U.S. Food and Drug Administration • Center for Drug Evaluation
and Research |
U.S. Food and Drug Administration • Center for Drug Evaluation
and Research |
 |
About CDER |
 |
Summary Context: There continues to be a perception that women are under-represented in clinical trials of new drugs. Few data are available from formal evaluations of approved new drug applications. This study reviews women's participation in clinical trials and gender-related labeling for new molecular entities approved during a five-year period by the Food and Drug Administration's (FDA) Center for Drug Evaluation and Research (CDER). Objectives: The objectives of the study were to assess to what extent women participated in clinical trials, to describe what FDA medical officers concluded about gender differences in sponsors' evaluations, and to determine to what extent sponsors described gender-related information in the approved product labeling. Design: This was a retrospective review of clinical trial protocols and labeling for 185 new molecular entities approved by CDER between January 1, 1995, and December 31, 1999. Enrollment data were derived from medical officers' reviews and tabulated according to gender. The approved product labeling was searched for statements related to product use in humans. All data were compiled and analyzed using Microsoft Access. Main Outcome Measures: This study quantifies women's participation as described by enrollment data in clinical trials by year, phase of study, and product type as defined by the CDER division responsible for the product review. Additionally, the study categorizes labeling based on the inclusion of gender-specific information. Results: Overall, women appear to participate in the clinical trials at nearly the same rate as men even when gender-specific products are excluded. Some differences in participation are seen when year-to-year or division-to-division comparisons are made. Labeling for two-thirds of the products contained some statement about gender, although only 22% described actual gender effects. Four labels described more than one gender effect. Ninety percent (90%) of the effects discussed in the labeling were pharmacokinetic, 12% were safety, and 5% were efficacy. No product recommended a change in dosage for women. Conclusions: Women are participating in clinical trials of new drugs in approximate proportion to their representation in the population. The majority of product labeling contains references to gender evaluation. Most of the gender effects were pharmacokinetic. Few products demonstrated safety or efficacy effects, and none recommended changes in dosage based on gender effects. Background Attention to potential gender differences is part of a larger effort by the FDA to ensure that the safety and efficacy of drugs are adequately studied in people who represent the full range of patients who will receive them upon marketing. FDA's 1977 "General Considerations for the Clinical Evaluation of Drugs" restricted participation by most women of childbearing potential in phase 1 and early phase 2 trials for non-serious diseases. More recent guidelines and regulations encourage the participation of women in all phases of drug development, promote collection of gender-related data during research and development, and recommend analysis of the data for gender effects. Following is a brief summary: · The 1988 "Guideline for the Format and Content of the Clinical and Statistical Sections of New Drug Applications" emphasizes the importance of including analysis of demographic subset data in new drug applications. · The 1993 "Guideline for the Study and Evaluation of Gender Differences in the Clinical Evaluation of Drugs" was developed amidst concerns that the drug development process may not provide adequate information about the effects of drugs in women. It called for attention to possible pharmacokinetic differences and for analyses of data to see whether men and women responded differently. Accompanying the guideline was a notice revoking the 1977 "General Considerations" restriction, reflecting the general consensus that women should be allowed to determine for themselves the appropriateness of participating in early clinical trials. The guidance made clear the agency position that women should be included in all phases of drug development. · The 1998 "Final Rule on Investigational New Drug Applications and New Drug Applications (Demographic Rule) requires that analyses of effectiveness and safety data for important demographic subgroups, including gender, be submitted to NDAs and that enrollment of subjects into clinical studies for drug and biological products be tabulated by important demographic subgroups in investigational new drug (IND) annual reports. · The 1999 Final Rule "Investigational New Drug Applications: Amendment to Clinical Hold Regulations for Products Intended for Life-Threatening Diseases" allows FDA to place an IND, or specific protocols under an IND, on clinical hold if a sponsor proposes to exclude from a study of a drug intended for a life-threatening disease women or men with reproductive potential because of a risk, or a potential risk, of reproductive or developmental toxicity from the use of an investigational drug product. Draft guidance documents on the adverse reactions section of the labeling and population pharmacokinetics also consider gender specific issues as summarized below. · The Guidance for Industry "Content and Format of the Adverse Reactions Section of Labeling for Human Prescription Drugs and Biologics" is intended to assist sponsors in developing the adverse reactions section of labeling for human prescription drugs and biologics. It includes specific recommendations for presenting subgroup (e.g. gender) specific information. · The Guidance for Industry "Population Pharmacokinetics" makes recommendations on the use of population pharmacokinetics in the drug development process to help identify differences in drug safety and efficacy among population subgroups. The Food and Drug Modernization Act (FDAMA) of 1997 prompted FDA to examine issues related to the inclusion of women in clinical trials. Section 115 of the Act required the Secretary, "in consultation with the Director of the National Institutes of Health and the representatives of the drug manufacturing industry, review and develop guidance, as appropriate, on the inclusion of women and minorities in clinical trials..." FDA established the FDAMA Women and Minorities Working Group to review and implement this section of FDAMA. In a report issued on July 20, 1998, the Working Group concluded that no additional guidance was needed. The report stated that the agency would implement procedures to enhance its ability to gather and evaluate demographic data and then decide whether additional guidance should be developed in the future. FDA's Office of Special Health Issues (OSHI) initiated two projects to gather and evaluate demographic data. The results of the first study were presented at the 2001 FDA Science Forum in a poster presentation entitled "Race, Age, and Gender: A Review of Demographic Subgroups in Clinical Trials of FDA-Regulated Drugs and Biologics." The study examined the extent to which IND sponsors complied with the reporting requirements of the 1998 Demographic Rule and found that required information was not submitted for 85% of IND protocols with data reported. This report presents the results of the second OSHI study "Women's Participation in Clinical Trials and Gender-Related Labeling: A Review of New Molecular Entities Approved 1995-1999." Objectives The objectives of this study were to assess the following: 1. To what extent women participated in clinical trials of drug products approved between 1995 and 1999, 2. What FDA medical officers concluded about gender differences in sponsors' evaluations, and 3. To what extent sponsors described gender-related information in the approved product labeling. Methods Data Sources and Collection Data for this study were obtained from two primary sources. Information about clinical trial enrollment data was obtained from FDA medical officers' reviews of 185 new molecular entities approved by CDER between 1995 and 1999. The reviews contained descriptions of 2,581 clinical trial protocols. The reviews were searched to code FDA comments about the gender of study participants and conclusions about the sponsors' gender analyses. Data were recorded on a protocol extraction sheet. Study participant data were tabulated according to male, female, or gender unspecified. Tabulations were quantified according to year of product approval, review division, and to the extent possible, phase of study. To capture comments regarding the sponsors' gender analyses, medical officers' statements were categorized into one of the following categories: gender differences were found, no differences were found, analysis is not applicable, sponsors' studies were inadequate to determine differences, or the medical officers' review had no information on the sponsors' analysis or lack of analysis. Information about the sponsors' findings of gender differences was obtained from the labeling at the time of product approval. The original package inserts were reviewed for references to human data on gender. Excluded from the labeling reviews were any statements about teratogenicity from animal studies or common pregnancy statements. Exact wording from the label was captured. Products with gender differences were reviewed to determine the type and significance of the difference and whether the difference required a change in dosage for women. All data extracted were captured on a protocol extraction form and entered into a Microsoft Access database. An initial audit of each data field for 250 protocols assured the quality of the extracted data. A second audit was conducted for each data field for 100% of the protocols once the data were entered into the database. Analysis Gender data were analyzed from clinical trial protocols described in 185 NMEs. Excluded from some of the analyses were 14 products identified as gender-specific. A product was considered gender-specific if it met either of the following two criteria: 1. The labeling indication stated that the product was exclusively for one gender (n=8) or 2. The condition that the drug is intended to treat occurs almost exclusively in one gender (n=6). As a result, 32,333 individuals were excluded from some data presentations: 12,647 males, 19,335 females, and 351 individuals whose gender could not be determined. The products excluded are listed in Table 1. Table 1. Gender-Specific Products Notes used in Table 1:
Female-specific products
Results Participation of Women Overall Participation A total of 493,600 individuals were described in the medical officers' reviews as being enrolled in the 2,581 clinical trials for all products examined in this study. Gender could be determined from the medical officers' reviews for almost three-fourths (74%) of the participants (Fig. 1). Thirty six percent were female and 38% were male. Gender could not be determined for the remaining 26%. In a few instances, medical officers described gender by the percentage of males only or females only. When the total enrollments are adjusted, representation by both genders remained the same. When evaluating only those participants for whom gender could be determined, overall the numbers of males and females participating were nearly equal at 51% and 49%, respectively. When the gender-specific products such as those for ovarian or prostate cancer listed in Table 1 are excluded from the analysis, participation by women (48%) is slightly less than that of men (52%) (Fig. 2). Figure 1 Figure 2
Participation by Year The percentage of women participating in the trials varied from year to year. Where gender could be determined from the total enrolled, women appear to be represented least in 1995 (20%), the greatest in 1996 (45%), and approximately 40% in each year from 1997 to 1999 (Fig. 3). It must be noted, however, that in 1995 gender could not be determined for as many as 52% of the participants. When evaluating only those participants for whom gender could be determined, and eliminating those enrolled in gender-specific products, women ranged from 42% in 1995 to 55% in 1998 (Fig 4). Women comprised 51%, 44%, and 44% of the participants in 1996, 1997, and 1999, respectively. Figure 3. Figure 4.
*excludes 159,049 (32%) in gender-specific products and whose gender is unspecified. Participation by Phase Of the 2,581 protocols reviewed, 72% of them did not have a phase specified. Where phase was specified, the majority were described in the medical officers' reviews as phase 3 (8%), followed by phase 1 and 2 at 7% each. A few others were described as phase 1/2 and phase 2/3 (Fig. 5) & (Table 2). Figure 5. Table 2.
Table 2. Total Protocols by Phase
Where gender could be determined, women participated to a greater extent in later phases of study than earlier phases (Figs. 6a & 6b). Figure 6a.
Figure 6b. *excludes 159,049 (32%) in gender-specific products and whose gender is unspecified
Participation by Product Type Product types were defined according to the division within the agency responsible for the product review. The categories of products for each division are listed in Table 3. Overall, women appeared to have participated in lesser proportions in trials for cardio-renal, anesthetic, and antiviral products and in greater proportions for neuropharmacological, gastrointestinal, anti-infectives, analgesics, and reproductive and urologic products (Fig. 7a). No statistical analyses were done. In many instances in which women appear to be underrepresented, there are substantial numbers of individuals whose gender could not be determined. For example, gender could not be determined for as many as 80% of those enrolled in the trials for the anesthetic products and 65% of those in trials for the antiviral products. In comparing only those whose gender could be determined, women's participation ranged from 32% for antiviral products to 83% for reproductive and urologic products (Fig. 7b). Women comprised more than half of the participants in six areas: neuropharmacology, anesthetic, gastrointestinal and coagulation, anti-infective, analgesic, and reproductive/urologic drug products. Women represented about half of the participants in three areas: metabolic/endocrine, pulmonary/allergy, and special pathogens. In four areas, women comprised less than half of the participants: cardio-renal, oncology, medical imaging, and antiviral drug products. Table 3. FDA CDER Product Review Divisions
Figure 7a.
Figure 7b.
*excludes 159,049 (32%) participants in gender-specific products and whose gender is unspecified. Medical Officers' Findings FDA medical officers' evaluations of the products were reviewed to code their comments about whether the sponsors' studies revealed gender differences. In 32/185 (17%) of the total reviews, the medical officer commented that gender differences were noted. In 78/185 (42%) of the reviews, the medical officer indicated that no gender differences were found. Gender analysis was described by medical officers as not applicable for 13/185 (7%) of the products. In 10/185 (5%) instances, medical officers noted the studies were inadequate to determine differences. In another 10/185 (5%) medical officers commented that the sponsors did not conduct a gender analysis. The medical officers' reviews made no comment about gender analysis for 42/185 (23%) of the products (Fig. 8). Interviews of medical officers were not conducted to determine reasons for why they did not make comments. Figure 8.
Of the 32 products described by the medical officers as demonstrating gender differences, the cardio-renal and metabolic/endocrine products occurred most frequently, each representing 6/32 (19%) of all products described as showing gender differences. This was followed by neuropharmacological 5/32 (16%), antiviral products 5/32 (16%), oncologic products 3/32 (9%), and special pathogens 3/32 (9%) (Fig. 9). Figure 9.
In 26 (81%) of the 32 products identified by the medical officers as having gender differences, the product labeling statements were consistent with the medical officers' findings. In the remaining six (19%), there were no statements in the labeling describing gender differences. In these six instances, no copies of memos, supervisor reviews, or meetings records were reviewed to determine whether the differences were significant or if a decision was made not to include gender differences in the label. Label Review Findings Review of the 185 product labels revealed that 125/185 (68%) contained some type of statement related to gender. Fifty of those labels (27%) indicated that there were no differences between the genders. Fifteen (8%) indicated that there were similar effects of the drug in men and women, and 41 (22%) indicated that there were actual gender differences. Nineteen (10%) of the labels stated either no studies were performed, studies were inadequate, retrospective review showed no differences, or the product was not indicated in a specific gender. The remaining 60 (32%) had no statements about gender. A summary of the labels with gender statements is listed in Table 4. Table 4. Label Review Findings * Rounded to the nearest whole number. **An additional nine products are designated as gender-specific products based on indications described in the product label.
When reviewing differences found in the label by the product type, the greatest proportion of gender-based differences occurred in special pathogens (7/8 or 88%), neuropharmacology (9/21or 43%), and metabolic and endocrine (7/25 or 28%) and cardio-renal (6/22 or 27%). There were no differences noted among anesthetic, gastrointestinal, dermatological and dental, pulmonary and allergy, or reproductive and urologic drug products (See Table 5). Table 5. Frequency of Gender Differences by Product Type
Of the 41 products for which labeling described gender effects, 90% (37/41) were pharmacokinetic, 12% (5/41) were safety, and 5% (2/41) were efficacy. These represent 20%, 3%, and 1% of the total product labels reviewed. Four product labels described more than one effect related to gender. No products required a change in the dosage based on gender differences. Examples of Labeling Statements Listed below are examples of gender-related labeling statements No differences between genders ex: "Patient race and gender did not influence the blood pressure lowering effect of SULAR." ex: "The pharmacokinetics of docetaxel were not influenced by age or gender and docetaxel total body clearance was not modified by pretreatment with dexamethasone." Differences described ex: "Plasma concentrations of atorvastatin in women differ from those in men (approximately 20% higher for Cmax and 10% lower for AUC); however, there is no clinically significant difference in LDL-C reduction with LIPITOR between men and women." ex: " Female patients constituted 32% of the patients in the placebo-controlled trials of TIKOSYN. ...TIKOSYN was associated with a greater risk of torsade de points in female patients than in male patients. ...the risk of torsade de points in females was approx. 3 times the risk in males. ...the incidence of other ventricular arrhythmias was similar in female patients receiving TIKOSYN and patients receiving placebo... in post-hoc analyses, no increased mortality was observed in females on TIKOSYN compared to females on placebo." Similar responses ex: "Analysis of age, gender, and race subgroups of patients showed that men and women, and patients over and under 65, had generally similar responses." (drug: losartan) ex: "Pharmacokinetics of indinavir appear to be comparable in men and women based on pharmacokinetic studies including 32 women (15 HIV positive)." No studies performed: ex: "The effects of age, gender, and race on cidofovir pharmacokinetics have not been investigated." Studies inadequate/insufficient to determine differences ex: "There has not been sufficient pharmacokinetic study to make a definitive statement as to whether the pharmacokinetics of nalmefene differs between the genders." No studies to assess, but retrospective review showed no differences ex: "No specific pharmacokinetic study was conducted to investigate the effects of gender and race on the disposition of ARICEPT. However, retrospective pharmacokinetic analysis indicates that gender and race (Japanese and Caucasian) did not affect the clearance of ARICEPT." ex: "No specific pharmacokinetic study was conducted to investigate gender effects. Retrospective analysis of pharmacokinetic data, however, following single and multiple dose administration of 4 mg tizanidine showed that gender had no effect on the pharmacokinetics of tizanidine." Product not indicated for women ex: "Because of the mechanism of action and the indication, bicalutamide has not been studied in women or pediatric subjects." ex: "FLOMAX is used by men only. It is not indicated for use in women." Not evaluated in men ex: "Safety and effectiveness in male osteoporosis have not been established." (drug: FOSAMAX) Discussion Women's Participation Overall, during the five-year period examined, women participated in clinical trials at nearly the same rate as men. In fact, in 1996 and 1998, women comprised more than half of the participants whose gender could be determined. No comparisons were done to determine the types and proportions of products approved that might provide insight into year-to-year differences. No comparison was done, for example, to determine whether more neuropharmacological products (where women are commonly enrolled) were approved in years where women were present in larger numbers. Women appear to be participating in all phases of study development. Women participated to a greater extent in later phases of study than earlier phases. This study does not distinguish data from trials for products intended exclusively for pediatric or adult use. However, among participants whose age could be determined, fewer than 1% (.0088) were 17 years old or younger. Therefore, it is unlikely that the results described in this study are affected by the inclusion of this age group. Medical Officers' Reviews Medical officer reviews were not written in a standard format and contained varying amounts of information related to gender. The information was not always presented in a way that was readily transferable to this study. In some reviews demographics were clearly presented and easy to locate, and discussions of gender related issues were highlighted. It is likely that some reviews did not reference all studies submitted by the sponsors. Some of the reviews noted gender effects that did not appear in the labeling. Internal memos, meeting records, or other documents that would indicate why such statements do not appear in the labeling were not reviewed. Label Reviews The review of the labels showed that most products contained gender-related statements that indicate sponsors are aware of the need to look for gender effects. The majority of the products did not demonstrate a gender difference. The labels described gender effects related primarily to pharmacokinetic differences. No product recommended a different dosage for women. Excluded from the labeling review are any statements about teratogenicity from animal studies or statements about pregnant or lactating women. Limitations of the Study The limitations of this study are listed below: · The study was conducted using medical officers' reviews and product labeling only. There were large numbers of participants whose gender could not be determined from the reviews. The data presented do not reflect the gender representation and analysis in the complete submission by the sponsors. · The documents used may not represent the complete FDA analysis. Additional information, such as meeting records, memos, and supervisory, clinical pharmacology, or statistical reviews, was not included. · The study reviewed only approved products and are a subset of all products reviewed by FDA staff. · Study phase data were captured only if specifically described in the review. More accurate data might have been obtained if other bases for categorizing studies (e.g., study size) had been established. Conclusions and Recommendations This study implements, in part, the July 1998 FDAMA Working Group initiative to gather and evaluate demographic data. Overall participation by women and men was similar even when trials for gender-specific products were excluded. The majority of products contain labeling statements about whether the effect of gender was known but safety and efficacy data are analyzed to varying degrees for the specific influence of gender. Attention to gender-related issues by CDER medical officer reviews is variable. The recently launched Clinical Review Template will be implemented over the next year and help facilitate consistency among reviews. The template is intended primarily to standardize the ordering and placement of topics within reviews. The section on Special Populations should discuss how many males and females were exposed to the investigational product during clinical trials and how well the data were analyzed for differences between them. These issues will also be addressed through reviewer education by a web based self-learning program on gender differences. Additional studies using actual sponsor submissions may be needed to determine if FDA should provide additional guidance to industry regarding demographic subgroup analysis in NDAs.
FDA/Center for Drug Evaluation and Research |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
