Epithelial neoplasms are by far the most common primary malignancies in the lung. Worldwide, lung cancer is the leading cause of illness and death in patients with neoplastic disease as reflected in morbidity and mortality rates for this population. Environmental hazards, including tobacco use, have been conclusively linked to the development of pulmonary carcinoma. Recently, several important features have been observed in the epidemiology of lung cancer, including an increased frequency among African-American men compared with white men, and an increased frequency of squamous cell carcinoma among white women compared with African-American women. Histopathologically, adenocarcinoma appears to be the most common lung cancer diagnosis; squamous cell carcinoma appears to be more commonly associated with tobacco use. In 2007, Wahbah and associates demonstrated that adenocarcinoma diagnoses have become far more frequent over the past 3 decades: In 1980, adenocarcinomas in both men and women accounted for approximately 29% of the cases, and squamous cell carcinoma accounted for approximately 39%. By 2003, adenocarcinomas had increased to represent 40% of cases, whereas squamous cell carcinomas had decreased to 29%. In a Japanese study from 1958 to 1997, the investigators determined that the average age at death from lung cancer is 71.6 years for males and 73 years for females. Regardless of the histologic type and ethnic group affected, the survival rate after 5 years is poor; as better techniques for early detection are developed, however, survival rates may be expected to improve.
Non–small cell carcinomas of the lung are by far the most common malignant tumors, usually appearing in the sixth and seventh decades of life. The study by Wahbah and colleagues reported the average age at diagnosis for patients with lung carcinoma to be 59 years for both men and women. The patient’s symptomatology will depend largely on the anatomic location and the size of the tumor. Tumors that are centrally located are more likely to manifest earlier in the clinical course, typically with signs and symptoms of pulmonary obstruction such as cough, dyspnea, wheezing, hemoptysis, and pneumonia. Tumors that are located in the periphery of the lung will not cause symptoms until they reach a relatively large size. Some clinical signs and symptoms may be correlated with a particular type of tumor—for example, bronchorrhea (expectoration of large amounts of mucus) most commonly is seen in bronchioloalveolar carcinoma (BAC). Depending on the extent of the tumor within the thorax, clinical manifestations may include pleuritic pain, the Pancoast syndrome, or superior vena cava syndrome. Paraneoplastic syndromes such as inappropriate secretion of antidiuretic hormone, Cushing’s syndrome, or acromegaly also may be associated with some non–small cell carcinomas of the lung. Infectious processes such as tuberculosis and other primary pulmonary conditions such as pulmonary fibrosis and bronchiectasis also have been reported in association with these tumors.
The histopathologic classification of non–small cell carcinomas of the lung can be extensive. Table 3-1 presents a histopathologic classification based on the different growth patterns that have been described for these tumors. For some of these growth patterns, a rather specific clinical behavior has been recognized. An important point to keep in mind, however, is that the staging of non–small cell carcinomas is paramount in assessing outcome. Therefore, although it is important to properly classify a tumor from its histopathologic features, it also is crucial to demonstrate all of the features necessary for accurate staging of the tumor.
|Squamous cell carcinoma|
|Large cell carcinoma|
SQUAMOUS CELL CARCINOMA
Squamous cell carcinoma is currently less common than adenocarcinoma; however, squamous cell carcinoma is more commonly associated with tobacco use. Unlike adenocarcinoma, squamous cell carcinoma may show a progressive disease pattern, ranging from mild dysplasia (atypia) to carcinoma in situ to invasive squamous cell carcinoma.
After diagnosis of dysplasia or squamous cell carcinoma in situ of the lung, in most cases a clearly identifiable tumor mass is absent. The great majority of patients come to medical attention because of a history of tobacco use or human papillomavirus infection. Some patients may provide a history of cough and respiratory difficulty.
Squamous Cell Dysplasia/Carcinoma In Situ
Although no foolproof mechanism has been found to distinguish between mild and moderate dysplasia on histopathologic grounds, and because such distinction may be rather arbitrary, a reasonable approach is to follow criteria similar to those used for other organ systems, such as the uterine cervix. In many cases, however, the diagnosis of high-grade dysplasia is synonymous with that of carcinoma in situ.
The diagnosis of mild dysplasia is made when cellular atypia is observed at, or just above, the basal cell layer. The nuclei may be hyperchromatic, and mitotic figures may be observed. No mitotic activity or cellular atypia is found in the midportion of the mucosa, which may show normal maturation.
For a diagnosis of moderate dysplasia, the findings are similar to those of mild dysplasia; however, the cellular atypia and mitotic activity are not limited to the lower portion of the mucosa. In addition, the cellular atypia may be more pronounced and mitotic figures may be seen in the midportion of the mucosa. Normal maturation may still be identified in the upper portion of the mucosa, however.
Severe Dysplasia/Carcinoma In Situ
Severe dysplasia and carcinoma in situ are intrinsically associated conditions and represent the same pathologic process. Histologically, the mucosa shows full-thickness involvement: The cellular changes and mitotic figures occupy the entire mucosa; maturation is essentially absent, atypical mitotic figures may be seen at any level, and cellular atypia is much more pronounced ( Fig. 3-1A ). In some cases, the process may extend to the endobronchial glands; such extension, however, does not imply that the process is invasive ( Fig. 3-1B ). Careful evaluation is needed in cases with marked chronic inflammation, which may either obscure the true nature of the process or mimic an invasive neoplasm. In some unusual cases of in situ squamous cell carcinoma, changes may be observed in the epithelium that are histologically similar to those caused by Paget’s disease ( Fig. 3-1C ). Although this pagetoid change may be extensive, it is still possible to find areas of more conventional squamous cell differentiation.
Invasive Squamous Cell Carcinoma
Traditionally, squamous cell carcinomas have been classified as well-, moderately, or poorly differentiated. Most of the factors influencing this classification involve the presence of either keratinization or intercellular bridges. Therefore, the closer a tumor mimics normal squamous epithelium, the better the grade of differentiation will be.
As with other non–small cell carcinomas, patients may complain of cough, dyspnea, hemoptysis, or thoracic pain. Presence of such signs and symptoms generally reflects the anatomic location of the tumor. Tumors that are centrally located are more likely to produce signs and symptoms associated with obstruction, whereas tumors that are in the periphery of the lung are more likely to produce thoracic pain, owing to the large size they may attain.
Squamous cell carcinomas may be seen as polypoid tumors obstructing the lumen of the airway ( Fig. 3-2 ). As they reach a larger size, they may be seen extrinsically pushing into airway structures. Squamous cell carcinomas may be observed in the periphery of the lung, in a subpleural location ( Fig. 3-3 ), or infiltrating the pleura with direct invasion into the soft tissues of the chest wall.
Tumor size may range from smaller than 1 cm to larger than 10 cm in greatest dimension. These tumors are white-tan and well demarcated, but not encapsulated. The cut surface may be homogeneous or show extensive areas of necrosis or hemorrhage. In some instances the tumor takes the form of a cavitary tumor mass ( Fig. 3-4 ).
Well-Differentiated Squamous Cell Carcinoma
Well-differentiated tumors show areas of keratinization and easily identifiable intercellular bridges ( Fig. 3-5 ). The tumor may grow in sheets or ribbons, destroying the normal lung parenchyma. The neoplastic cells are oval, with moderate amounts of eosinophilic cytoplasm, small nuclei, and inconspicuous nucleoli. Mitotic figures are present and can be numerous. Areas of necrosis or hemorrhage usually are absent or may be only focally present.
Moderately Differentiated Squamous Cell Carcinoma
Moderately differentiated tumors may still show areas in which either keratinization or intercellular bridges can be identified. However, more extensive areas of hemorrhage and necrosis, which may obscure the true nature of the neoplasm, also may be observed. Mitotic figures and cellular pleomorphism are more prominent ( Fig. 3-6 ).
Poorly Differentiated Squamous Cell Carcinoma
Poorly differentiated squamous cell carcinomas have a tendency to grow in sheets, in which definitive evidence of squamous differentiation may be lacking. Extensive necrosis and hemorrhage may be present. The presence of cellular pleomorphism and mitotic activity is marked. In focal areas, however, it may still be possible to find unequivocal areas showing squamous cell differentiation ( Fig. 3-7 ).
Squamous cell carcinomas show a variety of different growth patterns, which are summarized in Table 3-2 and illustrated throughout the chapter ( Figs. 3-8 to 3-16 ). Some of the more important variants that have been presented in the literature are described next.
The unusual exophytic variant of squamous cell carcinoma also is known as Sherwin tumor in honor of one of the first investigators to assess this particular growth pattern. Sherwin and colleagues documented the existence of this neoplasm after reviewing the cases of 85 patients diagnosed with squamous cell carcinoma. These workers isolated nine patients between 50 and 74 years of age in whom the tumor resembled an exophytic neoplasm. These patients all had undergone resection of the tumor; those who were treated with pneumonectomy had a more favorable outcome than those who were treated with lobectomy. Histologically, the tumors grew in an exophytic manner, obstructing the airway. They were of the well-differentiated grade, and none of the patients had metastasis to lymph nodes. The investigators noted that this type of squamous cell carcinoma may potentially have a more favorable outcome than other, more infiltrative variants.
Dulmet-Brender and associates also presented their experience with 34 cases collected over 35 years and concluded that these tumors are almost always of the T1N0 stage at presentation. In their experience, however, the prognosis was not any better than it would be for another non–small cell carcinoma at the same stage. These workers concluded that these tumors were not a more common squamous cell carcinoma detected at an earlier stage, but rather that they represented a distinct variant.
More recently, Cooper and associates documented a similar occurrence in a 75-year-old woman. The initial chest radiographic appearance was unremarkable, but an endobronchial tumor became evident on computed tomography (CT) scan. On resection of the mass, a papillary endobronchial neoplasm that did not extend into the adjacent lung parenchyma was found.
Histologically, all of these lesions appear to have the same characteristics: an exophytic tumor with squamous appearance, in some cases with extensive keratinization. In some areas, the tumor may acquire a papillary growth pattern in which the papillary projections show only a thin line of fibroconnective tissue, with no invasion. The most important histologic characteristic is the limited behavior of the tumor; it extends exophytically into the lumen without invasion into the bronchial wall. Rarely, the endobronchial glands also may show extension, but bronchial wall invasion should not be observed. The papillary fronds show cellular atypia, nuclear pleomorphism, and mitotic activity with some atypical mitotic figures.
Although the diagnosis of exophytic squamous cell carcinoma should not pose a problem from a histologic standpoint, some other lesions that manifest in similar fashion may be considerations in the differential diagnosis. In 1969, Laubscher described a tumor referred to as “solitary squamous cell papilloma.” Histologically, the tumor is a keratinizing, noninvasive squamous cell neoplasm that, based on the illustrations provided, shares some features with the cases described by Sherwin and colleagues. The existence of noninvasive papillary neoplastic disease, such as solitary condylomatous papilloma and lower respiratory tract papillomatosis, is well recognized in the literature. These lesions will show squamous epithelium with some evidence of maturation remaining and in focal areas may demonstrate areas of transition between squamous and nonsquamous epithelium.
The cystic variant usually is recognized from the gross appearance of the tumor. It is characterized by prominent cystically dilated spaces, which may be filled with necrotic or inflammatory debris. In the periphery of the tumor, strands of squamous cell carcinoma, which may resemble buds or ribbons, are typical features. Usually, these tumors are well differentiated and have undergone cystic and necrotic changes (see Fig. 3-8 ).
The small cell growth pattern is unusual and represents a very small percentage of squamous cell carcinomas. Currently, no single series of cases has been presented in the literature; most reports are of single anecdotal cases described in textbooks. Histologically, the tumor shows a proliferation of small cells with oval nuclei and inconspicuous nucleoli. Unlike in true small cell carcinomas, the chromatin displayed in the small cell variant of squamous cell carcinoma is not the typical “salt and pepper” type; rather, a smoother type of scant cytoplasm is observed (see Fig. 3-9 ). In cases in which a transbronchial biopsy has been performed, the finding of squamous cell carcinoma in situ should raise clinical suspicion for the small cell variant. When squamous cell carcinoma in situ is not evident, it is important to identify areas in which the tumor may show focal keratinization or other clues to squamous cell differentiation. In more complicated cases, the use of immunohistochemical studies may be helpful. The small cell variant of squamous cell carcinoma may show positive staining for keratin 5/6 and p63, with negative staining for neuroendocrine markers; however, some squamous cell carcinomas also may show focal positive staining for some of the neuroendocrine markers—namely, synaptophysin.
The spindle cell growth pattern also is rather unusual and can be confused with other spindle cell neoplasms, especially sarcomas. It is characterized by fascicles of fusiform cells with elongated nuclei and inconspicuous nucleoli, increased mitotic figures, and atypical mitotic figures (see Fig. 3-10 ). The tumor may show interlacing fascicles of cells mimicking sarcomas. In focal areas, however, keratin pearls, single cell keratinization, or frank focus of keratinization may be observed. Currently, the use of immunohistochemical studies including immunostaining for epithelial markers such as broad-spectrum keratin, low-molecular-weight keratin, keratin 5/6, or p63 may help in distinguishing these tumors from primary pulmonary sarcomas. Electron microscopic studies also may be useful, because spindle cell carcinomas may manifest epithelial differentiation with the presence of intercellular junctions and tonofilaments.
The basaloid growth pattern is an unusual variant of squamous cell carcinoma and resembles that seen in basal cell carcinomas of the skin. The tumor cells are arranged in islands, with palisading of the nuclei (see Fig. 3-11 ). Mitotic figures are numerous, and nuclear atypia is prominent. The cells may acquire a fusiform appearance, with elongated nuclei and inconspicuous nucleoli. In some cases, areas of keratinizatin are present, whereas in others, the squamous cell differentiation may be subtler. Because this basaloid pattern may be present in other tumors, such as neuroendocrine carcinomas or the so-called basaloid carcinomas of the lung, it is important to ensure that the tumor under investigation represents a true basaloid squamous cell carcinoma.
Immunohistochemical and Molecular Features
Currently, the literature describes numerous immunohistochemical markers that may have the potential to distinguish squamous cell carcinomas from other types of non–small cell carcinomas, including keratin 5/6, p63, and thyroid transcription factor-1 (TTF-1). Immunostaining for these markers is particularly useful in cases in which the patient has a history of squamous cell carcinoma of extrathoracic origin, and the lung biopsy seems to show another non–small cell carcinoma. Also, when a patient with a previous carcinoma of the lung presents with a different pulmonary neoplasm, these markers may be used to determine whether the tumor is a recurrence or a separate primary lung neoplasm. Other immunohistochemical markers have shown positive staining in squamous cell carcinomas, including CD117; however, the significance of the positive staining is not yet clear.
Yoshino performed molecular studies in 22 cases of squamous cell carcinomas of the lung and found that loss of heterozygosity (LOH) was more frequently observed in squamous cell carcinomas than in adenocarcinomas of the lung, and that LOH tends to be associated with tobacco use. In a comparative genomic hybridization analysis, Chujo and coworkers encountered over-representation of the long arm of chromosome 3 (3q) in squamous cell carcinomas of the lung and concluded that an increased copy number at 3q may contribute to the development of squamous cell carcinoma of the lung.
Many controversial issues associated with the diagnosis of bronchioloalveolar carcinoma (BAC) have emerged. Multiple definitions have been applied to this tumor over the years, and the interpretation of these definitions varies from individual to individual. The currently accepted definition of BAC is considerably different from those of previous decades. This definition is so restrictive that BAC can now be considered to be adenocarcinoma in situ. Many years ago, the descriptions of BAC were much broader; in fact, most of the series presented in the literature on BAC have described tumors that had metastasized, or that had involved other thoracic structures.
In 1876, Malassez was the first to describe a primary pulmonary neoplasm with an unusual alveolar pattern of growth. In Malassez’s description, the tumor appeared to be multinodular and therefore was named “Cancer Encephaloide of the Lung (Epithelioma).” Twenty-seven years later, Musser described a similar tumor under the designation “Primary Cancer of the Lung.” Both of these descriptions are essentially similar in terms of histopathology; however, the gross features were quite different. Malassez described a multinodular tumor replacing lung parenchyma, whereas Musser described a diffuse process, similar to an infectious process such as bronchopneumonia. These two representations demonstrate what are currently acknowledged to be the two main growth patterns of BAC. The presence of metastatic disease was noted in both of these descriptions.
During the first half of the 20th century, a sizable number of cases with histopathologic and macroscopic characteristics similar to those described by Malassez and Musser were presented in the literature. The gross features of these cases were either multinodular or diffuse; histopathologically, the tumors showed the alveolar lining replaced either by columnar or low cuboidal cellular proliferation. Numerous terms were used to describe this neoplasm, including alveolar cell cancer, primary multiple carcinoma, multiple nodular carcinoma, diffuse lung carcinoma, alveolar carcinoma, carcinosis, carcinomatoides alveogenica multicentrica, alveolar cell tumor, and pulmonary alveolar adenomatosis.
Among these early investigators, the main point of debate centered on the existence of alveolar epithelium. Those who denied the existence of alveolar epithelium did not believe that epithelial neoplasm could arise from alveolar structures, whereas those who supported the existence of alveolar epithelium were ready to classify these neoplasms as alveolar carcinomas. Nevertheless, the concept of a tumor growing along alveolar structures remained. Almost all of the cases described during the first half of the 20th century involved tumors that extended beyond the lung parenchyma. In a majority of cases, tumors seeding the pleura or peribronchial lymph nodes, or tumors outside of the thoracic cavity, were documented. Despite the discussion surrounding alveolar epithelium, some researchers suggested a possible infectious origin for these tumors, likening them to similar tumors seen in sheep (the so-called Jaagsiekte). As a result of the lack of consensus on whether these tumors originated from alveolar epithelium or from finer terminal bronchioles, the term bronchioloalveolar carcinoma was established.
In 1960, Liebow not only coined the term bronchioloalveolar carcinoma but also provided a definition for the diagnosis of this neoplasm. Liebow defined BAC as a well-differentiated adenocarcinoma and distinguished it from “ordinary” adenocarcinoma on the basis of lack of evidence regarding the cell of origin. In addition, Liebow identified three main forms of this neoplasm: single nodular, disseminated nodular, and diffuse. He also noted that the tumor is capable of invading lymph nodes, pleura, and extrathoracic organs—a feature that is observed in greater than 50% of autopsy cases. In Liebow’s opinion, the tumors may have a long dormancy or slow growth, particularly those presenting as isolated nodules. He also pointed out that 50% of patients present with bilateral disease. In 1980, the Armed Forces Institute of Pathology defined BAC as “a lesion with relatively bland cytologic features that arises in the periphery of the lung and spreads on the walls of the distal air spaces.” In the second published series of the Armed Forces Institute of Pathology (1995), the definition is that of a subset of adenocarcinoma, common and distinctive enough to warrant separation from the other subtypes. In the two most recent publications of the World Health Organization (WHO), the tumor is defined as an adenocarcinoma with bronchioloalveolar pattern and no evidence of stromal, vascular, or pleural invasion. On the basis of these publications, it is apparent that the definition of BAC has changed to the point at which diagnosis is possible only by complete examination of the entire tumor in question, and not on biopsy material.
Analysis of the Literature
The literature on bronchioloalveolar carcinoma is filled with controversy, and many of the studies that address the controversial issues have only generated more debate. In 1955, Overholt and colleagues reported a study of 15 patients treated surgically, 11 of whom had no evidence of metastasis at follow-up. Of note, however, these workers also documented the development of metastatic disease to the brain in one patient and ipsilateral disease in another two patients. Belgrad and Munnel and their coworkers also made a similar claim in a study of patients with localized disease. According to both groups of investigators, surgery may have eradicated these patients’ tumors. However, Belgrad and colleagues also acknowledged that for tumors characterized by diffuse involvement of a lobe, the prognosis may not be as good. On the other hand, Watson and Farpour studied 265 patients in whom the clinical behavior of the tumor was, in some cases, more aggressive. In their series, 82 patients showed metastatic disease on autopsy, and only 16 of 82 patients who were treated with excisional surgery survived more than 5 years. Other researchers have questioned the validity of the BAC diagnosis, arguing that BAC represents a pattern rather than a specific entity, because many tumors of extrathoracic origin can metastasize to the lung in a manner indistinguishable from that characteristic of pulmonary BAC. In a study of 30 cases of BAC, Bennett and Sasser concluded that there is no morphologic, histogenetic, or clinical basis to separate “bronchioloalveolar carcinoma” from adenocarcinoma of the lung. Contrary to that opinion, Delarue and colleagues presented their views in a reappraisal of bronchioloalveolar carcinoma, and insisted on considering BAC a specific clinicopathologic entity. These workers’ criteria for the diagnosis of BAC were as follows: (1) absence of primary adenocarcinoma elsewhere; (2) absence of intrinsic tumor of bronchogenic origin; (3) peripheral location involving alveolar ducts and sacs; and (4) unaffected interstitium. However, they noted that metastatic adenopathies and malignant pleural effusions may occur. The overall survival rate for the 74 patients studied was 34% at 3 years. Marco and Galy presented their experience with BAC in 29 patients, separated into three main groups depending on the extent of their disease. Group 3 was composed of patients with pleural and lymph node involvement.
Over the years, some researchers have focused on differentiating BAC from conventional adenocarcinoma, whereas others have attempted to further divide BAC into subcategories. Singh and associates have argued that in addition to the peripheral lesion recognized as BAC, two other subtypes also may occur: the Clara cell and the type II pneumocyte subtypes. Other investigators have supported this opinion and have stated that BAC and peripheral bronchogenic adenocarcinoma are derived from secretory cells that resemble bronchiolar Clara cells, thereby separating them from conventional bronchogenic adenocarcinoma. Some workers also have attempted to correlate histopathologic features of BAC with survival. In a study of 34 cases in which BAC was separated into type 1 and type 2 tumors, Manning and coworkers stated that type 1 BAC is associated with mucus production and is most likely to be multicentric, whereas type 2 BAC does not show much mucus production and is most likely to be solitary. These investigators observed a 5-year survival rate of 72% for patients with the nonmucinous type (type 2), whereas patients with the mucinous type (type 1) had a 25% 5-year survival rate. Clayton arrived at a somewhat different conclusion in a study of 45 cases of BAC, finding that aerogenous spread had occurred in 24 of 36 nonmucinous tumors, and that at 5 years, these patients either had died from their disease or were alive with metastasis, whereas in 12 nonmucinous tumors without aerogenous spread, the 5-year survival rate was 61%. Clayton also found that smaller tumors carried a better prognosis, and that the presence of alveolar spread, rather than cell type, was the most important feature to predict prognosis.
In a 21-year retrospective study, Thomas and coworkers concluded that BAC is a valid term that represents a heterogeneous population of tumors; they argued that BAC should be retained as a term describing a growth pattern. In their experience, BAC carries a bad prognosis, which may be attributed to the fact that many patients are asymptomatic until the disease is advanced and inoperable. In a large study between 1968 and 1986, Elson and associates accumulated 193 cases of BAC, of which only 39 were selected as pure BAC. The type of material available for review in this study was unclear, however, and it is likely that the specimens were mixed among cytology smears, biopsy, and surgical resection. Thus, it is difficult to determine the validity of the designation “pure BAC.”
In 1991, Daly and colleagues of the Mayo Clinic presented a study of 134 patients with BAC. Ten of the patients described (7.5%) had lymph node metastasis, and although the great majority were stage I, several cases in stages II, IIIA, and IIIB also were included. The survival rate for patients who were T1N0M0 at 5 years was 90%, in contrast with patients who were T2, N0, M0, for whom the survival rate was 55%. These workers concluded that BAC has a unique natural history, which is more influenced by local neoplastic process than by lymph node metastases. From the same institution, Feldman and associates presented a study of 25 patients with metastatic BAC in which the response to chemotherapy was compared with that in patients with conventional adenocarcinoma. These investigators concluded that the chemotherapeutic response and the median rates of disease progression were similar in both groups and further stated that metastatic BAC is an aggressive disease that is associated with poor prognosis, similar to metastatic conventional adenocarcinoma of the lung. A similar experience was reported by Breathnach and colleagues in their study of 28 cases of BAC in stages IIIB and IV. In this particular report, patients with advanced BAC were more likely to have bilateral disease but were at lower risk for development of brain metastasis. In another study from Taiwan, the investigators collected 50 cases of BAC, including patients in different stages of disease. These workers concluded that BAC frequently manifests with lymphatic spread or systemic metastasis at diagnosis, and that in most cases patients with localized BAC fare better than those with the diffuse type of BAC. Also, Fujimoto and associates presented a study of 53 cases of BAC and found no difference in survival between patients who underwent resection for BAC and those who were operated on for non-BAC. Of note, however, these investigators found that patients with stage IV BAC had a better response to chemotherapy than that observed in patients with non-BAC tumors.
BAC has been perceived as a tumor of increasing frequency in younger persons, females, and nonsmokers, with a reported incidence as high as 14%. Read and coworkers presented an epidemiologic study of BAC over 2 decades (1979 to 1998) based on analysis of the SEER database and concluded that despite the apparent increase in BAC, this tumor represents less than 4% of all primary non–small cell carcinomas of the lung. In a different analysis of SEER database, Raz found that BAC is not associated with younger mean age at diagnosis and that the tumor is not associated with an age younger than 50 years at diagnosis. As suggested by some investigators, BAC may have an environmental etiology other than tobacco use. In addition, BAC in children is believed to carry a better prognosis, although the tumor is rare in the pediatric age group.
More recent literature on BAC, using the latest definition by the WHO, also has provided conflicting information. In a study focusing on the survival and recurrence of BAC stage I disease, Breathnach and coworkers found a 5-year survival rate for patients with BAC of 83%, as opposed to 63% for those with other types of adenocarcinoma. Rena and associates described similar findings in a study of 28 patients with stage I pure BAC, reporting a 5-year disease-free rate of 81% and a long-term survival rate of 86%, as opposed to the 5-year disease-free survival rate of 51% and long-term survival rate of 71% for conventional adenocarcinoma. Although these investigators claim that the WHO criteria were followed in these cases, 20 patients were diagnosed using fine needle aspiration biopsy, which is not permitted under the WHO’s current criteria. In a study of 20 cases of BAC, Gaeta and coworkers focused on pattern of recurrence after surgical resection, concluding that diffuse BAC may develop from previous focal carcinoma, and that the mucinous type is the one most likely to become diffuse. In this study, the investigators included three different types of BAC: mucinous, nonmucinous, and mixed adenocarcinomas with prominent bronchioloalveolar pattern. A study by Ebright and associates, however, concluded that the most important predictors of survival in BAC are clinical pattern and pathologic stage, rather than degree of invasion as seen on histologic examination.
In 2006, Travis and colleagues investigated the relevance of the WHO histopathologic criteria for the diagnosis of BAC. According to these workers, the existing evidence indicates that patients with solitary, small, peripheral BACs have a 100% survival at 5 years. They argue that the basis for the newly proposed classification of BAC is derived from a study by Noguchi and associates on the histologic characteristics and prognosis of small adenocarcinomas of the lung. In this study, the investigators reported 236 cases of resected peripheral adenocarcinomas of the lung in which the lesions were no more than 2 cm in greatest diameter. They separated the cases into different types designated A to F, and the ones belonging to types A and B were coded as localized BAC. Foci of structural collapse of alveoli were present in the type B cases but not the type A cases. The cases in the A and B groups together amounted to 28 (14 cases of each category). Closer analysis of the data showed that although general information is provided for the 236 case studies, the investigators did not provide a specific tumor size for types A and B, instead stating that “they are usually larger than 1 cm”—thus leaving open the possibility that some of these lesions may have been less than 1 cm. In addition, even though the investigators claimed that no lymph node metastasis was present in any of the 28 cases of localized BAC, they also stated that 3 of 34 (not the 28 cases specified) showed pleural involvement whereas 2 of 34 (again, not the initial 28 cases) showed vascular involvement.
Travis and colleagues also cite a study conducted by Zell and coworkers as evidence of the clinical impact of the WHO criteria on the diagnosis of BAC. In this study, Zell and coworkers presented a retrospective analysis of data from the population-based Cancer Surveillance Programs of three Southern California counties from 1995 to 2003, analyzing cases diagnosed as BAC before and after May 1999, when the WHO published the new criteria for the diagnosis of BAC. The investigators found that the overall survival period for patients with BAC diagnosed after 1999 was 53 months; before May 1999, it had been 32 months. At time of presentation, 48% of the patients had localized disease, 26% had regional spread, and 24% had metastatic disease. Furthermore, patients with BAC were found to have a “significantly” prolonged median overall survival (42 months), with 1-year survival rate of 69%, 2-year survival rate of 58%, and 5-year survival rate of 41%. When patients were stratified by extent of disease, the overall median survival rate for patients with localized disease was 98 months. The investigators also found that the incidence of BAC had increased from 5% before 1999 to 5.5% after 1999, raising the possibility that despite the WHO’s restrictive new criteria, the incidence of BAC was actually increasing.
Zell and coworkers also found that before the latest version of the WHO criteria, from January 1995 to May 1999, the 1-year survival rate was 66.5% and the 2-year survival rate was 54%. From June 1999 to December 2003, however, the 1-year survival rate was 72.5% and the 2-year survival rate was 63.3%. It is this 6% to 9% difference that the proponents of the WHO criteria use to support the current criteria for the diagnosis of BAC. Of the 1909 patients with a histologically confirmed diagnosis of BAC and complete TNM staging identified by Zell and coworkers in 2007, 627 patients (33%—probably the same cohort of patients used in this group’s previous publication) were found to have stage I disease, and 572 (30%) had stage II disease. Basing their analysis on current criteria established by the WHO, these investigators demonstrated that patients with stage I disease had a 1-year survival rate of 94% and a 5-year survival rate of 65%, and patients with stage II disease had a 1-year survival rate of 89% and a 5-year survival rate of 45%. This study has the benefit of a larger population than Noguchi’s, which claimed a 100% survival rate. The Zell investigators did assert, however, that survival had improved in patients in late stages of disease, a claim that would seem to contradict the WHO criteria, which limits the diagnosis of BAC to tumors that do not show pleural, lymphatic, and interstitial involvement. Even so, it is of great interest to review the stated survival rates in earlier publications on BAC, before the current WHO criteria took effect.
Grover and colleagues summarized the experience of the Lung Cancer Study Group by collecting a large series of 235 tumors diagnosed as pure BAC between the years 1977 and 1988. These workers concluded that the mortality rate for patients with BAC stage I (T1N0) was 7% per year, or approximately 35% at 5 years. These data support the findings of Zell and coworkers, rather than those of Noguchi. More recently, Garfield and colleagues raised concerns about the current definition of BAC, mainly in cases with multifocal involvement, and stated that the current definition is inapplicable for patients with stage IIIB and stage IV disease. Similar concerns also were raised by Damhuis and associates, who found an unfavorable prognosis for tage I BAC, with a 5-year survival rate of 24%. Although these investigators provided possible explanations for this rather poor outcome, they concluded that the current definition makes it difficult to establish the diagnosis of BAC before surgery, and they question whether the diagnosis of “BAC with invasive component” should be maintained.
Dissatisfaction with the WHO criteria for the diagnosis of BAC has been expressed by several groups of investigators. In an ultrastructural analysis of 155 cases of BAC, Sidhu and colleagues stated that the unique characteristic of BAC is its cell type, and that the extent of lepidic growth, degree of differentiation, and degree of stromal desmoplasia cannot be used as definitional requirements. These workers further characterized the current definition of the WHO as a form of in situ adenocarcinoma , in which BAC is defined as a pattern rather than an entity, being accepted as an entity only when the lesion is not invasive. The current criteria negate the notion that BAC has the potential to spread and makes the staging of tumor pathology impractical. Thus, under this current definition, BAC may become an extremely rare entity, if it continues to be diagnosed at all. Hajdu went even farther in his opinion about the current WHO definition of BAC by stating that when “a group of pathologists changed the definition of BAC, it was de facto implied that BAC is carcinoma in situ and that invasive BAC does not exist.”
Three main presentations for BAC are recognized:
Localized: In this form of the disease, a peripheral mass is present in the lung parenchyma, which may be indistinguishable from any other non–small cell carcinoma in the lung. Usually these tumors are smaller than 3 cm in greatest diameter and do not show areas of necrosis or hemorrhage. They are well-defined tumors without encapsulation. The cut surface appears homogeneous and is tan in color.
Multinodular: In this presentation, the tumor involves extensive areas of the lung parenchyma in a miliary fashion, almost mimicking metastatic disease. The nodules are of variable size but usually are less than 1 cm in greatest diameter. This type of presentation may involve a lobe or the entire lung parenchyma ( Fig. 3-17 ).
Diffuse: This presentation is similar in appearance to a pneumonic process. The tumor involves extensive areas of lung parenchyma that may encompass one lobe or the entire lung parenchyma. No tumor masses or nodules are identified in this form of the disease, and the appearance is that of a non-neoplastic process ( Fig. 3-18 ).
The histopathologic features of BAC echo those of its gross appearance. In the nodular form of the tumor, an almost intact, normal-appearing lung parenchyma is seen at light microscopy. Closer inspection, however, reveals areas in which the alveolar walls are being replaced by either a low cuboidal or a cylindrical type of epithelium, entirely or partly lining the alveolar wall and reminiscent of the outline of the normal alveolar wall ( Figs. 3-19 to 3-21 ). The tumor does not show increased mitotic activity or cellular pleomorphism with nuclear atypia. The proliferation is rather bland but is distinct from the normal alveolar lining.
The multinodular pattern of BAC resembles a meta- static tumor in terms of the extensive “skip” areas of normal lung parenchyma. The tumor nodules appear to be discretely affecting extensive areas of the lung parenchyma, but in a nodular pattern rather than as a continuous process ( Fig. 3-22 ). A low cuboidal or columnar type of mucinous epithelium lines the alveoli; in some areas, it is possible to identify normal alveoli that are filled with an acellular mucinous material. Mitotic figures and cellular pleomorphism with nuclear atypia are not common; nor are necrosis and hemorrhage.
The diffuse pattern of BAC is almost invariably of the mucinous type. In this pattern, two important features may be easily identifiable: extensive areas in which the alveoli are filled with mucinous material containing mucinophages, and the presence of alveoli that are being replaced by a columnar mucinous type of epithelium ( Fig. 3-23 ). At low magnification, this pattern can be easily misread as a pneumonic process; thus, it is referred to as the pneumonic type.
The most important consideration in the differential diagnosis for BAC is atypical adenomatous hyperplasia. The histopathologic characteristics of these two lesions are very similar, and in many situations the only way to separate them is strictly by size. Atypical adenomatous hyperplasia is defined as a lesion of no more than 0.5 cm in greatest diameter. In small core needle biopsy specimens, however, the distinction between these two conditions on histologic grounds may prove to be very difficult. One other lesion that may be important to include in the differential diagnosis is papillary adenoma of type II pneumocytes. This lesion is exceedingly rare and occurs in a central location.
ATYPICAL ADENOMATOUS HYPERPLASIA
Atypical adenomatous hyperplasia (AAH) is currently defined by the WHO as an adenomatous lesion no more than 0.5 cm in diameter ( Fig. 3-24 ). AAH frequently is associated with lung adenocarcinoma with a bronchioloalveolar component, and histologically, AAH may be indistinguishable from BAC. In a study of 3641 resections for lung adenocarcinoma, Koga and colleagues found that AAH was present in 57% of these tumors and proposed that these lesions may be a precursor of lung adenocarcinoma, specifically BAC. Of interest, Morandi and associates attempted to correlate the genetic relationship among atypical adenomatous hyperplasia, BAC, and adenocarcinoma, concluding that AAH and the associated cancer are genetically independent, and that less frequently, AAH foci may represent an early spread of cells from the main tumor, rather than a precursor tumor.
Adenocarcinomas are malignant epithelial tumors characterized by the presence of glandular differentia- tion. Depending on the degree of glandular differentiation, adenocarcinomas are grouped into three histopathologic grades: well-, moderately, and poorly differentiated. Traditionally, four major growth patterns have been recognized: acinar, solid, papillary, and bronchioloalveolar. Depending on the degree of differentiation, the intracellular mucin or the intraglandular mucinous content may vary. In well-differentiated adenocarcinomas, the glandular malignant component may show intraluminal collections of mucinous material, whereas in solid component and poorly differentiated tumors, the presence of mucin is more evident at the intracellular level. In these cases, histochemical stains such as mucicarmine may help to properly identify the mucinous content. Ultrastructural studies may reveal Clara cell granules in peripheral lung adeocarcinomas; Ogata and Endo suggest that peripheral adenocarcinomas may show Clara cell differentiation regardless of the histologic growth pattern. Hirata and colleagues suggest that bronchial gland cell–type adenocarcinoma occurs more often in younger patients than in older patients (mean age, 50 years).
Some researchers have classified adenocarcinomas by means of cellular morphology and anatomic site, rather than by histologic differentiation, dividing them into parenchymal adenocarcinomas, bronchial adenocarcinomas, and adenocarcinomas of uncertain origin. To some extent, these classifications correlate with the degree of differentiation of a particular tumor. In many instances, however, tumors will display varied histologic patterns ranging from well- to poorly differentiated. In small biopsy specimens, it may be impossible to assess the full spectrum of histologic variability.
Adenocarcinomas may manifest as central or peripheral tumors. Tumor size may range from 0.5 cm to larger than 10 cm in greatest dimension. These tumors appear to be well delimited but not encapsulated and are grayish to light brown in color ( Fig. 3-25 ). Areas of hemorrhage or necrosis may be present. When located centrally, the tumor may compress airway structures or in some cases actually obstruct the airway ( Fig. 3-26 ). Peripheral tumors may have the same macroscopic characteristics as those of central tumors but appear to be in a subpleural location. In some instances, puckering or retraction of the pleura may be present ( Fig. 3-27 ). In this setting, it is important to consider the possibility of gross pleural invasion, because tumors smaller than 3 cm with pleural invasion are placed in a different staging category. Inking the pleural surface is recommended in order to properly evaluate it under light microscopic examination.
Although in most instances lung carcinomas manifest as a solitary tumor, in a study of 50 cases of consecutive adenocarcinomas, Miller and colleagues documented that 12% were in fact multiple adenocarcinomas. It is important to assess the size of the adjacent nodules in this setting. If the lesions are smaller than 0.5 cm in diameter and display the proper histologic features, they may be classified as atypical adenomatous hyperplasia; however, if the nodules are larger than 0.5 cm, the possibility of multifocal adenocarcinoma must be considered.
Histopathologic findings will reflect the degree of tumor differentiation.
Well-differentiated adenocarcinoma: At low-power magnification, well-differentiated adenocarcinoma appears as an atypical glandular proliferation replacing normal lung parenchyma ( Figs. 3-28 to 3-34 ). These tumors are well defined but not encapsulated, and the glandular appearance is relatively easily identified by light microscopy. The malignant glands are composed of columnar or mucinous epithelium, with round to oval cells, ample cytoplasm, round nuclei, and prominent nucleoli. In some well-differentiated tumors, the cytologic features of the neoplastic glandular proliferation are bland, with virtual absence of mitotic activity. In others, mitotic figures are present, and areas of hemorrhage and necrosis may be conspicuous. The glandular proliferation may be embedded in dense areas of fibrocollagenous tissue. The acinar and the papillary growth patterns commonly are seen in well-differentiated adenocarcinomas.
Moderately differentiated adenocarcinoma: As in well-differentiated tumors, the glandular formation is apparent; however, the formation or partial formation of variably sized glandular structures also may be observed. The cellular component may exhibit more prominent nuclear atypia, and mitotic figures are more readily identifiable ( Figs. 3-35 to 3-40 ). Areas of necrosis or hemorrhage may be present, as may areas of inflammatory reaction. The acinar and papillary growth patterns often are observed with this degree of differentiation.
Poorly differentiated adenocarcinoma: These tumors are characterized by the presence of sheets of neoplastic cells with only focal areas of glandular differentiation ( Figs. 3-41 to 3-49 ). The solid growth pattern, in which focal areas of abortive glandular formation are present, is most commonly observed in tumors of this histologic grade. High-power magnification of the malignant cellular component may display cells with ample cytoplasm, round to oval nucleus, and prominent nucleolus. The cells may show vacuolization of the cytoplasm, and in such cases the use of histochemical studies, such as mucicarmine, may be helpful to demonstrate the presence of intracellular mucin. Areas of more conventional glandular differentiation also may be observed, which will assist in the classification of the tumor as adenocarcinoma.