Impact of Routine Surveillance Imaging on the Outcome of Patients with Relapsed Hodgkin Lymphoma.

To determine the role of asymptomatic post-remission surveillance imaging in the diagnosis of first relapse and its impact on outcomes in patients with Hodgkin lymphoma (HL).

The impact of surveillance imaging on diagnosis of first relapse and outcomes was determined by analysis of a defined patient population with relapsed HL who underwent high dose therapy followed by autologous stem cell rescue (HDT/ASCR). We retrospectively identified 114 patients from the HDT/ASCR database with biopsy-confirmed, relapsed HL; 94 patients had adequate data and follow-up for inclusion in the analysis. Details of surveillance imaging were obtained including: frequency; type; indication; and results of post-remission imaging. The indication for imaging was classified as asymptomatic surveillance (AS) or clinically indicated (CI; i.e. to investigate symptoms or physical examination findings). We have previously reported a validated prognostic model for relapsed HL that found time from initial therapy, presence of B symptoms at relapse and extra-nodal disease at relapse to be adverse factors (Moskowitz et al. Blood 97:616). We determined the prognostic risk group (PRG) (low [L] 0-1 factor, intermediate [I] 2 factors, high risk [H] 3 factors) for all patients. Overall and failure-free survivals were determined using the methods of Kaplan and Meier.

Patient characteristics included: median age 32 years; PRG L: 65%, I: 31%, H: 4%; AS 36 (38%), CI 58 (62%). The median follow-up for surviving patients was 7.4 years. The PRG (L, I/H) correlated to outcome, validating its applicability in this patient cohort: FFS at 5 years was 64.8% and 49.4% respectively, p=0.045. PRGs were evenly distributed between the AS and CI groups: L: 64% v 66%; I/H: 36% v 37% p=0.48. The FFS at 5 years for patients in the AS and CI groups was 58.4% and 59.3% respectively, p=0.9; similar there was no difference in 5 year OS, AS 62.4% and CI 73.3% p=0.6. Within a given risk group (L or I/H) patients in the AS group did not have a superior outcome compared to the CI group.

AS does not identify a group of patients with first relapse of HL with a more favorable risk profile according the MSKCC prognostic model. While nearly 40% of patients were identified with relapse as a consequence of AS imaging, this did not identify a group of patients with relapsed HL with improved outcomes compared to patients who had CI imaging. If these results are confirmed in a prospective study, AS may be safely eliminated in HL after remission to reduce cost and long-term risk of the diagnostic imaging.

No relevant conflicts of interest to declare.