ALK-positive advanced non-small cell lung cancer patients with poor performance status: Outcomes in a real-world scenario

Background Anaplastic lymphoma kinase (ALK) inhibitors have shown significant efficacy in ﻿ALK -rearranged non-small cell lung cancer (NSCLC) patients with good performance status (PS) in multiple randomised studies. However, there is limited data on patients with poor performance status. Patients and methods We carried out a retrospective analysis of prospectively collected data of patients with ALK-rearranged NSCLC and Eastern Cooperative Oncology Group (ECOG) PS of 2–4 treated at a single academic cancer centre from January 2013 to November 2018. The outcomes, progression-free survival (PFS) and overall survival (OS) were calculated from the date of diagnosis. SPSS version 20 was used for all statistical calculations. Results Out of the total 441 ALK-positive patients, 97 (21.9%) had ECOG PS 2–4 (poor PS). The median PFS was 9.3 months (95% CI = 6.6–12.0) as compared to 14.9 months (95% CI = 13.4–16.4) for patients with a PS of 0–1 (HR = 1.38, 95% CI = 1.04–1.84, p = 0.027). The corresponding median OS were 17.9 months (95% CI = 12.8–23.1) and 33.5 months (95% CI = 28.6–38.4), respectively (HR = 1.89, 95% CI = 1.36–2.62, p < 0.001). Among poor PS patients, a subgroup of patients with PS 2 had median OS of 20.6 months (95% CI = 10.8–47.3) as compared to 8.6 months for PS 3–4 (95% CI = 7.8–27.8) (HR = 1.79, 95% CI = 1.01–3.20, p = 0.047). The patients treated with upfront ALK inhibitors had better survival as opposed to those treated with chemotherapy. On multivariate analysis, PS 3–4, smoking, stage 4 and not using ALK inhibitors as first-line therapy were associated significantly with poor outcomes. Conclusion The ALK-rearranged NSCLC patients with poor PS derived significant benefits with ALK inhibitors. The outcomes were significantly poorer as compared to patients with PS 0–1; the subgroup of patients with PS 2 had better outcomes as compared to patients with PS 3–4.


Introduction
Poor performance status (PS) is one of the most important prognostic factors in non-small cell lung cancer (NSCLC). Very few studies have included NSCLC patients with poor PS in chemotherapy or targeted therapy trials. In a study by Rogerio et al [1], NSCLC had the highest prevalence of poor PS. It showed that poor PS prevalence was 34% when estimated by providers and 48% when estimated by the patients themselves.
Anaplastic lymphoma kinase (ALK) rearrangements are the most common targetable oncogenic driver mutations found in never smokers who are wild-type for epidermal growth factor receptor mutation [2]. In the past, before the availability of ALK inhibitors, most patients were treated with conventional chemotherapy with platin-based regimens. The chemotherapy studies included very few patients with PS 2, as chemotherapy is known to have significantly higher toxicities as compared to various ALK inhibitors, especially in the PS 2 subgroup [3][4][5].
In today's era, multiple ALK inhibitors, viz. crizotinib, ceritinib, alectinib, brigatinib and lorlatinib, have been approved. They have shown significant improvement in overall response rates (ORR), progression-free survival (PFS) and overall survival (OS) as compared to chemotherapy as the first-line or subsequent lines of therapy with less toxicity as compared to conventional chemotherapy [6][7][8][9] In most of these trials, only patients with good PS have been included. Very few studies included a limited number of patients with PS 2. There are very few prospective studies for the treatment outcomes in patients with poor PS, especially PS 3 or 4. Thus, it becomes difficult to have an evidence-based approach for these patients. In these situations, the real-world retrospective data can come to the rescue and help in the decision-making in these situations. Considering this, we carried out a retrospective analysis of the prospectively collected patient database of ALK-positive advanced lung cancer with poor PS.

Patients and methods
This study is a retrospective analysis of prospectively collected data of patients treated at a single academic cancer centre from January 2013 to November 2018. Out of all the patients from this database, patients with ALK rearrangement-positive advanced lung cancer having poor PS, defined as Eastern Cooperative Oncology Group (ECOG) PS 2-4, who were registered between January 2013 and September 2018, were included in this study. The patients who were not found to harbour ALK rearrangement were excluded. No sample size calculation was carried out as all the patients who were registered in the lung audit during the specified time period were evaluated for the study. The data included baseline demographics, comorbidities, habits, ALK testing method and disease sites, in addition to the dates of diagnosis, the start of first-line therapy, progression, the start of a subsequent line of treatment, last follow-up and death (if occurred). The best response to the treatment and sites of progression was also recorded. The primary outcome of this study was the estimation of PFS, while the secondary outcomes included OS and toxicities. The details of pre-treatment evaluation and ALK testing have been mentioned in detail in our previous publication [10].
Response assessment was carried out every 8-12 weeks of therapy or earlier as per the treating physician's discretion if clinically indicated and measured as per RECIST version 1.1 criteria. PFS was defined as the duration between the date of starting treatment and the date of disease progression or change in therapy or death, if it occurred before disease progression. OS was defined as time duration from the date of diagnosis of advanced-stage disease to date of death or last follow-up date. The data were analysed using SPSS for Windows version 20 (Armonk, NY, USA). Descriptive statistics were used to analyse demography data. The Kaplan-Meier method was used to calculate PFS and OS, and the comparisons of survival were made using the log-rank test. The Wilcoxon signed-rank test was carried out to check for the significance of change in ECOG PS post-first-line therapy.

0.049
Smoking and ECOG PS, on univariate analysis shown significant association with survival outcome, but not on multivariate analysis there was no statistical association. ALK inhibitor therapy has shown statistically significant association on both univariate and multivariate analysis Treatment details Figure 1 shows the consort diagram of the study; 43 (44%) patients received upfront TKI (41 crizotinib, 1 ceritinib and 1 alectinib), 18 (19%) patients received upfront crizotinib as switch after 1-2 cycles of pemetrexed-platinum-based chemotherapy, 6 (6%) patients received crizotinib as maintenance after 4-6 cycles of pemetrexed-platinum-based chemotherapy, 26 (26.8%) patients received upfront chemotherapy, while ALK TKI could be used in subsequent lines in seven of these patients. Two (2%) patients received only supportive care due to poor PS precluding any systemic therapy. Thus, ALK inhibitors could be used in 74 (76.3%) patients with poor PS in this study.
Univariate analysis of the various factors affecting OS found that ECOG PS 3-4 (versus PS 2), smokers and ALK inhibitors not received in first line led to significantly poorer outcomes. ECOG PS 3-4 was associated with HR = 1.79 (95% CI = 1.01-3.20, p = 0.047) as compared to PS 2; smokers had HR = 2.07 (95% CI = 1.0-4.3, p = 0.05) as compared to non-smokers, while the use of ALK inhibitors in first line was associated with HR = 0.49 (95% CI = 0.28-0.88, p = 0.014) as against not used in first line (Table 1). On multivariate analysis, the use of ALK inhibitors in the first line was the only significant factor with HR = 0.55 (95% CI = 0.31-0.99, p = 0.049).

Discussion
The prognosis of patients with ALK-rearranged NSCLC has improved in the last decade, and the median survival time is over 5 years in advanced stages with the use of third-generation ALK inhibitors [11]. Although chemotherapy is not recommended in patients with poor PS, it is feasible to use ALK inhibitors even in these patients. In this study, we report median OS of 21.4 months in patients with poor PS who could receive ALK inhibitors in any line. This is clearly much higher than any previously described results of palliative chemotherapy or supportive care alone [12,13]. In a phase 2 study (n = 18) by Iwama et al [14], alectinib showed an ORR of 72% and median PFS of 10.2 months in patients with PS 2-4. They found no significant difference in PS 2 versus PS 3-4. This is in contrast to our study in which there are significant differences in the two groups in terms of both PFS and OS. This can be explained by real-world data with intention-to-treat analysis in our study where all patients with poor PS are included, while in the study by Iwama et al [14], only patients with poor PS who received alectinib were included. The main concern with chemotherapy in poor PS patients remains the risk of increased toxicities [1]. However, in our study, we did not find any significant increase in toxicities with ALK inhibitors, despite the PS being 2-4. This is in agreement with the results of the use of alectinib in similar patients. Considering the available data, the National Comprehensive Cancer Network (NCCN) guidelines recommend the use of ALK inhibitors as first-line or subsequent therapy in patients with PS 0-4 [15].
The use of ALK inhibitors in first line came as the only significant factor in multivariate analysis for OS. The data of use in first line included patients who were initially started on chemotherapy but later shifted to ALK inhibitors after few cycles. This highlights the importance of using ALK inhibitors in first line instead of reserving them for second line or beyond. This is line with the data of PROFILE 1014 published in 2014, showing PFS at 18 months to be 31% (95% CI = 23-39) in the crizotinib arm and 5% (95% CI = 2-10) in patients who received chemotherapy in the first line [16]. This difference in PFS translated into OS benefit as well, with final analysis published in 2018 showing an improvement in OS favouring crizotinib (HR = 0.34; 95% bootstrap CI = 0.08-0.72) over chemotherapy after adjusting for cross-over [17].
Also, our study showed that the majority of poor PS patients had extensive metastatic disease at presentation which may be related to the EML4-ALK variant 3 [18]. This variant is associated with poor survival outcome, which may explain shorter OS and PFS in our poor PS population as opposed to patients with preserved PS at presentation. A limitation to our data is that the method of ALK testing was predominantly by IHC method and few by FISH method, hence it is difficult to make any final conclusion.
There are few important limitations of the study, which include the obvious drawbacks of a retrospective audit limiting the recording of toxicities, besides the treatment being heterogeneous. However, this study adds important real-world data in the treatment of ALK-rearranged NSCLC and gives confidence to the clinicians in prescribing ALK inhibitors to patients with poor PS at baseline.

Conclusion
ALK-rearranged NSCLC patients with poor PS derived significant benefits with ALK inhibitors. The outcomes were significantly poorer as compared to patients with PS 0-1; the subgroup of patients with PS 2 had better outcomes as compared to patients with PS 3-4.

Ethical statement
The study was conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonisation Guidelines for Good Clinical Practice.

Funding declaration
No funding was received for this study.