Review Of Individual Patient Data Research Articles

Effects of radiotherapy and of differences in the extent of surgery for early breast cancer on local recurrence and 15-year survival: an overview of the randomised trials

In early breast cancer, variations in local treatment that substantially affect the risk of locoregional recurrence could also affect long-term breast cancer mortality. To examine this relationship, collaborative meta-analyses were undertaken, based on individual patient data, of the relevant randomised trials that began by 1995. Information was available on 42,000 women in 78 randomised treatment comparisons (radiotherapy vs no radiotherapy, 23,500; more vs less surgery, 9300; more surgery vs radiotherapy, 9300). 24 types of local treatment comparison were identified. To help relate the effect on local (ie, locoregional) recurrence to that on breast cancer mortality, these were grouped according to whether or not the 5-year local recurrence risk exceeded 10% (<10%, 17,000 women; >10%, 25,000 women). About three-quarters of the eventual local recurrence risk occurred during the first 5 years. In the comparisons that involved little (<10%) difference in 5-year local recurrence risk there was little difference in 15-year breast cancer mortality. Among the 25,000 women in the comparisons that involved substantial (>10%) differences, however, 5-year local recurrence risks were 7% active versus 26% control (absolute reduction 19%), and 15-year breast cancer mortality risks were 44.6% versus 49.5% (absolute reduction 5.0%, SE 0.8, 2p<0.00001). These 25,000 women included 7300 with breast-conserving surgery (BCS) in trials of radiotherapy (generally just to the conserved breast), with 5-year local recurrence risks (mainly in the conserved breast, as most had axillary clearance and node-negative disease) 7% versus 26% (reduction 19%), and 15-year breast cancer mortality risks 30.5% versus 35.9% (reduction 5.4%, SE 1.7, 2p=0.0002; overall mortality reduction 5.3%, SE 1.8, 2p=0.005). They also included 8500 with mastectomy, axillary clearance, and node-positive disease in trials of radiotherapy (generally to the chest wall and regional lymph nodes), with similar absolute gains from radiotherapy; 5-year local recurrence risks (mainly at these sites) 6% versus 23% (reduction 17%), and 15-year breast cancer mortality risks 54.7% versus 60.1% (reduction 5.4%, SE 1.3, 2p=0.0002; overall mortality reduction 4.4%, SE 1.2, 2p=0.0009). Radiotherapy produced similar proportional reductions in local recurrence in all women (irrespective of age or tumour characteristics) and in all major trials of radiotherapy versus not (recent or older; with or without systemic therapy), so large absolute reductions in local recurrence were seen only if the control risk was large. To help assess the life-threatening side-effects of radiotherapy, the trials of radiotherapy versus not were combined with those of radiotherapy versus more surgery. There was, at least with some of the older radiotherapy regimens, a significant excess incidence of contralateral breast cancer (rate ratio 1.18, SE 0.06, 2p=0.002) and a significant excess of non-breast-cancer mortality in irradiated women (rate ratio 1.12, SE 0.04, 2p=0.001). Both were slight during the first 5 years, but continued after year 15. The excess mortality was mainly from heart disease (rate ratio 1.27, SE 0.07, 2p=0.0001) and lung cancer (rate ratio 1.78, SE 0.22, 2p=0.0004). In these trials, avoidance of a local recurrence in the conserved breast after BCS and avoidance of a local recurrence elsewhere (eg, the chest wall or regional nodes) after mastectomy were of comparable relevance to 15-year breast cancer mortality. Differences in local treatment that substantially affect local recurrence rates would, in the hypothetical absence of any other causes of death, avoid about one breast cancer death over the next 15 years for every four local recurrences avoided, and should reduce 15-year overall mortality.

Modified-Release Dipyridamole Combined with Aspirin for Secondary Stroke Prevention

Patients suffering from transient ischemic attack (TIA) or ischemic stroke have a high risk of suffering a first or recurrent stroke, with an annual risk of between 5 and 15%. The European Stroke Prevention Study (ESPS)2 was a randomized, double-blind, placebo-controlled trial involving 6602 patients with TIA or stroke comparing aspirin alone (50 mg daily), modified-release (MR) dipyridamole alone (200 mg twice daily), aspirin plus dipyridamole, and placebo. The 2-year relative risk reduction of stroke in the aspirin plus MR-dipyridamole group (37.0%) was significantly higher than in either the aspirin group (18.1%) or the MR-dipyridamole group (16.3%). The results of the comparison between aspirin plus MR-dipyridamole versus placebo confirmed the findings of ESPS1. The results of ESPS2 were at odds with all prior trials with dipyridamole alone or in association with aspirin, but it was also the only trial sufficiently powered to show a significant difference. For this reason, a meta-analysis was performed based on a systematic review of individual patient data from randomized controlled trials involving dipyridamole in patients with prior ischemic stroke or TIA. Recurrent stroke was reduced by dipyridamole compared with placebo, and by combined aspirin and dipyridamole versus aspirin alone, dipyridamole alone or placebo. In two post hoc analyses the efficacy of aspirin plus extended-release dipyridamole was compared with aspirin alone for the prevention of recurrent stroke among high-risk groups. Stroke models from the Framingham Study and the Stroke Prognostic Instrument II were applied to subjects in ESPS2 to assign patients to risk groups. Compared with aspirin alone, aspirin plus MR-dipyridamole demonstrated a more pronounced efficacy in reducing the risk for stroke and vascular events among patients younger than 70 years of age, with hypertension, or prior stroke; current smokers; and those with any prior cardiovascular disease. Another model (Essen risk score) showed that patients with a high risk of recurrent stroke show a much greater benefit from aspirin plus MR-dipyridamole compared with aspirin monotherapy than patients with a low risk.

Antiplatelet agents for prevention of pre-eclampsia and its consequences: a systematic review and individual patient data meta-analysis.

BackgroundThere is now good evidence that antiplatelet agents (principally low dose aspirin) prevent pre-eclampsia, a leading cause of morbidity and mortality for pregnant women and their babies. A Cochrane Review identified moderate, but clinically important, reductions in the relative risks of pre-eclampsia (19%), preterm birth (7%) and perinatal mortality (16%) in women allocated antiplatelets, rather than placebo or no antiplatelet.Uncertainty remains, however, about whether some women (in terms of risk) benefit more than others, what dose of aspirin is best and when in pregnancy treatment should ideally start. Rather than undertake new trials, the best way to answer these questions is to utilise existing individual patient data from women enrolled in each trial.Methods/DesignSystematic review with meta-analysis based on individual patient data. This involves the central collection, validation and re-analysis of thoroughly checked data from individual women in all the available randomised trials.The objective is to confirm that antiplatelet agents, given during pregnancy, will reduce the incidence of pre-eclampsia. The review will then determine the size of this effect, and whether antiplatelets delay the onset of pre-eclampsia or its impact on important outcomes for women and their babies. It will also explore whether the effect of antiplatelets differs by womens' risk profile; when commenced during pregnancy; and/or by dose.DiscussionThe PARIS (Perinatal Antiplatelet Review of International Studies) Collaboration has been formed to undertake the review. This will be the first individual patient data review in the perinatal field. Final results should be available by 2006–7.

Donepezil for the symptomatic treatment of patients with mild to moderate Alzheimer's disease: a meta-analysis of individual patient data from randomised controlled trials.

The objective was to evaluate the efficacy and tolerability of donepezil (5 and 10 mg/day) compared with placebo in alleviating manifestations of mild to moderate Alzheimer's disease (AD). A systematic review of individual patient data from Phase II and III double-blind, randomised, placebo-controlled studies of up to 24 weeks and completed by 20 December 1999. The main outcome measures were the ADAS-cog, the CIBIC-plus, and reports of adverse events. A total of 2376 patients from ten trials were randomised to either donepezil 5 mg/day (n = 821), 10 mg/day (n = 662) or placebo (n = 893). Cognitive performance was better in patients receiving donepezil than in patients receiving placebo. At 12 weeks the differences in ADAS-cog scores were 5 mg/day-placebo: - 2.1 [95% confidence interval (CI), - 2.6 to - 1.6; p < 0.001], 10 mg/day-placebo: - 2.5 ( - 3.1 to - 2.0; p < 0.001). The corresponding results at 24 weeks were - 2.0 ( - 2.7 to - 1.3; p < 0.001) and - 3.1 ( - 3.9 to - 2.4; p < 0.001). The difference between the 5 and 10 mg/day doses was significant at 24 weeks (p = 0.005). The odds ratios (OR) of improvement on the CIBIC-plus at 12 weeks were: 5 mg/day-placebo 1.8 (1.5 to 2.1; p < 0.001), 10 mg/day-placebo 1.9 (1.5 to 2.4; p < 0.001). The corresponding values at 24 weeks were 1.9 (1.5 to 2.4; p = 0.001) and 2.1 (1.6 to 2.8; p < 0.001). Donepezil was well tolerated; adverse events were cholinergic in nature and generally of mild severity and brief in duration. Donepezil (5 and 10 mg/day) provides meaningful benefits in alleviating deficits in cognitive and clinician-rated global function in AD patients relative to placebo. Increased improvements in cognition were indicated for the higher dose.

Phenobarbitone versus phenytoin monotherapy for partial onset seizures and generalized onset tonic-clonic seizures.

Worldwide, phenytoin and phenobarbitone are commonly used antiepileptic drugs. They are more likely to be used in the developing world than the developed world, primarily because they are inexpensive. The aim of this review is to summarise data from existing trials comparing phenytoin and phenobarbitone. To review the effects of phenobarbitone compared to phenytoin when used as monotherapy in patients with partial onset seizures or generalized tonic-clonic seizures with or without other generalized seizure types. Our search strategy has included: a) MEDLINE 1966 to 1998, b) the controlled trials register of the Cochrane Library, c) hand-searching relevant journals, d) the pharmaceutical industry, e) researchers in the field. Randomized controlled trials in children or adults with partial onset seizures or generalized onset tonic-clonic seizures. Trials must have included a comparison of phenobarbitone monotherapy with phenytoin monotherapy. This was an individual patient data review. Outcomes were time to a) withdrawal of allocated treatment, b) 12 month remission, and c) first seizure post randomization. Data were analysed using a stratified logrank analysis with results expressed as hazard ratios (HR) and 95% confidence intervals (95% CI), where a HR>1 indicates an event is more likely to occur earlier on phenobarbitone than phenytoin. To date, data have been obtained for four of ten studies meeting the inclusion criteria, amounting to 599 patients, or approximately 65% of the potential data. The main overall results (HR, 95% CI) were: a) time to treatment withdrawal 1.62 (1.22 to 2.14), b) time to 12 month remission 0.93 (0.70 to 1.23), c) time to first seizure 0.84 (0.68 to 1.05). These results indicate a statistically significant clinical advantage for phenytoin in terms of treatment withdrawal and a non-significant advantage in terms of 12 month remission. Results for time to first seizure suggest a non-significant clinical advantage for phenobarbitone. The results of this review favour phenytoin over phenobarbitone, as phenobarbitone was significantly more likely to be withdrawn than phenytoin. Given that no significant differences for seizure outcomes were found, the higher withdrawal rate with phenobarbitone may be due to side effects.

Selegiline for Alzheimer's disease.

Alzheimer's disease is the most common cause of dementia in older people accounting for some 60% of cases with late-onset cognitive deterioration. It is now thought that several neurotransmitter dysfunctions are involved from an early stage in the pathogenesis of Alzheimer's disease-associated cognitive decline. The efficacy of selegiline for symptoms of Alzheimer's disease remains controversial and is reflected by its low rate of prescription and the lack of approval by several regulatory authorities in Europe and elsewhere. Reasons for this uncertainty involve the modest overall effects observed in some trials, the lack of benefit observed in several trials, the use of cross-over designs which harbour methodological problems in a disease like dementia and the difficulty in interpreting results from trials when a variety of measurement scales are used to assess outcomes. The objective of this review is to assess whether or not selegiline improves the well-being of patients with Alzheimer's disease. The Cochrane Dementia and Cognitive Impairment Group Register of Clinical Trials, was searched using the terms 'selegiline', 'l-deprenyl', "eldepryl" and "monamine oxidase inhibitor-B". MEDLINE, PsycLIT and EMBASE electronic databases were searched with the above terms in addition to using the group strategy (see group details) to limit the searches to randomised controlled trials. All unconfounded, double-blind, randomised controlled trials in which treatment with selegiline was administered for more than a day and compared to placebo in patients with dementia. Data were extracted independently by the reviewers (JSB & LF), pooled where appropriate and possible, and the weighted or standardised mean differences (95%CI) estimated. Where possible, intention-to-treat data were used but usually the meta analyses were restricted to completers' data (data on people who completed the study). There are 15 included trials. All trials examined the cognitive effects of selegiline, and in addition, 12 trials examined the behavioural and mood effects. The results of 8 trials suggested some beneficial effect of selegiline in the treatment of cognitive deficits and in 3 trials in the treatment of behaviour and mood. The meta-analysis revealed benefits on memory function as evidenced by improvement in the memory tests from several cognitive tests. Pooling the data for all cognitive tests suggested significant benefits in those subjects treated with selegiline as compared to controls. There were benefits in mood and behaviour as demonstrated by measurements on the Brief Psychiatric Rating Scale and the Dementia Mood Assessment Scale. The global rating scales showed no effect of selegiline. Unfortunately, the evidence using standardised global cognitive scales was extremely limited, for both the MMSE and ADAS-cog. A variety of adverse effects were recorded, but very few patients left a trial as a direct result of the intervention. Although the evidence for a beneficial effect of selegiline on patients with Alzheimer's disease is promising there is not yet enough evidence to recommend its use routinely in practice. The individual patient data review will yield further evidence on the effects of selegiline compared to control as would additional studies evaluating the use of selegiline for the endpoints of standardised cognitive scales, clinician impression of global change, dependency and caregiver quality of life.

Publication bias and meta-analyses: a practical example.

Publication bias is widely appreciated, but considerable time and effort are needed to locate and obtain data from unpublished randomized controlled trials (RCTs), those published in non-English language journals or those reported in the gray literature; for this publication, we will call this collection of trials the "gray+literature." However, excluding such trials from systematic reviews could introduce bias and give rise to misleading conclusions. We aimed to explore and quantify the impact of inclusion of gray+ literature on the results of all completed individual patient data (IPD) reviews coordinated by our group (13 meta-analyses). For each IPD review, results were calculated for RCTs fully published in English language journals and RCTs fully published in English language journals and the gray+literature. The IPD meta-analyses based only on RCTs that were fully published in English language journals tended to give more favorable results than those that included RCTs from the gray+literature. Although in most cases the addition of gray+data gave less encouraging results, moving the estimated treatment effect toward a null result, the direction of effect was not always predictable. We recommend that all systematic reviews should at least attempt to identify trials reported in the gray+literature and, where possible, obtain data from them.

Radiotherapy for early breast cancer.

The long-term effects of radiotherapy on mortality from breast cancer and other causes remain uncertain. In this report the Early Breast Cancer Trialists' Collaborative Group present their systematic overview of treatment with radiotherapy. Trial identification procedures for the EBCTCG overviews have been described elsewhere. See under "EBCTCG" in the Breast Cancer Group module. A meta-analysis was done of 10-year and 20-year results from 40 unconfounded randomised trials of radiotherapy for early breast cancer. Radiotherapy fields generally included not only chest wall (or breast) but also axillary, supraclavicular, and internal mammary nodes. Data collection involved central review of individual patient data on recurrence and cause-specific mortality from 20,000 women. A reduction of approximately two-thirds in local recurrence was seen in all trials, largely independent of the type of patient or type of radiotherapy (8.8% vs 27.2% local recurrence by year 10). Hence, to assess effects on breast cancer mortality of substantially better local control, results from all trials were combined. Breast cancer mortality was reduced (2p=0.0001) but other, particularly vascular, mortality was increased (2p=0.0003), and overall 20-year survival was 37.1% with radiotherapy versus 35.9% control (2p=0.06). There was little effect on early deaths, but logrank analyses of later deaths indicate that, on average after year 2, radiotherapy reduced annual mortality rates from breast cancer by 13.2% (SE 2.5) but increased those from other causes by 21.2% (SE 5.4). Nodal status, age, and decade of follow-up strongly affected the ratio of breast cancer mortality to other mortality, and hence affected the ratio of absolute benefit to absolute hazard from these proportional changes in mortality. Radiotherapy regimens able to produce the two-thirds reduction in local recurrence seen in these trials, but without long-term hazard, would be expected to produce an absolute increase in 20-year survival of about 2-4% (except for women at particularly low risk of local recurrence). The average hazard seen in these trials would, however, reduce this 20-year survival benefit in young women and reverse it in older women.

Carbamazepine versus phenytoin monotherapy for epilepsy.

Worldwide, carbamazepine and phenytoin are commonly used antiepileptic drugs. This review summarizes evidence from randomized controlled trials in which these two drugs have been compared. To review the best evidence comparing carbamazepine and phenytoin when used as monotherapy in subjects with partial onset seizures, or generalized onset tonic-clonic seizures with or without other generalized seizure types. We searched: (a) the trial register of the Cochrane Epilepsy Group; (b) The Cochrane Controlled Trials Register (Cochrane Library Issue 4, 2001); (c) MEDLINE 1966-2001. In addition we hand searched relevant journals and contacted the pharmaceutical industry and researchers in the field to seek any ongoing or unpublished studies. Randomized controlled trials in children or adults with partial onset seizures or generalized onset tonic-clonic seizures. Trials must have included a comparison of carbamazepine monotherapy with phenytoin monotherapy. This was an individual patient data review. Outcomes were time to (a) withdrawal of allocated treatment, (b) 12 month remission, (c) six month remission, and (d) first seizure post randomization. Data were analysed using a stratified logrank analysis with results expressed as hazard ratios (HR) and 95% confidence intervals (95% CI), where a HR>1 indicates an event is more likely on phenytoin. Individual patient data are available for 551 participants from three trials, representing 63% of the participants recruited into the nine trials that met our inclusion criteria. By convention, for the outcomes time to six and 12 month remission HR>1 indicates a clinical advantage for phenytoin, whilst for time to withdrawal and first seizure HR>1 indicates a clinical advantage for carbamazepine. Results (HR (95% CI)) were: (i) time to withdrawal of allocated treatment 0.97(0.74 to 1.28), (ii) time to 12 month remission 1.00(0.78 to 1.29), (iii) time to six month remission 1.10(0.87 to 1.39), (iv) time to first seizure 0.91(0.74 to 1.12). The results suggest no overall difference between carbamazepine and phenytoin for these outcomes. We have not found evidence that a significant difference exists between carbamazepine and phenytoin for the outcomes examined in this review. Confidence intervals are wide and the possibility of important differences existing has not been excluded.

Phenytoin versus valproate monotherapy for partial onset seizures and generalized onset tonic-clonic seizures.

Phenytoin and valproate are commonly used antiepileptic drugs. It is generally believed that phenytoin is more effective for partial onset (simple partial, complex partial and secondary generalized tonic-clonic seizures) seizures whilst valproate is more effective in generalized onset seizures (generalized tonic-clonic seizures, absence, myoclonus) although there is no evidence from randomized controlled trials to support this belief. The use of individual patient data meta-analysis enabled us to examine time to event outcomes which are important in epilepsy monotherapy trials, and also to examine treatment-covariate interactions. To review the best evidence comparing phenytoin and valproate when used as monotherapy in subjects with partial onset seizures, or generalized onset tonic-clonic seizures with or without other generalized seizure types. Our search strategy included: (i) the Cochrane Epilepsy Group trial register, (ii) MEDLINE 1966-2000, (iii) hand-searching relevant journals, (iv) the pharmaceutical industry, and (v) researchers in the field. Randomized controlled trials in children or adults with partial onset seizures or generalized onset tonic-clonic seizures. Trials must have included a comparison of phenytoin monotherapy with valproate monotherapy. This was an individual patient data review. Outcomes were time to (i) withdrawal of allocated treatment, (ii) 12 month remission, (iii) six month remission, and (iv) first seizure post randomization. Data were analysed using stratified logrank analysis with results expressed as hazard ratios (HR) and 95% confidence intervals (95% CI), where a HR>1 indicates an event is more likely on phenytoin. A test for interaction between treatment and seizure type (partial onset versus generalized onset) was also undertaken for each outcome. Data were available for 669 subjects from five trials, representing 60% of the subjects recruited into the eleven trials that met our inclusion criteria. One important limitation of these data is that in four of the five trials, for patients classified as having generalized onset seizures, tonic-clonic seizures were the only seizure types recorded at follow up, despite the fact that some patients will have been experiencing other generalized seizure types such as absence or myoclonus. Their results for the generalized seizures therefore relate only to generalized onset tonic-clonic seizures. The main overall results were as follows (HR(95% CI), HR>1 indicates a clinical advantage for phenytoin for both remission outcomes and a clinical advantage for valproate for the outcomes time to withdrawal and time to first seizure) (i) time to withdrawal of allocated treatment 1.10(0.79-1.54), (ii) time to 12 month remission 1.04(0.78-1.38), (iii) time to six month remission 0.89(0.71-1.11), and (iv) time to first seizure 0.92(0.74-1.14). The results suggest no overall difference between drugs for these outcomes. The test for an interaction between treatment and seizure type (generalized versus partial onset) was non significant for all outcomes. We have not found evidence that a significant difference exists between phenytoin and valproate for the outcomes examined in this review. As generalized seizures such as absence and myoclonus were counted in only one trial, results do not address the treatment of these seizure types. We found no unequivocal evidence to overthrow or support the policy of using valproate in generalized onset tonic-clonic seizures and phenytoin in partial onset seizures.

Carbamazepine versus valproate monotherapy for epilepsy.

Carbamazepine and valproate are drugs of first choice for epilepsy. Despite the lack of hard evidence from individual randomized controlled trials, there is strong clinical belief that valproate is the drug of choice for generalized epilepsies and carbamazepine for partial epilepsies. To overview the best evidence comparing carbamazepine and valproate monotherapy Our search strategy included: (a) MEDLINE 1966-99, (b) The Cochrane Library 1999 issue 4, (c) The trial register of the Cochrane Epilepsy Group (d) the pharmaceutical industry. Randomized controlled trials comparing carbamazepine and valproate monotherapy for epilepsy. This was an individual patient data review. Outcome measures were time to withdrawal of allocated treatment, time to 12 month remission, and time to first seizure post randomization. Data were analysed using the stratified Logrank test with results expressed as hazard ratios (HR) (95% CI), where HR>1 indicates an event is more likely on valproate. A test for an interaction between treatment and epilepsy type (partial versus generalized) was also undertaken. Results Data were available for 1265 patients from five trials, representing 85% of the patients recruited into the eight trials that met our inclusion criteria. The main overall results (HR 95% CI) were: Time to treatment withdrawal 0.97 (0.79-1.18), 12 month remission 0.87 (0.74-1.02), first seizure 1.09 (0.96-1.25) suggesting no overall difference for these outcomes. The test for an interaction between treatment and epilepsy type was non significant for time to treatment withdrawal and 12 month remission, but significant for time to first seizure. The age distribution of adults classified as having a generalized epilepsy indicate that significant numbers of patients may have had their epilepsy misclassified. We have found some evidence to support the policy of using carbamazepine as the first treatment of choice in partial epilepsies, but no evidence to support the choice of valproate in generalized epilepsies, but confidence intervals are too wide to confirm equivalence. Misclassification of patients may have confounded our results, and has important implications for the design and conduct of future trials.