Although checkpoint blocking immunotherapy has significantly improved outcomes for patients with advanced and metastatic disease, researchers believe the modality also may be effective in the neoadjuvant setting.
Suzanne L.Topalian, MD
The neoadjuvant use of checkpoint blockade immunotherapy is showing promise in early clinical findings, opening up a potential new frontier for the game-changing modality to make an impact in earlier treatment settings in multiple solid tumor types.
Patients who received PD-1 or PD-L1 inhibitors before surgery or chemotherapy demonstrated positive outcomes in phase II studies in Merkel cell carcinoma (MCC), non—small-cell lung cancer (NSCLC), muscle-invasive bladder cancer (MIBC), and triple-negative breast cancer (TNBC), according to results reported at the 2018 American Society of Clinical Oncology (ASCO) Annual Meeting.
The rationale for earlier use of these agents stems from analyses of T-cell activity after the administration of the immune checkpoint blocking agents. Whereas conventional neoadjuvant chemotherapy is intended to shrink the tumor before surgery, immunotherapy serves as a primer for systemic antitumor responses, activating tumor-specific T cells that seek out distant micrometastases.1
Suzanne L. Topalian, MD, a pioneer in the development of anti—PD-1/PD-L1 immunotherapy, discussed the potential for the checkpoint blocking drugs in the neoadjuvant setting recently during the Noreen O’Neill Melanoma Research Symposium at the Wistar Institute in Philadelphia, Pennsylvania.
Melanoma and other skin cancers are serving as a proving ground for immunotherapy concepts, noted Topalian, who is associate director, Johns Hopkins Bloomberg—Kimmel Institute for Cancer Immunotherapy in Baltimore, Maryland. In MCC, a rare and aggressive skin cancer, investigators showed that neoadjuvant nivolumab (Opdivo) administered as 2 doses over approximately 4 weeks before surgery generated a pathologic complete response (pCR) or a major pathological response (MPR), defined as ≤10% residual viable tumor cells in the surgical specimen, in 65% of 17 patients whose samples were centrally reviewed.2
The strategy is being explored in other types of skin cancers, including melanoma. “The future development of anti—PD-1 in skin cancers is going to focus on earlier disease settings such as neoadjuvant therapy, adjuvant therapy, and on difficult-to-treat patient populations such as immunosuppressed patients,” Topalian said.
Topalian believes that neoadjuvant studies in melanoma will help pave the way for advancements in other tumor types, the pattern that has emerged over the past decade with checkpoint blocking agents. Melanoma “has always been the first tumor that people have looked at to develop immunotherapies for cancer. We realized a long time ago that melanoma was particularly responsive to immune therapies,” she said in an interview with OncologyLive®.
After decades of research, investigators believe that “melanomas are very immune-responsive because they contain a lot of mutations, and they contain a lot of mutations because many of them are caused by a carcinogen, which is ultraviolet light,” Topalian said. “Those mutations create new proteins that have never been seen by the immune system before. Because they are not normal, the immune system can recognize these mutant proteins very vigorously. This is one of the working theories for why melanoma is so immune-responsive, similar to lung cancer and bladder cancer. These are all carcinogeninduced cancers that have many mutations.”
Among the hundreds of ongoing clinical trials exploring anticancer immunotherapy regimens, many early studies are focusing on the neoadjuvant setting in a range of tumor types. These include several phase II and phase III studies (Table). Meanwhile, early-stage findings are being reported, including research presented at ASCO 2018.
More data from larger populations will be needed to draw firm conclusions about the efficacy of neoadjuvant checkpoint blocking immunotherapy, experts say. Future developments in new settings and with new combinations, Topalian noted, will increasingly focus on biomarkers. The FDA has approved the use of tumor PD-L1 expression level and microsatellite instability (DNA mismatch repair deficiency) as biomarkers to recommend patients for anti-PD-1/PD-L1 therapy in several settings and is considering tumor mutation burden as another marker.
“There’s been a lot of research activity around biomarkers and also around combination therapies,” she said. “Those 2 things are really linked, because the biomarker work can identify molecules associated with treatment resistance that we might want to direct drugs against in combination with anti—PD1 or anti–PD-L1 therapy.”The CheckMate-358 study represents the first trial of anti—PD-1 therapy in a neoadjuvant (presurgical) setting for patients with unresectable MCC, Topalian said in presenting the data at ASCO 2018. In 2017, avelumab (Bavencio), which targets PD-L1, gained the FDA’s approval for patients with unresectable metastatic MCC.
In moving immunotherapy forward in the treatment timeline, investigators were looking for signals of whether this approach would prevent the development of advanced, metastatic disease, which occurs in over 40% of patients with MCC, Topalian said at the Wistar symposium.
“Relapse comes from microscopic deposits of tumor in distant organs that we cannot see on [computed tomography] scans at the time that we are planning surgery. The idea here is to give anti—PD1 therapy before surgery so that, while the tumor is still in place and is infiltrated by the T cells, you are blocking PD-1 and allowing those T cells to become active and to proliferate. Then they leave the tumor, get into the bloodstream or the lymphatic system, enter the circulation, and disseminate systemically, where they can find sites of micrometastases and destroy them. This is the theory behind it.”
The study enrolled 29 patients with an average age of 69 years (range, 22-88) who had mostly stage III (69%) treatment-naïve MCC. The Merkel cell polyomavirus (MCPyV), which has been implicated in up to 80% of MCC cases, was observed in 71.4% of evaluable patients (15 of 21), and 30% (6 of 20) evaluable participants had tumor cell PD-L1 expression ≥1%. Overall, 26 patients received both doses of nivolumab and 27 proceeded to surgery.
The study met its primary endpoint of safety and tolerability. No new safety signals were observed and there were no serious treatment-related adverse events (AEs). All-grade AEs were experienced by 41.4% of patients, with fatigue (10.3%) and nausea (10.3%) as the most frequent reactions. There were 2 grade 3-4 AEs (rash and increased lipase).
The objective response rate (ORR) was high, Topalian said, with 40% of evaluable patients experiencing complete or partial tumor regression by RECIST. After a median 12 months’ follow-up, the progression-free survival rate was 72.6% among patients who had undergone surgery and none of those who had achieved pCR or MPR by site and/or central review had relapsed. At 24 months’ follow-up, the overall survival (OS) rate was 75.0% and the median had not yet been reached—a better outcome than previously observed in stage III patients.
Biomarker analyses are continuing, with investigators particularly interested in determining the role of MCPyV and PD-L1 status in outcomes. “We became interested in this cancer as part of an exploration of virus-associated cancers in general. Viruses are completely foreign to the immune system, and we thought virus-associated cancers would be particularly immunogenic and good targets for anti—PD-1 therapy,” Topalian said.
Patients responded regardless of their status of these markers in their tumors, but there were other interesting immunological findings, Topalian indicated. Although MCPyV-positive tumors have “an exceedingly low” mutational burden, the clonality of the T cells infiltrating these tumors is high, meaning that the few viral antigens that are being recognized can be the “right” ones to mediate tumor rejection. By contrast, MCPyV-negative tumors have a much higher mutational burden than MCPyV-positive MCC because they are caused by ultraviolet light, and they generate so many mutant tumor antigens that T cell responses are more diverse.In NSCLC, investigators from the Spanish Lung Cancer Group are evaluating the combination of nivolumab, paclitaxel, and carboplatin as neoadjuvant therapy in patients with stage IIIA resectable NSCLC in the phase II NADIM trial.3 Patients are receiving 3 cycles of nivolumab (360 mg) plus paclitaxel (200 mg/m2) and carboplatin (area under the curve 6) administered approximately 3 to 4 weeks before surgery, followed by adjuvant nivolumab for up to 1 year.
At ASCO 2018, Provencio- Pulla et al reported interim findings for 22 patients who received neoadjuvant therapy and continued on to surgery, with most undergoing a lobectomy (86.5%). The ORR reached 78% among these patients, consisting of 13 with a pCR and 4 with an MPR. Additionally, 5 patients experienced a partial response (22%).
In their poster, investigators said the study is the first multicentric test of the combination of immunotherapy and chemotherapy to show positive antitumor activity in the neoadjuvant NSCLC setting , and that the results thus far demonstrate a higher pCR rate than previously seen in resectable stage III disease.
Similarly, researchers have reacted positively to findings from a NSCLC study that Drew Pardoll, MD, PhD, and colleagues presented at this year’s American Association for Cancer Research Annual Meeting in April and published simultaneously in the New England Journal of Medicine.1,4 The pilot study was conducted in 21 patients with stage II or IIIA NSCLC who were eligible for surgical resection. After 2 preoperative doses of nivolumab (3 mg/kg), 9 of 20 resected tumors (45%) exhibited an MPR. At 18 months, the recurrence- free survival rate was 73% and the median had not yet been reached.
“Our result of a 45% major pathologic response rate is very encouraging, considering prior studies showing that a major pathologic response after neoadjuvant chemotherapy in lung cancer is associated with long-term survival,” said Pardoll, director of the Bloomberg—Kimmel Institute for Cancer Immunotherapy and director of Cancer Immunology at Johns Hopkins School of Medicine, in a statement.
By analyzing responses of T cell in blood samples at the time nivolumab therapy was initiated, and again 44 days after surgery, investigators found increased tumor-specific T cells in circulation.
“There was a big burst of tumor-specific T cells in the blood within, in most cases, 4 weeks of initiation of anti—PD-1 treatment, suggesting that neoadjuvant treatment may have enhanced antitumor immunity systemically,” Pardoll said. “While it’s still too early to tell whether our findings will translate into lower relapse rate and improved survival, pending confirmation in a larger study, we are very optimistic that this approach will eventually be practice-changing and may augment or even replace chemotherapy prior to surgical resection.”In MIBC, investigators evaluated atezolizumab (Tecentriq) as perioperative therapy for operable bladder cancer in patients with stage T2-4a N0 M0 disease in the phase II ABACUS study. Typically, eligible patients would be offered cisplatin chemotherapy before cystectomy.
That standard approach yields a pCR of 30% to 40% and a 5% improvement in OS, said Thomas Powles, MBBS, MRCP, MD, in presenting the study findings at 2018 ASCO.5 He is a professor of genitourinary oncology and lead for solid tumor research at Barts Cancer Institute, and director of Barts Cancer Centre at St Bartholomew’s Hospital in London. However, Powles said, many patients are not eligible for neoadjuvant cisplatin and are treated with cystectomy alone. Those who undergo transurethral resection of the bladder tumor (TURBT) without neoadjuvant therapy have a 15% pCR rate at cystectomy.
ABACUS enrolled patients who had transitional histology and residual disease after TURBT, and who were not fit for or had rejected cisplatinbased chemotherapy. Participants received 2 cycles of atezolizumab (1200 mg) after TURBT and waited a maximum of 8 weeks before proceeding to cystectomy.
In all, 74 patients received atezolizumab during the trial; 59 had 2 cycles and 15 had 1. Among these participants, 67 went on to receive a cystectomy; 7 patients did not proceed with the surgery because of treatment refusal (n = 1), clinical deterioration (n = 3), death (n = 2; pneumonia and myocardial infarction), or disease progression (n = 1).
The evaluable population (n = 68) was composed of those who underwent cystectomy and 1 patient who chose to be treated with chemotherapy instead. In the entire population, the pCR rate was 29% (95% CI, 19%-42%). The pCR rates were higher among those who were PD-L1— positive, defined as ≥5% immune component. Of the 56 patients who had PD-L1 analysis, the pCR rate was 40% (95% CI, 21%-61%) among those who tested positive (10 of 25 patients) versus 16% (85% CI, 5%-34%) for those who were PD-L1– negative (5 of 31). Moreover, there was a statistically significant increase (P <.05) in the mean CD8 count, a co-primary endpoint of the study, among 36 patients with paired samples available for testing at baseline and after cystectomy.
The phase II PURE-01 study explored the efficacy of therapy with the PD-1 inhibitor pembrolizumab (Keytruda) therapy before radical cystectomy in patients with cT ≤3B N0 stage, predominantly urothelial MIBC, and residual disease after TURBT. Participants, who were able to enroll regardless of cisplatin eligibility, received 3 cycles of 200-mg pembrolizumab.
In interim findings reported at ASCO 2018, the pCR rate among patients treated during the study (n = 43), the primary endpoint, was 39.5% (95% CI, 26.3%-54.4%).6 Pathologic downstaging to pT<2, the secondary endpoint, occurred in 51.2% of patients.
Additionally, investigators performed biomarker testing on tissue samples using a 22-gene “T-cell inflamed” panel designed to detail immune activity. They found that the pCR rates were higher for biomarker subgroups: 50.0% for those with a combined positive score (CPS) ≥20% on PD-L1 expression testing (11 of 22), 60.0% for those with DDR and/or RB1 genomic aberrations (15 of 25), and 90% for those with PD-L1 CPS ≥20% and DDR/RB1 alterations (9 of 10). Investigators said these markers may lead to an algorithm for a bladder- sparing approach after response assessment.
In commenting upon the results of the 2 studies, Matthew D. Galsky, MD, said the findings “were incredibly intriguing” but that many questions remain unanswered, particularly about the implications about pCR as a surrogate endpoint for survival in clinical trials versus in individual patients.
“The PURE and ABACUS trials undeniably establish single-agent activity for pembrolizumab and atezolizumab in the neoadjuvant setting, but if we’re honest with ourselves, we’re actually poorly equipped to understand what that activity means,” Galsky said during a presentation at ASCO 2018.7 He is the director of the Genitourinary Cancer Research Program and associate medical director of the Cancer Clinical Trials Office at Mount Sinai Health System, The Tisch Cancer Institute, in New York, New York.
“These are very important proof-of-concept studies testing whether single-agent immune checkpoint blockade can be administered in this setting,” Galsky said in an interview with OncologyLive®. “Both trials utilize pathologic complete response rate as the primary endpoint. The pCR rate with just 2 to 3 doses of immune checkpoint blockade prior to cystectomy was similar to what we achieve with combination cisplatin-based chemotherapy.
“These studies will inform further studies that have the potential to transform care,” he added. “Of course, as single-arm studies, they are not practice-changing, but they will impact further studies that may change practice in the future.”In TNBC, the phase II neoadjuvant GeparNuevo study evaluated the addition of durvalumab (Imfinzi), a PD-L1 inhibitor, to combination chemotherapy in patients with early-stage TNBC (cT1b-cT4d), irrespective of nodal involvement. Patients were randomized to receive durvalumab (1.5 g) or placebo during a “window of opportunity” stage.
After undergoing a core biopsy, patients went on to receive nab-paclitaxel (Abraxane; 125 mg/m2) with either durvalumab or placebo for 12 weeks. This was followed by epirubicin (90 mg/m2) plus cyclophosphamide (600 mg/m2) with either durvalumab or placebo for another 8 weeks. The primary endpoint was to compare rates of pCR, defined as ypT0 ypN0.
Investigators reported at ASCO 2018 that the trial enrolled 174 patients with a median age of 49.5 years (range, 23.0-76.0), with 61.5% of participants completing all treatments. Among the entire population, patients in the durvalumab group (n = 88) achieved a 53.4% pCR rate while the rate was 44.2% for those who received pacebo (n = 86).8 The difference was not statistically significant (P = .287), and the OR, adjusted for tumor-infiltrating lymphocyte stratification, was 1.53 (95% CI, 0.82-2.84; P = .182).
Notably, however, the pCR rates were significantly higher in several preplanned subgroups: patients in the window cohort who started with durvalumab before chemotherapy compared with those who received placebo (61.0% vs 41.4%); those with stage IIa and higher TNBC compared with lower-stage disease (55.8% vs 38.6%); and those age <40 years compared with older patients (69.2% vs 42.9%).
In terms of AEs, durvalumab was well tolerated, lead author Sibylle Loibl, MD, PhD, chair of the German Breast Group, reported. The most frequent grade 1-4 hematologic AEs in the durvalumab and placebo groups were anemia (94.6% vs 96.3%), neutropenia (77.2% vs 81.7%), and thrombocytopenia (38.0% vs 34.1%). The rate of grade 3-4 AEs was most pronounced for neutropenia for the durvalumab and placebo arms (37.0% vs 41.5%); other AEs of this severity were much lower for anemia (2.2% vs 2.4%), thrombocytopenia (1.1% vs 2.4%), and febrile neutropenia (4.3% vs 2.4%). Of note, immune-rated toxicities of any grade included hypothyroidism (6.5% vs 2.4%) and hyperthyroidism (7.6% vs 0%).
The findings were strong enough for Loibl and colleagues to conclude that durvalumab should be investigated further in primary TNBC, an assessment that Charles G. Drake, MD, PhD, shared during a discussion of the results of GeparNuevo and several other studies at ASCO 2018.9
“The trial was almost positive. It was actually pretty exciting,” said Drake, who is the co-director of cancer immunotherapy programs at Herbert Irving Comprehensive Cancer Center at NewYork-Presbyterian/Columbia University Medical Center in New York, New York. “The pathological complete response rate was trending in the right direction...The patients who got that short 2 weeks of immunotherapy monotherapy prior to starting combinatorial treatment seemed to do a little better.
“One thing that is clear is that the response rate was not worse with the addition of immunotherapy,” he added. “…This is a really interesting finding.”