News articles and press releases that focus on this year’s annual ASCO meeting include a plethora of predictions and analyses regarding immuno-oncology agents, including immune checkpoint inhibitors (ICIs) and therapeutic vaccines. Phrases like “new paradigm in cancer treatment,” and adjectives including “prominent”, “hot”, and “promising” pepper these headlines and commentaries. What exactly are immuno-oncology agents? How are they different from traditional immunotherapies, such as interleukin-2 and interferon? Why are they generating so much attention?
In a 2-hour educational session titled “Immunology for Nonimmunologists,” chairman Jedd D. Wolchok, MD, PhD, of Memorial Sloan Kettering Cancer Center, and his colleagues helped oncologists understand the immunologic principles that are being exploited by these new cancer immunotherapies. Panel members reviewed the historical development of cancer immunotherapy principles, described basic mechanisms of innate and adaptive immunity, and summarized clinical data for ICIs, therapeutic cancer vaccines (antigens, adjuvants, dendritic cells), and adoptive T-cell therapies.
In 1891, William Coley, a surgeon from New York, administered intratumoral injections of live or inactivated bacteria with the goal of generating spontaneous remissions in patients with sarcoma. These “Coley’s toxins” stimulated antibacterial phagocytes and, if the patient was lucky, killed bystander tumor cells. Although successes were recorded over the ensuing 40 years, they were sporadic and difficult to reproduce.1
Today, more than 120 years later, the efficacy (and complexity) of immunomodulatory strategies for cancer patients has improved exponentially. To ground the audience in key principles of immunotherapy, Dr Lawrence Fong of the University of California, San Francisco, introduced three basic themes that other immuno-oncology experts then explicated:
1. A given immunotherapy agent can be effective in many cancers, including tumor types with different origins and drivers. This is true because the immune system targets antigens (foreign or toxic cells, including cancer cells) to generate an immune response; they do not target cancer cell signaling pathways.
Dr Wolchok and other panelists enthusiastically explained that, in light of more sophisticated knowledge of the body’s response to cancer, use of immunotherapy strategies is no longer limited to “traditional” immuno-responsive tumors, renal cell cancer and melanoma. Clinical trials of immune modulators are demonstrating durable efficacy in many other hard-to-treat solid tumors, including lung cancer, triple negative breast cancer, prostate cancer, bladder cancer, and head and neck cancer.
To elucidate the mechanism by which immunotherapies engender tumor responses, Dr Fong outlined research demonstrating that various cancer cell mutations and their resulting proteins represent a pool of antigens that the immune system then recognizes and destroys. Specifically, tumor infiltrating lymphocytes (TILs) act in response to tumor-specific mutation proteins. He noted that various studies clearly demonstrate a correlation between improved clinical outcomes and the presence of TILs in tumors at the time of cancer diagnosis.
In light of the relationship between TILs and clinical outcomes, Sylvia Adams, MD of New York University explained that, upon cancer diagnosis, patients’ tumor samples can be assessed for the presence of TIL infiltration (ie, “TIL-rich” tumors, “non-infiltrated” tumors), in addition to standard parameters, such as histology and genetics. Like these more traditional tumor parameters, the presence of TILs can be used to stratify and assess patients in clinical trials of immunotherapies, and may serve as a clinical biomarker in the future.
Oncology journals, as well as panelists’ personal experiences, are now replete with examples of advanced cancer patients who continue to respond to immunotherapies, such as ipilimumab and interleukin-2 (IL-2), years after initiating treatment. In these cases, Dr Fong explained that the immune system develops “memory”, which is known as an adaptive immune response. When the same antigens are presented in the future, the body’s T cells recall them and respond to remove them. Dr Wolchok reinforced this unique aspect of immunotherapy treatment, quoting data presented in 2013 that demonstrated a 3-year OS rate of 22% and a 7-year OS rate of 17% with ipilimumab based on a pooled analysis of more than 1,800 patients with metastatic melanoma. These survival statistics were unaffected by patients’ prior treatments, the ipilimumab dose (3 mg/kg or 10 mg/kg), or receipt of maintenance therapy.2
To explain the durability of responses to immuno-oncology agents, Dr Fong explained that both T cell- and B cell-mediated immune responses can broaden beyond the targeted antigens. Immune cells can recognize and “adapt to” new antigens that arise within the tumor, a phenomenon known as “antigen spreading”. As such, patients’ responses to immunotherapy remain durable despite tumor cell mutations. A favorable side benefit of this long immune “memory” is the potential for administration of immunotherapy agents either one time or over a short timeframe. Most non-immunotherapy cancer-directed agents, including kinase inhibitors and radiation, require longer-term or chronic administration.
By its nature, chemotherapy agents directly target tumor cells. In contrast, immunotherapy agents “educate” patients’ immune systems (T cells) to target the tumor. Dr Fong stated that this inherent distinction among strategies has several implications for health care professionals who treat cancer:
Because immunotherapy acts to reduce tumor burden in completely different ways, Cassian Yee, MD of the University of Texas MD Anderson Cancer Center noted that use of immunotherapy does not inherently preclude concurrent or sequential use of chemotherapy, cell signaling inhibitors, monoclonal antibodies, or radiation therapy. Indeed, such multi-pronged approaches may represent the future of chronic tumor control. In his overview of tumor-targeting monoclonal antibodies and antibody drug conjugates, Louis Weiner, MD of Georgetown University provided provocative information regarding the ability of antibodies, such as trastuzumab and ado-trastuzumab emtansine, to alter tumor burden through at least two mechanisms: direct cytotoxicity and generation of an immune response.
While enthusiastic about future combination approaches, Dr Wolchok and his colleagues offered words of caution regarding use of multiple immunotherapies without clinical evidence of safety. He recalled studies in melanoma (ie, vemurafanib combined with ipilimumab, sunitinib with tremelimumab, an anti-CTLA-4 inhibitor), that resulted in severe, dose-limiting immune-related adverse events (irAEs). Dr Wolchok suggested, “We have to be careful about these intersections between targeted therapies and immunotherapies. The message that we put out after our experience with vemurafanib [combined with ipilimumab] was, ‘Just because two medicines are approved does not mean that they are safe to use together.’” Dr Wolchok also noted that immune-related AEs with single-agent ipilimumab and other checkpoint inhibitors can be problematic for patients, including rash, colitis, enteritis, elevated liver enzymes, and others. While many of these AEs are manageable and short-lived, endocrinopathies (such as adrenal insufficiency) can result in the need for long-term hormone replacement and other complications. Patients are best served by considering enrollment in one of the many clinical trials that evaluate immunotherapies and immunotherapy-based combinations in a controlled fashion.
Acknowledging the need for prudence, immunotherapy experts in this review session consistently predicted that multiple immune modulating strategies are needed “for immunotherapy to achieve its full potential”. Therapeutic cancer vaccines, antibodies that inhibit immune checkpoints, and adoptive T cell therapies are believed to work in concert, such that their combination represents the future of cancer management. Multiple abstracts and presentations during this year’s ASCO meeting will elucidate the latest findings and their implications for oncologists and their patients with cancer.
Glioblastoma multiforme (GBM), the most common primary brain tumor in adults, is highly aggressive.3 Median survival after maximal surgical resection, radiation, and temozolomide approaches 15 months in clinical trials.4 Bevacizumab is approved in the US for patients with recurrent GBM on the basis of response rate data, but no data are available to demonstrate durable improvement of disease-related symptoms or an overall survival (OS) benefit with its use.5 The need for effective alternatives in GBM remains dramatically high.
To determine whether immune checkpoint inhibitors might be clinically relevant in GBM, Berghoff and colleagues (ASCO 2014: Abstract 2011) assessed PD-1 and PD-L1 expression in 135 GBM specimens of 117 patients. The median age in this patient population was 60. Their median overall survival (OS) was 12 months (range: 0 to 86). In 18 of these 117 patients, tissue from both the initial tumor resection and the first local recurrence were available. Analyses of PD-1, PD-L1, CD3, and CD8 expression were performed by immunohistochemistry (IHC). Analysis of O6-methylguanine DNA methyltransferase (MGMT) promoter methylation was performed using pyrosequencing with an 8% cut-off.
Berghoff et al identified sparse to moderate density tumor-infiltrating lymphocytes (TILs) in 74% of the 135 GBM tissue samples (CD3+, 68%; CD8+, 47%). In 15% of specimens, PD-1 was expressed on scattered TILs in the perivascular compartment and within tumor tissue. In 86% of specimens, PD-L1 was expressed on tumor cells and macrophages/microglial cells throughout the tumor tissue. In almost half (45%) of the GBM samples, PD-L1 staining was observed in more than 50% of viable tumor tissue.
When assessing correlates of outcomes in these patients with GBM, Berghoff and colleagues discovered that younger age, better performance status, and MGMT hypermethylation were significantly related to favorable OS. In contrast, neither TIL density nor expression of PD-1 and PD-L1 associated with survival outcomes.
Although PD-1 and PD-L1 were immunohistochemically detectable in the GBM samples, the researchers did not observe clear up- or downregulation of PD-L1 or PD-1 TIL density. Their data suggest that the PD-1 axis may be one of several immunosuppressive pathways that glioma cells employ in order to avoid immune responses. They concluded that clinical study of immune checkpoint inhibitors (ICIs) appear to be warranted in patients with GBM, a discovery that undoubtedly represents welcome news for patients with GBM and their families who continue to hope for effective and safe new treatment options.
On the basis of findings like these, as well as preclinical studies that demonstrate benefits of combining PD-1 inhibitors with radiation in a mouse glioma model, Sampson and colleagues (ASCO 2014: Abstract TPS2101; Clinical trial identifier NCT02017717) described an ongoing phase 2b, randomized, open-label study that compares the efficacy and safety of nivolumab (NIV), a fully human IgG4 PD-1 antibody, given alone and with ipilimumab (IPI), with bevacizumab in patients with recurrent GBM. Patients whose first recurrence of GBM occurred within 28 days of randomization are eligible for this trial. Upon establishing safe and tolerable dosing for the NIV + IPI combination in GBM, the study will randomize a maximum of 240 patients to receive NIV (3 mg/kg x 4 doses given every 2 weeks), NIV + IPI, or bevacizumab (10 mg/kg every 2 weeks). Overall response rate (ORR), progression-free survival (PFS), and overall survival (OS) are the primary efficacy measures. The estimated completion date of this study is January 2018.6 As context, ORRs of 19.6% and 25.9% have been reported for bevacizumab in two separate phase 2 trials in patients with recurrent GBM.5
Renal cell carcinoma (RCC) is a heterogeneous disease with multiple histological variants and genetic changes. Three histological subtypes account for the majority of renal cancers: clear cell RCC (75% to 90%), papillary RCC (10% to 15%), and chromophobe RCC (4% to 5%).7 Because most clinical trials of novel agents, including tyrosine kinase inhibitors (TKIs), have prioritized testing in clear cell RCC (ccRCC) and excluded patients with non-clear cell RCC (non-ccRCC), optimal treatment of the latter patient subset remains unclear.8
PD-L1 expression has been observed in most ccRCC and is associated with poorer cancer-specific survival.9 However, PD-L1 expression and its relationship to clinical outcomes in non-ccRCC and benign kidney tumors (BKT) are unknown. Fay and colleagues (ASCO 2014: Abstract 4526) evaluated tumor specimens from 121 patients with non-ccRCC, including chromophobe RCC, papillary RCC, translocation Xp11.2 RCC, collecting duct carcinoma, oncocytoma, and angiomyolipoma. Using IHC, a mouse monoclonal anti-PD-L1 antibody, and ?5% tumor cell membrane staining as their definition of PD-L1 positivity, these researchers learned that PD-L1 expression varied in non-ccRCC and BKT and depended on histology.
Among the tumor samples from 121 non-ccRCC patients, 15% were PD-L1 positive in tumor cells and 59% were PD-L1 positive in immune cells. Table 1 summarizes IHC results by RCC histology.
Table 1: PD-L1 Expression in Non-ccRCC by Histology
PD-L1 expression in tumor cells was associated with higher disease stage (p=0.01) and grade (p=0.03), as well as lower OS (p<0.001). In patients with non-ccRCC whose immune cells expressed PD-L1, there was no correlation between PD-L1 positivity and disease stage or grade, and a non-significant trend toward lower overall survival (OS) was observed (p=0.08). Finally, PD-L1 expression in both tumor cells and immune cells was correlated with lower time to recurrence (TTR) (p=0.02 and p=0.03, respectively).
Because PD-L1 expression is detectable in histologic subsets and appears indicative of worse prognosis for these patients, future clinical trials of ICIs in RCC should not exclude patients with non-ccRCC.
