Vascular and upper gastrointestinal effects of non-steroidal anti-inflammatory drugs: meta-analyses of individual participant data from randomised trials.

BACKGROUND: The vascular and gastrointestinal effects of non-steroidal anti-inflammatory drugs (NSAIDs), including selective COX-2 inhibitors (coxibs) and traditional non-steroidal anti-inflammatory drugs (tNSAIDs), are not well characterised, particularly in patients at increased risk of vascular disease. We aimed to provide such information through meta-analyses of randomised trials. METHODS: We undertook meta-analyses of 280 trials of NSAIDs versus placebo (124,513 participants, 68,342 person-years) and 474 trials of one NSAID versus another NSAID (229,296 participants, 165,456 person-years). The main outcomes were major vascular events (non-fatal myocardial infarction, non-fatal stroke, or vascular death); major coronary events (non-fatal myocardial infarction or coronary death); stroke; mortality; heart failure; and upper gastrointestinal complications (perforation, obstruction, or bleed). FINDINGS: Major vascular events were increased by about a third by a coxib (rate ratio [RR] 1·37, 95% CI 1·14-1·66; p=0·0009) or diclofenac (1·41, 1·12-1·78; p=0·0036), chiefly due to an increase in major coronary events (coxibs 1·76, 1·31-2·37; p=0·0001; diclofenac 1·70, 1·19-2·41; p=0·0032). Ibuprofen also significantly increased major coronary events (2·22, 1·10-4·48; p=0·0253), but not major vascular events (1·44, 0·89-2·33). Compared with placebo, of 1000 patients allocated to a coxib or diclofenac for a year, three more had major vascular events, one of which was fatal. Naproxen did not significantly increase major vascular events (0·93, 0·69-1·27). Vascular death was increased significantly by coxibs (1·58, 99% CI 1·00-2·49; p=0·0103) and diclofenac (1·65, 0·95-2·85, p=0·0187), non-significantly by ibuprofen (1·90, 0·56-6·41; p=0·17), but not by naproxen (1·08, 0·48-2·47, p=0·80). The proportional effects on major vascular events were independent of baseline characteristics, including vascular risk. Heart failure risk was roughly doubled by all NSAIDs. All NSAID regimens increased upper gastrointestinal complications (coxibs 1·81, 1·17-2·81, p=0·0070; diclofenac 1·89, 1·16-3·09, p=0·0106; ibuprofen 3·97, 2·22-7·10, p<0·0001; and naproxen 4·22, 2·71-6·56, p<0·0001). INTERPRETATION: The vascular risks of high-dose diclofenac, and possibly ibuprofen, are comparable to coxibs, whereas high-dose naproxen is associated with less vascular risk than other NSAIDs. Although NSAIDs increase vascular and gastrointestinal risks, the size of these risks can be predicted, which could help guide clinical decision making. FUNDING: UK Medical Research Council and British Heart Foundation.

Considerations in the evaluation of surrogate endpoints in clinical trials. summary of a National Institutes of Health workshop.

We report on recommendations from a National Institutes of Health Workshop on methods for evaluating the use of surrogate endpoints in clinical trials, which was attended by experts in biostatistics and clinical trials from a broad array of disease areas. Recent advances in biosciences and technology have increased the ability to understand, measure, and model biological mechanisms; appropriate application of these advances in clinical research settings requires collaboration of quantitative and laboratory scientists. Biomarkers, new examples of which arise rapidly from new technologies, are used frequently in such areas as early detection of disease and identification of patients most likely to benefit from new therapies. There is also scientific interest in exploring whether, and under what conditions, biomarkers may substitute for clinical endpoints of phase III trials, although workshop participants agreed that these considerations apply primarily to situations where trials using clinical endpoints are not feasible. Evaluating candidate biomarkers in the exploratory phases of drug development and investigating surrogate endpoints in confirmatory trials require the establishment of a statistical and inferential framework. As a first step, participants reviewed methods for investigating the degree to which biomarkers can explain or predict the effect of treatments on clinical endpoints measured in clinical trials. They also suggested new approaches appropriate in settings where biomarkers reflect only indirectly the important processes on the causal path to clinical disease and where biomarker measurement errors are of concern. Participants emphasized the need for further research on development of such models, whether they are empirical in nature or attempt to describe mechanisms in mathematical terms. Of special interest were meta-analytic models for combining information from multiple studies involving interventions for the same condition. Recommendations also included considerations for design and conduct of trials and for assemblage of databases needed for such research. Finally, there was a strong recommendation for increased training of quantitative scientists in biologic research as well as in statistical methods and modeling to ensure that there will be an adequate workforce to meet future research needs.

Relation of gemfibrozil treatment and lipid levels with major coronary events: VA-HIT: a randomized controlled trial.

CONTEXT: A low plasma level of high-density lipoprotein cholesterol (HDL-C) is a major risk factor for coronary heart disease (CHD). A secondary prevention study, the Veterans Affairs High-Density Lipoprotein Intervention Trial (VA-HIT), demonstrated that CHD events were significantly reduced during a median follow-up of 5.1 years by treating patients with the fibric acid derivative gemfibrozil when the predominant lipid abnormality was low HDL-C. OBJECTIVE: To determine if the reduction in major CHD events with gemfibrozil in VA-HIT could be attributed to changes in major plasma lipid levels. DESIGN: Multicenter, randomized, double-blind, placebo-controlled trial conducted from September 1991 to August 1998. SETTING: The Department of Veterans Affairs Cooperative Studies Program, in which 20 VA medical centers were participating sites. PARTICIPANTS: A total of 2531 men with a history of CHD who had low HDL-C levels (mean, 32 mg/dL [0.83 mmol/L] ) and low low-density lipoprotein cholesterol (LDL-C) levels (mean, 111 mg/dL [2.88 mmol/L]). INTERVENTION: Participants were randomly assigned to receive gemfibrozil, 1200 mg/d (n = 1264), or matching placebo (n = 1267). MAIN OUTCOME MEASURE: Relation of lipid levels at baseline and averaged during the first 18 months of gemfibrozil treatment with the combined incidence of nonfatal myocardial infarction and CHD death. RESULTS: Concentrations of HDL-C were inversely related to CHD events. Multivariable Cox proportional hazards analysis showed that CHD events were reduced by 11% with gemfibrozil for every 5-mg/dL (0.13-mmol/L) increase in HDL-C (P =.02). Events were reduced even further with gemfibrozil beyond that explained by increases in HDL-C values, particularly in the second through fourth quintiles of HDL-C values during treatment. During gemfibrozil treatment, only the increase in HDL-C significantly predicted a lower risk of CHD events; by multivariable analysis, neither triglyceride nor LDL-C levels at baseline or during the trial predicted CHD events. CONCLUSIONS: Concentrations of HDL-C achieved with gemfibrozil treatment predicted a significant reduction in CHD events in patients with low HDL-C levels. However, the change in HDL-C levels only partially explained the beneficial effect of gemfibrozil.

Terms of reference for Data and Safety Monitoring Committees.

The Data and Safety Monitoring Committee (DSMC) constitutes one of the most complex elements of a clinical trial. For this reason, it is critical that all participants in and contributors to a trial understand the functions and responsibilities of the DSMC. Failure to understand the complexity of this group's mission can lead to substantial opportunities for confusion among investigators, sponsors, regulators, and the public, especially if the trial that it monitors must be modified or terminated early.

Efficacy of recombinant circumsporozoite protein vaccine regimens against experimental Plasmodium falciparum malaria.

After initial successful evaluation of the circumsporozoite-based vaccine RTS,S/SBAS2, developed by SmithKline Beecham Biologicals with the Walter Reed Army Institute of Research, protective efficacy of several regimens against Plasmodium falciparum challenge was determined. A controlled phase 1/2a study evaluated 1 or 2 standard doses of RTS,S/SBAS2 in 2 groups whose members received open-label therapy and 3 immunizations in blinded groups who received standard, one-half, or one-fifth doses. RTS,S/SBAS2 was safe and immunogenic in all groups. Of the 41 vaccinees and 23 control subjects who underwent sporozoite challenge, malaria developed in 7 of 10 who received 1 dose, in 7 of 14 who received 2 doses, in 3 of 6 who received 3 standard doses, in 3 of 7 who received 3 one-half doses, in 3 of 4 who received 3 one-fifth doses, and in 22 of 23 control subjects. Overall protective efficacy of RTS,S/SBAS2 was 41% (95% confidence interval, 22%-56%; P=.0006). This and previous studies have shown that 2 or 3 doses of RTS,S/SBAS2 protect against challenge with P. falciparum sporozoites.

An improved double sampling procedure based on the variance.

Sample size calculations for a continuous outcome require specification of the anticipated variance; inaccurate specification can result in an underpowered or overpowered study. For this reason, adaptive methods whereby sample size is recalculated using the variance of a subsample have become increasingly popular. The first proposal of this type (Stein, 1945, Annals of Mathematical Statistics 16, 243-258) used all of the data to estimate the mean difference but only the first stage data to estimate the variance. Stein's procedure is not commonly used because many people perceive it as ignoring relevant data. This is especially problematic when the first stage sample size is small, as would be the case if the anticipated total sample size were small. A more naive approach uses in the denominator of the final test statistic the variance estimate based on all of the data. Applying the Helmert transformation, we show why this naive approach underestimates the true variance and how to construct an unbiased estimate that uses all of the data. We prove that the type I error rate of our procedure cannot exceed alpha.

Long-term in vivo survival of receptor-modified syngeneic T cells in patients with human immunodeficiency virus infection.

To study human immunodeficiency virus (HIV)-specific cellular immunity in vivo, we transferred syngeneic lymphocytes after ex vivo expansion and transduction with a chimeric receptor gene (CD4/CD3-zeta) between identical twins discordant for HIV infection. Single and multiple infusions of 10(10) genetically modified CD8(+) T cells resulted in peak fractions in the circulation of approximately 10(4) to 10(5) modified cells/10(6) mononuclear cells at 24 to 48 hours, followed by 2- to 3-log declines by 8 weeks. In an effort to provide longer high-level persistence of the transferred cells and possibly enhance anti-HIV activity, we administered a second series of infusions in which both CD4(+ )and CD8(+) T cells were engineered to express the chimeric receptor and were costimulated ex vivo with beads coated with anti-CD3 and anti-CD28. Sustained fractions of approximately 10(3) to 10(4) modified cells/10(6) total CD4(+) or CD8(+) cells persisted for at least 1 year. Assessment of in vivo trafficking of the transferred cells by lymphoid tissue biopsies revealed the presence of modified cells in proportions equivalent to or below those in the circulation. The cell infusions were well tolerated and were not associated with substantive immunologic or virologic changes. Thus, adoptive transfer of genetically modified HIV-antigen-specific T cells was safe. Sustained survival in the circulation was achieved when modified CD4(+ )and CD8(+) T cells were infused together after ex vivo costimulation, indicating the important role played by antigen-specific CD4(+) T cells in providing "help" to cytotoxic effectors. (Blood. 2000;96:467-474)

Efficacy of vaccination with live attenuated, cold-adapted, trivalent, intranasal influenza virus vaccine against a variant (A/Sydney) not contained in the vaccine.

OBJECTIVE: To determine the safety, immunogenicity, and efficacy of revaccination of children with live attenuated influenza vaccine. STUDY DESIGN: A 2-year multicenter, double-blind, placebo-controlled, efficacy field trial of live attenuated, cold-adapted trivalent influenza vaccine administered by nasal spray to children. This report summarizes year 2 results, a year in which the epidemic strain of influenza A/Sydney was not well matched to the vaccine strains. Each year, vaccine strains were antigenically equivalent to the contemporary inactivated influenza vaccine. In year 2, a single intranasal revaccination was administered. Active surveillance for influenza was conducted during the influenza season by means of viral cultures. Influenza cases were defined as illnesses with wild-type influenza virus isolated from respiratory secretions. RESULTS: In year 2, 1358 (85%) children, 26 to 85 months of age, returned for revaccination. The intranasal vaccine was easily accepted, well tolerated, and immunogenic. Revaccination resulted in 82% to 100% of the vaccinated children in a subset studied for immunogenicity being seropositive as compared with 26% to 65% of placebo recipients, depending on the influenza strain tested. No serious adverse events were associated with the vaccine. In addition to the strains in the vaccine, antibody was induced to the variant strain A/Sydney/H3N2. In year 2, influenza A/Sydney/H3N2, a variant not contained in the vaccine, caused 66 of 70 cases of influenza A; nonetheless, intranasal vaccine was 86% efficacious in preventing A/Sydney influenza. Eight cases of lower respiratory tract disease were associated with A/Sydney influenza; all cases were in the placebo group. CONCLUSIONS: This live attenuated, cold-adapted influenza vaccine was safe, immunogenic, and efficacious against influenza A/H3N2 (including a variant, A/Sydney, not contained in the vaccine) and influenza B. The characteristics of this vaccine make it suitable for routine use in children to prevent influenza.

Internal pilot studies I: type I error rate of the naive t-test.

When sample size is recalculated using unblinded interim data, use of the usual t-test at the end of a study may lead to an elevated type I error rate. This paper describes a numerical quadrature investigation to calculate the true probability of rejection as a function of the time of the recalculation, the magnitude of the detectable treatment effect, and the ratio of the guessed to the true variance. We consider both 'restricted' designs, those that require final sample size at least as large as the originally calculated size, and 'unrestricted' designs, those that permit smaller final sample sizes than originally calculated. Our results indicate that the bias in the type I error rate is often negligible, especially in restricted designs. Some sets of parameters, however, induce non-trivial bias in the unrestricted design.

Internal pilot studies II: comparison of various procedures.

The two-stage design involves sample size recalculation using an interim variance estimate. Stein proposed the design in 1945; biostatisticians recently have shown renewed interest in it. Wittes and Brittain proposed a modification aimed at greater efficiency; Gould and Shih proposed a similar procedure, but with a different interim variance estimate based on blinded data. We compare the power of Stein's original test, an idealized version of the Wittes-Brittain test, and a theoretical optimal test which can be approximated in practice. We also compare two procedures that control the conditional type I error rate given the actual final sample size: Gould and Shih's procedure and a newly proposed 'second segment' procedure. The comparison among the first three procedures indicates that the Stein test is, unexpectedly, the test of choice under the original design alternative, whereas the approximate-optimal and Wittes-Brittain procedures appear to have superior power for detecting smaller treatment differences. As between the latter two procedures, the second segment procedure is more powerful when many observations are likely to be taken after the interim resizing, whereas otherwise the Gould-Shih procedure is superior.

Evaluation of trivalent, live, cold-adapted (CAIV-T) and inactivated (TIV) influenza vaccines in prevention of virus infection and illness following challenge of adults with wild-type influenza A (H1N1), A (H3N2), and B viruses.

Trivalent, live, cold-adapted influenza vaccine (CAIV-T) is highly effective in the prevention of influenza in children, and a variety of monovalent and bivalent cold-adapted influenza vaccines have been efficacious in adults. In order to determine the efficacy of CAIV-T in healthy adults, we administered CAIV-T, trivalent inactivated influenza vaccine (TIV) or placebo to 103 adults in randomized double-blind fashion, and then challenged those subjects who had pre-screening serum hemagglutination-inhibition antibody titers of 1:8 or less with wild-type influenza viruses corresponding to the strains contained in the vaccine. CAIV-T was well tolerated. Upon challenge with wild-type influenza virus, laboratory documented influenza illness (respiratory symptoms with either isolation of wild-type influenza virus from nasal secretions or 4-fold and/or greater HAI antibody response to challenge) occurred in 14/31 (45%) placebo recipients, 4/32 (13%) TIV recipients, and 2/29 (7%) CAIV-T recipients. The estimated protective efficacy of CAIV-T was therefore 85% and of TIV was 71%. These results are consistent with those of previous studies using monovalent preparations of cold-adapted influenza vaccine in this model, and indicate that CAIV-T will be an effective means to prevent influenza illness in adults.

The effect of spironolactone on morbidity and mortality in patients with severe heart failure. Randomized Aldactone Evaluation Study Investigators.

BACKGROUND AND METHODS: Aldosterone is important in the pathophysiology of heart failure. In a doubleblind study, we enrolled 1663 patients who had severe heart failure and a left ventricular ejection fraction of no more than 35 percent and who were being treated with an angiotensin-converting-enzyme inhibitor, a loop diuretic, and in most cases digoxin. A total of 822 patients were randomly assigned to receive 25 mg of spironolactone daily, and 841 to receive placebo. The primary end point was death from all causes. RESULTS: The trial was discontinued early, after a mean follow-up period of 24 months, because an interim analysis determined that spironolactone was efficacious. There were 386 deaths in the placebo group (46 percent) and 284 in the spironolactone group (35 percent; relative risk of death, 0.70; 95 percent confidence interval, 0.60 to 0.82; P<0.001). This 30 percent reduction in the risk of death among patients in the spironolactone group was attributed to a lower risk of both death from progressive heart failure and sudden death from cardiac causes. The frequency of hospitalization for worsening heart failure was 35 percent lower in the spironolactone group than in the placebo group (relative risk of hospitalization, 0.65; 95 percent confidence interval, 0.54 to 0.77; P<0.001). In addition, patients who received spironolactone had a significant improvement in the symptoms of heart failure, as assessed on the basis of the New York Heart Association functional class (P<0.001). Gynecomastia or breast pain was reported in 10 percent of men who were treated with spironolactone, as compared with 1 percent of men in the placebo group (P<0.001). The incidence of serious hyperkalemia was minimal in both groups of patients. CONCLUSIONS: Blockade of aldosterone receptors by spironolactone, in addition to standard therapy, substantially reduces the risk of both morbidity and death among patients with severe heart failure.

Gemfibrozil for the secondary prevention of coronary heart disease in men with low levels of high-density lipoprotein cholesterol. Veterans Affairs High-Density Lipoprotein Cholesterol Intervention Trial Study Group.

BACKGROUND: Although it is generally accepted that lowering elevated serum levels of low-density lipoprotein (LDL) cholesterol in patients with coronary heart disease is beneficial, there are few data to guide decisions about therapy for patients whose primary lipid abnormality is a low level of high-density lipoprotein (HDL) cholesterol. METHODS: We conducted a double-blind trial comparing gemfibrozil (1200 mg per day) with placebo in 2531 men with coronary heart disease, an HDL cholesterol level of 40 mg per deciliter (1.0 mmol per liter) or less, and an LDL cholesterol level of 140 mg per deciliter (3.6 mmol per liter) or less. The primary study outcome was nonfatal myocardial infarction or death from coronary causes. RESULTS: The median follow-up was 5.1 years. At one year, the mean HDL cholesterol level was 6 percent higher, the mean triglyceride level was 31 percent lower, and the mean total cholesterol level was 4 percent lower in the gemfibrozil group than in the placebo group. LDL cholesterol levels did not differ significantly between the groups. A primary event occurred in 275 of the 1267 patients assigned to placebo (21.7 percent) and in 219 of the 1264 patients assigned to gemfibrozil (17.3 percent). The overall reduction in the risk of an event was 4.4 percentage points, and the reduction in relative risk was 22 percent (95 percent confidence interval, 7 to 35 percent; P=0.006). We observed a 24 percent reduction in the combined outcome of death from coronary heart disease, nonfatal myocardial infarction, and stroke (P< 0.001). There were no significant differences in the rates of coronary revascularization, hospitalization for unstable angina, death from any cause, and cancer. CONCLUSIONS: Gemfibrozil therapy resulted in a significant reduction in the risk of major cardiovascular events in patients with coronary disease whose primary lipid abnormality was a low HDL cholesterol level. The findings suggest that the rate of coronary events is reduced by raising HDL cholesterol levels and lowering levels of triglycerides without lowering LDL cholesterol levels.

Effectiveness of live, attenuated intranasal influenza virus vaccine in healthy, working adults: a randomized controlled trial.

CONTEXT: Influenza virus is a major cause of illness, disruption to daily life, and increased use of health care in all age groups. OBJECTIVE: To assess the safety and effectiveness of intranasally administered trivalent, live, attenuated influenza virus (LAIV) vaccine for reducing illness, absenteeism, and health care use among healthy, working adults. DESIGN: Randomized, double-blind, placebo-controlled trial conducted from September 1997 through March 1998. SETTING: Thirteen centers across the United States. PARTICIPANTS: A total of 4561 healthy, working adults aged 18 to 64 years recruited through health insurance plans, at work sites, and from the general population. INTERVENTION: Participants were randomized 2:1 to receive intranasally administered trivalent LAIV vaccine (n = 3041) or placebo (n = 1520) in the fall of 1997. MAIN OUTCOME MEASURES: Episodes of febrile illness, severe febrile illness, febrile upper respiratory tract illness, work loss, and health care use during the peak and total influenza outbreak periods, and adverse events. RESULTS: Recipients of LAIV vaccine were as likely to experience 1 or more febrile illnesses as placebo recipients during peak outbreak periods (13.2% for vaccine vs 14.6% for placebo; P=.19). However, vaccination significantly reduced the numbers of severe febrile illnesses (18.8% reduction; 95% confidence interval [CI], 7.4%-28.8%) and febrile upper respiratory tract illnesses (23.6% reduction; 95% CI, 12.7%-33.2%). Vaccination also led to fewer days of illness across all illness syndromes (22.9% reduction for febrile illnesses; 27.3% reduction for severe febrile illnesses), fewer days of work lost (17.9% reduction for severe febrile illnesses; 28.4% reduction for febrile upper respiratory tract illnesses), and fewer days with health care provider visits (24.8% reduction for severe febrile illnesses; 40.9% reduction for febrile upper respiratory tract illnesses). Use of prescription antibiotics and over-the-counter medications was also reduced across all illness syndromes. Vaccine recipients were more likely to experience runny nose or sore throat during the first 7 days after vaccination, but serious adverse events between the groups were not significantly different. The match between the type A(H3N2) vaccine strain and the predominant circulating virus strain (A/Sydney/05/97[H3N2]) for the 1997-1998 season was poor, suggesting that LAIV provided substantial cross-protection against this variant influenza A virus strain. CONCLUSION: Intranasal trivalent LAIV vaccine was safe and effective in healthy, working adults in a year in which a drifted influenza A virus predominated.

Simultaneous screening for three correlated lipids in the VA HDL Intervention Trial.

We developed methodology to design the multistage lipid screen for the VA HDL Intervention Trial, a randomized double-blind placebo-controlled secondary prevention clinical trial of 2531 participants. The trial aimed to determine if HDL-raising therapy reduces coronary events in men with low HDL-cholesterol and desirable LDL-cholesterol. Joint lipid distributions for HDL-cholesterol, LDL-cholesterol, and triglycerides were derived on the basis of estimates from previous studies, and simulations were performed to determine the cutpoints for excluding screenees for the three lipid parameters to be used at each recruitment stage. Operating characteristics for different screening rules are presented. Comparisons between the predicted and actual study recruitment results show good agreement in lipid characteristics and underscore the complexity of simultaneously screening on correlated continuous physiologic parameters such as lipids.

Selecting screening criteria for clinical trials: an example from the Systolic Hypertension in the Elderly Program.

In clinical trials that study people with a continuous measure defined categorically, even repeated measurements within a visit and over successive visits do not prevent error-free classification. We describe the design of a screening procedure for the Systolic Hypertension in the Elderly Program (SHEP), a randomized clinical trial designed to test whether regular administration of antihypertensive medication reduces the risk of stroke in elderly persons with isolated systolic hypertension. Data from a pilot study performed before the inauguration of SHEP allowed empirical study of a variety of possible screening rules for SHEP. A desirable screening rule would require only two screening visits, would lead to a randomized cohort with high mean systolic blood pressure, and would not impede recruitment. We emphasize two classes of rules, "serial" and "conditional." A serial rule uses only the values observed at a given screen to determine eligibility to proceed to the next screen. A conditional rule uses the value observed at a given screen along with values already observed to determine eligibility to proceed. For the SHEP study, we chose a conditional rule for screening because of its efficiency in identifying eligible participants. Our approach to selection of screening rules should be applicable to other clinical trials in which the measurement that defines the primary entry criterion has considerable measurement error.