“PD-L1 expression remains the gold standard”

Jordi Remon

Jordi Remon, MD, PhD

Department of Medical Oncology, Centro Integral Oncológico Clara Campal, Hospital HM Nou Delfos, HM Hospitales, Barcelona, Spain.

What needs to be considered in the context of neoadjuvant and adjuvant immunotherapy in patients with resectable lung cancer?

Today we know that immunotherapy works in the perioperative therapeutic strategy of patients with early-stage NSCLC, both in the neoadjuvant setting in combination with chemotherapy and in the adjuvant setting. However, data from the metastatic setting show that patients with oncogenic-driven NSCLC do not benefit from checkpoint inhibition, at least when the treatment is administered as monotherapy. This does not completely translate to the adjuvant setting where data are somewhat conflicting, with IMpower010 showing no benefit of adjuvant atezolizumab in EGFR-mutant and ALK-positive tumors, while the KEYNOTE-091 trial conducted with pembrolizumab demon­strated an advantage in patients with EGFR mutations [1, 2]. However, this is a very small number of patients. From my point of view, most important before starting immunotherapy either in the neoadjuvant or adjuvant setting is the assessment of at least the most common genomic alterations including EGFR, ALK, ROS1, and BRAF. Taking into account the data from metastatic disease, especially for EGFR-, ALK-, and ROS1-positive tumors, these should be more or less excluded from treatment with adjuvant checkpoint inhibitors. In the neoadjuvant setting, this is less clear as patients with EGFR– and ALK-positive tumors were not eligible for the CheckMate 816 study [3].

In which ways can tyrosine kinase inhibitors contribute to effective peri­operative treatment?

Many trials clearly established that TKIs, at least EGFR TKIs, improve disease-free survival as compared to placebo in patients with EGFR-mutant, completely resected stage II and IIIA NSCLC. However, with respect to overall survival, the data are still immature. Obviously, we would like to convey this benefit to other oncogenic drivers in the adjuvant setting, which is complicated due to their low incidences. We would also like to have more early data on the efficacy of TKIs in the neoadjuvant setting, as the evidence is still very limited here. Although third-generation EGFR TKIs provide high response rates of approximately 70 % in EGFR-mutant tumors according to the RECIST criteria, the pathological complete response (pCR) rates with neoadjuvant TKI mono­therapy are below 10-15 %. We know that pCR with chemotherapy alone or plus immunotherapy is associated with disease-free survival benefit. Therefore, I am not completely sure that TKI mono­therapy in the neoadjuvant setting will be strong enough to achieve high pCR rates.

What can we achieve today with targeted agents in patients with oncogene-driven lung cancer who develop brain metastases?

This a very relevant clinical question, because we know that oncogenic-driven NSCLCs have a higher incidence of brain metastases than the wildtype tumors, and approximately 30 % of patients with oncogenic-driven NSCLC show brain metastases at baseline. Very potent next-generation TKIs with good brain penetration are available today. For patients with baseline brain metastasis, even for those that are minimally symptomatic, these targeted agents should be the standard of care. With this strategy, we can defer the use of radiotherapy that may have delayed cognitive side effects. Indeed, starting with a next-generation TKI, especially in EGFR– or ALK-positive tumors, does not negatively impact the outcome of these patients, allowing to defer the radiotherapy strategy in case of progression.

Which other treatment approaches are deemed promising in patients with CNS affliction whose tumors do not harbor genetic drivers?

Patients with baseline brain metastases and without oncogenic drivers have mostly been excluded from trials testing checkpoint inhibitors as monotherapy or in combination with chemotherapy. However, today we have data from different clinical trials, especially with combinations of chemotherapy and immunotherapy, showing that patients with previously treated and asymptomatic brain metastases who receive immunotherapy and chemotherapy obtain the same magnitude of benefit as those without brain metastases. In the phase II Atezo-Brain trial conducted by the Spanish Lung Cancer Group, atezolizumab plus chemotherapy showed high intracranial activity in patients with asymptomatic and untreated brain metastases, resulting in an intracranial response rate of 40 % [4]. Thirty-two percent of these patients were alive at 2 years, meaning that even patients with minimally symptomatic and untreated brain metastases may achieve long-term survival benefit when receiving the combination of chemotherapy and immunotherapy. This data must be validated in larger prospective trials.

What are the current challenges in selecting the best immunotherapeutic approach for the individual patient with metastatic NSCLC, and how do you handle them?

One of the most challenging questions is why we do not obtain the same results in daily clinical practice as are reported in clinical trials. The answer is that the population is completely different, with just 30 % of the patients we see in our daily practice mirroring those included in studies. I think that today the best strategy for patient selection is the assessment of the PD-L1 expression in tumors. PD-L1 is the most reliable predictive biomarker. For patients with high PD-L1 expression, immunotherapy as monotherapy is the standard of care. There is a debate if some of these tumors might benefit from the addition of chemotherapy. Perhaps those patients with high tumor volume, even if their tumors highly express PD-L1, could be more suitable to be treated with a combination strategy. The ongoing INSIGNA and PERSEE trials may help to answer this question. However, indirect comparisons of clinical trials testing immunotherapy as monotherapy (KEYNOTE-024) or in combination with chemotherapy (KEYNOTE-189) in tumors with high PD-L1 expression have reported the same median OS of 27 months and the same 3-year OS of 43.7 % [5, 6]. For the other group of tumors with PD-L1 expression below 50 %, any combination strategy regardless of the PD-L1 expression status or histological subtype could be suitable.

What are potential predictive biomarkers for checkpoint inhibitor therapy, as well as pitfalls encountered with their use and ways to optimize their implementation in clinical routine?

Although several predictive biomarkers have been assessed in NSCLC, the PD-L1 expression remains the gold standard. The association between tumor mutational burden in the blood and the efficacy of atezolizumab was tested prospectively in the BFAST trial. However, high tumor mutational burden was not a strong predictive biomarker for the selection of patients who benefited from atezolizumab compared with chemotherapy [7]. Today, we can only say that PD-L1 remains the standard of care. Probably the most important point for which predictive biomarkers could be relevant is the identification of patients who might develop more pronounced toxicity or a higher risk of early progressive disease on immunotherapy. To my mind, this is the more relevant challenge in our daily clinical practice, as well as the best therapeutic approach at progression under immunotherapy.


  1. Wakelee H et al., IMpower010: Primary results of a phase III global study of atezolizumab versus best supportive care after adjuvant chemotherapy in resected stage IB-IIIA non-small cell lung cancer (NSCLC). J Clin Oncol 2021; 39(suppl 15): 8500
  2. Paz-Ares L et al., Pembrolizumab versus placebo for early-stage non-small cell lung cancer following complete resection and adjuvant chemotherapy when indicated: randomized, triple-blind, phase III EORTC-1416-LCG/ETOP 8-15 – PEARLS/KEYNOTE-091 study. ESMO Virtual Plenary Abstract 2022; 33(4): P451-453
  3. Forde PM et al., Neoadjuvant nivolumab plus chemotherapy in resectable lung cancer. N Engl J Med 2022 Apr 11. doi: 10.1056/NEJMoa2202170
  4. Nadal E et al., Atezo-Brain: single arm phase II study of atezolizumab plus chemotherapy in stage IV NSCLC with untreated brain metastases. WCLC 2021, abstract OA09.02
  5. Reck M et al., KEYNOTE-024 3-year survival update: pembrolizumab vs. platinum-based chemotherapy for advanced non-small cell lung cancer. J Thorac Oncol 2019; 14(10): S243
  6. Gray J et al., Pembrolizumab + pemetrxed-platinum vs. pemetrexed-platinum for metastatic NSCLC: 4-year follow-up from KEYNOTE-189. J Thorac Oncol 2021; 16(3): S224
  7. Dziadziuszko R et al., Atezolizumab vs platinum-based chemotherapy in blood-based tumour mutational burden-positive patients with first-line advanced/metastatic NSCLC: results of the Blood First Assay Screening Trial (BFAST) phase III cohort C. Ann Oncol 2021; 32(suppl 5): S950-S951

© 2019 Springer-Verlag GmbH, Impressum