69: Role of the Pathologist and Pathology- Specific Treatment of Lung Cancer

A pathologist experienced in thoracic oncology is an essential member of the thoracic team. Surgical resection related to lung cancer and other pulmonary pathology accounts for the largest proportion of current thoracic practice. The goals of pathologic analysis of surgical lung specimens are to classify the lung cancer, determine the extent of its invasion (i.e., pleural, lymphovascular, soft tissue, or chest wall), and establish the status of the surgical margins for cancer involvement.¹ Accurate disease identification and staging are of pinnacle importance to the decision-making process and influence the diagnosis (benign or malignant), course of treatment, selection of optimal surgical approach, and pursuit of appropriate adjuvant and neoadjuvant therapies such as chemotherapy, radiation, and other innovative approaches to treatment. Further, determination of the specific molecular abnormalities of the tumor is critical for predicting sensitivity or resistance to a growing number of drugable targets primarily tyrosine kinase inhibitors (TKIs).^2,3 After a malignancy has been identified, the pathologist must determine whether the tumor is primary or metastatic. Most tumors found in the lung represent metastatic foci from distant primaries, such as breast and colon cancer, as opposed to a primary lung malignancy. While the pathologic features of metastatic versus primary adenocarcinoma may be similar, for example, the treatment course is not. Immunohistochemistry (IHC) is required to make the distinction and has proved to be an invaluable diagnostic adjunct. Primary malignant tumors of the lung are most often of epithelial or mesenchymal origin. The epithelial tumors are broadly divided into small-cell lung cancer (SCLC) and non–small-cell lung cancer (NSCLC). NSCLC is further classified as squamous cell carcinoma (SCC), adenocarcinoma, and large-cell carcinoma (LCC).⁴ The World Health Organization (WHO) tumor classification system has historically provided the foundation for the classification of lung tumors, including histologic types, clinical features, staging considerations as well as the molecular, genetic, and epidemiologic aspects of lung cancer.^4–6

The pathologist plays a fundamental role in the preoperative, intraoperative, and postoperative evaluation. The preoperative evaluation includes examination of one of the following specimens: bronchial brushings, bronchial washings, fine-needle aspiration biopsy, core needle biopsy, endobronchial biopsy, and transbronchial biopsy. Because lung tumors demonstrate a great deal of heterogeneity, accurate classification depends on sampling technique: If the pathology sample is limited, sometimes the only categorization that can be made is the distinction between NSCLC and SCLC. The generic term “non–small-cell lung cancer (NSCLC)” should be avoided as a single diagnostic term. In small biopsy samples of poorly differentiated carcinomas where IHC is used, the following terms are acceptable: “NSCLC favor adenocarcinoma” or “NSCLC favor SCC.”^6,7 Mutational testing (e.g., epidermal growth factor receptor [EGFR], anaplastic lymphoma kinases [ALKs]) should be performed in this setting. Lymph node status is one of the most important prognostic features in patients with NSCLC.^8,9 Since mediastinoscopy with pathologic examination of lymph nodes remains the “gold standard” for the evaluation of lymph node status in patients with NSCLC, mediastinal lymph nodes are sampled during the preoperative evaluation and provide information important to staging and therapeutic options.^10–13

The intraoperative evaluation of the surgical pathology specimen is performed by frozen-section examination, which can be analyzed immediately, and findings are communicated to the operating room. Lobectomy or pneumonectomy specimens are routinely evaluated intraoperatively to determine the status of the surgical resection margin, to diagnose incidental nodules discovered at the time of surgery, and to evaluate regional lymph nodes.

The postoperative evaluation reveals pathologic characteristics necessary for classification of tumor type, staging, and prognostic factors. The pathology diagnostic report should include the histologic classification as described by the WHO for carcinomas of the lung with squamous morphology, neuroendocrine differentiation, and other variant carcinomas. The recently published classification of adenocarcinoma should be used for this tumor subtype in resection specimens and small biopsy specimens.⁶ Use of bronchioloalveolar carcinoma (BAC) terminology is strongly discouraged. The parameters considered in the surgical pathology report are histologic type, histopathologic grade, visceral pleural invasion (Fig. 69-1), venous/lymphatic vessel invasion (Fig. 69-2), and involvement of mediastinal lymph nodes and TNM stage groupings.¹⁴

Although the WHO pathologic classification of lung tumors published in 2004 was the basis for categorizing lung tumors,⁵ there were a number of pitfalls that made it difficult to utilize the classification system: (1) the term bronchioloalveolar carcinoma (BAC) was used for widely divergent clinical, radiologic, pathologic, and molecular subsets of patients, (2) the “mixed subtype” category accounted for more than 90% of all resected lung adenocarcinomas, despite great heterogeneity in clinical, radiologic, pathologic, and molecular features, and (3) the WHO classification was based primarily on resection specimens, despite the fact that 70% of patients were presenting in advanced stages where the only type of tissue samples were obtained by biopsy or cytology.

The recent tumor classification system issued in 2011,⁶ addresses the inadequacies of the 2004 classification and provides the foundation for tumor diagnosis and patient therapy and a critical basis for epidemiologic, molecular, and clinical studies. The most important changes in the 2004 revised classification system were (1) eliminate the term, bronchioloalveolar carcinoma, (2) define the term adenocarcinoma in situ (AIS), (3) define the term minimally invasive adenocarcinoma (MIA), (4) revive the term lepidic, (5) promote comprehensive histologic subtyping, (6) emphasize and introduce the term micropapillary carcinoma, (7) detach the term mucinous adenocarcinoma, and (8) discourage use of the term NSCLC and subclassify the tumors in as much detail as possible.

For the first time in the history of lung cancer evaluation, the classification of lung cancer will be applicable to either (a) resected specimens or (b) small biopsy and cytology specimens. In the revised WHO classification system, (1) lesions attributed to preinvasive lung cancer now include atypical adenomatous hyperplasia (AAH) and AIS (formerly BAC); (2) the term bronchioloalveolar carcinoma (BAC) has been completely eliminated; (3) two new entities, both with favorable prognoses, were added to the category of adenocarcinoma, namely, AIS and MIA; and (4) the formerly known mucinous BAC category was renamed invasive mucinous adenocarcinoma and “shifted” to the category of “Variants of Invasive Adenocarcinoma” (Table 69-1). Since adenocarcinomas are extremely heterogeneous, extensive sampling is necessary to identify each of these histologies in the surgical pathology specimen. The surgical pathology report always should include the histologic classification published by the WHO for carcinomas of the lung (Table 69-2).

In contrast to surgical resections, the biopsy or cytology specimens have a limited amount of tissue and therefore the strategy and prioritization should focus on molecular tests that could identify potentially drugable molecular targets. For tumors that show clear morphologic features of adenocarcinoma or SCC, the standard terms are used. However, if the tumor only shows a carcinoma with no clear squamous or glandular features (NSCLC-NOS), a minimal immunohistochemical workup is recommended using a single adenocarcinoma marker and squamous marker. At the moment, the best markers for adenocarcinoma and SCC are TTF-1 and p63, respectively.¹⁵ One of the key aspects that potentially impacts radiologists is the need to obtain sufficient tissue not only for diagnosis, but also for molecular studies. This requires a strategic and multidisciplinary approach so the method of biopsy results in either a core biopsy or a cell block from tissue samples obtained for cytology.

Adenocarcinoma in situ (AIS), formerly BAC, is an important subtype of pulmonary adenocarcinoma. This cancer has received increasing attention in recent years owing to its increasing incidence and rate of sensitivity to epidermal growth factor—TKIs.¹⁶ AIS is a primary lung tumor with a peripheral location, well-differentiated cytology, lepidic growth pattern and a tendency for both aerogenous and lymphatic spread. The key feature is preservation of the underlying architecture of the lung with no invasion.

Minimally invasive adenocarcinoma (MIA) was introduced to define patients with a near 100% 5-year disease-free survival. It is defined as a lepidic predominant tumor measuring 3 cm or less that has an invasive component of 5 mm or less.¹⁷ MIA is characterized by a combination of ground glass opacity (GGO) and a central solid opacity, with the solid component measuring 5 mm or less. Nonmucinous MIA (Fig. 69-3), is more common than mucinous MIA and most often appears as a GGO. Mucinous MIA (Fig. 69-4) appears radiologically as a solid or part-solid nodule.

Invasive adenocarcinoma changes were inserted in the classification of invasive adenocarcinomas. Overtly invasive adenocarcinomas are classified according to the predominant subtype after the use of comprehensive histologic subtyping to estimate the percentages of the various components in a semiquantitative fashion in 5% to 10% increments. The term lepidic predominant adenocarcinoma consists of mixed subtype tumors containing a predominant lepidic growth pattern of type II pneumocytes and/or Clara cells that have an invasive component >5 mm. A micropapillary predominant subtype is added because it has been recognized as a poor prognostic category. Signet ring and clear cell carcinoma subtypes are now recorded as cytologic features whenever present with a comment about the percentage identified.

Tumors with Neuroendocrine Morphology

Neuroendocrine tumors of the lung are a distinctive subset of lung cancers characterized by varying degrees of neuroendocrine morphologic, immunohistochemical, and ultrastructural features.¹⁸ This category includes a wide spectrum of tumor types: low-grade typical carcinoid (TC) (Fig. 69-5), intermediate-grade atypical carcinoid (AC) (Fig. 69-6), and two high-grade tumors, large-cell neuroendocrine lung carcinoma (LCNEC) (Fig. 69-7) and SCLC (Fig. 69-8).¹⁹ Accurate classification of neuroendocrine tumors has prognostic importance. The grade of malignancy of neuroendocrine tumors progresses in the following order: TC, AC, LCNEC, and SCLC.¹⁹ No prognostic difference was noted between LCNEC and SCLC. The carcinoid nomenclature is preferred by the WHO over terms such as well-differentiated neuroendocrine carcinoma because it provides continuity with established terminology familiar to clinicians.⁵ In the 2004 WHO classification, TC and AC are categorized together under the heading of carcinoid tumors; LCNEC is listed as a subtype of LCC, and SCLC is retained as an independent category. Histologically, the neuroendocrine features consist of an organoid or trabecular growth pattern, peripheral palisading of tumor cells around the periphery of tumor nests, and the formation of rosette structures.

TC is defined as a neuroendocrine tumor with fewer than 2 mitoses per 2 mm² and no necrosis (Fig. 69-5). AC is defined as a neuroendocrine tumor that meets one of the two criteria: 2 to 10 mitoses per 2 mm² or necrosis (Fig. 69-6). In contrast to the high-grade neuroendocrine tumors, TC and AC do not occur in combination with other types of carcinoma. The number of mitoses and necrosis may be present only focally within a given tumor. Therefore, accurate classification of carcinoid tumors into TC or AC may not be possible in limited biopsy specimens with scant diagnostic material, and a definite diagnosis may require larger fragments of tumor. In these situations, it is recommended that small biopsies be signed as “carcinoid tumor” and the appropriate classification be performed on thorough examination of the resected specimens.^5,18

LCNEC is defined as a neuroendocrine tumor with more than 10 mitoses per 2 mm² and cytologic features of LCC (Fig. 69-7). These features include cells with polygonal shape, abundant cytoplasm, and prominent nucleoli. Evidence of neuroendocrine differentiation must be demonstrated by performing IHC for the specific neuroendocrine markers chromogranin and synaptophysin.⁴ Only tumors that show both neuroendocrine morphology and positive staining should be classified as LCNEC. It is important to note that up to 20% of conventional adenocarcinoma, small-cell carcinoma (SCC), or LCC will stain with neuroendocrine markers. Such tumors have been designated as NSCLC with neuroendocrine differentiation.

SCLC is defined as a neuroendocrine tumor with more than 10 mitoses per 2 mm² and small-cell cytologic features (Fig. 69-8). Cells have an oval or vaguely spindled shape and have scant cytoplasm. Nuclei are hyperchromatic and have absent or very small nucleoli (Fig. 69-8). Crush artifact may be prominent on small biopsies, but this is not pathognomonic for the diagnosis of SCC. In larger core biopsies or resected specimens, the cells may appear slightly larger than in a transbronchial biopsy and may have distinct cytoplasm. Numerous prominent nucleoli and large cells should not be seen.

LCNEC and SCLC may occur in combination with other NSCLCs as well as with each other. Such tumors are termed combined LCNEC, combined SCLC, and combined SCLC/LCNEC, respectively.^4,20 While the two high-grade neuroendocrine tumors show numerous similarities, they are retained in separate classifications because LCNEC currently has not been shown to respond to chemotherapy in the same fashion as SCLC. Surgery is the currently preferred treatment for LCNEC, although further studies are ongoing.^21–24

Immunohistochemical Staining

Although the concordance is generally good between the histologic subtype and the immunophenotype seen in small biopsies compared with surgical resection specimens, caution is advised in attempting to subtype small biopsies with limited material or cases with an ambiguous immunophenotype. IHC should be used to differentiate primary pulmonary adenocarcinoma from SCC or LCC, from metastatic carcinoma, and from malignant mesothelioma; and to determine whether neuroendocrine differentiation is present.^15,25,26 Limited use of IHC studies in small tissue samples is strongly recommended, thereby preserving critical tumor tissue for molecular studies particularly in patients with advanced stage disease. A limited panel of p63 and TTF-1 should suffice for most diagnostic problems.¹⁵

Differentiation between Primary Pulmonary Adenocarcinoma and Metastatic Adenocarcinoma

The morphologic features of primary adenocarcinoma of the lung may be similar to the features of an adenocarcinoma that is metastatic from a distant primary site. Although the presence of multiple nodules often leads to the presumptive diagnosis of metastases, multifocal adenocarcinoma is not rare and needs to be distinguished from metastases. Furthermore, patients with a solitary pulmonary nodule may have metastatic adenocarcinoma to the lung as the first presentation of disease.

TTF-1 is a homeodomain-containing transcription factor that regulates tissue-specific expression of surfactant apoprotein A, surfactant apoprotein B, surfactant apoprotein C, Clara cell antigen, and T1α. TTF-1 is very important in distinguishing primary from metastatic adenocarcinoma because most of the cases of primary carcinomas are positive, whereas metastatic adenocarcinoma to lung is virtually always TTF-1 negative (Fig. 69-9A). Lung cancer subtypes have different TTF-1 expression: 75% positive in adenocarcinoma, 85% positive in SCLC, and rarely in the SCC and LCC. Napsin A is an aspartic proteinase expressed in normal type II pneumocytes and in proximal and distal renal tubules and appears to be expressed in >80% of lung adenocarcinomas and may be a useful adjunct to TTF-1.²⁷ The panel of TTF-1 and p63 (or alternatively p40) may be useful in refining the diagnosis in small biopsy specimens previously classified as NSCLC, not otherwise specified⁶ to either adenocarcinoma or SCC.

Although SCC (Fig. 69-10) is usually p63-positive, there is no marker to date that is able to distinguish between primary and metastatic SCC. However, p63 is an important immunostain in cases of poorly differentiated NSCLC, where the distinction between adenocarcinoma and SCC is virtually impossible to make on the basis of H&E-stained slides alone (Fig. 69-11).

CDX-2 is a highly specific and sensitive marker for metastatic gastrointestinal malignancies that could be used to differentiate these gastric entities from primary lung tumors (Fig. 69-9B).

Determining the Neuroendocrine Status of Tumors

CD56, chromogranin (reacts with cytoplasmic neuroendocrine granules), and synaptophysin (reacts with a cell membrane glycoprotein) are used to diagnose the neuroendocrine tumors of the lung. All TC and AC tumors stain with chromogranin and synaptophysin, whereas SCC is negative in 25% of cases (Figs. 69-12 and 69-13). Neuroendocrine markers are valuable in diagnosing NSCLC with neuroendocrine differentiation, a specific group of poorly differentiated tumors without a neuroendocrine morphology.

Distinguishing between Malignant Mesothelioma and Lung Adenocarcinoma

IHC is most valuable in distinguishing between malignant mesothelioma and lung adenocarcinoma. The distinction between pulmonary adenocarcinoma and malignant mesothelioma (epithelioid type) is made by using a panel of markers including two with known immunopositivity in mesothelioma (but negative in adenocarcinoma) and two with known positivity in adenocarcinoma (but negative in mesothelioma).²⁵ Immunostains relatively sensitive and specific for mesothelioma include WT-1, calretinin, D2–40, HMBE-1, and cytokeratin 5/6 (negative in adenocarcinoma).^28,29 Antibodies immunoreactive in adenocarcinoma include CEA, B72.3, Ber-EP4, MOC-31, CD15, and TTF-1 (negative in mesothelioma). (Figs. 69-14–69-16)²⁵.

Discoveries in molecular pathogenetic mechanisms have had a major impact in the field of pulmonary disease, with important implications for the screening, diagnosis, and treatment of lung cancer. Small molecules or chemical compounds modeled on the protein structure of their targets are being actively developed and potentially are powerful therapeutic agents, designed to interfere with critical intracellular signaling pathways.

As a result of all these developments, physicians involved in the care of patients with lung cancer should have an understanding of molecular biology and related techniques to facilitate selection of the most appropriate therapies for their patients, to understand the novel agents and techniques that are under investigation in clinical trials, and to communicate adequately with their increasingly well-informed patients. In addition, it is imperative that institutions and pathology departments bank frozen tissue on a routine basis with patient informed consent because DNA, RNA, and protein extracted from such material is of paramount importance for evaluating genetic information. Lung cancer is one of the most challenging cancers to treat and the standard therapy includes surgical resection, platinum-based chemotherapy, and radiation therapy alone or in combination. New research and ongoing clinical trials involve new targeted therapies and milder treatment regimens that improve survival. Targeted therapeutic agents are based on the concept of discovering genetic alterations and the signaling pathways altered in cancer and have added significantly to our armamentarium to prolong patient survival and minimize drug toxicity. Although improvements seen in the trials are modest, the hope is that an increased number of biomarkers will be available in the near future to help clinicians predict which patients are most likely to benefit from such therapies.

EGFR is normally found on the surface of epithelial cells and is often overexpressed in a variety of human malignancies. Presence of EGFR-activating mutations represents a critical biological determinant for proper therapy selection in patients with lung cancer.

There is a significant association between EGFR mutations – especially exon 19 deletion, exon 21 (L858R), and exon 18 (G719X) mutations – and sensitivity to TKIs.^2,30,31 Exon 20 insertion mutation may predict resistance to clinically achievable levels of TKIs. EGFR and KRAS mutations are mutually exclusive in patients with lung cancer.³²

KRAS mutations are associated with intrinsic TKI resistance, and KRAS gene sequencing could be useful for the selection of patients as candidates for TKI therapy.³³ The prevalence of EGFR mutations in adenocarcinoma is 10% of Western and up to 50% of Asian patients, with a higher frequency of EGFR mutations in nonsmokers, women, and nonmucinous cancers. KRAS mutations are most common in non-Asians, smokers, and in mucinous adenocarcinoma.³⁴ The most common EGFR mutations result in an arginine for leucine substitution at amino acid 858 in exon 21 (L858R) and in frame deletions at exon 19. Mutations are more common in nonmucinous lung adenocarcinoma with lepidic pattern (formerly BAC pattern) and in lung adenocarcinoma with papillary (and or micropapillary) pattern.

Primary resistance to TKI therapy is associated with KRAS mutation. Acquired resistance is associated with second-site mutations within the EGFR kinase domain, amplification of alternative kinases (such as MET),³⁵ histologic transformation from NSCLC to SCLC, and epithelial to mesenchymal transition (EMT).

ALK gene rearrangements³⁶ represent the fusion between ALK and a variety of partners including echinoderm microtubule-associated protein-like 4 (EML4). ALK fusions have been identified in a subset of patients with NSCLC and represent a unique subset of NSCLC patients for whom ALK inhibitors may be a very effective therapeutic strategy.³⁷ Crizotinib is an oral ALK inhibitor that was approved by the FDA for patients with locally advanced or metastatic NSCLC who have the ALK gene rearrangement (i.e., ALK positive). ALK NSCLC occurs most commonly in a unique subgroup of NSCLC patients who share many of the clinical features of NSCLC patients likely to harbor EGFR mutations.^38–41 However, for the most part, ALK translocations and EGFR mutations are mutually exclusive.⁴⁰ ALK translocations tend to occur in younger patients and in those with more advanced NSCLC while this relationship has not been reported for EGFR mutant NSCLC.^42,43

The current standard method for detecting ALK NSCLC is fluorescence in situ hybridization (FISH), although other methods are currently being used for rapid screening of lung cancers with ALK rearrangements, including polymerase chain reaction (PCR) and IHC. A big advantage of FISH is that a probe set, developed for the diagnosis of ALK-rearranged anaplastic large-cell leukemia (ALCL), is applicable to the diagnosis of ALK-rearranged lung adenocarcinomas. Recently developed IHC tests are useful for the detection of the majority of ALK-rearranged lung adenocarcinomas.^38,40

Histology-Specific Surgery

Small-cell Lung Cancer

Historically, SCLC has been considered a nonoperative disease. Early randomized studies suggested no benefit to surgery over radiation or chemotherapy only. The Medical Research Council trial published in 1973 evaluated only patients with SCLC involving the bronchus, and complete resection was performed in only 34 of 71 (48%) patients.⁴⁴ The Lung Cancer Study Group, Eastern Cooperative Oncology Group (ECOG), and European Organization for Research and Treatment of Cancer published an intergroup trial in 1994, which randomized patients with limited SCLC who had responded to five cycles of chemotherapy (cyclophosphamide, doxorubicin, vincristine), but again these included only patients who had bronchoscopically visible disease. While most of the surgical patients (54 of 70, 77%) underwent complete resection, 35 (50%) surgical patients had cN2 disease, 24 (34%) had pN2 disease, and 12 (17%) were unresectable at the time of surgery.⁴⁵ Neither of these studies demonstrating lack of benefit of surgery in “limited” SCLC included patients with early SCLC as defined by modern-day staging techniques, including PET-CT, brain MR, and endobronchial ultrasound (EBUS) or mediastinoscopy for nodal evaluation.

Resection is recommended for early-stage SCLC. Early-stage SCLC comprises a small proportion of an already rare entity. Ten to fifteen percent of lung cancers are SCLC, and only 3% to 7% of these patients present with stage I or II disease, as determined in a SEER database study of all patients diagnosed with lung cancer between 1988 and 2005 and followed for at least 3 months.⁴⁶ Resection offers benefit for tumors that prove to be combined or mixed SCLC/NSCLC (10%–30% of SCLC cases). In addition, resection offers the best local control in limited disease. The same SEER database analysis⁴⁶ found lobectomy alone (47% 5-year overall survival) to be superior to radiation therapy (17% 5-year overall survival), a finding which was significant on multivariable analysis (HR = 0.56; 95% CI = 0.41–0.76). Other studies have demonstrated similar survival for lobectomy in early-stage SCLC.⁴⁷ The National Comprehensive Cancer Network (NCCN) currently recommends lobectomy with mediastinal nodal sampling or dissection for patients with clinical stage I disease to be followed by adjuvant chemotherapy, even if nodes are pathologically negative.⁴⁸ Patients treated with complete resection should also undergo prophylactic whole brain irradiation following adjuvant chemotherapy if performance status and neurocognitive function are not impaired, given the high propensity of SCLC to metastasize to the brain.

Carcinoid

Surgical resection is recommended for pulmonary carcinoid, which can be found at all levels from the trachea to the lung periphery.⁴⁹ The prognosis for resection of low-grade TC is excellent, with 82% 10-year survival in Wilkins series that spans a 50-year experience⁴⁹ as well as in a modern Australian series.⁵⁰ Most surgeons advocate a parenchymal-sparing approach, which often mandates sleeve resection (bronchus, carina, or even trachea) to preserve lung for tumors that are frequently central. Peripheral carcinoids can be managed with sublobar resection, including segmentectomy or wedge resection, with good results, as supported by a recent SEER database evaluation of 3270 pulmonary carcinoid patients, in which sublobar resection compared well to lobectomy when assessed for noninferiority in multivariable analysis.⁵¹

Non–small-cell Lung Cancer

The remaining primary epithelial lung malignancies are NSCLC, including adenocarcinoma, SCC, and others, such as LCC. Treatment for most NSCLCs depends on stage, with surgical resection generally the treatment of choice for early-stage disease. Chapter 71 describes the standard lobectomies for NSCLC isolated to individual lobes. SCCs are frequently central and may necessitate sleeve resection or even pneumonectomy if there is involvement of hilar vessels or bronchi.

In contrast to the other forms of NSCLC, adenocarcinomas are remarkably heterogeneous and recent advancements in molecular genetics, histopathology, and imaging have had significant impact on the treatment of these tumors. Adenocarcinoma represents the most common histology among NSCLC and lung cancer in the United States.⁶ Tumors previously identified as BAC, and now defined as AIS or MIA, often demonstrate indolent behavior and favorable long-term survival.⁵² Increased understanding of these aspects of tumor biology and the advent of targeted biologic therapy for certain genetic mutations⁵³ have revolutionized the treatment of patients with AIS, MIA, and adenocarcinoma. These tumors frequently present as nonsolid or subsolid nodules on low-dose chest computed tomography, presenting a conundrum for treating clinicians. Given indolent and multicentric behavior for subsolid nodules, parenchymal-sparing sublobar resection is likely the best surgical approach,⁵⁴ but these lesions are frequently not palpable, making wedge resection difficult if possible at all. Anatomic segmentectomy would be preferred, but when a nonsolid adenocarcinoma crosses segmental boundaries, it may necessitate lobectomy, which is suboptimal in the setting of potentially multicentric disease. Whether to follow these preinvasive and early invasive adenocarcinomas with surveillance imaging and when to treat with surgery, radiation, or biologic therapy are areas ripe for further research.⁵⁵

Stage-specific Surgery

Stage I–III NSCLC

Chapter 68 introduced the seventh edition staging system for NSCLC. As discussed above in this chapter, treatment for most early-stage NSCLCs involves surgical resection with lobectomy, segmentectomy, or wedge resection. More advanced stage disease warrants consideration of multimodality therapy, including chemotherapy and external beam radiation.

Adjuvant Chemotherapy Following Surgical Resection

Stage is the major factor influencing whether a patient who undergoes surgical resection for NSCLC should receive adjuvant chemotherapy. Early trials suggested no benefit to adjuvant chemotherapy following resection. The ECOG compared concurrent radiation and cisplatin–etoposide with radiation alone following resection of stage II or IIIA NSCLC and found no statistically significant difference in overall survival or recurrence.⁵⁶ The Adjuvant Lung Project, Italy randomized 1209 patients who underwent complete resection for stage I, II, or IIIA NSCLC to an adjuvant cisplatin-based regimen (MVP) or observation and likewise found no statistically significant difference in overall or progression-free survival between the two groups⁵⁷ following resection for more advanced disease. In the largest randomized-controlled study of adjuvant chemotherapy following surgery, the International Adjuvant Lung Cancer Trial randomized 1867 patients with stage I, II, or III disease to cisplatin-based chemotherapy or observation. Patients receiving adjuvant chemotherapy experienced a significantly higher overall and disease-free survival.⁵⁸ Two additional international multicenter randomized studies demonstrated similar benefit for adjuvant cisplatin–vinorelbine doublets: the National Cancer Institute of Canada/National Cancer Institute of the United States Intergroup JBR.10 trial for patients with stage IB or II NSCLC⁵⁹ and the Adjuvant Navelbine International Trialist Association (ANITA) Trial for patients with Stage IB, II, or IIIA disease.⁶⁰ Of note, the ANITA investigators performed subgroup analyses and survival benefit for adjuvant cisplatin–vinorelbine was only demonstrated with the stage II and IIIA patients (not in stage IA), but the authors felt the numbers were small and the study was not sufficiently powered to evaluate differences for the individual stages.

Therefore, it is generally agreed that for stage II disease (and for those with stage III disease diagnosed upon final pathology following resection), treatment should include anatomic surgical resection followed by adjuvant chemotherapy. Adjuvant chemotherapy is not recommended following resection for stage IA NSCLC. For stage IB disease, the role of adjuvant chemotherapy was evaluated by the Cancer and Leukemia Group B (CALGB) 9633 trial, in which 344 patients with stage IB NSCLC were randomized to receive carboplatin–paclitaxel or observation following lobectomy or pneumonectomy, and no difference in overall survival was found between the groups.⁶¹ On unplanned subgroup analysis, the investigators found a survival benefit of adjuvant chemotherapy for patients with tumors 4 cm or larger, but overall, the study was felt to be underpowered to detect subtle differences in survival given the relatively long survival of these patients compared to those with more advanced disease.

Superior Sulcus (Pancoast) Tumors

Although a few patients with superior sulcus (Pancoast) tumors will have stage II disease (if the lesion is 7 cm or smaller and the hilar and mediastinal nodes are negative), most will have stage IIIA or IIIB disease, depending on the status of the mediastinal nodes. Practice has evolved over time,⁶² but superior sulcus tumors are generally treated with induction chemoradiation followed by surgical resection, using the Southwest Oncology Group (SWOG) protocol of cisplatin–etoposide with 45 Gy concurrent radiation followed by en bloc resection of the tumor with lobectomy.⁶³

Stage III NSCLC

Stage IIIA NSCLC is a heterogeneous disease that includes four types of pathology: (1) bulky N2 disease; (2) microscopic N2 disease found on preoperative staging; (3) microscopic N2 disease found at the time of or following resection; and (4) T3N1 tumors (hilar nodal disease with tumor involving the chest wall or mainstem bronchus within 2 cm of the carina). Even among individual specialties, such as medical oncology⁶⁴ or thoracic surgery,⁶⁵ there is little consensus regarding how to treat patients with stage IIIA NSCLC. For patients with bulky N2 disease, this is particularly controversial.

Several randomized trials have been conducted to investigate the best multimodality approach. A recent phase III trial compared radiation to surgery following response to platinum-based induction chemotherapy for patients with histologically proved stage IIIA-N2 NSCLC. Three hundred and thirty-two patients were randomized and results suggested no difference between the two arms, with a median and 5-year overall survival of 16.4 months and 15.7% for resection and 17.5 months and 14% for radiation.⁶⁶ The Lung Intergroup Trial 0139 randomized 396 patients with Stage IIIA-N2 NSCLC who underwent two cycles of cisplatin–etoposide with concurrent 45 Gy radiation per the SWOG protocol and deemed fit for surgery to undergo either surgery followed by two more cycles of chemotherapy or additional chemoradiation (for a total of four cycles and 60 Gy). Although progression-free survival was higher for the surgical arm (median 12.8 months vs. 10.5 months, p = 0.017), there was no difference in overall survival (23.6 months vs. 22.2 months, for surgery compared to radiation, p = 0.24).⁶⁷ Nearly all (14 out of 16) of the treatment-related deaths in the surgical arm were following pneumonectomy, prompting a subgroup analysis for patients who underwent lobectomy and not pneumonectomy. In this subgroup, overall survival was higher for patients who underwent lobectomy following induction chemoradiation than for those who underwent definitive chemoradiation. This has led many clinicians to use caution in treating patients with stage IIIA-N2 disease with induction chemoradiation followed by pneumonectomy. The Intergroup 0139 trial also found that a single N2 nodal station (compared to multistation disease) was independently associated with higher survival among surgical patients (p = 0.009). This finding suggests that single-station disease behaves differently compared to multistation stage IIIA-N2 disease, supporting the practice of some surgeons to consider primary surgical resection with lymphadenectomy followed by adjuvant chemotherapy or chemoradiation for single-station stage IIIA-N2 disease.

There is even less consensus regarding the treatment of N3 disease, or stage IIIB. A phase II study of a trimodality approach using the SWOG protocol (two cycles of cisplatin–etoposide with concurrent 45 Gy radiotherapy) followed by resection for stable or improved disease was conducted for patients with stage IIIA-N2 and N3 disease.⁶⁸ No difference for survival was found for patients with N2 versus N3 disease at diagnosis, with a 3-year survival of 26%, and the strongest predictor of long-term survival was negative nodal status on resection. The SWOG investigators also examined the role of definitive chemoradiation (four cycles of cisplatin–etoposide with concurrent 61 Gy) for patients with stage IIIB NSCLC.⁶⁹ The 3-year and 5-year survival percentage for the 50 patients accrued were 17% and 15%, respectively.

A reasonable approach for the four types of N2 disease (Table 69-3) described above is (1) chemoradiation with consideration of surgical resection for bulky N2 disease depending on response to induction therapy; (2) chemotherapy or chemoradiation followed by surgical resection for micrometastatic disease found on preoperative staging; (3) surgical resection followed by chemotherapy or chemoradiation for single-station disease found at the time of surgery (particularly for the left upper lobe and station 4 L) or for multistation disease found on pathology following resection; and (4) chemotherapy or chemoradiation followed by resection if disease has not progressed for T3N1 disease, with chemoradiation followed by en bloc resection for superior sulcus (Pancoast) tumors.

Stage IV NSCLC

The role of surgery in stage IV disease is traditionally palliative, such as pleurodesis or pleurX catheter placement for NSCLC with malignant pleural effusion. There are particular circumstances, however, in which curative surgery is attempted in the setting of stage IV NSCLC. In a series of 28 patients who underwent extrapleural pneumonectomy (EPP) for pleural dissemination of NSCLC, the median survival for nine patients who had node-negative disease was 52 months (vs. 14 months for those with positive nodes, p = 0.0003), suggesting that EPP may prolong survival in NSCLC patients with node-negative malignant pleural effusion (stage IV).⁷⁰ Data gathered retrospectively from 14 SWOG trials involving 2531 patients suggest that approximately 7% of patients with metastatic NSCLC present with a solitary metastasis,⁷¹ often called oligometastatic NSCLC. This represents a unique situation in which a multidisciplinary approach of chemotherapy, surgical resection of the lung primary, and treatment of the metastasis with resection, radiotherapy, or both may be used.⁷² Generally, this is considered for oligometastases to the brain or the adrenal, but others have reported small series (fewer than 10 patients) of pulmonary resection with metastasectomy for nonbrain, nonadrenal oligometastasis.⁷³ The data regarding pulmonary resection and oligometastasectomy are retrospective and sparse, with most studies combining metachronous and synchronous presentation of metastases. Whether to perform surgical resection for NSCLC in the setting of oligometastatic disease should be considered on an individual basis in collaboration with other treating physicians and after thorough discussion of treatment options with the patient.

Surgical excision remains the only therapeutic modality that can cure selected lung cancer patients. The surgical strategy varies depending on histology and stage, in addition to other morphologic and molecular parameters. Pathologists play an important role in the surgical management of patients with lung cancer from preoperative diagnosis and staging, to intraoperative evaluation of the extent of distant disease and margin status, to postoperative assessment of tumor genetic alterations. They offer guidance in selecting appropriate surgical therapies and chemotherapeutic regimens, as well as in identifying morphologic and molecular prognostic markers and predictors of response to therapeutic agents. They are expected to provide accurate diagnosis and classification of the patient's lung cancer. Most important, pathologists are required to provide pathologic staging information in lung resection specimens, surgical resection margin status, and information about lung cancer subtype.

Tumor stage, as determined by current American Joint Committee on Cancer guidelines, is considered to be one of the most important prognostic factors for patients with lung neoplasms. Frozen-section examination of mediastinal lymph nodes obtained by mediastinoscopy is used routinely to determine whether NSCLC patients will undergo tumor excision. Determination of the pT and pN status of the resected surgical pathology specimen determines whether the patient will be treated with postoperative chemotherapy and/or radiation therapy. Furthermore, classification algorithms based on the presence of genetic alterations found in lung cancer can be used to identify drugs or therapeutic agents targeting these alterations.

Investigation and analysis of lung cancer for particular targeted molecular abnormalities expand the expertise of the pulmonary oncologic pathologist, who in addition to conventional pathologic analysis of surgical lung specimens will determine the molecular abnormalities of lung cancer that may be able to predict for sensitivity and resistance to various chemotherapeutic agents and targeted therapies.

With the new WHO criteria published in 2011, pathologists, pulmonologists, and thoracic surgeons are relearning the nomenclature that defines lung cancer. This is particularly pertinent in the era of molecular targeted therapy, where gene mutations, rather than pathology, better define the treatment to which an individual patient is likely to respond. That said, the differentiation between squamous NSCLC and other nonsquamous pathologies appears to be a crucial first step, as the types of chemotherapy chosen for first-line therapy differ between these entities. Although most surgeons still recommend surgery for all NSCLC patients, there is now an increasing awareness of the potential role of surgery even in some SCLC patients. Likewise, if a patient is found to have NSCLC with neuroendocrine features, some medical oncologists would suggest an SCLC-type regimen in the neoadjuvant or adjuvant setting. An additional area of major change is the reclassification of BAC as AIS, with different degrees of invasion and lepidic spread. This will probably have a huge impact on the management of GGO on radiographic imaging in the future. Finally, use of molecular marker panels are now recommended by NCCN guidelines, at least for patients with metastatic disease, both to determine prognosis and to help choose appropriate chemo/biologic therapy.