The Biostatistics Core provides statistical expertise for the planning, conduct, analysis, and reporting of clinical trials, epidemiologic and population based studies, experiments in the biology of cancer, including genomic experiments and laboratory and animal studies, and translational research studies with combined measurements on clinical and biological parameters. This includes consultation on all clinical protocols, and education in the areas of study design, electronic data capture, case report form design, and statistical methods.
One research project that illustrates the collaborations of core biostatisticians and computational biologists is Veno-occlusive disease of the liver after stem cell transplantation. Core statisticians Kristen Stevenson and Haesook Kim collaborated with Corey Cutler, MD, MPH, FRCPC (DFCI), Paul Richardson, MBBS (HMS), and colleagues to conduct a large retrospective analysis investigating the role of sirolimus in its potential association with veno-occlusive disease (VOD) of the liver after myeloablative allogeneic stem cell transplantation. This investigation retrospectively reviewed the records of 488 patients who underwent myeloablative transplantation between January 2000 and May 2007 and examined the influence of sirolimus on VOD incidence. After performing descriptive analysis including Kaplan-Meier and cumulative incidence curves of transplant related mortality and relapse, Stevenson and colleagues used multivariable exact logistic regression models to examine sirolimus use as a potential risk factor of VOD, Cox regression models to investigate the impact of sirolimus use on overall survival and progression-free survival, and competing risks regression analysis to investigate the impact of sirolimus use on transplant-related mortality in the presence of a competing risk, relapse.
They found that the incidence of VOD was significantly higher in the sirolimus group compared to the non-sirolimus group (15.8 percent vs. 7.4 percent; unadjusted OR=2.35; 95 percent CI 1.28-4.30, P<0.001; adjusted OR=2.14; 95 percent CI 1.12-4.09; p=0.02). However, this interestingly did not affect the cumulative incidence of transplant related mortality (100 day TRM 12 percent in sirolimus vs 18 percent in non-sirolimus, p=0.38) nor did it reduce overall survival (3-yr OS 51 percent in sirolimus and 47 percent in non-sirolimus, p=0.29). Their findings suggest that sirolimus use is associated with VOD after myeloablative transplantation; however, this did not translate into inferior survival nor increased risk of transplant related mortality. These findings were submitted for publication to Blood.