Reply to Potential underestimation of cerebrovascular events in the PROVENGE Registry for the Observation, Collection, and Evaluation of Experience Data

2935 Cancer June 15, 2020 Registry for the Observation, Collection, and Evaluation of Experience Data (PROCEED) registry. This prospective registry was required by the US Food and Drug Administration to evaluate a potential cerebrovascular event (CVE) signal observed in the sipuleucel-T clinical program through the use of prespecified methodology with which to assess CVEs in a larger population. With respect to their concern about study methodology, men could be enrolled either prospectively before receiving sipuleucel-T or “retrospectively,” which was defined as “within 6 months after the first leukapheresis.” Retrospective enrollment was allowed mainly to enhance timely accrual to the registry by allowing patients who had initiated sipuleucel-T but were not enrolled in the registry to be included. In all, 711 men (27%) were retrospectively and 1191 men (73%) were prospectively enrolled. Of those who enrolled retrospectively, approximately onehalf enrolled within 28 days after the first leukapheresis, 75% enrolled within 69 days, and only 1 patient enrolled 199 days after the first infusion. Understanding this distribution should lessen concern that patients who enrolled retrospectively would have passed the usual time frame in which CVEs have been noted (ie, 6-7 months after infusion of sipuleucel-T). Furthermore, CVE rates per 100 person-years were found to be similar for those enrolled prospectively or retrospectively at 1.293 (95% CI, 0.8101.958) and 1.376 (95% CI, 0.974-1.889), respectively. Another concern raised by Dores et al was the definition of CVE beyond that described in our article. Unlike previous studies, CVEs in PROCEED were defined prospectively in the protocol as was the adjudication process: “...CVEs will include all completed strokes, both ischemic and hemorrhagic in etiology, intracranial hemorrhage and transient ischemic attacks (TIA)...” The adjudicator applied similar, prospectively defined criteria: major, nonfatal CVE, nonfatal stroke (acute, focal neurologic event that persisted for >24 hours ideally confirmed by imaging) with evaluation of ischemic stroke subtypes as per TOAST criteria and primary intracranial hemorrhage as per Stroke Data Bank criteria; and nonmajor CVEs including TIAs without acute infarction and fatal CVEs. Hence, embolic strokes were included; however, bleeding due to head trauma or metastatic lesions and neurological symptoms due to anoxia or hypoxia were not considered to be CVEs. The third concern relates to the impact of the adjudication process on the reported CVE rate. The adjudication algorithm was prespecified and all CVEs and serious adverse events that might have been a CVE were reviewed. The adjudicator received relevant source documentation and could request additional details to be able to identify The opinions and information in this article are those of the authors, and do not represent the views and/or policies of the US Food and Drug Administration.