Prognostic Markers of Survival among Japanese Patients with Anaplastic Lymphoma Kinase-Positive Non-Small-Cell Lung Cancer Receiving First-Line Alectinib

The prognoses of patients with non-small-cell lung cancer (NSCLC) harboring anaplastic lymphoma kinase (ALK) gene rearrangement have dramatically improved with the use of ALK tyrosine kinase inhibitors. Although immunological and nutritional markers have been investigated to predict outcomes in patients with several cancers, their usefulness in targeted therapies is scarce, and their significance has never been reported in patients receiving first-line treatment with alectinib. Meanwhile, neutrophil-to-lymphocyte ratio and platelet-to-lymphocyte ratio (PLR) has been investigated during crizotinib treatment. This multicenter retrospective study evaluated 42 consecutive Japanese patients with ALK-positive NSCLC who received first-line treatment with alectinib. Immunological and nutritional markers were evaluated at baseline and 3 weeks after alectinib introduction, and their significance in predicting progression-free survival (PFS) was explored. PFS duration was significantly associated with baseline PLR (hazard ratio (HR): 2.49, p = 0.0473), systemic immune-inflammation index (SII; HR: 2.65, p = 0.0337), prognostic nutrition index (PNI; HR: 4.15, p = 0.00185), and the 3-week values for SII (HR: 2.85, p = 0.0473) and PNI (HR: 3.04, p = 0.0125). Immunological and nutritional markers could be useful in predicting the outcomes of first-line treatment with alectinib. Since PLR and SII consist of platelet counts, platelet count could be an important constituent of these markers.

Alectinib demonstrated superior overall survival (OS) and PFS and less toxicity compared to crizotinib in two randomized phase III trials [7,8]. The lower toxic property of alectinib made it suitable for ALK-positive NSCLC patients with poor Eastern Cooperative Oncology Group performance status (ECOG-PS) [11]. However, some patients experienced early progressive disease (PD) during treatment with alectinib [12]. Therefore, it would be useful to identify predictive markers for alectinib, which could help identify patients who would benefit from alectinib continuation and patients who should be switched to second-line treatment.
However, previous reports on the usefulness of these markers during targeted therapies are scarce. The usefulness of NLR and PLR has been reported in epidermal growth factor receptor-TKI treatment [26,27]. In contrast, PLR could reportedly be a predictive marker in patients with ALK-positive NSCLC [28]. Furthermore, NLR and PLR can predict outcomes during crizotinib treatment [29]. However, the investigation of these markers during first-line alectinib treatment has not been explored. To our knowledge, this is the first study to investigate immunological and nutritional markers in first-line alectinib treatment. Considering how less toxic the profile of alectinib is, compared to crizotinib, the association between these markers and outcomes during alectinib treatment could differ from those observed during crizotinib treatment. Therefore, we aimed to identify the predictors of outcomes during first-line alectinib treatment for ALK-positive NSCLC and examined candidate markers at baseline and mid-treatment.

Patients
This multicenter retrospective study evaluated consecutively 42 Japanese patients with metastatic or recurrent ALK-positive NSCLC who received first-line alectinib at six hospitals in Japan between 1 September 2015, and 31 March 2021. The ALK rearrangement was diagnosed by immunohistochemistry (IHC) in 11 patients, both IHC and fluorescence in situ hybridization (FISH) in 27, FISH in 1, reverse transcriptase-polymerase chain reaction (RT-PCR) in 1, both IHC and RT-PCR in 1, and next-generation sequencing (Oncomine Dx Target Test Multi-CDx system, Thermo Fisher Scientific Life Technologies Corporation Japan, Tokyo, Japan) in 1. All patients received oral alectinib at a dose of 300 mg twice daily, in compliance with the authorized dosage in Japan. The study protocol complied with the Declaration of Helsinki and was approved by the Ethics Committee of the Japanese Red Cross Kyoto Daini Hospital (4 June 2021; S2021-09) and of each participating hospital. The requirement for informed consent was waived as this was a retrospective analysis of anonymized patient data. Patients were allowed to opt out of the research's use of their data, and related information is publicly available on each hospital's website.

Data Collection
Baseline characteristics (age, sex, smoking history, ECOG-PS score, height, body weight, weight loss of ≥5% 6 months before alectinib initiation, body mass index (BMI), histology, blood examination, and the efficacy of alectinib, including objective response and PFS) were collected from electronic medical records.

Study Variables
Previous reports were used to define cut-off values for the following immunological and nutritional markers: NLR, PLR, SII, ALI, and PNI. The NLR was calculated as the ratio of absolute neutrophil count (/µL) to lymphocyte count (/µL), and patients were grouped according to two cut-off values for NLR < 5 or ≥5, which is usually adopted in studies, and <3 or ≥3, which was adopted in the study dealing with ALK-positive NSCLC [28]. The PLR was calculated as the ratio of absolute platelet count (/µL) divided by absolute lymphocyte count (/µL) and grouped according to PLR values of <250 or ≥250. The cut-off value for PLR was set at <250 or ≥250 according to the previous report adopting <246.36 or ≥246.36 [18]. The SII values were calculated as absolute platelet count (/L) × NLR, and patients were grouped according to SII values of <1000 or ≥1000. The cut-off values for PLR and SII were originally set based on the receiver operating characteristic curve, based on data of the current study participants. ALI was calculated as (BMI × albumin (g/dL))/NLR, and patients were grouped according to ALI values of <18 or ≥18. The PNI was calculated as 10 × albumin (g/dL) + 0.005 × absolute lymphocyte count (/µL), and patients were grouped according to PNI values of ≥40 or <40.
These markers were evaluated at baseline and 3 weeks after alectinib initiation, for associations with PFS at these two time points. Patients who experienced PD within 3 weeks were excluded from subsequent time point analyses.
After the evaluation of candidate markers at baseline, the composite markers were evaluated, combining one nutritional marker (ALI or PNI) and one immunological marker (NLR, PLR, or SII). At this time, the combination of either of the nutritional markers and the statistically significant immunological markers, as well as the combination of 2 statistically significant immunological markers were evaluated.

Assessments
Tumor response to treatment was assessed using the Response Evaluation Criteria in Solid Tumors (version 1.1). Chest radiography and computed tomography were used to evaluate treatment response, which was determined by the treating physicians and radiologists.

Statistical Analysis
Statistical analyses were performed using EZR [30], a graphical user interface for R software (The R Foundation for Statistical Computing, Vienna, Austria). Baseline characteristics were compared using Pearson's chi-squared test. The median PFS interval, with corresponding 95% confidence intervals (CIs) and an objective response rate (ORR), was calculated. Curves for PFS were evaluated using the Kaplan-Meier method and log-rank test. Univariate Cox proportional hazards regression analyses were performed for each potential marker, although multivariate analyses were not performed because of the small sample size. Differences were considered statistically significant at p < 0.05.

ORR and PFS Outcomes
The treatment responses evaluated after 3 weeks of alectinib treatment were classified as complete response in 5 (11.9%), partial response in 31 (73.8%), stable disease in 2 (4.8%), and PD in 4 (9.5%). The median PFS was 44.2 months (95% CI: 14.6 months, not reached). The Kaplan-Meier curve for PFS in all patients is shown in Figure 1.

Relationships between the Baseline Characteristics or Candidate Predictive Biomarkers and PFS
Univariate analyses were performed to predict PFS for baseline characteristics and the markers at baseline and 3 weeks after alectinib initiation ( Table 2)

Discussion
The discovery of ALK gene rearrangement and the introduction of ALK-TKIs have substantially improved the prognosis of patients with ALK-positive NSCLC. However, patients treated with ALK-TKIs ultimately face resistance. The resistance mechanisms have been vigorously investigated. Mutations in target tyrosine kinases (second-site gatekeeper mutation), activation of alternative pathways, epithelial-to-mesenchymal transition, phenotypical changes, and drug transporter activation are the major mechanisms [31]. The different resistance mechanisms in first-and second-generation ALK-TKIs have been documented in terms of second-site gatekeeper mutations [32].
Meanwhile, biomarkers to predict outcomes of ALK-TKIs are limited. B cell lymphoma-2-like protein 11 (known as BIM) deletion polymorphism [33] and differential protein stability [34] are related to the outcomes of ALK-TKIs on a molecular basis, while PLR can predict the prognosis of patients with ALK-positive NSCLC [28]. NLR and PLR are associated with the outcomes of crizotinib treatment on a practical basis [29]. Immunological and nutritional markers are suitable for repetitive evaluation, which reflects dynamically changing host-tumor interactions. Considering the difference in therapeutic effects, and adverse events between crizotinib and alectionib, it is expected that the biomarkers could be different between these ALK-TKIs.
Contrary to our prospect of the different biomarker profiles between them, baseline PLR and SII, both comprising platelet count, were significantly associated with the outcomes of alectinib treatment in the current study. This finding was consistent with previous reports on crizotinib treatment [29], although alectinib had different effects on the efficacy and adverse events of crizotinib [7,8,11]. The observed findings suggest that platelet count could influence the prognosis of patients with ALK-positive NSCLC.
Platelets are considered favorable in tumor spread by stimulating cancer cell proliferation and facilitating metastasis [35]. Considering the relationship between ALK-positive NSCLC and platelets, EML4-ALK rearrangements were found in platelets. Their detection was associated with poor outcomes during crizotinib treatment [36]. This suggests that tumor-educated platelets play a key role and could be predictive markers during the treatment of ALK-positive NSCLC.
Platelet-derived growth factors (PDGFs) and their receptors (PDGFRs) have been found to correlate with tumor progression. Phosphorylated stem cell growth factor receptor kit and PDGF-α have been reported as predictive and prognostic factors in ALK-positive NSCLC [37]. Platelets contain several growth factors, including PDGFs and transforming growth factor-β (TGF-β), which favor tumor progression [38]. Platelet counts are regulated by thrombopoietin, which has close interaction with PDGFs and TGF-β [39]. The fact that ALK-positive NSCLC has a close relationship with platelets and PDGFR-α could explain the observed outcome in biomarkers composed of platelet count and the outcomes of alectinib treatment.
Conversely, ALK-positive NSCLC is associated with a high rate of venous thromboembolism [40]. Platelets have been found to have a closer relationship with venous thromboembolism than previously anticipated [41]. Thus, the role of platelets in ALKpositive NSCLC could be of clinical importance, which reinforces the aforementioned mechanism of platelet and platelet-containing PDGFs and TGF-β.
Regarding the importance of the host's nutritional status, the current study demonstrated the efficacy of baseline ECOG-PS and PNI, as well as 3-week values of PNI in predicting PFS. This result suggests the importance of the baseline nutritional status during alectinib treatment.
The timing of the evaluation is another scientific interest. Our results demonstrated the importance of baseline PLR, SII, and PNI values. Contrary to our expectations, the immunological and nutritional markers did not change in favor of better outcomes after 3-week treatment with alectinib, which is considered a sufficient treatment period for the efficacy of alectinib in tumors. PLR was the only marker that lost significance after 3 weeks of alectinib treatment. These observations suggest that the immunological and nutritional status of ALK-positive NSCLC does not change dramatically after alectinib treatment, irrespective of treatment efficacy. Thus, the baseline values of PLR, SII, and PNI could be of the utmost importance in predicting outcomes.
The current study has several limitations. Firstly, a retrospective design is prone to bias. Secondly, the small sample size due to the rarity of ALK-positive NSCLC is another potential source of bias. The small sample size hampered multivariate analyses. Multivariate analyses would have allowed us to investigate the significance of SII and PNI, since these two markers are closely related to ECOG-PS. Thirdly, we selected the candidate immunological and nutritional markers based on the results of the original articles. For example, we did not adopt lung immune prognostic index (LIPI), because LIPI was associated with the outcomes of immune checkpoint inhibitors but not with chemotherapy [42]. Therefore, there could be some potential markers, which we did not analyze. Finally, we selected cut-off values for the various predictive markers from previous reports, save for PLR and SII. Therefore, the utility of markers and optimal cut-off values should be clarified in large prospective studies.

Conclusions
The present study revealed that the PFS of Japanese patients with ALK-positive NSCLC who received first-line treatment with alectinib could be related to immunological and nutritional status. PLR and SII at baseline and SII after 3 weeks were significantly associated with PFS. Since these markers consist of platelet counts, the current study suggests the importance of platelet count as a constituent of immunological markers in predicting the outcome of ALK-positive NSCLC. Baseline ECOG-PS and PNI are also important in predicting PFS, which reflect the nutritional status of the host. Informed Consent Statement: Patient consent was waived as this was a retrospective analysis of anonymized patient data. Patients were allowed to opt out of the research's use of their data, and related information is publicly available on each hospital's website.
Data Availability Statement: Not applicable.