Diagnostic Utility of the 18f-FDG PET/CT Scan in Solid Organ Transplant Patients with EBV Reactivation at Risk for Post-Transplant Lymphoproliferative Disorder

Background: Post-transplant lymphoproliferative disorder (PTLD) is a major complication following solid organ transplantation (SOT). Epstein-Barr virus (EBV) infection is a significant risk factor for PTLD. Unfortunately, PTLD spans a heterogenous spectrum of subtypes often with nonspecific presentations. ¹⁸FDG-positron emission tomography/computed tomography (PET/CT) is a powerful diagnostic modality because of its sensitivity for lymphomatous infiltration. PET/CT scans are often performed in the setting of EBV reactivation and/or clinical suspicion for PTLD, but SOT patients (pts) are also at risk for other post-transplant complications. Here, the goal was to define the utility of PET/CT to diagnose PTLD in EBV+ patients following SOT.

Methods: Pts with history of EBV reactivation after SOT with both a PET/CT (indication for work-up or staging of PTLD) and biopsy performed were included in this single center study. Pts with PET/CT done for unrelated reasons, after PTLD treatment, or after complete resection of disease were excluded. Pt characteristics at time of SOT and PET/CT were abstracted. PET/CTs were interpreted by 2 radiologists blinded to clinical data assessing for PTLD, FDG-uptake and volumetric measures of the dominant lesion, and locations and numbers of lesions. Characteristics were associated with likelihood of biopsy-proven PTLD by binary logistic regression or Fischer exact test. Differences in continuous variables were assessed with Wilcoxon rank sum test.

Results: 118 pts were included; 82 (70%) were >18 yrs at SOT, median time from SOT to PET/CT was 6 yrs (range 0-31), and most common SOT organs were kidney (n=35, 30%), heart (29, 25%), and liver (27, 23%). 69 (59%) pts were EBV seropositive before SOT, and 74 (63%) and 57 (48%) patients had EBV viremia and elevated LDH at PET/CT, respectively. Median SUVmax of the dominant lesion in 100 pts was 9.85 (range 0-66.1) and was not measured for 18 with diffuse GI luminal uptake; 14 pts had no abnormal FDG uptake. FDG-avid lymphadenopathy most commonly was on both sides of the diaphragm (39, 33%) and extra-nodal disease most commonly involved multiple organs (40, 34%). PET/CTs were assessed as positive for PTLD in 67 (57%) cases and required adjudication by a third reviewer in 17 (14%) cases due to discordance. PTLD was diagnosed in 66 (56%) pts; subtypes were monomorphic B-cell in 28 (42%), polymorphic B-cell in 21 (32%), Hodgkin lymphoma in 9 (14%), monomorphic T-cell in 4 (6%), and other in 4 (6%). PTLD was EBV+ in 36 (55%) cases. Non-PTLD findings in the remaining 52 (44%) pts were reactive changes/follicular hyperplasia in 16 (14%), benign in 18 (15%), other malignancy in 12 (10%), or infection in 6 (5%).

The sensitivity and specificity of blinded radiologist review for PTLD were 76% (95% CI 65-86) and 67% (95% CI 55-80), respectively. By univariable analysis, age>18 yrs at SOT (OR 0.37 p=0.02), recipient EBV seropositivity (OR 0.15 p<0.001), EBV viremia (OR 4.08 p=0.001), and SUVmax > median (OR 2.94 p=0.01) were associated with likelihood of PTLD, however by multivariable analysis only EBV seropositivity (p=0.049) and viremia (p=0.027) were statistically significant. Metabolic tumor volume and total lesion glycolysis of the dominant lesion were not associated with PTLD likelihood. FDG-avid lymphadenopathy was more likely to be associated with PTLD if located on both sides of the diaphragm or subdiaphragmatic (p=0.010). SUVmax was not different between PTLD subtypes (p=0.44) but was different when comparing PTLD (median 13.0) to non-PTLD findings (p=0.007), in particular reactive changes/hyperplasia (median 6.1) or benign findings (median 6.5).

Conclusion: This study assessed the diagnostic utility of PET/CT in pts with EBV reactivation and clinical suspicion of PTLD. Due to the heterogeneity of presentation and risk of other etiologies, pathologic confirmation of PTLD is still recommended even in SOT pts with abnormal PET/CT. Blinded radiologist PET/CT review was moderately sensitive but less specific for PTLD, potentially because SOT pts are also at risk for non-PTLD causes of FDG-avid lesions. The degree of metabolic activity of the dominant lesion was useful in discriminating PTLD from other etiologies, though EBV seropositivity and viremia were the most important characteristics. This data provides guidance for PTLD work-up and highlights the importance of incorporating multiple diagnostic modalities and biomarkers.