Criteria
AJCC staging, 7th edition
Primary lung tumor with intrapulmonary metastases
Similar morphology and molecular characteristics
T3: Additional nodules in the same lobe are considered T3
T4: Additional tumors in the ipsilateral non-primary tumor-bearing lobe
M1a: Involvement of pleural fluid or contralateral lung
Synchronous, independent pulmonary tumors
Tumors with different morphology, immunoprofile, or molecular characteristics occurring at the same time
Stage each separately with parentheses to indicate number
Metachronous, independent pulmonary tumors
Tumors with different morphology, immunoprofile, or molecular characteristics occurring at a different time
Stage each separately with parentheses to indicate number
In some cases, morphological comparison of tumors in small biopsies with the prior primary tumor can be helpful in determining if the two have similar or divergent features (e.g., squamous cell carcinoma and adenocarcinoma). However, a morphological comparison between two adenocarcinomas can be difficult, given that the majority of primary lung adenocarcinomas have heterogeneous histology with a variety of different patterns (e.g., solid, papillary, lepidic, micropapillary, acinar) [11], and metastasis can have similar overlapping morphological features.
In morphologically ambiguous cases, immunohistochemical stains and/or molecular studies are helpful. Positivity for site-specific immunohistochemical markers is useful in establishing a pulmonary or non-pulmonary origin (Fig. 9.1). However, for two lung tumors that show similar morphology and immunophenotype, a comparison of molecular studies, which includes mutational profiles, loss of heterozygosity (LOH) profile, and clonality assays [12], is often the preferred approach to determine if they are synchronous, independent primaries.
Fig. 9.1
Adenocarcinoma of the lung with double immunostaining for TTF1 and NapsinA [(a) DQ stain, high power; (b) H&E stain, medium power; (c) TTF1 and NapsinA double stain]. The cytology touch preparation (a) shows cohesive clusters of cells with nuclear enlargement, atypia, and moderate cytoplasm, forming vague gland-like structures. The H&E stain (b) of the core biopsy shows an adenocarcinoma, which is highlighted in the image by TTF1 (brown, nuclear positivity) and napsinA (red, cytoplasmic positivity)
The emphasis on mutational status and clonality of tumors has become prevalent since more than two-thirds of second lung cancers (synchronous or metachronous) have similar histological features as the prior lung tumor [5, 13]. In studies focusing on mutational profiles of different lung tumors with similar histology, molecular differences that are genetically unique and independent from the patient’s prior tumor have been recognized [7]. However, even at the molecular level, establishing a relationship between two tumors can be difficult given that metastatic tumors may have a genetic profile that differs from the primary due to intratumoral heterogeneity or tumor progression with acquisition of additional genetic abnormalities [8, 14]. This is an important consideration particularly when molecular testing of the primary tumor is performed on cytology or small biopsy specimens, which may not be representative of the larger tumor mass.
In general, multiple tumors should be considered synchronous primaries if they have different histology or if there are morphological (e.g., associated in situ carcinoma), immunohistochemical, or molecular differences that support diagnosing two discrete primaries. In addition, if the tumors are located in different locations and temporally distinct without systemic metastases, then they may also be considered independent primary tumors. Thus, if deemed synchronous primaries based on clinical or pathological assessment, then each primary should be staged separately (Table 9.1). Intrapulmonary metastases tend to be tumors that have the same histology and occur in the same lobe without systemic metastases [5]. Given the importance of this determination for staging, meticulous description of the number of nodules, size of the nodules, and location of the nodules by the submitting clinician or radiologist that is procuring the small biopsy is crucial.
Pathological Evaluation of Small Specimens in Patients with Multiple Lung Tumors
Small biopsies and cytology specimens obtained with minimally invasive procedures are effective at providing sufficient tissue for an accurate diagnosis and ancillary studies with little risk to the patient [15–17]. In some instances, cytological specimens have an advantage over small biopsies, given that cytological specimens utilize multiple different preparations (e.g., smears, cell blocks, ThinPrep) and stains (e.g., Diff-Quik, Papanicolaou, hematoxylin and eosin). Well-prepared cytological specimens provide better morphology without the artifactual distortion seen with formalin-fixed small biopsies and are less dependent on immunostaining for the subclassification of non-small cell carcinomas, given that keratinizing squamous cell carcinomas are easily recognized by their orangeophilia on Papanicolaou staining, which is not utilized in the analysis of small biopsies [18, 19]. In addition, endobronchial biopsies are superficial and may miss a deep tumor, which is better sampled with a transbronchial FNA or Wang needle biopsy. This is especially relevant in cases of small cell carcinomas and other tumors causing large hilar masses that compress but do not involve the superficial bronchial wall.
However, small biopsies have the advantage of potentially providing greater tumor architecture for more accurate histological subtyping. This is important because identification of adenocarcinoma subtypes may have prognostic importance. In a recent study assessing the cytological subtyping of lung adenocarcinomas using the International Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society (IASLC/ATS/ERS) classification, it was shown that when assigning the dominant pattern subtype, there was only 40 % concordance between preoperative cytological samples and resections [20]. However, the low concordance may result from sampling. Also, formal studies comparing accuracy of dominant pattern classification based on small biopsy is lacking, and cytological patterns corresponding to known histological patterns have not been formally defined.
Reported adequacy rates of molecular studies for small specimens are variable as are preferences for cytology and/or small biopsy. For instance, one study described that approximately 15 % of tissue blocks had insufficient material for molecular studies after morphological and immunohistochemical stains are performed and an additional 15 % had insufficient or poor-quality nucleic acid for molecular studies, resulting in about 30 % of small biopsy and cytology specimens that are potentially inadequate for molecular studies [21]. Another study focused on molecular testing on cytology specimens reported that approximately 30 % of lung carcinomas with <300 tumor cells fail PCR-based studies for EGFR and KRAS mutations [22]. In contrast to these, successful molecular testing of EGFR, KRAS, or ALK was achieved in 93 % of all reflexively tested EBUS FNAs [23], and no statistically significant difference in results of ancillary studies between FNAs, core biopsies, or a combination of FNAs and core biopsies was noted on transthoracic CT-guided samples [24]. The variability among studies may be multifactorial.
The promising news is that molecular testing with whole-genome or targeted sequencing yields impressive results when detecting mutations on cytology specimens and small biopsies, which have less tissue and lower tumor cellularity than resections [15, 25]. Next-generation sequencing with targeted sequencing has enabled greater resolution, improved sensitivity, and lower cost compared to traditional sequencing methods and will likely have a significant role on small specimens in the future [26]. In addition, the combination of multiple sampling modalities, such as cytology specimens with small core biopsies, may help maximize the diagnostic yield of these procedures, and adequacy assessment during the procedure may help to ensure sufficient material. This underscores the importance of radiological-pathological correlation, immediate evaluation during the procedure (e.g., frozen section or on-site evaluation), and the use of multiple different sampling modalities in difficult or suboptimal cases.
For every case, sample management for ancillary studies is critical and begins with procuring sufficient material at the time of the procedure. Instituting an algorithm to cut blank slides upfront in the laboratory minimizes loss of tissue (Fig. 9.2). Now, to avoid the need for repeat biopsies for potential molecular studies, the trend is to conserve tissue by using focused and limited immunopanels in a staged approach guided by the clinical and radiological findings [27–29]. If there are atypical findings (e.g., TTF1 and p63 or p40 negativity), then additional immunostains are warranted to exclude a neuroendocrine tumor (e.g., synaptophysin), non-pulmonary adenocarcinoma (e.g., CDX2, PAX8, GATA3), or non-epithelial tumor.
Fig. 9.2
An example of an algorithm illustrating the triage of small biopsy and cytology materials for ancillary studies
Adenocarcinoma
Clinical Presentation
Cytology specimens and small biopsies are useful for diagnosing metastatic non-pulmonary adenocarcinomas within the lung or mediastinum, particularly when there are characteristic morphological or immunophenotypic findings. Clinically, an extrapulmonary primary should be considered in patients with a history of a prior malignancy, coincident mass lesion outside the lung, or radiological imaging showing multiple peripheral and/or bilateral lung nodules without a dominant lung mass. Nevertheless, a solitary mass can also represent a metastasis.
Some metastatic carcinomas can also present as endobronchial masses, particularly metastases from the breast, colon/rectum, kidney, and skin (e.g., melanoma) [30, 31]. By morphology, pulmonary adenocarcinomas can have morphological overlap with extrapulmonary adenocarcinomas and benign entities. Thus, a thorough clinical history is essential in distinguishing among these entities, in addition to performing focused immunostaining, conserving tissue in cases of atypical cytomorphological features or unexpected immunohistochemical staining results (e.g., TTF1 negative), and conducting potential molecular testing (Table 9.2).
Table 9.2
Differential diagnosis of pulmonary adenocarcinoma in small biopsies and cytology specimens
Neoplastic |
Poorly differentiated, nonkeratinizing squamous cell carcinoma |
Large cell neuroendocrine carcinoma |
Adenosquamous carcinoma |
Salivary gland-type tumors of the bronchial glands |
Non-epithelial tumors (e.g., sclerosing hemangioma, epithelioid sarcoma, angiosarcoma, melanoma) |
Extrapulmonary metastatic adenocarcinomas (e.g., breast, colon) |
Nonneoplastic |
Benign/reactive epithelial cells (e.g., bronchial cells, squamous cells, pneumocytes) |
Treatment-related atypia (e.g., chemotherapy, radiation) |
Granulomatous inflammation and epithelioid histiocytes |
Cytomorphological Features
Primary and secondary lung adenocarcinomas have overlapping cytomorphological features, and both possibilities have to be entertained, especially in the presence of signet-ring, mucinous, or papillary features (Fig. 9.3).
Fig. 9.3
Differential diagnoses for adenocarcinoma involving the lung
Adenocarcinomas with Signet-Ring Cells
Signet-ring morphology is well described in adenocarcinomas arising in the lung, stomach, and other gastrointestinal sites (Fig. 9.4). These tumors tend to be discohesive in cytological specimens and infiltrate as single cells or small nests on biopsies. With abundant cytoplasm, low nuclear-to-cytoplasmic ratios, and lack of hyperchromasia, signet-ring cells may be overlooked or dismissed as histiocytes on low magnification. Examination on high power demonstrates their subtle nuclear atypia and peripheral displacement and indentation of the nucleus by mucin, which can be highlighted by a mucicarmine stain.
Fig. 9.4
Signet-ring adenocarcinoma [(a) Pap stain, high power; (b) H&E stain, high power; (c) TTF1 stain, high power; (d) Pap stain, low power]. The slide (a) shows tumor cells with nuclear pleomorphism and eccentrically placed cytoplasm containing targetoid mucin droplets. The cell block (b) shows the same tumor cells with signet-ring type cells. The TTF1 stain (c) shows nuclear staining. Extrapulmonary sites (d) also have signet-ring cells [Photographs courtesy of Dr. Anjali Saqi (a–c) and Dr. John Crapanzano (d)]
Determining the site of origin in the setting of signet-ring cells requires immunohistochemical confirmation. Typically TTF1 and CK7 staining confirms a primary lung adenocarcinoma, while CK7 and CDX2 positivities are characteristic of the upper gastrointestinal tract and CK20 and CDX2 positivities of colorectal primaries (Table 9.3a). Other signet-ring mimics include plasma cell neoplasms (e.g., CD138) and malignant melanoma (e.g., S100, MelanA, HMB45).
Table 9.3a
Immunophenotype of primary and secondary adenocarcinomas with signet-ring cells
Primary | CK7 | CK20 | TTF1 | CDX2 |
|---|---|---|---|---|
Lung | + | − | + | − |
Upper GI tract | + | − | − | + |
Colorectal | − | + | − | + |
Mucinous (Colloid) Adenocarcinoma
Mucinous tumors arise in the lung, female genital tract, gastrointestinal tract, appendix, and pancreatobiliary region. Akin to signet-ring cell tumors, cells of mucinous adenocarcinomas may have low nuclear-to-cytoplasmic ratios due to their voluminous mucinous cytoplasm, but they are more cohesive and lack the indention of the nucleus in signet-ring cells. In cytology specimens, there is typically thick mucinous material in the background and a “drunken honeycomb” pattern. In adenocarcinomas with mild cytological atypia, these changes can be misinterpreted as goblet cell metaplasia, which often has interspersed ciliated bronchial cells and lacks nuclear atypia, or an aggregate of histiocytes, which lacks atypia. On surgical biopsies, neoplastic cells either line alveolar walls or are suspended within mucin pools. As in cytology specimens, mucinous adenocarcinoma may go unnoticed on a small biopsy (Fig. 9.5); however, the identification of large pools of mucin should raise the possibility of the diagnosis and a search for malignant cells. Metastatic colorectal adenocarcinomas may have columnar cells and mucinous features but more commonly have oval, palisading nuclei and a background of inflammatory necrosis (Fig. 9.6).
Fig. 9.5
Mucinous adenocarcinoma of the lung [(a) DQ stain, high power; (b) Pap stain, high power; (c) cell block H&E stain, low power, (d) core biopsy, low power]. The aspirate (a and b) shows clusters of tumor cells with moderate amounts of clear, vacuolated cytoplasm and round nuclei with mild pleomorphism. The cell block (b) and core biopsy (d) show similar tumor cells with mucinous cytoplasm and basally located nuclei [Photographs courtesy of Dr. Anjali Saqi]
Fig. 9.6
Metastatic colonic adenocarcinoma [(a) Pap stain, high power; (b) DQ, low power of touch imprint; (c) Cell block H&E stain, high power; (d) Core biopsy H&E stain, low power; (e) CDX2 immunostain]. The aspirates from adenocarcinomas of the colon typically show cohesive cells with prominent nucleoli and columnar morphology, with palisading around the edges (a). In addition, there is usually background necrosis (a) with inflammatory cells (B). The columnar morphology is usually more readily identified on the cell block sections (c) and core biopsy (d). CDX2 immunostain is useful in establishing the diagnosis (e) [Photographs (b) and (d) courtesy of Dr. Anjali Saqi]
The immunophenotype of mucinous adenocarcinomas can be ambiguous and staining has to be interpreted with caution in conjunction with the clinical history. CDX2 is a relatively specific marker generally associated with gastrointestinal or pancreatobiliary primaries. Nonetheless, primary mucinous adenocarcinomas arising in extra-gastrointestinal sites, such as the lung, ovary, and pancreas, can also express CDX2 and CK20 [32, 33]. Positivity for CK20 and negativity for TTF1 in a mucinous tumor may be seen in primary lung adenocarcinomas. However, villin and PAX8 are typically negative in these cases [34, 35]. Villin and CDX2 are typically positive in metastatic colorectal carcinomas, and PAX8 highlights ovarian mucinous tumors (Fig. 9.6). Making the distinction between synchronous and metachronous morphologically similar mucinous adenocarcinomas in the pancreas and lung may be impossible to separate either morphologically or by molecular tests as similar KRAS mutation have been described (Table 9.3b) [33].
Table 9.3b
Immunophenotype of primary and secondary mucinous adenocarcinomas
CK20 | CDX2 | VILLIN | PAX8 | |
|---|---|---|---|---|
Lung | + | +/− | − | − |
Colorectal | + | + | + | − |
Ovary | − | + | – | + |
Papillary Tumors
Papillary tumors typically arise from the lung, thyroid, kidney, mesothelium, or ovary. These tumors have fibrovascular cores and occasional psammomatous calcifications. Papillary tumors of the thyroid and lung are also associated with intranuclear inclusions, while ovarian serous papillary tumors typically have more nuclear pleomorphism (Fig. 9.7).
Fig. 9.7
Papillary adenocarcinoma of the lung [(a) DQ stain, high power; (b) H&E stain, high power; (c) TTF1 and NapsinA double stain]. Adenocarcinomas of the lung with a papillary pattern can have transgressing vessels associated with the tumor cells on the aspirates (a), and show fibrovascular cores on the cell block (b). TTF1 and NapsinA double immunostain is helpful to conserve tissue in small specimens and shows positivity for TTF1 (brown, nuclear positivity) and napsinA (red, cytoplasmic positivity) in this lung adenocarcinoma
A limited immunostain panel can resolve the differential of a papillary tumor (Table 9.3c). TTF1 is typically used to confirm lung origin but is ubiquitously expressed by papillary thyroid carcinomas. Performing additional markers, such as NapsinA for lung origin and thyroglobulin or PAX8 for thyroid origin, can discriminate between the two entities [36]. Also, in a peripheral lung mass or in a fluid cytology specimen, the possibility of a papillary epithelioid mesothelioma or papillary mesothelial hyperplasia should be considered. Given that mesothelial cells and lung adenocarcinomas both commonly express CK7 [37], TTF1 positivity can resolve the dilemma. In the absence of TTF1 staining, a panel of immunostains (e.g., for mesothelial origin: calretinin, CK5, WT1, and D2-40 and for adenocarcinoma: B72.3, BerEP4, and CEA] is recommended when lung adenocarcinoma and mesothelial origin are in consideration.
Table 9.3c
Immunophenotype of primary and secondary papillary tumors in the lung
Primary site | TTF1 | NapsinA | CK7 | Thyroglobulin | PAX8 | WT1 |
|---|---|---|---|---|---|---|
Lung | + | + | + | – | – | –/+ |
Thyroid | + | – | + | + | + | – |
Ovary | – | – | + | – | + | + |
Kidney | – | –/+ | +/– | – | + | – |
Mesothelium | – | – | + | – | – | + |
Additional Considerations
Likewise, other carcinomas metastatic to the lung may not have specific features, which make their identification difficult and underscores the relevance of clinical history. This is particularly true for metastatic breast carcinomas and prostate carcinomas that can present as single or multiple lung masses and have an acinar morphology or eosinophilic granular appearance (e.g., apocrine) similar to that seen in some lung carcinomas. Clear cells suggestive of conventional (clear) cell renal carcinoma are easily identified on biopsies and cell blocks, but the recognition of cells with abundant foamy cytoplasm and discrete punched-out vacuoles associated with thin blood vessels on other cytology preparations is more subtle (Fig. 9.8).
Fig. 9.8
Metastatic renal cell carcinoma [(a) DQ stain, high power; (b) cell block H&E stain, high power; (c) PAX8 immunostain]. Renal cell carcinomas have tumor cells with moderate-to-abundant cytoplasm that contains discrete “punched-out” vacuoles (a) on the air-dried Romanowsky-stained slides. These are more obvious on cell block sections and biopsies. The morphology and immunostaining of the cell block (b) show strong PAX8 positivity, supporting a diagnosis of conventional (clear cell) carcinoma of the kidney [Photographs courtesy of Dr. Anjali Saqi]
In the absence of TTF1 staining on the initial limited panel, additional immunostains, to exclude the possibility of a poorly differentiated squamous cell carcinoma, with p40, p63, or CK5 should be performed. When reactivity with these markers is also lacking, immunostains such as CK7 and CK20 [38] and site-specific markers need to be pursued to ascertain the primary origin (Tables 9.4 and 9.5, Fig. 9.9). Also important to consider is that TTF1 can be expressed rarely by non-pulmonary and non-thyroid carcinomas, including those arising from the ovary, endometrium, and endocervix, making interpretation in conjunction with clinical history essential [41, 42].
Table 9.4
Summary of CK7 and CK20 profile of primary and metastatic tumors
CK20 | |||
|---|---|---|---|
Positive | Negative | ||
CK7 | Positive | GI (upper) Ovary (mucinous) Pancreatobiliary Urothelial | Lunga Breast GI (upper) Gynecological tractb Kidneyc Pancreaticobiliary Salivary gland Thyroid Mesothelial |
Negative | Colorectal Merkel cell carcinoma Small cell carcinoma of salivary gland | Adrenal Kidney (clear cell) Liver (hepatocellular carcinoma) Prostate Squamous cell carcinomad Thymic carcinoma Non-epithelial tumorse | |
Table 9.5
Summary of commonly used markers for carcinoma of unknown primarya
Primary tumor site | Immunohistochemical stains |
|---|---|
Breast | Hormone receptors (ER, PR, AR) GATA3, mammaglobin, GCDFP15 |
Gastrointestinal | CDX2, CK20 |
Lung | TTF1, NapsinA |
Ovary | PAX8, WT1 |
Pancreatobiliary | CA19-9 |
Prostate | NKX3.1, PSAP, PSA |
Renal | PAX8, RCC, CD10 |
Thyroid | PAX8, thyroglobulin, TTF1 |
Urothelial | CK5, p63, GATA3 |
Fig. 9.9
Differential diagnosis based on select site-specific markers. (a): TTF1. (b) CDX2. (c) GATA3. (d) PAX8
Molecular Testing of Adenocarcinomas
Although there are some histological associations with particular molecular abnormalities, triage based on the histological pattern of lung adenocarcinoma identified is not recommended. Thus, regardless of histological pattern, it is recommended that at a minimum all adenocarcinomas of the lung are tested for EGFR mutations and ALK rearrangements [26, 43]. Allocation of material for potential molecular studies is important, particularly on small specimens. The recommendation is to preserve tissue for the testing of EGFR mutations and ALK rearrangements [26, 43–45], after an extrapulmonary adenocarcinoma has been excluded, for the diagnosis of lung adenocarcinoma, adenosquamous carcinoma, and non-small cell lung carcinoma not otherwise specified, or when the possibility of an adenocarcinoma cannot entirely be excluded. In many instances, clinical management of site-specific extrapulmonary adenocarcinomas is also dependent upon ancillary testing for clinical management, such as testing for Her2/neu in gastric adenocarcinomas [46] and BRAF and KRAS mutations in colorectal primaries.
Squamous Cell Carcinomas
Squamous cell carcinomas have different morphological subtypes, including keratinizing, nonkeratinizing, and basaloid, but the variants are not specific to a particular primary site, such that pulmonary and extrapulmonary squamous cell carcinomas can appear morphologically similar (Fig. 9.10). In addition, atypical squamous cells may represent a component of other entities, such as mucoepidermoid carcinoma or reactive processes, and resemble other neoplasms including urothelial carcinomas.
Fig. 9.10
Differential diagnoses for squamous cell carcinoma involving the lung
Well-Differentiated Squamous Cell Carcinoma
Squamous pearls, keratinization, and intercellular bridges are the sine qua non of squamous differentiation and are readily identifiable in well-differentiated carcinomas. When present, all of these features can be seen on core biopsies. Intercellular bridges, however, are not evident on cytology smears, only on cell blocks. Meanwhile, keratinization is easily recognized as orangeophilia (i.e., orange-staining cytoplasm) on Papanicolaou-stained cytology preparations but not on hematoxylin and eosin-stained core biopsies. Additional cytomorphological features favoring squamous cell carcinoma over adenocarcinoma include greater nuclear pleomorphism, coarser chromatin with inconspicuous nucleoli, and denser cytoplasm (Fig. 9.11).
Fig. 9.11
Squamous cell carcinoma [well-differentiated squamous cell carcinoma: (a) DQ stain, high power; (b) Pap stain, high power; (c) core biopsy H&E stain, high power; poorly differentiated squamous cell carcinoma: (d) Pap stain, high power; (e) cell block H&E, high power; (f) cell block p63, immunostain; basaloid squamous cell carcinoma: (g) DQ stain, high power; Pap stain, high power; (h) Pap stain, high power; (i) core biopsy H&E, high power.]. This example of well-differentiated keratinizing squamous cell carcinoma shows discohesive single cells (a) with dense cytoplasm showing cytoplasmic orangeophilia on the Pap-stained slide (b). The accompanying core biopsy shows tissue fragments with nests of an invasive squamous cell carcinoma (c). Poorly differentiated squamous cells arranged as a syncytial sheet (d) with the corresponding cell block (e) and p63 immunostain (f). Basaloid squamous cell carcinomas can mimic small cell carcinomas. The cells of basaloid squamous cell carcinomas are more cohesive and demonstrate peripheral palisading and crush artifact (g). The presence of occasional large cells is a clue to the diagnosis (h). Necrosis and apoptosis are associated with basaloid carcinomas (i) [Photographs courtesy of Dr. Anjali Saqi (a, d, e, f, g, h and i) and Dr. John Crapanzano (b)]
Poorly Differentiated Squamous Cell Carcinoma
Based on the above morphological criteria alone, nonkeratinizing, poorly differentiated squamous cell carcinomas can be difficult to differentiate from other carcinomas, including poorly differentiated adenocarcinomas and extrapulmonary metastases [47]. On FNAs, poorly differentiated adenocarcinomas and squamous cell carcinomas yield syncytial sheets or groups [48], whether they are arising from the lung or other sites, such as the head or neck (Fig. 9.11). Likewise, a core biopsy through an adenocarcinoma with a solid pattern can mimic squamous cell carcinoma.
Basaloid Squamous Cell Carcinoma
Squamous cell carcinomas with basaloid features are not unique to the lung and also arise in the head and neck. Squamous cell carcinomas with basaloid features also coincide morphologically with a subtype of pulmonary large cell carcinoma, basaloid carcinoma, which is devoid of morphological squamous differentiation. Necrosis, brisk mitotic activity, apoptosis, scant cytoplasm, inconspicuous-to-pinpoint nucleoli, and nuclear molding can be seen in the basaloid variant of squamous cell carcinoma [18, 19] and pulmonary large cell basaloid carcinoma. These features can be misinterpreted as small cell carcinoma on cytology and small biopsy [49] (Fig. 9.11). Palisading of nuclei at the edge of cell clusters, the lack of prominent nuclear molding, and tightly cohesive clusters favor basaloid features over small cell carcinoma [49].
Characteristically squamous cell carcinomas are positive for p63, p40, and/or CK5/6. The sensitivity and specificity for detecting a squamous cell carcinoma is higher for p40 than for p63, which make p40 the marker of choice [50, 51]. In contrast to lung adenocarcinomas, squamous cell carcinomas do not express a site-specific marker like TTF1, which makes differentiating a primary from a metastatic squamous cell carcinomas difficult.
However, it is now well established that HPV-related squamous cell carcinomas of the head and neck, cervix, and anorectal area are positive for p16 immunostain and HPV, which can be detected by chromogenic in situ hybridization (CISH) and other methods (Fig. 9.12) [52]. Since HPV is not linked to squamous cell carcinomas of the lung, its identification supports a HPV-related metastatic squamous cell carcinoma [53]. However, some squamous cell papillomas, which are typically endobronchial and can have parenchymal involvement, are HPV(+) [54]. Positivity for p16 is defined as strong and diffuse staining of the cytoplasm and nucleus in over 70 % of the tumor cells [55]. For HPV CISH performed on formalin-fixed paraffin-embedded tissue, a dot-like pattern confers positivity but requires meticulous high-power examination to prevent a false negative interpretation [55]. The reactivity of these two markers, p16 and HPV, in a new lung squamous cell carcinoma is useful in establishing metastatic disease in the setting of a known HPV-related carcinoma or identifying an occult squamous cell carcinoma, such as a tonsillar or base of tongue primary [56]. Conversely, the absence of p16 staining and HPV identification does not exclude the possibility of a metastasis, since many squamous cell carcinomas arise secondary to non-HPV etiologies.
