Lung cancer is the common carcinoma worldwide in both men and women, and is the leading cause of carcinoma-related death (1). Lung cancer is classified into two main subtypes: Non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC), which are identified in 85 and 15% of diagnosed patients, respectively. NSCLC may be classified further into three subtypes: Adenocarcinoma (ADC) is the most common type (60%), followed by squamous cell carcinoma (SCC) and large cell carcinoma (LCC). The incidence of ADC has increased greatly over the course of the last few decades (2). Several previous studies have identified genetically driven mutations associated with the tyrosine kinase receptor pathway in patients with NSCLC, especially in the case of ADC, including mutations of the genes for epidermal growth factor receptor (EGFR), Kirsten rat sarcoma viral oncogene homolog (KRAS) and v-raf murine sarcoma viral oncogene homolog B1 (BRAF), in addition to translocation of the anaplastic lymphoma kinase (ALK), c-ros oncogene 1 (ROS1) and rearranged during transfection (RET) genes (3-8), which have led to novel treatment outcomes involving specific targeted tyrosine kinase inhibitors (TKIs). In terms of the histological evaluation of ADC, the recently formed International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS) classified invasive ADC according to its predominant growth pattern, including lepidic, acinar, papillary, micropapillary and solid subtypes, as well as ADC variants, including mucinous, colloid, enteric and fetal ADC (9). Recently published studies have demonstrated that the histological subtypes of pulmonary ADC are associated with 5-year survival, overall survival and disease-free survival rates across all tumor stages during treatment with adjuvant therapy (4,10-11). Histological subtypes and molecular factor correlations have also been reported on. Previous studies revealed that the EGFR mutation is associated with the lepidic and papillary subtypes (4,12), whereas solid subtypes are predominantly associated with the ALK mutation (13). Another study suggested that the predominance of papillary features in ADC is predictive of the response to gefitinib, a finding that implied that the EGFR mutation could be more common among this subtype (14).

In advanced stage lung cancer, a large body of evidence derived from clinical trials has led to an understanding that immunotherapy is more effective than systemic chemotherapy in terms of treating unresectable or advanced NSCLC (15-19). In the field of cancer immunotherapy, an understanding of the complex interplay between the tumor and the immune systems has been proposed. Programmed death 1 (PD-1) protein, a T-cell co-inhibitory receptor that binds to its ligand, programmed death-ligand 1 (PD-L1), serves a pivotal role in regulating T-cell activation and proliferation, functioning as an immune checkpoint. The mechanism of PD-1/PD-L1 interaction provides one of the major pathways used by certain tumors to escape immune surveillance (20,21). Studies using various PD-L1 detection antibodies and immunohistochemistry (IHC) assays have identified that a high level of PD-L1 expression in tumor cells is associated with poor prognosis in NSCLC (22,23), whereas in other studies, PD-L1 expression was associated with longer survival rates (24,25). The association between PD-L1 expression and clinicopathological characteristics, including histopathological parameters, revealed no significant correlations in certain studies (23,26), whereas another study identified that PD-L1 expression was associated with more advanced tumor status, node involvement status and the pathological stage. A further study suggested that PD-L1 expression was likely to be predominantly associated with the solid subtype (27).

This aim of the present study was to identify the association between histological subtypes of pulmonary ADC in the Thai population and genetic alterations identified by next-generation sequencing (NGS), as well as PD-L1 expression identified by PD-L1 (clone 22C3) IHC. The association among clinicopathological parameters and PDL1 expression was also explored.

The present study has revealed that the most common histological subtype identified in lung ADC regarding the IASLC/ATS/ERS classification system was the acinar subtype (43.0%). The EGFR mutation was the most common mutation, which was detected in 76 out of 136 (55.8%) cases. The histological subtype with the highest level of significance in terms of its association with the EGFR mutation, based on the present study, was the papillary subtype (P=0.038; OR 2.793). This finding concurred with those of previous studies: A significant association between EGFR and micropapillary, lepidic, acinar or papillary subtypes were identified in studies by Wang et al (29) who analyzed 153 cases, Li et al (30) who investigated 230 cases, Yoshizawa et al (31) who analyzed 440 cases, and Dong et al (32) who analyzed 330 cases. In investigating KRAS-mutated NSCLC, several studies have demonstrated that the KRAS mutation is associated with mucinous ADC, including studies by Dong et al (32) (330 cases) and Marchetti et al (33) (58 cases). A study by Kadota et al (34), that included 864 patients with ADC revealed that mucinous ADC, as well as extracellular mucin production, were also significantly associated with the KRAS mutation. The KRAS mutation in the present study was significantly associated with mucinous ADC variant (P=0.037), a finding that was in accordance with all mentioned studies (32-34). On the other hand, in the present study, the subtype for which the predominantly significant association was identified was the lepidic (P<0.001) subtype, as opposed to the solid subtype. There was a possibility of contrast outcome, which might have been influenced by the KRAS mutation identified in our study: In 23 of 46 cases, KRAS was mainly co-mutated with BRAF and/or EGFR mutations, resulting in predominant patterns that differed with respect to other co-mutations, especially EGFR. Moreover, the acinar predominant subtype was also identified to be statistically significant as a protective prediction factor, and it was unlikely to harbour the KRAS mutation (P=0.002; OR, 0.301). The acinar pattern was predominantly found in 12 cases that harboured the BRAF mutation; however, no statistically significant association was identified in these cases (P=0.393).

PD-1/PD-L1 expression in NSCLC has been widely studied in terms of exploring significant associations with clinicopathological parameters and prognostic indicators. A meta-analysis of 11,444 cases from 47 studies by Zhang et al (35) revealed that the expression of PD-L1 was increased in males, smokers, patients with SCC, and patients with a higher histological grade, larger tumor size, nodal metastasis, and later clinical stage. Another meta-analysis of 1,157 patients from 6 studies by Wang et al (23) concluded that PD-L1 expression was significantly associated with poorly differentiated tumors, resulting in poor overall survival of the patients with NSCLC, whereas a meta-analysis of 1,550 patients from 9 studies by Pan et al (36) demonstrated that the majority of the clinicopathological features, including histological type, smoking status, tumor depth, lymph node status and tumor-lymph node-metastasis (TNM) stage, were not significantly associated with PD-L1 expression, with the exception of tumor differentiation in NSCLC. A meta-analysis of 3,128 patients from 11 studies by Yang et al (37), also failed to reveal any significant associations among clinicopathological characteristics and PD-L1 expression in patients with NSCLC. These data may be compared with the present study, in which specimen size was the only clinicopathological parameter that exerted a significant impact on PD-L1 expression. The larger tumor size (≥0.5 cm in the greatest dimension) was more significantly associated with PD-L1 expression compared with the smaller tumor size (<0.5 cm; P=0.037). This could be accounted for by the fact that, the larger the specimen size, the more likely it is that there will be enhanced PD-L1 expression due to tumor heterogeneity. The remaining clinical parameters, including smoking status, age, sex, stage, and specimen type, were not found to be significantly associated with PD-L1 expression, and these results were in accordance with those of Pan et al (36) and Yang et al (37). The discrepancies identified in the results obtained from these meta-analyses might have been due to the differing sample sizes, differences in subtype among the patients with NSCLC, variable levels of PD-L1 expression detected using different assays, and other interpretation criteria.

More recent studies on PD-L1 expression in pulmonary ADC have also been published. PD-L1 expression in these studies was detected according to standardized Food and Drug Administration (FDA) approved assays, including the 22C3 PharmDx assay (Agilent Technologies/Dako) (38,39), the Ventana PD-L1 SP142 assay (Ventana Medical Systems Inc.) (40), and the Ventana PD-L1 SP263 assay (Ventana Medical Systems Inc.) (41,42). Each IHC antibody clone is associated with a specific type of targeted therapy. In the present study, Dako PD-L1 IHC 22c3 pharmDx was used, which is the IHC study that makes use of pembrolizumab. Pembrolizumab has been evaluated in large-scale clinical trials; patients with TPS≥50% are eligible to receive pembrolizumab as first-line therapy, and these have been shown to have significantly improved progression-free survival and overall survival rates compared with those patients with TPS<1% or 1-49% (43). Patients with TPS≥1% subsequently treated in a trial with pembrolizumab were shown to have significantly improved overall survival rates and risk-benefit profiles that reached the level of clinical relevance compared with standard care chemotherapy (43). Among the PD-L1 expression studies in patients with pulmonary ADC using PD-L1 IHC clone 22C3, Pan et al (38) identified PD-L1 expression (TPS≥1%) in 4.1% of the total of 221 cases, including 1 (0.5%) case with TPS≥50%. Miyazawa et al (39) identified PD-L1 expression (TPS≥1%) in 21% of 78 cases, including TPS≥50% in 7 (9%) cases, percentages that were more comparable with those in our study (TPS≥1%, 21 vs. 23%; and TPS≥50%, 9 vs. 5%). They also reported that PD-L1 tends not to be expressed in early invasive ADC or well differentiated subtypes, similar to our finding that negative PD-L1 (TPS<1%) expression was found to be associated with the lepidic subtype (P=0.007). On the other hand, Driver et al (40) identified a larger number of cases of PD-L1 expression (44% of 125 cases) compared with our study, possibly due to their use of a different PD-L1 IHC clone (SP142), and the criteria for positive PD-L1 expression from their study also included positivity of infiltrating immune cells.

In histological terms, solid predominant ADC was found to be the cancer type that was consistently associated with high PD-L1 expression. The studies by Pan et al (38) and Driver et al (40) demonstrated significantly higher PD-L1 expression levels in solid vs. non-solid subtypes of ADC. Our results also revealed a significant association between the solid predominant histological subtype and PD-L1 expression, consistent with those two studies. Moreover, the large genetic and immune profiles of solid ADC published by Dong et al (42) from 194 patients with lung ADC also demonstrated that solid ADC is significantly associated with a higher smoking index, tumor mutation burden and adaptive immune resistance (highly PD-L1-expressing tumor cells and highly CD-8 positive tumor-infiltrating lymphocytes), findings which draw attention to the benefits associated with using PD-1/PD-L1 immune checkpoint inhibitors in patients with solid predominant ADC.

It should be noted that our results did reveal certain differences compared with those of previous studies, possibly resulting from several limitations. Firstly, in pattern recognition that exist in small biopsies, which may be distortions of the true predominant pattern one would identify from the whole resection specimen. Secondly, dynamic changes in expression occurring between the primary tumour and the metastatic sites may also affect the histological pattern and/or genetic alterations or PD-L1 expression, and these possibilities should be investigated in future studies.

In conclusion, in the present study patients from a Thai population with lung ADC were collected and analyzed according to the histological subtypes in terms of their genetic mutations and PD-L1 expression levels. The EGFR and KRAS mutations were the first and second most common genetic alterations identified in the present study. The papillary histological subtype was associated with a high incidence of EGFR mutations. The lepidic subtype and mucinous ADC variant were significantly associated with the KRAS mutation. PD-L1 expression (TPS≥1%) was identified in 26% of the patients, and the solid subtype was associated with PD-L1 protein expression. Genetic mutations and the PD-L1 expression level therefore may serve as representative predictive factors for lung ADC, and may be of use in terms of targeted immunotherapeutic strategies. The histological subtype associated with the lung tumor is worthy of investigation in terms of identifying the genotype-phenotype association. These results may be helpful in selecting appropriate areas of the heterogeneous tumor feature for the patients who require personalized therapy, and predicting their prognosis.