Brigatinib for Pretreated, ALK-Positive, Advanced Non-Small-Cell Lung Cancers: Long-Term Follow-Up and Focus on Post-Brigatinib Lorlatinib Efficacy in the Multicenter, Real-World BrigALK2 Study

Simple Summary This analysis assesses the efficacy of brigatinib, a next-generation ALK inhibitor in ALK+ advanced non-small cell lung cancer (aNSCLC) included in the brigatinib French Early-Access Program (1 August 2016–21 January 2019), with a focus on post-brigatinib lorlatinib efficacy. With a median follow-up of 40.4 months (95% CI, 38.4–42.4), the median investigator-assessed PFS of the 183 included patients was 7.4 months (5.9–9.6) and overall survival from brigatinib initiation was 20.3 (15.6–27.6) months. For patients who received 1 (n = 23), 2 (n = 146) or 3 (n = 14) ALKi(s) before brigatinib, the median overall survival was 33 (9.7—not reached), 20.3 (15.7–28.7) and 18.1 (3.3–24.5) months, respectively. Ninety-two (50.3%) patients received one agent(s) post-brigatinib; 68 (73.9%) of them received lorlatinib: 51 (75%) immediately post-brigatinib. With a median follow-up of 29.9 months (25.7–33.1), the median overall survival from lorlatinib initiation was 14.1 months (10.3–19.2). Analysis results confirmed brigatinib effectiveness in a population of heavily pretreated ALK+ positive aNSCLC patients and the activity of lorlatinib after brigatinib. We confirm that neither the manuscript nor any parts of its content are currently under consideration or published in another journal. All authors have approved the manuscript and agree with its submission to cancers. Abstract Brigatinib is a next-generation ALK inhibitor (ALKi) that shows efficacy in ALK inhibitor naïve and post-crizotinib ALK+ advanced NSCLCs (aNSCLCs). The efficacy of brigatinib was retrospectively assessed in patients with aNSCLCs included in the brigatinib French Early-Access Program (1 August 2016–21 January 2019). The primary endpoint was investigator-assessed progression-free survival (invPFS) and the primary analysis was updated in 2021 with a longer follow-up, focused on post-brigatinib lorlatinib efficacy. Sixty-six centers included 183 patients: median age 60 ± 12.7 years; 78.3% never/former smokers; median of 3 ± 1 previous lines and 2 ± 0.5 ALKis; 37.1% ECOG PS 2 and 55.6% >3 metastatic sites. The median follow-up from brigatinib initiation was 40.4 months (95% CI 38.4–42.4). InvPFS was 7.4 months (95% CI 5.9–9.6), median duration of treatment (mDOT) was 7.3 months (95% CI 5.8–9.4) and median overall survival (mOS) was 20.3 months (95% CI 15.6–27.6). The median DOT and OS from brigatinib initiation tend to decrease with the number of ALK inhibitors used in previous lines of therapy. Based on the data collected, 92 (50.3%) patients received ≥1 agent(s) post-brigatinib and 68 (73.9%) of them received lorlatinib, with 51 (75%) immediately receiving it post-brigatinib, 12 (17.6%) receiving it after one and 5 (7.4%) after ≥2 subsequent treatments. The median follow-up was 29.9 (95% CI 25.7–33.1) months. Lorlatinib mDOT was 5.3 (95% CI 3.6–7.6) months with a median OS from lorlatinib initiation of 14.1 (95% CI 10.3–19.2) months. The results of the brigALK2 study confirm the efficacy of brigatinib in a population of heavily pretreated ALK+ aNSCLC patients and provide new data on the activity of lorlatinib after brigatinib.

BrigALK2 covered the entire FEAP period; an updated analysis examined brigatinib efficacy. Subsequent treatments were collected, focusing on post-brigatinib lorlatinib efficacy.

Study Design and Patients
The objective of this non-interventional study was to evaluate, in the real-world setting, brigatinib efficacy in ALK + advanced NSCLC. Inclusion criteria were: 18 years old; ALK + advanced NSCLCs assessed by fluorescence in situ hybridization and/or immunohistochemistry or NGS in each participating center; prior treatment with 1 ALK inhibitor(s) (ALKi(s)), including crizotinib; FEAP-provided brigatinib (1 August 2016 to 21 January 2019).

Data Collection
Patient data, collected retrospectively from medical files, included demographics, NSCLC characteristics, numbers and localizations of metastatic sites, previous treatments, tumor response to brigatinib, resistance mutations before starting brigatinib or after progression and post-progression treatments. All consecutive patients meeting inclusion criteria were enrolled without selection in each participating center.

Endpoints
The primary endpoint was invPFS from brigatinib onset, i.e., from the first brigatinib dose to first documentation of objective disease progression or death from any cause. Secondary endpoints included objective response rate (ORR), median duration of treatment (DOT), median OS and analysis of subsequent treatment(s). Analysis focused particularly on lorlatinib efficacy (median DOT and OS from lorlatinib initiation) after progression on brigatinib.

Statistical Analyses
Comparisons between patient characteristics were performed using the chi-square test or Fisher's exact test for discrete variable. The Kaplan-Meier method was used to estimate PFS and OS for the entire population and defined subgroups according to the number of treatment lines. The log-rank test compared survival by treatment category. Best response to treatment was assessed according to RECIST 1.1 criteria. Statistical analyses were computed with SAS v9.4 software (SAS Institute, Cary, NC, USA).
The study was conducted in accordance with French laws and regulations in force

Discussion
With a median follow-up of 40.4 months, the results of this retrospective, multicenter, real-life study confirmed brigatinib efficacy at managing heavily pre-treated advanced ALK + NSCLCs with median invPFS and OS from brigatinib initiation at 7.4 and 20.3 months, respectively. As noted previously and in another analysis [20][21][22], brigatinib effectiveness (DOT and OS) tends to decline, depending on the line at which it is used. Most patients were able to receive the standard regimen, with dose reduction for 20% of them and a 10% discontinuation rate, thereby confirming intermediate analysis findings [19].
Several recent studies examined brigatinib efficacy in heavily pretreated patients ALK + advanced NSCLCs. According to a retrospective chart review of 104 brigatinibtreated patients in Italy, Norway, Spain and the UK, ORR was 39.8%, median PFS was 11.3 (95% CI 8.6-12.9) months and median OS lasted 23.3 (95% CI 16.0-NR) months [22]. Based on 604 patients (from 21 countries), including those previously given next-generation ALKis, median brigatinib DOTs for patients with prior crizotinib, alectinib, ceritinib or lorlatinib were 10.0, 8.7, 10.3 or 7.5 months, respectively [21]. Finally, a phase 2, singlearm study analyzed brigatinib efficacy and safety in 47 Japanese patients with ALK + advanced NSCLCs that progressed from previous alectinib or other ALKis [20]. Their ORR and DCR were 34% (16/47) and 79% (37/47), respectively, with median PFS lasting 7.3 months. These summarized results differ slightly from ours and this difference is likely attributable, in part, to the limitations of retrospective chart data but also, and perhaps  Finally, the median OS from the initial NSCLC diagnosis was 75.3 months (95% CI 38.2-174.6), knowing that 86.4% of the patients had an advanced stage at diagnosis.

Discussion
With a median follow-up of 40.4 months, the results of this retrospective, multicenter, real-life study confirmed brigatinib efficacy at managing heavily pre-treated advanced ALK + NSCLCs with median invPFS and OS from brigatinib initiation at 7.4 and 20.3 months, respectively. As noted previously and in another analysis [20][21][22], brigatinib effectiveness (DOT and OS) tends to decline, depending on the line at which it is used. Most patients were able to receive the standard regimen, with dose reduction for 20% of them and a 10% discontinuation rate, thereby confirming intermediate analysis findings [19].
Several recent studies examined brigatinib efficacy in heavily pretreated patients ALK + advanced NSCLCs. According to a retrospective chart review of 104 brigatinib-treated patients in Italy, Norway, Spain and the UK, ORR was 39.8%, median PFS was 11.3 (95% CI 8.6-12.9) months and median OS lasted 23.3 (95% CI 16.0-NR) months [22]. Based on 604 patients (from 21 countries), including those previously given next-generation ALKis, median brigatinib DOTs for patients with prior crizotinib, alectinib, ceritinib or lorlatinib were 10.0, 8.7, 10.3 or 7.5 months, respectively [21]. Finally, a phase 2, single-arm study analyzed brigatinib efficacy and safety in 47 Japanese patients with ALK + advanced NSCLCs that progressed from previous alectinib or other ALKis [20]. Their ORR and DCR were 34% (16/47) and 79% (37/47), respectively, with median PFS lasting 7.3 months. These summarized results differ slightly from ours and this difference is likely attributable, in part, to the limitations of retrospective chart data but also, and perhaps more importantly, to imbalances between the populations analyzed. Patients' health status was better in the studies by Popat et al. [22] and Lin et al. [21], with respective ECOG PS 0/1 for 84% and 85.9%, and they were less intensively pretreated (median 2 lines and 1 TKI before brigatinib) in Popat et al.'s study [22] compared to medians of 3 lines and 2 TKIs herein.
Access to treatment post-brigatinib progression was also an important factor: 58% of our patients had access to post-progression therapy. Among them, nearly two-thirds received the third-generation ALKi lorlatinib with a median DOT of 5.3 months and median OS from lorlatinib initiation of 14.1 months. These results are close to those of Popat et al., who reported that 69% (53/77) of their patients who had progressed at the time of the analysis had received systemic therapy after brigatinib, 42 with an ALKi. 34 treated with lorlatinib. The median DOT for evaluable patients was 2.57 months [22].
Lorlatinib efficacy in this context was also analyzed in several studies [23][24][25][26]. Lorlatinib was accorded marketing authorization for second-line treatment after failure of a first-line second-generation TKI, regardless of the existence of a resistance mechanism. In a phase 2 trial, lorlatinib efficacy was evaluated in different populations based on treatment history before lorlatinib. Of the five cohorts, three included a second-generation TKI in their treatment sequence. Analysis of data from those three populations (comprised of 159 patients) showed an ORR of 39.6%, median PFS lasting 6.6 months and median OS 20.7 months after starting lorlatinib. These observations are quite similar to ours and are further supported by real-life data. In a multicenter retrospective analysis of lorlatinib in 37 heavily pretreated, ALK + advanced NSCLC patients, median lorlatinib DOT was 4.4 months, with 43.2% ORR and median OS from lorlatinib onset lasting 10.2 months [24]. Another analysis of 22 patients in the same setting found 35.7% ORR and 64.3% DCR, with median PFS at 6.2 months. PFS was longer for patients who benefited from prior ALKi(s) than those who did not (6.5 vs. 3.5 months, respectively) [24]. In a real-world analysis of 76 ALK + NSCLC patients enrolled in early or expanded access programs for lorlatinib in Asia and the United States, respective ORR and median PFS for those treated with <2 previous TKIs, 2 previous TKIs and 3 previous TKIs, were 42% and NR months, 35% and 11.2 months and 18% and 6.5 months [27].
There is a growing body of literature on the search for mechanisms of resistance to ALKi but this practice is not mandatory for progression management [28][29][30]. In our retrospective series, rebiopsy was performed in approximately 25% and 20% of pre-and post-brigatinib patients. Though these rates may seem low, they represent an accurate reflection of a period of management of ALK+ aNSCLC that corresponds to the brigatinib EAP between August 2016 and January 2019. When genotyping was performed, it led to the identification of an ALK secondary mutation in 37.5% and 40%, pre-and post-brigatinib, respectively, but we could not know whether the identification of the mutations was considered for the therapeutic decision. These rates of resistance mutations are demonstrated in a series of patients pre-treated with a median of two pre-brigatinib TKIs, mainly the crizotinib-ceritinib sequence. They are quite similar to those reported by Gainor et al. [31]: after two TKIs, including a second-generation TKI, secondary mutations were detected in approximately 50% of cases. Moreover, among mutations detected when patients experienced progression, G1202R was the most common. This is also consistent with previously known data that have shown the high frequency of the G1202R mutation after second-generation TKIs [32].
Our study has several limitations. First of all, it should be noted that treatment sequences are currently different from those analyzed in the brigALK study, as most patients are treated with a second-generation TKI in the first line (alectinib or brigatinib). Other limitations are those inherent in this type of retrospective study without data monitoring. Treatment efficacy was assessed by the investigators, without any independent review committee. Investigator assessment bias cannot be excluded. In this real-life study, it was not always possible to obtain complete information from patients' medical records. That was the case, for example, for patients with a dose reduction or those who discontinued brigatinib before disease progression. Another limitation to point out is the small number of events on lorlatinib that could explain the CI 95% of DOT and OS that are a little too wide. One of the strengths of the study is the absence of stringent criteria for study inclusion, meaning the population is representative of real-life, heavily treated patients with ALK + advanced NSCLCs.

Conclusions
This analysis of FEAP data confirmed brigatinib effectiveness in a population of heavily pretreated ALK + advanced NSCLCs and provided informative real-life observations about lorlatinib efficacy post-brigatinib.

Informed Consent Statement:
Patients received written and oral information on the study and gave their consent to participate in the study and for the use of their medical data for research purposes.

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