Morphological sub-classification of non-small cell carcinoma in small biopsy specimens presents difficulties for pathologists and recent advances in chemotherapy have resulted in increased pressure to more robustly differentiate between squamous carcinoma and adenocarcinoma. The results of audits examining classification of non-small cell lung carcinoma by pathologists working in a specialist team within a regional centre and the effect of introducing adjunct immunohistochemistry into the reporting pathway are presented. It is concluded that the use of a limited immunohistochemical panel substantially reduces the number of cases when a specific cell type cannot be identified or ‘favoured’ (34% to 6%) and that the classification obtained correlates well with that found in subsequent resection specimens. In addition the introduction of immunohistochemistry substantially reduces the variability in reporting practice between pathologists.
- lung cancer
Statistics from Altmetric.com
Lung carcinomas are classified according to the WHO classification,1 which describes a complex variety of subsets that can be identified in resection specimens when multiple blocks of tissue are available for histological assessment. The majority of lung carcinomas are, however, not resected and are diagnosed and classified using small biopsies or cytology specimens.2 It has been recognised for many years that this causes significant problems for pathologists in robustly identifying sub-types of non-small cell lung carcinoma (NSCLC), as the features which allow identification of squamous carcinoma or adenocarcinoma may be absent in these small samples.3 4 This has lead to the use of the term NSCLC (NOS) (not otherwise specified), reflecting this difficulty.5
Recent advances in chemotherapy have however placed a greater requirement on pathologists to accurately subclassify NSCLC and in particular differentiate squamous carcinomas from non-squamous carcinomas.6 Recent studies have shown that in bronchial biopsies obtained from NSCLCs showing no obvious evidence of either squamous or glandular differentiation, a prediction of cell type can be made with a high degree of accuracy using immunohistochemical (IHC) staining.7 This approach has been widely adopted and recently has been recommended in an international IASLC/ATS/ERS publication on adenocarcinomas.8 The effect on reporting patterns of adopting such a strategy in routine practice has, however, not been documented. We describe the outcome of audits carried out in 2007–08, looking at reporting patterns for NSCLC in a busy regional respiratory pathology service. Following the adoption of a protocol requiring IHC staining of bronchial and percutaneous lung biopsies diagnosed as NSCLC when no definitive cell typing was possible on morphological criteria, the audit cycle was completed by examining reporting patterns for 2010.
Material and methods
The Pathology Department in the Royal Infirmary of Edinburgh provides diagnostic histopathology services to respiratory physicians across four acute hospitals in two health boards (NHS Lothian and the NHS Borders). The respiratory samples are all reported by a team of three consultant pathologists with a specific interest in the area.
An audit was carried out examining the reporting patterns of NSCLC in bronchial and percutaneous lung biopsies over a 2-year period (2007 and 2008). Cases of NSCLC were identified from the departmental database using searches to identify specific SNOMED codes. All cases had been stained routinely with H&E and AB/PAS. In cases where neuroendocrine differentiation was suspected, appropriate immunohistochemistry had been undertaken. The reports of the cases were read and the conclusion of the report recoded as adenocarcinoma, squamous carcinoma, NSCLC with neuroendocrine features, carcinoid tumour, other or NSCLC (NOS).
Following this audit the clinical requirement for more specific subclassification of NSCLC was discussed. A protocol was instigated such that in cases where the morphological features were insufficient to allow specific classification, additional IHC staining with p63 (Dako UK clone 4A4) and TTF1 (Dako UK clone 8G7G3/1) would be undertaken. These biopsies where then reported as ‘NSCLC favouring squamous differentiation’ if p63 was positive or ‘NSCLC favouring adenocarcinoma’ if TTF1 was positive. In cases where neuroendocrine morphology was suspected, appropriate confirmatory stains would be performed. Cases where IHC was performed to confirm a primary lung adenocarcinoma in contrast to a metastasis, rather to establish the cell type, were not included.
The audit cycle was completed by reviewing all bronchial and percutaneous lung biopsies reported in 2010. The final reported cell type and whether this was obtained by morphology alone or with the aid of IHC was noted. If the patient had then gone on to resection, the final cell type based on this was recorded.
The results are summarised in table 1. The 2007–08 audit revealed a total of 420 bronchial and percutaneous lung biopsies which were reported as NSCLC. Of these, 125 (30%) were classified as adenocarcinoma, 131 (31%) as squamous carcinoma and 141 (34%) as NSCLC (NOS). In addition it was noted that there appeared to be a high degree of variability between consultants in the proportion of cases where no cell type was specified (table 2).
In 2010 a total of 189 cases were identified. Immunohistochemistry was performed on a total of 63 cases (34%) in order to clarify cell type. Classification as adenocarcinoma or ‘adenocarcinoma favoured’ was obtained in 64 (34%) of the cases and squamous carcinoma or ‘squamous carcinoma favoured’ in 98 (52%). Only 12 (6%) cases were reported as NSCLC (NOS). In all these cases IHC had been performed. In nine, the tumour cells were negative for both p63 and TTF1, while in two there was variable staining for both antibodies. In the final case no residual tumour was present in the sections cut for IHC. This repeat audit also showed that implementation of the new approach to cell typing resulted in a more uniform reporting pattern by the different pathologists in the team (table 2).
Forty-one of the 189 cases identified in the 2010 audit went on to have surgical resections. The results are shown in table 3. The initial classification given on the small biopsy was regarded as ‘correct’ if it matched with the classification given in the resection report or ‘not incorrect’ if a mixed pattern of differentiation was seen, one of which was identified in the diagnostic biopsy. The classification was regarded as ‘incorrect’ when the final cell type was different from that suggested on the initial biopsy.
Of the 12 resections with a preoperative diagnosis of adenocarcinoma or ‘adenocarcinoma-favoured’, only one case was regarded having been incorrectly classified. For the cases diagnosed as squamous carcinoma or ‘squamous carcinoma favoured’, only 2 of the 21 resection cases were incorrectly classified, giving an overall ‘correct’ or ‘not incorrect’ classification rate for this group of biopsies of 30/33 (91%).
Of the 12 cases reported as NSCLC (NOS), two were subsequently resected, with one being classified as a pleomorphic carcinoma (large cell undifferentiated carcinoma with spindle cell carcinoma) and the other as a large cell neuroendocrine carcinoma.
The 2007–08 rate of around 33% of NSCLC being classified as ‘not otherwise specified’ is in line with previously published experience and reflects the difficulty in classification using morphology alone.5 Our experience demonstrates that the addition of IHC coupled with an increased awareness of the need for more information on subclassification can result in a dramatic decrease in the use of this term and greater consistency in reporting patterns between pathologists. Our data also supports previous evidence validating the use of IHC in this context with reference to the final resection histology.7
The data presented relates solely to histological specimens, and thus the rate of NSCLC (NOS) obtained in all patients with lung cancer will be higher than this, given that many are diagnosed by cytology. We and others have, however, recently shown that similar IHC approaches can be employed with final needle aspirates processed as cell blocks, allowing good concordance with histological classification,9 10 suggesting that such use of IHC will be able to provide robust data on cell type for the majority of patients with NSCLC.
In summary, our data indicates that in routine practice the use of immunohistochemistry can dramatically reduce the NSCLC (NOS) rate in small histological samples submitted for the diagnosis of lung carcinoma. This simple change in protocol has improved the quality of information given to oncologists and thus helped direct and refine their treatment choices for patients. There will, however, always remain a small subset of tumours where classification beyond NSCLC (NOS) remains impossible. Further refining of classification may be possible using wider panels of antibodies,11 12 but given the relatively small amount of tissue which is often available and the requirement to preserve this for potential molecular studies, we believe a pragmatic approach is required. Investigations should be selected to answer specific issues related to individual patient management through discussion at multidisciplinary meetings.
Subclassification of NSCLC in diagnostic samples is important in patient management.
In cases where morphological features in small biopsies fail to allow classification as a squamous carcinoma or adenocarcinoma immunohistochemistry provides good predictive evidence of cell likely cell type.
The use of adjunct immunohistochemistry signficantly reduces the proportion of cases of NSCLC where a sub-type cannot at least be favoured and reduces variability in the reporting parctices between pathologists.
Interactive multiple choice questions
This JCP article has an accompanying set of multiple choice questions (MCQs). To access the questions, click on BMJ Learning: take this module on BMJ Learning from the content box at the top right and bottom left of the online article. For more information please go to: http://jcp.bmj.com/education. Please note: the MCQs are hosted on BMJ Learning the best available learning website for medical professionals from the BMJ Group. If prompted, subscribers must sign into JCP with their journal's username and password. All users must also complete a onetime registration on BMJ Learning and subsequently log in (with a BMJ Learning username and password) on every visit.
Competing interests WAW has received sponsorship to attend meetings in the UK and Ireland from Lilly UK.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.