2 Clarke Drive
Suite 100
Cranbury, NJ 08512
© 2024 MJH Life Sciences™ and OncLive - Clinical Oncology News, Cancer Expert Insights. All rights reserved.
With the proliferation of agents for cancer over the past few years, including both targeted and immunotherapies, the number of clinical questions surround- ing optimal sequencing and combinations approaches infinity.
Jason J. Luke, MD, FACP
The past year has seen an explosion in the development of immune-oncology (IO) treatments for cancer. In solid tumors, immune-checkpoint blocking antibodies are being explored in many tumors while recent reports from the American Society of Hematology meeting suggest that immune-checkpoint blockade and adoptive cellular transfer will also have an important role in hematologic malignancies. These treatments are for the most part much less toxic than historical combination chemotherapy regimens and ad- ditionally appear to have the potential for long-term durable response. Clearly, IO has a significant and expanding role in cancer treatment.
Despite the recent amazing success and hype surrounding IO, immunotherapy for cancer is not at all a new concept. Cancer vaccines and cytokines have been of interest for decades. Unfortunately, the limited success observed with these approaches made them relevant only to a small group of patients—predominantly melanoma, renal cell carcinoma, and patients who had progressed through standard chemotherapies. The one significant exception to this is the field of hematologic bone marrow transplantation where the utility of the approach has been predicated on inducing an anti-tumor effect by the transplanted immune system.
Moving beyond transplant however, it appears that we are on the cusp of an age of IO therapy drug development in all cancers. In solid tumors, immune-checkpoint blocking antibodies, such as anti-CTLA4 (ipilimumab), anti-PD1 (nivolumab, pembrolizumab) and anti-PD-L1 (MPDL3280a) have or are close to becoming the standard of care for melanoma, bladder, and lung cancers and promising results have been observed in several other tumors types. In fact, multiple anti-PD1 or anti-PD-L1 agents have been granted breakthrough status and are likely to come into clinical practice within the next few months.
Similarly in hematologic malignancies, especially Hodgkin’s lymphoma, previously unknown clinical activity has been observed in refractory patient populations.1 Immune-checkpoint blocking antibodies are only part of the story, however, as other immunotherapies have also shown impressive results. Examples include bispecific antibodies, such as blinatumomab, and antibody drug conjugates (ADCs), such as brentuximab vedotin, demonstrating marked activity in chemotherapy and stem cell transplant refractory patient populations. In solid tumors, antibody drug conjugates are also changing the standard of care with T-DM1 becoming an important therapy in breast cancer, and other molecules such as glenbatumumab in melanoma, suggesting a potentially important role in the future. Finally, adoptive cellular therapy with chimeric antigen receptor (CAR) T-cells has received breakthrough status for acute lymphocytic leukemia and CARs for other hematologic malignancies as well as some solid tumors are rapidly being developed.
Clearly then, IO therapy is revolutionizing cancer therapeutics and may be positioned to become the underlying backbone of therapy in all cancers. A limitation of further development, however, is that a substantial proportion of cancer patients still do not respond to these agents and much work remains to determine both why this is the case and what measures may be taken to facilitate a more robust immune response in these patients.
Integral to understanding both of these issues will be the development of reliable biomarkers of anti-tumor immune activation and treatment response. With immunotherapies such as ADCs and CAR T-cells, this is less of an issue as the biomarker target is an essential element to the drug design. That being said, the number of common, tumor specific (or significantly overexpressed) antigens that can facilitate the development of these treatments is likely limited. In contrast, biomarkers of effect for immune-checkpoint blocking antibodies have been difficult to develop to date. There was initial optimism that expression of PD-L1 in the tumor microenvironment may be a useful biomarker. Unfortunately, the current generation of assay technology certainly does not support this, however, given that approximately 10% to 20% of patients whose tumors test “PD-L1 negative” have been shown to experience Response Evaluation Criteria In Solid Tumors (RECIST) responses.
Given this, better predictive and prognostic biomarkers are needed. Recent studies have suggested that perhaps a more individual tumor specific biomarker development strategy may be useful. A recent elegant report detailed that nearly all patients responding to the anti-PD1 antibody pembrolizumab had expression of PD-1/PD-L1 in a population of T cells existing specifically at the tumor margin.2 Therefore, PD-L1 testing would need to take into account the anatomy of the tumor and biopsy techniques would need to be adapted as such. Alternatively, a model of understanding the immune response in the context of the tumor microenvironment has been proposed in which tumors can be determined as having an “inflamed” or “noninflamed tumor phenotype.” This phenotype can be predicted by gene expression profiling of a 13-gene signature from a tumor biopsy and has been very highly correlated with response to various immunotherapies from vaccines, cytokines, and immune-checkpoint blocking antibodies.3
Returning to the role of IO in cancer therapy—perhaps it’s time to consider flipping our standard paradigm of adding new drugs, such as IO agents, to existing treatments and instead consider whether the existing treatments can augment the immune response. An emerging field of preclinical research suggests that in fact this may be the best way to use the drugs we have already. Studies detailing the utility of radiation in augmenting immunotherapy4 or BRAF inhibitor driving a T cell infiltrate into the tumor5 suggest that there is an intersection between our existing treatments and the immune response. Other examples including VEGF-directed agents modifying immunosuppressive cell populations6 and “immunogenic” chemotherapies priming an immune response.7
With the development of these exciting new categories of IO therapies, the future seems bright for the development of vastly more efficacious treatment strategies that are much less toxic. Those treatment regimens will likely include a little of the old (chemotherapy, radiation, etc) and a little of the new (immune-checkpoint blockade, CAR T-cells, etc). To reach this future, however, the oncology community will have to consider new models for drug development. The days of conducting clinical trials by adding new drug X to old drug Y just because we have them should be abandoned. Instead, we need scientifically justified and rationale clinical trials. An example of such a trial might be 1 week versus 1 month of drug X, in combination with immunotherapy drug Y, to determine the maximal immune effect achieved. With the proliferation of agents for cancer over the past few years, including both targeted and immunotherapies, the number of clinical questions surrounding optimal sequencing and combinations approaches infinity. Therefore, we need a future of smart, biomarker driven trials to guide the optimal treatment of our patients.
Related Content: