Analysis of a pathology review of patients with lung tumors

Reproducibility of histopathological diagnosis in poorly differentiated NSCLC: an international multiobserver study

INTRODUCTION

The 2004 World Health Organization classification of lung cancer contained three major forms of non-small-cell lung cancer: squamous cell carcinoma (SqCC), adenocarcinoma (AdC), and large cell carcinoma. The goal of this study was first, to assess the reproducibility of a set of histopathological features for SqCC in relation to other poorly differentiated non-small-cell lung cancers and second, to assess the value of immunohistochemistry in improving the diagnosis.

METHODS

Resection specimens (n = 37) with SqCC, large cell carcinoma, basaloid carcinoma, sarcomatoid carcinoma, lymphoepithelial-like carcinoma, and solid AdC, were contributed by the participating pathologists. Hematoxylin and eosin (H&E) stained slides were digitized. The diagnoses were evaluated in two ways. First, the histological criteria were evaluated and the (differential) diagnosis on H&E alone was scored. Second, the added value of additional stains to make an integrated diagnosis was examined.

RESULTS

The histologic criteria defining SqCC were consistently used, but in poorly differentiated cases they were infrequently present, rendering the diagnosis more difficult. Kappa scores on H&E alone were for SqCC 0.46, large cell carcinoma 0.25, basaloid carcinoma 0.27, sarcomatoid carcinoma 0.52, lymphoepithelial-like carcinoma 0.56, and solid AdC 0.21. The κ score improved with the use of additional stains for SqCC (combined with basaloid carcinoma) to 0.57, for solid AdC to 0.63.

CONCLUSION

The histologic criteria that may be used in the differential diagnosis of poorly differentiated lung cancer were more precisely refined. Furthermore, additional stains improved the reproducibility of histological diagnosis of SqCC and AdC, uncovering information that was not present in routine H&E stained slides.

A systematic review of the interobserver variability for histology in the differentiation between squamous and nonsquamous non-small cell lung cancer

INTRODUCTION

The importance of identifying non-small cell lung cancer (NSCLC) histologic subtype has increased recently because of the development of target-specific chemotherapeutic agents. This systematic review was undertaken to examine the interobserver variability for histology in differentiating between subtypes of NSCLC, specifically the ability to differentiate squamous from nonsquamous histology.

METHODS

A systematic literature search was undertaken to identify studies that evaluated the reproducibility of histologic diagnosis by pathologists in their reporting of NSCLC subtypes. Studies were screened using a priori defined eligibility criteria. The National Health and Medical Research Council diagnostic levels of evidence were applied and quality assessed using the Quality Assessment of Diagnostic Accuracy Studies tool. Data were extracted and reanalyzed to permit comparison of agreement in nonsquamous and squamous cell carcinoma by 2 × 2 tables. Percentage agreement and kappa statistics were calculated for each included study.

RESULTS

Out of 1480 articles identified through the literature search, six were eligible for inclusion. The percentage agreement for all subtypes of NSCLC in the included studies ranged from 67.1 to 89.6% (κ, 0.42-0.84). Based on the primary reanalysis of data (reanalysis 1), agreement between pathologists in differentiating nonsquamous and squamous histology ranged from 77.0 to 94.2% (κ = 0.48-0.88) indicating a moderate to high level of agreement.

CONCLUSION

The reasonably high agreement and kappa statistics for the included studies suggest that pathologists can reproducibly differentiate between nonsquamous and squamous NSCLC. This is clinically important in guiding oncologist decision making in choosing the most appropriate therapy for their patients.

Tobacco smoking as a risk factor of bronchioloalveolar carcinoma of the lung: pooled analysis of seven case-control studies in the International Lung Cancer Consortium (ILCCO)

BACKGROUND

The International Lung Cancer Consortium (ILCCO) was established in 2004, based on the collaboration of research groups leading large molecular epidemiology studies of lung cancer that are ongoing or have been recently completed. This framework offered the opportunity to investigate the role of tobacco smoking in the development of bronchioloalveolar carcinoma (BAC), a rare form of lung cancer.

METHODS

Our pooled data comprised seven case-control studies from the United States, with detailed information on tobacco smoking and histology, which contributed 799 cases of BAC and 15,859 controls. We estimated the odds ratio of BAC for tobacco smoking, using never smokers as a referent category, after adjustment for age, sex, race, and study center.

RESULTS

The odds ratio of BAC for ever smoking was 2.47 (95% confidence interval [CI] 2.08, 2.93); the risk increased linearly with duration, amount, and cumulative cigarette smoking and persisted long after smoking cessation. The proportion of BAC cases attributable to smoking was 0.47 (95% CI 0.39, 0.54).

CONCLUSIONS

This analysis provides a precise estimate of the risk of BAC for tobacco smoking.

Diagnostic agreement in the histopathological evaluation of lung cancer tissue in a population-based case-control study

Only few studies have compared the agreement of histological lung carcinoma diagnosis of a population-based case series and an independent pathology review. We analyzed data of our population-based lung cancer case-control study to determine the agreement in the histopathological evaluation of lung cancer. Six-hundred and sixty-eight out of 1004 interviewed male and female lung cancer cases were histologically evaluated according to the 1981 WHO classification by regional pathologists and a central pathologist who was blinded to the evaluations of the regional pathologists. The observed agreement was 0.65 with kappa = 0.54 (95% CI: 0.49-0.58). It was highest for small-cell carcinoma (0.94; kappa = 0.82) and lower for squamous-cell carcinoma (0.81; kappa = 0.58) and adenocarcinoma (0.81; kappa = 0.55). Agreement was slightly higher among women than men. The observed agreement among non-smoking cases was 58% as compared to 67% heavy smoking cases. The moderate agreement for squamous-cell and adenocarcinoma complicates epidemiological studies that address these histological subtypes.

The health-related quality of life and survival of small-cell lung cancer patients: results of a companion study to CALGB 9033

The purposes of this study were 2-fold: to evaluate the impact of the schedule dependency of etoposide (3-day IV short course vs. a 21-day oral prolonged course) with cisplatin on the quality of life of small-cell lung cancer (SCLC) patients; and to examine the effect of baseline quality of life variables on long-term survival, after adjustment for known demographic and clinical prognostic factors. Participants were 70 patients enrolled in the cancer and leukemia group B (CALGB) protocol 9033. Quality of life was assessed at baseline, 6 and 12 weeks by: the EORTC QLQ-30, the Centers for epidemiology studies--Depression short form, the medical outcomes study (MOS) social support questionnaire, and a scale of sleep quality. Contrary to expectations, study results suggested no significant differences in the patients' life quality and treatment response based on whether they received etoposide in a 3-day IV vs. a 21-day oral regimen. The use of the baseline variables in predicting overall survival indicated that patients who were non-white and with liver involvement had decreased survival. Brain involvement, being male, and higher depressive symptoms were also found to be borderline significant in predicting decreased survival in this patient population.

Lung cancer in nonsmoking women. Histology and survival patterns

BACKGROUND

Despite the widespread view that important clinical and etiologic differences exist between histologic categories of lung cancer, few studies have examined the accuracy of hospital-reported pathologic diagnoses of lung cancer.

METHODS

A review of pathologic material and an assessment of survival patterns were conducted in conjunction with a recently completed case-control study of lung cancer among nonsmoking women in Missouri. Using established protocols, tissue slides from tumors of 482 patients were reviewed by 3 pathologists.

RESULTS

Adenocarcinoma was the most common histologic type among former smokers and lifetime nonsmokers. The overall agreement rate between the original and review diagnoses was 65.6%. The positive predictive value ranged from 0.33 for bronchioalveolar carcinomas to 0.84 for adenocarcinomas. Agreement rates for small, medium, and large hospitals were 63.1, 66.6, and 66.2%, respectively. Survival rates were highest for bronchioalveolar carcinoma and lowest for small cell carcinoma.

CONCLUSION

Given the importance of lung cancer to public health and the need to examine risk by histologic type, these data indicate that pathologic review of registry-reported lung cancer cases may be an important component of large scale studies of etiology.

Interobserver variability in histopathologic subtyping and grading of pulmonary adenocarcinoma

BACKGROUND

The classification of lung tumors by the World Health Organization (WHO 1981) describes subtyping of adenocarcinoma of the lung (ACL) into acinar adenocarcinoma and papillary adenocarcinoma, bronchioloalveolar carcinoma, and solid carcinoma with mucus formation. Acinar and papillary adenocarcinoma may be graded as well-, moderately, or poorly differentiated. This study evaluated the interobserver variability in the subtyping and grading of ACL according to the WHO classification.

METHODS

Histologic specimens from 211 patients with disease of Stages IIIa-IV were classified in a blind manner by three panelists. All available paraffin-embedded tissue blocks, including metastatic tumors, were sampled, and new slides were made for the study.

RESULTS

Twenty-two ACL tumors could not be assigned a subtype by Panelist 1, which left 189 tumors as the basis for additional evaluation. Overall agreement for the three panelists regarding subtypes was 41%. Nonchance agreement was evaluated by kappa statistics, which may vary between -1 in the event of agreement that is less than that expected by chance and /1 in the event of full agreement. The kappa value regarding subtypes was 0.18 (95% confidence limits, 0.14-0.23). Overall observed agreement regarding degree of differentiation was 43%, with a kappa value of 0.12 (95% confidence limits, 0.06-0.17). Histologic material was obtained by thoracotomy in a subgroup of 53 patients, and in this patient group unanimity among two or more panelists was 88% for subtyping and 100% for degree of differentiation.

CONCLUSIONS

The degree of agreement in subtying and grading of ACL in Stages IIIa-IV is low, suggesting that more objective criteria is needed before a prognostic impact of such variables can be assessed. The quality and quantity of material available for subtyping obviously influence the results, which is reflected in a better agreement when histologic material is obtained by thoracotomy.

The accuracy of the histological classification of lung carcinoma and its reproducibility: a study of 75 archival cases of adenosquamous carcinoma

The classification of lung carcinoma into a small cell lung carcinoma (SCLC) and non-small cell lung carcinoma (NSCLC), is highly reproducible. There are few studies on the reproducibility of subtyping NSCLC, with anecdotal reports on the subtype of adenosquamous carcinoma. We undertook a study of 75 cases archived as adenosquamous carcinoma. All cases were accepted as NSCLC on independent review by three pathologists utilizing the 1982 World Health Organization (WHO) classification criteria. The acceptance rate of adenosquamous carcinoma by the three pathologists was 65%, 28%, and 65%. Cases not accepted as adenosquamous fell into the different subtypes of NSCLC, with a concordance rate between each pair of pathologists of 49%, 61% and 43%, indicating poor agreement between pathologists. The study confirms the high reproducibility of the classification into SCLC and NSCLC; it shows poor concordance for all subtypes of NSCLC with poor reproducibility of diagnosis of adenosquamous carcinoma as specified by the 1982 WHO classification. Physicians dealing with lung carcinoma should be aware of this limitation of classification.

Further evaluation of quantitative nuclear image features for classification of lung carcinomas

The usefulness of quantitative nuclear image features (QNI) for the histological classification of lung carcinomas was investigated. As no clear distinction could be established between the distributions of these features for the nuclei of squamous cell, adenocarcinoma, and large cell carcinoma, the attention was restricted to the discrimination between small cell lung carcinoma (SCLC) and non-small cell carcinoma (NSCLC). This discrimination is the crucial one in discussions about the choice of treatment. The differences between SCLC and NSCLC are statistically highly significant for various QNI features. The use of more than one QNI feature hardly raised the discriminatory performance with respect to the distinction between SCLC and NSCLC. Inferences were made about the probability and confidence interval of SCLC for a given QNI feature. It is concluded that in cases of uncertainty or disagreement, nuclear characteristics are useful for the discrimination between SCLC and NSCLC.

Relation of bronchioloalveolar carcinoma to tobacco

OBJECTIVE

To determine whether bronchioalveolar carcinoma is related to tobacco use.

DESIGN

Case-control study.

SETTING

11 teaching hospitals of Chicago, Long Island, New York, and Philadelphia, 1977-89.

SUBJECTS

87 patients with histologically diagnosed bronchioloalveolar carcinoma (cases) and 286 non-cancer and 297 cancer patients matched to cases on age, sex, race, hospital, and date of admission.

RESULTS

10% of male cases and 25% of female cases had never smoked. Relative risks of bronchioloalveolar carcinoma (as estimated by the relative odds) were greater for subjects who started smoking at a younger age, smoked for a longer time, or smoked more cigarettes per day. Relative risks decreased proportionally to the duration of smoking cessation.

CONCLUSION

Smoking plays an important part in the aetiology of bronchioloalveolar carcinoma but is not the only potential cause because of the large proportion of never smokers among patients with this disease.

Cigarette smoking and lung cancer cell types

The role of tobacco smoke in the development of lung cancer is well known for squamous and small cell types, somewhat less so for adenocarcinoma, and not specifically assessed for large cell carcinoma. In the current analysis, based on 851 men and 507 women with lung cancer and their matched controls (888 men and 608 women), smoking was associated with each lung cancer cell type, and differences in smoking habits by cell types were small. However, the increase of lung cancer risk according to number of cigarettes per day was stronger for small cell and oat cell carcinoma than for adenocarcinoma. There was no increase for large cell carcinoma. For squamous cell carcinoma, this dose response was weak among men and strong among women. The strength of the association between smoking and lung cancer cell types may be related to cancer location, with more peripheral lung cancer types (such as adenocarcinoma and large cell carcinoma) showing weaker associations than more central tumors (such as squamous or small cell and oat cell carcinoma).

Lung cancer histology in Jews and Arabs in Israel, 1962-1982

Lung cancer rates in Israel are lower than in Western countries, not explainable by smoking habits. Because of the different relations of squamous cell carcinoma and adenocarcinoma to smoking it was of interest to study the histologic distribution in Israel. A total of 7,871 histologically confirmed lung cancer cases were studied in the period 1962-1982. Squamous cell carcinoma was the leading tumor type in Jewish men and adenocarcinoma in Jewish women. Rates of both adenocarcinoma and squamous cell carcinoma increased throughout the period in both Jewish men and women, but the increase in adenocarcinoma was more pronounced in the last study period than that in squamous cell carcinoma. In 1977-1982 the rate ratio of squamous cell carcinoma to adenocarcinoma among Jewish men was 1.7. In Arab men it was 2.9, and in Jewish women 0.57. The Kreyberg I/II ratio among Jewish men was about 2.7 with no clear trend throughout the study period, and among Arab men this gradually decreased from 8.1 to 3.5. Jewish women had a constant Kreyberg I/II ratio of about 1 through the whole study period, but the ratio in Arab women was significantly higher than 1, with a mean overall ratio of 3.2. Jews and Arabs in Israel are different from each other in their patterns of lung cancer histology and are different to some extent from other populations in the Western world.

The reliability of routine pathologic diagnosis of colorectal adenocarcinoma

The diagnosis of colorectal adenocarcinoma is generally considered as being reliable. However, the reproducibility of the classification of specific histologic patterns and the interrater agreement on the gradings have not been firmly established. A panel of three independent expert pathologists reviewed histologic sections from 128 patients selected among 1848 with colorectal cancer, diagnosed in 11 hospitals of the same region. The panel agreed with 92.6% of the original diagnoses of colorectal adenocarcinoma. As for agreement between panel members, the kappa value was 0.78 for the diagnosis of adenocarcinoma and 0.62 for confirmation of colorectal origin. The intraclass correlation coefficient for tumor differentiation features was 0.75. The proportion of villous and adenomatous components also generated good agreement. However, the grading of mucin secretion showed poor agreement (intraclass correlation coefficient = 0.44). Results confirm the reliability of routine pathological diagnosis and also demonstrate the reproducibility of basic diagnostic categories and pathognomonic features. Thus, to obtain reliable information from medical records for epidemiologic and clinical studies, data should be limited to well-defined diagnostic and histopathologic categories.

Reproducibility of tumour typing of lung carcinomas performed according to WHO's recommendation

Tumour sections from 139 patients with bronchial carcinoma were classified blindly by a panel of three pathologists according to the WHO recommendation of 1981. Only sections from resected specimens were included. There was agreement among all reviewers with regard to the main WHO tumour type of 71% of the specimens. For the remainder, two reviewers agreed upon 19% whereas the tumour classification differed substantially in 10% of all cases. About half of the original diagnoses were maintained when the latter group was re-evaluated blindly. Best agreement in the whole material was noted for squamous cell carcinoma. The main difficulty was differentiation between adenocarcinoma and squamous cell carcinoma with a high malignancy grading score. Despite the improvements of the WHO classification system, our group of diagnostic problem cases was not reduced when compared with earlier classification studies.

Lung carcinoma in the elderly population. Influence of histology on the inverse relationship of stage to age

Unlike most cancers, lung carcinoma is more likely to be localized at the time of diagnosis in older age groups when compared to those who are middle-aged. In an attempt to explain this inverse age-stage relationship we studied 9062 histologically confirmed cases of lung carcinoma occurring from 1975 to 1984 obtained from the regional cancer registry for Kansas and western Missouri. They were analyzed according to histologic type, age, sex, and stage at diagnosis. The data suggest that the proportion of squamous cell carcinoma rises and that of adenocarcinoma and small cell undifferentiated carcinoma falls with increasing age. The proportion of staged lung carcinoma with local disease at the time of diagnosis increases with age. In males this trend occurred in all cell types except large cell undifferentiated carcinoma but was most significant for squamous cell carcinoma. Squamous cell carcinoma was the only group to show a significant trend among females. The rise in squamous cell and fall in small cell carcinoma may partially explain the increased prevalence of local stage disease with advancing age.

Lung cancer classification: the relationship of disease extent and cell type to survival in a clinical trials population

The staging and histologic cell type of patients in the Lung Cancer Study Group (LCSG) clinical trials program are reviewed and confirmed or resolved at the reference center for anatomic and pathologic classification of lung cancer. A high level of consistency in classification has been achieved through the use of criteria that minimize intraobserver variability. The data obtained from the review project have been used to characterize the relationship of disease extent and cell type to survival in the clinical trials population. Survival characteristics were generated for 1,121 patients who underwent apparent complete resection of nonsmall cell lung cancer and were subsequently entered into various protocols to receive either adjuvant treatment or no further therapy. The end results study provides some insight regarding the biological behavior of squamous cell carcinoma and adenocarcinoma of the lung in terms of the anatomic extent of disease at the time of apparent complete resection. Patients with squamous cell carcinoma had an outcome superior to that of patients with adenocarcinoma in every TNM subset. The differences in survival according to these major cell types were significant overall and in the T1 N0, T1 N1, and T2 N1 subsets but not in the TNM subsets in stage III disease. Histologic cell type and extent of disease are important factors in survival expectations; thus the accuracy and reproducibility of these classifications plays a significant role in the evaluation of differing modalities of treatment.

The 'grey area' between small cell and non-small cell lung carcinomas. Light and electron microscopy versus clinical data in 14 cases

We studied 14 lung tumours which on light microscopy had posed difficulties on classification as either small cell or non-small cell carcinomas. The light and electron microscopical features were compared with patient follow-up data. Electron microscopy showed neuroendocrine granules in 12 cases, and adeno- and squamous cell differentiation but no neuroendocrine granules in the remaining two cases. The latter two cases showed prolonged patient survival (both patients alive after 2 1/2 and 2 years, respectively). Ten of the cases with neuroendocrine granules showed a rapid course of disease (death between 2 1/2 weeks and 15 months after diagnosis) and marked initial response to multiagent chemotherapy. Thus, the clinical impression of these cases was that of small cell carcinoma. The remaining two cases with neuroendocrine granules showed a more protracted course, with death after 1 1/2 and 2 1/2 years. These two tumours did not show the light microscopical features of atypical carcinoid. The results illustrate the value of electron microscopy in predicting clinical behaviour of carcinomas difficult to place into small cell or non-small cell carcinoma groups. They also point to the existence of neuroendocrine carcinomas other than carcinoids with a more protracted course than small cell carcinomas.

Prevalence of neuroendocrine granules in small cell lung carcinoma. Usefulness of electron microscopy in lung cancer classification

Fifty-four lung carcinomas submitted for routine electron microscopy under the light microscopical diagnosis of small cell carcinoma were investigated. In 42 or of 45 evaluable cases, neuroendocrine granules were considered definitely or probably present, the modification 'probably' being necessary in suboptimal material. In three cases, squamous cell differentiation was seen, but no neuroendocrine granules were found. On revision of these three cases, the light microscopical diagnosis was changed to squamous cell carcinoma in one instance, and neither of the other two cases was considered classical for small cell carcinoma. Patient follow-up of these three cases showed tumour behaviour indicative of non-small cell carcinoma in two evaluable cases, the third case yielding no significant data. These results indicate that neuroendocrine granules can generally be found in small cell lung carcinomas, provided the material is sufficient for evaluation. When these granules are absent, and other differentiation is found on electron microscopy, the final classification of the tumour should incorporate this finding, to warn the clinician that the tumour will not necessarily behave as a small cell carcinoma.

Interobserver reproducibility of histopathological features in stage II breast cancer. An ECOG study

Eleven surgical pathologists studied microscopic sections from 45 mastectomy specimens of node positive breast cancer patients who had been entered into ECOG clinical trials. Inter-observer reproducibility for histoprognostic features was examined as a prerequisite before a subsequent evaluation of their possible clinical applicability could be undertaken. Histological type, nuclear grade, tubular formation, and lymphoid reactions were studied in the cancerous tissues. Lymph nodal responses (follicular and pulp prominence, sinus histiocytosis) were also examined in a manner that stimulated slide review in routine surgical pathology practice. Numerous two-way comparisons of the pathologists' findings resulted in low levels of agreement (usually much less than 90%). The degree of inter-observer reliability is clinically unacceptable using customary slide review analysis. New ways of examining breast cancer tissues need to be explored in the search for prognostic features which can be applied to the clinical management of breast cancer patients.

Measuring repeatability and validity of histological diagnosis--a brief review with some practical examples

Evaluation of histological diagnosis requires an index of agreement (to measure repeatability and validity) together with a method of assessing bias. Cohen's kappa statistic appears to be the most suitable tool for measuring levels of agreement, which if unsatisfactory may be caused by bias. Further study of bias is possible by examining levels of agreement for each diagnostic category or by searching for categories of disagreement in which more observations occur than would be expected by chance alone. This article gives reasons for choosing the kappa statistic, with examples illustrating its calculation and the investigation of bias.

Carcinoma of the lung: evaluation of histological grade and factors influencing prognosis

The results in 96 patients with lung cancer who underwent lobectomy or pneumonectomy were analyzed. In reviewing the case histories of these patients, it became apparent that those with poorly differentiated tumor (grade 3) have an increased likelihood of positive lymph node metastases compared with those with well-differentiated (grade 1) or moderately differentiated (grade 2) tumor. Poor differentiation of the tumor, vascular invasion, and lymph node metastases appear to represent poor prognostic indices in patients undergoing operation. Compared with patients with grade 1 and grade 2 tumors, patients with a grade 3 adenocarcinoma had more local recurrences, while those with grade 3 squamous cell carcinoma had more distant metastases. The findings suggest that histological grading is an important adjunct to the clinical evaluation of and planning of treatment for patients with lung cancer.