Identifying biomarkers to predict immune-checkpoint therapy response—is it a reality or a distant dream?

Immune checkpoint inhibitors (ICIs) are an important component of the systemic therapeutic strategy for non-small cell lung cancer (NSCLC). While its role was initially limited to metastatic NSCLC, in recent times, there have been multiple approvals for its use in both adjuvant and neoadjuvant limited-stage setting (1). It is apparent that not all patients derive benefit from ICIs. In the phase 3 study involving the direct comparison of Pembrolizumab, a programmed cell death 1 (PD-1) inhibitor, to chemotherapy in advanced and unresectable NSCLC with a high programmed death-ligand 1 (PD-L1) staining, i.e., a PD-L1 ≥50%, the response rate for pembrolizumab from a cohort of 154 patients was 45%. Although this number was impressive and was significantly better than chemotherapy (27%), it also meant that >50% of the patients did not respond to the drug (2). Traditional biomarkers for ICI like PD-L1 and tumor mutational burden (TMB) do have predictive clinical utility but are not completely reliable, leaving clinicians longing for a biomarker with better specificity (3). An example of this is even more evident in the limited stage setting, as in the case of the KEYNOTE-091 study where pembrolizumab was studied for its role in the adjuvant setting after resection for limited stage NSCLC. While the PD-L1 1–49% sub cohort showed a statistically significant disease-free survival (DFS) difference, the ≤1% and ≥50% did not reach significance (4). While one can argue that limited stage and metastatic disease are biologically different, the need for accurate predictors for ICI response in both settings cannot be denied. Besides, ICIs do have the potential to have life-threatening and high-grade adverse effects, necessitating the development for reliable tools beyond PD-L1 and TMB that enable better patient selection (5). Looking beyond PD-L1 and TMB, several molecular biomarkers like STK11 have been studied in retrospective studies to predict ICI response, but none in a prospective manner that would result in clinical use (6).

van de Haar et al. conducted a retrospective study utilizing a diverse cohort from multiple academic centers in Europe to answer if combining genomic alterations with markers like TMB and PD-L1 have any significant ICI predictive value in advanced NSCLC (7). The study involved analysis of whole genome sequencing (WGS), TMB and PD-L1 done on histologic biopsies from advanced NSCLC treated with PD-1/PD-L1 ICI monotherapy. The study worked on the hypothesis that while biomarkers like STK11, KEAP1, and EGFR have known correlation to ICI resistance, their specificity for this is poor with multiple studies revealing contradictory results. Similarly, TMB and PD-L1, while a good marker of ICI response, is also not specific as some patients with low TMB and PD-L1 show a response to ICI (7). Seventy-five patients underwent WGS in the study; 63% were adenocarcinomas and 73% of the cohort received ICI as 2^nd line therapy, suggesting that they may have received chemotherapy as first line; 18.7% had PD-L1 ≥50%, while 21.3% had a PD-L1 of 1–50%. The authors use durable clinical benefit (DCB) as the primary endpoint. This was defined as the best overall response (complete or partial response or stable disease) that lasted at least 6 months. This was done to homogenize the results between patients. The overall DCB was 37% (28/75). In the KEYNOTE-010 study where pembrolizumab was used a 2^nd line agent against docetaxel, the overall response rate (ORR) was 21.2% (8). In CheckMate-057 with nivolumab, the ORR was 19%, while the disease control was close to 44.5% (9). As anticipated, patients with high TMB (≥10 mut/Mb) had a higher DCB (66 vs. 20%, P=0.000087), as well as statistically significant higher progression free survival (PFS) and overall survival (OS). This was observed despite the relatively small numbers (TMB high 28, low 47). On combining TMB with PD-L1 and studying the DCB, intermediate (1–50%) and high PD-L1 (>50%) with TMB did show a marginally significant improvement compared to low PD-L1 (<1%) and high TMB, despite the small sample size. However, some response was noted even in the low PD-L1 and low TMB sub-cohort. Eighteen patients had a pathogenic STK11 alteration, among which, none of the 9 patients with low TMB had DCB, while 7 out of 9 high TMB patients had a DCB with a P of 0.0023. STK11 alteration with low TMB had a lower PFS and OS, but this not observed with high TMB. Similarly, for KEAP1 altered/low TMB, none out of 7 patients had DCB, while all 5 KEAP1 altered/high TMB patients exhibited DCB (P=0.0013). KEAP1 altered/low TMB had a poor PFS which was not seen with high TMB. OS showed no significance. Only 4 patients had an EGFR alteration of which 1 had a DCB and had high TMB, while 3 who did not have DCB were low TMB. On combing all 3 genomic markers, DCB was seen in none of the 15 low TMB patients, while 11/13 patients with high TMB had a DCB that amounted to 85%. When all 3 were altered, PFS was lower with low TMB but not with high TMB. However, when altered, low TMB had a poor OS, while high TMB continues to have an improved OS. The study included a larger validation cohort that was larger with 169 patients from published literature. Several findings were very similar between the cohorts. When all three markers described are altered with low TMB, PFS and OS were significantly lower. High TMB resulted in better OS even when all three markers were altered (7).

STK11 is a tumor suppressor gene seen in 10% of NSCLC and was initially thought as a definite predictor ICI resistance (10). KEAP1 regulates oxidate damage response and occurs in 20% of NSCLC, usually resulting in an immunologically cold tumor (11). A large study, comprising several hundred patients evaluated the impact of STK11 and KEAP1 in the context of KRAS alteration for predicting response to PD-1 ICI. Both STK11 and KEAP1 were associated with a worse PFS and OS when KRAS was mutated. STK11 and KEAP1 also resulted in poor PFS and OS independently (12). But, in an exploratory analysis of the KEYNOTE-042 study, OS benefit for pembrolizumab was seen regardless of the STK1, KEAP1 or KRAS mutational status (13). The case of EGFR is more established as there is strong evidence to support that EGFR driven NSCLC respond poorly to ICI (14). The FLAURA study in the advanced setting (15), ADURA in the resected space (16), and more recently, the LAURA study post chemoradiotherapy (17) has established the role EGFR tyrosine kinase inhibitors like Osimertinib in EGFR mutated NSCLC. Similarly, ALK+ NSCLC also respond poorly to ICI (18). Given this data, it is standard clinical practice to use a tyrosine kinase inhibitor over ICI when targetable alterations exist such as EGFR, ALK, RET, ROS1, MET, NTRK, or BRAF (1). Other genes that have been evaluated as potential candidates to predict ICI response include PTEN, IFNGR1, TP53 SPIN, SETD7, FGFR3, YAP1, TEAD3, BCL, BLCAP, CD8A, CD8B, GZMA, GZMB, and PRF1 (19).

Despite being severely limited by the sample size with most sub-cohorts in single digits, the study raises an interesting hypothesis of combining genomic markers with TMB to improve the specificity of predicting ICI response in metastatic NSCLC (7). While no definite conclusions can be made from the study itself, it draws attention to the potential for combination strategies in the future (7). The study hypothesizes that TMB may supersede other markers in predicting ICI response. The authors suggest that in patients with high TMB, the immune response may be strong enough to overcome the resistance caused by other alterations, which may not be the case with low TMB. They also conclude that ICI therapy could be reconsidered when all three biomarkers (STK11/KEAP1/EGFR) are altered and TMB is low, which would need to be confirmed via larger studies. The study also highlights the advantages and disadvantages of using WGS TMB measurement, by emphasizing that the risk of misclassification and eventual treatment error should be weighed against cost and logistical challenges (7). It is well known that tumors with high TMB have develop a robust anti-tumor pro-inflammatory response resulting in response to ICI (20). Attempts using combination strategies have been studied with conflicting results (21). Combining PD-L1 and TMB promises a high predictive ability for ICI response in advanced NSCLC (21). The LACE Bio consortium queried the PD-L1/TMB combination approach in early-stage non metastatic NSCLC to prognosticate and predict response. PD-L1 negative/TMB high group had the best prognosis, while high TMB predicted poor response to adjuvant chemotherapy, hypothesizing that ICI may be better suited for this group (22). Genomic alterations have been combined to develop an immune-checkpoint therapy signature profile that may predict response (23). In the era of artificial intelligence (AI), studies are underway to see if machine learning can analyze genomic signatures and enable response prediction (24). Invariably, all these studies are handicapped by the limitation of being retrospective in nature, uncontrolled/unadjusted with confounders and limited in sample number. Future research should focus on prospective clinical studies that account for these biomarkers from the outset (7,21).

In conclusion, we believe that with the currently available data, none of the genomic markers have clinical utility at this point of time. While PD-L1 and TMB are widely adopted in practice and does have clinical utility, there is a desire to improve the predictive ability which led to studies like the one described. Technologies like circulating tumor DNA and detecting minimal residual disease are in the forefront of NSCLC research which may provide a viable answer to this question (25).

Acknowledgments

Funding: None.

Footnote

Provenance and Peer Review: This article was commissioned by the editorial office, Translational Cancer Research. The article has undergone external peer review.

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1383/prf

Conflicts of Interest: The author has completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-24-1383/coif). The author has no conflicts of interest to declare.

Ethical Statement: The author is accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Riely GJ, Wood DE, Ettinger DS, et al. Non-Small Cell Lung Cancer, Version 4.2024, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2024;22:249-74. [Crossref] [PubMed]
2. Reck M, Rodríguez-Abreu D, Robinson AG, et al. Pembrolizumab versus Chemotherapy for PD-L1-Positive Non-Small-Cell Lung Cancer. N Engl J Med 2016;375:1823-33. [Crossref] [PubMed]
3. Yarchoan M, Albacker LA, Hopkins AC, et al. PD-L1 expression and tumor mutational burden are independent biomarkers in most cancers. JCI Insight 2019;4:e126908. [Crossref] [PubMed]
4. O'Brien M, Paz-Ares L, Marreaud S, et al. Pembrolizumab versus placebo as adjuvant therapy for completely resected stage IB-IIIA non-small-cell lung cancer (PEARLS/KEYNOTE-091): an interim analysis of a randomised, triple-blind, phase 3 trial. Lancet Oncol 2022;23:1274-86. [Crossref] [PubMed]
5. Ashok Kumar P, Basnet A. The role of laboratory tests as a prognostic marker for immune-checkpoint therapy in non-small cell lung cancer. Transl Lung Cancer Res 2023;12:1838-41. [Crossref] [PubMed]
6. Catalano M, Iannone LF, Nesi G, et al. Immunotherapy-related biomarkers: Confirmations and uncertainties. Crit Rev Oncol Hematol 2023;192:104135. [Crossref] [PubMed]
7. van de Haar J, Mankor JM, Hummelink K, et al. Combining Genomic Biomarkers to Guide Immunotherapy in Non-Small Cell Lung Cancer. Clin Cancer Res 2024;30:1307-18. [Crossref] [PubMed]
8. Herbst RS, Garon EB, Kim DW, et al. Five Year Survival Update From KEYNOTE-010: Pembrolizumab Versus Docetaxel for Previously Treated, Programmed Death-Ligand 1-Positive Advanced NSCLC. J Thorac Oncol 2021;16:1718-32. [Crossref] [PubMed]
9. Borghaei H, Paz-Ares L, Horn L, et al. Nivolumab versus Docetaxel in Advanced Nonsquamous Non-Small-Cell Lung Cancer. N Engl J Med 2015;373:1627-39. [Crossref] [PubMed]
10. Li A, Wang Y, Yu Z, et al. STK11/LKB1-Deficient Phenotype Rather Than Mutation Diminishes Immunotherapy Efficacy and Represents STING/Type I Interferon/CD8(+) T-Cell Dysfunction in NSCLC. J Thorac Oncol 2023;18:1714-30. [Crossref] [PubMed]
11. Kansanen E, Kuosmanen SM, Leinonen H, et al. The Keap1-Nrf2 pathway: Mechanisms of activation and dysregulation in cancer. Redox Biol 2013;1:45-9. [Crossref] [PubMed]
12. Ricciuti B, Arbour KC, Lin JJ, et al. Diminished Efficacy of Programmed Death-(Ligand)1 Inhibition in STK11- and KEAP1-Mutant Lung Adenocarcinoma Is Affected by KRAS Mutation Status. J Thorac Oncol 2022;17:399-410. [Crossref] [PubMed]
13. Mok TSK, Lopes G, Cho BC, et al. Associations of tissue tumor mutational burden and mutational status with clinical outcomes in KEYNOTE-042: pembrolizumab versus chemotherapy for advanced PD-L1-positive NSCLC. Ann Oncol 2023;34:377-88. [Crossref] [PubMed]
14. Shi C, Wang Y, Xue J, et al. Immunotherapy for EGFR-mutant advanced non-small-cell lung cancer: Current status, possible mechanisms and application prospects. Front Immunol 2022;13:940288. [Crossref] [PubMed]
15. Ramalingam SS, Vansteenkiste J, Planchard D, et al. Overall Survival with Osimertinib in Untreated, EGFR-Mutated Advanced NSCLC. N Engl J Med 2020;382:41-50. [Crossref] [PubMed]
16. Tsuboi M, Herbst RS, John T, et al. Overall Survival with Osimertinib in Resected EGFR-Mutated NSCLC. N Engl J Med 2023;389:137-47. [Crossref] [PubMed]
17. Lu S, Kato T, Dong X, et al. Osimertinib after Chemoradiotherapy in Stage III EGFR-Mutated NSCLC. N Engl J Med 2024;391:585-97. [Crossref] [PubMed]
18. Schenk EL. Narrative review: immunotherapy in anaplastic lymphoma kinase (ALK)+ lung cancer-current status and future directions. Transl Lung Cancer Res 2023;12:322-36. [Crossref] [PubMed]
19. Hellmann MD, Nathanson T, Rizvi H, et al. Genomic Features of Response to Combination Immunotherapy in Patients with Advanced Non-Small-Cell Lung Cancer. Cancer Cell 2018;33:843-852.e4. [Crossref] [PubMed]
20. Sinha N, Sinha S, Valero C, et al. Immune Determinants of the Association between Tumor Mutational Burden and Immunotherapy Response across Cancer Types. Cancer Res 2022;82:2076-83. [Crossref] [PubMed]
21. Castellanos E, Snider J, Ali SM, et al. Tumor mutational burden and PD-L1 expression as predictors of response to immunotherapy in NSCLC. ASCO Annual Meeting 2019;37:2630.
22. Ashok Kumar P, Karimi M, Basnet A, et al. Association of Molecular Profiles and Mutational Status With Distinct Histological Lung Adenocarcinoma Subtypes. An Analysis of the LACE-Bio Data. Clin Lung Cancer 2023;24:528-40. [Crossref] [PubMed]
23. Jiang Z, Zhou Y, Huang J. A Combination of Biomarkers Predict Response to Immune Checkpoint Blockade Therapy in Non-Small Cell Lung Cancer. Front Immunol 2021;12:813331. [Crossref] [PubMed]
24. Mei T, Wang T, Zhou Q. Multi-omics and artificial intelligence predict clinical outcomes of immunotherapy in non-small cell lung cancer patients. Clin Exp Med 2024;24:60. [Crossref] [PubMed]
25. Pan Y, Zhang JT, Gao X, et al. Dynamic circulating tumor DNA during chemoradiotherapy predicts clinical outcomes for locally advanced non-small cell lung cancer patients. Cancer Cell 2023;41:1763-1773.e4. [Crossref] [PubMed]