FDG and FCH PET/CT of Multiple Myeloma at Various Clinical Situations: Lesion Detection, Proposal for a Patient-Based "Summ" Score and Reproducibility of Scoring

Introduction: In a previous pilot study, PET/CT performed for suspected relapsing or progressive multiple myeloma (MM) revealed more lesions with FCH than with FDG. We then launched a prospective comparison of FDG and FCH PET/CT in MM at initial staging, restaging and during therapeutic follow-up. The IMPeTUs scoring system was proposed in 2016 by Nanni et al. to summarize the results of FDG PET/CT. However, the IMPeTUs scoring system is not designed to reflect by one single patient-based score the extent of the disease, as determined on radiotracer uptake. To compare the performance of those two PET radiotracers, we set a single patient-based score, derived from the IMPeTUs approach, and we applied it to the comparison of FDG and FCH PET/CT at various clinical situations of MM.

Aim: To compare (1) the reproducibility of reading and scoring of FDG and FCH PET/CT in initial staging, restaging and therapeutic follow-up of MM, (2) the correlation between the scores on FDG or on FCH PET/CT and serum concentration of M-protein, a quantitative criterion reflecting the secretory activity of MM, recommended at diagnosis and as a part of evaluation of therapeutic response.

Patients and methods: FDG and FCH PET/CT were performed in 105 patients (20 at initial staging, 37 at relapse and 48 during follow-up).The FDG and FCH PET/CTs were read in a random order with a wash out period by two experienced readers blinded to any patient's clinical data. They scored according to IMPeTUs scoring system, evaluating the number and intensity of foci of radiotracer uptake in 5 sites (spine, skull, extraspinal, visceral and nodal), the presence of diffuse axial and/or appendicular bone marrow uptake, of PET-positive lytic lesions, or of fractures. According to the number of visible foci in each site, the site-score was: 1: 0 focus, 2: 1-3 foci, 3: 4-10 foci or 4: >10 lesions. In case of diffuse bone marrow uptake by the axial or appendicular skeleton, the site-score 4 was attributed to the corresponding site, to integrate this pattern in one single score per site. A "SUMM" score, theoretically ranging from 5 to 20, was obtained by summing up the site-scores of the 5 sites. The kappa coefficient was calculated to evaluate the inter-reader agreement in determining the SUMM score.A consensus reading was finally performed, to solve discordances in scoring between the two readers. The paired Wilcoxon test was used to compare the consensual SUMM scores of PET/CT with FDG and FCH; the correlation coefficient between the consensual SUMM scores and M-protein level was calculated.

Results:Before final consensus, the inter-reader agreement in determining the SUMM score was better with FCH (Kappa=0.90; CI=0.85-0.95) than with FDG (Kappa=0.75; CI=0.64-0.86). The analysis was then continued on the data of consensus reading. FCH showed more foci in 28 patients (4 at initial staging, 13 at relapse, 11 during follow up) and FDG in 14 patients (3 at initial staging, 4 at relapse, 7 during follow up). The SUMM score was significantly greater with FCH than with FDG (p=0.0027).M-protein was present in the serum of 95/105 patients. The correlation with serum M-protein level was significantly >0 for both SUMM scores, but loose for FDG (r=0.3, p=0.0029) and stronger for FCH (r=0.7, p<0.001), the difference being significant (p=0.0001).Among 8 patients without M-protein detected in their serum, a negative PET/CT with FDG (SUMM FDG=5) but positive with FCH (SUMM FCH>5) was observed in 2 patients (1 referred for evaluation of relapsing MM and in one referred during therapeutic follow-up of MM).

Conclusion: A patient-based "SUMM" score is proposed to characterize the extension and the number of foci on PET/CT imaging of MM. Using this SUMM score, a better inter-reader reproducibility, lesion detectability and correlation with serum M-protein was observed for FCH than for FDG, confirming FCH as a potential tracer in MM deserving further research.