Limited diagnostic utility of bone marrow sampling in <scp>PET</scp> ‐defined limited‐stage follicular lymphoma

Bone marrow (BM) involvement is present in up to 40%–55% of patients with follicular lymphoma (FL) and has negative prognostic implications.1 Historically, BM aspiration and biopsy (A&B) were routinely performed to complete staging in newly diagnosed FL,2 although accumulating data suggest that routine BM A&B at diagnosis or follow-up rarely influences therapeutic decisions and may be reasonably omitted for many patients with advanced-stage disease.3-5 Even so, international guidelines continue to endorse BM sampling together with 18F-fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT) imaging to confirm cases of suspected limited-stage disease,3, 4, 6 especially in cases where definitive treatment with radiation therapy (RT) is planned.7 However, the diagnostic utility of routine BM A&B in patients with PET-defined limited-stage FL remains uncertain. This retrospective population-based study included consecutive patients diagnosed with PET-defined stage I or II FL grade 1–3A in Alberta, Canada, between 2017 and 2023 who underwent BM sampling at diagnosis. The objective of this study was to assess the diagnostic yield of BM A&B in this setting. Individuals with PET-defined advanced-stage FL, grade 3B FL or histological transformation were excluded. PET scans were interpreted by experienced nuclear medicine radiologists as part of routine clinical practice. Patients were required to fast for 6 h before injection of 18F-FDG tracer (median dose 503 MBq, range 333–558) and subsequently underwent PET imaging from the skull base to mid-thigh with low-dose CT for localization and attenuation correction. Of note, due to the relatively limited access to PET scans in Alberta during the study period, a staging PET was often omitted when it was unlikely to alter clinical management, for example, in patients proceeding with active surveillance or chemoimmunotherapy based on staging CT findings alone. According to provincial practice guidelines, unilateral iliac crest BM A&B was recommended to confirm limited-stage disease or to investigate unexplained cytopenias. However, the decision to perform BM sampling was ultimately at the discretion of the treating physician and was often reserved for patients with PET-defined limited-stage disease prior to referral for RT. BM specimens were reviewed by experienced haematopathologists in routine practice using standard immunohistochemistry and flow cytometry panels. Among 1209 patients diagnosed with FL during the study period (Figure S1), 295 (24%) were classified as having radiographically defined limited-stage disease based on PET (n = 114) or CT (n = 181). BM A&B was performed at diagnosis in 84 (28%) of these 295 cases, including 60 (53%) PET-staged and 24 (13%) CT-staged patients. Unexpectedly, a staging BM A&B was performed in only 50 (44%) of 113 patients who received RT, largely reflecting variability in practice between physicians. The study cohort included 60 patients with PET-defined limited-stage FL who underwent BM aspiration (n = 59) and/or biopsy (n = 60). Median age was 65 years (range 31–82). Lugano stage by PET was I in 31 (52%), IE in 5 (8%) and II in 24 (40%) patients. Other baseline characteristics are described in Table 1. The median core biopsy length was 1.65 cm (range 0.25–3.1 cm), and 19 (32%) were ≥2 cm in length. Initial management for FL included ≥24 Gy of RT (n = 44), active surveillance (n = 11), chemoimmunotherapy (n = 4) or rituximab monotherapy (n = 1). As displayed in Figure 1, the staging BM A&B was negative for FL in 56 (93%) cases, including 3 patients in whom flow cytometry of the BM aspirate detected a minute CD10-negative clonal B-cell population unrelated to FL (e.g. marrow equivalent of monoclonal B-cell lymphocytosis). Only 4 (7%) patients had FL detected in the BM by flow cytometry (n = 1), biopsy (n = 1) or both (n = 2), despite having no PET findings of marrow involvement. Routine laboratory parameters (e.g. full blood count, lactate dehydrogenase, β2-microglobulin and serum protein electrophoresis) were unremarkable in these four patients, except a single patient with a monoclonal protein due to a concurrent plasma cell neoplasm. The number of patients required to undergo BM A&B to identify one case of marrow involvement with FL was 15. Notably, in no case did these findings influence clinical management, as two patients proceeded with RT despite minimal BM infiltration, while the remaining two patients were already deemed ineligible for RT due to a concurrent plasma cell neoplasm or disease extending beyond a single radiation field. Although provincial and international guidelines continue to recommend BM A&B for confirmation of limited-stage FL, the findings of this study demonstrate that many clinicians consider the procedure to have low diagnostic yield and frequently omit it in routine practice, including in more than half of patients treated with definitive RT. Additionally, among the 60 PET-defined limited-stage cases in this cohort, BM A&B revealed unsuspected FL in only 1 (7%) of 15 and did not alter clinical management in any case. These results challenge the necessity of routine BM sampling in this setting, particularly given its invasiveness, patient discomfort, resource requirements and limited clinical impact. Prior evidence supporting recommendations for staging BM A&B in limited-stage FL included a National LymphoCare Study which observed superior progression-free survival among patients with stage I FL who underwent BM sampling compared to those who did not, implying that comprehensive staging may help exclude occult systemic disease.8 However, that analysis was confounded by heterogeneous treatment approaches and the frequent use of CT rather than PET for staging. More recent data suggest that PET has only modest sensitivity for detecting BM involvement in FL.9 Nevertheless, the relevance of these findings to the minority of patients with presumed limited-stage disease remains uncertain, as this subgroup likely has a substantially lower pretest probability of marrow infiltration than patients presenting with more advanced disease. Thus, the clinical value of PET in this context may stem less from its ability to directly detect marrow involvement and more from the fact that patients with localized nodal disease by PET have a low overall frequency of BM involvement, potentially reducing the incremental yield and necessity of routine BM sampling in this population. Few studies, aside from the present analysis, have specifically examined the added value of BM A&B in patients with PET-defined limited-stage FL. Consistent with our findings, prior series have reported that BM A&B identified unsuspected marrow involvement in only 6 (6%) of 100, 0 (0%) of 102 and 0 (0%) of 27 patients with PET-defined limited-stage FL.10-12 Moreover, an international study of 512 patients with PET-defined limited-stage FL treated with RT found no difference in outcomes between those staged with versus without BM A&B, suggesting that occult marrow involvement in this population is rare.7 Similarly, in a cohort of 311 patients with marginal zone lymphoma staged with both PET and BM sampling, marrow involvement was not associated with progression or survival outcomes, and the authors concluded that staging BM A&B may be unnecessary in routine practice.13 Potential limitations of this study include its relatively modest sample size despite a large number of patients screened for eligibility, variable adherence to staging guidelines, possible interobserver variability in PET interpretation and the limited proportion of BM biopsy cores exceeding the recommended 2 cm length.14 Although we cannot exclude the possibility that interpreting radiologists were not blinded to BM A&B results, this is unlikely to have meaningfully influenced the findings given that PET imaging was performed before BM sampling in most cases. Additionally, the proportion of patients who underwent staging BM A&B and the median biopsy core lengths observed in this study are consistent with those reported in routine clinical practice in other series.8, 15 Further research is warranted to validate these findings in larger cohorts and to determine whether more sensitive molecular techniques could offer greater accuracy in the staging of FL in the future. In conclusion, routine BM A&B rarely detects unexpected marrow involvement and is unlikely to affect the clinical management of patients with PET-defined limited-stage FL. These findings argue against the routine use of BM sampling in this population and may facilitate informed discussions between physicians and patients regarding the value of this procedure in contemporary staging practice. MS performed the research, analysed data and wrote the manuscript. CO, CS, JQC, AB, SP, MS and LS revised the manuscript. RP conceived the study, analysed data and wrote the manuscript. This study did not receive funding from any source. The authors have no relevant conflicts of interest to disclose for this study. This study was approved by the Health Research Ethics Board of Alberta. Informed consent was not required for this retrospective study as it posed minimal risk to participants. Data may be made available by the authors upon reasonable request. Figure S1. Staging of patients with newly diagnosed follicular lymphoma (FL) during the study period. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.