Assessing the impact of restricted follow-up and small sample sizes on survival estimations in prostate cancer using registry data.

2016 ◽  
Vol 34 (7_suppl) ◽  
pp. 294-294 ◽  
Author(s):  
Dhvani Shah ◽  
Victoria Paly ◽  
Andrew Briggs ◽  
Manpreet Sidhu ◽  
Esprit Ma ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Clinical Trials ◽  
Small Sample ◽  
Sample Sizes ◽  
Reference Case ◽  
Term Survival ◽  
Small Sample Sizes ◽  
Log Normal

294 Background: Economic evaluations in oncology aim to assess the value of new therapies in the long term based on clinical trial data that often have restricted follow-up times (< 5 years) and small sample sizes (< 500 patients). This requires the use of extrapolation assumptions on long-term survival that go beyond the observed data. In this analysis, differences in survival extrapolation methods are tested in samples of sizes and follow-up reflecting typical clinical trials against a background of known survival in prostate cancer from a US based cancer registry. Methods: Data from the National Cancer Institute's Surveillance Epidemiology and End Results (SEER) registry on long-term survival in patients with stage IV prostate cancer were employed. The data set comprised those patients diagnosed between 1988 and 2003, with follow-up data available until 2012. Additional survival for those who received surgery (compared to those who did not), was estimated based on extrapolations using standard parametric statistical models (exponential, Weibull, log-logistic, log-normal, Gamma) fitted to the observed data. Survival analyses were run for 5 sample size scenarios (n = 27,670, 1000, 500, 200, 50) and 6 follow-up scenarios (follow-up years = 25, 20, 10, 5, 2, 1) yielding 30 combination scenarios. Performance of the methods was tested relative to the maximum follow-up, maximum sample size scenario (i.e. reference case) from the SEER registry. Results: Log-logistic and log-normal models were associated with flat tails which led to inflated survival estimations. For scenarios with smaller sizes, gamma models often did not converge. Exponential models were the most frequently reported as best model fit (in approximately 50% of scenarios). Also, gains in OS were consistent when exponential models were selected, and closely matched gain in OS from the reference case. Conclusions: Since clinical trials in oncology are often associated with small patient sample sizes and restricted follow-up, selecting an exponential model may lead to the most consistent and stable results based on the experiment constructed here. Further research should confirm these results for other types of cancer.

Independence Estimating Equations for Controlled Clinical Trials with Small Sample Sizes

2006 ◽  
Vol 45 (04) ◽  
pp. 430-434 ◽  
Author(s):  
G. Dahmen ◽  
A. Ziegler
Keyword(s):  
Clinical Trials ◽  
Estimating Equations ◽  
Small Sample ◽  
Repeated Measurements ◽  
Robust Estimator ◽  
Variance Estimator ◽  
Sample Sizes ◽  
Controlled Clinical Trials ◽  
Parallel Group ◽  
Small Sample Sizes

Summary Objectives: The application of independence estimating equations (IEE) for controlled clinical trials (CCTs) has recently been discussed, and recommendations for its use have been derived for testing hypotheses. The robust estimator of variance has been shown to be liberal for small sample sizes. Therefore a series of modifications has been proposed. In this paper we systematically compare confidence intervals (CIs) proposed in the literature for situations that are common in CCTs. Methods: Using Monte-Carlo simulation studies, we compared the coverage probabilities of CIs and non-convergence probabilities for the parameters of the mean structure for small samples using modifications of the variance estimator proposed by Mancl and de Rouen [7], Morel et al. [8] and Pan [3]. Results: None of the proposed modifications behave well in each investigated situation. For parallel group designs with repeated measurements and binary response the method proposed by Pan maintains the nominal level. We observed non-convergence of the IEE algorithm in up to 10% of the replicates depending on response probabilities in the treatment groups. For comparing slopes with continuous responses, the approach of Morel et al. can be recommended. Conclusions: Results of non-convergence probabilities show that IEE should not be used in parallel group designs with binary endpoints and response probabilities close to 0 or 1. Modifications of the robust variance estimator should be used for sample sizes up to 100 clusters for CI estimation.

Continuous circular cycling as a predictor of treatment response in bipolar disorders: a comprehensive review of the current literature

CNS Spectrums ◽  
2017 ◽  
Vol 23 (1) ◽  
pp. 24-28 ◽  
Author(s):  
Antonio Tundo ◽  
Paola Cavalieri
Keyword(s):  
Bipolar Disorders ◽  
Poor Response ◽  
Term Treatment ◽  
Small Sample ◽  
Long Term Treatment ◽  
Group Members ◽  
Short And Long Term ◽  
Small Sample Sizes

Evidence from the literature suggests that, on average, 27% of patients with a bipolar disorder (BD) experience a continuous cycling course (CCC) and that this subgroup differs significantly from patients with a noncontinuous cycling course (N-CCC) with respect to sociodemographic characteristics and clinical presentation. The aim of the present paper is to review the studies that evaluated short- and long-term treatment responses in BD patients with CCC. The retrieved studies indicate that CCC is a significant predictor of poor response to long-term treatment with lithium (the odds of a response in the CCC group were 57% less than in the N-CCC group; p<0.01), as well as to polytherapies including lithium and/or an antiepileptic augmented, when necessary, with an antipsychotic and/or antidepressant. The percentage of patients without new episodes during follow-up was significantly lower in the CCC group compared with the N-CCC group (15.4 vs. 37.6% , p<0.01). Compared with patients in the N-CCC group, members of the CCC group had a poorer response and lower remission rates after 12-week antidepressant treatments for a major depressive episode (82.3 vs. 50%, p =0.002; 69.6 vs. 40.9%, p=0.013). These findings, underlining that CCC is a predictor of poor response to short- and long-term treatment in BD, should be interpreted considering the limitations of the reviewed studies (the small sample sizes, the small number of trials and their observational nature, the lack of randomization or placebo controls, and the unblinded nature of the outcomes). Clinical trials and observational studies with larger samples are warranted to confirm the conclusions of our review.

Small sample sizes in clinical trials: a statistician’s perspective

2012 ◽  
Vol 2 (7) ◽  
pp. 655-657 ◽  
Author(s):  
Lucinda Billingham ◽  
Kinga Malottki ◽  
Neil Steven
Keyword(s):  
Clinical Trials ◽  
Small Sample ◽  
Sample Sizes ◽  
Small Sample Sizes

Ten Year-Long Term Survival After up-Front Autologous Stem Cell Transplantation in Multiple Myeloma: Results From Two Prospective Clinical Trials

Blood ◽  
2012 ◽  
Vol 120 (21) ◽  
pp. 594-594
Author(s):  
Elena Zamagni ◽  
Francesco Di Raimondo ◽  
Francesca Patriarca ◽  
Patrizia Tosi ◽  
Annalisa Pezzi ◽  
...  
Keyword(s):  
Multiple Myeloma ◽  
Clinical Trials ◽  
Regression Analysis ◽  
Stem Cell ◽  
High Quality ◽  
Term Survival ◽  
Long Term Survival ◽  
Long Term Survivors

Abstract Abstract 594 Survival of patients with multiple myeloma (MM) has been extended with the introduction of autologous stem cell transplantation (ASCT). More recently, availability of highly effective novel agents has further improved patient outcomes. However, it is still the matter of debate whether a proportion of patients treated with ASCT can enjoy a long term survival, while sustaining prolonged high quality response. To address this issue and to identify those variables which were related to long-term survival, we performed a post-hoc analysis of two large prospective clinical trials of ASCT in newly diagnosed MM patients, the first one comparing single versus double ASCT and the second one incorporating thalidomide-dexamethasone (TD) into double ASCT. A total of 321 patients were randomly assigned in the first study to receive either a single or double ASCT, as previously described (Cavo M et al, JCO 2007). Three hundred and fifty seven patients were enrolled in the subsequent multicenter phase 2 study incorporating TD from the outset until the second ASCT; details of the protocol were previously reported (Cavo et al, J. Clin. Oncol 2009). Results were updated as of 30 March 2012 and compared with those previously reported. All the analyses were performed on an intention-to-treat basis. After a median follow-up of 61 months for the entire treatment population of the first study, PFS remained significantly longer with tandem versus single ASCT (median 37 vs 25 months, P= 0.012), while OS was similar in the two groups (median 71 vs 67 months). 47% and 33% of the patients in the double and single ASCT group achieved a CR+nCR (P= 0.008). Overall, in 24% and 11% of the patients, CR+nCR was sustained for more than 5 and 10 years, respectively. In a multivariate Cox regression analysis, best response (CR+nCR) ever achieved was the most important variable significantly extending PFS (P= 0.003) and OS (P=0.050); random assignment to double ASCT was an additional variable predicting for prolonged PFS(P= 0.026). After a median follow-up of 84 months from starting TD in the second study, median values of PFS and OS were 47.2 and 109.6 months, respectively. The final rate of CR+nCR was 34%, which was maintained for a median of 53 months. Overall, in 42.1% and 9.1% of the patients CR+nCR was sustained for more than 5 and 8 years, respectively. On multivariate analysis, failure to ever achieve at least CR+nCR, low Hb, high β2-m and t(4;14)±del(17p) were found to be independent variables predicting for poorer outcomes. In particular, a shorter OS was seen for patients ever lacking high-quality responses (HR: 0.35, 0.23–0.54, p<0.0001) and with t(4;14)±del(17p) (HR: 0.51, 0.33–0.79, p=0.0030). Overall, 23% and 20% of patients in the first and second study were alive over 10 or 8 years, respectively (long-term survivors). Median PFS of long-term survivors in the 2 studies were 74 and 87.7 months, respectively, versus 25 and 37 months for the rest of the population (P= 0.0000). Median duration of CR+nCR were 70 and 78 months in the long-term survivors group for the first and second study, respectively, in comparison with 21 and 49 months in the remaining patients (P<0.001 for both). The 10 and 8-year estimates of OS after relapse or progression in the long-term survivors of the two protocols were 58% and 72%, respectively, in comparison to a median value of 24 and 23 months for the control group (p<0.0001 for both). In a logistic regression analysis, attainment of high-quality responses was independently associated with long-term survival in both the studies (first study: OR: 1.8, 1.06–3.01, P= 0.03; second study: OR: 4.3, 2.17–8.60, P= 0.000). In conclusion, although the comparison between TD incorporated into ASCT and ASCT without thalidomide was not directly addressed by this analysis, TD + ASCT was associated with extended PFS and OS. Approximately 20% of the patients undergoing up-front ASCT can achieve long term survival (8–10 years from start of treatment), with 33% of them remaining relapse free. Attainment of sustained high-quality responses was the leading independent variable predicting for long-term OS. Prolonged survival after relapse was a contributing factor to long-term OS. Disclosures: Off Label Use: One of the 2 protocols discussed includes the use of thalidomide as induction prior to ASCT.

Sign in / Sign up

Export Citation Format

Share Document