Worse disease-free, tumor-specific, and overall survival in surgically-resected lung adenocarcinoma patients with ALK rearrangement

Introduction This study determined the prevalence of anaplastic lymphoma kinase (ALK) rearrangement, and identified the associations of ALK rearrangement with clinicopathologic characteristics and treatment outcomes in patients with surgically-resected stage I-III lung adenocarcinoma. Methods A total of 534 surgically-resected lung adenocarcinoma patients were studied. The prevalence of ALK protein over-expression was determined by a fully-automated immunochemistry assay (with mouse monoclonal Ventana D5F3 antibody), and the associations of ALK rearrangement with clinicopathologic characteristics and treatment outcomes were analyzed. Results Forty-two (7.9%) of the 534 lung adenocarcinoma patients were ALK IHC-positive. ALK rearrangement was significantly associated with younger age (P = 0.011), high T-stage (P = 0.025), high pathologic stage (P = 0.002), solid predominant adenocarcinoma with mucin production (P = 0.006), invasive mucinous adenocarcinoma (P = 0.009), and receipt of adjuvant therapy after surgery (P = 0.036), but no significant associations were found between the ALK rearrangement and sex or smoking status. ALK IHC-positivity was significantly associated with a shorter disease-free survival, tumor-specific survival, and overall survival (P = 0.001, 0.026, and 0.007, respectively). Multivariate analysis showed that ALK IHC-positivity was an adverse prognostic factor for disease-free survival (HR, 1.80; 95% CI 1.18-2.77; P = 0.007), tumor-specific survival (HR, 2.59; 95% CI 1.35-4.97; P = 0.004), and overall survival (HR, 1.92; 95% CI 1.07-3.44; P = 0.030). Conclusion The clinical characteristics of patients with ALK-positive lung adenocarcinoma were similar to those of EGFR-mutated patients. ALK rearrangement was an adverse prognostic factor in surgically-resected lung adenocarcinoma patients.


INTRODUCTION
Lung cancer is the leading cause of cancer-related deaths in China [1]. Globally, non-small-cell lung cancer (NSCLC) accounts for 85% of lung cancer cases [2]. Although surgical resection may be curative in some patients, it is achievable only for early-stage NSCLC, and the majority of cases are diagnosed at an advanced stage. For advanced cases, cytotoxic chemotherapy is still the firstline treatment. In the last 5 years, evidence has accumulated that there are genetically-defined molecular subsets of lung adenocarcinoma, the most remarkable subset being patients with epidermal growth factor receptor (EGFR) gene mutations, which defines a small subset of patients with NSCLC who have high sensitivity to EGFR-tyrosine kinase inhibitors (TKIs) such as gefitinib and erlotinib [3,4].
Patients with echinoderm microtubule-associated protein-like 4 gene and anaplastic lymphoma kinase gene (EML4-ALK) fusion is another important subset of NSCLC. This is caused by a small inversion of chromosome 2p (the N-terminal half of EML4 encompassing the basic region) and the hydrophobic echinoderm microtubule-associated protein-like protein (HELP) domain, and a portion of the WD-repeat region becomes fused to the intracellular juxtamembrane region of ALK [5]. The EML4-ALK fusion gene possesses powerful oncogenic activity, both in vivo and in vitro [5,6]. It has been reported that there are 10 or more subtypes of the EML4-ALK fusion gene, with the E13:A20 and E6a/b:A20 types being the most common ones (incidence rates, 33% and 29%, respectively) [7,8].
Methods of detecting ALK gene fusion include immunohistochemistry (IHC), reverse transcriptasepolymerase chain reaction (RT-PCR) technology, and fluorescence in situ hybridization (FISH). In 2013, the National Comprehensive Cancer Network (NCCN) stated that FISH was the 'gold standard' method to detect ALK fusion genes. However, FISH is expensive and it is difficult to determine the overall tumor morphology and heterogeneity with its use [15], while RT-PCR requires high-quality primers and more RNA [16]. In contrast, IHC is economical, practical, and efficient, and this method is now widely used in routine pathology laboratory testing. However, there is some subjectivity in evaluating staining results in IHC, and the accuracy of the method depends largely on the quality of the antibodies used [17]. Therefore, antibodies with high specificity and sensitivity are an important requirement. A recent study that compared 4 different ALK antibodies -D5F3 (Ventana), D5F3 (CST), 1A4/1H7 (OriGene Tech), and 5A4 (Abcam) -reported that their sensitivities were 93.8%, 84.4%, 93.8%, and 56.3%, respectively [18]. Notably, the newly developed Ventana monoclonal antibody (D5F3) has greatly improved the specificity and sensitivity of IHC testing [19], and one study has suggested that it can be used as a stand-alone test in cases displaying an unequivocal staining pattern [20]. Recently, based on a fully automated IHC assay developed by Ventana Medical Systems, the Ventana ALK (D5F3) IHC kit was approved to detect ALK fusion genes by the US Food and Drug Administration (FDA). The sensitivity and specificity of this IHC assay have been reported to be 100% and 98%, respectively [21].
The prevalence of ALK rearrangement in patients with NSCLC has been found to range from 1.4% to 13% [22][23][24][25], and to be most common in those with a young age, a never or light smoking history, an abundant signet ring cell or solid pattern histology, and wild-type EGFR or KRAS gene mutations [22][23][24][25][26][27][28][29][30][31][32]. The incidence rate of ALK rearrangement in NSCLC with wild-type EGFR or KRAS gene mutations has been reported to range from 25.7% to 34% [22,29,33]. Although crizotinib, a small-molecular TKI, is now approved for the treatment of advanced ALK-positive NSCLC in view of its favorable therapeutic effect and safety in clinical trials [34,35], the prognostic influence of ALK rearrangement in early-stage NSCLC in the absence of crizotinib treatment remains unclear.
The aim of the present study was to detect overexpression of ALK protein with the Ventana IHC test and to examine the associations of ALK rearrangement with clinicopathologic characteristics and treatment outcomes in patients with early-stage lung adenocarcinoma.

Prevalence and clinicopathologic characteristics of patients harboring ALK rearrangement
Data on a total of 534 completely-resected lung adenocarcinoma patients were analyzed. The Ventana IHC test for ALK rearrangement was performed in all patients. Fortytwo (7.9%) of the 534 patients were IHC-positive for ALK.
Mean overall survival was significantly shorter in ALK IHC-positive patients than in the ALK IHC-negative group (46.0 months vs 57.5 months, respectively; P = 0.007). In addition, the overall mortality rate was significantly higher in the ALK IHC-positive group than in the ALK IHC-negative group (33.3% vs 14.0%, respectively; P = 0.003), and the 2-year overall survival rate after surgery was significantly lower (78.5% vs 90.2%, respectively; P = 0.032). Moreover, mean tumorspecific survival was significantly shorter in the ALK IHCpositive group than in the ALK IHC-negative group (47.8 months vs 58.1 months, respectively; P = 0.026), and the 2-year tumor-specific survival rate of ALK IHC-positive patients was lower than that of ALK IHC-negative patients (82.7% vs 91.3%, respectively; P = 0.092).
Tumor recurrences occurred in a total of 199 patients (37.3%), 25 of whom (12.6%) were in the ALK IHC-positive group and 174 (87.4%) in the ALK IHCnegative group. The mean disease-free survival was 39.0 months in ALK IHC-negative patients versus 26.3 months in ALK IHC-positive patients (P = 0.001), and the tumor recurrence rate was significantly higher in the ALK IHCpositive group (59.5% vs 35.4%, respectively; P = 0.003).
Clinical outcomes, including disease-free survival (DFS), tumor-specific survival (TSS), and overall survival (OS) in the 2 patient groups are shown in Figure 1, and the clinicopathologic characteristics of the 199 patients with recurrences are shown in Table 2. The median age of the ALK IHC-positive group was significantly lower than that of the ALK IHC-negative group (53 years vs 60 years, respectively; P = 0.002). However, there were no significant differences for other clinicopathologic characteristics, such as sex (P = 0.391), smoking status (P = 0.513), tumor status (P = 0.700), lymph node status (P = 0.061), histopathologic subtypes (P = 0.070), or adjuvant therapy (P = 0.266). In addition, there were no statistically significant differences between the 2 groups for the first recurrence site (P = 0.865) and the number of recurrence sites (P = 0.328) [ Table 2].
With multivariate analysis, lymph node involvement was found to be an independent adverse prognostic factor for disease-free survival in ALK IHC-negative patients (HR, 2.32; 95% CI 1.64-3.29; P < 0.001).

DISCUSSION
Previous studies have shown that the prevalence of ALK rearrangement in early-stage NSCLC ranges from 2.4% to 8.6%, depending on the detection method used and the population studied [28][29][30][31][32]. We found that the prevalence was 7.9%, which is consistent with other studies. ALK rearrangement has previously been shown to be more common in light smokers or never-smokers [22][23][24][25][26][27][28][29][30], females [23][24][25]28], the solid predominant with mucin production adenocarcinoma subtype [37][38][39], and young patients [22,24,26,28,30,31]. However, ALK rearrangement has been reported in males (P = 0.039) [22] and some studies have observed no sex difference [23,26,[29][30][31]. Furthermore, one study concluded that ALK rearrangement is not related to smoking status [31]. A recent meta-analysis [40] showed that the prevalence of ALK rearrangement in stage III-IV NSCLC is higher than that  in stages I-II, indicating that ALK rearrangement is more common with higher pathologic stages, which is consistent with our results (IIIB, P = 0.002). Although a previous study [28] showed that ALK-positive lung cancers have unique biologic features with early nodal metastasis despite a small-sized primary tumor (pT1, P = 0.02), we found that ALK IHC-positive lung adenocarcinoma is associated with a higher tumor stage (pT4, P = 0.025). As factors such as race, population, detection method, and the subsequent statistical analysis were not the same in the above studies, this may be the cause of the reported differences. Further larger-scale studies are necessary to resolve these differences. The univariate analysis for disease-free survival in our study showed that patients receiving adjuvant therapy had a higher risk of disease recurrence (P < 0.001). However, this does not mean that they would not benefit from adjuvant therapy. Among the 199 patients with recurrences in our study population, 129 (64.8%) had received adjuvant therapy. Of these 129 patients, 25 (19.4%) were in pathologic stage I, and 104 (80.6%) were in pathologic stages II and III. As the use of adjuvant therapy is stage-dependent, patients with a high pathological stage comprised a high proportion (80.6%) of all patients who received adjuvant therapy. These patients are known to have a poor prognosis.
There is controversy as to whether ALK rearrangement is associated with the prognosis of patients with early-stage NSCLC. A retrospective study from South Korea [29] reported that ALK rearrangement was not associated with overall survival in lung adenocarcinoma patients who were never-smokers (P = 0.720), but disease-free survival was shorter in ALKpositive patients (P = 0.022). Another study, also from South Korea [28], concluded that ALK rearrangement was not a prognostic factor in early-stage NSCLC patients, and a study conducted in Japan [32] came to the same conclusion. The populations of these studies were all Asian. In comparison, a study from European Thoracic Oncology Platform Lungscape Project [30] found that ALK-positivity was related to an improved overall survival in early-stage lung adenocarcinoma patients. However, we found that ALK rearrangement was associated with worse disease-free survival, tumor-specific survival, and overall survival in surgically-resected lung adenocarcinoma patients. Many studies have shown that the prognosis of patients with solid predominant with mucin production adenocarcinoma is poor [41][42][43][44][45][46]. In our study, 15 (35.7%) of the 42 ALK IHC-positive patients had solid predominant with mucin production adenocarcinoma, and analysis of clinicopathologic characteristics showed that  ALK rearrangement was significantly more common in patients with solid predominant with mucin production adenocarcinoma than in patients with other subtypes (P = 0.006). In addition, univariate analysis showed that solid predominant with mucin production adenocarcinoma was associated with a shorter DFS, TSS, and OS, and multivariate analysis showed that this subtype was an adverse prognostic factors for TSS (HR, 1.82; 95% CI 1.07-3.07; P = 0.026) and OS (HR, 1.77; 95% CI 1.08-2.90; P = 0.020). Therefore, whether the poor prognosis of ALK-positive patients is associated with the histopathologic subtype is uncertain.
To our knowledge, this is the largest cohort of patients in which ALK rearrangements have been evaluated by a fully-automated immunochemistry assay in patients with early-stage lung adenocarcinoma, unlike most other ALK-related studies focusing on late-stage patients. We concluded that ALK-positivity is correlated with worse disease-free, tumor-specific, and overall survival. In addition, this is the first study to conduct a detailed clinicopathologic analysis of Chinese patients, and its findings complement existing information on the influence of ALK rearrangement on the prognosis of patients with early-stage NSCLC. However, like most retrospective studies of rare genes, our study has some limitations including observational assessment of tumor Table 4: Disease-free survival analysis for the ALK IHC-positive and ALK IHC-negative resected lung adenocarcinoma patients recurrence, unknown EGFR or KRAS status, and a short follow-up time, Therefore, in a future study, we will obtain treatment information after tumor recurrences, analyze other rare genes, and also analyze the relationship of the solid predominant with mucin production and invasive mucinous adenocarcinoma subtypes with ALK rearrangement and treatment outcome. In addition, the follow-up duration will be extended to obtain more mature data for the overall survival analysis.
In summary, the prevalence of ALK rearrangement detected by Ventana IHC testing in Chinese patients with early-stage lung adenocarcinoma was 7.9%. ALK rearrangement was significantly associated with younger age, and with the solid predominant with mucin production and invasive mucinous adenocarcinoma subtypes. ALK rearrangement was an adverse prognostic factor in surgicallyresected, early-stage lung adenocarcinoma patients.

Patients and specimen characteristics
Patients with lung adenocarcinoma stages I-III who underwent complete surgical resection and detection of ALK rearrangement at Tianjin Medical University Cancer Institute and Hospital between January 2011 and December 2014 were studied. All patients had an adequate quality and quantity of formalin-fixed paraffin-embedded tissue sections available for analysis. Patients who did not undergo curative resection or had prior histories of cancer or neoadjuvant therapy were excluded. None of the patients in the study cohort had received ALK-targeted therapy before tumor recurrence. Data on the patients' clinicopathologic characteristics were retrospectively obtained from the hospital's medical recording system. The study was approved by the Research Ethics Committee of Tianjin Medical University Cancer Institute and Hospital, Tianjin, China.

Clinicopathological evaluation and detection of ALK rearrangement
Pathologic staging of the patients' NSCLC was based on the 7th edition of the tumor-node-metastasis (TNM) classification for lung cancer [47]. Histologic subtypes were determined according to the International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ ATS/ERS) classification of lung adenocarcinoma [36]. ALK rearrangement was detected by the fully-automated Ventana IHC assay using the pre-diluted Ventana anti-ALK (D5F3) rabbit monoclonal primary antibody, together with the Optiview DAB IHC detection kit and Optiview Amplification kit on the Benchmark XT stainer. Evaluation of the test results was determined by a binary scoring system (positive or negative for ALK status), as previously reported [21]. The patients' ALK rearrangement and histopathologic types were determined by 2 experienced pathologists from Tianjin Lung Cancer Center (Z.Z. and Q.Y.) on a blinded basis.

Assessment of clinical outcomes
Clinical outcomes were assessed by determining the patients' overall survival (OS), defined as the time from the date of surgical resection to death from any cause; tumorspecific survival (TSS), defined as the time from date of surgical resection to tumor-related death; and disease-free survival (DFS), defined as the time from the date of surgical resection to disease recurrence or death from any cause. If death or recurrence was not observed in any patient, the censoring date was the last day of follow-up.

Statistical analysis
The prevalence of ALK rearrangement was determined by Ventana IHC testing, and the associations of ALK rearrangement with the patients' clinicopathologic characteristics and treatment outcomes were analyzed using Fisher's exact test (for categorical data) or the Wilcoxon rank-sum test (for continuous data). The DFS, TSS, and OS were plotted using the Kaplan-Meier method and compared by a log-rank test. Multivariate analysis of treatment outcomes was performed using Cox's proportional hazards model. All P-values were based on a 2-sided hypothesis. P-values less than 0.05 were considered statistically significant. All analyses were performed using SPSS ® , version 20.0 (SPSS Inc., Chicago, IL, USA).