|Year : 2019 | Volume
| Issue : 1 | Page : 16-20
Determination of ROS1 positivity by immunohistochemistry in a multicentric cohort of 426 non-small-cell lung cancer cases in India
Jugnu Jain1, Deepthi Chinta1, Umeshnandan B Jayaraman1, Neha Pathak1, Manpreet Kaur1, Soma Chatterjee1, Meenakshi Swain2, Vijay Anand Reddy3
1 Sapien Biosciences Private Limited, Apollo Health City, Hyderabad, Telangana, India
2 Department of Pathology, Apollo Hospitals, Apollo Health City, Hyderabad, Telangana, India
3 Department of Radiation Oncology, Apollo Cancer Hospital, Apollo Health City, Hyderabad, Telangana, India
|Date of Web Publication||9-Sep-2019|
Sapien Biosciences Private Limited, Apollo Health City, Jubilee Hills, Hyderabad - 500 096, Telangana
Source of Support: None, Conflict of Interest: None
Introduction: ROS1 is a receptor tyrosine kinase of the insulin receptor family that acts as a driver oncogene in 1%–2% of non-small cell lung carcinoma (NSCLC) patients via a gene translocation between ROS1 and other genes, the most common one being CD74. Histologic and clinical features that are associated with ROS1 translocations include never smokers, younger age, and adenocarcinoma (AC) histology. The identification of ROS1 fusion is important because it responds to targeted therapies such as crizotinib that can greatly improve the survival of the patient. The goal of this study was to determine the percentage of population harboring ROS1 mutation in India.
Materials and Methods: ROS immunohistochemistry (IHC) method was optimized using Cell Signalling Technology, ROS antibody (ROS1 clone D4D6 rabbit mAb), and appropriate positive and negative controls in a fully automated procedure using Ventana automated stainer. Approximately 2000 retrospective lung cancer cases from six different Apollo hospitals were identified. The formalin-fixed paraffin-embedded blocks and corresponding diagnostic and demographic data were retrieved. A total of 426 cases with confirmed diagnosis of NSCLC were then tested by IHC for ROS1.
Results: Given the retrospective nature of the cancer cases and diagnostic data being used for the study, there was high attrition in the availability of the blocks and data. Among the 2000 cases examined, there were 604 cases that were initially reported as lung cancer in the hospital medical records and for whom corresponding blocks were available, but blocks for only 485 cases had sufficient tumor content, as judged by hematoxylin and eosin staining and microscopic examination, to enable testing for ROS1 staining. Of these 485 cases, 225 were confirmed AC cases. A single case stained positive for ROS1 among the 225 AC cases. The percentage is calculated to be 0.44% based on this single positive case. No positive case was observed in squamous cell carcinomas or other cancers of the lung such as neuroendocrine tumors, germ cell tumors, and adenoid cystic tumors.
Conclusion: The percentage positivity for ROS1 in the study comprising 225 Indian NSCLC AC cases appeared similar to the 0.3%–1.6% range reported for smokers and lower than 0.9%–1.7% positivity reported in mixed patient populations.
Keywords: Crizotinib, immunohistochemistry, NSCLC, ROS positivity, targeted therapy
|How to cite this article:|
Jain J, Chinta D, Jayaraman UB, Pathak N, Kaur M, Chatterjee S, Swain M, Reddy VA. Determination of ROS1 positivity by immunohistochemistry in a multicentric cohort of 426 non-small-cell lung cancer cases in India. Cancer Res Stat Treat 2019;2:16-20
|How to cite this URL:|
Jain J, Chinta D, Jayaraman UB, Pathak N, Kaur M, Chatterjee S, Swain M, Reddy VA. Determination of ROS1 positivity by immunohistochemistry in a multicentric cohort of 426 non-small-cell lung cancer cases in India. Cancer Res Stat Treat [serial online] 2019 [cited 2020 Jul 4];2:16-20. Available from: http://www.crstonline.com/text.asp?2019/2/1/16/266440
| Introduction|| |
ROS1 positivity is a receptor tyrosine kinase of the insulin receptor family that acts as a driver oncogene in 1%–2% of non-small-cell lung cancer (NSCLC) patients via a gene translocation between ROS1 and other genes, the most common one being CD74., Histologic and clinical features that are associated with ROS1 translocations include adenocarcinoma (AC) histology, younger age, and never-smokers, a profile similar to ALK-positive cases.
The interest in ROS1 gene arrangement arises because crizotinib, an approved small molecule oral therapy for ALK mutation-positive NSCLC patients, is also effective for patients with these ROS1 mutations., Hence, ROS1 gene arrangement is now identified as an actionable target in NSCLC because the identification of patients harboring this ROS1 gene rearrangement may benefit from crizotinib.
At present, there are no data available on the percentage of population harboring ROS1 mutation in India. Sapien Biobank has access to a wide variety and number of cancer specimens and associated data. Using these specimens and data, Sapien screened 485 retrospective samples of lung cancer patients using immunohistochemistry (IHC), for ROS1 mutation to assess the percentage of patients who may benefit from targeted therapy. In addition to determining the percentage positivity in our collection, the IHC method standardized in this study may also help in the development of a less expensive and more readily available diagnostic method for detecting ROS1 mutation in future NSCLC patients.
| Materials and Methods|| |
Ethical approval for the study
The samples and associated data were obtained by Sapien Biobank with approval from the individual institutional ethics committees (IECs) of each Apollo hospital from where the blocks were retrieved, namely Ahmedabad, Bangalore, Chennai, Delhi, Kolkata, and Hyderabad. Use of the archived samples for this project was further approved with waiver of informed consent requirement by the IEC of the biobank, given the retrospective nature of analysis (Protocol SBS-IEC-2017-01 titled “Determination of the frequency of different types of mutations in non-small cell lung carcinomas [NSCLC] in Indian patient population”).
NSCLC formalin-fixed paraffin-embedded blocks and data
The biobank team identified 2000 retrospective lung cancer cases drawn from multiple hospitals from 1997 to 2016. Sapien had already archived and annotated cancer samples from 2006 to 2016 at Hyderabad, 2007–2013 from Chennai, and 2009–2013 from Delhi; hence, the corresponding pathological diagnosis of lung cancer was captured in the biobank database and readily identified. The remaining cases were identified using keywords such as lung cancer, NSCLC, ICD code from hospital information and technology, and medical records department records manually for the years that blocks had been preserved in different Apollo hospitals. The FFPE samples and data were coded to protect patient identity and privacy as per the Indian Council of Medical Research guidelines. Most of the shortlisted cases (1396) did not have blocks available because the sample (s) had been returned to the patient over the years or exhausted during clinical investigations and most samples in lung cancer tend to be small biopsy specimens. Therefore, only 604 cases which had blocks available were used for further investigations. Basic demographic data such as age and sex of the patient and diagnostic data for the biopsy or surgical tissue samples were largely available and retrieved from the hospital medical records.
The samples comprised a spectrum of different stages and grades, as well as histological classifications including AC and squamous cell carcinoma (SCC), as listed in [Table 1]. Most cases (458) were mentioned as primary lung cancers in the histopathology reports, with 27 being classified as metastatic tumors without specifying the site of the primary tumor. Such metastatic samples (27), along with histologies other than confirmed NSCLCs such as germ cell tumors (2), adenoid cystic tumors (4), benign tumor (1), suspicious of lung malignancy (12), mesothelial carcinoma (1), and neuroendocrine tumors (11), totaling 59 cases, were also included in the ROS staining, but excluded from the deeper analysis of the 426 confirmed NSCLC cases.
|Table 1: Non-small-cell lung cancer cases, grouped by the clinical diagnosis data available with them|
Click here to view
The quantity and quality of the blocks and tumor tissue were checked with a hematoxylin and eosin (H and E) stain by a trained clinical pathologist. This was important because the majority of the blocks comprised biopsies with small amount of tissue. The H and E stain also served to confirm the diagnosis, thereby eliminating any possible error in diagnostic details that may have crept in due to the data having been retrieved from old records.
Validation of the ROS1 antibody for immunohistochemistry
Based on the literature comparing IHC method favorably with fluorescence in situ hybridization (FISH) for ROS1 detection with a concordance of well above 90%, in this study, Cell Signalling Technology (CST), ROS antibody (ROS1 clone D4D6 rabbit mAb, cat No. 3287) was purchased. The recommended antibody dilution buffer (SignalStain antibody diluent, cat No. 8112) and positive and negative control slides for validation of IHC with this antibody were obtained from CST. The positive and negative controls comprised slides made from cell pellet blocks, with the positive control being HCC78 cells overexpressing ROS1 mutation and the negative control being HeLa cells lacking expression of Ros1 protein (references and data on CST website).
The manufacturer recommended 1:200 dilution of the antibody. Previous publications using this antibody reported either a 1:200 or a 1:100 dilution for IHC. Hence, three dilutions of the ROS1 antibody, 1:200, 1:150, and 1:100, were tested with the positive and negative control slides provided by the manufacturer. The 1:150 dilution was judged to be optimal by the pathologist, with the positive control staining well and no signal being observed in the negative control. Subsequent NSCLC slides were stained at 1:150 dilution in the SignalStain diluent recommended by the manufacturer.
For NSCLC paraffin-embedded blocks, IHC was performed on unstained tissue sections of 4 microns each using poly-L-lysine-coated slides. The slides were stained with a fully automated procedure using Ventana Benchmark XT automated stainer (Ventana Medical Systems, Tucson, AZ) at pH 9.0 in an ethylenediaminetetraacetic acid buffer. The primary antibody was manually applied at 1:150 dilution, and the slides were incubated at 37°C for 1 hour. Amplification and detection were performed using the UltraView Ventana Amplification and DAB detection kit (cat No. 760-500).
| Results|| |
A total of 604 lung cancer cases, including germ cell tumors, neuroendocrine tumors, suspicious of lung malignancy, adenoid cystic tumors, etc., blocks and associated diagnostic data were retrieved for this project. Of the 426 NSCLC cases, majority of them, 310 (72.7%), were male patients. The median age was 60 years (range, 7–87). Most cases (66%) were in the age group of 51–70 years, similar to previous reports of lung cancer patients.
The quality of the FFPE blocks was checked by H and E staining and reviewed by experienced clinical pathologists. Of the 604 cases tested, 119 were rejected due to inadequate amount of tissue in block, or lack of tumor cells in the block, or due to the block itself being damaged.
Biopsies comprised majority of the cases available for this study (77%), as is common for lung cancer patients. Among the 225 ACs where detailed histopathology report or diagnosis was available, 4 adenosquamous, 2 infiltrating, 3 invasive, 5 mucinous, 3 papillary, and 1 signet-ring ACs were mentioned. Well-differentiated AC was mentioned for 35 cases, moderately differentiated for 17, and poorly differentiated for 22 cases.
Several dilutions of the ROS antibody, 1:200 (recommended by CST) and lower ones, 1:150 and 1:100, were tested with the positive and negative control slides [Figure 1], with the 1:150 dilution being selected based on good sensitivity and specificity. Using this optimized automated staining protocol, 485 NSCLC cases were stained with ROS1 antibody. Diffuse staining was observed in the cytoplasm of many cancer cells [Figure 2], in agreement with cancer-specific expression of the ROS1 fusion protein. Only one case was observed to be strongly and reproducibly positively stained. The corresponding H and E images of the case and ROS staining are shown in [Figure 2]a and [Figure 2]b, respectively. [Figure 2]b, [Figure 2]c and [Figure 2]d are higher magnification images of the ROS1-positive section at 10, 20 and 40X magnification to show its cytoplasmic staining. Another section from the same block was restained in a subsequent run and confirmed to be positive. [Figure 3] shows a representative negatively stained case.
|Figure 1: Representative image for positive and negative controls used to optimize immunohistochemistry staining protocol for ROS1. Positive (left side) and negative (right side) controls captured at ×20. These were stained with 1:200 (top panel), and 1:150 (bottom panel), dilution of ROS antibody|
Click here to view
|Figure 2: SB-32158 adenocarcinoma, positive for ROS1 staining by immunohistochemical analysis. Hematoxylin and eosin image of the section at ×40 (a). Multiple images of the ROS1-stained tissue captured at ×10 (b), ×20 (c), and ×40 (d)|
Click here to view
|Figure 3: Image of an NSCLC adenocarcinoma case that was negative for ROS1 staining. Left: Hematoxylin and eosin image, Right: ROS1 immunohistochemistry at ×20|
Click here to view
The strongly positive NSCLC case was that of a 35-year-old female, a known case of metastatic AC with deposits that were progressively increasing in size on follow-up. The sample comprised a guided biopsy of a large mass, filling the entire right upper zone of the lung, with central necrosis and peripheral vascularity with a pleural-based lesion. The pathology diagnosis was that of a poorly differentiated carcinoma. The sections showed multiple fragmented linear cores predominantly consisting of necrosis with a few islands of tumor arranged in nests and sheets. The tumor cells were moderately pleomorphic with a high nuclear-to-cytoplasmic ratio, moderate eosinophilic cytoplasm, and eccentric nuclei that were hyperchromatic.
As reported in literature, a few NSCLC cases with faint staining that localized to macrophages but not to tumor cells were also observed. These cases were scored as negative cases.
| Discussion|| |
Among the 426 NSCLC cases stained here with ROS1, 225 cases were AC and 136 SCC cases. One case among ACs was found to be positive; thus, the percentage positivity rate was 0.44% in ACs and 0.23% among all NSCLC cases.
The smoking history of patients was not available in the patient records, so the association of lung cancer with smoking history could not be determined in this study. The frequency of ROS1 staining, in mixed populations of smokers and non-smokers in AC cases, has been reported at 0.9%–1.7% in the Western populations using large datasets such as BIDMC and TCGA.,, In our study, a lower frequency of 0.44% was observed with the caveat that it is calculated from only one case of ROS1 gene rearrangement, from 225 confirmed AC cases with adequate tumor tissue, and needs to be confirmed in larger cohorts.
A strong association between the presence of ROS1 rearrangements in non-smokers has also been reported, but the percentage varied from study to study (1-6% in non-smokers),,, compared to heavy and light smokers (0.3% and 1.6%, respectively). As mentioned above, in our study, the smoking history was not available in the patients' records, but it is likely that the cases evaluated here comprise a high proportion of smokers, given that 71% of cases were male and smoking has been reported to be common among Indian male lung cancer patients.,,
The lower frequency of ROS1 mutations could have been due to the age of the archived blocks used for the study, with the oldest block being 12 years old. The sole case which was strongly ROS1 positive was from the year 2015. However, that did not appear to be the case because many of the same NSCLC AC cases from the years 2014–2015 had been tested for ALK mutation positivity, also by IHC, in a research study conducted at the same time at one of the partner Apollo sites. In that study, ALK expression was reported in 12 out of 56 cases or 21.4% using an anti-ALK antibody from Ventana, indicating that the blocks were in good condition with sufficient tumor content for detecting proteins such as ALK by IHC. The percentage positivity could have been lower due to the use of IHC in this study compared to FISH used in other studies, but Shan et al.'s study  had reported that IHC detected more cases as compared to FISH.
The IHC method has been compared favorably with FISH  and next-generation sequencing (NGS) for ROS1 detection with a concordance of well above 90%, using sixty cases. For NGS, an anchored multiplex polymerase chain reaction (PCR) technique, using a panel of gene rearrangements that also contained the MSN-ROS1 gene, has also been tested on 319 FFPE samples. The performance was tested with a standard FISH assay to return 100% sensitivity and 100% specificity, though one of the samples was detected to be a CD74-ROS1 fusion.
Among NSCLC cases, ROS1 mutations have usually been reported in ACs, not in SCCs., Previous publications using FISH and nanostring have reported zero positivity in 59 or 214 SCCs of the lung, respectively. Rarely, cases of ROS1 mutation have been reported in non-ACs. Our results confirm these findings in that ROS mutations were not observed by IHC in the 136 confirmed SCC lung cancer cases tested here.
| Conclusion|| |
ROS1-positive tumors are very sensitive to treatment with crizotinib; hence, detecting this rare genetic alteration is an important step in the diagnostic workup of a lung AC patient. IHC offers an economical alternative for routine screening of ROS1 compared to FISH, which is not as accessible, or to NGS or PCR that require more tissue. The ROS1 IHC method, successfully optimized using commercially available positive and negative controls here, can help increase the testing for ROS1 and improve treatment for patients who are positive for it.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Matsuura S, Shinmura K, Kamo T, Igarashi H, Maruyama K, Tajima M, et al.
CD74-ROS1 fusion transcripts in resected non-small cell lung carcinoma. Oncol Rep 2013;30:1675-80.
Rimkunas VM, Crosby KE, Li D, Hu Y, Kelly ME, Gu TL, et al.
Analysis of receptor tyrosine kinase ROS1-positive tumors in non-small cell lung cancer: Identification of a FIG-ROS1 fusion. Clin Cancer Res 2012;18:4449-57.
Bubendorf L, Büttner R, Al-Dayel F, Dietel M, Elmberger G, Kerr K, et al.
Testing for ROS1 in non-small cell lung cancer: A review with recommendations. Virchows Arch 2016;469:489-503.
Bergethon K, Shaw AT, Ou SH, Katayama R, Lovly CM, McDonald NT, et al.
ROS1 rearrangements define a unique molecular class of lung cancers. J Clin Oncol 2012;30:863-70.
Jain J, Natarajan S, Chatterjee S. Using human medical waste to build biobanks. Indian J Med ImmunoOncol 2016;1:24-7.
Shan L, Lian F, Guo L, Qiu T, Ling Y, Ying J, et al.
Detection of ROS1 gene rearrangement in lung adenocarcinoma: Comparison of IHC, FISH and real-time RT-PCR. PLoS One 2015;10:e0120422.
Fu S, Liang Y, Lin YB, Wang F, Huang MY, Zhang ZC, et al.
The frequency and clinical implication of ROS1 and RET rearrangements in resected stage IIIA-N2 non-small cell lung cancer patients. PLoS One 2015;10:e0124354.
Jorge SE, Schulman S, Freed JA, VanderLaan PA, Rangachari D, Kobayashi SS, et al.
Responses to the multitargeted MET/ALK/ROS1 inhibitor crizotinib and co-occurring mutations in lung adenocarcinomas with MET amplification or MET exon 14 skipping mutation. Lung Cancer 2015;90:369-74.
Li C, Fang R, Sun Y, Han X, Li F, Gao B, et al.
Spectrum of oncogenic driver mutations in lung adenocarcinomas from East Asian never smokers. PLoS One 2011;6:e28204.
Kim HR, Lim SM, Kim HJ, Hwang SK, Park JK, Shin E, et al.
The frequency and impact of ROS1 rearrangement on clinical outcomes in never smokers with lung adenocarcinoma. Ann Oncol 2013;24:2364-70.
Sharma PK, Bansal R. Profile of lung cancer in predominantly bidi smoking rural population of Northern Himachal Pradesh. Indian J Chest Dis Allied Sci 2013;55:75-8.
Noronha V, Pinninti R, Patil VM, Joshi A, Prabhash K. Lung Cancer In The Indian Subcontinent. South Asian J Cancer 2016;5:95-103.
] [Full text]
Kaur H, Sehgal IS, Bal A, Gupta N, Behera D, Das A, et al.
Evolving epidemiology of lung cancer in India: Reducing non-small cell lung cancer-not otherwise specified and quantifying tobacco smoke exposure are the key. Indian J Cancer 2017;54:285-90.
] [Full text]
Gupta P, Swarnalatha G, Swain M. Comparison of histopathological findings and epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) expression in lung adenocarcinoma. Indian J Pathol Microbiol 2019;62:24-30.
] [Full text]
Zheng Z, Liebers M, Zhelyazkova B, Cao Y, Panditi D, Lynch KD, et al.
Anchored multiplex PCR for targeted next-generation sequencing. Nat Med 2014;20:1479-84.
Zhao W, Choi YL, Song JY, Zhu Y, Xu Q, Zhang F, et al.
ALK, ROS1 and RET rearrangements in lung squamous cell carcinoma are very rare. Lung Cancer 2016;94:22-7.
Song Z, Yu X, Zhang Y. Clinicopathological characteristics and survival of ALK, ROS1 and RET rearrangements in non-adenocarcinoma non-small cell lung cancer patients. Cancer Biol Ther 2017;18:883-7.
[Figure 1], [Figure 2], [Figure 3]