Purpose
This prospective single-institution study examines the impact of positron emission-tomography (PET) using 2-[18F] fluoro-2-deoxyglucose and CT scan radiation treatment planning (TP) on target volume definition in lymphoma.
Methods and Materials
118 patients underwent PET/CT TP during 6/2007-5/2009. Gross tumor volume (GTV) was contoured on CT-only and PET/CT studies by radiation oncology (RO) and nuclear medicine (NM) for 95 patients with positive PET scans. Treatment plans and dose-volume histograms were generated for CT-only and PET/CT for 95 evaluable sites. Paired t-test statistics and Pearson correlation coefficients were used for analysis.
Results
70 (74%) patients had non-Hodgkin lymphoma, 10 (11%) had Hodgkin lymphoma, 12 (10%) plasma-cell neoplasm, and 3 (3%) other hematologic malignancies. Forty-three (45%) presented with relapsed/refractory disease. Forty-five (47%) received no prior chemotherapy. The addition of PET increased GTV as defined by RO in 38 patients (median, 27%; range, 5-70%) and decreased GTV in 41 (median, 39.5%; range, 5-80%). The addition of PET increased GTV as defined by NM in 27 patients (median, 26.5%; range, 5-95%) and decreased GTV in 52 (median, 70%; range, 5-99%). Intra-observer correlation between CT-GTV and PET-GTV was higher for RO than for NM (0.94, P < .01 vs. 0.89, P < .01). Based on Bland-Altman plots, the PET-GTVs defined by RO were larger than NM. On evaluating clinical TPs, only four (4%) patients had inadequate target coverage (D95<95%) of the PET-GTV defined by NM.
Conclusions
Significant differences between RO and NM volumes were identified when PET was co-registered to CT for radiation planning. Despite this, the PET-GTV defined by RO and NM received acceptable prescription dose in nearly all patients. However, given the potential for a marginal miss, consultation with an experienced PET reader is highly encouraged when delineating PET/CT volumes, particularly for questionable lesions and to assure complete and accurate target volume coverage.