to report on long-term survival benefits and whether screening lowers the risk of metastasis.

Imaging-guided prostate biopsy

Once a patient is found to have an elevated PSA level, standard practice has been to perform transrectal ultrasonography to obtain 12 core biopsy samples. The results indicate whether the prostate contains cancer, how aggressive the cancer is (Gleason score), and whether there is extracapsular extension.

In the past, magnetic resonance imaging (MRI) of the prostate before biopsy was thought to be too costly, and many insurance plans do not currently cover it.

Pahwa et al,¹⁹ however, in a cost-effectiveness study using a decision-analysis model, found that using MRI to detect lesions and then guide biopsy by triaging patients into proper treatment pathways added health benefits in a cost-effective manner in 94.05% of simulations. These benefits were found across all age groups.

This study demonstrated that doctors could use MRI to better evaluate patients for potentially harmful lesions. If a focus of cancer is found, it can be biopsied; if no cancer is seen on MRI, the patient can avoid biopsy completely. Additionally, though MRI tended to miss low-risk cancers, these cancers are thought to disproportionately lead to higher healthcare costs through unnecessary treatment. Therefore, a negative MRI study was believed to be an excellent sign that the patient does not have aggressive prostate cancer. This approach led to a net gain of 0.251 additional quality-adjusted life years compared with the standard biopsy strategy.

The Prostate MRI Imaging Study²⁰ also found MRI to be effective in the prostate cancer workup. In this trial, 576 men who had never undergone biopsy underwent multiparametric MRI, transrectal ultrasonography-guided biopsy, and the reference standard, ie, transperineal template prostate mapping biopsy. Of those who underwent biopsy, 71% received a diagnosis of prostate cancer, and 40% had clinically significant disease. In patients with clinically significant disease, MRI was more sensitive than ultrasonography-guided biopsy (93% vs 48%, P < .0001) but less specific (41% vs 96%, P < .0001).

Based on these findings, if biopsy were performed only in those who had suspicious lesions on MRI, 27% of men with elevated PSA could avoid biopsy and its potential complications such as bleeding and sepsis, which occurred in 5.9% of the biopsy group.

The Prostate Evaluation for Clinically Important Disease: Sampling Using Image Guidance or Not? trial²¹ more recently studied MRI with or without targeted biopsy vs standard transrectal ultrasonography-guided biopsy in 500 men who had not undergone biopsy before, and reported similar results. MRI with or without biopsy led to fewer biopsies and less overdetection of clinically insignificant prostate cancers compared with the standard approach. Furthermore, those in the MRI-targeted biopsy group were 13% less likely to receive a diagnosis of clinically insignificant cancer than those who received the standard biopsy (adjusted difference −13 percentage points, 95% confidence interval [CI] −19 to −7, P < .001).

Together, these data provide another argument for adding multiparametric MRI to the workup of men with an elevated PSA level.

Surveillance vs treatment for prostate cancer

Once prostate cancer is diagnosed, surveillance is becoming an increasingly common management strategy.

The Prostate Cancer Intervention Versus Observation Trial (PIVOT),²² one of the largest and longest trials involving cancer patients, offered further evidence that active surveillance and less intervention for men with prostate cancer is a better approach in many cases. This trial compared prostatectomy and observation alone in a randomized fashion. Inclusion for the study required men to be medically fit for radical prostatectomy, along with having histologically confirmed localized prostate cancer (stage T1-T2NxM0 in the tumor-node-metastasis classification system) of any grade diagnosed within the last 12 months.

During 19.5 years of follow-up, 223 (61.3%) of the 364 men randomly assigned to radical prostatectomy died, compared with 245 (66.8%) of 367 men in the observation group; the difference was not statistically different (P = .06). Only 9.4% of the deaths were due to prostate cancer, 7.4% in the surgery group and 11.4% in the observation group (P = .06).

Surgery was associated with a lower all-cause mortality rate than observation in the subgroup of patients with intermediate-risk prostate cancer (defined as PSA 10–20 ng/mL and a Gleason score of 7). Surgery was also associated with less disease progression.²²

This finding is in line with previous data from the Scandinavian Prostate Cancer Group Study Number 4,²³ as well as the much larger Prostate Testing for Cancer and Treatment (ProtecT) trial,²⁴ both of which reported that metastasis was 1.5 and 2.6 times as common, respectively, in participants in the active surveillance groups. However, in the PIVOT trial, those in the surgery group were significantly more likely than those in the observation group to have erectile dysfunction and urinary incontinence at 10 years.

Therefore, in men with localized disease and in those with low-risk PSA levels, both the PIVOT and ProtecT trials suggest that death from prostate cancer is uncommon and that observation may be more appropriate.

Prostate cancer: Take-home points

• A new look at 2 large trials of PSA screening strengthened evidence that testing in the right patient population can reduce deaths from prostate cancer, but a third recently published trial that found no benefit from 1-time screening may reopen debate on the topic.
• MRI offers a better method than ultrasonography-guided biopsy to triage patients thought to be at high risk of prostate cancer and tends to limit costly overtreatment of disease that likely would not cause death.
• Surgery for prostate cancer may not prolong life but could reduce disease progression, at the risk of more adverse effects.
• Shared decision-making should be practiced when deciding whether to use active surveillance or active treatment of diagnosed prostate cancer.

A 62-year-old man with hypertension, hyperlipidemia, peripheral artery disease, and type 2 diabetes presents for a 6-month follow-up. His medications include aspirin, metformin, lisinopril, and atorvastatin, all of which he takes without problems. Over the past several months, he has noticed that his erections are not adequate for sexual intercourse. He recently heard that a generic version of sildenafil has just become available, and he wonders if it might benefit him.

Erectile dysfunction is common, associated with chronic diseases

Erectile dysfunction, ie, persistent inability to obtain and maintain an erection sufficient to permit satisfactory sexual intercourse,^25,26 is estimated to affect nearly 20% of men over the age of 20 and 75% of men over the age of 75.²⁷

In age-adjusted models, erectile dysfunction has been shown²⁸ to be associated with:

• History of cardiovascular disease (odds ratio [OR] 1.63, 95% CI 1.02–2.63)
• Diabetes (OR 3.90, 95% CI 2.16–7.04)
• Treated hypertension vs no hypertension (OR 2.22, 95% CI 1.30–3.80)
• Current smoking vs never smoking (OR 1.63, 95% CI 1.01–2.62)
• BMI greater than 30 kg/m² vs less than 25 kg/m² (OR 1.80, 95% CI 1.03–3.14).

Because of the strong association between cardiovascular disease and erectile dysfunction, the presence of one often suggests the need to screen for the other.²⁹ While tools such as the International Index of Erectile Function (IIEF-5) have been developed to evaluate erectile dysfunction, it is most often diagnosed on the basis of clinical impression, while validated assessment methods are reserved for clinical trials.²⁸

Multiple causes of erectile dysfunction

Erectile dysfunction arises from inadequate penile tissue response to a sexual signal. The response can be disrupted at several points. For example, damage to vascular smooth muscle cells (eg, from age or obesity) and endothelial cells (from smoking or diabetes) and narrowing of the vascular lumen (from atherosclerosis or hypertension) have all been shown to impair engorgement of the corpus cavernosum.³⁰ In addition, denervation from prostate surgery or spinal trauma and psychogenic causes should be recognized in discussions with patients.

Drugs for erectile dysfunction

Pharmacologic management of erectile dysfunction includes oral, sublingual, intracavernosal, and intraurethral therapies.³¹ Treatment in primary care settings usually includes addressing underlying chronic diseases³² and prescribing phosphodiesterase-5 inhibitors (sildenafil, tadalafil, vardenafil, and avanafil). These drugs work by increasing local concentrations of cyclic guanosine monophosphate in the corpus cavernosum to induce vasodilation.³³

While these 4 drugs are still patent-protected, a manufacturer has been allowed to introduce a generic version of sildenafil into US markets, and a generic version of tadalafil is expected to be available soon.

Sildenafil, tadalafil, and vardenafil have been studied and found to have some degree of effectiveness in erectile dysfunction caused by damage to the penile vasculature, denervation, and spinal cord injury.³⁴ All drugs of this class have adverse effects including headache, facial flushing, and nasal congestion, but the drugs are generally well tolerated.³⁵

Sildenafil and tadalafil improve IIEF-5 scores by a similar margin, raising scores on the erectile domain subsection from approximately 14 of a possible 30 to approximately 24 of 30 in a trial of both drugs.³⁶ However, multiple crossover studies comparing the 2 drugs have shown that nearly 75% of patients prefer tadalafil to sildenafil,^36,37 perhaps because of tadalafil’s longer duration of action.³⁴

There is little evidence to suggest that vardenafil is more effective or more often preferred by patients than tadalafil or sidenafil.^34,38 And though data on the newest drug on the market, avanafil, are limited, a meta-analysis concluded that it may be less effective than tadalafil and without significant differences in terms of safety.³⁹

Other treatments

Lifestyle modifications, especially smoking cessation and exercise, have been shown to reduce the risk of erectile dysfunction with varying effect sizes across studies.^40–42 Moreover, factors such as obesity, alcohol use, and smoking may cause irreversible harm, and thus a healthy lifestyle should be encouraged.⁴¹

While there is only weak evidence for the use of psychological interventions alone for treating most types of erectile dysfunction, one meta-analysis found that the combination of psychological intervention and a phosphodiesterase-5 inhibitor improved sexual satisfaction more than drug therapy alone.⁴³

Erectile dysfunction: Take-home points

• Erectile dysfunction is common, affecting nearly 20% of men over the age of 20 and over 75% of men over the age of 75.
• Erectile dysfunction is often associated with chronic disease and may suggest the need to screen for cardiovascular disease.
• Treating underlying chronic diseases may help, and phosphodiesterase-5 inhibitors are effective; tadalafil may be most often preferred.

A 68-year-old man with a history of hypertension, BPH, and erectile dysfunction presents for a 6-month follow-up. His medication use includes lisinopril, which he takes without problems. He denies any new physical symptoms. His physical examination is unremarkable. He says he has heard about supplements that might help with his sexual performance and hopes to discuss recommendations during the visit.

A burgeoning, unregulated industry

Since the passage of the Dietary Supplement and Health Education Act in 1994, a law that decreased oversight of the supplement industry, spending on supplements has skyrocketed to over $41.1 billion each year.⁴⁴ Advertisements for these products typically claim that they improve general mental and physical health, sexual and romantic performance, leanness, and muscularity.⁴⁵ A national survey of men ages 57 and older reported that the most popular products were aimed at nutrition (such as multivitamins), cardiovascular health (such as omega-3 fatty acids), and chronic conditions (such as saw palmetto for BPH).⁴⁶

Little evidence of efficacy

There is little evidence to support the use of most supplements to improve men’s health. For example, a study in 82,405 men found no association between mortality rates and multivitamin use (hazard ratio [HR] 1.07, 95% CI 0.96–1.19).⁴⁷ Even for specific uses, such as cognitive performance, randomized trials exploring the effects of multivitamins in men have been largely negative.⁴⁸

The positive trials that have been reported are often of low quality and are funded by supplement manufacturers. For example, one of the few trials that reported a positive association between multivitamin supplementation and cognition in men was underpowered (N = 51) and found improvement in only 1 of 19 cognitive domains.⁴⁹ Despite the poor design and results to the contrary, this industry-funded study nevertheless concluded that multivitamins may play a role in improving elements of memory.

Evidence of possible harm from antioxidants

While not always specific to men, many meta-analyses have explored the effects of antioxidant supplements on cardiovascular and mortality risk. Most of them concluded that antioxidant supplements have no benefit and that some may actually be harmful.

For example, multiple meta-analyses of vitamin E supplementation found no cardiovascular benefit but possible increases in all-cause mortality rates in those taking high doses (risk ratio 1.04, 95% CI 1.01–1.07).^50,51

Another meta-analysis of 180,938 partici­pants in high-quality studies found an increased risk of all-cause mortality associated with independent intake of several antioxidant vitamins, including beta-carotene (risk ratio 1.07, 95% CI 1.02–1.11) and vitamin A (risk ratio 1.16, 95% CI 1.10–1.24), while intake of vitamin C and selenium had no impact on mortality.⁵²

Similarly, although nearly 10% of US adults report taking omega-3 fatty acid supplements, a review of 24 randomized controlled trials and meta-analyses published between 2005 and 2012 concluded that only 2 supported the use of these supplements for any health benefit.⁵³

Can supplements improve sexual function, prostate health?

To improve sexual function. A 2015 narrative review of the ingredients in General Nutrition Center’s top 30 best-selling products targeted at improving men’s sexual performance (including improving libido and erectile dysfunction) found only poor evidence for any efficacy.⁵⁴ The few studies that did support the use of select supplements, including B vitamins in people with diabetes, L-arginine, and yohimbine, were deemed to be of poor quality or showed a smaller effect size compared with standard medical therapy.

To prevent prostate cancer. Studies of supplement use to improve prostate health have had mixed results. For example, multiple large case-control studies have suggested that taking vitamin D^55,56 or vitamin C⁵⁷ is not associated with prostate cancer risk, while increased vitamin A^58,59 and E^60,61 intake is associated with inconsistent increases in prostate cancer risk.

In the Selenium and Vitamin E Cancer Prevention Trial,⁶² a randomized controlled trial in 35,533 men, those assigned to receive vitamin E supplementation were 17% more likely to get prostate cancer than were those assigned to placebo (HR 1.17, 99% CI 1.004–1.36, P = .008).

However, there are plausible biologic links between nutraceuticals and prostate cancer. For example, studies have linked genetic polymorphisms in vitamin D receptors⁶³ as well as intake of natural androgen receptor modulators, such as the most active polyphenol in green tea,⁶⁴ to prostate cancer risk and aggressiveness in certain populations. This led a recent review to conclude that there is some biologic plausibility, but at present little epidemiologic evidence, to support any dietary supplement’s ability to broadly affect prostate cancer risk.⁶⁵

Interest continues in exploring the targeted use of nutraceuticals as adjuvant therapy in specific populations at risk of prostate cancer.^66,67

To treat BPH. There is a similar dearth of clinical or population-based evidence that supplements can broadly affect BPH symptoms. For example, in a 2012 Cochrane review of Serenoa repens (saw palmetto) utilizing only high-quality evidence, there was no evidence that supplement use significantly reduced lower urinary tract symptoms, nocturia, or peak urine flow in BPH patients, and this was true even when the supplement was taken at triple-strength doses.⁶⁸

For other diseases. There is also limited evidence that supplements can affect other chronic diseases. For example, a meta-analysis of 3,803 patients found that glucosamine, chondroitin, and their combination had no impact on joint pain or joint space narrowing in patients with osteoarthritis of the knee or hip.⁶⁹

Even when there is some evidence to suggest benefit from supplementation, study heterogeneity and varying evidence quality limit confidence in the conclusions. For example, meta-analyses suggest garlic may improve blood pressure control in those with hypertension⁷⁰ and improve lipid and blood glucose control in type 2 diabetes.⁷¹ However, most of the trials included in those systematic reviews were underpowered, with samples as low as 10 patients, and many suffered from improper design, such as inadequate blinding of researchers. In addition, these meta-analyses often do not report adverse events, suggesting that higher quality studies would be needed to adequately measure event rates. As such, there is need for caution and a case-by-case review before recommending even a seemingly benign supplement like garlic to patients.

In total, there is only limited evidence to support the efficacy of supplements across many diseases and concerns common to men in primary care. This includes improving general health, cardiovascular health, sexual functioning, or other chronic diseases. While a supplement’s placebo effect may at times provide some benefit, supplements are much less strictly regulated since the passing of the 1994 act, and even vitamin supplementation has been shown to be associated with negative health outcomes. As such, a patient’s use of supplements requires careful consideration and shared decision-making.

Supplements: Take-home points

• Supplements are only loosely regulated by the federal government.
• There is some biologic but limited epidemiologic evidence for the use of multivitamins to improve cognition or mortality rates; for the use of antioxidant vitamins or omega-3 fatty acids to improve cardiovascular health; for the use of any of the top-selling sexual enhancement supplements to improve libido or erectile function; and for the use of vitamins or other supplements for improving BPH or reducing prostate cancer risk. Using supplements may in some cases be harmful.
• Given the heterogeneity of studies of supplements to manage chronic diseases and a lack of reporting of adverse events, careful consideration is needed when recommending supplements to patients.

Primary care physicians are tasked with a wide variety of issues affecting men. This article reviews the latest research in 4 areas of men’s health commonly addressed in primary care:

• Medical management of benign prostatic hyperplasia (BPH)
• Prostate cancer screening and treatment
• Medical management of erectile dysfunction
• Use of supplements.

See related commentary

MEDICAL MANAGEMENT OF BPH

An 84-year-old man with a history of hypertension, type 2 diabetes, hyperlipidemia, BPH, mild cognitive impairment, and osteoarthritis presents for a 6-month follow-up, accompanied by his son.

Two years ago he was started on a 5-alpha reductase inhibitor and an alpha-blocker for worsening BPH symptoms. His BPH symptoms are currently under control, with an American Urological Association (AUA) symptom index score of 7 of a possible 35 (higher scores being worse).

However, both the patient and son are concerned about the number of medications he is on and wonder if some could be eliminated.

Assessment tools

BPH is a common cause of lower urinary tract symptoms in older men. Evidence-based tools to help the clinician and patient decide on when to consider treatment for symptoms are:

• The AUA symptom index¹
• The International Prostate Symptom Score (IPSS).²

An AUA symptom index score or IPSS score of 8 through 19 of a possible 35 is consistent with moderate symptoms, while a score of 20 or higher indicates severe symptoms.

Combination therapy or monotherapy?

Monotherapy with an alpha-blocker or a 5-alpha reductase inhibitor is often the first-line treatment for BPH-related lower urinary tract symptoms.³ However, combination therapy with both an alpha-blocker and a 5-alpha reductase inhibitor is another evidence-based option.

The Medical Therapy of Prostatic Symptoms study,⁴ a randomized controlled trial, reported that long-term combination therapy reduced the risk of BPH clinical progression better than monotherapy. The same trial also found that either combination therapy or finasteride alone (a 5-alpha reductase inhibitor) reduced the risk of acute urinary retention and the future need for invasive therapy.

Monotherapy after a period of combination therapy?

There is also evidence to support switching from combination to monotherapy after an initial treatment period.

Matsukawa et al⁵ examined the effects of withdrawing the alpha-blocker from BPH combination therapy in a study in 140 patients. For 12 months, all patients received the alpha-blocker silodosin and the 5-alpha reductase inhibitor dutasteride. At 12 months, the remaining 132 patients (8 patients had been lost to follow-up) were randomized to continue combination therapy or to take dutasteride alone for another 12 months. They were evaluated at 0, 12, and 24 months by questionnaires (the IPSS and Overactive Bladder Symptom Score) and urodynamic testing (uroflowmetry, cystometrography, and pressure-flow studies).

There were no significant differen­ces in subjective symptoms and bladder outlet obstruction between patients who continued combination therapy and those who switched to dutasteride monotherapy. In the monotherapy group, those whose symptoms worsened weighed more (68.8 kg vs 62.6 kg, P =.002) and had a higher body mass index (BMI) (26.2 kg/m² vs 22.8 kg/m², P < .001) than those whose symptoms stayed the same or got better.

These findings of successful alpha-blocker withdrawal were consistent with those of other studies.

The Symptom Management After Reducing Therapy study⁶ showed that 80% of men with an IPSS score less than 20 who changed to dutasteride monotherapy did not have a noticeable worsening of their symptoms.

Baldwin et al⁷ noted similar success after withdrawing the alpha-blocker doxazosin in patients on finasteride.

Review all medications

The National Health and Nutrition Examination Survey noted that the estimated prevalence of polypharmacy increased from 8% in 1999 to 15% in 2011.⁸ Many commonly used medications, such as decongestants, antihistamines, and anticholinergic agents, can worsen BPH symptoms,⁹ so it is reasonable to consistently review the patient’s medications to weigh the risks and benefits and determine which ones align with the patient’s personal care goals.

BPH: Take-home points

• Combination therapy with an alpha-blocker and a 5-alpha reductase inhibitor is an effective regimen for BPH.
• Polypharmacy is a significant problem in the elderly.
• Withdrawing the alpha-blocker component from BPH combination therapy can be considered after 1 year of combination therapy in patients whose symptoms have been well controlled.

A 60-year-old patient calls you after receiving his laboratory testing report from his insurance physical. His prostate-specific antigen (PSA) level is 5.1 ng/mL, and he has several questions:

• Should he have agreed to the screening?
• How effective is the screening?
• What are the next steps?

Is PSA screening useful?

Over the last few years, there has been great debate as to the utility of screening for prostate cancer.

The US Centers for Disease Control and Prevention¹⁰ reported that in 2014, an estimated 172,258 men in the United States were diagnosed with prostate cancer, but only 28,343 men died of it. These statistics support the notion that screening programs may be detecting what might otherwise be a silent disease.

The US Preventive Services Task Force (USPSTF)¹¹ recommends against blanket PSA screening, in view of the low probability that it reduces the risk of death from prostate cancer. For men ages 55 through 69, current guidelines give a grade C recommendation to PSA screening, meaning there is moderate agreement that the benefit is likely small, and screening should be selectively offered based on professional judgment and patient preference. In men ages 70 and older who are not at high risk, the guideline gives screening a grade D recommendation, meaning there is moderate evidence that there is no benefit from the practice. This is a change from the 2012 USPSTF guidelines,¹² which gave a grade D recommendation to PSA screening for all ages.

The American Urological Association¹³ recommends against PSA screening in men under age 40 or ages 70 and older. It does not recommend routine screening in those ages 40 to 54 at average risk, but it says the decision should be individualized in this age group in those at higher risk (eg, with a positive family history, African American). At ages 55 through 69, it recommends shared decision-making, taking into account cancer risk and life expectancy. In those who opt for screening, an interval of 2 years or more may be preferred over annual screening to reduce the risk of overdiagnosis.

The USPSTF recommendations rely heavily on data from 2 trials: the European Randomized Study of Screening for Prostate Cancer (ERSPC)¹⁴ and the Prostate, Lung, Colorectal, and Ovarian Screening (PLCO) trial.¹⁵

The ERSPC¹⁴ demonstrated that screening for prostate cancer reduced deaths from prostate cancer by 20%, with an absolute risk difference of 0.71 deaths per 1,000 men; 1,410 men would need to be screened and 48 additional cases of prostate cancer would need to be treated to prevent 1 death from prostate cancer. Screening also decreased the risk of developing metastatic disease by 30%.¹⁶ On the negative side, screening increased the risk of overdiagnosis and other harms such as bleeding, sepsis, and incontinence.

The PLCO trial,¹⁵ in contrast, found no difference in death rates between men randomly assigned to annual screening and those assigned to usual care. Differences between the trial results were thought to be due to different practice settings as well as study implementation and compliance.

Tsodikov et al¹⁷ reanalyzed data from the ERSPC and the PLCO trial using 3 different mathematical models to estimate the effects of screening in both trials compared with no screening. The analysis found no evidence that the effects of screening vs not screening differed between the 2 trials, ultimately concluding that PSA screening reduced prostate cancer deaths by 25% to 32%, which the authors inferred was primarily a result of earlier detection of cancer.

The Cluster Randomized Trial of PSA Testing for Prostate Cancer,¹⁸ published in March 2018, explored the effect of single PSA screening vs no screening on prostate cancer mortality rates in 419,582 men ages 50 through 69. Although screening detected more cases of low-risk prostate cancer, there was no significant difference in prostate cancer mortality rates after a median follow-up of 10 years. However, 10% to 15% of the control group was estimated to have also been screened, and these results do not directly speak to the efficacy of serial PSA screening.

Extended follow-up of this trial is planned to report on long-term survival benefits and whether screening lowers the risk of metastasis.

Imaging-guided prostate biopsy

Once a patient is found to have an elevated PSA level, standard practice has been to perform transrectal ultrasonography to obtain 12 core biopsy samples. The results indicate whether the prostate contains cancer, how aggressive the cancer is (Gleason score), and whether there is extracapsular extension.

In the past, magnetic resonance imaging (MRI) of the prostate before biopsy was thought to be too costly, and many insurance plans do not currently cover it.

Pahwa et al,¹⁹ however, in a cost-effectiveness study using a decision-analysis model, found that using MRI to detect lesions and then guide biopsy by triaging patients into proper treatment pathways added health benefits in a cost-effective manner in 94.05% of simulations. These benefits were found across all age groups.

This study demonstrated that doctors could use MRI to better evaluate patients for potentially harmful lesions. If a focus of cancer is found, it can be biopsied; if no cancer is seen on MRI, the patient can avoid biopsy completely. Additionally, though MRI tended to miss low-risk cancers, these cancers are thought to disproportionately lead to higher healthcare costs through unnecessary treatment. Therefore, a negative MRI study was believed to be an excellent sign that the patient does not have aggressive prostate cancer. This approach led to a net gain of 0.251 additional quality-adjusted life years compared with the standard biopsy strategy.

The Prostate MRI Imaging Study²⁰ also found MRI to be effective in the prostate cancer workup. In this trial, 576 men who had never undergone biopsy underwent multiparametric MRI, transrectal ultrasonography-guided biopsy, and the reference standard, ie, transperineal template prostate mapping biopsy. Of those who underwent biopsy, 71% received a diagnosis of prostate cancer, and 40% had clinically significant disease. In patients with clinically significant disease, MRI was more sensitive than ultrasonography-guided biopsy (93% vs 48%, P < .0001) but less specific (41% vs 96%, P < .0001).

Based on these findings, if biopsy were performed only in those who had suspicious lesions on MRI, 27% of men with elevated PSA could avoid biopsy and its potential complications such as bleeding and sepsis, which occurred in 5.9% of the biopsy group.

The Prostate Evaluation for Clinically Important Disease: Sampling Using Image Guidance or Not? trial²¹ more recently studied MRI with or without targeted biopsy vs standard transrectal ultrasonography-guided biopsy in 500 men who had not undergone biopsy before, and reported similar results. MRI with or without biopsy led to fewer biopsies and less overdetection of clinically insignificant prostate cancers compared with the standard approach. Furthermore, those in the MRI-targeted biopsy group were 13% less likely to receive a diagnosis of clinically insignificant cancer than those who received the standard biopsy (adjusted difference −13 percentage points, 95% confidence interval [CI] −19 to −7, P < .001).

Together, these data provide another argument for adding multiparametric MRI to the workup of men with an elevated PSA level.

Surveillance vs treatment for prostate cancer

Once prostate cancer is diagnosed, surveillance is becoming an increasingly common management strategy.

The Prostate Cancer Intervention Versus Observation Trial (PIVOT),²² one of the largest and longest trials involving cancer patients, offered further evidence that active surveillance and less intervention for men with prostate cancer is a better approach in many cases. This trial compared prostatectomy and observation alone in a randomized fashion. Inclusion for the study required men to be medically fit for radical prostatectomy, along with having histologically confirmed localized prostate cancer (stage T1-T2NxM0 in the tumor-node-metastasis classification system) of any grade diagnosed within the last 12 months.

During 19.5 years of follow-up, 223 (61.3%) of the 364 men randomly assigned to radical prostatectomy died, compared with 245 (66.8%) of 367 men in the observation group; the difference was not statistically different (P = .06). Only 9.4% of the deaths were due to prostate cancer, 7.4% in the surgery group and 11.4% in the observation group (P = .06).

Surgery was associated with a lower all-cause mortality rate than observation in the subgroup of patients with intermediate-risk prostate cancer (defined as PSA 10–20 ng/mL and a Gleason score of 7). Surgery was also associated with less disease progression.²²

This finding is in line with previous data from the Scandinavian Prostate Cancer Group Study Number 4,²³ as well as the much larger Prostate Testing for Cancer and Treatment (ProtecT) trial,²⁴ both of which reported that metastasis was 1.5 and 2.6 times as common, respectively, in participants in the active surveillance groups. However, in the PIVOT trial, those in the surgery group were significantly more likely than those in the observation group to have erectile dysfunction and urinary incontinence at 10 years.

Therefore, in men with localized disease and in those with low-risk PSA levels, both the PIVOT and ProtecT trials suggest that death from prostate cancer is uncommon and that observation may be more appropriate.

Prostate cancer: Take-home points

• A new look at 2 large trials of PSA screening strengthened evidence that testing in the right patient population can reduce deaths from prostate cancer, but a third recently published trial that found no benefit from 1-time screening may reopen debate on the topic.
• MRI offers a better method than ultrasonography-guided biopsy to triage patients thought to be at high risk of prostate cancer and tends to limit costly overtreatment of disease that likely would not cause death.
• Surgery for prostate cancer may not prolong life but could reduce disease progression, at the risk of more adverse effects.
• Shared decision-making should be practiced when deciding whether to use active surveillance or active treatment of diagnosed prostate cancer.

A 62-year-old man with hypertension, hyperlipidemia, peripheral artery disease, and type 2 diabetes presents for a 6-month follow-up. His medications include aspirin, metformin, lisinopril, and atorvastatin, all of which he takes without problems. Over the past several months, he has noticed that his erections are not adequate for sexual intercourse. He recently heard that a generic version of sildenafil has just become available, and he wonders if it might benefit him.

Erectile dysfunction is common, associated with chronic diseases

Erectile dysfunction, ie, persistent inability to obtain and maintain an erection sufficient to permit satisfactory sexual intercourse,^25,26 is estimated to affect nearly 20% of men over the age of 20 and 75% of men over the age of 75.²⁷

In age-adjusted models, erectile dysfunction has been shown²⁸ to be associated with:

• History of cardiovascular disease (odds ratio [OR] 1.63, 95% CI 1.02–2.63)
• Diabetes (OR 3.90, 95% CI 2.16–7.04)
• Treated hypertension vs no hypertension (OR 2.22, 95% CI 1.30–3.80)
• Current smoking vs never smoking (OR 1.63, 95% CI 1.01–2.62)
• BMI greater than 30 kg/m² vs less than 25 kg/m² (OR 1.80, 95% CI 1.03–3.14).

Because of the strong association between cardiovascular disease and erectile dysfunction, the presence of one often suggests the need to screen for the other.²⁹ While tools such as the International Index of Erectile Function (IIEF-5) have been developed to evaluate erectile dysfunction, it is most often diagnosed on the basis of clinical impression, while validated assessment methods are reserved for clinical trials.²⁸

Multiple causes of erectile dysfunction

Erectile dysfunction arises from inadequate penile tissue response to a sexual signal. The response can be disrupted at several points. For example, damage to vascular smooth muscle cells (eg, from age or obesity) and endothelial cells (from smoking or diabetes) and narrowing of the vascular lumen (from atherosclerosis or hypertension) have all been shown to impair engorgement of the corpus cavernosum.³⁰ In addition, denervation from prostate surgery or spinal trauma and psychogenic causes should be recognized in discussions with patients.

Drugs for erectile dysfunction

Pharmacologic management of erectile dysfunction includes oral, sublingual, intracavernosal, and intraurethral therapies.³¹ Treatment in primary care settings usually includes addressing underlying chronic diseases³² and prescribing phosphodiesterase-5 inhibitors (sildenafil, tadalafil, vardenafil, and avanafil). These drugs work by increasing local concentrations of cyclic guanosine monophosphate in the corpus cavernosum to induce vasodilation.³³

While these 4 drugs are still patent-protected, a manufacturer has been allowed to introduce a generic version of sildenafil into US markets, and a generic version of tadalafil is expected to be available soon.

Sildenafil, tadalafil, and vardenafil have been studied and found to have some degree of effectiveness in erectile dysfunction caused by damage to the penile vasculature, denervation, and spinal cord injury.³⁴ All drugs of this class have adverse effects including headache, facial flushing, and nasal congestion, but the drugs are generally well tolerated.³⁵

Sildenafil and tadalafil improve IIEF-5 scores by a similar margin, raising scores on the erectile domain subsection from approximately 14 of a possible 30 to approximately 24 of 30 in a trial of both drugs.³⁶ However, multiple crossover studies comparing the 2 drugs have shown that nearly 75% of patients prefer tadalafil to sildenafil,^36,37 perhaps because of tadalafil’s longer duration of action.³⁴

There is little evidence to suggest that vardenafil is more effective or more often preferred by patients than tadalafil or sidenafil.^34,38 And though data on the newest drug on the market, avanafil, are limited, a meta-analysis concluded that it may be less effective than tadalafil and without significant differences in terms of safety.³⁹

Other treatments

Lifestyle modifications, especially smoking cessation and exercise, have been shown to reduce the risk of erectile dysfunction with varying effect sizes across studies.^40–42 Moreover, factors such as obesity, alcohol use, and smoking may cause irreversible harm, and thus a healthy lifestyle should be encouraged.⁴¹

While there is only weak evidence for the use of psychological interventions alone for treating most types of erectile dysfunction, one meta-analysis found that the combination of psychological intervention and a phosphodiesterase-5 inhibitor improved sexual satisfaction more than drug therapy alone.⁴³

Erectile dysfunction: Take-home points

• Erectile dysfunction is common, affecting nearly 20% of men over the age of 20 and over 75% of men over the age of 75.
• Erectile dysfunction is often associated with chronic disease and may suggest the need to screen for cardiovascular disease.
• Treating underlying chronic diseases may help, and phosphodiesterase-5 inhibitors are effective; tadalafil may be most often preferred.

A 68-year-old man with a history of hypertension, BPH, and erectile dysfunction presents for a 6-month follow-up. His medication use includes lisinopril, which he takes without problems. He denies any new physical symptoms. His physical examination is unremarkable. He says he has heard about supplements that might help with his sexual performance and hopes to discuss recommendations during the visit.

A burgeoning, unregulated industry

Since the passage of the Dietary Supplement and Health Education Act in 1994, a law that decreased oversight of the supplement industry, spending on supplements has skyrocketed to over $41.1 billion each year.⁴⁴ Advertisements for these products typically claim that they improve general mental and physical health, sexual and romantic performance, leanness, and muscularity.⁴⁵ A national survey of men ages 57 and older reported that the most popular products were aimed at nutrition (such as multivitamins), cardiovascular health (such as omega-3 fatty acids), and chronic conditions (such as saw palmetto for BPH).⁴⁶

Little evidence of efficacy

There is little evidence to support the use of most supplements to improve men’s health. For example, a study in 82,405 men found no association between mortality rates and multivitamin use (hazard ratio [HR] 1.07, 95% CI 0.96–1.19).⁴⁷ Even for specific uses, such as cognitive performance, randomized trials exploring the effects of multivitamins in men have been largely negative.⁴⁸

The positive trials that have been reported are often of low quality and are funded by supplement manufacturers. For example, one of the few trials that reported a positive association between multivitamin supplementation and cognition in men was underpowered (N = 51) and found improvement in only 1 of 19 cognitive domains.⁴⁹ Despite the poor design and results to the contrary, this industry-funded study nevertheless concluded that multivitamins may play a role in improving elements of memory.

Evidence of possible harm from antioxidants

While not always specific to men, many meta-analyses have explored the effects of antioxidant supplements on cardiovascular and mortality risk. Most of them concluded that antioxidant supplements have no benefit and that some may actually be harmful.

For example, multiple meta-analyses of vitamin E supplementation found no cardiovascular benefit but possible increases in all-cause mortality rates in those taking high doses (risk ratio 1.04, 95% CI 1.01–1.07).^50,51

Another meta-analysis of 180,938 partici­pants in high-quality studies found an increased risk of all-cause mortality associated with independent intake of several antioxidant vitamins, including beta-carotene (risk ratio 1.07, 95% CI 1.02–1.11) and vitamin A (risk ratio 1.16, 95% CI 1.10–1.24), while intake of vitamin C and selenium had no impact on mortality.⁵²

Similarly, although nearly 10% of US adults report taking omega-3 fatty acid supplements, a review of 24 randomized controlled trials and meta-analyses published between 2005 and 2012 concluded that only 2 supported the use of these supplements for any health benefit.⁵³

Can supplements improve sexual function, prostate health?

To improve sexual function. A 2015 narrative review of the ingredients in General Nutrition Center’s top 30 best-selling products targeted at improving men’s sexual performance (including improving libido and erectile dysfunction) found only poor evidence for any efficacy.⁵⁴ The few studies that did support the use of select supplements, including B vitamins in people with diabetes, L-arginine, and yohimbine, were deemed to be of poor quality or showed a smaller effect size compared with standard medical therapy.

To prevent prostate cancer. Studies of supplement use to improve prostate health have had mixed results. For example, multiple large case-control studies have suggested that taking vitamin D^55,56 or vitamin C⁵⁷ is not associated with prostate cancer risk, while increased vitamin A^58,59 and E^60,61 intake is associated with inconsistent increases in prostate cancer risk.

In the Selenium and Vitamin E Cancer Prevention Trial,⁶² a randomized controlled trial in 35,533 men, those assigned to receive vitamin E supplementation were 17% more likely to get prostate cancer than were those assigned to placebo (HR 1.17, 99% CI 1.004–1.36, P = .008).

However, there are plausible biologic links between nutraceuticals and prostate cancer. For example, studies have linked genetic polymorphisms in vitamin D receptors⁶³ as well as intake of natural androgen receptor modulators, such as the most active polyphenol in green tea,⁶⁴ to prostate cancer risk and aggressiveness in certain populations. This led a recent review to conclude that there is some biologic plausibility, but at present little epidemiologic evidence, to support any dietary supplement’s ability to broadly affect prostate cancer risk.⁶⁵

Interest continues in exploring the targeted use of nutraceuticals as adjuvant therapy in specific populations at risk of prostate cancer.^66,67

To treat BPH. There is a similar dearth of clinical or population-based evidence that supplements can broadly affect BPH symptoms. For example, in a 2012 Cochrane review of Serenoa repens (saw palmetto) utilizing only high-quality evidence, there was no evidence that supplement use significantly reduced lower urinary tract symptoms, nocturia, or peak urine flow in BPH patients, and this was true even when the supplement was taken at triple-strength doses.⁶⁸

For other diseases. There is also limited evidence that supplements can affect other chronic diseases. For example, a meta-analysis of 3,803 patients found that glucosamine, chondroitin, and their combination had no impact on joint pain or joint space narrowing in patients with osteoarthritis of the knee or hip.⁶⁹

Even when there is some evidence to suggest benefit from supplementation, study heterogeneity and varying evidence quality limit confidence in the conclusions. For example, meta-analyses suggest garlic may improve blood pressure control in those with hypertension⁷⁰ and improve lipid and blood glucose control in type 2 diabetes.⁷¹ However, most of the trials included in those systematic reviews were underpowered, with samples as low as 10 patients, and many suffered from improper design, such as inadequate blinding of researchers. In addition, these meta-analyses often do not report adverse events, suggesting that higher quality studies would be needed to adequately measure event rates. As such, there is need for caution and a case-by-case review before recommending even a seemingly benign supplement like garlic to patients.

In total, there is only limited evidence to support the efficacy of supplements across many diseases and concerns common to men in primary care. This includes improving general health, cardiovascular health, sexual functioning, or other chronic diseases. While a supplement’s placebo effect may at times provide some benefit, supplements are much less strictly regulated since the passing of the 1994 act, and even vitamin supplementation has been shown to be associated with negative health outcomes. As such, a patient’s use of supplements requires careful consideration and shared decision-making.

Supplements: Take-home points

• Supplements are only loosely regulated by the federal government.
• There is some biologic but limited epidemiologic evidence for the use of multivitamins to improve cognition or mortality rates; for the use of antioxidant vitamins or omega-3 fatty acids to improve cardiovascular health; for the use of any of the top-selling sexual enhancement supplements to improve libido or erectile function; and for the use of vitamins or other supplements for improving BPH or reducing prostate cancer risk. Using supplements may in some cases be harmful.
• Given the heterogeneity of studies of supplements to manage chronic diseases and a lack of reporting of adverse events, careful consideration is needed when recommending supplements to patients.