Fax: (410) 614-9556
The utility of napsin-A in the identification of primary and metastatic lung adenocarcinoma among cytologically poorly differentiated carcinomas
Version of Record online: 9 SEP 2010
Copyright © 2010 American Cancer Society
Volume 118, Issue 6, pages 441–449, 25 December 2010
How to Cite
Stoll, L. M., Johnson, M. W., Gabrielson, E., Askin, F., Clark, D. P. and Li, Q. K. (2010), The utility of napsin-A in the identification of primary and metastatic lung adenocarcinoma among cytologically poorly differentiated carcinomas. Cancer Cytopathology, 118: 441–449. doi: 10.1002/cncy.20108
- Issue online: 12 DEC 2010
- Version of Record online: 9 SEP 2010
- Manuscript Accepted: 20 JUL 2010
- Manuscript Revised: 2 JUL 2010
- Manuscript Received: 4 MAY 2010
- thyroid transcription factor-1;
- poorly differentiated adenocarcinoma;
- lung carcinoma;
New developments in the treatment of lung cancer have necessitated the further histologic and cytologic subtyping of nonsmall cell lung carcinomas. Thyroid transcription factor-1 (TTF-1) long has served as the predominant marker for demonstrating lung origin. However, it is also expressed in a variety of other tumors, particularly neuroendocrine neoplasms and, to a much lesser degree, squamous cell carcinoma of the lung. Napsin-A, which is expressed in lung tissue, is a relatively new marker for lung adenocarcinoma. In this study, the authors examined the utility of napsin-A compared with TTF-1 in cytologic specimens of primary and metastatic, poorly differentiated lung adenocarcinomas.
The archives of the Department of Pathology at The Johns Hopkins Hospital were searched for cytologic cases of poorly differentiated lung adenocarcinoma that were histologically confirmed. In total, 75 patients (cases) along with 95 controls were included, each of whom had adequate cell block material for TTF-1 and napsin-A staining. Tissue microarrays of lung adenocarcinoma also were examined.
TTF-1 and napsin-A were detected in 61 of 75 cases (81.3%) and in 49 of 75 cases (65.3%), respectively. The sensitivity and specificity of TTF-1 were 81% each; and napsin-A had a greater specificity of 96%, and sensitivity of 65%. Napsin-A was not detected in small cell carcinomas or in other carcinomas of nonlung origin except for renal cell carcinoma.
Although TTF-1 had a higher sensitivity, napsin-A was useful as a surrogate marker when encountering a poorly differentiated lung adenocarcinoma or an unknown primary tumor, particularly in cytologic specimens and difficult cases. The current results indicate that the dual use of both markers may be necessary to improve diagnostic accuracy. Cancer (Cancer Cytopathol) 2010. © 2010 American Cancer Society.
Lung cancer is the leading cause of cancer-related deaths worldwide, and approximately 70% to 80% of cases fall under the classification of nonsmall cell carcinoma.1-4 Within this category, adenocarcinoma is the predominant subtype.5, 6 Approximately 40% to 60% of patients present with locally advanced or metastatic disease at the time of diagnosis. Current conventional chemoradiation therapies have begun to reach their plateaus of effectiveness; thus, increasing efforts have been made to develop and use novel targeted or personalized therapies that potentially may lead to improved survival and prognosis.7-10 Targeted therapies dedicated to specific nonsmall cell carcinomas (ie, squamous cell carcinoma vs adenocarcinoma vs large cell carcinoma) have generated a demand for the further subtyping of nonsmall cell carcinomas in pathology.
The cytologic sampling of lesions plays a critical role in the diagnosing and staging of lung cancer, particularly for patients with advanced disease (stage III and IV), because the majority of these individuals are not candidates for invasive or surgical procedures. Although lung adenocarcinoma can be differentiated and diagnosed in most cases, a significant number of cases still are difficult to classify for a variety of reasons, such as sampling error, poor specimen preparation, and tumor differentiation. Poorly differentiated carcinomas are particularly difficult to classify, because they may lack specific architectural and cytomorphologic characteristics of either adenocarcinoma or squamous cell carcinoma. This is particularly common in cytologic specimens because of the nature of their scant material. Under these circumstances, immunohistochemistry is invaluable for determining lung origin in both primary and metastatic, poorly differentiated carcinomas and for distinguishing the subtype of carcinoma, (ie, squamous cell, adenocarcinoma, etc).
Thyroid transcription factor-1 (TTF-1) has been the predominant immunohistochemical (IHC) marker used to identify lung origin and has a reported sensitivity of 75% to 80% for lung adenocarcinomas.11, 12 However, TTF-1 also stains other tissues and tumors, such as thyroid tissue; metastatic breast carcinoma; neuroendocrine tumors, such as small cell lung carcinoma (SCLC) and carcinoid; and, to a lesser degree, primary lung squamous cell carcinoma.12-14 In addition, its expression reportedly decreases inversely in relation to the degree of tumor differentiation (ie, poorly differentiated adenocarcinomas are less likely to express TTF-1 compared with well differentiated carcinomas).11-13 Thus, the search for a surrogate marker with greater sensitivity and specificity for lung adenocarcinoma, along the lines of tumor differentiation, has been initiated.
Napsin-A, an aspartic proteinase involved in the maturation of the surfactant protein B, recently has come to the attention of investigators as a potential marker of lung adenocarcinoma.2, 15 The proteinase is expressed abundantly in the cytoplasm of normal lung cells (type II pneumocytes and Clara cells) and kidney cells (proximal and convoluted tubules) and in lung adenocarcinomas and renal cell carcinomas.15, 16 It has been demonstrated that the expression of napsin-A is regulated by TTF-1, a member of the Nkx2 family of transcription factors that also regulate the expression of surfactant protein B.16 Previous studies using resected tumor tissues demonstrated that napsin-A was equal to or better than TTF-1 and surfactant protein A immunostains for determining lung origin in well to moderately differentiated adenocarcinomas.17, 18 Furthermore, previous studies demonstrated that napsin-A and TTF-1 expression levels were decreased in poorly differentiated lung adenocarcinomas compared with the levels in well differentiated carcinomas.14, 19
Few studies have investigated the role of napsin-A in daily cytology practice or its sensitivity and specificity for the detection of poorly differentiated lung adenocarcinomas.18 In the current study, we investigated the utility of napsin-A and compared its sensitivity and specificity with those of TTF-1 for the identification of lung adenocarcinoma in cytologic specimens, particularly in so-called “poorly differentiated” carcinomas.
MATERIALS AND METHODS
The archives of the Department of Pathology at The Johns Hopkins Hospital were searched for “poorly differentiated adenocarcinoma” or “poorly differentiated carcinoma favor adenocarcinoma” of lung origin over 10 years from January 1, 2000 to February 1, 2010. The cytologic features of these tumors consisted of discohesive cells that lacked acinar formation or overt mucin production. Cells exhibited eccentric nuclei with prominent nucleoli as well as significant anisonucleosis and anisocytosis. There was a noted absence of opaque, “hard” cytoplasm, intracytoplasmic processes, or other features characteristic of squamous cell carcinoma. All patients who had their cytologic specimens included in this study were restricted to those who had corresponding surgical/histologic confirmation of lung adenocarcinoma and adequate cell block material for IHC study. In total, 75 patients with cytologic specimens were identified, and the specimens included both primary lung lesions (n = 28) and metastatic lung lesions (n = 47). Ninety-five cytologic specimens of other cancers, including 16 SCLCs, were used as controls. In addition, 2 tissue microarrays (TMAs), including 1 of bronchioloalveolar adenocarcinoma (BAC) and 1 of true pulmonary papillary adenocarcinoma, also were used. All specimens were stained subsequently for TTF-1 and napsin-A.
Processing of Cytologic Material
The cytologic specimens were collected using fine-needle aspiration (FNA) with ultrasound or computed tomography guidance and core biopsies (obtained with a 19-gauge needle). The material was prepared using both air-dried and wet-fixed methods. The air-dried slides were stained with Diff-Quik and used for immediate, on-site evaluation. The wet-fixed slides were stained using the Papanicolaou method (Pap stain) in the cytology laboratory. Other cytologic materials, such as bronchial washings and body fluids, were prepared by using the cytospin method and cell block preparation.
Cell Block and Core-Needle Biopsy Material Preparations
For cell block preparations, the aspiration needle was rinsed with 10 to 20 mL of Hanks balanced salt solution (Sigma Chemical Company, St. Louis, Mo) into a 50-mL centrifuge tube. Other types of specimens were collected in a centrifuge tube. Then, the cellular material was harvested using a microcentrifuge at 1870 revolutions per minute for 10 minutes (Hettich, Beverly, Mass). Next, 1 to 2 mL of 10% neutral buffered formalin were added to pellet, and the pellet was placed into a cassette, embedded with paraffin, and processed in the histology laboratory. The specimen was cut into 5-μm-thick sections and stained with hematoxylin and eosin (H&E). The core biopsy materials were fixed in 10% neutral buffered formalin and processed in the histology laboratory with subsequent H&E staining.
Tissue Microarray Construction
The TMAs (1 mm in greatest dimension; 3 cores per case) were constructed using specimens retrieved from the surgical pathology archives of The Johns Hopkins Hospital. Represented on the arrays were 26 BACs, including 19 of the nonmucinous subtype and 7 of the mucinous subtype, and 27 true pulmonary papillary adenocarcinomas. Twenty-three control tissues were used on each microarray and consisted of normal kidney tissue, gastrointestinal tissue, lung tissue, and lymph node. All tissues were obtained from routine 10% formalin-fixed and paraffin-embedded tissue blocks.
The IHC studies were performed on cell blocks, core biopsy material, and TMAs. Cell blocks and TMAs were used for TTF-1 and napsin-A staining, and core biopsy material was used for additional marker studies, such as renal cell carcinoma (RCC) and the paired box (PAX) genes PAX2 or PAX8 in specimens of RCC.
The specimens were cut into at 5-μm-thick sections and deparaffinized before incubation with primary antibodies. Napsin-A was stained using a monoclonal antihuman napsin-A antibody (Novocastra, Newcastle, United Kingdom) at 1:800 dilution with a Leica Bondmax autostainer (Leica Microsystem, Bannockburn, Ill). TTF-1 was stained using a monoclonal antibody (Cell Marque, Rocklin, Calif) at 1:500 dilution with a Ventana XT autostainer (Ventana Medical System, Tucson, Ariz). Staining characteristics were reviewed by the authors (L.M.S. and Q.K.L.) who considered the intensity and breadth of staining patterns. TTF-1 is a known nuclear stain, and napsin-A is a known cytoplasmic stain. Staining for napsin-A and TTF-1 was scored semiquantitatively using a 3-tier scoring system in which 0 indicated undetectable staining, 1+ indicated weakly positive staining, and 2+ indicated moderately to intensely positive staining. Focal positivity was defined as staining in ≤25% of cells. Appropriate positive and negative controls were included in the assay. Care also was taken not to interpret entrapped normal lung epithelium or pulmonary macrophages as positive for tumor staining.
Evaluation of Data
We prepared 2 × 2 contingency tables to calculate the sensitivity and specificity of TTF-1 and napsin-A. The Fisher exact test and the chi-square test were calculated by using Prism 4 software (GraphPad, Inc., La Jolla, Calif). If the α value was <.05, then it was considered statistically significant (P < .05).
In this study of primary or metastatic lung adenocarcinoma, the patients ranged in age from 28 years to 87 years (median age, 65 years), and the ratio of men to women was 1:1.2. Of 75 patients who had cytologic material available, 56 cases were diagnosed as “poorly differentiated carcinoma with glandular features” or “poorly differentiated carcinoma favor adenocarcinoma,” and 19 cases were diagnosed as “poorly differentiated carcinoma, not otherwise specified.” All of the patients who had cytologic material available in our study had corresponding surgical/histologic confirmation of lung adenocarcinoma. Forty-seven of 75 cytologic cases were diagnosed as metastatic lung adenocarcinoma. Sites of metastasis in descending order of frequency included neighboring lymph nodes (n = 23), pleural cavity (n = 13), soft tissue (n = 5), bone (n = 2), and others (n = 4).
The overall staining characteristics of TTF-1 and napsin-A in our patients are summarized in Tables 1 and 2. Among 75 specimens that were identified cytologically as poorly differentiated carcinoma but were confirmed on histology as primary lung adenocarcinoma, TTF-1 and napsin-A staining was positive in 61 of 75 specimens (81%) and 49 of 75 specimens (65%), respectively (Table 1) (Fig. 1). Within primary lung adenocarcinomas, TTF-1 and napsin-A staining was positive in 20 of 28 specimens (71%) and 16 of 28 specimens (57%), respectively, whereas 8 of 28 specimens (29%) were negative for both markers. Four specimens displayed TTF-1 expression with absent staining for napsin-A; however, TTF-1 staining was focal in 2 of these specimens (Table 2). Among the metastatic lung adenocarcinomas, TTF-1 and napsin-A staining was positive in 41 of 47 specimens (87%) and 33 of 47 specimens (70%), respectively, and 5 of 47 specimens (11%) were negative for both markers. Eight specimens expressed TTF-1 with negative napsin-A staining (Table 2). An additional specimen had only napsin-A expression with focal cytoplasmic staining. Among the specimens that were stained only for TTF-1, expression was focal in 5 of 8 tumor cells (63%) and weak in 1 of 8 tumor cells (12.5%). A comparison of the expression of TTF-1 and napsin-A between primary and metastatic carcinomas revealed a slight difference (P < .05 and P = .0415, respectively). Taken together, concordant staining results were observed in 60 of 75 specimens (80%). The overall sensitivity and specificity of TTF-1 were 81% and 81%, respectively, whereas napsin-A had a sensitivity of 65% and a specificity of 96% (Table 3).
|Diagnoses||Total No./No. Positive (%)|
|TTF-1 Positivity||Napsin-A Positivity|
|Primary lung adenocarcinoma||20/28 (71.4)||16/28 (57.1)|
|Metastatic lung adenocarcinoma||41/47 (87.2)||33/47 (70.2)|
|Total||61/75 (81.3)||49/75 (65.3)|
|Nonmucinous BAC||17/19 (89.5)||18/19 (94.7)|
|Mucinous BAC||5/7 (71.4)||6/7 (85.7)|
|PPA||21/27 (77.8)||26/27 (96.3)|
|Total||43/53 (81.1)||50/53 (94.3)|
|SCLC||13/16 (81.3)||0/16 (0)|
|SCC, lung||2/35 (5.7)||0/35 (0)|
|SCC, anus and skin||0/3 (0)||0/3 (0)|
|Gynecologic/breast carcinoma||0/8 (0)||0/8 (0)|
|RCC||0/14 (0)||4/14 (28.6)|
|Clear cell type||0/10 (0)||1/10 (10)|
|Papillary type||0/4 (0)||3/4 (75)|
|Pancreatic carcinoma||0/13 (0)||0/13 (0)|
|Gastric carcinoma||0/1 (0)||0/1 (0)|
|Large cell neuroendocrine carcinoma||2/2 (100)||0/2 (0)|
|Unknown primary carcinoma||0/3 (0)||0/3 (0)|
|Case No.||Cytologic Diagnosis||TTF-1||Napsin-A|
|Primary lung carcinomas|
|1||PD adenocarcinoma||F. positive||0|
|2||PD adenocarcinoma||F. positive||2+|
|3||PD adenocarcinoma||2+||F. positive|
|5||PD carcinoma with glandular and squamous features||2+||0|
|6||PD carcinoma with glandular and squamous features||F. positive||0|
|Metastatic lung carcinomas|
|7||PD adenocarcinoma||F. positive||0|
|9||PD adenocarcinoma||0||F. positive|
|10||PD adenocarcinoma||F. positive||0|
|14||PD adenocarcinoma||F. positive||0|
|15||PD carcinoma||F. positive||0|
|Results||No. of Cytologic Cases, n=75||No. of Cytologic Controls, n=95|
Analysis of the TMAs revealed TTF-1 nuclear staining in 22 of 26 BAC specimens (85%) and in 21 of 27 papillary adenocarcinoma specimens (78%) (Table 1). Specifically, 17 of 19 nonmucinous BAC specimens (89%) and 5 of 7 mucinous BAC specimens (71%) displayed intense TTF-1 expression. Napsin-A staining was cytoplasmic in 24 of 26 BAC specimens (92%) and in 26 of 27 papillary adenocarcinoma specimens (96%) (Table 1). Eighteen nonmucinous BAC specimens (95%) and 6 of 7 mucinous BAC specimens (86%) expressed napsin-A. Among the control tissues, TTF-1 was absent in normal kidney, lymph node, and gastrointestinal sections. TTF-1 was expressed in normal lung tissue, as expected. Napsin-A was strongly expressed in the normal lung tissue (macrophages and pneumocytes) and in tubules from normal kidney tissue. Napsin-A expression was absent in control lymph node and gastrointestinal sections.
In total, 95 cytologic specimens were included in our study as controls. They consisted of SCLC (n = 16), poorly differentiated carcinomas that arose in sites other than the lung, or poorly differentiated lung masses that were diagnosed previously as neoplasms other than adenocarcinoma (n = 79; ie, 8 gynecologic/breast primary tumors, 38 squamous cell carcinomas [including 35 lung primary tumors and 3 metastases of anal or skin origin], 13 pancreatic primary tumors, 14 RCCs [10 clear cell carcinomas and 4 papillary carcinomas], 1 gastric primary tumor, 3 unknown primary tumors, and 2 metastatic large cell neuroendocrine carcinomas of lung origin). Within this collection of neoplasms, 4 of 79 specimens (5.1%) stained with TTF-1 (2 poorly differentiated squamous cell carcinomas of the lung and 2 metastatic large cell neuroendocrine carcinomas of lung primaries). Napsin-A staining was present only in 4 of 14 RCCs (28.6%), including 1 clear cell carcinoma (1 of 10 RCCs; 10%) and 3 of 4 papillary carcinomas (75%) (Table 1). Among the SCLCs, 13 of 16 specimens (81%) stained with TTF-1, whereas napsin-A was negative in all 16 specimens (Fig. 2). Specimens of metastatic carcinomas of the pancreas, stomach, gynecologic/breast origin, and of unknown primary origin were negative for both TTF-1 and napsin-A (Table 1).
The distinction of squamous and nonsquamous cell carcinoma, including adenocarcinoma and large cell carcinoma, can be made in the majority of patients in daily cytology practice. However, in certain patients, the distinction cannot be made by assessment of morphology alone, because small samples usually are obtained by FNA procedures. In this situation, the profile of IHC markers in tumor cells may provide additional differential diagnostic information. In the current study, we compared the immunoreactivity of napsin-A versus TTF-1 in cytologically so-called “poorly differentiated” carcinomas.
Napsin-A is a 35-kilodalton protein that is expressed abundantly in certain cells, such as type II pneumocytes, alveolar macrophages, and renal tubular cells.2, 6 In the lung, it functions as an aspartate proteinase and regulates the maturation and activation of surfactant protein B. It was described first in the late 1990s; since then, it has been used increasingly as a marker for both lung adenocarcinoma and RCC.2, 6 Previous studies using histologic specimens indicated that napsin-A has a sensitivity equal to or greater than that of TTF-1 in well to moderately differentiated lung adenocarcinomas.2, 14, 20, 21 Therefore, its use has been advocated in conjunction with TTF-1 in the differential diagnosis of lung adenocarcinomas. However, the utility of napsin-A as a marker in poorly differentiated adenocarcinomas, particularly within cytologic specimens, has not been well studied. Demjeck et al published a study in which they examined TTF-1 and napsin-A in 50 specimens of pleural effusions.21 In that study, the authors observed that napsin-A and TTF-1 were positive in 8 and in 10 of 12 lung adenocarcinoma specimens, respectively; and both napsin-A and TTF-1 were negative in other types of adenocarcinomas.21
Among the 75 specimens from our patients, we observed that 81% (61 of 75 specimens) stained for TTF-1, and 65% (49 of 75 specimens) stained for napsin-A. In the specimens that we studied, the overall sensitivity was 81% for TFF-1 and 65% for napsin-A. However, we observed that TTF-1 was detected in 2 of 2 specimens of metastatic large cell neuroendocrine carcinoma from a lung primary, in 2 of 35 specimens (5.7%) of lung squamous cell carcinoma, and in 13 of 16 specimens (81%) of SCLC. This correlates with previous results, which demonstrated TTF-1 expression in pulmonary neuroendocrine tumors, specifically in SCLC (with a reported staining rate of 70% to 80%); large cell neuroendocrine carcinoma, and rare cases of squamous cell carcinomas22-24; however, napsin-A expression was not detected in the aforementioned cases. Conversely, napsin-A expression was not detected in our SCLC specimens, consistent with previous reports.12, 14 SCLC is a diagnosis typically made on H&E slides with immunohistochemistry used as supporting evidence. However, in poorly differentiated carcinomas, the distinction between SCLC and nonsmall cell lung carcinoma on histology/cytology alone may be more difficult, which paves the way for using napsin-A in combination with neuroendocrine marker panels. Our data indicate that TTF-1 and napsin-A appear to stain more metastatic lung adenocarcinomas (87% and 70%, respectively) than primary lung adenocarcinomas (71% and 57%, respectively). The reason for this observation is unclear and may be related to an experimental bias, because we only included a small number of patients in the current investigation. The observation needs to be examined and validated in a large-scale study. Taken together, our results indicate that TTF-1 has a greater sensitivity and that napsin-A has a better specificity of 96%; thus, their dual use may be necessary for the evaluation of poorly differentiated carcinomas of probable lung origin or unknown primary carcinomas.
Our results indicated that sensitivity was greater for napsin-A than for TTF-1 among our specimens of well to moderately differentiated lung adenocarcinoma (TMAs). The sensitivity of napsin-A was 96% in specimens of adenocarcinoma that had papillary features compared with 78% sensitivity for TTF-1. The sensitivities for napsin-A and TTF-1 were relatively similar in specimens of mucinous and nonmucinous BAC. Our finding of higher sensitivity for napsin-A compared with TTF-1 for the detection of well to moderately differentiated adenocarcinoma has been documented previously.6, 14, 19, 21 The data reported by Bishop et al14 and Hirano et al19 also demonstrated decreased expression of napsin-A and TTF-1 in poorly differentiated carcinomas.14, 19 This trend was reflected to some degree in our current data. TTF-1 was expressed in 89% of nonmucinous BAC/well differentiated adenocarcinoma, in 78% of papillary lung adenocarcinomas, and in 81% of poorly differentiated lung adenocarcinomas. Napsin-A was expressed in 95% of nonmucinous BAC, in 96% of papillary lung adenocarcinomas, and in 65% of poorly differentiated lung adenocarcinomas. Although napsin-A demonstrated high expression in well differentiated adenocarcinomas and greatly diminished expression in poorly differentiated adenocarcinomas, TTF-1 staining did not exhibit a consistent trend. Indeed, TTF-1 expression was higher in nonmucinous BAC specimens compared with more poorly differentiated adenocarcinomas; however, only 78% of papillary adenocarcinomas had TTF-1 staining. The reason for the TTF-1 “irregularities” is uncertain and may be related to sampling issues. Documentation of TTF-1 staining in papillary lung adenocarcinomas is lacking in the literature for the purposes of comparison. In our study, the trend of diminished expression along the lines of tumor cell differentiation was consistent with previously published observations. However, the hypothesis of higher expression levels in well differentiated specimens compared with poorly differentiated specimens still needs to be tested in a large-scale study.
Our data demonstrate that TTF-1 has higher positivity (81%) than napsin-A (65%) in cytologically so-called “poorly differentiated” carcinomas. However, the data reported by Bishop et al14 demonstrated that napsin-A had more positivity (69%) than TTF-1 (44%) among poorly differentiated lung adenocarcinomas. There are several reasons for this discrepancy. In the study by Bishop et al,14 the data were collected using TMA, in which the tissue sections may have had higher quantities of tumor cells for staining and visual detection. In our study, the observations were made on cytologic material. The poorly differentiated adenocarcinomas consisted entirely of cytologic specimens that had matching cell blocks, which may have contained fewer tumor cells and/or poorly preserved cells. Hence, the IHC results may have been hampered by the quantity of cells available for staining. However, a review of cell blocks and a comparison of staining patterns revealed no significant difference in staining patterns based on the amount of tissue present on the slides. Second, sampling error also may have played a role in our study, because cytologic specimens were obtained by FNA, whereas TMA data were based on surgically resected specimens. Finally, both the current study and the study by Bishop et al had relatively small numbers of “poorly differentiated” carcinomas, which included 75 specimens in the current study and 16 specimens in the study by Bishop et al. Admittedly, the differential expression of TTF-1 and napsin-A in different studies may have been caused by experimental bias and limitations. We believe that this observation needs to be examined further in a large-scale study.
Napsin-A is expressed within normal proximal and convoluted tubules of the kidney along with RCCs, as reported previously.15, 16 In our study, we included normal renal tissues and RCC specimens as controls. Napsin-A staining was noted in the renal tubule cells on the TMAs and was observed in 3 of 4 papillary renal carcinoma specimens (75%) and in 1 of 10 clear cell carcinoma specimens (10%). Larger studies have demonstrated napsin-A staining in 34% of clear cell renal carcinomas and in 79% of papillary RCCs.14 It is important to keep this cross-staining in mind when a patient presents with a lung mass with or without a known history of RCC. In addition, napsin-A is a granular cytoplasmic stain that has a propensity for higher background staining and cross-reactivity in pulmonary macrophages (presumably because of the phagocytosis of pneumocytes). In our experience, pulmonary macrophages often were present in association with tumor cells/fragments on cytologic preparation; thus, it is also important not to interpret these cells as lung adenocarcinoma cells within cytologic specimens.
Finally, we produced no evidence of napsin-A expression or TTF-1 expression in our specimens of adenocarcinoma of breast or gynecologic origin. Because breast carcinomas are 1 of the more common tumors that metastasize to the lung, the absence of staining for both immunomarkers is reassuring. It is known that a small percentage of lung carcinomas will stain for gross cystic disease fluid protein and estrogen receptors (ERs).13 Because both tumor types express cytokeratin 7 (CK7) with absent staining for CK20, napsin-A and TTF-1 are useful markers for identifying a lung primary tumor, particularly in patients with a known history of breast carcinoma. However, a note of caution must be raised when evaluating a carcinoma of unknown origin. In such patients, it must be kept in mind that TTF-1 has a propensity to cross-react with neuroendocrine tumors, such as SCLC, and with thyroid neoplasms. Conversely, napsin-A can be expressed within clear cell and papillary variants of RCC. In such instances in which thyroid or renal primary is considered, other stains, such as either thyroglobulin and CD10 (a type II transmembrane protein) or PAX8/PAX2 may be used, respectively.
In conclusion, specificity was higher for napsin-A in our study, which advocates for its use in conjunction with TTF-1 for evaluating a poorly differentiated lung carcinoma or a metastatic lesion of probable lung origin. TTF-1 remains a desirable marker for poorly differentiated adenocarcinoma of the lung given its greater sensitivity and staining characteristics. However, we believe that this observation needs to be tested in a larger study.
CONFLICT OF INTEREST DISCLOSURES
Supported in part by the American Society of Cytopathology Foundation Cytopathology Research Seed Grant (to Q.K.L.).
- 3Global Cancer Facts and Figures 2007. Atlanta, GA: American Cancer Society; 2007., , , et al.
- 7Erlotinib in nonsmall-cell cancer: a review of the clinical and economic evidence. Expert Rev Pharamacoecon Outcomes Res. 2009; 9: 409-416..
- 10Female sex and bronchioloalveolar pathologic subtype predict EGFR mutations in nonsmall-cell lung cancer. Chest. 2005; 128: 217-321., , , et al.
- 18TKTL1 is overexpressed in a large portion of nonsmall cell lung cancer specimens. Diagn Pathol. 2008; 12: 1-5., , , et al.