Introduction

Anaplastic lymphoma kinase (ALK) rearranged non-small cell lung carcinoma (NSCLC) has emerged as a distinct entity with a growing number of potent ALK tyrosine kinase inhibitors (TKIs). Crizotinib, a first-generation TKI, emerged as the forerunner among these, and has shown remarkable responses as evidenced in the PROFILE study.1 However, resistance mechanisms ensue and newer generation alectinib and ceritinib were developed to overcome these.23 Despite showing durable responses and promising survival rates, resistance to these have also been reported. The newest third-generation lorlatinib has now gained approval for use in the first-line setting (Food and Drug Administration data sheet 2021),4 and has been described to overcome resistance to the second-generation TKIs.

Resistance to TKIs can broadly be categorized as on-target alterations which include ALK kinase domain alterations (the most common being G1202R, L1196M, G1269A) and ALK amplification; and the second category includes off-target mechanisms involving upregulation of other bypass pathways (EGFR, SRC, MEK/ERK, KIT, and others).5678910 Description of resistance mechanisms to crizotinib have been extensively reported in literature with almost one-third cases showing on-target resistance mechanisms.11 However, the same for second-generation drugs is limited to in vitro studies, few case series and reports, and anecdotal cohorts. Resistance to lorlatinib has also emerged and still needs real-world elucidation.1213

In this study, we present the largest series of repeat biopsies (tissue and/or liquid) from patients of ALK-positive NSCLC who have progressed on ALK-directed therapy from this part of the world. Using a combinatorial approach of genomics and histology, we describe the spectrum of various resistance mechanisms encountered as patients relapse on ALK TKIs.

Methods

Results

Between 2015 and 2021, a total of 32 patients with ALK-positive NSCLC, who progressed on any ALK TKI, were enrolled in this study based on the availability of a rebiopsy specimen at disease progression. The median age was 53 years (range: 36–75 years) with a male predilection (male:female 1.3:1). ALK immunohistochemistry was positive in all the 32 cases at diagnosis, whereas on NGS an ALK fusion was detected in 30 out of the 32 cases. The two cases which did not reveal a fusion on NGS, were confirmed on fluorescence in situ hybridization and showed break-apart signals. Table 1 depicts the basic demographic details, preenrolment, and postenrolment treatment details.

Of these 32 cases, 27 (84.4%) cases were found to harbor an additional resistance mechanism. Among these 27 cases, 12 (44.4%) received crizotinib as the first line, 4 (%) received alectinib, and 11 received ceritinib. Eighteen of these 27 cases harbored an on-target ALK alteration, with L1196M gatekeeper mutation being the most common, seen in 11 cases, and G1202R alteration seen in 3 cases. One case harbored the I1171T mutation after first-line alectinib, and three cases showed ALK amplification. In 9 cases a potential off-target alteration was detected, the most frequent being TP53 mutation in 8 cases, followed by KRAS mutation in 4 cases and MET amplification in 3 cases. Fig. 1 depicts the spectrum of resistance alterations detected along with their frequencies and related drugs.

Close

Fig. 1: Spectrum of on-target and off-target resistance mechanisms encountered in the study cohort posttreatment with anaplastic lymphoma kinase (ALK) tyrosine kinase inhibitor therapy.

Export to PPT

A total of four patients underwent serial rebiopsies after subsequent lines of treatment.Fig. 2 Three out of these four patients were treated with crizotinib in the first-line setting, of which two developed L1196M solvent front mutation, one patient developed a G1202del, and one patient treated with alectinib developed KRAS mutation and MET dysregulation. The two patients who developed L1196M were offered ceritinib, and post-ceritinib progression NGS revealed additional D1203N mutation in one patient along with increased allele frequency of the existing L1196M mutation, and the other developed an ALK amplification with a copy number of 8.4. The patient with G1202del was offered alectinib, and post-alectinib progression revealed stable allele frequency of the G1202del mutation, with an additional TP53 alteration (allele frequency: 2.3%, possibly indicative of clonal hematopoiesis of indeterminate potential) on liquid biopsy-based NGS profiling. The patient with KRAS (p.G12V), MET exon 14 skipping mutation, showed an increase in variant allele frequency of the KRAS alteration (from 3.2 to 4.6%).

Close

Fig. 2: Alluvial diagram depicting the sequence of alterations and sequential tyrosine kinase inhibitor (TKI) therapy in the four patients who underwent serial biopsies after every line of treatment.

Export to PPT

Discussion

This is a real-world experience of ALK TKI resistance from this part of the world. A total of 32 patients who progressed on any ALK TKI and who underwent NGS-based testing for detection of potential resistance mechanisms were enrolled. Using targeted NGS, resistance mutations were detected in 27 (84.4%) cases, with 18 (56.3%) being on-target ALK mutations and 9 (28.1%) being secondary off-target mechanisms.

Our positive rate of resistance mutations in rebiopsied samples (84.4%) is substantially higher when compared to contemporary real-world studies, or controlled clinical trials. Lin et al reported a positive rate of 39%, as well as Gainor et al reported a positive rate of 38%.5614 This may be attributed to the enriched nature of the population in our study, as only those who underwent NGS-based testing were included. This proposition has also been exemplified by the high 52% positive rate from Guardant database where mutations were encountered in 16/31 patients enrolled.15 Additionally, ethnic and geographic differences, analogous to EGFR-mutated NSCLC, may also contribute to the same.1617

As reported by other real-world studies and trials, L1196M, G1202R, and D1203N are known mechanisms of resistance to crizotinib.1819 Also, crizotinib was the most common drug, after which resistance mechanisms developed. While G1202 mutations have been reported to be common after second-generation ALK TKIs, in our study, this was not encountered as most cases were post-crizotinib. Secondary ALK mutations like L1196M and I1171T2021 have been shown to be sensitive to ceritinib and not alectinib. One patient subsequent to treatment with first-line alectinib, developed I1171T mutation, which has been reported to be sensitive to ceritinib. The response to ceritinib in this patient was partial response and the patient is currently on second-line ceritinib therapy with an ongoing response.

The role of rebiopsy versus a liquid biopsy-based assessment of circulating tumor DNA and newly emerged mutations have been widely discussed in literature.22232425 Whereas tissue is the gold-standard material to ascertain any emergent and evolved clones/mutations, liquid biopsy has emerged as an effective surrogate for the same, especially in the metastatic setting where the disease burden is high. Issues related to tumor heterogeneity with dynamic clones present in different areas of a tumor are also taken care of when using liquid biopsy-based testing.2627 In a study by Shaw et al,28 on sequencing of ALK TKIs, they reported that in cases who progress on alectinib, the underlying cause in almost 50% cases is a secondary ALK mutation, and hence rebiopsy in such cases revealed both pan-TKI resistant mutations as well as secondary mutations which were proven sensitive to ceritinib as well as crizotinib, both in vitro and in vivo.29

This is a single-center real-world experience, the largest from the Indian peninsula, depicting resistance profiles to ALK TKIs. Owing to ethnic and geographic heterogeneity in terms of prevalence of biomarker-driven process in both EGFR and ALK altered NSCLC in the Asian population, this study becomes relevant depicting the clinical behavior and responses, including the need for rebiopsy (tissue/liquid) to ascertain the underlying molecular causes for resistance or disease progression. Future artificial intelligence-based algorithms may be developed to predict these mechanisms of resistance for instituting optimal therapy.30