Impact of non-canonical ALK fusion on the efficacy of targeted therapy in non-small cell lung cancer

To the Editor: With the wide application of next-generation sequencing (NGS), a number of EML4-ALK fusion partners or coexisting (non)-EML4-ALK and non-EML4-ALK have been detected in non-small cell lung cancer (NSCLC). Although targeted therapies for patients with metastatic NSCLC and actionable mutations dramatically change how these patients are treated, studies have demonstrated differences in progression-free survival (PFS) among patients based on different ALK fusion partners.^[1]

More importantly, few studies have assessed the impact of non-canonical ALK fusion on the overall survival (OS) of NSCLC patients. Additionally, most studies to date have only included patients treated with crizotinib. Wang et al^[2] reported that alectinib prolonged PFS in the EML4 fusion group compared to the non-EML4 fusion group. However, the impact of non-canonical ALK fusion on OS in NSCLC patients treated with second-generation anaplastic lymphoma kinase (ALK)-tyrosine kinase inhibitors (TKIs) is still unknown. In this study, we comprehensively investigated the impact of non-canonical ALK fusion on the efficacy of ALK-TKI treatment in NSCLC.

Our study was carried out according to the Declaration of Helsinki and approved by the West China Hospital Ethics Committee (No. 2022-1085). Individual inform consent was waived by the Ethics Committee of West China Hospital, as the privacy of the patients was not disclosed.

We retrospectively screened NSCLC patients who were analyzed by deoxyribonucleic acid (DNA)-based NGS from January 2016 to December 2020 at West China Hospital. The following inclusion criteria were used: (1) pathologically confirmed NSCLC; (2) ALK rearrangements identified by DNA-based NGS; and (3) treatment with ALK-TKI. Patients who were lost to follow up within 1 month after receiving ALK-TKI treatment were excluded. Data of clinicopathological characteristics and survival information were collected through medical records. The cutoff date was January 2023, and patients without radiographic disease progression at the latest date were censored. The details of the identification of the ALK variant and statistical analysis are shown in Supplementary Material, http://links.lww.com/CM9/B844.

Among the 2231 screened consecutive patients with NSCLC, ALK fusion was detected in 125, revealing an overall mutation rate of 5.6%. Finally, 56 patients who received ALK-TKI treatment were enrolled [Supplementary Figure 1, http://links.lww.com/CM9/B791]. Supplementary Table 1, http://links.lww.com/CM9/B791 shows the baseline characteristics of the 56 NSCLC patients with ALK rearrangement. These patients had a median age of 51.5 years, ranging from 42.1 years to 63.4 years; 29 (52%) were female, and 54 (96%) had adenocarcinoma. Fifteen (27%) patients were treated with second-generation ALK-TKIs, and the second-generation ALK-TKI used was alectinib in all the 15 cases. The clinical characteristics of the patients with canonical ALK fusion and non-canonical ALK fusion were similar (all P >0.05).

Among the 36 patients harboring canonical ALK fusions, the most frequent EML4-ALK variants were V1, in 12 (33%) patients, and V3, in 11 (31%) patients. Among the 20 patients who harbored noncanonical ALK fusions, fusion partner genes included STRN, PDK1, STK17B, and ACTN1 [Supplementary Figure 2, http://links.lww.com/CM9/B791]. Among the 20 patients with non-canonical ALK fusion, 11 rare partners/fusion variants were identified in this study. Additionally, fusions of intergenic sequences with ALK were found in 7 patients [Supplementary Table 2, http://links.lww.com/CM9/B791].

All 56 patients who received ALK-TKIs were evaluated for treatment efficacy. In the canonical ALK fusion group, the objective response rate (ORR) and disease control rate (DCR) were 58.3% and 94.4%, respectively. In the non-canonical ALK fusion group, the ORR and DCR were 70.0% and 85.0%, respectively [Supplementary Figure 3, http://links.lww.com/CM9/B791]. No significant difference was found in ORR or DCR in the canonical ALK fusion vs. the non-canonical ALK fusion group, both P >0.05.

The median follow-up period was 30.9 months. The median PFS and OS of the whole cohort were 16.4 (95% CI: 8.4, 24.4) months and not reached (NR) (95% confidence interval [CI]: 61.6, NR) months, respectively. For canonical ALK fusion patients, the median PFS was 15.6 (95% CI: 10.7, 29.5) months; for non-canonical ALK fusion patients, the median PFS was 21.1 (95% CI: 9.0, NR) months. No significant differences in PFS were observed between the two groups (hazard ratio [HR] = 0.794; 95% CI: 0.402, 1.571; P = 0.509) [Figure 1A]. Consistently, no significant differences in OS were observed between canonical ALK fusion and non-canonical ALK fusion patients (HR = 1.179; 95% CI: 0.407, 3.411; P = 0.762), with a median OS of NR and 61.6 (95% CI: 52.0, NR) months, respectively [Figure 1B].

Figure 1.

Efficacy assessment of patients with the non-canonical ALK fusion and canonical ALK fusion. Kaplan-Meier estimates of PFS (A) and OS (B). Subgroup analysis of PFS (C) and OS (D) according to baseline characteristics. CI: Confidence interval; ECOG PS: Eastern Cooperative Oncology Group Performance Status; HR: Hazard ratio; NSCLC: Non-small cell lung cancer; OS: Overall survival; PFS: Progression-free survival.

Among patients with no brain metastasis at baseline, no significant differences in PFS and OS were observed between the canonical ALK fusion and non-canonical ALK fusion groups (median PFS, 15.9 [95% CI: 10.7, NR] months vs. 15.5 [95% CI: 6.2, NR] months, P = 0.318, Supplementary Figure 4A, http://links.lww.com/CM9/B791; median OS, NR [95% CI: 45.5, NR] months vs. NR [95% CI: 52.0, NR] months, P = 0.976, Supplementary Figure 4B, http://links.lww.com/CM9/B791). Similarly, for patients presenting with brain metastasis at baseline, there were no significant differences in PFS and OS between canonical ALK fusion and non-canonical ALK fusion (median PFS, 11.5 [95% CI: 8.1, NR] months vs. 15.5 [95% CI: 6.2, NR] months, P = 0.579, Supplementary Figure 4C, http://links.lww.com/CM9/B791; median OS, NR [95% CI: 9.2, NR] months vs. 61.6 [95% CI: 24.0, NR] months, P = 0.615, Supplementary Figure 4D, http://links.lww.com/CM9/B791).

The PFS of canonical ALK fusion and non-canonical ALK fusion patients treated with different generations of ALK-TKIs was compared. The results showed no significant differences in PFS between canonical ALK fusion and non-canonical ALK fusion patients treated with first-generation ALK-TKIs (median PFS, 11.2 months vs. 14.2 months, HR = 0.907, P = 0.786 [Supplementary Figure 5A, http://links.lww.com/CM9/B791]. Consistently, OS did not differ significantly between canonical ALK fusion and non-canonical ALK fusion patients treated with first-generation ALK-TKIs (median OS, NR vs. 61.6 months, HR = 1.139, P = 0.824) [Supplementary Figure 5B, http://links.lww.com/CM9/B791]. Similarly, there were no significant differences in PFS (NR vs. NR, P = 0.371) [Supplementary Figure 5C, http://links.lww.com/CM9/B791] or OS (median OS, NR vs. NR months; P = 0.919) [Supplementary Figure 5D, http://links.lww.com/CM9/B791] between the two groups receiving second-generation ALK-TKIs. More data on analyses of the effect of non-canonical ALK fusion on PFS and OS in other subgroups are presented in Figure 1C and Figure 1D, which also showed consistent results across the subgroups.

We demonstrate that non-canonical ALK fusion had no effect on efficacy in patients with ALK rearrangement NSCLC treated with ALK-TKIs, either in patients with or without brain metastasis or in patients receiving first-generation or second-generation inhibitors.

Targeted-capture DNA-based NGS panels are usually able to target exonic and selected intronic regions of kinase genes, and may effectively detect kinase fusions. However, inherent limitations exist. For example, if the oncogenic fusion is caused by complex DNA rearrangements, genomic rearrangement may not be fully captured. As RNA-based NGS focuses on exons postsplicing, it is a more direct technology for determining clinically actionable fusions. However, the sampling requirements and quality metrics for RNA-based NGS are stricter than those for DNA-based NGS.^[3]

In our study, the therapeutic benefits for patients harboring non-canonical fusions indicated that non-canonical fusions detected by DNA-based NGS can serve as therapeutic targets for ALK inhibitors. Hence, most uncommon ALK fusions detected by DNA-based NGS may generate transcripts and proteins, and these cases may respond to ALK-TKIs as those involving canonical fusions do. As Kang et al^[4] showed most samples with pure non-canonical ALK fusions identified by DNA-based NGS have positive immunohistochemistry, and thus most non-canonical ALK fusions are likely to express the fusion products in the tumor.

Our study has several limitations. First, this was a retrospective study with a relatively small sample size, particularly regarding patients with non-canonical ALK fusions who were treated with second-generation ALK-TKIs. Prospective studies with large cohorts are needed to validate the findings. Second, non-canonical ALK fusion detected by DNA-based NGS was not validated at the RNA or protein level. Future studies are needed to confirm whether these non-canonical ALK fusions form functional products in the tumor. However, Rosenbaum et al^[5] demonstrated that RNA-based NGS and DNA-based NGS show a high degree of concordance for the detection of rearrangement (92%).

In conclusion, we provide knowledge of the impact of non-canonical ALK fusion on efficacy based on different generations of ALK-TKIs and whether brain metastasis is present at baseline. The findings will assist clinicians in the decision-making process in clinical practice when managing an NSCLC patient harboring an unknown non-canonical ALK fusion.

Funding

This work was supported by grants from the National Natural Science Foundation of China (Nos. 82072598, 81871890, and 91859203), 1-3-5 project for disciplines of excellence, West China Hospital, Sichuan University, China (No. ZYJC21052), National Key Development Plan for Precision Medicine Research of China (No. 2017YFC0910004), Science and Technology Program of Sichuan, China (No. 2020YFS0572), Major Science and Technology Innovation Project of Chengdu City, China (No. 2020-YF08-00080-GX), and Central Guide Place-Free Exploration Project, Sichuan Provincial Department of Science and Technology, China (No. 2020ZYD005).

Conflicts of interest

None.

Supplementary Material

cm9-137-1468-s001.pdf ^{(88.1KB, pdf)}

cm9-137-1468-s002.pdf ^{(1MB, pdf)}