Liver SUV versus stage in Hodgkin's lymphoma: the total amount of uptake may play a role in the inverse relationship.

Abstract

Dear Sir, We read with great interest the paper by Chiaravalloti et al. [1], who investigated factors affecting intrapatient variability of F-FDG uptake in the liver and in the mediastinum of patients with Hodgkin’s lymphoma (HL) treated with doxorubicin (Adriamycin), bleomycin, vinblastine and dacarbazine (ABVD) chemotherapy. The authors convincingly showed that liver SUV variability should be carefully considered in assessing response to therapy on interim PET. Chiavaralloti et al. also found that, at staging, a higher stage is related to a lower liver SUVmax and SUVmean. To explain this result, the authors suggested that the liver F-FDGmetabolismmight be changed in patients with HL and a detailed liver function assessment should be performed soon after diagnosis of the disease (and also during treatment). We would like to focus on this finding and we suggest that, besides possible metabolic changes in the liver in patients with HL at staging, the total amount of F-FDG uptake in malignant tissues of these patients may also play a role in its origin. Indeed, it is known that if tissues show intense tracer uptake, the amount of tracer available to another tissue will be reduced [2]. Therefore, we suggest that the greater the total amount of F-FDG uptake in patients with HL at staging the lower the liver SUV. We have recently published a formula to estimate the percentage of injected F-FDG molecules (P) that are irreversibly trapped in an F-FDG-positive tissue during a PET examination: P ≅ 0.17*exp(−λtacq)*TLG/W, where λ (min) is the F physical decay constant, tacq (min) is the injection–acquisition time delay which is in the range 55 – 110 min, TLG (g) is total lesion glycolysis (i.e. the product of metabolic tumour volume and associated SUVmean) and W is the patient’s weight (kg) [3, 4]. This formula has been used to assess physiological uptake that may occur in current F-FDG PET imaging, such as in brown fat, myocardium and bowel, but it is also relevant to assess the total amount of uptake in malignant lesions over the whole body that usually trap F-FDG in an irreversible manner [5]. Examples of different amounts of pathological uptake are shown in Fig. 1 in two patients with Ann Arbor stage II (bulky) HL (Fig. 1a) and stage IV HL (Fig. 1b). PET imaging and data processing have been described in detail previously [4]. Briefly, intense F-FDG-positive pathological tissues were automatically outlined in 3-D on an Advantage 4.6 workstation (GE Medical Systems) using a thresholding based segmentation method (SUV threshold set to 4 g.mL, i.e. greater than the liver SUVmax). To avoid segmentation errors, preliminary manual outlining had first to be performed so that some F-FDG-positive tissues such as the heart (Fig. 1a), kidneys and bladder (Fig. 1b), located close to the relevant pathological tissues, could be discarded. In patients with stage II and IV HL, P and liver SUVmean (obtained from a 50-mL spherical volume in the right liver) were estimated to be 2.2 – 16.1 % and 2.1 – 1.5 g.mL, respectively. The main comparisons of the study by Chiaravalloti et al. showed an increase in liver SUV (both maximal and mean) between PET2 and PET0 (PET/CT scan performed after two ABVD cycles and at staging, respectively) and between PET2 and PET6 (PET/CT scan performed after six ABVD cycles). The former may be consistent with the role of the total amount of uptake at staging (PET0: reduced liver SUV), which would A reply to these comments can be found at DOI 10.1007/s00259-0142926-0