Doctors Debate: Should Immunotherapy Be Given Before or After Surgical Resection in Locally Advanced NSCLC?

Valerie W. Rusch, MD, FACS

Immunotherapy (IO) has revolutionized the treatment of locally advanced non–small cell lung cancer (NSCLC), offering an effective and noninvasive treatment option to patients being treated in this setting. For those whose treatment plan involves surgery, physicians have options to administer IO either before or after surgery. Preference regarding timing of administration led to a debate at the International Association for the Study of Lung Cancer 2020 Targeted Therapies in Lung Cancer Conference.

Valerie W. Rusch, MD, FACS, Miner Family Chair in Intrathoracic Cancers and vice chair for Clinical Research in the Department of Surgery at Memorial Sloan Kettering Cancer Center in New York, was tasked with informing the meeting attendees her reason for using IO in the neoadjuvant setting with evidence to support her choice. Her opponent, Jessica Donington, MD, professor and chief of Thoracic Surgery at the University of Chicago, provided evidence for the use IO after surgery, which she considers to be the most optimal treatment sequence for this patient population.

IO Before Surgery

The controversy surrounding neoadjuvant versus adjuvant treatment in general for patients with locally advanced NSCLC is best illustrated by a phase III study led by Enriqueta Felip, MD, PhD, that examined outcomes in patients with stage IA tumors who had surgery with or without either adjuvant or neoadjuvant chemotherapy, Rusch argued.¹

In the study, there was no significant difference in disease-free survival (DFS), the primary end point of the study, or overall survival between patients who received surgery alone and those who receive chemotherapy at any point during their treatment. Superiority trended towards the preoperative chemotherapy arm versus the surgery arm, but this was not statistically significant.

There were 374 events of disease progression and death in the study overall, which occurred most commonly in the surgery arm (n = 132), followed by the adjuvant arm (n = 125) then the neoadjuvant arm (n =117). The hazard ratio for disease progression or death related to the comparison of neoadjuvant chemotherapy and surgery was 0.92 (95% CI, 0.81-1.04; P = .176). The hazard ratio for adjuvant chemotherapy compared with surgery was 0.96 (95% CI, 0.75-1.22; P = .74).²

The five-year OS rates were 44% in the surgery arm, 46.6% in the neoadjuvant arm, and 45.5% in the adjuvant arm.

Aside from efficacy, Rusch found it notable that both chemoradiotherapy arms had similar grade greater than 3 toxicities. Additionally, a larger percentage of patients in the neoadjuvant arm (90.4%) completed the planned amount of treatment compared with only 60.9% in the adjuvant arm.^1,2

Although the study missed its primary end point of demonstrating an absolute improvement

in disease-free survival of 15% at 5 years with the addition of chemotherapy, due to the lack of significant difference in DFS and OS, Rusch holds that physicians should reconsider shifting systemic therapy to the neoadjuvant setting.

Factors at play may explain the favorability of neoadjuvant IO in this setting. Rush stated that administration of IO before surgery may offer greater benefit by:

• attacking micrometastases early on
• helping to assess the sensitivity and resistance of agents to inform the design of future trials
• shortening clinical trial timelines
• assesses the sensitivity of agents used in induction
• providing a complete and major pathologic response (MPR) outcome surrogate
• leading to better chemotherapy drug delivery and tolerability
• inducing better compliance with subsequent therapies
• allowing time to identify unsuspected metastases and comorbidities before local therapy
• creating an opportunity for “prehab” before surgery.

Despite the known potential of neoadjuvant IO, questions continue to circulate among experts about the toxicities neoadjuvant therapy can cause and their potential to sabotage surgery outcomes or cause delays. Rusch believes these concerns may not be fairly characterized in the oncologic surgery community.

“I have seen this only in patients who had protracted treatment, especially dual agent treatment. I have not seen it to a degree that makes the operation hazardous in patients who’ve had monotherapy with the usual 2 or 3 cycles of IO.” Rusch told Targeted Therapies in Oncology (TTO), in an interview. “The toxicity of neoadjuvant IO to-date has not been an issue,” she added.

Clinical Evidence for Neoadjuvant IO

As an example, Rusch shared results from a pilot study of nivolumab (Opdivo) in patients with untreated, surgically resectable early stage NSCLC (NCT02259621), which evaluated the safety and feasibility of the PD-1 inhibitor as neoadjuvant therapy. This study showed that nivolumab led to a limited number of predominantly low-grade adverse effects. Only 1 grade 3/4 toxicity was observed in patients. The investigators concluded that surgery was not delayed following neoadjuvant therapy and that treatment elicited an MPR in 45% of resected tumors.³

Investigators of the pilot study lead by Patrick Forde, MB, BCh, noted that their findings signal that some patients with NSCLC may derive benefit from neoadjuvant IO without radiographic tumor shrinkage due to immune-cell infiltration into the tumor rather than an observed change in the tumor size.

Outcomes

In terms of perioperative outcomes in patients who have had neoadjuvant IO, Rusch detailed results from a trial that showed neoadjuvant therapy does not lead to increased morbidity or mortality with a single-group assignment study (NCT02259621), which assessed safety and feasibility of preoperative nivolumab with or without ipilimumab (Yervoy) as treatment of patients with high-risk resectable NSCLC.⁴

Altogether there were 20 resection surgeries, which included 15 lobectomies, 2 pneumonectomies, 1 bilobectomy, 1 sleeve lobectomy, and 1 wedge resection. Of those procedures, 7 (54%) required thoracotomy. No operative mortality was observed in the study and morbidity occurred in 50% of patients. Additionally, 45% of patients had a MPR. The study authors concluded from this research that neoadjuvant IO had no association with perioperative mortality or morbidity.

In the phase II, open-label, multicenter, single-arm LCMC3 study, in which Rusch was a participating investigator, the safety and efficacy of neoadjuvant atezolizumab (Tecentriq) was evaluated in patients with stage IB, II, IIIA, or selected IIIB resectable NSCLC. The study was positive for its primary end point of MPR, which was 19% (95% CI, 11%-30%) and included 4 pathologic complete responses (5%). Partial responses were seen in 6 patients (7%), stable disease in 80 (89%), and progressive disease in 4 (4%). What was notable about this study in terms of the debate, was that a 10-day window to surgery was achieved following up to 2 cycles of atezolizumab in the neoadjuvant setting.⁵

“I think the people who are fearful of the neoadjuvant approach are surgeons who have not yet had a lot of experience operating after this kind of induction therapy.”

Rusch does believe that many questions still exist on the matter, which calls for more clinical research.

“We don’t have a good understanding of why some patients have a robust response to immunotherapy and others do not. That’s important, not only in terms of selecting patients but also in terms of the neoadjuvant therapy approach, Rusch stated. This is one reason why neoadjuvant clinical trials are incredibly important.”

IO After Surgery

Donington used her time to highlight that although only 10% to 30% of patients do achieve remission, patients in this setting have the option of receiving therapy with this outcome as the goal. The use of systemic therapy for these patients puts them at high risk of distant relapse, she argued, while surgery plus radiotherapy may lead to more loco-regional control.

Jessica Donington, MD, MSCR

Strategies for treatment of locally advanced NSCLC are dependent on whether the patient has node-positive disease and the number of nodes are present. The validity IO in the adjuvant setting is best supported with the phase III PACIFIC trial, which Donington started her argument with.⁶

Interim analysis data published in the New England Journal of Medicine showed that durvalumab (Imfinzi) significantly improved PFS compared with placebo in patients with stage III resectable NSCLC. From baseline, the median PFS observed with durvalumab was 16.8 months (95% CI,13.0-18.1) with durvalumab versus 5.6 months (95% CI, 4.6-7.8) with placebo (stratified HR for disease progression or death, 0.52; 95% CI, 0.42-0.65; P <.001). At 12 months, the PFS rate was 55.9% versus 35.3%, and the18-month PFS rate was 44.2% versus 27.0%. In addition, patients who received durvalumab had a longer DOR, with 72.8% versus 46.8%, respectively, having ongoing responses at 18 months.⁷

Survival data for durvalumab was published a year later and demonstrated a 12-month OS rate of 83.1% (95% CI, 79.4%-86.2%) compared with 75.3% (95% CI, 69.2%-80.4%) in the placebo group. The 24-month OS was 66.3% (95% CI, 61.7%-70.4%) in the durvalumab group versus 55.6% (95% CI, 48.9%-61.8%) in the placebo group (two-sided P =.005). Overall, durvalumab showed a significant improvement in OS compared with placebo (HR, 0.68; 99.73% CI, 0.47-0.997; P =.0025). The benefit was seen across the prespecified subgroups which included patients with and without EGFR mutations or unknown EGFR mutation status, nonsquamous and squamous histology, and PD-L1 expression levels.⁸

In this study, durvalumab was administered after chemoradiotherapy in the adjuvant setting. Scott J. Antonio, MD, PhD, and colleagues wrote that based on the results, durvalumab may be an effective adjuvant therapy in this patient population after standard therapy.

The phase III ANVIL study (NCT02595944) of nivolumab after surgery and chemotherapy, part of the ALCHEMIST lung cancer trials, is underway to further confirm the benefit of adjuvant IO in patients with locally advanced NSCLC. Donington noted that many clinical trials are ongoing that can defend the use of neoadjuvant IO; but considering that these trials include patients who would not traditionally be eligible for chemotherapy in the clinical setting, physicians should be cautious about using IO agent in the neoadjuvant space.

“By giving induction therapy prior to resection, you do take on risks. One risk is that you can have attrition, meaning that patients who go for induction therapy do not come back. Then, you’ve taken away the chance for cure because local regional control is what’s most important for cure,” Donington explained in an interview with TTO. “Second, you could cause surgery delays [or] make surgery more complex.”

Adding to her stance, Donington went into detail about clinical trial end points and their significance. The end points that pertain to surgical success and may be considered most relevant for this argument include: surgical delays, operative morbidity, and operative mortality, which are essential for induction strategy, and minimally invasive resection, which is important for short-term outcomes.⁶

Attrition with Induction Therapy

As examples, the single-group assignment study of neoadjuvant nivolumab, the LCM3 study of neoadjuvant atezolizumab, the phase II NEOSTAR study (NCT03158129) of nivolumab with or without ipilimumab, and the phase II NADIM study (NCT03081689) we examined in Donington’s presentation. All of the studies that the use of induction therapy caused attrition. For the single-group assignment study, there was 10% attrition; the LCM3 study had 7% attrition; NEOSTAR had 16% attrition; and NADIM had 11% attrition.⁶

Surgery Delays

Resection delays associated with induction therapy of any kind should be prevented, said Donington. It should be a significant enough length to allow patients to respond, but not delay resection.

In terms of toxicity leading to surgery delays, the most common is pneumonitis, explained Donington, which occurs in 5% to 19% of patients. However, it is more likely that surgery is delayed with combination IO strategies than with monotherapy.⁶

This argument regarding surgery delays was a key point of contention during the debate.

“Donington’s argument presupposes that the impact of surgery is so substantial that a narrow window of roughly 6 weeks is going to make a difference in OS,” said Rusch.

Increased Toxicity

Studies like LCM3, NEOSTAR, and others have shown that minimally invasive resection, which is the standard of care, decreases morbidity and hospital length-of-stay, and improves tolerability and recovery. Donington queries, is the benefit of IO in the neoadjuvant setting worth losing the positive results of minimally invasive resection?

Donington also questions the claim that neoadjuvant therapy improves survival. She highlighted that early clinical trials do not show a correlation between induction therapy and radiographic and pathologic responses.

The final point of that argument against IO before surgery was the idea that mandatory biomarker analyses to determine PD-L1 expression ahead of neoadjuvant therapy may lead to unnecessary surgery delays and increased costs, as seen in the ALCHEMIST lung cancer trials.

In conclusion, Donington stated that any use of neoadjuvant IO should be balanced against any potential benefit.

Common Ground

Both surgeons realize the potential of IO, in general, for the treatment of patients surgically resectable patients with NSCLC. The unanswered question whether it’s appropriate for patients with early disease. In terms of giving IO in the neoadjuvant setting, Rusch and Donnington agree that there may be potential but are uncertain of which patients would derive the most benefit from IO in this setting.

“We understand that giving treatment before surgery may have many benefits. For one, treatments tend to be better tolerated and there’s a better chance of immune priming with a big tumor in place. The uncertainty is whether this is appropriate for patients with stage I or II disease,” Donington stated in the interview.

_References:

_{1. Rusch VW. Should IO be given before or after surgery? BEFORE! Presented at: International Association for the Study of Lung Cancer. 2020 Targeted Therapies of Lung Cancer Meeting; February 19–22, 2020; Santa Monica, California}

_{2. Felipe E, Roswell R, Maestre JA, et al. Preoperative chemotherapy plus surgery versus surgery plus adjuvant chemotherapy versus surgery alone in early-stage non–small-cell lung cancer. J Clin Oncol. 28(19);3138-3145. doi: 10.1200/JCO.2009.27.6204}

_{3. Forde PM, Chaft JE, Smith KN, et al. Neoadjuvant PD-1 blockade in resectable lung cancer. N Engl J Med. 2018;378(21):1976-1986. doi: 10.1056/NEJMoa1716078}

_{4. Bott MJ, Yang SC, Park BJ, et al. Initial results of pulmonary resection after neoadjuvant nivolumab in patients with resectable non–small cell lung cancer. Thorac Cardiovasc Surg. 2019;158(1):269-276. doi: 10.1016/j.jtcvs.2018.11.124.}

_{5. Kwiatkowski PJ, Rusch VN, Chaft JE, et al. Neoadjuvant atezolizumab in resectable non-small cell lung cancer (NSCLC): Interim analysis and biomarker data from a multicenter study (LCMC3). J Clin Oncol. 2019; 37(15): 8503-8503. doi: 10.1200/JCO.2019.37.15.}

_{6. Donington, J. Surgery Before Immunotherapy. Presented at: International Association for the Study of Lung Cancer. 2020 Targeted Therapies of Lung Cancer Meeting; February 19–22, 2020; Santa Monica, California.}

_{7. Antonio SJ, Willegas A, Daniel D, et al. Durvalumab after chemoradiotherapy in stage III non–cmall-cell lung cancer. N Engl J Med. 2017;377(20):1919-1929. doi: 10.1056/NEJMoa1709937}

_{8. Antonia SJ, Villegas A, Daniel D, et al. Overall survival with durvalumab after chemoradiotherapy in stage iii nsclc. N Engl J Med. 2018;379(24):2342-50. doi: 10.1056/NEJMoa1809697.}