Multiplicity issues in exploratory subgroup analysis.

The general topic of subgroup identification has attracted much attention in the clinical trial literature due to its important role in the development of tailored therapies and personalized medicine. Subgroup search methods are commonly used in late-phase clinical trials to identify subsets of the trial population with certain desirable characteristics. Post-hoc or exploratory subgroup exploration has been criticized for being extremely unreliable. Principled approaches to exploratory subgroup analysis based on recent advances in machine learning and data mining have been developed to address this criticism. These approaches emphasize fundamental statistical principles, including the importance of performing multiplicity adjustments to account for selection bias inherent in subgroup search. This article provides a detailed review of multiplicity issues arising in exploratory subgroup analysis. Multiplicity corrections in the context of principled subgroup search will be illustrated using the family of SIDES (subgroup identification based on differential effect search) methods. A case study based on a Phase III oncology trial will be presented to discuss the details of subgroup search algorithms with resampling-based multiplicity adjustment procedures.

General guidance on exploratory and confirmatory subgroup analysis in late-stage clinical trials.

This article focuses on a broad class of statistical and clinical considerations related to the assessment of treatment effects across patient subgroups in late-stage clinical trials. This article begins with a comprehensive review of clinical trial literature and regulatory guidelines to help define scientifically sound approaches to evaluating subgroup effects in clinical trials. All commonly used types of subgroup analysis are considered in the article, including different variations of prospectively defined and post-hoc subgroup investigations. In the context of confirmatory subgroup analysis, key design and analysis options are presented, which includes conventional and innovative trial designs that support multi-population tailoring approaches. A detailed summary of exploratory subgroup analysis (with the purpose of either consistency assessment or subgroup identification) is also provided. The article promotes a more disciplined approach to post-hoc subgroup identification and formulates key principles that support reliable evaluation of subgroup effects in this setting.

The levonorgestrel-releasing intrauterine system provides a reliable, long-term treatment option for women with idiopathic menorrhagia.

OBJECTIVES: Idiopathic menorrhagia (IM) is an important clinical challenge. The levonorgestrel-releasing intrauterine system (LNG IUS) provides an effective treatment option as shown by multiple small clinical studies. In this analysis of combined data, we describe the time course of relative change in menstrual blood loss (MBL) from baseline up to 5 years. The results of two different methods to assess MBL were merged. METHODS: We pooled and analyzed five prospective, randomized clinical studies investigating the effect of the LNG IUS on IM in a total of 230 women. Four studies assessed MBL by using the pictorial blood loss assessment chart (PBAC) and one study used the alkaline hematin method. We gathered data on percentage change from baseline after 3 and 6 months, and annually up to 5 years. In addition we analyzed results on hemoglobin (Hb) and serum ferritin (S-Fe). RESULTS: MBL data was available after 3 and 6 months from 165 and 152 patients, respectively, and after 1 year from 51 patients. Long-term data up to 3 and 5 years was available for 28 and 10 patients, respectively. Not all studies provided data for all time points. Median (interquartile range) MBL decreased from baseline by -84.5% (-93.3; -63.6%) after 3 months, by -92.9% (-97.6; -81.1%) and by -93.8% (-98.8; -81.1%) after 6 months and 1 year, respectively (P < 0.0001, all time points). After 2 and 5 years the decrease was more than 96%. In parallel, Hb and S-Fe increased significantly. CONCLUSION: The LNG IUS rapidly induced clinically and statistically significant long-term reductions in MBL, paralleled by increases in Hb and S-Fe levels.

A novel concept of screening for subgrouping factors for the association between socioeconomic status and respiratory allergies.

BACKGROUND: The new subgroup screening tool "subscreen" aims to understand the unclear and complex association between socioeconomic status (SES) and childhood allergy. This software R package has been successfully used in clinical trials but not in large population-based studies. OBJECTIVE: To screen and identify subgrouping factors explaining their impact on the association between SES and respiratory allergies in childhood and youth. METHODS: Using the national German childhood and youth survey dataset (KiGGS Wave 2), we included 56 suspected subgrouping factors to investigate the association between SES (low vs. high) and allergic rhinitis and/or asthma in an exploratory manner. The package enabled a comprehensive overview of odds ratios when considering the SES impact per subgroup and analogously all disease proportions per subgroup. RESULT: Among the 56 candidate factors, striking subgrouping factors were identified; e.g., if mothers were younger and in the low SES group, their children had a higher risk of asthma. In addition children of the teen's age were associated with increased risks in the low SES group. For the crude proportions, factors such as (parental) smoking or having had no "contact with farm animals" were identified as strong risk factors for rhinitis. SIGNIFICANCE: The "subscreen" package enabled the detection of notable subgroups for further investigations exemplarily for similar epidemiological research questions.

A Systematic Approach for Post Hoc Subgroup Analyses With Applications in Clinical Case Studies.

BACKGROUND: The analysis of subgroups in clinical trials is essential to assess differences in treatment effects for distinct patient clusters, that is, to detect patients with greater treatment benefit or patients where the treatment seems to be ineffective. METHODS: The software application subscreen (R package) has been developed to analyze the population of clinical trials in minute detail. The aim was to efficiently calculate point estimates (eg, hazard ratios) for multiple subgroups to identify groups that potentially differ from the overall trial result. The approach intentionally avoids inferential statistics such as P values or confidence intervals but intends to encourage discussions enriched with external evidence (eg, from other studies) about the exploratory results, which can be accompanied by further statistical methods in subsequent analyses. The subscreen application was applied to 2 clinical study data sets and used in a simulation study to demonstrate its usefulness. RESULTS: The visualization of numerous combined subgroups illustrates the homogeneity or heterogeneity of potentially all subgroup estimates with the overall result. With this, the application leads to more targeted planning of future trials. CONCLUSION: This described approach supports the current trend and requirements for the investigation of subgroup effects as discussed in the EMA draft guidance for subgroup analyses in confirmatory clinical trials (EMA 2014). The lack of a convenient tool to answer spontaneous questions from different perspectives can hinder an efficient discussion, especially in joint interdisciplinary study teams. With the new application, an easily executed but powerful tool is provided to fill this gap.

A Systematic Approach for Post Hoc Subgroup Analyses With Applications in Clinical Case Studies.

BACKGROUND: The analysis of subgroups in clinical trials is essential to assess differences in treatment effects for distinct patient clusters, that is, to detect patients with greater treatment benefit or patients where the treatment seems to be ineffective. METHODS: The software application subscreen (R package) has been developed to analyze the population of clinical trials in minute detail. The aim was to efficiently calculate point estimates (eg, hazard ratios) for multiple subgroups to identify groups that potentially differ from the overall trial result. The approach intentionally avoids inferential statistics such as P values or confidence intervals but intends to encourage discussions enriched with external evidence (eg, from other studies) about the exploratory results, which can be accompanied by further statistical methods in subsequent analyses. The subscreen application was applied to 2 clinical study data sets and used in a simulation study to demonstrate its usefulness. RESULTS: The visualization of numerous combined subgroups illustrates the homogeneity or heterogeneity of potentially all subgroup estimates with the overall result. With this, the application leads to more targeted planning of future trials. CONCLUSION: This described approach supports the current trend and requirements for the investigation of subgroup effects as discussed in the EMA draft guidance for subgroup analyses in confirmatory clinical trials (EMA 2014). The lack of a convenient tool to answer spontaneous questions from different perspectives can hinder an efficient discussion, especially in joint interdisciplinary study teams. With the new application, an easily executed but powerful tool is provided to fill this gap.

Levonorgestrel-releasing intrauterine system or medroxyprogesterone for heavy menstrual bleeding: a randomized controlled trial.

OBJECTIVE: To compare the efficacy and safety of the levonorgestrel-releasing intrauterine system and oral medroxyprogesterone acetate in the treatment of idiopathic heavy menstrual bleeding. METHODS: In this multicenter, randomized, controlled study, women aged 18 years or older with heavy menstrual bleeding (menstrual blood loss 80 mL or more per cycle) were randomly assigned to six cycles of treatment with either levonorgestrel-releasing intrauterine system or oral medroxyprogesterone acetate (10 mg daily for 10 days beginning on day 16 of each cycle). The primary efficacy variables were the absolute change in menstrual blood loss from baseline to end of study and the proportion of women with successful treatment (defined as menstrual blood loss less than 80 mL and a 50% or greater reduction in menstrual blood loss from baseline). RESULTS: Of 807 women screened, 165 were randomly assigned to treatment (levonorgestrel-releasing intrauterine system n=82, oral medroxyprogesterone acetate n=83). At the end of the study, the absolute reduction in median menstrual blood loss was significantly greater in the levonorgestrel-releasing intrauterine system group (-128.8 mL, range -393.6 to +1242.2 mL) than in the medroxyprogesterone acetate arm (-17.8 mL, range -271.5 to +78.6 mL, P < .001), and the proportion of women with successful treatment was significantly higher for the levonorgestrel-releasing intrauterine system (84.8%) than for medroxyprogesterone acetate (22.2%, P < .001). CONCLUSION: In women with idiopathic heavy menstrual bleeding, the levonorgestrel-releasing intrauterine system reduces menstrual blood loss more effectively and has a higher likelihood of treatment success than oral medroxyprogesterone acetate. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov, www.clinicaltrials.gov, NCT00360490. LEVEL OF EVIDENCE: I.

Effect of microdose transdermal 17beta-estradiol compared with raloxifene in the prevention of bone loss in healthy postmenopausal women: a 2-year, randomized, double-blind trial.

OBJECTIVE: Declining estrogen levels after menopause result in bone loss and increased fracture risk. This study investigated whether transdermal microdose 17beta-estradiol (E2) has efficacy and safety comparable to those of raloxifene, a selective estrogen-receptor modulator approved for the prevention and treatment of postmenopausal osteoporosis. METHODS: This study involved a multicenter, randomized, double-blind, active-controlled, noninferiority trial in 500 osteopenic postmenopausal women comparing transdermal microdose E2 (0.014 mg/d) versus oral raloxifene (60 mg/d), administered for 2 years. Percent change from baseline in bone mineral density at the lumbar spine was measured after 2 years of treatment. Secondary endpoints included proportion of women with no loss of bone mineral density in lumbar spine, change in bone mineral density at hip, biochemical markers of bone turnover, and safety parameters. RESULTS: In the per protocol set, lumbar spine bone mineral density increased by 2.4% (95% CI, 1.9-2.9) with microdose E2 versus 3.0% (95% CI, 2.5-3.5) with raloxifene after 2 years; 77.3% of E2 recipients and 80.5% of those taking raloxifene had no bone loss in the lumbar spine. Both treatments were well tolerated. Most women (99% in the E2 group and 100% in the raloxifene group) showed no histological evidence of endometrial stimulation after 2 years. Mean dense area in breast mammograms was 19.8% in the E2 group versus 19.0% in the raloxifene group after 2 years. CONCLUSIONS: Transdermal microdose E2 was similarly effective as raloxifene in preventing bone loss at the lumbar spine. Both treatments were well tolerated, with no clinically significant effect on endometrium or breast density.