P21 activated kinase 6: a promising tool for predicting small cell lung cancer diagnosis and treatment response

Patients and methods

Between March 2021 and March 2024, a total of 201 patients were newly diagnosed with lung cancer at the Second Affiliated Hospital of Nanchang University. This cohort included 109 patients with SCLC and 92 patients with non-small cell lung cancer (NSCLC). Among the SCLC group, paired archived pre-treatment and post-treatment serum samples were collected from 56 patients. Progression-free survival (PFS) is formally defined as the time interval from the randomization date (or treatment initiation date in non-randomized trials) to the first documented occurrence of either: Radiographic disease progression as per RECIST 1.1 criteria (Eisenhauer et al., 2009), or Death from any cause, irrespective of radiographic confirmation of progression with data gathered through case review and telephone follow-up. The NSCLC group consisted of 67 cases of lung adenocarcinoma, 23 cases of lung squamous cell carcinoma, and two cases of large cell carcinoma. Additionally, 85 patients with benign pulmonary nodules (Group PN) and 94 healthy individuals (Group NC) were included as normal controls.

The inclusion and exclusion criteria were as follows:

Inclusion criteria: (1) Newly diagnosed patients who had not undergone surgery, chemotherapy, or radiotherapy prior to the study; (2) Patients with a confirmed diagnosis based on comprehensive clinical, imaging, and histopathological data; (3) Patients without other autoimmune diseases or cancers; (4) Healthy volunteers in Group NC were screened for pulmonary disease via CT examination. (5) Histologically confirmed SCLC/NSCLC; (6) treatment-naive patients; (7) availability of pre-treatment serum.

Exclusion criteria: (1) Unknown tumor type or non-primary tumor; (2) Presence of other lung diseases; (3) Concurrent severe heart, kidney, liver, thyroid, blood, or systemic diseases; and (4) Pregnancy or breastfeeding. (5) Mixed histology; prior anticancer therapy; (6) severe comorbidities.

SCLC patients were classified according to the Veteran’s Administration (VA) staging standard. Limited-stage SCLC (LS-SCLC) was defined as stage I to III (T_any, N_any, M0), and 37 patients were included in this group. Extensive-stage SCLC (ES-SCLC) was defined as stage IV (T_any, N_any, M1a/b/c), with 72 patients included. This study was approved by the Ethics Committee of the Second Affiliated Hospital of Nanchang University and exempted from informed consent review.

Specimen collection

Fasting blood samples were collected by venipuncture into EDTA tubes, stored at 4 °C for less than 4 h, and centrifuged at 1,026 × g for 15 min. Serum was then collected and stored at −80 °C until further analysis.

Methods and instruments

PAK6 levels were measured using an enzyme-linked immunosorbent assay (ELISA) kit from Shanghai Enzyme-linked Biotechnology Co., Ltd. (Shanghai, China), following the manufacturer’s protocol. The experiment included blank control wells and standard wells, and absorbance was measured at 450 nm. Serum concentrations of NSE, CEA, CA19-9, and ProGRP were determined using a chemiluminescence method. Detection kits for these biomarkers were provided by Siemens Medical Diagnostic Products Co., Ltd. (Munich, Germany), and the measurements were performed using the ADVIA Centaur fully automated chemiluminescence immunoanalyzer. All procedures were conducted according to the reagent instructions and the quality management standard documents of the Second Affiliated Hospital of Nanchang University.

Statistical analysis

Statistical analyses were carried out using SPSS 25.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism 8.0, with sample sizes determined using PASS 15.0. X-tile software was used to calculate the cut-off values for K–M analysis. The one-sample Kolmogorov–Smirnov test was applied to assess data normality. Measurement data were presented as Median (P25–P75). For comparisons between two independent samples, the Mann–Whitney U test was used, and the Kruskal–Wallis H test was employed for multiple independent sample comparisons. Categorical data were expressed as rates and analyzed using the χ² test. Receiver operating characteristic (ROC) curves were generated, and the area under the curve (AUC) was calculated. The cut-off corresponding to the maximum value of Youden’s Index was used as the optimal clinical diagnostic threshold. DeLong’s test was used to compare the diagnostic efficacy of two variables for predicting outcomes. The diagnostic value of combined tumor marker detection was assessed using binary logistic regression analysis. A p value of less than 0.05 was considered statistically significant. Confirmed outliers were excluded from the analysis. Spearman’s correlation test was performed to examine the associations between PAK6 and other indicators in the SCLC group. Visual analyses were performed using the Xiantao platform (Xiantao, Shanghai, China), available at https://www.xiantao.love.

Study population

A χ² test analysis revealed that there were significantly more males than females across all groups (p < 0.05). However, no significant differences were observed in age and gender distributions among the groups (p > 0.05, Table 1).

Concentration of STMs in each group

At the group level, serum PAK6 levels in patients with SCLC (56.44 ng/L) were significantly higher than in those with NSCLC (41.06 ng/L), PN (37.82 ng/L), and the NC group (34.75 ng/L) (p < 0.01, Fig. 1A). In the SCLC group, the concentrations of ProGRP and NSE were also significantly elevated, with levels notably higher than in the other three groups (p < 0.01, Figs. 1B and 1C). The CEA level in the SCLC group was significantly lower than in the NSCLC group, but slightly higher than in the NC and PN groups (p < 0.01, Fig. 1D). The distribution of CA19-9 levels mirrored that of CEA, with levels in the SCLC and NSCLC groups significantly higher than those in the PN and NC groups (p < 0.01, Fig. 1E). Additional data for these groups are provided in Table SI. When comparing the LS-SCLC and ES-SCLC groups, no significant differences were observed in any of the STMs (Figs. S1A–S1E).

The correlation between PAK6 and other indicators in SCLC group

Given the similar trends in PAK6 expression across different groups, we investigated the correlation between PAK6 and other indicators, including VA stage, smoking status, gender, age, CEA, CA19-9, NSE, and ProGRP. No significant correlations were found between PAK6 and these indicators (Fig. 2).

Diagnostic value of STMS in SCLC

The ROC curves for diagnosing SCLC using PAK6, ProGRP, NSE, CEA, and CA19-9 were individually plotted (Fig. 3). As shown in Fig. 3A, the AUC values for these five STMs in diagnosing SCLC were 0.892, 0.935, 0.834, 0.676, and 0.611, respectively. The 95% CI for the AUC values were 0.857–0.927 for PAK6, 0.909–0.960 for ProGRP, 0.778–0.890 for NSE, 0.622–0.731 for CEA, and 0.547–0.675 for CA19-9. The diagnostic efficiency of ProGRP was significantly higher than that of NSE and PAK6, with no significant difference between the latter two. PAK6 demonstrated diagnostic efficacy second only to ProGRP, as detailed in Table 2. Among the five diagnostic indicators, NSE had the highest sensitivity (0.96), while ProGRP had the highest specificity (0.89). Based on the maximum value of Youden’s index, the optimal cut-off value for PAK6 in diagnosing SCLC was determined to be 47.30 ng/L, with a sensitivity of 0.82 and a specificity of 0.86. When combining the indicators NSE, ProGRP, and PAK6 in various pairs (Fig. 3B), there was no significant difference in the diagnostic efficacy between the PAK6 & NSE combination, NSE & ProGRP combination, and the PAK6 & ProGRP combination. The AUC values for these combinations were 0.95, 0.96, and 0.97, respectively, while Youden’s Index values were 0.76, 0.81, and 0.87, respectively. The diagnostic efficiency of combining all three indicators together reached 0.98, which was significantly higher than that of any two-indicator combination.

The serum level of PAK6 can predict the therapeutic effect of SCLC

In this section, we divided the patients into two groups: the pre-treatment group and the post-treatment group, based on the levels of markers after the initial diagnosis and three months of follow-up (Fig. 4). We then analyzed differences in serum STM levels among these groups. The results showed that serum levels of PAK6, ProGRP, and NSE significantly decreased following SCLC treatment (p < 0.001, Figs. 4A–4C), while CEA and CA19-9 levels showed no significant changes (Figs. 4D, 4E). Detailed data are available in Tables SII–SIII.

High expression of PAK6 is associated with poor prognosis of SCLC

We conducted a retrospective follow-up of newly diagnosed SCLC patients to explore the relationship between STMs expression levels and PFS. Using a cut-off value for PAK6 (60.10 ng/L), NSE (3.01 ng/mL), ProGRP (53.1 pg/mL), CEA (2.83 ng/mL) and CA19-9 (13.15 U/mL) determined by X-tile, we divided the SCLC group into high and low STMs expression groups. The results of PAK6 indicated that the PFS of the high-expression group was significantly lower than that of the low-expression group, with a significant difference between the two (HR = 2.02 [1.33−3.07], p < 0.01, Fig. 5A). The median PFS of the low expression group and the high expression group were 194 [153–267] and 92 [91–134] days, respectively. In contrast, no significant differences in prognosis were observed between the high and low expression groups for the other four STMs (Figs. 5B–5C). Detailed data are available in Table SIV.