High mesothelin expression in advanced lung adenocarcinoma is associated with KRAS mutations and a poor prognosis.

Mesothelin is a cell surface glycoprotein which is highly expressed in several epithelial cancers and may have a role in cell adhesion and metastases. In this study, we used prospectively obtained clinical and pathological data to characterize mesothelin expression in advanced lung adenocarcinoma. Tissue was obtained from patients who underwent molecular profiling of potentially actionable genes on a trial of molecular profiling and targeted therapies in advanced thoracic malignancies. We immunohistochemically evaluated the intensity, and the percentage of cells expressing mesothelin in 93 advanced lung adenocarcinomas. The evaluation was blinded for molecular data and outcome. Mutations of EGFR, KRAS, BRAF, AKT1, PIK3CA and HER2 were assessed by pyrosequencing; HER2 amplification and ALK translocation were assessed by fluorescence in situ hybridization. 53% of advanced lung adenocarcinomas expressed mesothelin to some degree; high mesothelin expression, defined as mesothelin positivity in more than 25% of cells, was found in 24% of patients. High mesothelin expression was associated with inferior survival (median 18.2 months vs. 32.9 months; P = 0.014). High mesothelin expression was strongly associated with mutant KRAS (P < 0.0001) and wild-type EGFR (P = 0.002). Our results provide strong rationale to explore anti-mesothelin targeted therapies in advanced lung adenocarcinoma especially in the KRAS-mutant subgroup.


INTRODUCTION
Lung cancer is the leading cause of cancer-related death worldwide, accounting for more than one million deaths every year. [1,2] Non-small-cell lung cancer (NSCLC) constitutes approximately 85% of lung cancers and about 40% of patients with newly diagnosed NSCLC have advanced disease. [1] In the past decade, the standard of care for patients with advanced disease was platinumbased chemotherapy, which improved survival, quality of life and symptom control compared with supportive care. [3] However, the median overall survival is only about a year; only 3.5% of patients are alive five years after diagnosis. Chemotherapy is also associated with high morbidity. [4] In recent years, identification of "druggable" oncogenic alterations such as mutations in EGFR and ALK translocations, and development of drugs that specifically target these mutations have led to a substantial improvement in the prognosis of patients with advanced lung cancer. [5] However, "druggable" alterations have www.impactjournals.com/oncotarget been detected in less than half of all advanced NSCLC patients. [6] Mutations in the KRAS oncogene, for example, accounts for 20-30% of lung adenocarcinomas, yet no targeted agents are currently available. Hence there is an unmet need to develop new, effective and minimally toxic targeted therapies in advanced NSCLC.
Mesothelin is a 40-kDa cell surface glycoprotein that is present on normal mesothelial cells lining the pleura, peritoneum and pericardium. [7] Mesothelin expression in normal human tissues is observed only in a single layer of mesothelial cells lining the pleura, peritoneum and pericardium, surface epithelial cells of the ovary, tunica vaginalis, rete testis and the tonsilar and fallopian tube epithelial cells. [8] However, mesothelin is highly expressed in several cancers, including epitheloid mesotheliomas, pancreatic, biliary adenocarcinomas, gastric and ovarian cancers. [8][9][10][11] The high expression of mesothelin in cancers have prompted its therapeutic targeting using a variety of strategies including immunotoxins, monoclonal antibodies, antibody drug conjugates, vaccines and adoptive T cell therapy. [12][13][14] We recently demonstrated major and durable tumor regressions in chemotherapy-refractory patients with advanced epitheliod mesothelioma using the antimesothelin immunotoxin SS1P. [15] Conceptually, SS1P and other mesothelin-targeted therapies might also confer efficacy in other tumor types that over-express mesothelin. Identifying these cancers could thus expand the therapeutic utility of these therapies. Mesothelin expression has been demonstrated in approximately 30-70% of lung adenocarcinoma. [16][17][18][19][20] However, these studies were retrospective, analyzed a limited number of samples, did not provide clinical information and did not study the patterns of expression in detail.
Given the paucity of data and the heterogeneous and conflicting results of prior investigations, we sought to determine the expression patterns and prognostic value of mesothelin in advanced lung adenocarcinoma and the association of mesothelin expression with other molecular alterations and clinico-pathologic variables. We demonstrate here that 24% of advanced lung adenocarcinoma express high levels of mesothelin, and that high mesothelin expression is associated with EGFR wild-type and mutant KRAS and, independent of the mutation status, is associated with decreased overall survival. Our results suggest that mesothelin targeted therapies could be useful in patients with KRAS mutant lung cancer, a subtype for which no targeted therapies are currently available.

RESULTS
From February 2011 to December 2012, 272 patients with NSCLC enrolled and underwent molecular profiling in the pilot trial of molecular profiling and targeted therapies in advanced thoracic malignancies at the Center for Cancer Research, National Cancer Institute. [21] Two hundred and eleven had adenocarcinoma histology with 179 having advanced disease (stages III or IV) at diagnosis. Ninety three patients had adenocarcinoma histology, stage III or IV at diagnosis and had adequate FFPE samples available for further studies.

Patient characteristics
The clinicopathological characteristics are summarized in Table 1. The median age of all patients was 61 years and 53 (57%) patients were female. The patient population was predominantly Caucasian (74%) and 35% were never-smokers. Oncogenic alterations in EGFR, KRAS mutations and ALK translocations were found in 25%, 29% and 11% patients respectively.

Mesothelin expression in lung adenocarcinoma
The intensity and percentage of cells expressing mesothelin in advanced lung adenocarcinoma is shown in Table 2. Of the 93 tumors tested, any expression of mesothelin was observed in 49 (53%) and high expression in 22 (24%). Figure 2 shows representative images of mesothelin expression in lung cancers. Mesothelin expression was observed in the cytoplasm alone in 10 (20%), membrane alone in 14 (29%) and in both membrane and cytoplasm in 25 (51%).

Association of mesothelin expression with clinico-pathological variables
The association between mesothelin expression and clinico-pathological characteristics in advanced lung adenocarcinoma is shown in Table 3. There was no association between mesothelin expression and age, sex, or race. There was a strong association between KRAS mutation and mesothelin expression. Twenty one of 49 (43%) of tumors that expressed at least some mesothelin (> = 1% cells) had a KRAS mutation whereas only 6 of 43 (14%) mesothelin negative tumors expressed this mutation (P = 0.003; Figure 3A). The association with KRAS mutation was stronger for high mesothelin expressors (> = 25% cells): 15 of 22 (68%) mesothelin positive tumors with high mesothelin expression had a RAS mutation compared with only 12 of 70 (17%) of mesothelin negative tumors (P < 0.0001; Figure 3B).
Mesothelin expression in more than 25% of cells was also significantly associated with smoking and wildtype EGFR. Tumors from 19 of 60 (32%) current or exsmokers expressed mesothelin in more than 25% cells, whereas tumors from 3 of 33 (9%) never-smokers which expressed mesothelin (P = 0.02). Eight of 49 (16%) of mesothelin positive (> = 1% cells) patients had EGFR mutations compared to 12 of 41 mesothelin negative patients (P = 0.20; Figure 3C). None of the 22 mesothelinexpressing tumors with high mesothelin expression www.impactjournals.com/oncotarget (> 25% cells) had EGFR mutations compared to 20 with EGFR mutations among 68 mesothelin negative samples (P = 0.0024; Figure 3D). No association was found between mesothelin expression and the other molecular alterations evaluated: amplification of HER2, ALK translocation and mutations in BRAF, AKT1, PIK3CA or HER2. By logistic regression analysis, the association between KRAS mutation and high mesothelin expression was independent of co-variates including smoking, sex, stage at biopsy, and age.

Mesothelin expression and survival
There was no difference in survival from diagnosis between patients with any mesothelin expression and patients with non-mesothelin expressing tumors. (median 32.4 months vs. 26.2 months; P = 0.29) ( Figure 3E). However, patients with high mesothelin-expressing tumors had significantly shorter overall survival compared with patients with low or no mesothelin expression. (median 18.2 months vs. 32.9 months; P = 0.014) ( Figure 3F). The median potential followup (from date of diagnosis until analysis, February 1, 2014) for all patients was 38.6 months. There was no association between the pattern of mesothelin expression (cytoplasmic vs. membranous) and survival.
A Cox proportional hazards regression model was used to evaluate the association between mesothelin expression and overall survival from date of diagnosis. The following covariates were included: age, smoking, KRAS and EGFR mutations. Mesothelin vs. nonmesothelin expression remained not significantly associated with survival after adjusting for clinical covariates. However, high mesothelin expression remained significantly associated with survival when adjusted for covariates. Specifically, in a model beginning with the parameters as stated above, backward selection resulted in a model including only high mesothelin expression (P = 0.015; Hazard Ratio (HR)=1.94; 95% confidence interval (CI), 1.14-3.30) and smoking history (P = 0.045; HR = 0.57; 95% CI, 0.33-0.99) as parameters retaining their joint statistical significance.

DISCUSSION
In this study, we used prospectively obtained clinical and pathological data to characterize mesothelin expression in lung cancer. We evaluated the intensity, and the percentage of cells expressing mesothelin in tissue obtained from 93 patients with advanced lung adenocarcinoma who underwent molecular profiling for potentially actionable genes using a multi-platform approach. We found that approximately 50% of advanced lung adenocarcinomas express mesothelin and high mesothelin expression, defined as mesothelin positivity in more than 25% of cells, was associated with inferior survival. Importantly, we found that high-expression of mesothelin was strongly associated with mutant KRAS and wild-type EGFR.
In early-stage lung adenocarcinoma, a recent retrospective study found mesothelin expression in 69% of tumors. Patients with high mesothelin expression were more likely to have KRAS mutations, compared with patients with low expression [22]. In the advanced lung cancer setting, prior studies of mesothelin expression are retrospective, involved small number of patients and a very heterogeneous population. [16][17][18][19][20] In addition to being the largest series of prospective assessment of mesothelin expression in lung cancer, the strengths of our study include a uniform patient population, all of whom enrolled in a trial of tumor molecular profiling, availability of robust clinical, pathologic, immunohistochemical, and mutational data and long-term follow-up.
The mechanistic association of KRAS mutation to mesothelin is not delineated by the present study. The observed enrichment for mesothelin expression within the KRAS mutated lung adenocracinoma population could occur due to a direct or indirect regulatory relationship between the two proteins. Transcription enhancer factor (TEF-1) is known to be upregulated in tumors from a KRAS mutated, genetically engineered mouse model of lung cancer. [23] TEF-1 has also been shown to directly bind an upstream enhancer sequence in the mesothelin gene, causing upregulation of mesothelin transcription. [24] However, concordance of KRAS mutation and mesothelin expression have not been observed in mesothelioma and high grade serous ovarian cancers where robust mesothelin expression is nearly universal [8,17] and KRAS mutation quite rare. [25,26] By contrast, almost all pancreatic adenocarcinomas express both mesothelin and mutated KRAS. However, the temporal relationship of their expression in tumor development is not suggestive of a regulatory relationship since KRAS mutation is one of the earliest detectable changes in premalignant PanIN lesions while mesothelin expression does not occur until much later in adenocarcinoma development. [27,28] These data from other tumor types suggest that KRAS mutation alone is neither necessary nor sufficient to universally induce mesothelin expression and that other pathways must also participate in regulation of mesothelin. Our data are consistent with these observations since complete concordance between KRAS mutation status and mesothelin expression was not observed in our study. Nevertheless, this does not preclude the possibility that there may be a direct regulatory relationship in some lung tumors. Further studies would be required to establish whether this does occur. The results of this study and previous work by others provide strong rationale that anti-mesothelin targeted therapy should be explored as a therapeutic modality in advanced lung adenocarcinomas. Our prior studies have demonstrated that mesothelin mRNA and protein are present in a substantial number of lung adenocarcinoma cell lines and that SS1P, an antimesothelin recombinant immunotoxin, was cytotoxic to mesothelin expressing lung cancer cell lines with IC50 values ranging from 2 to 5 ng/mL. [29] We have recently shown that in patients with chemotherapy refractory malignant mesothelioma, SS1P in combination with pentostatin and cyclophosphamide leads to durable tumor regression. [15] These lines of compelling evidence suggest that SS1P may confer clinical activity in mesothelin-expressing lung adenocarcinomas. Several other mesothelin-targeted therapies including a vaccine [30], antibody drug conjugate [31] and a monoclonal antibody [32] are undergoing phase I/II clinical trials.
The current study provides a strong rationale to target mesothelin in advanced lung adenocarcinoma and suggest that clinical trials of mesothelin-directed therapies in lung cancer should focus on patients with  Representative images are depicted (original magnification x400). Focal cytoplasmic immunostaining of 2+ intensity in 15% cells (A), membranous and cytoplasmic immunostaining of 2+ intensity in 1% cells (B), membranous immunostaining in of 3+ intensity in 30% cells (C), membranous and cytoplasmic immunostaining of 3+ intensity in 60% cells (D), membranous and cytoplasmic immunostaining of 3+ intensity in 80% cells (E), membranous and cytoplasmic immunostaining of 3+ intensity in 100% cells (F). www.impactjournals.com/oncotarget KRAS mutations, the most commonly mutated oncogene in NSCLC and one which has proven intractable even in the era of targeted therapy.

Patients
Patients were prospectively enrolled in a pilot trial of molecular profiling and targeted therapies in advanced thoracic malignancies (Figure 1) (ClinicalTrials.gov Identifier: NCT01306045). [21] Main eligibility criteria were histologically confirmed, advanced stage lung cancer and Eastern Cooperative Oncology Group Performance Status ≤ 2. Patients had to have biopsiable disease and be willing to undergo biopsy for molecular profiling or have formalin-fixed paraffin-embedded (FFPE) tissue blocks suitable for molecular profiling analysis. Patients underwent new tumor biopsies when possible but archival tumor samples were also used. Tumor samples were screened for mutations in AKT1, BRAF, EGFR, HER2, KRAS, NRAS, and PIK3CA by pyrosequencing; HER2 amplification and ALK translocation were assessed by Fluorescence in situ hybridization (FISH). Patients who did not have a FFPE sample available for mesothelin immunohistochemistry (IHC) were excluded from this  analysis. All patients were followed for survival. The protocol was approved by the National Cancer Institute Institutional Review Board.

Tumor mesothelin expression
Mesothelin IHC was performed on tumor samples obtained using monoclonal antibody 5B2 (Novocastra/ Leica, Bannockburn, IL) at 1:40 dilution. One slide was tested for each case. Heat-induced epitope retrieval (20 mins in citrate buffer, pH 6.0) we performed prior to incubation with primary antibody.. The detection was performed with Ventana Ultra View detection kit with DAB chromogen. All immunostaining was done using positive and negative controls and results were highly concordant.
Immunohistochemical staining was evaluated by a pathologist (MM) with special expertise in evaluating anti-mesothelin IHC who was blinded to the clinical and molecular data. The positivity (strength of labeling) was assessed as negative (no labeling), weak 1+, moderate 2+, and strong 3+, and the percentage of positive cells was also estimated. High mesothelin expression was defined as mesothelin expression in > 25% cells.

Statistical analysis
The association of dichotomous parameters with mesothelin positivity was determined using Fisher's exact test. The association of race with mesothelin positivity was determined by Mehta's modification to Fisher's exact test. [34] The difference in age between the two mesothelin groups was determined by a Wilcoxon rank sum test. Overall survival time was defined as the time from date of diagnosis of metastatic cancer to date of death or last follow-up. The association between mesothelin and survival was presented using a Kaplan-Meier curve and a log-rank test. The association of mesothelin and survival adjusting for demographic, genetic, and clinical parameters was determined by a Cox proportional hazards model. The association between KRAS mutation and mesothelin expression after adjusting for other clinical parameters was determined by logistic regression analysis. All p-values are two-tailed and presented without adjustment for multiple comparisons.