Epidemiological study of HER-2 mutations among EGFR wild-type lung adenocarcinoma patients in China

The inclusion criteria were as follows: patients over 18 years old; confirmed diagnosis of lung adenocarcinoma; wild-type EGFR gene; Eastern Cooperative Oncology Group score of 0–2 points; sufficient formalin-fixed, paraffin-embedded samples for HER-2 mutation detection and DNA sequencing; and patient who did not receive preoperative neoadjuvant therapy. Exclusion criteria were as follows: patients with other tumors or severe diseases, including psychiatric illnesses; and pregnancy. The study protocol was approved by the Ethical Review Committee of the Shanghai Pulmonary Hospital, and the clinical registration number was ChiCTR-OCH-13003507. The study design is outlined in Fig. 1.

A DNA/RNA Isolation kit (Amoy Diagnostics Co., Xiamen, China) was used according to the manufacturer’s instructions to extract DNA from formalin-fixed, paraffin-embedded tissue samples. Mutations were detected using the Human EGFR Mutation, Human echinoderm microtubule-associated protein-like 4–ALK Fusion Gene, Human ROS1 Fusion Gene, and Human HER-2 Mutation Detection kits (Amoy Diagnostics Co.). PCR amplification conditions were as follows: 95 °C 5 min; 15 cycles of 95 °C for 25 s and 72 °C for 20 s; and 31 cycles of 93 °C for 25 s, 60 °C for 35 s, and 72 °C for 20 s. All samples positive for HER-2 mutations were confirmed by DNA sequencing using primers with the following sequences: 5'GCC ATG GCT GTG GTT TGT GAT GG3' (forward) and 5'ATC CTA GCC CCT TGT GGA CAT AGG3', which amplified a 342-bp fragment in exon 20 of the HER-2 gene.

HER-2 expression in tissue was detected using a rabbit monoclonal anti-HER-2/Erbb2 (29D8) antibody (diluted 1:200) (Cell Signaling Technology, Danvers, MA, USA). The intensity of the signal was determined as previously described [10], which was as follows: 0, no staining; 1+, < 5 % of cells with strong staining or 5 %–10 % of cells with weak staining; 2+, 5 %–10 % of tumor cells with strong staining or > 10 % of tumor cells with weak staining; and 3+, > 10 % of cells with strong staining. Membranous staining was set as the criteria, and cytoplasmic staining was excluded from the scoring of stain intensity. Scores of 0/1+ and 2+/3+ were counted as negative and positive HER-2 expression, respectively. Samples were independently evaluated by two pathologists.

SPSS v.17.0 software (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Disease-free survival (DFS) was defined as the time between the start of surgery to the first recorded disease recurrence/metastasis. Overall survival (OS) was defined as the time between the start of surgery to disease-related death. The correlations between HER-2 mutation and clinical characteristics, as well as between HER-2 mutation and protein expression, were evaluated with Pearson’s χ² test or Fisher’s exact test. Survival was analyzed with the Kaplan-Meier test, and survival comparisons were carried out with the log-rank test. The two-sided significance level was set at P < 0.05.

Patient characteristics are summarized in Table 1. A total of 1316 formalin-fixed, paraffin-embedded samples from surgical patients at Shanghai Pulmonary Hospital were initially screened, including 1107 cases of lung adenocarcinoma, of which 456 wild-type EGFR lung adenocarcinoma cases were analyzed for HER-2 mutation. The final sample set included 250 males and 206 females; the median age was 60 years (range: 26–82 years), with 135 and 321 cases ≥ 65 and < 65 years, respectively. In addition, 295 patients had never smoked or were light smokers while 161 were heavy smokers. In terms of disease staging, 262 cases were Stage I, 46 were Stage II, 91 were Stage III, 23 were Stage IV, and 34 were unconfirmed. The 437 cases were classified according to lung adenocarcinoma subtype based on the 2011 criteria. There were 40 cases of lepidic predominant adenocarcinoma (LPA), 187 of acinar predominant adenocarcinoma (APA), 100 of papillary predominant adenocarcinoma (PPA), 35 of solid predominant adenocarcinoma (SPA), 31 of invasive mucinous adenocarcinoma (IMA), 14 of adenocarcinoma in situ (AIS), 14 of minimally invasive adenocarcinoma (MIA), 14 of micropapillary predominant adenocarcinoma (MPA), and two of enteric adenocarcinoma (EA).

ARMS analysis was used to detect HER-2 exon 20 mutations in 456 cases of wild-type EGFR lung adenocarcinoma; 22 (4.8 %) were found to be positive. The presence of an echinoderm microtubule-associated protein-like 4/anaplastic lymphoma kinase (ALK)–ROS proto-oncogene (ROS) 1 fusion gene was also examined in 390 cases; 331 were found to harbor wild-type ALK/ROS1. The HER-2 mutation rate in patients with wild-type EGFR/ALK/ROS1 (i.e., triple-negative samples) was 6.7 %; the 22 samples were analyzed by PCR combined with direct sequencing. For three of the samples, sequencing failed due to insufficient sample quantity. Results for the remaining 19 samples were consistent with those obtained by ARMS: 13 cases were 2324_2325 ins12 (atacgtgatggc), three were 2326_2327insTGT, two were 2325_2326 ins12 (tacgtgatggct), and one was 2331_2332 ins9 (ggctcccca) (Fig. 2).

HER-2 protein expression was assessed in 357 samples by immunohistochemistry. Results were interpreted as previously described [10]. There were 55 positive samples (15.4 %), of which 43 were 2+ (12.0 %) and 12 were 3+ (3.4 %) (Fig. 3).

The presence of HER-2 mutations was not correlated with clinical characteristics, including gender, age, smoking status, histological subtype, and disease stage. However, the mutation rate differed among lung adenocarcinoma subtypes: the 22 cases of HER-2 mutation were observed in APA (36.4 %), PPA (36.4 %), MIA (13.6 %), SPA (9.2 %), and IMA (4.5 %), whereas no mutations were detected in LPA, MPA, EA, or AIS (Table 1 and Fig. 4). Furthermore, there was no correlation between HER-2 mutation and protein expression. Simultaneous HER-2 mutation and HER-2 overexpression was observed in only two samples.

At the last follow-up, 123 patients achieved DFS, and OS data were collected from 39 patients. Most of the remaining patients continued to survive, and follow-up data are needed. A subset of patients was lost to follow-up. HER-2 mutation was not associated with DFS (median: 16.7 vs. 17.0 months; P = 0.966). The difference with OS was also not statistically significant (median: 34.2 vs. 47.6 months; P = 0.844). DFS data were obtained from 79 and 17 patients who were negative and positive for HER-2, respectively, by immunohistochemistry. However, DFS did not differ significantly between the two groups (median: 14.4 vs. 12.8 months, P = 0.238). The 14 HER-2-negative and one HER-2-positive patients had similar OS (median: 24.9 vs. 48.0 months, P = 0.075) (Fig. 5). We also analyzed the DFS data of the HER2 IHC(-) versus HER2 IHC (+) in stage I/II and stage III/IV cases, respectively (Additional file 1: Figure S1) and found that none of them had a significantly difference.