One of the exciting developments in IO research has been the investment by industry in prospective clinical trials to help define best practice in the care of our patients. Nordion, Merit/Biosphere, Sirtex, BTG/Biocompatibles, Siemens, and Delcath all sponsor multicenter Phase 2 and 3 trials in the US and globally. (If I left anyone out, please add a comment below!) This represents a major transformation in behavior for the device industry (see The Pot and the Kettle, Sept. 4, 2012).
Such high quality science remains the exception rather than the rule. Recently, one of our corporate partners reached out to WCIO leaders individually to do single institution, investigator initiated/investigator IDE studies with their product. This might sound like a good thing -- more industry-sponsored research, right? Wrong!
Clinical research in oncology follows a standardized structure to establish safety, effectiveness, and comparative efficacy. Phase 1 safety trials typically require 10-20 patients. Phase 2 effectiveness trials with sufficient power to enable statistical design of subsequent phase 3 randomized trials require 40+ patients. Definitive randomized controlled trials require hundreds of patients.
So what is the problem with investigator-initiated single institution prospective trials? They are OK if all you are doing is a Phase 1 safety study. Once you move on to a Phase 2 efficacy study, it takes a single institution years years to accrue the required 40 patients. That is too long in the life cycle of the device industry, and investigators are too quick to publish early rather than completing a robust cohort.
Single institution reports are the bane of the IO literature. In the past year, JVIR and CVIR published 70 clinical IO papers. 84% were single institution reports, 76% were retrospective, and 57% had <40 patients. Among the papers reporting more than 40 patients, 20 were from Japan, Korea, or China; only seven were from Europe; and a pitiful three from the US. Only four papers used a prospective, Phase1/2 design. Seventeen publications were purported efficacy reports on <15 patients -- mostly from leading western cancer centers! These underpowered series do nothing to advance interventional oncology as the fourth pillar of cancer care. No oncology journal would publish such garbage.
A powerful solution is staring us in the face. If we band together to do multicenter Phase 2 trials, then we can accrue rapidly, and the results will be stronger for being generalized across centers. This comes at a cost in terms of complexity, audits and core labs to ensure standardization, and liability for the investigator holding the IDE. Yet the end product is surely worth it -- a credible trial that can influence oncologic care. Three centers doing separate investigator IDE studies yield three worthless papers; the same three centers working together advance the field of IO. I believe that our industry colleagues recognize the value of multicenter trials, and would bear the added expense if we have the skill and the will to deliver.
So are there any good IO articles? Sure, just not in IR journals. There is no way to keep up with all the oncologic and related journals. IO Central and IO Insights do the work for you. Our abstraction service pulls in all the news and publications related to key IO terms. Subscribe to IO Insights, so you can read the good, not just the bad and the ugly.