SUVR (SUVmax Lymphoma/SUVmax Liver) Vs Deauville Score for Predicting Relapse in Hodgkin Lymphoma

Background

The Deauville score (DS) assessed by fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) is the current standard method for evaluating treatment response in patients with Hodgkin lymphoma (HL). However, DS is a qualitative and somewhat subjective evaluation. Recent evidence supports the use of the standardized uptake value ratio (SUVR) as a more quantitative and objective assessment of treatment response.

Objective

To compare the diagnostic performance of DS and SUVR at the EOT FDG PET/CT in predicting PFS in patients with HL.

Methods

We included patients with classical HL treated with curative intent between 2000 and 2019 who had EOT PET/CT scans. Each EOT scan was re-scored according to DS and SUVR. SUVR was defined as lesional SUVmax divided by liver SUVmax. Receiver operator curve (ROC) analysis determined the optimal SUVR cut-off value using Youden's index. We computed sensitivity, specificity, positive, and negative predictive values (PPV & NPV) for DS (positive ≥ 4) and SUVR (positive ≥ 1.13). Kaplan-Meier curves and Cox-regression analysis evaluated the PFS predictive ability of the two response-assessment modalities. Informed consent was obtained from all patients.

Results

157 patients had available data at the EOT timepoint. Median age was 31, and 30% had limited favorable disease at diagnosis, defined as Stage I-IIA, non-bulky. Most patients received frontline ABVD chemotherapy (4-6 cycles) and 16 patients (10%) received radiation therapy. There were 35 PFS events, including 5 deaths. The optimal SUVR cut-off at EOT was 1.13. Median PFS among patients with positive DS or SUVR was 8.1 and 8.2 months, respectively. Diagnostic parameters for DS and SUVR were similar (PPV: 77% and 78%, respectively). Both DS and SUVR predicted PFS (HR 2.46 [95% CI 0.29-20.66] and HR 9.06 [95% CI 1.08-75.87], respectively).

Conclusion

Both SUVR with a positivity threshold of 1.13 and DS are predictive of PFS at EOT and have similar diagnostic parameters in patients with HL. However, SUVR offers a more objective assessment of treatment response.

No relevant conflicts of interest to declare.