ORIGINAL ARTICLE
Lung Tumors Masquerading as Desquamative Interstitial Pneumonia (DIP) Report of 7 Cases and Review of the Literature Kirtee Raparia, MD,* James Ketterer, BS,w Mercedes L. Dalurzo, MD,z Yu-Hui Chang, PhD,w Thomas V. Colby, MD,w and Kevin O. Leslie, MDw
Abstract: Malignant tumors in the lung (both primary and metastatic) rarely may be associated with markedly discohesive tumor cells, resulting in airspace filling reminiscent of “desquamative interstitial pneumonia” (DIP) on histopathology evaluation. A peculiar aspect of this growth pattern is the relatively bland appearance of the tumor cells, in many cases simulating benign alveolar macrophages at scanning magnification. We searched the Charles Carrington Memorial consultation files in the Department of Laboratory Medicine and Pathology at Mayo Clinic Arizona for instances of malignant tumors in lung simulating DIP, from 1992 to 2011. We identified 7 cases involving transbronchial biopsies, needle core samples, or resected lung specimens. Clinical, histopathologic, and immunohistochemical analyses of these 7 patients were performed, including detailed morphometric analysis of the individual tumor cells using calibrated measurement tools on digital images. We compared the results with those of a control group of 4 patients with benign DIP-macrophage reactions in smoking-related lung disease. The study group comprised 5 male and 2 female patients, 48 to 86 years in age (median: 67 y). The radiologic findings included lobar consolidation, localized groundglass opacities, and 1 or more nodules. None of the patients had typical bilateral infiltrates of DIP. Microscopically, the lung parenchyma was dominated by the presence of prominent tumor cells filling alveolar spaces. Four patients had primary lung carcinoma (adenocarcinoma), whereas 3 had metastases from other sites, including a melanoma. Immunohistochemical staining studies were performed on 6 of 7 cases to establish the diagnosis. Nuclear diameter, cytoplasmic diameter, and nuclear/ cytoplasmic (N/C) ratios in patient and control groups were compared using the Wilcoxon rank sum test. No significant difference in the diameters of nucleus and cytoplasm between cases and control groups (P = 0.3447 and 0.7055, respectively) was seen, and only a marginally significant difference in N/C ratios (P = 0.0890) was seen. A more complex analysis, generFrom the *Department of Pathology, Northwestern University, Chicago, IL; wMayo Clinic, Scottsdale, AZ; and zItalian Hospital of Buenos Aires, Buenos Aires, Argentina. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Kevin O. Leslie, MD, Mayo Clinic Arizona, 13400 East Shea Blvd, Scottsdale, AZ 85259 (e-mail: [email protected]). Copyright r 2014 by Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 7, July 2014
alized estimating equation analysis, showed a significant difference in N/C ratio between the 2 groups (P = 0.0278). A “DIPgrowth pattern” of malignant tumors in the lung is presented. Although the N/C ratio differences approached statistical significance when compared with controls, the key to diagnosis is the recognition of the malignant cytology of the tumor nuclei. Immunohistochemical studies (keratin or other markers) are helpful in establishing an accurate diagnosis in this setting. Key Words: lung adenocarcinoma, desquamative interstitial pneumonia (DIP), metastasis, cytology (Am J Surg Pathol 2014;38:921–924)
T
he term “desquamative interstitial pneumonia” (DIP) was first proposed by Liebow in 1962 for a diffuse lung disease occurring in patients who often were smokers.1–3 It was initially believed that these intra-alveolar cells were desquamated pneumocytes and of epithelial origin, hence the name. The major histologic feature in DIP is the presence of large numbers of macrophages within alveoli. These macrophages have abundant eosinophilic cytoplasm, often with fine granular light-brown pigment. The alveolar architecture is generally well preserved with a mild chronic inflammatory cell infiltrate within the interstitium. The intra-alveolar cells in DIP were later confirmed to be macrophages containing membrane-bound dense bodies, devoid of intercellular junctions and lacking microvilli by electron microscopy.4,5 These cells stain positively with the macrophage marker CD68.3 A peculiar characteristic of malignant epithelioid cells in lung parenchyma is their presence on occasion within alveolar spaces on histopathologic examination. In most instances this is a focal phenomenon, but occasionally this growth appearance can be prominent and even dominant within the tumor. It is mentioned in the Armed forces of Institute of Pathology (AFIP) fascicle (third series) on tumors of the lower respiratory tract that some pulmonary adenocarcinomas have discohesive zones with large numbers of cells in the airspaces.6 We highlight this peculiar growth pattern of tumors in the lung (both primary and metastatic) in cases in which tumor extension into the alveolar spaces produces a pattern reminiscent of “desquamative interstitial pneumonia.” www.ajsp.com |
921
Raparia et al
Am J Surg Pathol
The relatively bland appearance of the tumor cells, in many cases simulating macrophages, can be a potential mimicker of benign disease, especially in small biopsy samples. Although the cells involved in our cases were discernible on cytopathologic grounds as being malignant, knowledge of the radiologic findings also may help exclude serious consideration of an idiopathic interstitial pneumonia such as DIP.
DESIGN We searched the Charles Carrington Memorial consultation files in the Mayo Clinic Arizona Pathology Department for instances of malignant tumors in the lung simulating DIP, from 1992 to 2011. We identified 7 cases involving the transbronchial biopsies (n = 1), needle core samples (n = 1), and resected lung specimens (n = 5). One case was referred to the study after its inception by one of the authors (M.L.D.). Clinical, histopathologic, and immunohistochemical analyses of these 7 patients were performed, including detailed morphometric analysis of the individual tumor cells using calibrated measurement tools on digital images. We compared the results with similar measurements taken from a control group of 4 patients with DIP-reactions in smoking-related disease. To better understand the incidence of this phenomenon, we reviewed histopathologic data from 150 random resected lung adenocarcinomas for the occurrence of a DIPlike growth pattern. The average nuclear and cytoplasmic diameters of 100 intra-alveolar cells were measured in each study case and control sample using a morphometric software program (Olympus “cellSens Standard”) on digital images captured with an Olympus microscope equipped with a 12 megapixel digital camera. Nuclear-cytoplasmic (N/C) ratios were calculated for the data set derived from each subject and control sample. The medians of averaged data (maximum nuclear dimension, same cell cytoplasmic dimension, and calculated N/C ratio) were compared between cases and controls using the Wilcoxon rank sum test. Generalized estimating equations (GEE) were also
Volume 38, Number 7, July 2014
TABLE 2. Detailed Morphometric Analysis of the Tumor Cells and Control Group Using Calibrated Measurement Tools on Digital Images Mean (SD) Cases
Controls
1 2 3 4 5 6 7 1 2 3 4
Nuclear Size 8.62 8.49 9.51 6.35 6.54 8.08 10.51 8.36 7.89 7.53 7.41
(1.46) (1.50) (1.58) (1.26) (1.38) (1.91) (1.52) (1.35) (1.33) (1.44) (1.61)
Cytoplasmic Size 17.83 18.97 17.25 11.15 11.48 12.02 18.59 15.72 15.62 20.29 18.59
(4.25) (4.01) (3.15) (2.27) (2.08) (2.54) (3.96) (2.78) (2.80) (4.53) (2.87)
N/C Ratio 0.50 0.46 0.56 0.58 0.58 0.68 0.59 0.54 0.51 0.38 0.48
(0.11) (0.09) (0.11) (0.11) (0.11) (0.14) (0.12) (0.10) (0.08) (0.10) (0.10)
used to evaluate whether there was a difference in the diameters and N/C ratio between groups. The advantage of using GEE was that rather than computing a single mean for each patient, all the measurements from each patient were included in the analysis, and the correlation of the ratios within each patient was taken into account. Hence, this method was more appropriate for correlated data. Using a 2-sided test, a P-value 80 > 80
Met
Poorly differentiated adenocarcinoma Adenocarcinoma
78
Atelectasis and pleural effusion, left M Transthoracic needle core Chronic RLL infiltrate biopsy F Bronchoscopic biopsy Pneumonia in RUL
Prim
6
78
M Lobectomy
Mass in RLL
Prim
7
64
M Lobectomy
LLL consolidation
Prim
Prim
100
Poorly differentiated adenocarcinoma Poorly differentiated adenocarcinoma Mucinous adenocarcinoma
IHC Ker+ S100+ MART-1+ Ker+
> 80
Ker+
100
None
> 80
Ker+
100
Ker+
IHC, immunohistochemistry; Ker, keratin; MART-1, melanoma-associated antigen recognized by T cells; RLL, right lower lobe; RUL, right upper lobe.
922 | www.ajsp.com
r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 7, July 2014
Lung Tumors Masquerading as DIP
TABLE 3. Statistics Based on the Mean Values From the Patients for Cases and Controls Groups Case (n = 7)
Control (n = 4)
Mean SD Median Mean SD Median Nuclear 8.30 1.50 Cytoplasm 15.33 3.59 N/C ratio 0.56 0.07
8.49 17.25 0.50
7.80 0.43 16.84 2.30 0.48 0.07
7.71 15.73 0.50
P*
Pw
0.3447 0.0278 0.7055 0.3669 0.0890 0.3441
*Wilcoxon rank sum test. wThe Wald test from GEE.
was identified in 27 (18%) of 150 lung adenocarcinoma samples, ranging in extent from 5% to 25%. Four of our patients had primary lung carcinoma (adenocarcinoma), whereas 3 had metastatic tumor from other sites (cases 1, 2, and 3). Immunostaining analyses were performed on 6 of 7 cases to confirm the diagnosis. The case (case 5) in which immunostaining analyses were not performed showed tumor with focal areas of glandular differentiation. The clinical, histopathologic, and immunohistochemical characteristics of the 7 patients are summarized in Table 1. The means and
SDs of nuclear diameter, cytoplasmic size, and N/C ratio for each patient are presented in Table 2. The summary statistics based on the mean values from the patients for cases and controls are presented in Table 3. Using the Wilcoxon rank sum test, no significant differences were seen in the nuclear or cytoplasmic diameters when comparing study cases with control groups (P = 0.3447 and 0.7055, respectively), and only a marginal significant difference in N/C ratios (P = 0.0890) was seen (Table 3). The group difference was also evaluated by applying GEE with normal link. These results showed a significant difference in the N/C ratio between the 2 groups, with study cases being on average 0.08 units (P = 0.0278) higher than controls. The diameters of nuclei and cytoplasm did not differ between groups (P = 0.3669 and 0.3441, respectively).
DISCUSSION Our small series of cases highlights an uncommon but real problem related to the diagnosis of lung tumors. When tumors in the lung grow mainly within the alveolar spaces, the process still should be diagnosable if the cytology of the tumor cells is definitively malignant. In
FIGURE 1. DIP-pattern. A, Pigmented “smokers-type” alveolar macrophages filling alveolar spaces. B, Pulmonary adenocarcinoma. C, Metastatic renal cell carcinoma. D, Metastatic melanoma. Hematoxylin and eosin stains. r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
923
Raparia et al
some instances, however, the diagnosis can be challenging if the cytology of tumor cells is relatively bland. The problem is magnified in small samples in which DIP reaction of the tumor may be dominant. Both primary and metastatic tumors in the lung can show this phenomenon. The “DIP carcinomas” described in the 1995 AFIP fascicle illustrated this phenomenon in addition to areas of conventional adenocarcinoma with characteristic gland formation or mucin production.7 The radiology findings in these cases do not always show a discrete mass-like lesion, making the diagnosis more challenging. To better understand this histopathologic growth pattern, we evaluated biopsies from 6 patients with adenocarcinoma and 1 patient with metastatic melanoma who had a “DIP-growth pattern” and compared the observations with lung biopsies from 4 patients who had smoking-related DIP alone (Fig. 1). On the basis of hematoxylin and eosin examination, we encountered some difficulty in confirming the diagnosis of malignant tumor. With the hypothesis that malignant exfoliated intraalveolar cells should be measurably different from alveolar macrophages, we compared the mean and SD of nuclear diameter, cytoplasmic size, and N/C ratio for cases and controls, using morphometry. We were surprised to find that, even though the nuclear features of the malignant cases allowed recognition of a neoplastic process, using the Wilcoxon rank sum test, no significant difference in the diameters of individual cell nuclei and cytoplasm between cases and control groups was seen, and only a marginal significant difference in N/C ratios (P = 0.0890) was identified. We further evaluated the group difference by applying GEE with normal link, which did show a significant difference in N/C ratio between the cases and controls; with tumor cases on average having higher ratios compared with control by 0.08 units (P = 0.0278). The phenomenon we describe is not new. Mutton et al3 published a case of poorly differentiated lung adenocarcinoma, which was mistakenly diagnosed as DIP on open lung biopsy, and resulted in the death of the patient. The authors reviewed the case and compared it with 3 cases of DIP and found that the intra-alveolar cell populations in both DIP and adenocarcinoma have similar features on hematoxylin and eosin examination. However, they identified focal areas of lymphovascular invasion in the patient with adenocarcinoma, and the diagnosis was confirmed only on autopsy material. Yousem7 showed a similar growth pattern (alveolar filling) in cases of peripheral squamous cell carcinoma, wherein cohesive aggregates of tumor cells filled airspaces formed by intact nondisrupted alveolar septa. A recent study by
924 | www.ajsp.com
Am J Surg Pathol
Volume 38, Number 7, July 2014
Onozato et al8 identified a large collection of tumor cells present within alveolar spaces, which they defined as tumor islands in patients with lung adenocarcinoma, exhibiting a micropapillary and solid growth pattern. The authors mentioned that these tumor cells composing the islands were cytologically similar to those of the main tumor and could easily be differentiated from alveolar macrophages given a high N/C ratio and cytologic atypia.8 Although the mechanism of this “DIP-growth” of tumor is unknown, it is tempting to consider this growth pattern as a means of tumor spread in high-grade lung adenocarcinomas and other discohesive tumors.
CONCLUSIONS A “DIP-growth pattern” of malignant tumor growth in the lung is presented. This pattern can be seen in both primary and metastatic lung tumors, in our study adenocarcinomas and a melanoma, but in the literature also peripheral squamous cell carcinomas. Although measurement of the N/C ratio difference approached statistical significance, the key to the diagnosis was the individual cytology of the tumor cell nuclei. Immunohistochemical studies (keratin or other markers) are helpful in confirming the diagnosis in these tumors. We propose that the “DIP-growth pattern” might be regarded as an important morphologic feature of epithelioid tumors and that pathologists should be aware of this peculiar growth pattern in practice and always consider the radiologic findings before considering a diagnosis of an idiopathic interstitial pneumonia. REFERENCES 1. Leibow AA. Desquamative interstitial pneumonia. Am J Pathol. 1962;41:127. 2. Liebow AA, Steer A, Billingsley JG. Desquamative interstitial pneumonia. Am J Med. 1965;39:369–404. 3. Mutton AE, Hasleton PS, Curry A, et al. Differentiation of desquamative interstitial pneumonia (DIP) from pulmonary adenocarcinoma by immunocytochemistry. Histopathology. 1998;33:129–135. 4. Corrin B, Price AB. Electron microscopic studies in desquamative interstitial pneumonia associated with asbestos. Thorax. 1972;27:324–331. 5. Brewer DB, Heath D, Asquith P. Electron microscopy of desquamative interstitial pneumonia. J Pathol. 1969;97:317–323. 6. Colby TV, Koss MN, Travis WD. Tumors of the Lower Respiratory Tract. Washington, DC: Armed Forces Institute of Pathology; 1995. 7. Yousem SA. Peripheral squamous cell carcinoma of lung: patterns of growth with particular focus on airspace filling. Hum Pathol. 2009;40:861–867. 8. Onozato ML, Kovach AE, Yeap BY, et al. Tumor islands in resected early-stage lung adenocarcinomas are associated with unique clinicopathologic and molecular characteristics and worse prognosis. Am J Surg Pathol. 2013;37:287–294.
r
2014 Lippincott Williams & Wilkins
Lung Tumors Masquerading as Desquamative Interstitial Pneumonia (DIP) Report of 7 Cases and Review of the Literature Kirtee Raparia, MD,* James Ketterer, BS,w Mercedes L. Dalurzo, MD,z Yu-Hui Chang, PhD,w Thomas V. Colby, MD,w and Kevin O. Leslie, MDw
Abstract: Malignant tumors in the lung (both primary and metastatic) rarely may be associated with markedly discohesive tumor cells, resulting in airspace filling reminiscent of “desquamative interstitial pneumonia” (DIP) on histopathology evaluation. A peculiar aspect of this growth pattern is the relatively bland appearance of the tumor cells, in many cases simulating benign alveolar macrophages at scanning magnification. We searched the Charles Carrington Memorial consultation files in the Department of Laboratory Medicine and Pathology at Mayo Clinic Arizona for instances of malignant tumors in lung simulating DIP, from 1992 to 2011. We identified 7 cases involving transbronchial biopsies, needle core samples, or resected lung specimens. Clinical, histopathologic, and immunohistochemical analyses of these 7 patients were performed, including detailed morphometric analysis of the individual tumor cells using calibrated measurement tools on digital images. We compared the results with those of a control group of 4 patients with benign DIP-macrophage reactions in smoking-related lung disease. The study group comprised 5 male and 2 female patients, 48 to 86 years in age (median: 67 y). The radiologic findings included lobar consolidation, localized groundglass opacities, and 1 or more nodules. None of the patients had typical bilateral infiltrates of DIP. Microscopically, the lung parenchyma was dominated by the presence of prominent tumor cells filling alveolar spaces. Four patients had primary lung carcinoma (adenocarcinoma), whereas 3 had metastases from other sites, including a melanoma. Immunohistochemical staining studies were performed on 6 of 7 cases to establish the diagnosis. Nuclear diameter, cytoplasmic diameter, and nuclear/ cytoplasmic (N/C) ratios in patient and control groups were compared using the Wilcoxon rank sum test. No significant difference in the diameters of nucleus and cytoplasm between cases and control groups (P = 0.3447 and 0.7055, respectively) was seen, and only a marginally significant difference in N/C ratios (P = 0.0890) was seen. A more complex analysis, generFrom the *Department of Pathology, Northwestern University, Chicago, IL; wMayo Clinic, Scottsdale, AZ; and zItalian Hospital of Buenos Aires, Buenos Aires, Argentina. Conflicts of Interest and Source of Funding: The authors have disclosed that they have no significant relationships with, or financial interest in, any commercial companies pertaining to this article. Correspondence: Kevin O. Leslie, MD, Mayo Clinic Arizona, 13400 East Shea Blvd, Scottsdale, AZ 85259 (e-mail: [email protected]). Copyright r 2014 by Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 7, July 2014
alized estimating equation analysis, showed a significant difference in N/C ratio between the 2 groups (P = 0.0278). A “DIPgrowth pattern” of malignant tumors in the lung is presented. Although the N/C ratio differences approached statistical significance when compared with controls, the key to diagnosis is the recognition of the malignant cytology of the tumor nuclei. Immunohistochemical studies (keratin or other markers) are helpful in establishing an accurate diagnosis in this setting. Key Words: lung adenocarcinoma, desquamative interstitial pneumonia (DIP), metastasis, cytology (Am J Surg Pathol 2014;38:921–924)
T
he term “desquamative interstitial pneumonia” (DIP) was first proposed by Liebow in 1962 for a diffuse lung disease occurring in patients who often were smokers.1–3 It was initially believed that these intra-alveolar cells were desquamated pneumocytes and of epithelial origin, hence the name. The major histologic feature in DIP is the presence of large numbers of macrophages within alveoli. These macrophages have abundant eosinophilic cytoplasm, often with fine granular light-brown pigment. The alveolar architecture is generally well preserved with a mild chronic inflammatory cell infiltrate within the interstitium. The intra-alveolar cells in DIP were later confirmed to be macrophages containing membrane-bound dense bodies, devoid of intercellular junctions and lacking microvilli by electron microscopy.4,5 These cells stain positively with the macrophage marker CD68.3 A peculiar characteristic of malignant epithelioid cells in lung parenchyma is their presence on occasion within alveolar spaces on histopathologic examination. In most instances this is a focal phenomenon, but occasionally this growth appearance can be prominent and even dominant within the tumor. It is mentioned in the Armed forces of Institute of Pathology (AFIP) fascicle (third series) on tumors of the lower respiratory tract that some pulmonary adenocarcinomas have discohesive zones with large numbers of cells in the airspaces.6 We highlight this peculiar growth pattern of tumors in the lung (both primary and metastatic) in cases in which tumor extension into the alveolar spaces produces a pattern reminiscent of “desquamative interstitial pneumonia.” www.ajsp.com |
921
Raparia et al
Am J Surg Pathol
The relatively bland appearance of the tumor cells, in many cases simulating macrophages, can be a potential mimicker of benign disease, especially in small biopsy samples. Although the cells involved in our cases were discernible on cytopathologic grounds as being malignant, knowledge of the radiologic findings also may help exclude serious consideration of an idiopathic interstitial pneumonia such as DIP.
DESIGN We searched the Charles Carrington Memorial consultation files in the Mayo Clinic Arizona Pathology Department for instances of malignant tumors in the lung simulating DIP, from 1992 to 2011. We identified 7 cases involving the transbronchial biopsies (n = 1), needle core samples (n = 1), and resected lung specimens (n = 5). One case was referred to the study after its inception by one of the authors (M.L.D.). Clinical, histopathologic, and immunohistochemical analyses of these 7 patients were performed, including detailed morphometric analysis of the individual tumor cells using calibrated measurement tools on digital images. We compared the results with similar measurements taken from a control group of 4 patients with DIP-reactions in smoking-related disease. To better understand the incidence of this phenomenon, we reviewed histopathologic data from 150 random resected lung adenocarcinomas for the occurrence of a DIPlike growth pattern. The average nuclear and cytoplasmic diameters of 100 intra-alveolar cells were measured in each study case and control sample using a morphometric software program (Olympus “cellSens Standard”) on digital images captured with an Olympus microscope equipped with a 12 megapixel digital camera. Nuclear-cytoplasmic (N/C) ratios were calculated for the data set derived from each subject and control sample. The medians of averaged data (maximum nuclear dimension, same cell cytoplasmic dimension, and calculated N/C ratio) were compared between cases and controls using the Wilcoxon rank sum test. Generalized estimating equations (GEE) were also
Volume 38, Number 7, July 2014
TABLE 2. Detailed Morphometric Analysis of the Tumor Cells and Control Group Using Calibrated Measurement Tools on Digital Images Mean (SD) Cases
Controls
1 2 3 4 5 6 7 1 2 3 4
Nuclear Size 8.62 8.49 9.51 6.35 6.54 8.08 10.51 8.36 7.89 7.53 7.41
(1.46) (1.50) (1.58) (1.26) (1.38) (1.91) (1.52) (1.35) (1.33) (1.44) (1.61)
Cytoplasmic Size 17.83 18.97 17.25 11.15 11.48 12.02 18.59 15.72 15.62 20.29 18.59
(4.25) (4.01) (3.15) (2.27) (2.08) (2.54) (3.96) (2.78) (2.80) (4.53) (2.87)
N/C Ratio 0.50 0.46 0.56 0.58 0.58 0.68 0.59 0.54 0.51 0.38 0.48
(0.11) (0.09) (0.11) (0.11) (0.11) (0.14) (0.12) (0.10) (0.08) (0.10) (0.10)
used to evaluate whether there was a difference in the diameters and N/C ratio between groups. The advantage of using GEE was that rather than computing a single mean for each patient, all the measurements from each patient were included in the analysis, and the correlation of the ratios within each patient was taken into account. Hence, this method was more appropriate for correlated data. Using a 2-sided test, a P-value 80 > 80
Met
Poorly differentiated adenocarcinoma Adenocarcinoma
78
Atelectasis and pleural effusion, left M Transthoracic needle core Chronic RLL infiltrate biopsy F Bronchoscopic biopsy Pneumonia in RUL
Prim
6
78
M Lobectomy
Mass in RLL
Prim
7
64
M Lobectomy
LLL consolidation
Prim
Prim
100
Poorly differentiated adenocarcinoma Poorly differentiated adenocarcinoma Mucinous adenocarcinoma
IHC Ker+ S100+ MART-1+ Ker+
> 80
Ker+
100
None
> 80
Ker+
100
Ker+
IHC, immunohistochemistry; Ker, keratin; MART-1, melanoma-associated antigen recognized by T cells; RLL, right lower lobe; RUL, right upper lobe.
922 | www.ajsp.com
r
2014 Lippincott Williams & Wilkins
Am J Surg Pathol
Volume 38, Number 7, July 2014
Lung Tumors Masquerading as DIP
TABLE 3. Statistics Based on the Mean Values From the Patients for Cases and Controls Groups Case (n = 7)
Control (n = 4)
Mean SD Median Mean SD Median Nuclear 8.30 1.50 Cytoplasm 15.33 3.59 N/C ratio 0.56 0.07
8.49 17.25 0.50
7.80 0.43 16.84 2.30 0.48 0.07
7.71 15.73 0.50
P*
Pw
0.3447 0.0278 0.7055 0.3669 0.0890 0.3441
*Wilcoxon rank sum test. wThe Wald test from GEE.
was identified in 27 (18%) of 150 lung adenocarcinoma samples, ranging in extent from 5% to 25%. Four of our patients had primary lung carcinoma (adenocarcinoma), whereas 3 had metastatic tumor from other sites (cases 1, 2, and 3). Immunostaining analyses were performed on 6 of 7 cases to confirm the diagnosis. The case (case 5) in which immunostaining analyses were not performed showed tumor with focal areas of glandular differentiation. The clinical, histopathologic, and immunohistochemical characteristics of the 7 patients are summarized in Table 1. The means and
SDs of nuclear diameter, cytoplasmic size, and N/C ratio for each patient are presented in Table 2. The summary statistics based on the mean values from the patients for cases and controls are presented in Table 3. Using the Wilcoxon rank sum test, no significant differences were seen in the nuclear or cytoplasmic diameters when comparing study cases with control groups (P = 0.3447 and 0.7055, respectively), and only a marginal significant difference in N/C ratios (P = 0.0890) was seen (Table 3). The group difference was also evaluated by applying GEE with normal link. These results showed a significant difference in the N/C ratio between the 2 groups, with study cases being on average 0.08 units (P = 0.0278) higher than controls. The diameters of nuclei and cytoplasm did not differ between groups (P = 0.3669 and 0.3441, respectively).
DISCUSSION Our small series of cases highlights an uncommon but real problem related to the diagnosis of lung tumors. When tumors in the lung grow mainly within the alveolar spaces, the process still should be diagnosable if the cytology of the tumor cells is definitively malignant. In
FIGURE 1. DIP-pattern. A, Pigmented “smokers-type” alveolar macrophages filling alveolar spaces. B, Pulmonary adenocarcinoma. C, Metastatic renal cell carcinoma. D, Metastatic melanoma. Hematoxylin and eosin stains. r
2014 Lippincott Williams & Wilkins
www.ajsp.com |
923
Raparia et al
some instances, however, the diagnosis can be challenging if the cytology of tumor cells is relatively bland. The problem is magnified in small samples in which DIP reaction of the tumor may be dominant. Both primary and metastatic tumors in the lung can show this phenomenon. The “DIP carcinomas” described in the 1995 AFIP fascicle illustrated this phenomenon in addition to areas of conventional adenocarcinoma with characteristic gland formation or mucin production.7 The radiology findings in these cases do not always show a discrete mass-like lesion, making the diagnosis more challenging. To better understand this histopathologic growth pattern, we evaluated biopsies from 6 patients with adenocarcinoma and 1 patient with metastatic melanoma who had a “DIP-growth pattern” and compared the observations with lung biopsies from 4 patients who had smoking-related DIP alone (Fig. 1). On the basis of hematoxylin and eosin examination, we encountered some difficulty in confirming the diagnosis of malignant tumor. With the hypothesis that malignant exfoliated intraalveolar cells should be measurably different from alveolar macrophages, we compared the mean and SD of nuclear diameter, cytoplasmic size, and N/C ratio for cases and controls, using morphometry. We were surprised to find that, even though the nuclear features of the malignant cases allowed recognition of a neoplastic process, using the Wilcoxon rank sum test, no significant difference in the diameters of individual cell nuclei and cytoplasm between cases and control groups was seen, and only a marginal significant difference in N/C ratios (P = 0.0890) was identified. We further evaluated the group difference by applying GEE with normal link, which did show a significant difference in N/C ratio between the cases and controls; with tumor cases on average having higher ratios compared with control by 0.08 units (P = 0.0278). The phenomenon we describe is not new. Mutton et al3 published a case of poorly differentiated lung adenocarcinoma, which was mistakenly diagnosed as DIP on open lung biopsy, and resulted in the death of the patient. The authors reviewed the case and compared it with 3 cases of DIP and found that the intra-alveolar cell populations in both DIP and adenocarcinoma have similar features on hematoxylin and eosin examination. However, they identified focal areas of lymphovascular invasion in the patient with adenocarcinoma, and the diagnosis was confirmed only on autopsy material. Yousem7 showed a similar growth pattern (alveolar filling) in cases of peripheral squamous cell carcinoma, wherein cohesive aggregates of tumor cells filled airspaces formed by intact nondisrupted alveolar septa. A recent study by
924 | www.ajsp.com
Am J Surg Pathol
Volume 38, Number 7, July 2014
Onozato et al8 identified a large collection of tumor cells present within alveolar spaces, which they defined as tumor islands in patients with lung adenocarcinoma, exhibiting a micropapillary and solid growth pattern. The authors mentioned that these tumor cells composing the islands were cytologically similar to those of the main tumor and could easily be differentiated from alveolar macrophages given a high N/C ratio and cytologic atypia.8 Although the mechanism of this “DIP-growth” of tumor is unknown, it is tempting to consider this growth pattern as a means of tumor spread in high-grade lung adenocarcinomas and other discohesive tumors.
CONCLUSIONS A “DIP-growth pattern” of malignant tumor growth in the lung is presented. This pattern can be seen in both primary and metastatic lung tumors, in our study adenocarcinomas and a melanoma, but in the literature also peripheral squamous cell carcinomas. Although measurement of the N/C ratio difference approached statistical significance, the key to the diagnosis was the individual cytology of the tumor cell nuclei. Immunohistochemical studies (keratin or other markers) are helpful in confirming the diagnosis in these tumors. We propose that the “DIP-growth pattern” might be regarded as an important morphologic feature of epithelioid tumors and that pathologists should be aware of this peculiar growth pattern in practice and always consider the radiologic findings before considering a diagnosis of an idiopathic interstitial pneumonia. REFERENCES 1. Leibow AA. Desquamative interstitial pneumonia. Am J Pathol. 1962;41:127. 2. Liebow AA, Steer A, Billingsley JG. Desquamative interstitial pneumonia. Am J Med. 1965;39:369–404. 3. Mutton AE, Hasleton PS, Curry A, et al. Differentiation of desquamative interstitial pneumonia (DIP) from pulmonary adenocarcinoma by immunocytochemistry. Histopathology. 1998;33:129–135. 4. Corrin B, Price AB. Electron microscopic studies in desquamative interstitial pneumonia associated with asbestos. Thorax. 1972;27:324–331. 5. Brewer DB, Heath D, Asquith P. Electron microscopy of desquamative interstitial pneumonia. J Pathol. 1969;97:317–323. 6. Colby TV, Koss MN, Travis WD. Tumors of the Lower Respiratory Tract. Washington, DC: Armed Forces Institute of Pathology; 1995. 7. Yousem SA. Peripheral squamous cell carcinoma of lung: patterns of growth with particular focus on airspace filling. Hum Pathol. 2009;40:861–867. 8. Onozato ML, Kovach AE, Yeap BY, et al. Tumor islands in resected early-stage lung adenocarcinomas are associated with unique clinicopathologic and molecular characteristics and worse prognosis. Am J Surg Pathol. 2013;37:287–294.
r
2014 Lippincott Williams & Wilkins
