Interim PET Scan Does Not Predict Outcome in Primary Mediastinal B Cell Lymphoma Treated with R DA EPOCH

Primary mediastinal B cell lymphoma (PMBL) is an aggressive subtype of diffuse large B-cell lymphoma derived from thymic B cells. As data presented by Dunleavy et al has shown remarkable overall survival with rituximab and dose-adjusted EPOCH (R DA EPOCH) (1), we adopted this strategy as standard of care for fit patients with PMBL in our centres from 2010. Although interim PET after 2 cycles (PET2) is of demonstrable predictive value in activated and germinal centre B-cell DLBL (ABC/GCB DLBL) and has become routine practice in many centres, its role in monitoring PMBL response to treatment is unclear. We therefore decided to evaluate the utility of PET2 in prognosticating PMBL outcome.

A multicenter retrospective analysis was performed on PMBL patients treated with R DA EPOCH (6 to 8 cycles) and evaluated with PET at baseline (PET0), after two chemotherapy cycles (PET2), and post treatment. Radiotherapy (RT) treatment was decided on a case by case basis. A cohort of ABC/GCB DLBL patients treated with R-CHOP and evaluated with PET2 served as an internal control. Histological diagnosis with reviewed by a haematopathologist and PET scans were assessed by a nuclear medicine physician specialising in lymphoma imaging. Fisher's exact test was used to compare categorical variables, and positive predictive value of PET2 for relapse was calculated.

Ten cases of PMBL were identified. Median age was 35 years (17 – 49) and 62% were female. Three (30%) patients had stage 1 disease, 4 (40%) were stage 2, and 3 (30%) were stage 4. Eight (80%) patients had PET0. Two patients had no PET0 due to clinical urgency of presentation. PET2 was negative in only 1 (10%) patient (Deauville 2) and positive in 9 (90%) patients (Deauville 4). SUVmax reduction PET0 to PET2 for the single PET2 negative case was 24.7 to 2.7. Mean SUVmax reduction for the other 7 paired PET0 to PET2 cases was 17.6 to 6.0. At the time of analysis, all 10 patients are alive. Seven patients have completed treatment with median follow up was 16 months. All 7 patients (100%) achieved an end of treatment PET negative remission (Deauville 1 or 2) with 3 (43%) patients receiving RT. Of the 9 PET2 positive patients, 1 (11%) relapsed at 17 months post PET2 but remains in remission following salvage treatment and autograft.

The ABC/GCB DLBL internal control group consisted of 36 patients treated uniformly with R-CHOP. In contrast to PMBL, PET2 was positive in only 10 (28%) cases and negative in 26 (82%) cases. PET2 positivity was significantly more frequent in the PMBL cohort (p = 0.007). The positive predictive value for future relapse of PET2 in PMBL patients who have completed treatment was 16%.

Excellent results were achieved with R DA EPOCH, in line with previous data(1). Similar to prior studies, a high rate of PET2 negativity was achieved in ABC/GCB DLBL treated with R-CHOP (82%). However, only a minority of PMBL patients treated with R DA EPOCH were PET2 negative (10%), yet all patients who completed therapy were PET negative at the end of treatment. PET2 had a very low positive predictive value in identifying patients at risk of relapse (16%).

No relevant conflicts of interest to declare.