Definition Of Diffuse Micronodular Spleen Infestation (Dmsi) In Fdg-Pet Of Pediatric Patients With Hodgkin'S Lymphoma (Hl)

21 Aim: The spleen is an important, but often underestimated part of the immune system. Involvement by HL may indicate further chemotherapy cycles and irradiation of this organ, which may impair its proper functioning. Therefore, specific staging of the spleen is essential for risk-adapted treatment. Among the available imaging techniques FDG-PET detects focal HL manifestations with high sensitivity and indicates response to treatment at an early stage. However, DMSI is difficult to determine and may appear as an increased uptake of the entire organ. In this context some publications suggest DMSI if the uptake of the spleen exceeds the liver uptake. This thesis was examined based on a large cohort of pediatric HL patients. Methods: 490 patients (196 with and 294 without spleen infestations) of the EuroNet-PHL-C1 trial who received central reference review of their imaging were considered. In the staging FDG-PET the SUVmean quotient spleen [7.5 ml VOI] / liver [30 ml VOI] (SQSL) was calculated by avoiding inclusion of focally uptaking splenic lesions (for all of the 490 patients). In case of spleen involvement, the respective pattern (focal / multifocal or diffus-micronodular) was recorded from morphological imaging. SQSLs were grouped (no involvement - focal / multifocal involvement - diffus-micronodular involvement) and plotted. Results: 183 patients had either focal or multifocal involvement, 13 had a DMSI. The SQSL at staging was on average 1.02 (+/- 0.19) in patients without infestations, 1.52 (+/- 0.72) in patients with focal/multifocal involvement and 1.76 (+/- 0.73) in case of DMSI. SQSL values for patients with DMSI overlapped SQSL values of patients without spleen involvement in the range of 1.06 to 1.58. Thus, deduction of DMSI is only reasonable if SQSL exceeds 1.58 (specificity 99.6%; 95%-CI 0.98 - 1.0). Using this cut-off 6 / 13 patients with DMSI were diagnosed correctly. Applying a SQSL cut-off at 1.0 would have led to a false positive diagnosis in 148 of 294 patients without spleen infestation. Conclusion: In about half of the patients without spleen involvement the spleen uptake exceeded that of the liver. Thus, the criterion (SQSL>/= 1.0 => involvement) is not feasible to diagnose DMSI. To avoid unnecessary upstaging, DMSI should only be diagnosed in a high specific manner which may be defined by an SQML > 1.58.