Bone marrow biopsy does not add value to PET/CT staging of DLBCL
The use of PET/CT imaging accurately assessed bone marrow involvement in most patients with diffuse large B-cell lymphoma, according to retrospective study results.
Thus, the addition of bone marrow biopsy to PET/CT staging does not add relevant diagnostic or prognostic value for most patients, according to the researchers.
Because bone marrow involvement occurs in up to 25% of newly diagnosed patients with diffuse large B-cell lymphoma (DLBCL) and is associated with poor outcomes, random iliac crest bone marrow biopsy has been a standard component of routine staging workup.
“However, it is well recognized that bone marrow biopsy has several limitations, including the possibility of missing a patchy pattern of marrow involvement … [and] risks include pain, anxiety, infection and bleeding,” Diego Villa, MD, FRCPC, associate professor of medicine at University of British Columbia and postgraduate site director at BC Cancer Agency, and colleagues wrote. “Thus, readdressing the value of this traditional staging investigation is relevant with the advent of advanced imaging technologies, including PET/CT.”
Villa and colleagues retrospectively identified 530 patients (median age, 65 years; 56% male) from Canada and Denmark with newly diagnosed DLBCL, who underwent PET/CT staging and a bone marrow biopsy. They used original written PET/CT and pathology reports to determine Ann Arbor staging and outcomes, with and without the contribution of bone marrow biopsy.
Twenty-eight percent of patients (n = 146) had focal bone marrow lesions on PET/CT and 16% (n = 87) had a positive bone marrow biopsy.
Of the 146 patients with a positive PET/CT, 36% (n = 52) had a positive biopsy (DLBCL, n = 39; indolent non-Hodgkin lymphoma [iNHL], n =13). Nine percent (n = 35 of 384) of patients with negative PET/CT results had a positive biopsy (DLBCL, n = 12; iNHL, n = 23).
Bone marrow biopsy results upstaged 12 of 209 patients (6%) with stage I or stage II disease to stage IV; however, only three of these patients had DLBCL. Sixteen percent of patients with stage III disease (n = 14 of 92) were upstaged to stage IV (DLBCL, n = 5; iNHL, n = 9).
PET/CT identified bone marrow involvement with a sensitivity of 60% (95% CI, 49-70) and a specificity of 79% (95% CI, 75-83). The positive predictive value was 36% (95% CI, 28-44) and the negative predictive value was 91% (95% CI, 88-94).
Median follow-up was 2 years (range, 3 months-6.5 years).
Among patients with stage IV disease, patients with a negative vs. positive PET/CT for bone marrow involvement demonstrated a higher rate of 2-year OS (79% vs. 63%; P = .001) and PFS
(74% vs. 53%; P < .001).
Patients with negative PET/CT and bone marrow biopsy for bone marrow involvement had higher rates of 2-year OS (82% vs. 65%) and PFS (75% vs. 57%; P = .003 for both).
Further, patients with concordant bone marrow involvement (n = 51) experienced worse OS and PFS outcomes than those with discordant involvement (n = 36) or no histological involvement (n = 443).
The researchers acknowledged the lack of immunohistochemistry and cytogenetics data, as well as the lack of central pathology review, as potential study limitations.
“It is difficult to identify a subgroup of patients with negative PET/CT in whom treatment and prognosis would have been altered by bone marrow biopsy, as even in those with stage I/II by PET/CT, the likelihood of upstaging with a meaningful change in treatment is small,” Villa and colleagues wrote. “Even though discordant iNHL is missed without routine bone marrow biopsy, the favorable outcome of patients with discordant iNHL supports that they are managed safely with standard DLBCL therapy … Our data suggest routine bone marrow biopsy is not a critical element of disease staging in all patients with newly diagnosed DLBCL undergoing upfront PET/CT staging.” – by Cameron Kelsall
Disclosure: The researchers report no relevant financial disclosures.