Primary Pulmonary Hodgkin Lymphoma: A Cytologic Diagnostic Challenge

1. Case Presentation

A 57‐year‐old man presented with cough for about a month, associated with intermittent hemoptysis, mild shortness of breath on exertion, left chest soreness and a single episode of night sweats. He had no significant past medical history but had an 18 pack‐year smoking history, having quit 6 years prior. Apart from tenderness to palpation of the left chest wall, physical examination of all other systems was unremarkable. Chest x‐ray showed a left lung mass, following which computed tomography (CT) scan of the thorax confirmed a 5 cm mass that was centered in the left upper lobe with extension into the anterior mediastinum (Figure 1A). The mass was hypermetabolic on positron emission tomography (PET) scan (Figure 1B). The overall radiologic impression was that of a primary lung malignancy. While there was also a mildly enlarged aortopulmonary lymph node, no abnormal metabolic activity was detected in it and in the rest of the body.

FIGURE 1.

(A) Computed tomography and (B) positron emission tomography scans showing hypermetabolic, left upper lung mass (arrow) extending into anterior mediastinum.

The patient underwent CT‐guided fine needle aspiration (FNA) and biopsy of the mass. The cytologic smears showed predominantly acute inflammation with occasional admixed enlarged atypical cells that showed rounded to irregular lobated nuclear contours, open chromatin, variably prominent nucleoli, and scant to moderate eosinophilic cytoplasm (Figure 2A,B). Occasional binucleated forms with prominent nucleoli were present (Figure 2C). The corresponding core biopsy mirrored the cytologic findings, with some atypical cells displaying irregular hyperchromatic nuclei that were attributable to a degree of degenerative change (Figure 2D). Ancillary immunohistochemical stains were not performed in view of the sparsity of the atypical cells and obscuring florid inflammation, and the case was finalized as atypical but not further diagnostic. Concurrent microbiological studies were negative. A 4‐week interval CT scan showed no change in the nature of the lung mass despite a course of antibiotics. An endobronchial ultrasound (EBUS) guided FNA and biopsy was then performed, which yielded cytology and biopsy findings similar to the prior CT‐guided procedure.

FIGURE 2.

(A to C) Hematoxylin and eosin (H&E) stained cytologic smear of lung mass: (A) Highly cellular yield showing abundant inflammatory cells that mostly comprise neutrophils (100×); (B) Atypical mononucleated cells with large vesicular nuclei, rounded or lobated nuclear contours and variably prominent nucleoli were present which in retrospect are reminiscent of Hodgkin cells (arrows) (400×); (C) Occasional binucleated form with prominent nucleoli, in retrospect suggestive of Reed‐Sternberg (RS) cell (arrow), was identified (600×). (D) Core biopsy showing sparse, atypical cells with irregular hyperchromatic nuclei (arrows) on a background of predominantly neutrophils (400×).

Given the ambiguity of the disease, the decision was made to perform a surgical resection to clarify the diagnosis and to rule out malignancy. A robotic video‐assisted thoracoscopic (VATS) wedge resection of the left upper lobe was performed. Histologic examination showed effacement of the pulmonary parenchyma by an extensive, nodular inflammatory infiltrate with surrounding and interspersed fibrotic bands (Figure 3A). The inflammation was predominantly comprised of abundant mature neutrophils, but eosinophils were also present, with scattered small lymphocytes and plasma cells. There were scattered cells with morphologic features similar to the atypical cells seen on cytology, including large mononuclear cells with a single, prominent nucleolus (Figure 3B) and occasional binucleated forms, mummified cells with degenerative nuclear change, and cells with clear, retracted cytoplasm. The neoplastic cells were positive for CD30, MUM1, and PAX5 (Figure 3C,D), negative for several other lymphocytic and pan‐cytokeratin immunohistochemical markers, and negative for Epstein–Barr virus encoded RNA by in situ hybridization, confirming a diagnosis of classic Hodgkin lymphoma (HL) and favoring nodular sclerosis subtype. Excised mediastinal (levels 5 and 9L) lymph nodes and biopsy of adherent pleural tissue were negative for malignancy. Subsequent bone marrow biopsy and PET‐CT staging scan were also negative for systemic lymphomatous involvement.

FIGURE 3.

(A) Low‐magnification view of lung resection showing nodular inflammatory infiltrate (right) containing the neoplastic cells, separated from lung parenchyma (left) by thick bands of fibrous stroma. Reactive lymphoid aggregates/follicles are present in the fibrotic stroma and lung parenchyma (arrowheads) (40×). (B) Scattered large neoplastic mononucleated cells, including Hodgkin cells (arrowheads), were admixed with numerous neutrophils, eosinophils, and a small number of mononuclear cells. Occasional RS cells were also present (arrows) (600×). (C) The neoplastic cells were positive for CD30 and MUM1 (inset) (600×). (D) The neoplastic cells also showed weak to moderate positivity for PAX5 (600×).

2. Discussion

As in this case, the diagnosis of primary pulmonary Hodgkin lymphoma poses a significant challenge for a number of reasons. Primary pulmonary HL, defined as HL that is confined to the lung, with or without minimal hilar involvement and in the absence of other extranodal/distant disease, is an exceptionally rare form of HL with less than 150 cases reported in the literature [1, 2, 3]. Reported cases show that it can present with nonspecific respiratory symptoms such as cough, chest pain, dyspnea, fever, hemoptysis, or may be asymptomatic, and imaging may show a solitary mass, consolidation, ground‐glass opacity, nodules, or cavitary lesions. The nonspecific clinical presentation and radiologic findings result in a broad clinical differential diagnosis that includes infectious, inflammatory, and neoplastic processes [3, 4, 5].

Pulmonary HL is particularly challenging to recognize on cytology and small biopsies and may lead to delayed or misdiagnosis. In the only published cytology case report of primary pulmonary HL, the tumor was initially interpreted as non‐small cell carcinoma [6]. The relative paucity of malignant cells and their variable obscuration by a prominent reactive inflammatory background can mimic an inflammatory/infectious process. Depending on the severity of the atypia of the malignant cells, diagnostic considerations may include non‐lymphoid malignancy and non‐neoplastic processes. For example, mildly atypical tumor cells may be mistaken for reactive pneumocytes or neoplastic Langerhans cells (Langerhans cell histiocytosis) in a background of florid mixed inflammation, while tumor cells with a greater degree of nuclear atypia can mimic non‐small cell carcinoma or melanoma. Contributing to the diagnostic challenge, the tumor cells may exhibit degenerative nuclear change on cell block and core biopsies, which can preclude recognition of classic Hodgkin and Reed‐Sternberg cells. The limited number of malignant cells also potentially complicates the interpretation of immunohistochemistry, which is pivotal to diagnosis. Unsurprisingly, because imaging and cytomorphology are often nonspecific, diagnosis often depends on diagnostic pulmonary resection or sufficiently large core biopsy [4].

Given these diagnostic challenges, a high index of suspicion for HL must be maintained when evaluating cellular pulmonary aspirates composed predominantly of inflammatory cells with careful scrutiny for Hodgkin and Reed‐Sternberg‐like cells. Additional passes to procure adequate lesional material for ancillary immunohistochemical studies are highly recommended. Furthermore, in patients whose pulmonary lesions fail to resolve or respond to empirical antimicrobial treatment, it is prudent to pursue more substantial tissue biopsy for histologic examination early in the workup. Timely and accurate diagnosis is critical, as prompt identification allows for the early initiation of highly effective, standard chemotherapy regimens.

Author Contributions

Nicholas Jin Hong Tan: writing, original draft; visualization. Vanda F. Torous: writing, review and edits; visualization.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.