The occurrence of secondary pulmonary lymphoma (SPL) in patients diagnosed with non-Hodgkin's lymphoma (NHL) is not rare. Apart from the gastrointestinal tract, the lungs are one of the most frequent extra-nodal sites invaded by NHL, with 25-40% reported in the literature (1,2). SPL is part of systemic lymphoma and can invade lung tissue through blood dissemination or transfer to the lung through hilar and mediastinal lymph nodes. Clinically, patients frequently present with an occasional cough and hemoptysis, the same symptoms as those of pneumonia (3). Therefore, when CT scans demonstrate pneumonia-like imaging features, distinguishing between SPL and infectious pneumonia poses notable diagnostic challenges. The risk of opportunistic infections is influenced by the degree of pathogen exposure, host immune status and pathogen-host interactions (4). Patients with hematologic malignancies are prone to pulmonary opportunistic infections, which often deteriorate rapidly and evolve into respiratory failure (5). However, diagnosis is often delayed and occasionally completely overlooked, especially with invasive bronchopulmonary aspergillosis. The commonly used diagnostic methods for SPL include transbronchial lung biopsy or CT-guided lung biopsy and surgical interventions. The present report outlines 2 cases of coexistent SPL and opportunistic infections in patients with NHL diagnosed by minimal residual disease (MRD) detected in bronchoalveolar lavage fluid (BALF), as well as a review of the relevant literature.

A combination of search terms (lymphoma and pulmonary infiltrates) was used to screen abstracts and titles under case reports in the PubMed database (https://pubmed.ncbi.nlm.nih.gov/). A total of 386 relevant studies were identified. After evaluating the titles and abstracts, 299 irrelevant reports and duplicate cases were excluded. Within the remaining 87 studies, 56 consisted of primary pulmonary lymphoma and 31 contained SPL. Of these 31 cases, 11 were excluded due to non-English language use or a misdiagnosis of SPL. An additional four were excluded due to being dated before the 21st century. Ultimately, 16 patient case reports were enrolled for analysis (6-21). Fig. 8 exhibits the screening process and Table I outlines the basic information of 18 patients who experienced SPL (including the 2 cases outlined in the present report).

Within the 18 cases, SPL occurred in individuals of both sexes and varying ages. The ratio of male to female occurrence was 1:1 with a higher frequency among individuals aged >60 years (Table II). The three most prevalent clinical manifestations in these patients were dyspnea (66.7%), cough (33.3%) and a fever (27.8%). These symptoms are consistent with the 2 cases outlined in the present report. However, the patients in the present report manifested a cough with a large amount of pulmonary secretion, but did not exhibit hemoptysis. According to the literature, the proportion of asymptomatic patients was as high as 27.8%. It is worth noting that within these 18 patients, only 2 (11.1%) were associated with pulmonary infection. Among the chest CT findings, the majority of SPL exhibited bilateral lung infiltration, with only 2 cases showing unilateral infiltration (Table III). The majority of chest imaging findings suggested ground glass density (50%), lymph node enlargement (33.3%), multiple pulmonary nodules (44.4%) and lung consolidation (27.8%). Other presentations included single lymph node enlargement, pleural effusion and cavities.

Pathological characteristics were determined by a pathologist in all 18 cases (Table IV). There were no cases of Hodgkin's lymphoma. However, 4 cases were confirmed to contain diffuse large B cell lymphoma and there were 3 cases of mucosa-associated lymphoid tissue (MALT) and adult T cell lymphoma/leukemia as well as 2 cases of peripheral T cell lymphoma and follicular lymphoma, which were confirmed separately. The remaining cases were confirmed as angioimmunoblastic T cell lymphoma, T-acute lymphoblastic leukemia/lymphoblastic lymphoma, anaplastic large cell lymphoma and chronic lymphocytic leukemia.

There were a number of differential diagnoses of secondary pulmonary infiltration in patients with lymphoma, including SPL, drug-related lung injury, connective tissue disease and pulmonary infections, especially pulmonary fungal infections and tuberculosis. During the diagnostic process, the persistent clinical symptoms and the radiographic manifestations served as key indicators for clinicians to initiate therapeutic interventions. Diagnostic methods often involved numerous types of invasive methods, such as transbronchial lung biopsy (50%), surgery or thoracoscopic lung biopsy (27.8%), autopsy (16.7%) and BALF immunophenotyping test (27.8%) (Table V). On occasion, multiple biopsy methods were involved in the diagnostic process. Furthermore, it was observed that although transbronchial lung biopsy was used more frequently, the success rate was only 55.6%. Almost half of the patients needed to undergo other methods at the same time due to negative pathological findings or suboptimal biopsy specimen quality. In addition, unlike traditional invasive methods, in the present report, lymphoma cells were detected in BALF in only 5 cases, which has rarely been reported before, despite 1 case initially failing to diagnose SPL and requiring surgical biopsy for a definitive answer (case no. 11).

With regard to treatment and prognosis, 7 patients received combination chemotherapy after SPL diagnosis, 2 patients received single-agent chemotherapy with chlorambucil or cladribine respectively, 2 patients received only glucocorticoid therapy, 1 patient received prednisolone followed by allogeneic hematopoietic stem cell transplantation and 4 patients received no treatment or only palliative care. The overall prognosis of SPL was poor, with only 44.6% of enrolled patients exhibiting ongoing survival (Table II). However, early diagnosis and treatment has the potential to improve prognosis. As for the 2 cases outlined in the present report, the first patient, who had just achieved complete remission 2 months prior and developed a pulmonary infection during the treatment course of admission, received only methylprednisolone for antitumor therapy. The second patient received antitumor therapy with dexamethasone, cyclophosphamide and bendamustine after admission and continued with methylprednisolone treatment after an SPL diagnosis was confirmed. Both patients eventually progressed rapidly to mortality. The marked differences in treatments among patients may be attributed to individualized risk-benefit assessments based on a number of key factors differing between each patient. These may include disease characteristics (pathological confirmation, stage, tumor burden and aggressiveness of the lesion) and patient-specific factors (age, performance status, comorbidities, organ function and previous treatment history). Additionally, the treatment preferences of patients themselves also serve an important role in the selection of treatment regimens. Between the 2 cases in the present report, the first declined following further antitumor treatment. By contrast, the second patient underwent combination chemotherapy immediately upon confirmation of relapsed lymphoma after admission.

Within the present report, 2 cases of patients diagnosed with NHL, but having achieved complete remission following CHOP treatment, were outlined. The patients subsequently developed respiratory distress. Due to the pneumonia-like symptoms, signs and chest CT scan findings, the patients were initially diagnosed with pulmonary infection. Following combination treatment with multiple antibiotics, their sputum culture and mNGS of BALF were both negative. However, the patients continued to develop a fever and the imaging findings continued to progress in both lungs. The MRD tests of BALF were performed to establish the existence of SPL. Based on the diagnosis of SPL, a glucocorticoid-based treatment regimen was implemented, a method consistent with a number of reported cases (18,20). However, both patients ultimately succumbed to respiratory failure within 1 month. The lung is a common secondary site of involvement in lymphoma, accounting for ~24% of cases in NHL (22). The clinical symptoms of SPL are non-specific, except for respiratory and extrapulmonary symptoms. However, common symptoms include dyspnea (66.7%), cough (33.3%) and fever (27.8%). The proportion of asymptomatic patients was as high as 27.8% in the reported literature. In previous reports, the imaging findings of SPL have included ground glass density (50%), lymph node enlargement (33.3%), multiple pulmonary nodules (44.4%) and lung consolidation (27.8%). Due to the atypical clinical manifestations, the majority of cases were initially misdiagnosed as tuberculosis, pneumonia or deep fungal infections.

The lung is a predominant site for opportunistic infections. The risk of such infections is determined by the extent of pathogen exposure, host immunological competence and specific pathogen-host interaction mechanisms (4). Among patients with hematological malignancies, pulmonary opportunistic infections deteriorate rapidly and evolve into respiratory failure. In the cases outlined in the present report, infection was initially suspected based on sputum culture and mNGS results of BALF, which revealed a number of opportunistic pathogens. Following anti-infective treatment, the clinical manifestations of the patients showed marked improvement during the initial phase of treatment. No pathogens were found in the subsequent sputum culture and mNGS of BALF. However, the symptoms exhibited by the patients and the imaging findings worsened afterwards, which prompted an alternative diagnosis. Regardless of the ultimate etiology, prompt definitive diagnosis and targeted therapeutic intervention are key in optimizing clinical outcomes.

Histopathological workup is often required to establish a correct diagnosis. In current clinical practice, the diagnostic evaluation of pulmonary lesions primarily relies on minimally invasive techniques, such as bronchoscopy and CT-guided percutaneous lung biopsy, thereby markedly reducing the need for more invasive surgical interventions (23-25). However, it has been established that, although transbronchial lung biopsy is conducted more frequently, the success rate lies at only 55.6%. Furthermore, although surgical lung biopsy and autopsy often achieve an approximate diagnostic success rate of 100%, both are associated with notably higher risks and cost, given that a lung biopsy is often difficult to perform and requires patient cooperation. In hematological patients, such as patients with thrombocytopenia and coagulation disorders, these procedures may notably increase the risk of complications associated with invasive diagnostic interventions, and when sedation or general anesthesia is required, patients are exposed to additional risks. The literature demonstrates that a total of 27.8% of patients underwent surgery or thoracoscopic lung biopsy. Therefore, non-invasive tools to improve the differentiation of unclear pulmonary lesions are desirable.

Regarding diagnostic choices for the included cases from the literatures, the majority of patients underwent biopsy procedures, including repeated transbronchial lung biopsy, surgical lung biopsies and autopsies (Table V). Immunohistochemical analysis of BALF was conducted in only 5 patients. Of these, 4 patients were directly diagnosed through BALF immunohistochemistry, while the remaining patient still required surgical biopsy for definitive diagnosis of SPL. This corresponds with a diagnostic success rate of only 80% for BALF immunohistochemistry. Biopsy is the well-established standard for diagnosis, but if lymphoma cells can be identified through simple flow cytometry-based immunohistochemistry of BALF, this would potentially spare patients from the discomfort and risks associated with invasive biopsies. For the 2 patients in the present report who did not undergo biopsy, the lymphoma clinicians had initially planned CT-guided lung biopsies. However, both patients exhibited marked coagulation abnormalities and low oxygenation indices, rendering the procedure too high-risk. Ultimately, the clinicians decided against biopsy; therefore, to confirm the diagnosis, alternative approaches had to be explored. Both of the cases presented in the present study happened to produce large amounts of daily secretion from the lungs, prompting the collection of BALF for relevant diagnostic tests.

Flow cytometry analysis (Table SII) of the alveolar cell composition in BALF, particularly the percentages of CD4⁺ and CD8⁺ T cells, has long been applied in the diagnosis of diseases, including hypersensitivity pneumonitis, idiopathic non-specific interstitial pneumonia, idiopathic pleuroparenchymal fibroelastosis and unclassifiable idiopathic interstitial pneumonia (26). It may also predict the progression and prognosis of fibrotic diseases such as interstitial lung disease (27). Additionally, literature has reported that flow cytometry-based analysis of inflammatory cells in BALF can be indirectly used to assess the degree of lung tissue damage and macrophage subtyping to evaluate the pulmonary immune status (28). The use of flow cytometry depends on the selection of antigens. In the majority of lymphoproliferative disorders, lymphocyte subset analysis of BALF demonstrates a T cell predominance (>90%) with fewer B cell populations (<10%). Of note, lymphocytic alveolitis is a characteristic feature of pulmonary lymphoma, which can be examined by BALF cytological analysis (29). Particularly in pulmonary B cell lymphomas, flow cytometry of BALF has revealed both a notable elevation in B cell proportion (>10%) and immunophenotypic evidence of clonal proliferation, which may contribute to a diagnosis (30,31). Similar findings have been observed in pulmonary T cell lymphomas (32).

The basic principle of flow cytometry has been described for multiple purposes, including MRD detection, primarily depending on the selection of antibodies labelled with immunofluorescence (33). MRD testing is a diagnostic tool for patients with hematological diseases, mainly those with acute myeloid leukemia (AML). It refers to the small number of residual leukemia cells in patients with newly diagnosed or refractory/relapsed disease after achieving complete hematological remission through treatment. This encompasses the specific expertise in multiparameter flow cytometry (MFC)-based MRD, molecular MRD, NGS and clinical considerations. Methods for MRD detection using MFC include leukemia-associated immunophenotype (LAIP) and ‘different from normal’ (D-F-N) approaches. The former involves identifying the LAIP of the patient at initial diagnosis and using it for subsequent MRD monitoring during treatment. The latter is applicable to patients lacking an initial LAIP and can also detect antigen shift occurring during treatment. The combination of LAIP and D-F-N is suitable for both patients without an initial LAIP and for detecting newly emerging phenotypic abnormalities or antigen shift of existing LAIP (34). With regard to antibody selection, experts from diagnostic groups (34) recommend using MFC with ≥8-color fluorescence labeling for MRD detection to improve specificity. The antibody panels for MRD detection include: i) Core antibodies CD45, CD117, CD34, CD13 and CD33; ii) other antibodies including CD4, CD11b, CD14, CD64 and human leukocyte antigen-DR to assess MRD in monocytic or myelomonocytic AML; iii) CD7, CD19 and CD56 to evaluate cross-lineage antigen expression; and iv) CD133, CD38 and CD123 to detect leukemia stem/progenitor cells. A number of different marker panels have been used to assess MRD. To minimize the number of panels employed, experts recommend (35) the design and validation of a single common panel assay for all MRD studies. Each center should select the appropriate antibody panel for MRD detection based on the disease subtype. For the 2 cases outlined in the present report, both patients underwent bone marrow flow immunophenotyping prior to BALF MRD detection. Therefore, the BALF MRD detection in both cases was performed using fluorochrome-conjugated antibodies specifically tailored to their baseline B cell or T cell lineage.

Among the numerous cases of SPL diagnosed through BALF reported in the literature, only 1 case was found to be confirmed using flow cytometry of BALF, which showed λ-restricted B cells with dim CD20 expression and co-expression of CD5 and CD23, consistent with chronic lymphocytic leukemia. The majority of cases are still diagnosed based on the fractionation results of lymphocytes and neutrophils, cytological findings, combined with clinical characteristics and other biopsy data. Therefore, to the best of our knowledge, the present study was the first to report an SPL diagnosis that relied on MRD testing in BALF. The premise of this examination was that there must be sufficient living cells in BALF for detection. Overall, the minimum number of cells needed for accurate reporting of MRD is 500,000-1,000,000 excluding CD45^- cells and debris (35). These high numbers enable the assessment of possible MRD <0.1%. Furthermore, except for the definite MRD cell population identified during specimen collection, other residual leukemia cell subsets must be confirmed by skilled operators or personnel with professional knowledge of flow cytometry. However, in the cases outlined in the preset report, only 140,000 nucleated cells were found in Case 2 and even fewer in Case 1, but a positive status was still exhibited based on the 0.1% threshold.

To provide guidance for early detection and diagnosis, the present report recommends the following clinical strategies: i) In patients with a prior confirmed diagnosis of NHL, when respiratory symptoms emerge, the possibility of secondary pulmonary infection and SPL should both be considered in the differential diagnosis; ii) when chest CT reveals pneumonia-like changes, differentiating SPL from pneumonia is challenging, therefore a comprehensive evaluation incorporating bronchoalveolar lavage, serum testing and tissue biopsy is advised when necessary; and iii) in certain cases, if pathological tissue biopsy cannot be performed, MRD detection of BALF can serve as an alternative diagnostic method.

The present study has certain limitations. First, only 2 cases were reported in this study. From these, 1 patient did not receive standardized treatment due to personal reasons, thereby weakening the reliability and statistical power of the study results. Second, the generalizability of the study conclusions is limited as MRD detection in BALF requires collecting a sufficient number of cells for staining, and most patients may not meet this criterion. Third, this study excluded cases before the 21st century, which may introduce temporal selection bias. However, cases before the 21st century are relatively rare, which may mitigate this type of bias.