Molecular Profile of Advanced Non-Small Cell Lung Cancers in Octogenarians: The Door to Precision Medicine in Elderly Patients

Background: There is a pressing need to expand the evidence base in geriatric lung oncology. Most non-small cell lung cancers (NSCLCs) are diagnosed in the elderly, with approximately 15% of cases affecting octogenarians. Treatment-related decisions are challenging in this population, and the role of biologically driven therapies is still underrated. Methods: A single-institution cohort of 76 NSCLCs from octogenarian patients was submitted to molecular analysis using a next-generation sequencing (NGS) multigene panel, fluorescence in situ hybridization (FISH) analyses, and immunohistochemistry for PD-L1 assessment. Treatment and clinical outcome data were available for 33 patients. Results: Most cases (n = 66, 87%) harbored at least one genomic alteration. EGFR and KRAS mutations were detected in 18 (24%) and 20 (26%) patients, respectively. No ALK alterations were found, but in two patients ROS1 translocation was identified. Of 22 cases tested, 17 were positive for PD-L1 staining. Octogenarian patients who received tyrosine kinase inhibitors (TKIs) based on molecular analysis showed clinical benefits, with long progression-free survival as expected in TKI-treated younger cohorts. Conclusions: This study highlights the utility of molecular profiling in all advanced-stage NSCLCs, regardless of the age at diagnosis, to drive personalized treatment. The prevalence of druggable alterations and the clinical benefits obtained by biologically-driven therapies in octogenarians were comparable to those of the younger NSCLC population.


Introduction
Lung cancer is the leading cause of cancer-specific mortality worldwide [1], and is typically a disease of the elderly. The median age at diagnosis is 70 years, and almost 10% of patients are older than 84 years [2]. Since the geriatric population has been increasing, the management of elderly patients affected by lung cancer is of growing concern [3]. Old patients with cancer usually have comorbidities and age-related physiological characteristics that make the choice of treatment challenging. Systemic cytotoxic chemotherapy, which has been the standard treatment of advanced lung cancer for decades, tends to be less-well-tolerated by elderly patients than by younger ones. The development of specific therapies targeting driver alterations (e.g., EGFR mutations or ALK and ROS1 rearrangements) has changed the treatment paradigm and natural history of non-small cell lung cancer (NSCLC) harboring these aberrations [4,5]. To date, limited data are available regarding the safety and efficacy of these agents in the elderly population, and above all in octogenarian patients, since they are underrepresented in clinical trials [6,7]. Nevertheless, in clinical practice, the evaluation of tumor molecular features together with the clinical characteristics of octogenarian patients with NSCLC may broaden the treatment options and drive a tailored clinical management of these patients.
In the present study, we report the molecular characterization of advanced NSCLC from 76 consecutive octogenarian patients who were referred to our institution over 19 months for molecular diagnosis, following clinical requests. The molecular testing was performed using a next-generation sequencing (NGS) panel including EGFR, KRAS, and BRAF genes, in addition to fluorescence in situ hybridization (FISH) analyses for ALK, ROS1, MET, and immunohistochemical (IHC) evaluation of PD-L1 expression. We sought to (i) evaluate the biological characteristics of NSCLC occurring in octogenarian patients, which could lead to personalized treatments; and (ii) investigate the potential clinical benefit of molecularly driven therapies in a subgroup of patients treated with targeted treatment.

Patients
Of the 890 patients with advanced-stage (IIIb and IV) NSCLCs referred to the Molecular Diagnostics Unit of the European Institute of Oncology from August 2016 to February 2018, 76 patients over 80 years were included in this study. Molecular tests and IHC evaluation were required by thoracic oncologists, and included the analysis of EGFR, KRAS, BRAF, ALK, ROS1, MET, and PD-L1 according to the international guidelines for molecular testing in lung cancer [8,9]. Patients' age, gender, smoking history, performance status (PS), and treatment regimens were collected from medical records and are summarized in Table 1. Based on their smoking status, patients were categorized as smokers, recent ex-smokers (more than 6 months but less than 15 years), long-term ex-smokers (more than 15 years), and never-smokers (Table 1). Thirty-three patients (43%) were inpatients and treated at the Division of Thoracic Oncology of the European Institute of Oncology. PS at diagnosis was evaluated according to the Eastern Cooperative Oncology Group (ECOG) criteria [10]. Patients' treatments included chemotherapy (vinorelbine, pemetrexed, or gemcitabine), local radiotherapy, and anti-EGFR tyrosine kinase inhibitor (TKI) (erlotinib, gefitinib, or osimertinib) ( Table 1). The best response to treatment was evaluated according to Response Evaluation Criteria in Solid Tumors (RECIST) [11].
All patients gave written informed consent regarding the storage of any biological specimens collected in the course of diagnosis and the use of these samples for research purposes.

Fluorescence in Situ Hybridization Analysis
ALK and ROS1 gene rearrangements and MET amplification were evaluated by the standard FISH method. Briefly, unstained sections obtained from FFPE blocks or cytoblocks were incubated with an ALK and ROS1 dual-color probe (IQFISH Break Apart Probe Agilent Technologies, Santa Clara, CA, USA). In each case, at least 100 tumor nuclei were evaluated. Cells were considered positive if a break-apart pattern of orange and green signals, at least one single orange signal, or a combination of both patterns were seen. Tumors with at least 15% of cells with ALK or ROS1 rearrangements were defined as positive. In ambiguous or equivocal cases, ALK or ROS1 immunohistochemical stains (clone D5F3, Ventana, Tucson, AZ, USA and clone D4D6, Cell Signaling, Danvers, MA, USA, respectively) were performed. The presence of MET gene amplification was evaluated using the MET IQFISH Probe with CEP7 (Agilent Technologies, Santa Clara, CA, USA). Amplification was reported in cases with a MET Probe/CEP7 Ratio ≥2 and/or gene copy number ≥5.

Molecular Profile of Advanced Non-Small Cell Lung Cancer in Octogenarian Patients
In our study population, 66 of 76 (87%) cases harbored at least one gene alteration, including clinically relevant and cancer-related variants. The most frequently mutated genes were TP53 (37%), KRAS (26%), and EGFR (24%) ( Figure 1 and Table 2

Molecular Profile of Advanced Non-Small Cell Lung Cancer in Octogenarian Patients
In our study population, 66 of 76 (87%) cases harbored at least one gene alteration, including clinically relevant and cancer-related variants. The most frequently mutated genes were TP53 (37%), KRAS (26%), and EGFR (24%) ( Figure 1 and Table 2).   Level I and level II alterations, defined by high clinical relevance for prognosis and/or target therapy available in a clinical setting (level I) or clinical trials (level II), accounted for 62/106 (58.5%) of genomic aberrations identified and involved EGFR, ROS1, KRAS, BRAF (1 case V600E), MET, NRAS, ERBB2, PIK3CA, and AKT1 genes ( Figure 1 and Table 2).
In detail, KRAS and EGFR mutations were reported in 20 and 18 NSCLC specimens, respectively. All KRAS mutations identified affected exon 2.
Mutations in EGFR included both TKI-sensitive mutations (n = 16), such as exon 19 deletions (n = 8) and exon 21 L858R point mutation (n = 8); and TKI-resistant mutations, such as exon 20 insertions (n = 2). In four cases, a secondary T790M resistant mutation was detected during TKI treatment. Concurrent EGFR, T790M, and L858R mutations were present at diagnosis in one patient with no history of TKI therapy.
No ALK fusions were detected in our study cohort. ROS1 gene rearrangements were identified in two cases, with a concurrent NRAS mutation or KRAS mutation.
Overall concurrent genetic aberrations were found in 33 cases (Figure 1), including TP53 mutations with either EGFR (n = 7) or KRAS (n = 4) mutations. Moreover, EGFR mutations were found in association with alterations of MET (n = 2) and STK11 (n = 1) genes. Concurrent KRAS and STK11 mutations were identified in two cases.

Treatment Regimens and Clinical Outcome of Octogenarian Patients with Advanced Non-Small Cell Lung Cancers
Treatment and follow-up data were available for thirty-three inpatients. Twenty-three of these patients received an active treatment ( Table 2). In ten cases, only radiotherapy or best supportive care was given because of the presence of comorbidities or poor performance status, including four patients with a PS of two, five patients with a PS of three, and one patient with a PS of one who had previously undergone right nephrectomy for renal cell carcinoma. Data about targeted treatments were available only for patients with EGFR mutations. The other actionable alterations (in ROS1, ERBB2, MET, and BRAF) were found in NSCLC outpatients or inpatients with a poor PS, not eligible for enrollment in a clinical trial or active treatment.
Among the patients with EGFR mutations, the majority were never-smokers or long-term ex-smokers (n = 13/18 cases, 72.2%). Smoking habits were not assessed in the remaining cases. A different distribution between males and females was not seen (8 males and 12 females). Treatment and follow-up data were available for ten patients who received tyrosine kinase inhibitors (TKIs), including erlotinib, gefitinib, or osimertinib, when a concurrent EGFR T790M mutation was detected. One patient was first-line treated with osimertinib because of the co-occurrence of L858R and T790M mutations at diagnosis (patient #3, Figure 2 and Table 3). The median progression-free survival (PFS) for patients receiving TKI as first-line therapy was 17 months, including five patients in whom treatment was ongoing ( Figure 2). One patient died of progressive disease after 4 months of treatment. During first-line TKI therapy, two patients developed the EGFR T790M resistance mutation and received osimertinib as second-line treatment. The best responses observed were complete responses in two patients (#7 and #8, Figure 2), reached after four months of gefitinib and 5 months of osimertinib, respectively. Both patients were still receiving treatment after 14 months of gefitinib and 15 months of osimertinib, respectively. Thirteen patients received chemotherapy-based regimens, including metronomic vinorelbine (n = 10) or platinum-based therapy (n = 3). The median PFS of these patients was 2 months (Figure 2).

Discussion
Treatment-related decisions in octogenarians affected by NSCLC are challenging: the elderly could have reduced tolerance of treatments because of limited physiological reserve or multiple comorbid conditions, but they could also be under-treated based solely on biological age [6]. Case-by-case clinical assessment in addition to tumor molecular analyses could lay the basis for tailored treatment choices. However, few data are available regarding the real-life management of octogenarian patients with NSCLC.
In the present study, we focused on a single-institution cohort of 76 consecutive octogenarian patients affected by advanced NSCLC. All tumors were analyzed using an NGS multi-gene panel that expands the tumor genetic portrait, including genes such as BRAF, ERBB2, KRAS, and MET in addition to EGFR and ALK/ROS1, in accordance with recently updated recommendations [8,9]. Most cases (86.8%) harbored at least one genetic alteration, and in only ten cases no genomic aberration was detected in the genes tested. These data were consistent with our recent findings obtained in a larger cohort of NSCLC patients, not selected by age at diagnosis [12]. Indeed, among 535 NSCLC patients, we found that 82.4% harbored at least one NGS-detected mutation, compared to 85.5% within this study cohort. In the present study, we reported level I and II variants in 65.8% of the cases, including EGFR (24%) and KRAS (26%) mutations. In our previous series, level I and II variants were detected in 63.6% of cases, with KRAS and EGFR mutations present in 31% and 22% of the cases, respectively [12]. The differences in mutation distribution between the two series were not statistically significant ( 2 test p > 0.05) and showed that the presence of specific tumor genetic aberrations was not related to the age of the patients. Conversely, ALK fusions were not detected in our cohort, consistent with the reported association between ALK translocation and the patients' younger age [17,18]. Surprisingly, we found ROS1 translocations in two patients. This rarest druggable alteration has been reported in 1%-2% of NSCLCs [19,20], usually affecting never-smoker young female patients [19]. In our cohort, two octogenarian males harbored a ROS1 translocation, and both had concurrent clinically relevant mutations (ROS1 + /KRAS + and ROS1 + /NRAS + ). Similar

Discussion
Treatment-related decisions in octogenarians affected by NSCLC are challenging: the elderly could have reduced tolerance of treatments because of limited physiological reserve or multiple comorbid conditions, but they could also be under-treated based solely on biological age [6]. Case-by-case clinical assessment in addition to tumor molecular analyses could lay the basis for tailored treatment choices. However, few data are available regarding the real-life management of octogenarian patients with NSCLC.
In the present study, we focused on a single-institution cohort of 76 consecutive octogenarian patients affected by advanced NSCLC. All tumors were analyzed using an NGS multi-gene panel that expands the tumor genetic portrait, including genes such as BRAF, ERBB2, KRAS, and MET in addition to EGFR and ALK/ROS1, in accordance with recently updated recommendations [8,9]. Most cases (86.8%) harbored at least one genetic alteration, and in only ten cases no genomic aberration was detected in the genes tested. These data were consistent with our recent findings obtained in a larger cohort of NSCLC patients, not selected by age at diagnosis [12]. Indeed, among 535 NSCLC patients, we found that 82.4% harbored at least one NGS-detected mutation, compared to 85.5% within this study cohort. In the present study, we reported level I and II variants in 65.8% of the cases, including EGFR (24%) and KRAS (26%) mutations. In our previous series, level I and II variants were detected in 63.6% of cases, with KRAS and EGFR mutations present in 31% and 22% of the cases, respectively [12]. The differences in mutation distribution between the two series were not statistically significant ( 2 test p > 0.05) and showed that the presence of specific tumor genetic aberrations was not related to the age of the patients. Conversely, ALK fusions were not detected in our cohort, consistent with the reported association between ALK translocation and the patients' younger age [17,18]. Surprisingly, we found ROS1 translocations in two patients. This rarest druggable alteration has been reported in 1%-2% of NSCLCs [19,20], usually affecting never-smoker young female patients [19]. In our cohort, two octogenarian males harbored a ROS1 translocation, and both had concurrent clinically relevant mutations (ROS1 + /KRAS + and ROS1 + /NRAS + ). Similar anecdotal cases of concurrent ROS1 rearrangements and KRAS mutations were recently reported [21,22], whereas no cases of ROS1 translocations and NRAS variants have been described so far. ROS1 gene rearrangements could be targeted by TKIs, but KRAS/NRAS gene point mutations could lead to ineffective inhibition through inactivation of the RAF/MEK/ERK signaling pathway. Given that these patients were both outpatients, no data about therapeutic regimens or clinical outcomes were available.
In a subgroup of octogenarian patients, we also identified alterations in MET, BRAF, and ERBB2 genes, which may be targeted by specific therapies in a clinical setting or clinical trials [23,24]. Moreover, we reported level III mutations that could be clinically significant. TP53 mutations have recently been associated with prognosis or therapy response in patients with concomitant EGFR or KRAS mutations, or ALK rearrangement [25]. Patients whose tumors harbored concurrent EGFR and TP53 mutations showed a worse response rate to TKIs and worse progression-free survival [26]. KRAS + -TP53 + tumors displayed a higher level of inflammatory markers and improved relapse-free survival [27].
In the octogenarians included in this study, EGFR mutated status was associated with non-smoking, as in the general population of NSCLC patients. A TKI treatment was administered to ten EGFR mutation-positive patients, who showed a PFS consistent with data from larger series of NSCLC patients [28][29][30][31][32][33][34][35]. Although these series included patients ranging from 24 to 88 years old, the elderly were underrepresented and specific data regarding octogenarians are limited. Previous studies have reported on the safety and efficacy of TKIs in elderly patients, as shown by osimertinib treatment for T790M-positive NSCLC [36] or erlotinib treatment for NSCLC after one or more platinum-based chemotherapy regimens [37]. However, the definition of "elderly" is highly heterogeneous among these studies, including patients ranging from 65 [37] to 75 years old [36]. Recently, a multicenter real-world study (OCTOMUT) evaluated the tolerability and efficacy of TKIs in Caucasian octogenarian patients with EGFR-mutated NSCLC [38]. The authors reported clinical outcomes and toxicity profiles comparable to those in younger patients. The median PFS observed in that retrospective study was 11.9 months for patients treated with TKIs. Although the number of patients with follow-up data was limited in our cohort, we observed a prolonged response in six of ten patients treated with TKI therapy. Moreover, seven patients are receiving ongoing treatment.
During the past few years, the spectrum of therapeutic options for NSCLC patients with no identified genomic aberrations has been enlarged. Besides standard chemotherapy regimens, treatments with immune checkpoint inhibitors alone or in combination with chemotherapy have been shown to be effective in selected patients with advanced NSCLC enrolled in clinical trials. However, in the largest series, elderly patients were underrepresented, especially octogenarians [39]. The ELDERS Study is a prospective pilot study evaluating the role of checkpoint inhibitors in patients ≥70 years old [40]. Although there are some promising results, the real benefit of immunotherapy in this specific population is still under debate. In this study population, only a few tumors were tested for tumor PD-L1 expression (the only approved biomarker to select patients for immunotherapy in our country) and were found to have the whole range of TPS scores. Retrospective collections of clinical data from a large series of elderly individuals who received immunotherapies and, above all, prospective studies focused on NSCLC patients ≥80 years old are warranted in order to clarify the role of immunotherapy in the octogenarian population.
A major limitation of the present study is the relatively small cohort of patients analyzed. However, we collected consecutive octogenarian patients affected by NSCLC who underwent tumor molecular analyses for diagnostic purposes in a single institution. Another limitation is that treatment and follow-up data were not available for all patients and were collected retrospectively from patients' medical records. We reported data for thirty-three inpatients of the Division of Thoracic Oncology of the European Institute of Oncology, including twenty-three patients who received molecularly driven treatments. Although the small cohort prevented identification of any clinical or statistical correlations and the retrospective analysis was not selected for the probability of survival, we reported interesting insights into the clinical course of octogenarian patients actively treated based on tumor molecular characteristics.

Conclusions
This study highlights the utility of molecular analyses for all patients with advanced-stage NSCLC, regardless of age at diagnosis, to guide clinical decisions. Except for ALK translocation, the distribution of molecular alterations in the octogenarian population under investigation was very similar to that in a younger population. Molecular analyses should be performed with multi-gene panels, following the recent recommendations for molecular testing in NSCLC [8,9], to widen the number of predictive markers that may be detected. Indeed, most of the NSCLCs analyzed in this study harbored at least one actionable or clinically-relevant gene alteration. This information may guide clinical decisions about active treatments for these patients with clinical benefits. In the era of precision medicine, age itself may not represent a contraindication to evaluate the molecular characteristics of a tumor. Rather, a comprehensive assessment of clinical, pathological, and molecular features of octogenarian patients with NSCLC may improve the selection of tailored and effective treatments for this population.