Evaluation of the Molecular Landscape in PD-L1 Positive Metastatic NSCLC: Data from Campania, Italy

Background: Immune-checkpoint inhibitors (ICIs) have increased and improved the treatment options for patients with non-oncogene-addicted advanced stage non-small cell lung cancer (NSCLC). However, the role of ICIs in oncogene-addicted advanced stage NSCLC patients is still debated. In this study, in an attempt to fill in the informational gap on the effect of ICIs on other driver mutations, we set out to provide a molecular landscape of clinically relevant oncogenic drivers in programmed death-ligand 1 (PD-L1) positive NSCLC patients. Methods: We retrospectively reviewed data on 167 advanced stage NSCLC PD-L1 positive patients (≥1%) who were referred to our clinic for molecular evaluation of five driver oncogenes, namely, EGFR, KRAS, BRAF, ALK and ROS1. Results: Interestingly, n = 93 (55.7%) patients showed at least one genomic alteration within the tested genes. Furthermore, analyzing a subset of patients with PD-L1 tumor proportion score (TPS) ≥ 50% and concomitant gene alterations (n = 8), we found that n = 3 (37.5%) of these patients feature clinical benefit with ICIs administration, despite the presence of a concomitant KRAS gene alteration. Conclusions: In this study, we provide a molecular landscape of clinically relevant biomarkers in NSCLC PD-L1 positive patients, along with data evidencing the clinical benefit of ICIs in patient NSCLC PD-L1 positive alterations.


Introduction
Lung cancer represents the leading cause of cancer deaths worldwide [1]. About 85% of lung cancers are non-small cell lung cancer (NSCLC) [2,3]. In recent years, several efforts have been made to improve clinical outcomes of advanced stage NSCLC patients. Central to these efforts has been the advent of precision medicine. This approach, which involves the identification of actionable oncogenic driver alterations, has spurred the development of specific therapeutic agents capable of thwarting the molecular pathways involved in cancer progression. Among these agents are tyrosine kinase inhibitors (TKIs). Remarkably, these agents are able to target a long series of recently discovered oncogenetic mutations involved in several driver genes, namely, Epidermal Growth Factor Receptor (EGFR) [4][5][6][7], V-Raf Murine Sarcoma Viral Oncogene Homolog B1 (BRAF) [8,9], Kirsten Rat Sarcoma Viral Oncogene Homolog (KRAS) exon 2 p.G12C [10,11] and gene fusions in Anaplastic Lymphoma Receptor Tyrosine Kinase (ALK) [12][13][14][15][16] and ROS Proto-Oncogene 1, Receptor Tyrosine Kinase (ROS1) [17][18][19]. Another milestone in the clinical management of advanced stage NSCLC patients has been the development of immune-checkpoint inhibitors (ICIs) [20]. Currently, the evaluation of programmed death-ligand 1 (PD-L1) expression levels is the most widely adopted and standardized tool for ICI administration [21,22]. ICIs have indeed increased and improved the treatment options for non-oncogene-addicted advanced stage NSCLC patients [23][24][25][26]. However, the role of ICIs in oncogene-addicted advanced stage NSCLC patients is still debated [27]. For example, a recent review has highlighted the lack of efficacy of pembrolizumab in naïve EGFR-mutated advanced stage NSCLC patients expressing low levels of PD-L1 (1%) [28]. However, even less is known about the effect of ICIs on other clinically relevant biomarkers. Undoubtedly, paucity of data in this specific field is a major setback for lung cancer treatment. Indeed, evaluating PD-L1 expression levels and the genomic assessment of clinically relevant oncogenic targetable drivers would be crucial to broaden the treatment options for NSCLC patients. In our referral laboratory experience at the Molecular Predictive Pathology Laboratory at the Department of Public Health of the University of Naples Federico II, we routinely perform immunohistochemistry/immunocytochemistry (IHC/ICC) to evaluate PD-L1 expression [29,30]. In addition, we perform both DNA-based next generation sequencing (NGS) and fully automated realtime polymerase chain reaction (RT-qPCR), namely, Idylla™ (Biocartis, Mechelen, Belgium) to evaluate point mutations, deletions and insertions [31][32][33] and IHC/ICC and RNA-based NGS analysis to identify gene fusions [34].
In this study, in an attempt to fill in the informational gap on the effect of ICIs on other driver mutations, we set out to provide a molecular landscape of clinically relevant oncogenic drivers in PD-L1 positive NSCLC patients. To this aim, we retrospectively evaluated data collected from our archives of advanced stage NSCLC patients with positive PD-L1 expression (≥1%) who were referred to our clinic for evaluation of at least five of the most common driver mutations, namely, EGFR, KRAS, BRAF, ALK and ROS1. In addition, in a subset of patients, we were also able to retrieve information about patients' medical treatments and performance status.
Results are summarized in Table 1, Figures 1-3 and Supplementary Table S1.
Results are summarized in Table 1, Figures 1-3 and Supplementary Table S1.
Results are summarized in Table 1, Figures 1-3 and Supplementary Table S1. Results are summarized in Table 1, Figures 1-4 and Supplementary Table S1.
Results are summarized in Table 1, Figures 1-3 and Supplementary Table S1.
Regarding the evaluation of ALK and ROS1 gene rearrangements, IHC/ICC was employed in the vast majority of cases (n = 74, 89.2%), whereas RNA-based NGS analysis was adopted in only n = 9 (10.8%) instances. Interestingly, whereas ALK fusions were identified in n = 4 (4.8%) cases, ROS1 fusions were detected in only n = 1 (1.2%) case. As for the biomarker analyses, at least one genomic alteration was detected in n = 50 (60.2%) cases, whereas no concomitant clinically relevant biomarker alterations were detected in the remaining n = 33 (39.8%) cases.
Results are summarized in Table 1, Figures 1-3 and Supplementary Table S1.
Results are summarized in Table 2.
In this setting, despite the role of ICIs has been clearly demonstrated in the treatment of high PD-L1 expressers [23], little is known about the role of concomitant genomic alterations on this regimen. Lee et al. showed that ICI administration in KRAS mutated patients may determine an overall survival (OS) benefit respect to KRAS wild-type patients [41]. Similarly, Bodor et al. highlighted that KRAS-mutated NSCLC patients with PD-L1 TPS ≥1% had a longer progression-free survival respect to PD-L1 negative patients (4.1 vs. 3.2 months, p = 0.001) [42]. A possible explanation may be the presence of a specific interaction between the tumor microenvironment and ICIs for this specific subset of patients as demonstrated by Falk et al. [43]. Similarly, the adoption of front-line pembrolizumab in PD-L1 positive advanced stage NSCLC patients harboring a KRAS exon 2 p.G12C point mutation seemed to be predictive of higher objective response rate (ORR, 57% versus 29%), median progression free survival (PFS, 12 versus 6 months) and OS (28 versus 15 months) [44]. Different from KRAS exon 2 p.G12C, the identification of other concomitant driver mutations is predictive of poor response to ICIs administration in the PD-L1 positive population [27]. The limited efficacy of ICIs in patients harboring EGFR mutations has been widely demonstrated [45]. In a phase II study, Lisberg et al. highlighted the absence of response to pembrolizumab as first line approach in advanced stage PD-L1 positive EGFR-mutant NSCLC patients naïve to TKI administration [28]. Similar data have been reported for other ICI drugs, such as atezolizumab and durvalumab [46,47]. The role of ICIs is controversial in BRAF-mutated patients [48]. In fact, in a multicentric retrospective cohort, Dudnik et al. showed promising data in terms of clinical efficacy of ICIs in BRAFmutated advanced stage NSCLC [49]. Conversely, in a small retrospective study, Tan et al. highlighted an inferior OS in BRAF-mutated patients receiving ICI respect to those treated with front-line chemotherapy [50]. Regarding gene rearrangements, a very limited efficacy of ICIs in ALK- [47,[51][52][53][54], ROS1- [55,56], RET- [57] and NTRK-rearranged [27] NSCLC patients has been highlighted. Considering MET exon 14 skipping, despite some evidence reporting response to ICIs [58], the overall efficacy of immunotherapy respect to target therapy is quite modest [59].
In conclusion, in this study we have provided a real-world practice experience on the molecular landscape of clinically relevant biomarkers in NSCLC PD-L1-positive patients. The most significant limitations of our study were the limited number of cases, the absence of molecular data on PD-L1 negative patients, the limited number of gene alterations analyzed and clinical data on progression-free survival and overall survival and the lack of clinical data on the vast majority of patients. Further studies are thus needed to better assess the role of the complex genomic landscape in advanced stage NSCLC patients.

Study Design
In this study, we retrospectively reviewed cases referred to our clinic from 1 January 2018 to 30 June 2021 for molecular evaluation of at least five driver druggable oncogenes, namely, EGFR, KRAS, BRAF, ALK, ROS1 and PD-L1 expression assessment; PD-L1 positive cases (expression in ≥1% tumor cells) were selected. Information regarding sex, median age, sample type and subtype and diagnosis was also retrieved. ( Figure 5) Furthermore, for a subset of patients, data related to the duration of the first-line treatment, or until the loss of data for any causes, were also gathered.
data have been reported for other ICI drugs, such as atezolizumab and durvalumab [46,47]. The role of ICIs is controversial in BRAF-mutated patients [48]. In fact, in a multicentric retrospective cohort, Dudnik et al. showed promising data in terms of clinical efficacy of ICIs in BRAF-mutated advanced stage NSCLC [49]. Conversely, in a small retrospective study, Tan et al. highlighted an inferior OS in BRAF-mutated patients receiving ICI respect to those treated with front-line chemotherapy [50]. Regarding gene rearrangements, a very limited efficacy of ICIs in ALK- [47,[51][52][53][54], ROS1- [55,56], RET- [57] and NTRK-rearranged [27] NSCLC patients has been highlighted. Considering MET exon 14 skipping, despite some evidence reporting response to ICIs [58], the overall efficacy of immunotherapy respect to target therapy is quite modest [59].
In conclusion, in this study we have provided a real-world practice experience on the molecular landscape of clinically relevant biomarkers in NSCLC PD-L1-positive patients. The most significant limitations of our study were the limited number of cases, the absence of molecular data on PD-L1 negative patients, the limited number of gene alterations analyzed and clinical data on progression-free survival and overall survival and the lack of clinical data on the vast majority of patients. Further studies are thus needed to better assess the role of the complex genomic landscape in advanced stage NSCLC patients.

Study Design
In this study, we retrospectively reviewed cases referred to our clinic from 1 January 2018 to 30 June 2021 for molecular evaluation of at least five driver druggable oncogenes, namely, EGFR, KRAS, BRAF, ALK, ROS1 and PD-L1 expression assessment; PD-L1 positive cases (expression in ≥1% tumor cells) were selected. Information regarding sex, median age, sample type and subtype and diagnosis was also retrieved. ( Figure 5) Furthermore, for a subset of patients, data related to the duration of the first-line treatment, or until the loss of data for any causes, were also gathered.  All information regarding human material was managed using anonymous numerical codes, and all samples were handled in compliance with the Declaration of Helsinki (http://www.wma.net/en/30publications/10policies/b3/, last accessed on 30 June 2022).

IHC/ICC Analysis
PD-L1 IHC/ICC evaluation was performed with a validated laboratory developed test (LDT), consisting of the use of Dako's concentrate 22C3 anti-PD-L1 primary antibody with a Ventana's detection systems on the BenchMark XT platform, or by using the companion diagnostic kit SP263 assay (Ventana Medical Systems, Tucson, AZ, USA) [29,30]. The level of PD-L1 expression was determined by using tumor proportion score (TPS). PD-L1 positive cases were classified either as low-positive PD-L1 expression (1-49%) or as high-positive PD-L1 expression (≥50%) [29,30].
ALK IHC/ICC evaluation was performed by using the Ventana ALK D5F3 companion diagnostic (CDx) assay (Ventana Medical Systems) together with the OptiView (Ventana) detection system. The latter system features a tyramide-based amplification phase in addition to the polymeric step. In particular, by increasing the signal difference between the specific immunoreaction of neoplastic cells and the background, the amplification phase significantly reduces equivocal results. Thus, only positive or negative ALK cases can be reported. Typically, only strong and granular cytoplasmic signals are scored as positive, regardless of the percentage of stained neoplastic cells [60][61][62].
ROS1 IHC/ICC evaluation was carried out with the D4D6 (Cell Signaling Technology, Inc., Danvers, MA, USA) clone. Generally, only tumors with a moderate-to strong staining intensity signal (2+ or 3+ scores) in more than half of the neoplastic cells are considered positive [60,63,64].
Finally, ALK and ROS1 IHC/ICC assays were adopted to confirm RNA-based NGS positive cases.

Molecular Testing
DNA-and RNA-based analyses of samples were carried out. DNA-based NGS analysis was performed with our narrow NGS panel, namely, SiRe ® [65]; this panel was designed, developed and validated in the Molecular Predictive Pathology Laboratory of the Department of Public Health at the University of Naples Federico II [65]. SiRe ® can simultaneously detect multiple hotspot gene alterations in seven genes (EGFR, KRAS, BRAF, NRAS, KIT, PDGFRα, and PIK3CA) [31,65]. In the present study, only variants with allele coverage >20X and a quality score >20, with an amplicon coverage of at least 500X alleles, were called.
RNA-based NGS analysis was performed with a narrow NGS panel, namely, SiRe fusion [34]. This panel was also designed, developed, and validated in the Molecular Predictive Pathology Laboratory of the Department of Public Health at University of Naples Federico II [34]. It simultaneously detects alterations in six oncogenic genes, namely, ALK, ROS1, RET, NTRK gene rearrangements, MET exon 14 skipping alterations [34]. In all the study cases, ALK and ROS1 status was further confirmed with IHC/ICC.
In a limited number of cases, the fully automated Idylla™ RT-qPCR platform was adopted to evaluate the molecular status of EGFR, KRAS and BRAF [32,33,66].  Institutional Review Board Statement: Ethical review and approval were waived for this study due to written informed consent was obtained from all patients and documented in accordance with the general authorisation to process personal data for scientific research purposes from 'The Italian Data Protection Authority' (http://www.garanteprivacy.it/web/guest/home/docweb/-/ docwebdisplay/export/2485392). All information regarding human material was managed using anonymous numerical codes, and all samples were handled in compliance with the Helsinki Declaration (http://www.wma.net/en/30publications/10policies/b3/).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.