Leisha A. EmensEmail author,
Lisa H. Butterfield,
F. Stephen Hodi Jr.,
Francesco M. Marincola and
Howard L. Kaufman
Journal for ImmunoTherapy of Cancer20164:42
DOI: 10.1186/s40425-016-0146-9
© The Author(s). 2016
Received: 15 April 2016
Accepted: 6 July 2016
Published: 19 July 2016
Cancer immunotherapy has a long history in the clinic. William Coley was an orthopedic surgeon in the late 1800’s who noted marked tumor regression in a patient with severe erysipelas. Based on this observation, he hypothesized that stimulating the immune system could effectively treat cancer. Thus, over the next decades he gave various bacterial products, including live bacteria, to patients with advanced cancers [1]. Although toxicity was often substantial, he and his colleagues noted tumor regressions in some patients, particularly those with bone and soft tissue sarcomas. His work was groundbreaking in strategy and method, but eventually fell out of favor due to lack of scientific control and the advent of radiotherapy and chemotherapy for cancer treatment. Despite this, he was a pioneering physician-scientist. His work is notable because his hypothesis that the immune system can be harnessed to treat cancer is the bedrock of the striking success of cancer immunotherapy today.
https://jitc.biomedcentral.com/articles ... 016-0146-9Clinical endpoints
Advances in tumor immunotherapy have been dramatic over the last five years. We now understand that the kinetics of immune-mediated anti-tumor activity may be delayed compared to that of cytotoxic chemotherapy or targeted therapy. Immunotherapy mediates tumor regression indirectly through activation of immune responses and/or inhibition of suppressive immune elements. This may result in delayed tumor regression, with some patients even experiencing the progression of existing disease or the appearance of new lesions prior to eventual disease regression. Furthermore, many immunotherapy agents do not impact progression-free survival, but are associated with significant improvements in overall survival [20]. Standard metrics for tumor response assessment thus may not capture the true clinical benefit of cancer immunotherapies, and the selection of clinical endpoints may require re-thinking when immunotherapy agents are being used as part of a cancer treatment regimen.
The use of alternative clinical endpoints for immunotherapy has already been widely discussed. First, the atypical response patterns of cancer immunotherapy led to the development of the immune-related response criteria (irRECIST), which captures disease in bidimensional measurements and allows for continued treatment in the face of clinically insignificant disease progression [21]. Immune-related RECIST (irRECIST) was subsequently simplified to modified RECIST criteria (mRECIST) capture disease in unidimensional measurements while still allowing for the unique response patterns of cancer immunotherapies with apparent disease progression by standard criteria that is clinically insignificant (delayed response and/or pseudoprogression) [22]. A recent report from the pembrolizumab Phase 1b study demonstrated a significant rate of atypical responses, where standard RECIST criteria appeared to underestimate the benefit of pembrolizumab in about 15% of patients [23]. Thus, accounting for the atypical response patterns of cancer immunotherapy might prevent the cessation of treatment in patients who will go on to benefit from cancer immunotherapy. Second, the durable response rate is a novel clinical endpoint already successfully incorporated into the registration strategy for T-VEC. It captures response rates over a pre-specified time frame based on anticipated time to disease progression, which may differ across various cancer types. Importantly, the strong association between stable disease and improved overall survival increasingly reported for cancer immunotherapy further supports disease control rate as a highly clinically relevant endpoint. Disease control rate may be a particularly appropriate endpoint in early studies of novel single agent and combination cancer immunotherapies [24]. A caveat to the use of disease control rate as an endpoint for clinical trials is that it may be confounded by indolent disease biology: in the absence of a randomized control arm, it could overestimate clinical benefit. One way to mitigate this risk is the use of an imbalanced randomization such as that used in the pivotal trial leading to the approval of T-VEC. Third, the unusually high rate of complete responses seen in some recent combination immunotherapy trials suggests that complete response could emerge as an appropriate clinical endpoint for selected cancer immunotherapy strategies [25]. Finally, composite endpoints that capture both the depth and duration of response may be most appropriate for some cancer immunotherapies. Efforts to develop novel endpoints appropriate to the mechanism of action of the particular immunotherapy strategy being tested are ongoing