Tumoral Maximum Standardized Uptake Value > 10 Measured on 18 f-FDG-PET: A New Valid Marker to Discriminate Richter Syndrome

Background: One of the main complications in CLL is Richter syndrome (RS). RS derives from the rare transformation of chronic lymphocytic leukemia (CLL) into an aggressive lymphoma, most commonly of the diffuse large B-cell lymphoma (DLBCL) type. RS occurs in 2.2% to 8% of patients with CLL and the prognostic is poor with a median survival from 5 to 8 months. Detection of RS by imaging has resulted in conflicting and non-significant results. Aim: The objective of this study was to validate recent findings correlating FDG/PET imaging and histological features in CLL and to demonstrate that a tumoral maximum standardized uptake value > 10 measured on 18F-FDG-PET is a new valid marker to discriminate RS. Results: From June 2006 through December 2012, 240 patients from the Division of hematology of Centre Hospitalier Lyon Sud and Créteil and from the Mayo Clinic Rochester have been analyzed with a mean age of 62 years (21-91). Clinical, histological (confirm by biopsy) and biological parameters have been identified with 10% of the patients as having RS, 34% stable CLL disease; 42% of rapid CLL progression (histological features of progression defined as increased large cell number, large confluent proliferation centers or high proliferation rate assessed by Ki-67 but not meeting criteria for diffuse large B-cell lymphoma/RS) and 14% with others diseases (e.g. infection and or cancers). For patients with stable CLL disease, the median tumoral SUV max was 2 (range: 0-2.4). Among patients with a rapid progression of CLL, 90% had a tumoral SUV max <10 but greater than the liver median SUV max (4.5 range: 1-11). In contrast, 90% of patients with RS had a median tumoral SUV max >10 (12.9; range: 5-27). A statistically significant difference between SUV max of CLL patients with stable disease and RS was observed (2.2 vs. 12.9; p< 0.0001) and similarly for SUV max of CLL patients with rapid disease progression and RS (4.5 vs 12.9; p< 0.0001). Regardless of the RS prevalence (2.2% to 8%), statistical tests identified a threshold of tumor SUV max > 10 as the more discriminating cut off. Using this threshold, the sensitivity and specificity of PET to identify RS in our cohort are 91% and 95% respectively. Assuming an RS prevalence of 2.2%, positive predictive value (PPV) and negative predictive value (NPV) using the >10 threshold were 28.7% and 99.8%; for an 8% prevalence of RS, the PPV and NPV are 60.8 and 99.2% respectively.The proportion of correctly classified patients with RS is more accurate using a threshold of tumoral SUV max > 10 than 5 (2.2% RS prevalence: 94.8% versus 71.8%; 8% RS prevalence: 94.6% versus 73.5%). Finally analysis of the area under the curve (AUC) reveals a value of 0.95 (95% confidence interval: 0.89- 0.99). Recently, Falchi et al, reported on 332 patients with CLL classified as 95 RS, 117 rapid progression and 120 stable disease and demonstrated a strong correlation between histological features and PET imaging. A SUV max ≥10 strongly correlated with a shorter overall survival. Similarly, in our study we have shown excellent sensitivity and negative predictive value estimated at 100%, revealing the ability of 18F-FDG-PET-CT to rule out the diagnosis of RS if the tumor SUV max is less than 10. Conclusion: 18F-FDG-PET-CT is a valuable tool in the diagnostic evaluation of RS for patients with CLL. It is not only useful to identify RS syndrome and guide the site of biopsy but also to identify CLL patients who will experience more rapid disease progression. An SUV max > 10 is the optimal threshold to distinguish RS in CLL. Disclosures No relevant conflicts of interest to declare.