Missing data and censoring in the analysis of progression-free survival in oncology clinical trials.

Progression-free survival (PFS) is increasingly used as a primary endpoint in oncology clinical trials. However, trial conduct is often such that PFS data on some patients may be partially missing either due to incomplete follow-up for progression, or due to data that may be collected but confounded by patients stopping randomized therapy or starting alternative therapy prior to progression. Regulatory guidance on how to handle these patients in the analysis and whether to censor these patients differs between agencies. We present results of a reanalysis of 28 Phase III trials from 12 companies or institutions performed by the Pharmaceutical Research and Manufacturers Association-sponsored PFS Expert Team. We show that analyses not adhering to the intention-to-treat principle tend to give hazard ratio estimates further from unity and describe several factors associated with this shift. We present illustrative simulations to support these findings and provide recommendations for the analysis of PFS.

Attenuation of treatment effect due to measurement variability in assessment of progression-free survival.

For normally distributed data analyzed with linear models, it is well known that measurement error on an independent variable leads to attenuation of the effect of the independent variable on the dependent variable. However, for time-to-event variables such as progression-free survival (PFS), the effect of the measurement variability in the underlying measurements defining the event is less well understood. We conducted a simulation study to evaluate the impact of measurement variability in tumor assessment on the treatment effect hazard ratio for PFS and on the median PFS time, for different tumor assessment frequencies. Our results show that scan measurement variability can cause attenuation of the treatment effect (i.e. the hazard ratio is closer to one) and that the extent of attenuation may be increased with more frequent scan assessments. This attenuation leads to inflation of the type II error. Therefore, scan measurement variability should be minimized as far as possible in order to reveal a treatment effect that is closest to the truth. In disease settings where the measurement variability is shown to be large, consideration may be given to inflating the sample size of the study to maintain statistical power.

Blinded independent central review of progression in cancer clinical trials: results from a meta-analysis.

PURPOSE: Progression free survival (PFS) is increasingly used as a primary end-point in oncology clinical trials. This paper provides observations and recommendations on the use of a blinded independent central review (BICR) for progression. PATIENTS AND METHODS: The findings and recommendations are based on extensive simulations and a meta-analysis based on 27 previously conducted randomised phase III trials with BICR performed by the Pharmaceutical Research and Manufacturers Association (PhRMA) sponsored PFS Independent Review Working Group. RESULTS: Results of the meta-analysis demonstrate a strong correlation between LE and BICR estimates of treatment effect (R=0.947). Further, differences between treatment groups in discordance rates predict the differences between LE and BICR estimates of treatment effect supporting the use of a sample-based BICR on a subgroup of patients. CONCLUSION: The meta-analysis demonstrates that local evaluation (LE) provides a reliable estimate of the treatment effect with little evidence for systematic evaluation bias. Therefore, when a trial is blinded or a large effect on PFS is observed a BICR may not be warranted. When a BICR is warranted, a sample-based BICR may provide a reduction in operational complexity without compromising the credibility of trial results. While for large trials that are not adequately blinded a sample-based BICR may be recommended. A full BICR should be considered in the case of smaller trials or in situations in which there is a particular need to increase the confidence in the LE results.

Research outcomes and recommendations for the assessment of progression in cancer clinical trials from a PhRMA working group.

PURPOSE: Progression free survival (PFS) is increasingly used as a primary end-point in oncology clinical trials. This paper provides recommendations for optimal trial design, conduct and analysis in situations where PFS has the potential to be an acceptable end-point for regulatory approval. PATIENTS AND METHODS: These recommendations are based on research performed by the Pharmaceutical Research and Manufacturers Association (PhRMA) sponsored PFS Working Group, including the re-analysis of 28 randomised Phase III trials from 12 companies/institutions. RESULTS: (1) In the assessment of PFS, there is a critical distinction between measurement error that results from random variation, which by itself tends to attenuate treatment effect, versus bias which increases the probability of a false negative or false positive finding. Investigator bias can be detected by auditing a random sample of patients by blinded, independent, central review (BICR). (2) ITT analyses generally resulted in smaller treatment effects (HRs closer to 1) than analyses that censor patients for potentially informative events (such as starting other anti-cancer therapy). (3) Interval censored analyses (ICA) are more robust to time-evaluation bias than the log-rank test. CONCLUSION: A sample based BICR audit may be employed in open or partially blinded trials and should not be required in true double-blind trials. Patients should be followed until progression even if they have discontinued treatment to be consistent with the ITT principle. ICAs should be a standard sensitivity analysis to assess time-evaluation bias. Implementation of these recommendations would standardize and in many cases simplify phase III oncology clinical trials that use a PFS primary end-point.

Impact of diabetes mellitus on the severity of erectile dysfunction and response to treatment: analysis of data from tadalafil clinical trials.

AIMS/HYPOTHESIS: A retrospective analysis of pooled data from twelve placebo-controlled trials was conducted to characterise the efficacy and safety of tadalafil for the treatment of erectile dysfunction in men with diabetes compared with that in men without diabetes. METHODS: Patients were randomly allocated to tadalafil 10 mg, 20 mg, or placebo, taken as needed for 12 weeks. The study population comprised 637 men with diabetes (mean age 57 years) and 1681 men without diabetes (mean age 56 years). RESULTS: At baseline, patients with diabetes had more severe erectile dysfunction than patients without diabetes, with mean International Index of Erectile Function (IIEF) erectile function domain scores of 12.6 and 15.0 respectively (p<0.001). Compared with placebo, tadalafil 10 mg and 20 mg improved all primary efficacy outcomes in both patient groups (p<0.001). Men with diabetes receiving tadalafil 20 mg experienced a mean improvement of 7.4 in their IIEF erectile function domain score against baseline versus 0.9 for placebo (p<0.001). This group reported on average that 53% of their attempts at intercourse were successful, compared with 22% for placebo (p<0.001 for the change from baseline). Baseline IIEF erectile function domain scores correlated inversely with baseline HbA(1)c levels. The responses to tadalafil were similar regardless of levels of baseline glycaemic control, diabetic therapy received, or previous use of sildenafil. CONCLUSIONS/INTERPRETATION: Despite more severe baseline erectile dysfunction in men with diabetes, tadalafil was efficacious and well tolerated in this population. As reported for other phosphodiesterase 5 inhibitors, the response to tadalafil was slightly lower in men with diabetes than in men without diabetes.

The efficacy and safety of tadalafil: an update.

OBJECTIVE: To provide an update on the efficacy and safety of tadalafil, a phosphodiesterase-5 inhibitor, in the treatment of erectile dysfunction (ED). PATIENTS AND METHODS: In all, 2102 men (mean age 56 years) with mild-to-severe ED of various causes were randomized to placebo or tadalafil, taken as needed with no food restrictions, at fixed 'on-demand' doses of 10 or 20 mg in 11 randomized, double-blind, placebo-controlled trials lasting 12 weeks. The three co-primary outcomes were changes from baseline in the erectile function domain of the International Index of Erectile Function (IIEF) and the proportion of 'yes' responses to questions 2 and 3 of the Sexual Encounter Profile (SEP). Additional efficacy instruments included a Global Assessment Question (GAQ). RESULTS: Compared with placebo, tadalafil gave significantly better outcomes. Patients receiving either dose of tadalafil had a significant mean improvement of 6.5 and 8.6, respectively, in the IIEF erectile function domain score from baseline (P < 0.001 vs placebo). At both doses the mean success rate for intercourse attempts (SEP-Q3) was 58% and 68%, respectively, compared with 31% in the placebo group (P < 0.001), and 71% and 84% reported improved erections at the endpoint (GAQ), vs 33% on placebo (P < 0.001). Tadalafil was effective up to 36 h after dosing and was effective regardless of disease severity and causes, and in patients of all ages. The most frequent adverse events were headache, dyspepsia, back pain and myalgia. CONCLUSION: Tadalafil was an effective and well-tolerated treatment for ED.

Long-term safety and tolerability of tadalafil in the treatment of erectile dysfunction.

OBJECTIVE: To assess the long-term safety and tolerability of tadalafil for patients with erectile dysfunction (ED). PATIENTS AND METHODS: This was a multicentre, open-label, 24-month extension trial involving 1173 men with ED. The mean age was 57 (range 23-83) years and 74.8% of patients were taking concomitant medications for comorbid conditions, including diabetes mellitus in 30.5% of men and hypertension in 29.5%. These patients had participated in 1 of 5 previous 8-week or 12-week randomised, double-blind, placebo-controlled tadalafil studies. In the present trial, the starting 10mg dose of tadalafil could be increased to 20mg if the patient could not achieve satisfactory intercourse or reduced to 5mg for an adverse event that was persistent, intolerable and judged by the investigator to be related to tadalafil. RESULTS: Four hundred ninety-three (42.0%) men completed 24 months of treatment. In addition, a further 234 (19.9%) completed 18 months of treatment due to a sponsor decision to reduce the study duration. The total tadalafil exposure was 1676.0 patient-years. Tadalafil was safe and well tolerated. Headache (15.8%), dyspepsia (11.8%), nasopharyngitis (11.4%), and back pain (8.2%) were the most common treatment-emergent adverse events. The rate of discontinuations due to adverse events for this 18-24-month study was 6.3% and the rate for any individual event was <1%. Serious adverse events occurred in 8.6% of patients. No consistent pattern of serious adverse events assessed as causally associated with tadalafil administration was observed. None of the four deaths that occurred during the study was assessed as tadalafil related. There were no clinically significant laboratory or electrocardiographic findings or changes in vital signs in mean baseline-to-endpoint analysis attributable to tadalafil. Tadalafil administration was not causally associated with drug-induced hepatotoxicity, neutropenia, thrombocytopenia, or renal dysfunction. CONCLUSION: Tadalafil at doses of 5, 10, or 20mg taken as needed up to once daily for 18 to 24 months was safe and well tolerated. These findings support the long-term use of tadalafil in the clinical management of erectile dysfunction.

Monitoring a clinical trial with multiple hypotheses concerning the treatment effect on a single primary endpoint.

In a number of clinical trials there is interest in testing more than one hypothesis concerning the treatment effect on a single primary endpoint. For instance, a sponsor of a trial to demonstrate that a test treatment (T) is non-inferior to an active control (R) may also be interested in showing that T is superior to R, if this is the case. Using the closed testing method for constructing tests of multiple hypotheses which control the multiple level of significance, we provide a framework for testing these hypotheses sequentially during a trial at pre-planned interim analyses.

Sample size recalculation using conditional power.

The sample size required to achieve a given power at a prespecified absolute difference in mean response may depend on one or more nuisance parameters, which are usually unknown. Proposed methods for using an internal pilot to recalculate the sample size using estimates of these parameters have been well studied. Most of these methods ignore the fact that data on the parameter of interest from within this internal pilot will contribute towards the value of the final test statistic. We propose a method which involves recalculating the target sample size by computing the number of further observations required to maintain the probability of rejecting the null hypothesis at the end of the study under the prespecified absolute difference in mean response conditional on the data observed so far. We do this within the framework of a two-group error-spending sequential test, modified so as to prevent inflation of the type I error rate.

Estimation following extension of a study on the basis of conditional power.

Proschan and Hunsberger (1) propose a method based on conditional power for designed extension of a study beyond its originally intended sample size. Their data-dependent sampling method can be viewed as a two-stage procedure in which the target total sample size is dependent upon the data observed at the first stage. We demonstrate that the maximum likelihood estimate of the parameter of interest upon completion may be biased, and that this bias is similar in direction and magnitude to that commonly associated with estimation following a group sequential test with predetermined target total sample size. Furthermore, we show how a bias adjusted estimate may be formed.

Estimating the sample size for a t-test using an internal pilot.

If the sample size for a t-test is calculated on the basis of a prior estimate of the variance then the power of the test at the treatment difference of interest is not robust to misspecification of the variance. We propose a t-test for a two-treatment comparison based on Stein's two-stage test which involves the use of an internal pilot to estimate variance and thus the final sample size required. We evaluate our procedure's performance and show that it controls the type I and II error rates more closely than existing methods for the same problem. We also propose a rule for choosing the size of the internal pilot, and show that this is reasonable in terms of the efficiency of the procedure.

The determinants of iron status in early pregnancy.

In pregnancy, the additional demands for Fe are thought to be met principally through increased maternal dietary Fe absorption and by mobilization of maternal Fe stores. In a general population sample of 576 women we examined the maternal and dietary characteristics which influenced Fe stores (assessed by serum ferritin concentration) in early pregnancy. The effects of these characteristics on two measures of functional Fe status (mean cell volume and haemoglobin concentration) were also considered. Serum ferritin concentrations were lower in multiparous women (P < 0.0001) and in those with a lower BMI (P = 0.01), and rose with increasing alcohol intake (P < 0.0001). Ferritin concentrations fell with increasing Ca intake (P < 0.0001); the proportion of women with serum ferritin values < or = 12 micrograms/l rose from 14% of the women in the lowest quarter of Ca intake to 29% of the women in the highest quarter. Mean cell volume and haemoglobin concentration were not related to Ca intake in early pregnancy. Although Ca added to test-meals reduces Fe absorption, long-term Ca supplementation has not been shown to lower plasma ferritin concentration, suggesting that high habitual Ca intakes would be unlikely to influence Fe status in non-pregnant individuals. Our findings show that in early pregnancy there is an association between high dietary Ca intake and lower Fe stores. This effect of Ca on one aspect of Fe status may result from its influence on Fe bioavailability.

A survey of patient preference for insulin jet injectors versus needle and syringe.

Many studies have tested the insulin absorption rate and mechanical reliability of jet injectors. However, no published papers have dealt exclusively with patient preference for this method of administering insulin over a period of years. This paper reports the results of a survey done to determine if use of a jet injector for delivering insulin has an effect on acceptance of and adherence to a regimen of multiple doses of insulin. Over the past 15 years, the authors have instructed approximately 70 patients to use various jet injectors. A questionnaire was developed and sent to 75 patients. Of these, 42 completed and returned the questionnaire. Survey results indicate that even though some problems with the injectors were identified, 70% of those responding still preferred to take insulin by jet injector.