Clinical Trials of Heart Failure: Is There a Question of Sex?

The syndrome of heart failure (HF) exhibits profound sex differences in clinical presentation, predisposing factors, untreated disease progression, and prognosis.1Lam CSP Arnott C Beale AL et al.Sex differences in heart failure.Eur Heart J. 2019; 40 (3859-68c)Crossref Scopus (111) Google Scholar, 2Dewan P Rørth R Raparelli V et al.Sex-related differences in heart failure with preserved ejection fraction.Circ Heart Fail. 2019; 12e006539Crossref PubMed Scopus (28) Google Scholar, 3Eisenberg E Di Palo KE Piña IL Sex differences in heart failure.Clin Cardiol. 2018; 41: 211-216Crossref PubMed Scopus (42) Google Scholar, 4Cannatà A Fabris E Merlo M et al.Sex differences in the long-term prognosis of dilated cardiomyopathy.Can J Cardiol. 2020; 36: 37-44Abstract Full Text Full Text PDF PubMed Scopus (21) Google Scholar Overall, a higher prevalence of hypertension, previous myocardial infarctions, obesity, and alcohol use predispose men to HF; whereas common risk factors among women include diabetes, uncontrolled hypertension, smoking, as well as sex-specific conditions such as peripartum cardiomyopathy and that induced by breast cancer treatment.1Lam CSP Arnott C Beale AL et al.Sex differences in heart failure.Eur Heart J. 2019; 40 (3859-68c)Crossref Scopus (111) Google Scholar,3Eisenberg E Di Palo KE Piña IL Sex differences in heart failure.Clin Cardiol. 2018; 41: 211-216Crossref PubMed Scopus (42) Google Scholar,5Lawson CA Zaccardi F Squire I et al.20-Year trends in cause-specific heart failure outcomes by sex, socioeconomic status, and place of diagnosis: a population-based study.Lancet Public Health. 2019; 4: e406-e420Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar Additionally, women and men differ according to type of HF; women classically experience HF with preserved ejection fraction, and HF with reduced ejection fraction (HFrEF) predominates among men.6Vasan RS Larson MG Benjamin EJ et al.Congestive heart failure in subjects with normal versus reduced left ventricular ejection fraction: prevalence and mortality in a population-based cohort.J Am Coll Cardiol. 1999; 33: 1948-1955Crossref PubMed Scopus (1171) Google Scholar,7Senni M Tribouilloy CM Rodeheffer RJ et al.Congestive heart failure in the community: a study of all incident cases in Olmsted county, Minnesota, in 1991.Circulation. 1998; 98: 2282-2289Crossref PubMed Scopus (929) Google Scholar Evidence for treatment of HF comes from large-scale clinical trials. Unfortunately, most of these trials are conducted among men with HFrEF. The recently published Prospective Comparison of ARNI With ARB Global Outcomes in HF With Preserved Ejection Fraction (PARAGON-HF) trial, however, is among the first to study HF with preserved ejection fraction and to include an equal proportion of men and women. In PARAGON-HF the efficacy of sacubitril–valsartan was compared vs valsartan for a composite of total hospitalizations for HF and death from cardiovascular causes. The trial showed that although treatment with sacubitril–valsartan did not lower the rate of the primary composite outcome in the overall population, there appeared to be a statistically significant benefit of the angiotensin receptor-neprilysin inhibitors among women, which was not seen in men. Because the results come from a subgroup analysis, this trial underscores an important question: should subgroups studying a sex effect be conducted and if so, how? In this commentary, we explore how population selection, reduced power associated with subgroup analyses, and inappropriate statistical techniques have limited the ability to make valid conclusions regarding sex-specific effects. Additionally, we provide recommendations for the conduct of future trials to improve our understanding of sex differences. Although these recommendations are discussed in the context of HF, they can be applied to randomized controlled trials (RCTs) across the field of cardiovascular medicine. Large-scale epidemiologic studies show a comparable lifetime risk of HF between the sexes, but higher overall survival among men. In the Framingham Heart Study it was estimated that 21% of men and 20% of women are at risk of developing HF at age 40 years.8Kenchaiah S Vasan RS Heart failure in women – insights from the Framingham Heart Study.Cardiovasc Drugs Ther. 2015; 29: 377-390Crossref PubMed Scopus (34) Google Scholar On the basis of data from 56,000 primary care patients between 2000 and 2017 in the United Kingdom, overall survival was 75.9% at 1 year and 12.7% at 15 years. Women had worse survival compared with men (1-year survival 74.5% vs 77.2% and 15-year survival 11.0% vs 14.1%).9Taylor CJ Ordóñez-Mena JM Roalfe AK et al.Trends in survival after a diagnosis of heart failure in the United Kingdom 2000-2017: population based cohort study.BMJ. 2019; 364: 223Crossref PubMed Scopus (117) Google Scholar Similar trends have been observed in the Rotterdam and Olmsted County studies.7Senni M Tribouilloy CM Rodeheffer RJ et al.Congestive heart failure in the community: a study of all incident cases in Olmsted county, Minnesota, in 1991.Circulation. 1998; 98: 2282-2289Crossref PubMed Scopus (929) Google Scholar,10Bleumink GS Knetsch AM Sturkenboom MCJM et al.Quantifying the heart failure epidemic: prevalence, incidence rate, lifetime risk and prognosis of heart failure - The Rotterdam Study.Eur Heart J. 2004; 25: 1614-1619Crossref PubMed Scopus (767) Google Scholar Women also experience greater exercise limitations compared with men; in fact, New York Heart Association class IV is an important predictor of mortality in acute decompensated HF among women with HFrEF, but not men.11Cai S Gong IY Gale CP Yan AT Sex-specific differences in New York Heart Association classification and outcomes of decompensated heart failure.Can J Cardiol. 2020; 36: 4-6Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar,12Kajimoto K Sato N Sex differences in New York Heart Association functional classification and survival in acute heart failure patients with preserved or reduced ejection fraction.Can J Cardiol. 2020; 36: 30-36Abstract Full Text Full Text PDF PubMed Scopus (9) Google Scholar Additionally, the incidence of adverse drug effects among women is 1.5‐ to 1.7‐fold higher in than men; hospitalizations due to adverse drug effects occur more frequently in women.3Eisenberg E Di Palo KE Piña IL Sex differences in heart failure.Clin Cardiol. 2018; 41: 211-216Crossref PubMed Scopus (42) Google Scholar These differences might exist in part because current guidelines for HF treatment are on the basis of clinical trials in which women are largely under‐represented, comprising an average of 30% of the trial population. Variation in clinical progression of disease and outcomes between the sexes in conjunction with under-representation of women in HF that define management of this syndrome underscores the importance of sex-specific subgroup analyses. To this end, there is perhaps a role for trials conducted among women only to answer questions specific to the efficacy of therapeutics for this sex; existing examples of these include the Women's Health Initiative and Women's Health Study.13Liu KA Dipietro Mager NA Women's involvement in clinical trials: historical perspective and future implications.Pharm Pract (Granada). 2016; 14: 708Crossref PubMed Scopus (94) Google Scholar A review of 325 clinical trials in cardiovascular medicine between 1997 and 2009 estimated that approximately 30% of study participants were women.14Tsang W Alter DA Wijeysundera HC Zhang T Ko DT The impact of cardiovascular disease prevalence on women's enrollment in landmark randomized cardiovascular trials: a systematic review.J Gen Intern Med. 2012; 27: 93-98Crossref PubMed Scopus (30) Google Scholar In a more comprehensive review of 740 trials registered on ClinicalTrials.gov between 2010 and 2017, of which 102 were conducted among patients with HF, a global participation rate for women of 28.6% was reported.15Jin X Chandramouli C Allocco B et al.Women's participation in cardiovascular clinical trials from 2010 to 2017.Circulation. 2020; 141: 540-548Crossref PubMed Scopus (48) Google Scholar Women comprise almost 57% of the HF population.16Popovic JR Kozak LJ National Hospital Discharge Survey: annual summary, 1998.Vital Health Stat 13. 2000; 148: 1-194PubMed Google Scholar Taking the population prevalence of diseases into consideration, women were represented at a rate lower than their share in the respective disease population for HF (participation prevalence ratio = 0.48; a ratio of 0.8-1.2 would suggest good representation).15Jin X Chandramouli C Allocco B et al.Women's participation in cardiovascular clinical trials from 2010 to 2017.Circulation. 2020; 141: 540-548Crossref PubMed Scopus (48) Google Scholar Although women are not specifically excluded from trials, the lower rates of inclusion beg the question, why women do not participate in trials with equal frequency? Extensive literature reviews on the topic highlight greater distrust of the medical system, poor communication, precarious employment, among others as possible reasons.17Myles S Tocci C Falk M et al.A multicenter investigation of factors influencing women's participation in clinical trials.J Womens Health. 2018; 27: 258-270Crossref PubMed Scopus (9) Google Scholar, 18Walsh E Sheridan A Factors affecting patient participation in clinical trials in Ireland: a narrative review.Contemp Clin Trials Commun. 2016; 3: 23-31Crossref PubMed Scopus (38) Google Scholar, 19Coakley M Fadiran EO Parrish LJ et al.Dialogues on diversifying clinical trials: successful strategies for engaging women and minorities in clinical trials.J Womens Health. 2012; 21: 713-716Crossref PubMed Scopus (81) Google Scholar, 20Ding EL Powe NR Manson JAE Sherber NS Braunstein JB Sex differences in perceived risks, distrust, and willingness to participate in clinical trials: a randomized study of cardiovascular prevention trials.Arch Intern Med. 2007; 167: 905-912Crossref PubMed Scopus (89) Google Scholar Ding et al. conducted a randomized study of patient willingness to participate in cardiovascular prevention trials and reported that men had 15% greater willingness to participate than women.20Ding EL Powe NR Manson JAE Sherber NS Braunstein JB Sex differences in perceived risks, distrust, and willingness to participate in clinical trials: a randomized study of cardiovascular prevention trials.Arch Intern Med. 2007; 167: 905-912Crossref PubMed Scopus (89) Google Scholar The authors reported lower rates of participation and trust in research among women with precarious employment and lower income. Jin and colleagues surveyed the literature comprising 740 trials including 862,652 patients with cardiovascular disease15Jin X Chandramouli C Allocco B et al.Women's participation in cardiovascular clinical trials from 2010 to 2017.Circulation. 2020; 141: 540-548Crossref PubMed Scopus (48) Google Scholar and reported that participation rate of women was higher among trials conducted in North America and lowest in the Western Pacific region; higher in trials sponsored by private industry compared with government-sponsored trials, and particularly low in trials in which the average participant age was between 61 and 65 years. These findings might suggest a role of intention and knowledge on the part of the trialists—trialists of large North American private institutions might be specifically focused on recruiting a balanced, generalizable cohort. Boston Scientific Corporation's Women Opt-In for Heart Research (WIN-Her) initiative included surveys and interviews of women with cardiovascular disease to understand their perceptions of trials. Their work identified minimal understanding of the trial process and logistics, limited clinical trial information from physicians, and misperceptions around the and of participation as to X Chandramouli C Allocco B et al.Women's participation in cardiovascular clinical trials from 2010 to 2017.Circulation. 2020; 141: 540-548Crossref PubMed Scopus (48) Google Scholar of the data use of among women come from and of E Di Palo KE Piña IL Sex differences in heart failure.Clin Cardiol. 2018; 41: 211-216Crossref PubMed Scopus (42) Google Scholar clinical trials of angiotensin inhibitors and angiotensin showed a benefit only among men. and studies suggest a comparable mortality benefit among women with HFrEF. subgroup of in the Study and Heart and Study trials showed a mortality benefit among women as E Di Palo KE Piña IL Sex differences in heart failure.Clin Cardiol. 2018; 41: 211-216Crossref PubMed Scopus (42) Google Scholar mortality data according to sex from the in Heart Study and Prospective trials showed and significant survival in women and men. conclusions are to make because and of trials are by and are M et of women, and and minorities in heart failure clinical Cardiol. 2018; 3: PubMed Scopus Google Scholar In this conducted primary subgroup can provide data to sex differences in treatment and a for of for women with E Di Palo KE Piña IL Sex differences in heart failure.Clin Cardiol. 2018; 41: 211-216Crossref PubMed Scopus (42) Google Scholar Sex-specific effects of a treatment are in A large proportion of landmark HF trials have conducted subgroup according to F et in patients with heart failure and J Med. 364: PubMed Scopus Google Scholar, JB et and with in patients with heart failure and left ventricular results of the ventricular ejection fraction 2004; PubMed Scopus Google Scholar, M et of on survival in heart J Med. PubMed Scopus Google Scholar, et in heart failure with preserved ejection J Med. 2019; PubMed Scopus Google Scholar, M et versus in heart J Med. PubMed Scopus Google Scholar, B et for heart failure with preserved ejection J Med. PubMed Scopus Google Scholar, K M M et and outcomes in heart failure a Full Text Full Text PDF PubMed Scopus Google Scholar, et in patients with heart failure and preserved ejection J Med. PubMed Scopus Google Scholar of subgroup can significant because findings can inappropriate or of in subgroup from trials do not statistically significant to In a systematic review and colleagues and reported only statistically significant of which treatment Sex based subgroup differences in randomized controlled trials: from 2016; PubMed Scopus Google Scholar review of published between and in showed that of the statistically significant subgroup but only were to a only sex as a E in cardiovascular clinical trials: are most of Heart J. PubMed Scopus Google Scholar The of effects suggest that exist or underscores conduct of subgroup analyses. To answer a effect and to the effect well conducted and subgroup are In the we common among subgroup in this HF trials are limited by low prevalence of women, as and that of patients from inclusion the to patients seen in the an of HF that compared trial participants with in the patients in trials were more to have ejection fraction, and were more of patients were compared with estimated in the A CP of the women, and minorities in heart failure clinical trials.Arch Intern Med. PubMed Scopus Google Scholar the average age of trial participants was compared with a age of at first diagnosis of HF on the basis of the M Tribouilloy CM Rodeheffer RJ et al.Congestive heart failure in the community: a study of all incident cases in Olmsted county, Minnesota, in 1991.Circulation. 1998; 98: 2282-2289Crossref PubMed Scopus (929) Google Scholar of the trials included patients than years. of more to be women, was part of the study in of the trials, which for greater than A CP of the women, and minorities in heart failure clinical trials.Arch Intern Med. PubMed Scopus Google Scholar of participants with was also a common such as that are more common among T et al.Sex differences in outcomes of heart failure in an population-based cohort from 2009 to 2018; PubMed Google Scholar, N A T et al.A review of clinical guidelines that they were based on of to primary care Clin Full Text Full Text PDF PubMed Scopus Google Scholar, J P et patients in heart failure trials of primary care A systematic 2018; PubMed Scopus Google Scholar A to under-representation of women in clinical trials is whereas this would the in the it would not systematic by that women are not of the HF population. A common to subgroups is to the cohort is into and a treatment effect is to from more findings compared with of and S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google Scholar conducted to significant subgroups were compared with of data with subgroup but an overall significant treatment to of subgroups had at 1 that was statistically is with data with subgroup but an overall treatment only of subgroups had 1 significant is compared with of which had a rate of S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google Scholar Although the of subgroup is to the treatment effect across sex and of a the to be conducted with power to Because subgroups the overall trial population into the power to a treatment effect of the as seen in the overall trial is With the in there is a in the risk of a type or S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google Scholar The more is a of because this the between treatment effects across subgroups of the of S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google Scholar, CJ Scopus Google Scholar, RJ for subgroup in clinical PubMed Scopus Google Scholar In or In a trial at a subgroup that of the approximately power a rate power can only be if the type I rate is to which is With more subgroup power can be as low as 30% for a type I rate of because effect are to statistical in findings more frequently in a subgroup because of uncontrolled from of or from due to compared with Although of are for power compared with S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google E et in controlled trials: the of and PubMed Scopus Google Scholar the of power A trial classically at only be to of significant results with an effect of the S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google Scholar the of the effect was the overall effect or the had at the power as that for the overall treatment subgroup the power for an effect of the as the overall the power for the is reduced from approximately for a subgroup ratio of to approximately for a for and for S E M et in randomized trials: of power and for the Clin 2004; Full Text Full Text PDF PubMed Scopus Google Scholar is to our to that because trials are not to study specific subgroups and are for subgroups analyses, findings from these should only be and significant might provide a that the effect of treatment is the in subgroups in most of is due to a type of subgroups conducted in trials might provide a to the of however, can be limited by and of outcomes from the trials. data for from trials, it should be that might be at among the The for large-scale randomized on trials, or subgroup PubMed Scopus Google Scholar sex data are not reported in trials, which is common to of the treatment effect across a the more questions of a data the greater that at 1 of the subgroups an statistically is and by the landmark Study of trial of or among cases of acute myocardial Study of Google Scholar in which the authors benefit of in heart according to Their findings showed an overall benefit of but reported to or studying sex-specific it should be that a higher type I rate is if subgroups are also by the In these for can the findings of our sex-specific are to is not a common of among subgroup of sex might occur to in the of effect but it is to the or of subgroup effects. Additionally, there is systematic of an between data and sex; however, if authors are to that data in their subgroup are greater in proportion for sex compared with the study might benefit from of JB et for data in and clinical and PubMed Scopus Google Scholar, RJ R et prevention and treatment of data in clinical J Med. 2012; PubMed Scopus Google Scholar, J M data in clinical and Clin 2012; PubMed Scopus (94) Google Scholar is according to sex; of might occur with subgroup M K M et on subgroup in clinical 2015; Scopus (34) Google Scholar according to sex might in differences in exist according to Additionally, subgroup analyses, for should be represented in to a understanding of the effect of if findings of statistically significant In with by the National of of Health. Sex and Scholar and the of Health of Health the for and in clinical Scholar we of and outcome data as in C sex, or in clinical 2016; PubMed Scopus Google Scholar Although this been et al. of cardiovascular from 2010 and from 2017 and reported that sex-specific efficacy were reported in of the trials in 2010 and in The trials with to only in 2010 and in 2017 of according to E M et of sex-specific outcomes in trials of for cardiovascular there been Full Text Full Text PDF PubMed Scopus Google CM C et of sex-specific information into clinical guidelines for the management of myocardial and Am Heart 2019; PubMed Scopus Google Scholar data should be in the of the the results should be reported at if not in the significant or that have been conducted should be discussed in the all should have the of and the treatment effects by sex reported in the Similar to recommendations in the previous poor to also been and colleagues the and of subgroup in the New of from R JM in of subgroup in clinical J Med. 2007; PubMed Scopus Google Scholar reported that of the trials subgroups were reported for only the for treatment effects in but reported on results of The authors also that findings of the subgroup were in the for trials and that subgroup were reported in the and a or for trials were reported to have been to the of treatment effects for all subgroup in only trials and they were reported to be for but not subgroup in trials have E in cardiovascular clinical trials: are most of Heart J. PubMed Scopus Google Scholar to be more common in large than in can be for and of subgroups with a of the treatment and a however, these should not be to a treatment in a subgroup is statistically significant on the basis of the the on the basis of of treatment from the as the use of statistical The of subgroup J Med. PubMed Scopus Google Scholar they should be as a of treatment differences with the trial Additionally, such as for of subgroup and subgroup F M T A review of for subgroup in clinical 2020; PubMed Scopus Google Scholar should be in the clinical and rates of and mortality from HF among men and women have the for a sex-specific in the of guidelines the field of HF and cardiovascular C CM Sex-specific in guidelines and J Cardiol. 2019; Full Text Full Text PDF PubMed Scopus Google Scholar In our we have discussed the current literature on subgroup and recommendations for their These are in on the conduct of sex-specific subgroup in HF clinical the women to for at of patients women for to of the HF population in the to include women with a higher are more to and have a higher of prevalence of diabetes, hypertension, of HF between the the population, should be of for of uncontrolled more A A B I in findings and clinical 2012; 3: PubMed Scopus (34) Google of patients with also more women compared with men because is greater in associated with and J PubMed Scopus Google B B S for in J PubMed Scopus Google of with peripartum or use of only that do not include to women might not to participation because of of distrust of the medical system, and as well as more than the subgroups to be a according to sex to between men and and overall trial at to for greater power for subgroup of subgroup for P for study is a subgroup should be limited to of of should be included in the of trial data might be in power to answer particularly because trials have of R K studies for sex and how research can knowledge for women's Health 2019; 29: Full Text Full Text PDF PubMed Scopus Google angiotensin of HF, heart left ventricular ejection in a angiotensin of HF, heart left ventricular ejection a