Alectinib versus crizotinib as the first-line treatment in patients with advanced ALK-positive non-small cell lung cancer: a Chinese real-world cohort study.

Introduction
It has been reported that approximately 5% of patients with non-small cell lung cancer (NSCLC) have anaplastic lymphoma kinase fusion (ALK-positive) tumors (1-3). Crizotinib was the first tyrosine kinase inhibitor (TKI) developed for ALK. Multiple phase III clinical trials have shown that crizotinib significantly increases the objective response rate (ORR) and prolongs the progression-free survival (PFS) of patients with ALK-positive NSCLC compared to standard chemotherapy regimens (4,5). Subsequently, second-generation ALK-TKIs, such as alectinib, ceritinib, ensartinib, and brigatinib, along with the third-generation ALK-TKI lorlatinib, have been approved as first-line treatments due to their strong efficacy against crizotinib-resistant mutations, and improved ability to penetrate the blood-brain barrier (BBB) (6-14).
Recently, Wu et al. (15) reported that adjuvant alectinib significantly prolonged disease-free survival in patients with resected ALK-positive NSCLC compared to platinum-based chemotherapy. Moreover, Janson et al. reported a patient with stage IV ALK-rearranged NSCLC who attained a complete pathological response after administration of neoadjuvant alectinib (16). The ALEX study showed that alectinib exhibits significant superiority over crizotinib in terms of PFS (investigator-assessed PFS 34.8 vs. 10.9 months, P<0.0001) and overall survival (OS) [not reached (NR) vs. 57.4 months, P=0.04] (8). However, the J-ALEX study found that OS was comparable between alectinib and crizotinib treatment groups (alectinib vs. crizotinib, NR vs. NR, P=0.91; 5-year OS rates, 60.9% vs. 64.1%) (9). A systematic review and network meta-analysis of 14 randomized controlled clinical trials demonstrated that alectinib was the preferred choice for both first- and second-line treatments in terms of both safety and efficacy (17).
The patients enrolled in the clinical trials were strictly selected; however, they may not fully represent real-world populations. Due to the low incidence of ALK fusion and the extended duration of treatment follow-up, current real-world studies comparing different ALK-TKIs in the first-line treatment of ALK-positive NSCLC are limited. Research continues on the use of various strategies for different ALK-TKIs (18,19). Furthermore, given the divergent OS results from the ALEX trial (Western population; P=0.04) and the J-ALEX trial (Japanese population; P=0.91), a critical question remains whether the efficacy profile in a Chinese population will resemble that observed in the West or in Japan. We conducted a retrospective study using data from a real-world, multicenter observational cohort to investigate the efficacy and safety of first-line alectinib and crizotinib treatment for locally advanced or metastatic ALK-positive NSCLC in a real-world setting. Further, the current status and efficacy of the second-line and subsequent treatment regimens were evaluated. We present this article in accordance with the STROBE reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-1133/rc).

Methods

Patients
For this cohort study, patient data were retrospectively collected from the Peking Union Medical College Hospital and Beijing Chest Hospital between September 2013 and May 2023 using the CAPTRA-Lung (NCT03334864) database (20). To ensure participant representativeness, the inclusion and exclusion criteria were rigorously defined. Patients were included in the study if they met the following inclusion criteria: (I) had pathologically confirmed NSCLC; (II) had local advanced (inoperable or inradiable stage IIIB or IIIC) or metastatic (stage IV) NSCLC based on the 8th tumor-node-metastasis staging system (21); (III) had been confirmed to be ALK-positive via Ventana ALK (D5F3) immunohistochemical assay, polymerase chain reaction (PCR), fluorescence in situ hybridization (FISH), or next-generation sequencing (NGS); (IV) had received crizotinib (250 mg twice daily) or alectinib (600 mg twice daily) as first-line treatment; (V) had undergone at least one assessment according to the Response Evaluation Criteria in Solid Tumors (RECIST, version 1.1) (22); and (VI) had clinical and survival information available. Patients were excluded from the study if they met any of the following exclusion criteria: (I) had missing important clinical information or incomplete follow-up data; and/or (II) had co-mutated driver genes. Sample size was determined based on case availability at the two hospitals during the study period.
Observation commenced at the initiation of treatment with alectinib or crizotinib, with follow-up every two months until August 8, 2024, or until death or loss to follow-up.
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the ethics committee of Peking Union Medical College Hospital (No. I-23PJ1857). Beijing Chest Hospital was informed and agreed with this study. Informed consent was waived because of retrospective nature of the study.

Data collection
Data on the baseline characteristics of the patients, such as sex, age, history of smoking, Eastern Cooperative Oncology Group performance status (ECOG-PS), pathology, stage, metastatic site, ALK status, and treatment process, were reviewed and collected from electronic records. The best response of the tumor was determined using RECIST 1.1. Survival information, including PFS, OS, and adverse events (AEs), was obtained from outpatient, inpatient, and telephone follow-ups. During the data collection process, we regularly reviewed and verified the data to ensure their completeness and accuracy.

Evaluation of efficacy and safety
Progression-free survival 1 (PFS1) was defined as the interval between the first administration of crizotinib or alectinib and the first imaging-confirmed progressive disease (PD). Progression-free survival 2 (PFS2) was defined as the interval between the first administration of second-line treatment and the first imaging-confirmed PD. OS was defined as the period from the first day of first-line treatment to death or the last day of follow-up.

Statistical analysis
Patients were stratified into two groups based on age: <65 (younger), and ≥65 years (older). In order to control confounding factors, we conducted one-to-one propensity score matching (PSM) via the nearest neighbor matching. Propensity scores were generated based on clinically relevant variables, such as gender, age, ECOG-PS, smoking history, pathology, stage, ALK testing methodology, and baseline metastasis. The chi-squared test or Fisher’s exact test was used to examine the categorical variables. The Kaplan-Meier method was used to analyze the median PFS and OS, and the log-rank test was used to determine the P value. Factors with a P value <0.05 in the univariate Cox regression were included in the multivariate Cox regression to identify independent prognostic factors. Statistical significance was defined as a P value of less than 0.05. All analyses were performed using SPSS 26.0 (IBM Corp., Armonk, NY, USA) or R language (version 4.3.1, R Foundation for Statistical Computing, Vienna, Austria).

Results

Baseline characteristics of patients
A total of 352 patients with ALK-positive locally advanced or metastatic NSCLC who received first-line ALK-TKI treatment were identified. Of these, 91 patients were excluded from the study: 44 due to the receipt of other ALK-TKIs, 17 due to a lack of important clinical information, 14 due to an absence of efficacy evaluation, and 16 due to the presence of co-mutations. The remaining cohort comprised 261 patients, of whom 128 received crizotinib and 133 received alectinib (Figure S1).
The baseline characteristics were comparable between the two treatment groups (Table 1). However, the proportion of patients with baseline brain and bone metastases was significantly higher in the alectinib treatment group than the crizotinib treatment group. The vast majority (251/261, 96.2%) of patients were identified as ALK-positive through routine Ventana ALK (D5F3) immunohistochemistry testing. Of these, 173, 1, and 18 patients were further confirmed to be ALK-positive through NGS, FISH, and PCR testing, respectively. In addition, eight and two patients were confirmed to be ALK-positive using NGS and FISH testing alone, respectively. A significantly greater percentage of ALK rearrangements were confirmed through NGS in the alectinib treatment group than the crizotinib treatment group (Table 1).
After propensity score matching, 88 patients were included in each of the alectinib and crizotinib groups. The baseline characteristics of the two matched cohorts were comparable, including the ALK testing method and the presence of baseline brain and bone metastases (Table 1).

Efficacy
As of August 8, 2024, the median follow-up times for PFS were 59.6 months for the crizotinib treatment group and 34.6 months for the alectinib treatment group, with PD occurring in 102 and 47 patients, respectively. The alectinib treatment group had significantly longer median PFS than the crizotinib treatment group [45.5 months, 95% confidence interval (CI): 41.1–NR vs. 16.6 months, 95% CI: 11.6–21.0 months, hazard ratio (HR) =0.36, P<0.001] (Figure 1A).
Both alectinib and crizotinib demonstrated excellent performance in terms of both the ORR and disease control rate (DCR), which were comparable (ORR: 80.5% vs. 74.2%, P=0.23; DCR: 100.0% vs. 98.4%, P=0.24). Seven patients in the alectinib treatment group and one patient in the crizotinib treatment group achieved complete response (Table 2).
Among the patients with or without baseline central nervous system (CNS) metastasis, the incidence of CNS progression without prior non-CNS progression was significantly higher in the crizotinib treatment group than the alectinib treatment group, [71.4% (10/14) vs. 10.3% (3/29), P<0.001; 32.5% (37/114) vs. 1.9% (2/104), P<0.001].
The median follow-up times for OS were 61.7 and 34.4 months for the crizotinib and alectinib treatment groups, respectively. A total of 51 patients in the crizotinib treatment group died, compared with 21 patients in the alectinib treatment group. OS was comparable between the two groups (alectinib vs. crizotinib, NR vs. 71.2 months, P=0.23) (Figure 1B).
A subsequent survival analysis of the post-PSM cohort demonstrated that alectinib maintained a significant advantage in PFS compared to crizotinib (45.5 vs. 15.8 months; HR =0.37, 95% CI: 0.24–0.55; P<0.001). No statistically significant difference was observed in OS between the two groups (NR vs. 74.8 months; HR =0.66, 95% CI: 0.35–1.25; P=0.20) (Figure 1C,1D).
The subgroup analysis revealed that in most cases, alectinib was more effective than crizotinib in terms of PFS. However, in the patients with locally advanced disease and baseline adrenal or liver metastases, no statistically significant difference was observed between the two groups in terms of PFS (Figure 2).

Second-line and above therapy
The results of the patients who received second and later-line treatments are presented in Table S1 and Figure 3. In the crizotinib treatment group, 88 patients received second-line treatment. Most patients (86.4%) received ALK-TKIs as second-line therapy, with alectinib and ceritinib being the most commonly used agents, received by approximately 39.8% (35/88) and 18.2% (16/88) of patients, respectively. Eight patients participated in clinical trials on ALK-TKIs, and four patients received chemotherapy with or without anti-angiogenic agents or immunotherapy. The patients treated with ALK-TKIs in the crizotinib treatment group had significantly longer PFS2 than those treated with chemotherapy combined with anti-angiogenic agents or immunotherapy (15.2 vs. 5.9 months, P=0.01) (Figure 4A). Nevertheless, there was no significant difference between the different second-line ALK-TKIs in terms of PFS2 (P=0.15) (Figure 4B).
In the alectinib treatment group, 33 patients received second-line treatment. Most patients (29/33, 87.9%) were treated with ALK-TKIs, with lorlatinib being the most common (12/33, 36.4%). No difference was observed between the different second-line treatments in the alectinib treatment group in terms of PFS2 (P=0.56) (Figure 4C).
Notably, the PFS of the patients who received first-line alectinib therapy was comparable to that of the patients who received both first-line crizotinib therapy and second-line therapy (45.5 vs. 43.9 months, P=0.32) (Figure 4D).

Safety
The crizotinib treatment group had a significantly higher incidence of AEs (64/128, 50%) than the alectinib treatment group (46/133, 34.6%) (P=0.01). In the crizotinib treatment group, the most common AEs were elevated levels of alanine transaminase (ALT) or aspartate aminotransferase (AST) and gastrointestinal reactions, with occurrence rates of 26.6% (34/128) and 17.2% (22/128), respectively. Conversely, in the alectinib treatment group, the most common AEs were elevated ALT or AST levels and increased bilirubin levels, with occurrence rates of 13.5% and 6.8%, respectively. The alectinib treatment group exhibited significantly lower incidence rates of elevated ALT or AST levels and gastrointestinal reactions, while exhibiting a significantly higher incidence of elevated bilirubin levels. The occurrence rates of grade 3 or higher AEs in the crizotinib and alectinib treatment groups were 5.5% (7/128) and 4.5% (6/133), respectively, with no significant differences. QTc (QT interval corrected for rate) interval prolongation was more common in the crizotinib treatment group, while elevated bilirubin levels were more common in the alectinib treatment group (Table 3).

Prognostic factors for PFS and OS
In the overall population, the univariate Cox regression analysis revealed that receiving crizotinib as first-line therapy and having baseline bone, liver as well as adrenal metastases were significantly associated with shortened PFS. Moreover, the tumor stage IV, being male, and the presence of baseline bone, liver and adrenal metastases were significantly correlated with worse OS. The multivariate Cox regression analysis revealed that receiving crizotinib instead of alectinib as the first-line treatment and the presence of bone and adrenal metastases at the baseline were independent risk factors for PFS1. While the presence of bone, liver and adrenal metastasis were independent risk factors for OS (Table 4).

Discussion
This real-world study showed that compared to crizotinib, alectinib significantly prolonged the median PFS of the patients (HR =0.36, 95% CI: 0.26–0.51; log-rank P<0.001). This is consistent with the findings of the ALEX (investigator-assessed HR =0.43) (23), J-ALEX [independent review committee (IRC)-assessed HR =0.37] (24), and ALESIA studies (investigator-assessed HR =0.33) (25). Given that the lower rate of NGS-confirmed ALK fusions in the crizotinib group suggested a higher risk of false positives—a potential source of bias affecting efficacy assessment—we performed propensity score matching. Nevertheless, a significant PFS benefit for alectinib was maintained in the matched cohort (alectinib vs. crizotinib, 45.5 vs. 15.8 months; HR =0.37, 95% CI: 0.24-0.55; P<0.001), underscoring the robustness of this result. However, the median PFS of the crizotinib and alectinib treatment groups in this study was longer than that reported in the clinical trials. The median PFS of the crizotinib treatment group in this study was 16.6 months, which was longer than that reported in the ALEX study (investigator-assessed PFS =10.9 months) (23), the J-ALEX study (IRC-assessed PFS =10.2 months) (24), and the ALESIA study (investigator-assessed PFS =11.1 months) (26,27). This difference may be partially attributed to the higher proportion of patients without CNS metastasis at the baseline in this real-world study compared to the proportions in the ALEX (28), J-ALEX (29), and ALESIA (26) studies (89.1% vs. 62%, 72%, and 63%, respectively). In real-world practice, clinicians preferentially prescribe alectinib as first-line therapy for patients with brain metastases due to its superior blood-brain barrier penetration. This real-world treatment preference accounts for the lower baseline incidence of brain metastases in the crizotinib group compared to that typically reported in clinical trials. A recent report from the INSPIRE study indicated that the IRC-assessed median PFS for crizotinib was 14.6 months. At the baseline assessment, approximately 72.3% of patients exhibited no evidence of CNS metastasis (30). Additionally, in a real-world study conducted in Spain, 87.9% of the patients did not have brain metastases at the baseline, and the median PFS was 14.2 months (31). In real-world studies conducted in China, one study reported a median PFS of 15.5 months for patients who received first-line crizotinib therapy, among whom, 77.8% did not have brain metastasis (32), while another study reported a median PFS of 16.1 months (33).
Further, the median PFS of alectinib in this study also exceeded that reported in the ALEX (23) and J-ALEX studies (24) (45.5 months compared to the investigator-assessed PFS of 34.8 months and IRC-assessed PFS of 34.1 months). However, in a recent 5-year update of the ALESIA study, the investigator-assessed PFS for alectinib was 41.6 months (24). Additionally, a real-world study in Japan reported a median PFS of 40.11 months in the first-line alectinib treatment group (34). These data indicate that alectinib exhibits excellent efficacy in Asian populations with ALK-positive NSCLC.
Nevertheless, the median PFS1 of the patients who received the first-line alectinib therapy was comparable to that of the patients who received both the first-line crizotinib therapy and the second-line treatment (45.5 vs. 43.9 months, P=0.32). Similarly, another study reported comparable PFS (50.6 vs. 42.8 months, P=0.59) and OS (NR vs. NR, P=0.59) between the alectinib treatment group and the crizotinib-followed-by-alectinib treatment group (35). Further, no significant difference in OS was observed, which is consistent with the results of the J-ALEX study (9), and real-world studies in Japan (34) and Canada (19). In this study, the median OS was 5.9 years in the crizotinib treatment group, but it was NR in the alectinib treatment group. The similarity in OS may be attributed to the fact that over 80% of the patients in the crizotinib treatment group who experienced PD subsequently received next-generation ALK-TKIs as salvage therapy. These results provide further evidence of the superiority of alectinib over crizotinib as a first-line therapy, as it reduces the need for frequent changes in treatment. It should be noted that only 27.6% of the patients died during the study period. This indicates that the OS data are still preliminary and further follow-up is necessary.
Further analysis was conducted on the second-line and above therapies. In both the crizotinib and alectinib treatment groups, the next-generation ALK-TKIs were the preferred therapy. In the crizotinib treatment group, the use of next-generation ALK-TKIs as a second-line treatment in patients with PD resulted in a significantly longer PFS2 compared to that of the patients who received chemotherapy (15.2 vs. 5.9 months, P=0.01). Alectinib, ceritinib, and ensartinib were the most frequently selected second-line therapies with median PFS2 times of 15.2, 7.1, and 13.1 months, respectively. Moreover, a previous study demonstrated that brigatinib exhibits high efficacy in crizotinib-refractory patients, with a median PFS of 16.7 months (36). In the present study, the seven patients who received brigatinib had a median PFS2 of 17.4 months. Moreover, four patients received lorlatinib and only one experience PD at the end of the follow-up period. Compared with the other agents, ceritinib showed inferior efficacy, but the difference was not statistically significant (P=0.08).
A retrospective study reported a median PFS of 4.3 months in patients who had developed resistance to at least one second-generation ALK-TKI and were treated with chemotherapy as salvage therapy (37). Another study found that chemotherapy and lorlatinib had a similar median PFS when used in patients who were alectinib-refractory ALK-positive (6.9 vs. 6.2 months, P=0.83) (38). In this study, among patients in the alectinib group, the PFS2 was comparable between those who received second-line ALK-TKIs and those who received chemotherapy (P=0.56). The four patients who received chemotherapy with or without anti-angiogenic or immunotherapy as second-line treatment had a median PFS2 of 4.1 months. Only one patient who was refractory to alectinib received ceritinib as second-line therapy in this study, with a PFS2 of 2 months. This supports the findings of the ASCEND-9 study, which reported a median PFS of 3.7 months (95% CI: 1.9–5.3) (39). Three patients were treated with brigatinib and only one experienced PD, with a median PFS2 of NR. Conversely, PFS of 3.8 months and 4.4 months was reported in the ALTA2 study (40) and a multicenter real-world study, respectively (41). Twelve patients received lorlatinib as second-line therapy, achieving a median PFS of 5.0 months—a result consistent with the median PFS of 5.5 months reported in a phase II clinical trial (NCT01970865) (42). However, the limited sample size may compromise the reliability of these findings. Meanwhile, plasma cell-free DNA analysis has emerged as a valuable tool for elucidating resistance mechanisms to alectinib and guiding precision medicine-based treatment strategies (43).
In addition, we observed that the alectinib treatment group had a significant lower incidence of AEs of any grade than the crizotinib treatment group. However, the incidence of grade 3 or 4 AEs in the two groups was comparable. These results further support the use of alectinib as a first-line therapy. Further, multivariate Cox regression analysis demonstrated that baseline bone and adrenal metastases were independent risk factors for both PFS1 and OS in the overall population. Therefore, future research should be directed toward this high-risk population, investigating the potential of local therapies (e.g., radiotherapy) to improve survival outcomes. Caution is warranted, however, as the combined use of alectinib and thoracic radiotherapy has been reported to significantly increase the risk of treatment-related pneumonitis (44). This study had a number of limitations. First, it was a retrospective study, which may have led to missing data. Second, selection bias may have been introduced by the study’s restriction to two medical centers, a design that limits the representativeness of our cohort for the broader Chinese population. Third, the proportion of patients with baseline brain metastases in the crizotinib group was lower than that observed in clinical trials, likely reflecting real-world prescribing preferences for alectinib in this patient subgroup, thereby introducing a selection bias toward a lower-risk population. These limitations constrain the generalizability of our findings. Additionally, the follow-up time for the alectinib treatment group was relatively short, and the OS data were preliminary. The enrolled patients will be followed-up further in the future. Finally, a summary analysis of the repeated NGS results was not conducted of the patients who developed resistance to first-line treatment, which affected any further analysis of the efficacy of the second-line treatment regimens.

Conclusions
In conclusion, in a real-world setting, first-line treatment with alectinib significantly prolonged the PFS of patients with locally advanced or metastatic ALK-positive NSCLC compared to first-line treatment with crizotinib, and it had a better safety profile. However, OS did not differ significantly between the two treatment groups. Thus, alectinib should be the preferred drug over crizotinib for patients with ALK mutations. Further studies are necessary to validate these findings across diverse regions and ethnicities.

Supplementary
The article’s supplementary files as