Guidelines on how to assess the validity of results presented in subgroup analysis of clinical trials

In observational studies, identification of associations within particular subgroups is the usual method of investigation. As an exploratory method, it is the bread and butter of epidemiological research. Nearly everything that has been learned in epidemiology has been derived from the analysis of subgroups. In a randomized clinical trial, the entire purpose is the comparison of the test subjects and the controls, and when there is particular interest in the results of treatment in a certain section of trial participants, a subgroup analysis is performed. These subgroups are examined to see if they are liable to a greater benefit or risk from treatment. Thus, analyzing patient subsets is a natural part of the process of improving therapeutic knowledge through clinical trials. Nevertheless, the reliability of subgroup analysis can often be poor because of problems of multiplicity and limitations in the numbers of patients studied. The naive interpretation of the results of such examinations is a cause of great confusion in the therapeutic literature. We emphasize the need for readers to be aware that inferences based on comparisons between subgroups in randomized clinical trials should be approached more cautiously than those based on the main comparison. That is, subgroup analysis results derived from a sound clinical trial are not necessarily valid; one must not jump to conclusions and accept the validity of subgroup analysis results without an appropriate judgment.

Preinfarction angina prior to first myocardial infarction does not influence long-term prognosis: a retrospective study with subgroup analysis in elderly and diabetic patients.

BACKGROUND AND HYPOTHESIS Although prodromal angina occurring shortly before an acute myocardial infarction (MI) has protective effects against in-hospital complications, this effect has not been well documented after initial hospitalization, especially in older or diabetic patients. We examined whether angina 1 week before a first MI provides protection in these patients. METHODS A total of 290 consecutive patients, 143 elderly (>64 years of age) and 147 adults (<65 years of age), 68 of whom were diabetic (23.4%) and 222 nondiabetic (76.6%), were examined to assess the effect of preceding angina on long-term prognosis (56 months) after initial hospitalization for a first MI. RESULTS No significant differences were found in long-term complications after initial hospitalization in these adult and elderly patients according to whether or not they had prodromal angina (44.4% with angina vs 45.4% without in adults; 45.5% vs 58% in elderly, P < 0.2). Nor were differences found according to their diabetic status (61.5% with angina vs 72.7% without in diabetics; 37.3% vs 38.3% in nondiabetics; P = 0.4). CONCLUSION The occurrence of angina 1 week before a first MI does not confer long-term protection against cardiovascular complications after initial hospitalization in adult or elderly patients, whether or not they have diabetes.

Should Koos grade I vestibular schwannomas be treated early with gamma knife surgery? A subgroup analysis in a series of 190 consecutive patients.

INTRODUCTION: Gamma knife surgery (GKS) for vestibular schwannomas (VS) has a long-term clinical and scientific track record. After a period of de-escalation of dose prescription, results show a high rate of tumor control with improvement of clinical outcome (less than 1% facial palsy, 50-70% hearing preservation). Régis et al. (J Neurosurg 2013;119 Suppl.:105-11) suggested recently that proactive GKS management in intracanalicular tumors is better than a « wait and see » strategy when hearing is still useful at the time of diagnosis. MATERIALS AND METHODS: Based on these previous findings, we prospectively analyzed 190 vestibular schwannomas (VS), treated with GKS as first intention over a period of 4 years (2010-2014). We concentrated on patient, tumor and dosimetric characteristics. Special attention was given on the dose to the cochlea and its impact in maintaining serviceable hearing. RESULTS: The mean follow-up period was 1.3years (range 0.6-3.6). Preoperative serviceable hearing was present in 63.11% patients. The mean maximal diameter was 15.1mm (range 5-29.5). The size and volume of the tumor corresponded to Koos grade I, II, III and IV in 15.9%, 34.8%, 45.4% and 3.8% of the cases, respectively. The mean target volume was 1.24cm(3) (0.017-7.8). The mean prescription isodose volume was 1.6 cc (0.032-8.5). The mean marginal dose was 12Gy (11-12). The mean maximal dose received by the cochlea in patients with GR class 1 and 2 was 4.1Gy (1.5-7.6). Our preliminary neuroradiological follow-up shows 97% tumor control, with 45% shrinkage. Patients presenting with GR class 1 and class 2 at baseline retained serviceable hearing in 85% of cases. Among the patients with a follow-up of at least one year, those with Koos I tumors had the highest probability to maintain identical level of hearing after GKS. CONCLUSION: Our preliminary data suggest that Koos I patients should be treated early with GKS, before tumor growth and/or hearing deterioration, as they have the highest probability of hearing preservation.

Reporting on methods of subgroup analysis in clinical trials: a survey of four scientific journals

Results of subgroup analysis (SA) reported in randomized clinical trials (RCT) cannot be adequately interpreted without information about the methods used in the study design and the data analysis. Our aim was to show how often inaccurate or incomplete reports occur. First, we selected eight methodological aspects of SA on the basis of their importance to a reader in determining the confidence that should be placed in the author's conclusions regarding such analysis. Then, we reviewed the current practice of reporting these methodological aspects of SA in clinical trials in four leading journals, i.e., the New England Journal of Medicine, the Journal of the American Medical Association, the Lancet, and the American Journal of Public Health. Eight consecutive reports from each journal published after July 1, 1998 were included. Of the 32 trials surveyed, 17 (53%) had at least one SA. Overall, the proportion of RCT reporting a particular methodological aspect ranged from 23 to 94%. Information on whether the SA preceded/followed the analysis was reported in only 7 (41%) of the studies. Of the total possible number of items to be reported, NEJM, JAMA, Lancet and AJPH clearly mentioned 59, 67, 58 and 72%, respectively. We conclude that current reporting of SA in RCT is incomplete and inaccurate. The results of such SA may have harmful effects on treatment recommendations if accepted without judicious scrutiny. We recommend that editors improve the reporting of SA in RCT by giving authors a list of the important items to be reported.

Effective elements of cognitive behaviour therapy for psychosis: results of a novel type of subgroup analysis based on principal stratification

Background. Meta-analyses show that cognitive behaviour therapy for psychosis (CBT-P) improves distressing positive symptoms. However, it is a complex intervention involving a range of techniques. No previous study has assessed the delivery of the different elements of treatment and their effect on outcome. Our aim was to assess the differential effect of type of treatment delivered on the effectiveness of CBT-P, using novel statistical methodology. Method. The Psychological Prevention of Relapse in Psychosis (PRP) trial was a multi-centre randomized controlled trial (RCT) that compared CBT-P with treatment as usual (TAU). Therapy was manualized, and detailed evaluations of therapy delivery and client engagement were made. Follow-up assessments were made at 12 and 24 months. In a planned analysis, we applied principal stratification (involving structural equation modelling with finite mixtures) to estimate intention-to-treat (ITT) effects for subgroups of participants, defined by qualitative and quantitative differences in receipt of therapy, while maintaining the constraints of randomization. Results. Consistent delivery of full therapy, including specific cognitive and behavioural techniques, was associated with clinically and statistically significant increases in months in remission, and decreases in psychotic and affective symptoms. Delivery of partial therapy involving engagement and assessment was not effective. Conclusions. Our analyses suggest that CBT-P is of significant benefit on multiple outcomes to patients able to engage in the full range of therapy procedures. The novel statistical methods illustrated in this report have general application to the evaluation of heterogeneity in the effects of treatment.

Overall Survival Improvement in Patients with Lung Cancer and Bone Metastases Treated with Denosumab Versus Zoledronic Acid Subgroup Analysis from a Randomized Phase 3 Study

Introduction: Denosumab, a fully human anti-RANKL monoclonal antibody, reduces the incidence of skeletal-related events in patients with bone metastases from solid tumors. We present survival data for the subset of patients with lung cancer, participating in the phase 3 trial of denosumab versus zoledronic acid (ZA) in the treatment of bone metastases from solid tumors (except breast or prostate) or multiple myeloma. Methods: Patients were randomized 1:1 to receive monthly subcutaneous denosumab 120 mg or intravenous ZA 4 mg. An exploratory analysis, using Kaplan-Meier estimates and proportional hazards models, was performed for overall survival among patients with non-small-cell lung cancer (NSCLC) and SCLC. Results: Denosumab was associated with improved median overall survival versus ZA in 811 patients with any lung cancer (8.9 versus 7.7 months; hazard ratio [HR] 0.80) and in 702 patients with NSCLC (9.5 versus 8.0 months; HR 0.78) (p = 0.01, each comparison). Further analysis of NSCLC by histological type showed a median survival of 8.6 months for denosumab versus 6.4 months for ZA in patients with squamous cell carcinoma (HR 0.68; p = 0.035). Incidence of overall adverse events was balanced between treatment groups; serious adverse events occurred in 66.0% of denosumab-treated patients and 72.9% of ZA-treated patients. Cumulative incidence of osteonecrosis of the jaw was similar between groups (0.7% denosumab versus 0.8% ZA). Hypocalcemia rates were 8.6% with denosumab and 3.8% with ZA. Conclusion: In this exploratory analysis, denosumab was associated with improved overall survival compared with ZA, in patients with metastatic lung cancer.

Rivaroxaban compared with warfarin in patients with atrial fibrillation and previous stroke or transient ischaemic attack: a subgroup analysis of ROCKET AF

Background In ROCKET AF, rivaroxaban was non-inferior to adjusted-dose warfarin in preventing stroke or systemic embolism among patients with atrial fibrillation (AF). We aimed to investigate whether the efficacy and safety of rivaroxaban compared with warfarin is consistent among the subgroups of patients with and without previous stroke or transient ischaemic attack (TIA). Methods In ROCKET AF, patients with AF who were at increased risk of stroke were randomly assigned (1:1) in a double-blind manner to rivaroxaban 20 mg daily or adjusted dose warfarin (international normalised ratio 2-0-3.0). Patients and investigators were masked to treatment allocation. Between Dec 18,2006, and June 17,2009,14 264 patients from 1178 centres in 45 countries were randomly assigned. The primary endpoint was the composite of stroke or non-CNS systemic embolism. In this substudy we assessed the interaction of the treatment effects of rivaroxaban and warfarin among patients with and without previous stroke or TIA. Efficacy analyses were by intention to treat and safety analyses were done in the on-treatment population. ROCKET AF is registered with ClinicalTrials.gov, number NCT00403767. Findings 7468 (52%) patients had a previous stroke (n=4907) or TIA (n=2561) and 6796 (48%) had no previous stroke or TIA. The number of events per 100 person-years for the primary endpoint in patients treated with rivaroxaban compared with warfarin was consistent among patients with previous stroke or TIA (2.79% rivaroxaban vs 2.96% warfarin; hazard ratio [HR] 0-94,95% CI 0.77-1.16) and those without (1.44% vs 1.88%; 0.77, 0.58-1-01; interaction p=0.23). The number of major and non-major clinically relevant bleeding events per 100 person-years in patients treated with rivaroxaban compared with warfarin was consistent among patients with previous stroke or TIA (13.31% rivaroxaban vs 13.87% warfarin; HR 0.96,95% CI 0.87-1-07) and those without (16.69% vs 15.19%; 1.10, 0.99-1.21; interaction p=0.08). Interpretation There was no evidence that the relative efficacy and safety of rivaroxaban compared with warfarin was different between patients who had a previous stroke or TIA and those who had no previous stroke or TIA. These results support the use of rivaroxaban as an alternative to warfarin for prevention of recurrent as well as initial stroke in patients with AF.

Impact of vessel size on outcome after implantation of sirolimus-eluting and paclitaxel-eluting stents: a subgroup analysis of the SIRTAX trial

OBJECTIVES: We assessed the impact of vessel size on angiographic and long-term clinical outcome after percutaneous coronary intervention (PCI) with sirolimus-eluting stents (SES) and paclitaxel-eluting stents (PES) within a randomized trial (SIRTAX [Sirolimus-Eluting Stent Compared With Paclitaxel-Eluting Stent for Coronary Revascularization]). BACKGROUND: Percutaneous coronary intervention in small-vessel disease is associated with an increased risk of major adverse cardiac events (MACE). METHODS: A total of 1,012 patients were randomly assigned to treatment with SES (n = 503) or PES (n = 509). A stratified analysis of angiographic and clinical outcome was performed up to 2 years after PCI according to size of the treated vessel (reference vessel diameter < or =2.75 vs. >2.75 mm). RESULTS: Of 1,012 patients, 370 patients (37%) with 495 lesions underwent stent implantation in small vessels only, 504 patients (50%) with 613 lesions in large vessels only, and 138 patients (14%) with 301 lesions in both small and large vessels (mixed). In patients with small-vessel stents, SES reduced MACE by 55% (10.4% vs. 21.4%; p = 0.004), mainly driven by a 69% reduction of target lesion revascularization (TLR) (6.0% vs. 17.7%; p = 0.001) compared with PES at 2 years. In patients with large- and mixed-vessel stents, rates of MACE (large: 10.4% vs. 13.1%; p = 0.33; mixed: 16.7% vs. 18.0%; p = 0.83) and TLR (large: 6.9% vs. 8.6%; p = 0.47; mixed: 16.7% vs. 15.4%; p = 0.86) were similar for SES and PES. There were no significant differences with respect to death and myocardial infarction between the 3 groups. CONCLUSIONS: Compared with PES, SES more effectively reduced MACE and TLR in small-vessel disease. Differences between SES and PES appear less pronounced in patients with large- and mixed-vessel disease. (The SIRTAX trial; http://clinicaltrials.gov/ct/show/NCT00297661?order=1; NCT00297661).

Impact of local endothelial shear stress on neointima and plaque following stent implantation in patients with ST-elevation myocardial infarction: A subgroup-analysis of the COMFORTABLE AMI-IBIS 4 trial.

BACKGROUND Numerous studies have demonstrated an association between endothelial shear stress (ESS) and neointimal formation after stent implantation. However, the role of ESS on the composition of neointima and underlying plaque remains unclear. METHODS Patients recruited in the Comfortable AMI-IBIS 4 study implanted with bare metal stents (BMS) or biolimus eluting stents (BES) that had biplane coronary angiography at 13month follow-up were included in the analysis. The intravascular ultrasound virtual-histology (IVUS-VH) and the angiographic data were used to reconstruct the luminal surface, and the stent in the stented segments. Blood flow simulation was performed in the stent surface, which was assumed to represent the luminal surface at baseline, to assess the association between ESS and neointima thickness. The predominant ESS was estimated in 3-mm segments and was correlated with the amount of neointima, neointimal tissue composition, and with the changes in the underlying plaque burden and composition. RESULTS Forty three patients (18 implanted with BMS and 25 with BES) were studied. In both stent groups negative correlations were noted between ESS and neointima thickness in BMS (P<0.001) and BES (P=0.002). In BMS there was a negative correlation between predominant ESS and the percentage of the neointimal necrotic core component (P=0.015). In BES group, the limited neointima formation did not allow evaluation of the effect of ESS on its tissue characteristics. ESS did not affect vessel wall remodeling and the plaque burden and composition behind BMS (P>0.10) and BES (P>0.45). CONCLUSIONS ESS determines neointimal formation in both BMS and BES and affects the composition of the neointima in BMS. Conversely, ESS does not impact the plaque behind struts irrespective of stent type throughout 13months of follow-up.

Clinical Outcomes According to Diabetic Status in Patients Treated With Biodegradable Polymer Sirolimus-Eluting Stents Versus Durable Polymer Everolimus-Eluting Stents: Prespecified Subgroup Analysis of the BIOSCIENCE Trial.

BACKGROUND Ultrathin strut biodegradable polymer sirolimus-eluting stents (BP-SES) proved noninferior to durable polymer everolimus-eluting stents (DP-EES) for a composite clinical end point in a population with minimal exclusion criteria. We performed a prespecified subgroup analysis of the Ultrathin Strut Biodegradable Polymer Sirolimus-Eluting Stent Versus Durable Polymer Everolimus-Eluting Stent for Percutaneous Coronary Revascularisation (BIOSCIENCE) trial to compare the performance of BP-SES and DP-EES in patients with diabetes mellitus. METHODS AND RESULTS BIOSCIENCE trial was an investigator-initiated, single-blind, multicentre, randomized, noninferiority trial comparing BP-SES versus DP-EES. The primary end point, target lesion failure, was a composite of cardiac death, target-vessel myocardial infarction, and clinically indicated target lesion revascularization within 12 months. Among a total of 2119 patients enrolled between February 2012 and May 2013, 486 (22.9%) had diabetes mellitus. Overall diabetic patients experienced a significantly higher risk of target lesion failure compared with patients without diabetes mellitus (10.1% versus 5.7%; hazard ratio [HR], 1.80; 95% confidence interval [CI], 1.27-2.56; P=0.001). At 1 year, there were no differences between BP-SES versus DP-EES in terms of the primary end point in both diabetic (10.9% versus 9.3%; HR, 1.19; 95% CI, 0.67-2.10; P=0.56) and nondiabetic patients (5.3% versus 6.0%; HR, 0.88; 95% CI, 0.58-1.33; P=0.55). Similarly, no significant differences in the risk of definite or probable stent thrombosis were recorded according to treatment arm in both study groups (4.0% versus 3.1%; HR, 1.30; 95% CI, 0.49-3.41; P=0.60 for diabetic patients and 2.4% versus 3.4%; HR, 0.70; 95% CI, 0.39-1.25; P=0.23, in nondiabetics). CONCLUSIONS In the prespecified subgroup analysis of the BIOSCIENCE trial, clinical outcomes among diabetic patients treated with BP-SES or DP-EES were comparable at 1 year. CLINICAL TRIAL REGISTRATION URL: http://www.clinicaltrials.gov. Unique identifier: NCT01443104.

Effect of telmisartan on functional outcome, recurrence, and blood pressure in patients with acute mild ischemic stroke: a PRoFESS subgroup analysis

Background and Purpose—High blood pressure (BP) is common in acute ischemic stroke and associated independently with a poor functional outcome. However, the management of BP acutely remains unclear because no large trials have been completed. Methods—The factorial PRoFESS secondary stroke prevention trial assessed BP-lowering and antiplatelet strategies in 20 332 patients; 1360 were enrolled within 72 hours of ischemic stroke, with telmisartan (angiotensin receptor antagonist, 80 mg/d, n647) vs placebo (n713). For this nonprespecified subgroup analysis, the primary outcome was functional outcome at 30 days; secondary outcomes included death, recurrence, and hemodynamic measures at up to 90 days. Analyses were adjusted for baseline prognostic variables and antiplatelet assignment. Results—Patients were representative of the whole trial (age 67 years, male 65%, baseline BP 147/84 mm Hg, small artery disease 60%, NIHSS 3) and baseline variables were similar between treatment groups. The mean time from stroke to recruitment was 58 hours. Combined death or dependency (modified Rankin scale: OR, 1.03; 95% CI, 0.84–1.26; P0.81; death: OR, 1.05; 95% CI, 0.27–4.04; and stroke recurrence: OR, 1.40; 95% CI, 0.68–2.89; P0.36) did not differ between the treatment groups. In comparison with placebo, telmisartan lowered BP (141/82 vs 135/78 mmHg, difference 6 to 7 mmHg and 2 to 4 mmHg; P0.001), pulse pressure (3 to 4 mmHg; P0.002), and rate-pressure product (466 mmHg.bpm; P0.0004). Conclusion—Treatment with telmisartan in 1360 patients with acute mild ischemic stroke and mildly elevated BP appeared to be safe with no excess in adverse events, was not associated with a significant effect on functional dependency, death, or recurrence, and modestly lowered BP.

Superior outcome of women with stage I/II cutaneous melanoma: pooled analysis of four European Organisation for Research and Treatment of Cancer phase III trials.

PURPOSE: Several studies observed a female advantage in the prognosis of cutaneous melanoma, for which behavioral factors or an underlying biologic mechanism might be responsible. Using complete and reliable follow-up data from four phase III trials of the European Organisation for Research and Treatment of Cancer (EORTC) Melanoma Group, we explored the female advantage across multiple end points and in relation to other important prognostic indicators. PATIENTS AND METHODS: Patients diagnosed with localized melanoma were included in EORTC adjuvant treatment trials 18832, 18871, 18952, and 18961 and randomly assigned during the period of 1984 to 2005. Cox proportional hazard models were used to calculate hazard ratios (HRs) and 95% CIs for women compared with men, adjusted for age, Breslow thickness, body site, ulceration, performed lymph node dissection, and treatment. RESULTS: A total of 2,672 patients with stage I/II melanoma were included. Women had a highly consistent and independent advantage in overall survival (adjusted HR, 0.70; 95% CI, 0.59 to 0.83), disease-specific survival (adjusted HR, 0.74; 95% CI, 0.62 to 0.88), time to lymph node metastasis (adjusted HR, 0.70; 95% CI, 0.51 to 0.96), and time to distant metastasis (adjusted HR, 0.69; 95% CI, 0.59 to 0.81). Subgroup analysis showed that the female advantage was consistent across all prognostic subgroups (with the possible exception of head and neck melanomas) and in pre- and postmenopausal age groups. CONCLUSION: Women have a consistent and independent relative advantage in all aspects of the progression of localized melanoma of approximately 30%, most likely caused by an underlying biologic sex difference.