Characteristics of pulmonary lymphoma on baseline ¹⁸F-FDG positron emission tomography/computed tomography and their clinical value in predicting treatment response

Introduction

Pulmonary lymphomas can be classified according to site of origin into primary pulmonary lymphomas (PPLs) and secondary pulmonary lymphomas (SPLs) (1-4), with distinctions evident not only between these two categories but also across their respective histological subtypes. PPL is defined as a clonal lymphoproliferative neoplasm—excluding hilar lymph nodes—originating from the lungs, with no evidence of extrapulmonary infiltration for ≥3 months following initial clinical, radiological, or pathological diagnosis (1-7). It is a relatively rare condition, accounting for only approximately 0.5–1% of all primary lung tumors and 3–4% of extranodal primary non-Hodgkin lymphoma (NHL) cases (3-7). The most common subtype is mucosa-associated lymphoid tissue (MALT) lymphoma, which typically presents as nodules or consolidations with mild metabolic activity; the second most frequent subtype, diffuse large B-cell lymphoma (DLBCL), often manifests as nodules or masses, with central liquefactive necrosis observed in approximately half of the cases (4-7). SPL, which results from the pulmonary infiltration of systemic lymphoma, is the more common type. NHL constitutes 80–90% of all lymphomas, with intrathoracic involvement occurring in 50% of cases and pulmonary parenchymal involvement in 24%; meanwhile, Hodgkin lymphoma (HL) accounts for 10–20% of cases, with 85% involving the intrathoracic region and 38% affecting the lungs (3,4). Imaging features are nonspecific and may include nodules, masses, consolidations, cystic changes, lymphadenopathy, pleural effusion, and ground-glass opacities (GGOs) (1,4,8,9). Given this considerable imaging variability, a thorough understanding of the subtypes and imaging characteristics of pulmonary lymphoma is essential to avoiding misinterpretation.

Furthermore, the heterogeneous imaging manifestations of pulmonary lymphoma may complicate prognostic assessment. The evidence gathered thus far indicates that PPL and SPL differ in their clinical behavior and outcomes, with PPL associated with a comparatively favorable prognosis (10). In SPL, establishing the absence of an extrapulmonary primary site alongside confirmation of pulmonary infiltration is critical for staging, as such cases can be classified as stage IV according to the Lugano staging system and accrue an additional point within the International Prognostic Index (IPI), both of which are indicative of a potentially adverse prognosis (11,12). Regarding diagnostic biopsy, superficial lymph nodes are often preferred over pulmonary lesions due to the inherent risks of complications—such as pneumothorax following lung puncture—supporting the use of imaging in assisting lung assessment.

Whole-body positron emission tomography/computed tomography (PET/CT) with ¹⁸F-fludeoxyglucose (¹⁸F-FDG) tracer represents a crucial noninvasive imaging modality for the staging and response assessment in lymphoma. This technique is integral to the Lugano classification, enabling precise localization of lesions and the semiquantitative evaluation of metabolic activity via standardized uptake value (SUV) measurements (11,13,14).

Although previous research has sought to elucidate the relationships between PET/CT features and prognosis (15-18), the majority of these studies have been limited to common subtypes or have approached pulmonary involvement merely as part of systemic hematological disease. Consequently, a comprehensive investigation focused specifically on pulmonary lymphoma from a thoracic imaging perspective remains lacking. Earlier CT-based studies have reported correlations between imaging signs—such as halo sign, pleural involvement, number of lesions, and cross-lobe invasion—and progression-free survival (PFS) or overall survival (OS) (19), suggesting the potential value of expanded research on PET/CT. It is therefore essential to continually refine the understanding of imaging characteristics to enhance clinical utility. This study aimed to analyze the subtypes, PET/CT imaging findings, and corresponding treatment responses in pulmonary lymphoma, thereby contributing to a more comprehensive understanding of this entity. We present this article in accordance with the STROBE reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2869/rc).

Methods

Patients

This retrospective study was approved by the Institutional Review Board of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (2024 Clinical Ethics Review No. 70), and was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The requirement for written informed consent was waived by the Institutional Review Board due to the retrospective nature of the analysis.

This study retrospectively enrolled patients who underwent PET/CT examinations in the Department of Nuclear Medicine of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, between January 2015 and November 2023. The inclusion criteria were as follows: (I) adult patients with either primary lymphoma or relapse occurring >6 months after prior treatment; (II) availability of baseline assessment and PET/CT imaging showing suspected pulmonary lesions; (III) pathological confirmation of lymphoma obtained from Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (our institution) or another hospital within 3 months of the PET/CT examination; (IV) lymphoma diagnosis confirmed via bronchoscopy, lung puncture, or surgical biopsy; and (V) lymphoma diagnosed through lymph node or other extranodal biopsies, with pulmonary lesions considered representative of lymphomatous involvement based on clinical and follow-up imaging evaluations.

Classification into PPL and SPL groups was conducted as per the established diagnostic criteria derived from previous studies (1-7,9,10,19). (I) PPL cases were defined as those presenting with disease confined exclusively to the lungs (with or without hilar nodes) at initial diagnosis; pathologically confirmed via bronchoscopy, lung biopsy, or surgical resection; and with no extrapulmonary lesions detected during ≥3-month follow-up. (II) SPL cases were defined as those exhibiting extrapulmonary involvement confirmed within 3 months of diagnosis through pulmonary, lymph node, or other extranodal biopsies, with pulmonary lesions being identified as infiltration via histopathological evaluation and serial imaging review.

PET/CT image acquisition method

Patients fasted for ≥6 hours prior to the examination and received an intravenous injection of ¹⁸F-FDG (4–13 mCi, 148–481 MBq). With the patient in a supine position and arms elevated, images were acquired 60–90 minutes after injection with one of four devices (acquisition protocols detailed in Appendix 1, Table S1). Whole-body PET/CT scans extended from the skull to the thigh, and in some cases, total-body PET/CT was applied and extended from the hands to the ankles. Following the PET/CT acquisition, a breath-hold CT scan of the chest was obtained.

Image interpretation method

Image review and measurement of relevant parameters were performed via LIFEx version 7.4.0 software (Institut Curie, Paris, France) (20). Assessments were based on baseline examinations. Three independent observers interpreted the images, and any discrepancies were resolved by a senior radiologist with over 10 years of experience. The imaging findings were recorded based on the baseline examinations.

Lesion count, location, and morphology were recorded in accordance with standardized imaging reporting guidelines (21). Anatomical categories included nodule/mass (soft tissue opacity ≤3 or >3 cm), consolidation (dense parenchymal opacities due to alveolar infiltration), GGOs (area of hazy increased attenuation), air bronchogram (air-filled bronchi within consolidated lung), cavities (necrotic lesion with air-filled space), pleural involvement (pleural thickening or retraction), pleural effusion, and enlargement of hilar lymph nodes.

SUVs based on body weight (SUV_bw) were measured. Volumes of interest (VOIs) were delineated manually on chest CT-based anatomical contours. Reference values included the maximum SUV (SUV_max) of the liver blood pool and the mediastinal blood pool, which were obtained from a VOI with a 3-cm diameter placed in the right lobe of the liver and a VOI with a 1-cm diameter placed within the descending thoracic aorta, respectively. Target-to-background ratios [TBRs, target-to-mediastinal blood pool ratio (TBR_blood) and target-to-hepatic blood pool ratio (TBR_liver)] were subsequently calculated. Homogeneous metabolic uptake was defined as uniform metabolism throughout each lesion, while heterogeneous uptake was defined as the presence of both uniform and nonuniform metabolic lesions. Hilar lymph nodes were considered metabolically elevated when their activity substantially exceeded that of the blood pool. Representative examples of these features are illustrated in Figure 1.

Figure 1 Representative cases of PET/CT imaging features with arrow annotations. (A) A 68-year-old male with secondary pulmonary DLBCL. (B,C) A 60-year-old male with secondary pulmonary DLBCL. (D,E) A 58-year-old female with secondary pulmonary MALT lymphoma. (F) A 67-year-old male with secondary pulmonary PTCL. (G) A 45-year-old female with secondary pulmonary DLBCL. DLBCL, diffuse large B-cell lymphoma; FDG, fludeoxyglucose; GGO, ground-glass opacity; MALT, mucosa-associated lymphoid tissue; PET/CT, positron emission tomography/computed tomography; PTCL, peripheral T-cell lymphoma.

For follow-up studies, treatment response was assessed via the Deauville 5-point scale (5-PS), and assessment of Lugano classification criteria for treatment response was based on the most recent follow-up examination (11,22-24). In cases of discrepancy between metabolic and anatomical assessments, the 2023 PRoLoG (PINTad Response Criteria in Lymphoma Working Group) consensus guidelines were applied (24); otherwise, the worse rating was retained.

Statistical analysis

Data management and statistical analysis were performed with SPSS version 26.0 (IBM Corp., Armonk, NY, USA). Categorical variables, such as CT imaging features, were compared with the Chi-squared test, while continuous variables, including glycemia and SUV metrics, were assessed with the t-test. Given the predominance of categorical data, group comparisons involving treatment response and imaging characteristics were conducted with either the Chi-squared test or the Fisher’s exact test (the latter being applied when sample size <40 or >20% of cells had an expected count of less than 5). Associations between categorical and continuous SUV-based variables were evaluated via Spearman correlation coefficients.

Results

A total of 136 cases (66 males and 70 females) with pulmonary lymphoma involvement were included, with a mean age of 54.84±16.66 years. Diagnosis was confirmed via direct pulmonary biopsy in 69 cases (50.74%) and by lymph node or other extrapulmonary biopsy with subsequent follow-up confirmation in the other 67 cases (49.26%) (Table 1).

Subtypes and proportions of pulmonary lymphoma

All cases of PPL were NHL, with the majority being MALT lymphoma (31/35, 88.57%), while DLBCL accounted for only 4 cases (11.43%).

Among SPLs, 19 (18.81%) were HL, all of which were classical HL (cHL). These included 6 cases of mixed-cellularity HL (MCHL) and 12 cases of nodular sclerosis HL (NSHL). One cHL case, in which puncture was applied in other institution, lacked further subclassification. Of the 82 SPL cases of NHL origin, T-cell lymphomas constituted 20.73% (17/82), with natural killer T-cell lymphoma (NKTL) being the most prevalent (7/17, 41.18%). B-cell lymphomas were the most common subtype; DLBCL accounted for 78.46% (51/65) of B-cell cases and 50.50% (51/101) of all SPLs and was followed by follicular lymphoma (FL) and MALT lymphoma in incidence (Table 2).

Baseline PET/CT features and their association with pulmonary lymphoma subtype

Pulmonary lymphoma typically manifested as bilateral, multifocal, and hypermetabolic lesions, most frequently presenting as solid nodular masses or large consolidative patches. Among the cases, nodules were observed in 92 (67.65%), consolidation in 60 (44.12%), and large mass-like consolidations in another 60 (44.12%). Meanwhile, GGO was observed in 23 (16.91%) cases.

PPL was mainly characterized by consolidation (68.57%) and frequently accompanied by air bronchogram (62.86%). In contrast, SPL more commonly presented with nodules (76.24%), often demonstrating homogeneous hypermetabolism (77.23%) and increased metabolic activity in hilar lymph nodes (77.23%). Statistically significant distinctions between SPL and PPL were observed in terms of lesion size, hilar lymph node involvement, presence of consolidation and air bronchogram signs, distribution of FDG uptake, and TBRs (P<0.05; Table 3 and Figure 2).

Figure 2 Heatmap illustrating difference in the prevalence of PET and CT manifestations between SPL and PPL. CT, computed tomography; FDG, fludeoxyglucose; SPL, secondary pulmonary lymphoma; PET, positron emission tomography; PPL, primary pulmonary lymphoma.

Additionally, a notable difference was also observed in liver blood pool background metabolism between SPL and PPL (P<0.05), which may be attributable to variations between the acquisition systems. This underscores the importance of employing relative quantitative measures, such as TBR.

Metabolic characteristics

Lymphomatous pulmonary lesions were typically FDG-avid. SPLs demonstrated a significantly higher TBR_blood as compared to PPLs (P=0.038). Consistent and comparable values were observed for both SUV_max and TBR among the MALT-derived PPL, cHL, angioimmunoblastic T-cell lymphoma, anaplastic large-cell lymphoma, and T-cell lymphoblastic lymphoma subtypes, while considerable metabolic heterogeneity was observed between the other subtypes (Figure 3). However, the limited number of cases for certain subtypes precludes definitive conclusions (see Appendix 2, Table S2 for details).

Figure 3 Scatter and box plots illustrating metabolic parameters across different lymphoma subtypes. (A) SUV_max. (B) TBR_blood. (C) TBR_liver. −, negative; +, positive. AITL, angioimmunoblastic T-cell lymphoma; ALCL, anaplastic large-cell lymphoma; ALK, anaplastic lymphoma kinase; cHL, classical Hodgkin lymphoma; CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; MALT, mucosa-associated lymphoid tissue; MCL, mantle cell lymphoma; NKTL, natural killer T-cell lymphoma; NMZL, nodal marginal zone lymphoma; PPL, primary pulmonary lymphoma; PTCL, peripheral T-cell lymphoma; SPL, secondary pulmonary lymphoma; SUV_max, maximum standard uptake value; TBR_blood, target-to-mediastinal blood pool ratio; TBR_liver, target-to-hepatic blood pool ratio; TLBL, T-cell lymphoblastic lymphoma.

A notable deviation from the expected metabolic behavior was observed in one case of chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL) with 6-year follow-up (Figure 4). A hypermetabolic paraspinal mass was identified on PET/CT in the June 2023 scan, yet pathological analysis from a May 2023 biopsy reconfirmed the original CLL/SLL diagnosis.

Figure 4 PET/CT images from a 54-year-old male with an FDG-avid CLL/SLL lesion. (A) CT image showing a soft-tissue hemispherical mass in the right paravertebral region. (B) Fused image showing unusually high FDG avidity (SUV_max 31.01) for this subtype. CLL/SLL, chronic lymphocytic leukemia/small lymphocytic lymphoma; CT, computed tomography; FDG, fludeoxyglucose; PET, positron emission tomography; SUV_max, maximum standard uptake value.

Nodules and masses

The presence of nodules was the most common imaging feature of pulmonary lymphoma infiltration, particularly in cases with multiple lesions. Among PPL cases, 15 (42.86%) presented with nodules, including 1 case with a solitary nodule, 5 with multiple nodules, and 9 with consolidations and nodules. For SPLs, nodules were observed in 76.24% of cases, including 15 with a single nodule and 62 with multiple nodules or nodules coexisting with other manifestations (Figure 5A,5B). These were most frequently associated with DLBCL (39/77), followed by cHL (14/77) and NKTL (6/77). Masses were less common (35.29%) and were also primarily linked to DLBCL (23/48), with smaller proportions observed in cases of MALT (8/48), cHL (6/48), and NKTL (5/48).

Figure 5 Representative examples of PET/CT imaging features. (A,B) PET/CT images from a 20-year-old female with NKTL. (A) CT image showing a subpleural nodule in left lower lobe and a mass in right lower lobe. (B) Fused image showing homogeneous hypermetabolism in the nodule (SUV_max 12.89) and heterogeneously hypermetabolism in the mass (SUV_max 24.51), suggestive of central necrosis. (C,D) PET/CT images from a 50-year-old male with MALT lymphoma. (C) CT image showing a cavitated mass-like consolidation in the left upper lobe and consolidation with air bronchogram in the anterior segment of the right upper lobe. (D) Fused image showing mildly heterogeneous hypermetabolism in both lesions (left SUV_max: 7.58; right SUV_max: 4.32). (E,F) PET/CT images from a 62-year-old male with MALT lymphoma. (E) CT image showing a soft tissue density opacity with nodular pleural thickening in the right middle lung, accompanied by bilateral pleural effusion. (F) Fused image showing high FDG uptake (SUV_max 24.13) within the soft-tissue lesion and moderate uptake in the area of pleural thickening. CT, computed tomography; FDG, fludeoxyglucose; MALT, mucosa-associated lymphoid tissue; NKTL, natural killer T-cell lymphoma; PET, positron emission tomography; SUV_max, maximum standard uptake value.

Consolidation

Consolidation was the most common imaging finding in PPL. Among 35 PPL cases, 24 (68.57%) exhibited pulmonary consolidation, frequently accompanied by air bronchogram (21/24). The majority were MALT lymphomas (31/35), among which 20 cases presented as consolidations, 19 of these demonstrating air bronchogram.

GGO

GGO was an uncommon feature, observed in only 23 (16.91%) cases. MALT lymphoma was the most frequent subtype (11/23), followed by DLBCL (5/23). All 8 PPL cases presenting with GGOs were diagnosed as MALT lymphoma.

Cavitation

Cavitation was a rare finding, occurring in only 8 (5.88%) cases, all within masses or mass-like consolidations (Figure 5C,5D). SPL accounted for the majority (6/8), with DLBCL being the most frequent subtype (3/8).

Pleural involvement and effusion

Pleural involvement was observed in 86 (63.24%) cases. Furthermore, the 29 (21.32%) patients had varying degrees of pleural effusion, the majority of whom also showed pleural involvement (24/26, 92.31%) (Figure 5E,5F). Lymphoma cells were cytologically confirmed in the pleural fluid in 6 of these cases.

Hilar lymph node enlargement and hypermetabolism

Pulmonary lymphoma infiltration was frequently accompanied by reactive or infiltrative hilar lymph node involvement. Among 136 patients, hilar lymphadenopathy was identified in 68 (50%), particularly among those with SPL (62/101, 61.4%). Increased metabolic activity was observed in 88 (64.7%) cases, again predominantly in SPLs (78/101, 77.2%).

Among the 93 cases with hilar nodal abnormalities, 63 (67.7%) exhibited both enlargement and hypermetabolism. However, these features were not necessarily concurrent. Among the remaining cases, 25 (28.4%) showed hypermetabolism without significant morphological changes, while 5 (7.4%) had slightly enlarged nodes without significant hypermetabolism.

Correlation between baseline PET/CT features and treatment response

Among the cases of pathologically confirmed pulmonary lymphoma, 32 lacked follow-up PET/CT at our institution after baseline examination, including 14 PPLs and 18 SPL cases. After these were excluded, 104 cases (21 PPL and 83 SPL) were included in the treatment response analysis (treatment regimens are summarized in Appendix 3, Table S3). To address therapeutic heterogeneity, subgroup analysis was conducted. As all patients with PPL received conventional chemotherapy, targeted therapy, or a combination of chemotherapy and targeted therapy (without novel agents), no further stratification was required. In contrast, SPL cases and the overall pulmonary lymphoma cohort were stratified into conventional therapy and immunotherapy subgroups (25).

The Chi-squared and Fisher’s exact tests for categorical variables (P>0.05) indicated the following preliminary findings regarding treatment response in PPL and SPL: Among patients receiving conventional pharmacological therapies, neither primary nor secondary classification showed a statistically significant association with treatment response. Among the patients with SPL, no significant difference was observed between those treated with conventional pharmacological therapy and those treated with immunotherapy. Overall, treatment response did not significantly differ according to primary or secondary classification of pulmonary lymphoma (Appendix 3, Table S4).

The association between baseline imaging features and posttreatment 5-PS scores or Lugano response categories was evaluated with intergroup Chi-squared or Fisher’s exact tests (Appendix 3, Tables S5,S6). In the PPL group, 5-PS scores were significantly associated with nodular lesions (P=0.036). Among all patients with SPL, 5-PS scores correlated significantly with FDG distribution patterns (P=0.007), a finding that remained significant in the conventional therapy subgroup (P=0.002). In all patients with pulmonary lymphoma receiving conventional pharmacological therapy, 5-PS scores were significantly associated with the presence of nodular lesions (P=0.014), consolidations (P=0.032), and FDG distribution (P=0.025), and these three features were also significantly associated with 5-PS scores in the overall cohort (P=0.017, P=0.044, and P=0.010, respectively). Additionally, the Lugano response assessment was significantly associated with pleural involvement in PPL cases (P=0.001).

As correlation does not simply imply causation, ordered categorical variables were consolidated for the evaluation of potential causal relationships. The 5-PS scores were dichotomized into a lower group (score of 1–3: metabolism ≤ liver blood pool) and a higher group (score of 4–5: metabolism > liver blood pool). Similarly, response grades were categorized as complete remission/partial remission (CR/PR) or as stable disease/progressive disease (SD/PD). Upon re-evaluation, only pleural involvement in PPL remained significantly associated with Lugano response assessment (Table 4). All cases with pleural involvement achieved CR/PR, whereas 71.43% (5/7) of those without pleural involvement experienced SD/PD, suggesting a more favorable treatment response when pleural involvement is present. However, due to the limited sample size (n=21), this conclusion requires further validation. Other initially significant variables did not retain significance after categorical consolidation.

Spearman rank correlation was used to assess the relationships between continuous PET parameters and ordinal categorical variables (5-PS scores and Lugano response). A weak positive correlation was observed between baseline TBR_blood and 5-PS score in SPL (ρ=0.236; P=0.032). No significant correlations were found between SUV_max or other TBR values and the treatment response in the PPL and SPL subgroups (stratified by treatment type) or the overall cohort (Table 5).

Due to the metabolic and morphological heterogeneity across the different lymphoma subtypes, subgroup analyses were performed for those subtypes with a sufficient sample size (Appendix 3, Tables S7-S10). No significant associations were observed in HL. Overall, the findings remain consistent with those derived from the overall cohort analysis without subtype stratification. This may be explained by the predominance of DLBCL and MALT lymphoma within the SPL and PPL cohorts, respectively. However, the results indicated no significant correlation between SUV_max or other TBR values and treatment response in any of the subtypes examined, which contrasts the results from SPL. This discrepancy may be attributed to confounding factors introduced by other SPL subtypes or, alternatively, to the reduced sample size, which may have obscured potentially weak correlations that might have otherwise been detectable.

Discussion

This study analyzed the characteristics of pulmonary lymphoma, including its subtypes, relative frequency, imaging features, and treatment response.

With regard to classification and lymphoma subtypes, PPL is relatively uncommon, with fewer than 500 cases reported in the literature (3,26,27), accounting for less than 1% of all lymphomas (4,28,29). Our findings indicate that PPL constituted approximately 25.73% of lung lymphoma cases in this cohort, while SPL accounted for 74.26%. To enhance data reliability, suspected SPL cases lacking definitive pathological confirmation or long-term follow-up were excluded. Consequently, the actual proportion of SPL might have been higher. According to the World Health Organization, PPL is classified into three main subtypes, which, in descending order of prevalence, are MALT lymphoma, DLBCL, and lymphomatoid granulomatosis (2,5-7,22,23). MALT lymphoma is the most common subtype, constituting 70–90% of all PPL cases (5-7,30). Our results align with this, with MALT lymphoma accounting for 88.57% of PPL cases and DLBCL for 11.43%. SPL demonstrates a greater degree of diversity in its subtypes. Although pulmonary involvement is more frequently observed in HL than in NHL, the overall number of NHL cases is higher due to its greater incidence (4,26). DLBCL is the most common NHL subtype, representing 31% of cases (31). In our cohort, HL accounted for 13.97% of SPL cases and T-cell lymphoma for 12.50%, with the remainder comprising B-cell lymphomas—predominantly DLBCL—which is consistent with established distributions reported in the literature (3,4,31).

The imaging manifestations of pulmonary lymphomas demonstrated considerable heterogeneity. The majority of PPL cases, likely reflecting the high proportion of MALT lymphoma, predominantly presented as consolidation with air bronchogram and moderate FDG avidity, while nodules or GGO were less frequently observed. These findings are consistent with previous studies (1-7). In contrast, the imaging characteristics of SPL are less specific and vary depending on the site of involvement and the underlying lymphoma subtype. The most common CT features include lymphadenopathy and nodules distributed along lymphatic pathways (4,26). In our study, the majority of SPLs appeared with multiple hypermetabolic nodules, often accompanied by increased metabolic activity in hilar lymph nodes, irrespective of whether structural lymphadenopathy was present. This pattern may reflect the progressive invasive relationship between hilar nodal involvement and pulmonary infiltration in SPL. With the exception of MALT-derived PPL and cHL, notable differences in metabolic activity were observed across the subtypes, whereas the results for other subtypes remained inconclusive. The mean SUV_max of PPL-MALT lymphoma in this study was 7.50±5.11, which is comparable to previously reported ranges (32-35). Nevertheless, these imaging and metabolic features alone remain insufficient for reliably distinguishing pulmonary lymphomas from other pulmonary pathologies. For instance, as illustrated in Appendix 4, Table S11, there was considerable overlap between the imaging phenotypes of organizing pneumonia and PPL, as well as between pulmonary adenocarcinoma and SPL. Consequently, there is a need to develop more robust diagnostic approaches to improve the identification of pulmonary lymphoma and to differentiate multifactorial pulmonary.

With regard to treatment response, the IPI and MALT-IPI prognostic models were created several years ago, and ongoing research has led to the development of the revised IPI model and the establishment of β₂-microglobulin as a prognostic marker (12,36-38), indicating the continued emergence of novel biomarkers with potentially higher predictive utility. Although our findings are not conclusive, they contribute insights into the relationship between PET/CT imaging features and treatment response in pulmonary lymphoma. Specifically, the 5-PS score was statistically associated with several factors: the presence of nodular lesions in the PPL group, among pulmonary lymphoma patients receiving conventional therapies, and across the overall pulmonary lymphoma cohort; FDG distribution patterns within both the entire SPL group and SPL subgroup treated with conventional pharmacological therapies; and the presence of consolidations and FDG distribution patterns among conventional therapy recipients and the overall cohort (P<0.05). However, these associations lost statistical significance when contiguous categories were merged, likely due to the loss of information or limited statistical power inherent in the small sample size.

Although the Lugano response assessment showed a statistically significant association with pleural involvement, our data indicated that such involvement appeared to correlate with a more favorable treatment response in PPL. This observation appears inconsistent with a previous study that reported worse OS in patients with B-cell NHL and pleural involvement (19). It has been reported that interleukin-2 receptor and interleukin-8 levels are elevated in patients diagnosed with MALT lymphoma, a condition frequently associated with chronic infection or autoimmunity, and some presenting with clinical symptoms suggestive of pneumonia (39-43). Given the relatively indolent behavior of these tumors, we hypothesize that the elevated metabolic activity observed in most MALT lymphomas is not exclusively attributable to tumor glycolytic activity but may also reflect underlying inflammatory processes. Consequently, following chemotherapy and symptomatic treatment, patients with more pronounced pulmonary inflammatory reactions, manifesting as pleural involvement and higher metabolic activity, may exhibit more substantial metabolic improvements on posttherapy imaging. This suggests that a significant imaging response may not necessarily translate into improved OS or PFS. Therefore, further verification is required to ascertain the clinical interpretability of these observations, and investigation beyond the 5-PS framework may help refine evaluation scales and scoring systems for indolent PPLs or lymphomas with similar inflammatory components.

Analysis of PET-derived measurements indicated a positive correlation between baseline TBR_blood and posttreatment 5-PS scores in SPL, suggesting that higher baseline TBR_blood values are associated with a reduced degree of metabolic remission. In contrast, neither TBR_liver nor lesion SUV_max demonstrated a significant correlation with treatment response in the other subgroups, and the significance of TBR_blood was not retained in the subgroup analysis within SPL. The related literature indicates that the physiological mediastinal blood pool and liver exhibit relatively stable metabolic levels, with hepatic SUV normalized to lean body mass demonstrating slightly lower interindividual variability (44,45). However, intrapatient metabolic fluctuations in the hepatic blood pool of oncological patients between scans are more pronounced, which is potentially associated with previous chemotherapy (46). Additionally, studies have confirmed that hepatic steatosis can lead to a mild increase in hepatic FDG uptake, while chemotherapy can induce both inflammatory-reactive and steatosis-related hepatotoxicity (47-50). Our results also identified statistically significant heterogeneity in hepatic SUV between the PPL and SPL cohorts (Table 3). This variability may reflect elevated hepatic metabolic activity in relapsed SPL patients scanned >6 months after therapy and the effect of residual drugs, which could have artificially attenuated the differences in TBR_liver. In comparison, TBR_blood may be a more consistent indicator of tumor glycolytic activity against the background of circulating blood pool activity, exhibiting superior stability due to less interference. Meanwhile, the correlation between treatment response and baseline lung lesion characteristics in lymphoma with pulmonary infiltration may be obscured by the fact that pulmonary lesions alone may not fully represent the systemic disease or primary disease site in SPL. Consequently, the discrepancy in outcomes observed between TBR_blood and TBR_liver preclude definitive conclusions and require further validation.

This study integrated data from four scanning devices over an 8-year period, helping to reduce potential single-center bias. Although diagnosis of SPL prioritizes superficial sites such as subcutaneous lymph nodes, our study incorporated pulmonary lesions and treatment response assessments confirmed through long-term follow-up, providing a comprehensive characterization of subtype features, imaging manifestations, and treatment outcomes. Although this approach may introduce selection bias, such limitations are inherent to retrospective designs. Our study involved other limitations. First, certain subtypes are relatively rare, and large-scale datasets with systematic follow-up are currently lacking. Our sample, particularly within the PPL category, was limited. Consequently, the reliability of correlation analyses may be insufficient, and even small fluctuations in case numbers within subgroups during analyses stratified by treatment efficacy could lead to pronounced shifts in observation. The findings, which appear somewhat counterintuitive and diverge from previous studies, are nevertheless presented with respect to the factual results to maintain the study’s integrity and facilitate further discussion. Second, due to the constraints of the retrospective design, heterogeneity in treatment regimens might have confounded the treatment response analysis. Although treatments were categorized according to major therapeutic classes to address this, residual confounding might have contributed to outcome variability. Both limitations likely influenced the initial Chi-squared analyses, in which statistical significance frequently disappeared when adjacent categories were merged. In future research, we aim to adopt more refined inclusion and stratification criteria to clarify variable associations or employ meta-analyses to synthesize data from multiple studies, thereby yielding more robust conclusions.

Conclusions

This study summarized the PET/CT imaging characteristics of pulmonary lymphoma and investigated their association with treatment response. We found that PPLs and SPLs exhibit distinct, albeit nonspecific, PET/CT manifestations, which provide preliminary but inconclusive discriminatory value. Certain features, including FDG distribution patterns and lesion morphology, were correlated with treatment response, although causal relationships and trends remain unestablished. Pleural involvement in PPL was associated with a more favorable treatment response, while elevated baseline TBR_blood in SPL correlated with reduced metabolic remission, suggesting that baseline PET/CT parameters may possess predictive utility, although this remains to be more thoroughly validated. To this end, it may be productive to apply machine learning methods and identify discriminative features between pulmonary lymphomas and other pulmonary pathologies (51) and to quantitatively capture subtle imaging features.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2869/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2869/dss

Funding: This work was supported by the National Natural Science Foundation of China (No. 82171976), the Shanghai Pujiang Program (Class D) (No. 21PJD042), the Shanghai “New Star of Medical Field” for Youth Medical Talents [2021], and the Shanghai Municipal Key Clinical Specialty (No. shslczdzk03403).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2869/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This retrospective study was approved by the Institutional Review Board of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine (2024 Clinical Ethics Review No. 70). Written informed consent was waived by the Institutional Review Board due to the retrospective nature of the analysis. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Hare SS, Souza CA, Bain G, Seely JM, Gomes MM, Quigley M. The radiological spectrum of pulmonary lymphoproliferative disease. Br J Radiol 2012;85:848-64. [Crossref] [PubMed]
2. Borie R, Wislez M, Antoine M, Cadranel J. Lymphoproliferative Disorders of the Lung. Respiration 2017;94:157-75. [Crossref] [PubMed]
3. William J, Variakojis D, Yeldandi A, Raparia K. Lymphoproliferative neoplasms of the lung: a review. Arch Pathol Lab Med 2013;137:382-91. [Crossref] [PubMed]
4. Angirish B, Sanghavi P, Jankharia B. Pulmonary manifestations of lymphoma: A pictorial essay. Lung India 2020;37:263-7. [Crossref] [PubMed]
5. Piña-Oviedo S, Weissferdt A, Kalhor N, Moran CA. Primary Pulmonary Lymphomas. Adv Anat Pathol 2015;22:355-75. [Crossref] [PubMed]
6. Wu T, Huang Y, Wang Z, Cao H, Ding Q, Deng Z. Pulmonary MALT lymphoma: Imaging findings in 18 cases and the associated pathological correlations. Am J Med Sci 2022;364:192-7. [Crossref] [PubMed]
7. Borie R, Wislez M, Antoine M, Copie-Bergman C, Thieblemont C, Cadranel J. Pulmonary mucosa-associated lymphoid tissue lymphoma revisited. Eur Respir J 2016;47:1244-60. [Crossref] [PubMed]
8. Carter BW, Wu CC, Khorashadi L, Godoy MC, de Groot PM, Abbott GF, Lichtenberger JP 3rd. Multimodality imaging of cardiothoracic lymphoma. Eur J Radiol 2014;83:1470-82. [Crossref] [PubMed]
9. Bligh MP, Borgaonkar JN, Burrell SC, MacDonald DA, Manos D. Spectrum of CT Findings in Thoracic Extranodal Non-Hodgkin Lymphoma. Radiographics 2017;37:439-61. [Crossref] [PubMed]
10. Zhang MC, Zhou M, Song Q, Wang S, Shi Q, Wang L, Yan FH, Qu JM, Zhao WL. Clinical features and outcomes of pulmonary lymphoma: A single center experience of 180 cases. Lung Cancer 2019;132:39-44. [Crossref] [PubMed]
11. Cheson BD, Fisher RI, Barrington SF, Cavalli F, Schwartz LH, Zucca E, et al. Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol 2014;32:3059-68. [Crossref] [PubMed]
12. Zhou Z, Sehn LH, Rademaker AW, Gordon LI, Lacasce AS, Crosby-Thompson A, Vanderplas A, Zelenetz AD, Abel GA, Rodriguez MA, Nademanee A, Kaminski MS, Czuczman MS, Millenson M, Niland J, Gascoyne RD, Connors JM, Friedberg JW, Winter JN. An enhanced International Prognostic Index (NCCN-IPI) for patients with diffuse large B-cell lymphoma treated in the rituximab era. Blood 2014;123:837-42. [Crossref] [PubMed]
13. Barrington SF, Kluge R. FDG PET for therapy monitoring in Hodgkin and non-Hodgkin lymphomas. Eur J Nucl Med Mol Imaging 2017;44:97-110. [Crossref] [PubMed]
14. Cheson BD, Meignan M. Current Role of Functional Imaging in the Management of Lymphoma. Curr Oncol Rep 2021;23:144. [Crossref] [PubMed]
15. Zanoni L, Bezzi D, Nanni C, Paccagnella A, Farina A, Broccoli A, Casadei B, Zinzani PL, Fanti S. PET/CT in Non-Hodgkin Lymphoma: An Update. Semin Nucl Med 2023;53:320-51. [Crossref] [PubMed]
16. Al-Ibraheem A, Mottaghy FM, Juweid ME. PET/CT in Hodgkin Lymphoma: An Update. Semin Nucl Med 2023;53:303-19. [Crossref] [PubMed]
17. Sun Z, Yang T, Ding C, Shi Y, Cheng L, Jia Q, Tao W. Clinical scoring systems, molecular subtypes and baseline [18F]FDG PET/CT image analysis for prognosis of diffuse large B-cell lymphoma. Cancer Imaging 2024;24:168.
18. Ding J, Cheng X, Wang H, Sun Y, Yang Y, Qi N, Jiang Y, Chen X, Meng Q, You Z, Jiang J, Zhao J. Prognosis prediction in non-Hodgkin lymphoma by assessing lymphoid organs uptake patterns using baseline (18)F-FDG PET/CT. Int J Cancer 2025;157:183-92. [Crossref] [PubMed]
19. Wang Y, Pan ZC, Zhu L, Ma YY, Zhang MC, Wang L, Zhao WL, Yan FH, Song Q. The characteristic computed tomography findings of pulmonary B-cell non-Hodgkin’s lymphoma and their role in predicting patient survival. Quant Imaging Med Surg 2021;11:772-83. [Crossref] [PubMed]
20. Nioche C, Orlhac F, Boughdad S, Reuzé S, Goya-Outi J, Robert C, Pellot-Barakat C, Soussan M, Frouin F, Buvat I. LIFEx: A Freeware for Radiomic Feature Calculation in Multimodality Imaging to Accelerate Advances in the Characterization of Tumor Heterogeneity. Cancer Res 2018;78:4786-9. [Crossref] [PubMed]
21. Zhou Q, Fan Y, Wang Y, Qiao Y, Wang G, Huang Y, Wang X, Wu N, Zhang G, Zheng X, Bu H. China National Guideline of Classification, Diagnosis and Treatment for Lung Nodules (2016 Version). Zhongguo Fei Ai Za Zhi 2016;19:793-8. Erratum in: Zhongguo Fei Ai Za Zhi 2023;26:558. [Crossref] [PubMed]
22. Barrington SF, Qian W, Somer EJ, Franceschetto A, Bagni B, Brun E, Almquist H, Loft A, Højgaard L, Federico M, Gallamini A, Smith P, Johnson P, Radford J, O'Doherty MJ. Concordance between four European centres of PET reporting criteria designed for use in multicentre trials in Hodgkin lymphoma. Eur J Nucl Med Mol Imaging 2010;37:1824-33. [Crossref] [PubMed]
23. Barrington SF, Mikhaeel NG, Kostakoglu L, Meignan M, Hutchings M, Müeller SP, Schwartz LH, Zucca E, Fisher RI, Trotman J, Hoekstra OS, Hicks RJ, O'Doherty MJ, Hustinx R, Biggi A, Cheson BD. Role of imaging in the staging and response assessment of lymphoma: consensus of the International Conference on Malignant Lymphomas Imaging Working Group. J Clin Oncol 2014;32:3048-58. [Crossref] [PubMed]
24. Ricard F, Cheson B, Barrington S, Trotman J, Schmid A, Brueggenwerth G, et al. Application of the Lugano Classification for Initial Evaluation, Staging, and Response Assessment of Hodgkin and Non-Hodgkin Lymphoma: The PRoLoG Consensus Initiative (Part 1-Clinical). J Nucl Med 2023;64:102-8. [Crossref] [PubMed]
25. Tang L, Huang Z, Mei H, Hu Y. Immunotherapy in hematologic malignancies: achievements, challenges and future prospects. Signal Transduct Target Ther 2023;8:306. [Crossref] [PubMed]
26. Gozzi L, Cozzi D, Cavigli E, Moroni C, Giannessi C, Zantonelli G, Smorchkova O, Ruzga R, Danti G, Bertelli E, Luzzi V, Pasini V, Miele V. Primary Lymphoproliferative Lung Diseases: Imaging and Multidisciplinary Approach. Diagnostics (Basel) 2023;13:1360. [Crossref] [PubMed]
27. Travis WD, Brambilla E, Burke AP, Marx A, Nicholson AG. Introduction to The 2015 World Health Organization Classification of Tumors of the Lung, Pleura, Thymus, and Heart. J Thorac Oncol 2015;10:1240-2. [Crossref] [PubMed]
28. Lee KS, Kim Y, Primack SL. Imaging of pulmonary lymphomas. AJR Am J Roentgenol 1997;168:339-45. [Crossref] [PubMed]
29. Wang L, Ye G, Liu Z, Shi L, Zhan C, Gu J, Luo R, Lin Z, Ge D, Wang Q. Clinical characteristics, diagnosis, treatment, and prognostic factors of pulmonary mucosa-associated lymphoid tissue-derived lymphoma. Cancer Med 2019;8:7660-8. [Crossref] [PubMed]
30. Borie R, Wislez M, Thabut G, Antoine M, Rabbat A, Couderc LJ, Monnet I, Nunes H, Blanc FX, Mal H, Bergeron A, Dusser D, Israël-Biet D, Crestani B, Cadranel J. Clinical characteristics and prognostic factors of pulmonary MALT lymphoma. Eur Respir J 2009;34:1408-16. [Crossref] [PubMed]
31. Martelli M, Ferreri AJ, Agostinelli C, Di Rocco A, Pfreundschuh M, Pileri SA. Diffuse large B-cell lymphoma. Crit Rev Oncol Hematol 2013;87:146-71. [Crossref] [PubMed]
32. Albano D, Borghesi A, Bosio G, Bertoli M, Maroldi R, Giubbini R, Bertagna F. Pulmonary mucosa-associated lymphoid tissue lymphoma: 18F-FDG PET/CT and CT findings in 28 patients. Br J Radiol 2017;90:20170311. [Crossref] [PubMed]
33. Beal KP, Yeung HW, Yahalom J. FDG-PET scanning for detection and staging of extranodal marginal zone lymphomas of the MALT type: a report of 42 cases. Ann Oncol 2005;16:473-80. [Crossref] [PubMed]
34. Alinari L, Castellucci P, Elstrom R, Ambrosini V, Stefoni V, Nanni C, Berkowitz A, Tani M, Farsad M, Franchi R, Fanti S, Zinzani PL. 18F-FDG PET in mucosa-associated lymphoid tissue (MALT) lymphoma. Leuk Lymphoma 2006;47:2096-101. [Crossref] [PubMed]
35. Zinzani PL, Pellegrini C, Gandolfi L, Casadei B, Derenzini E, Broccoli A, Quirini F, Argnani L, Pileri S, Celli M, Fanti S, Poletti V, Stefoni V, Baccarani M. Extranodal marginal zone B-cell lymphoma of the lung: experience with fludarabine and mitoxantrone-containing regimens. Hematol Oncol 2013;31:183-8. [Crossref] [PubMed]
36. Kiesewetter B, Raderer M. How can we assess and measure prognosis for MALT lymphoma? A review of current findings and strategies. Expert Rev Hematol 2021;14:391-9. [Crossref] [PubMed]
37. Kim HD, Cho H, Jeong H, Bang K, Kim S, Lee K, Kang EH, Park JS, Park CS, Huh J, Ryu JS, Lee SW, Yoon DH, Oh SY, Suh C. A prognostic index for extranodal marginal-zone lymphoma based on the mucosa-associated lymphoid tissue International Prognostic Index and serum β2-microglobulin levels. Br J Haematol 2021;193:307-15. [Crossref] [PubMed]
38. Alderuccio JP, Reis IM, Habermann TM, Link BK, Thieblemont C, Conconi A, Larson MC, Cascione L, Zhao W, Cerhan JR, Zucca E, Lossos IS. Revised MALT-IPI: A new predictive model that identifies high-risk patients with extranodal marginal zone lymphoma. Am J Hematol 2022;97:1529-37. [Crossref] [PubMed]
39. Zhong H, Chen J, Cheng S, Chen S, Shen R, Shi Q, Xu P, Huang H, Zhang M, Wang L, Wu D, Zhao W. Prognostic nomogram incorporating inflammatory cytokines for overall survival in patients with aggressive non-Hodgkin’s lymphoma. EBioMedicine 2019;41:167-74. [Crossref] [PubMed]
40. Miyata-Takata T, Takata K, Toji T, Goto N, Kasahara S, Takahashi T, Tari A, Noujima-Harada M, Miyata T, Sato Y, Yoshino T. Elevation of serum interleukins 8, 4, and 1β levels in patients with gastrointestinal low-grade B-cell lymphoma. Sci Rep 2015;5:18434. [Crossref] [PubMed]
41. Du MQ. MALT lymphoma: A paradigm of NF-κB dysregulation. Semin Cancer Biol 2016;39:49-60. [Crossref] [PubMed]
42. Iftikhar A, Magh A, Cheema MA, Thappa S, Sahni S, Karbowitz S. Primary pulmonary MALT lymphoma presenting as non-resolving pneumonia. Adv Respir Med 2017;85:202-5. [Crossref] [PubMed]
43. Rechal R, Prasad VP, Sethi S, Maturu VN. Non-resolving pneumonia: primary pulmonary MALT lymphoma. BMJ Case Rep 2024;17:e255075. [Crossref] [PubMed]
44. Wahl RL, Jacene H, Kasamon Y, Lodge MA. From RECIST to PERCIST: Evolving Considerations for PET response criteria in solid tumors. J Nucl Med 2009;50:122S-50S. [Crossref] [PubMed]
45. Paquet N, Albert A, Foidart J, Hustinx R. Within-patient variability of (18)F-FDG: standardized uptake values in normal tissues. J Nucl Med 2004;45:784-8.
46. Boktor RR, Walker G, Stacey R, Gledhill S, Pitman AG. Reference range for intrapatient variability in blood-pool and liver SUV for 18F-FDG PET. J Nucl Med 2013;54:677-82. [Crossref] [PubMed]
47. Keramida G, Potts J, Bush J, Verma S, Dizdarevic S, Peters AM. Accumulation of (18)F-FDG in the liver in hepatic steatosis. AJR Am J Roentgenol 2014;203:643-8. [Crossref] [PubMed]
48. Keramida G, Potts J, Bush J, Dizdarevic S, Peters AM. Hepatic steatosis is associated with increased hepatic FDG uptake. Eur J Radiol 2014;83:751-5. [Crossref] [PubMed]
49. Calistri L, Rastrelli V, Nardi C, Maraghelli D, Vidali S, Pietragalla M, Colagrande S. Imaging of the chemotherapy-induced hepatic damage: Yellow liver, blue liver, and pseudocirrhosis. World J Gastroenterol 2021;27:7866-93. [Crossref] [PubMed]
50. Meunier L, Larrey D. Chemotherapy-associated steatohepatitis. Ann Hepatol 2020;19:597-601. [Crossref] [PubMed]
51. Hu J, Cheng R, Quan M, Peng Y, Yang Z, Zhang Q, Ji F, Chen Y, Li B, Wen N. Hypermetabolic pulmonary lesions detection and diagnosis based on PET/CT imaging and deep learning models. Eur J Nucl Med Mol Imaging 2025;52:3792-806. [Crossref] [PubMed]