Abstract
Clinical trials are underway using FDG PET imaging as a response surrogate for risk-adapted treatments to achieve favorable long term outcome while reducing therapy-related toxicities in patients with HL. The IHP provided “guidelines” to standardize FDG PET-based response evaluation. Before implementation in clinical practice, further validation of these guidelines is necessary. Our objective was to validate IHP criteria for response evaluation after 2 cycles of therapy with prospectively collected data in CALGB 50203, a trial of AVG for the initial treatment of stages I and II non-bulky HL. IHP-based PET interpretation was also compared with CT-based lesion size changes.
Methods: Treatment consisted of doxorubicin 25 mg/m2, vinblastine 6 mg/m2 and gemcitabine 800mg/m2 every 2 weeks for 12 treatments (6 cycles). Responses were assessed according to the International Workshop criteria (JCO1999:17: 1244–53). FDG-PET imaging (PET/CT: 60 pts, PET alone: 15 pts) and a separate dedicated diagnostic CT scan of involved sites was performed at baseline and after two cycles of AVG (PET-2). No treatment change was allowed on the basis of the PET-2 results. Of 99 assessable patients, 75 had PET-alone (14) or PET/CT (61) after 2 cycles; median age:34 yrs (18–80), 32% were ≥45 yrs, stages: 78% IA-IIA, 22% IIB. The primary interpretation of PET-2 studies was based on IHP criteria (JCO2007;25:571). The % change in the sum of the products of the diameter (%SPD) of all measurable lesions were compared between baseline and at cycle-2 CT. PET-2 and cycle-2 CT data were correlated with progression free survival (PFS).
Results: Fifty-six patients (75%) achieved a CR/CRu, 21% a partial response (PR), 4% had stable disease (SD). After a median follow up of 2.1 yrs (1.2–3.4 years), 19 of 75 patients relapsed/progressed, with an estimated 2-year PFS of 0.87 (95% CI [0.74,0.94]Only 10 of 56 patients (18%) with CR/CRu were PET positive at PET-2 compared to 13/19 (68%) of those with SD or PR (p<0.0001). Twelve of 23 (52%) PET-2 positive patients relapsed compared to 7 of 52 (13.5%) who were PET-2 negative. The 2-yr probability of PFS was 0.87, [95% CI (0.74,0.94)] among PET-2 negative patients vs. 0.47 [95% CI (0.26,0.66)] in those who were PET-2 positive, p=0.0001. In an exclusive analysis of PET/CT scans, the 2-yr probability of PFS was 0.90, [95% CI (0.75,0.96)] among PET-2 negative patients vs. 0.35 [95% CI (0.13,0.59)] in those who were PET-2 positive, p<0.0001. The best PFS cut-point for %SPD change at cycle-2 was 70%. After cycle-2, PET-negative patients had a higher %SPD change compared to PET-positive patients (74.5% vs. 64.5%, p=0.003). Adjusted for baseline SPD, patients with <70% change were at 5.3 times higher risk of relapse. The 2-yr probability of PFS was 0.89, [95% CI (0.73,0.95)] among patients with > 70% change vs. 0.55 [95% CI (0.36,0.71)] in those with < 70% change, p=0.003.
Conclusion: IHP-based interpretation of FDG PET after 2 cycles of chemotherapy yields a high correlation with 2-year PFS, in particular for combined PET/CT, thus, it could be used as a response surrogate for risk-adapted treatments. The prediction of PFS using FDG-PET is superior to %SPD change after 2 cycles of therapy. Ongoing studies will prospectively define the role of interim FDG PET in tailoring treatment to optimize benefits and minimize risks.
Analysis of glucose metabolism by 18F-FDG-PET imaging and glucose transporter expression in a mouse model of intracerebral hemorrhage
