Patients selected for PSMA radioligand therapy with two different PET tracers showed one clear difference in their baseline scans: total tumor volume, or TTV, was higher in the Ga-68 PSMA-11 group than in the F-18 DCFPyL group. The study found no significant difference in SUVmax or SUVmean, the two other quantitative PET measures used in the comparison. The TTV difference reached statistical significance, with p=0.038. The comparison involved 140 patients, including 51 scanned with Ga-68 PSMA-11 and 89 scanned with F-18 DCFPyL.
The question was whether the choice of tracer was linked to treatment response or overall survival after PSMA radioligand therapy, and whether numerical features from the scans helped explain any difference. Researchers also analyzed separately 80 patients from an earlier study who were excluded from the primary analysis. The main comparison therefore rested on the 140-patient cohort, rather than on a single combined sample.
The groups did not start from identical records
The two groups differed in another potentially important way. Other life-prolonging therapies were reported more often among patients in the Ga-68 group than among those in the F-18 group, 68.4% versus 44.3%, with p=0.033. Prior treatment history was unavailable for 41 patients, or 29.3% of the cohort. Those differences and missing records make a simple reading of tracer-related comparisons harder, because the treatment background was not fully alike.
How the scans were measured
To calculate the PET measures, the analysis used thresholds to mark lesions. The initial lesion threshold was at least 3.0 SUV. If needed, it could be lowered in steps of 0.5 SUV, but not below 1.0 SUV. For liver metastases, the threshold was set at 1.5 times the mean SUV of normal liver. A nuclear medicine physician reviewed the segmentation and corrected it manually when necessary, while blinded to the tracer type and clinical outcomes.
This meant that the volume measure combined a numerical rule with expert review. The approach was designed to keep the reviewer from knowing which tracer had been used or what happened to the patient, but the resulting measurements still depended on the selected thresholds and any manual corrections. That detail matters when comparing TTV between groups, because a volume estimate is tied to how lesions are outlined.
How response and survival were counted
Treatment response was assessed through PSA. The main response endpoint, called PSA50, required a decline of 50% from baseline PSA. Researchers used the maximum PSA decline recorded during radioligand therapy or within 12 weeks after treatment finished. Overall survival was measured from the first radioligand-therapy administration until death, with follow-up ending at the last recorded assessment for patients who had not died.
A comparison, not a verdict on tracer choice
For PSA50 response, the researchers used logistic regression, a statistical method for comparing the odds of reaching a defined response. For overall survival, they used Cox proportional-hazards regression, which compares the timing of deaths during follow-up. TTV was log-transformed for the analysis, and SUVmean was retained because it was correlated with SUVmax. The reported Spearman correlation was 0.723, with p<0.001.
The result is best read as an association within the patients included in the comparison, not as evidence that one tracer caused a different tumor burden or a different treatment outcome. The reported baseline signal was narrower: TTV was higher in the Ga-68 PSMA-11 group, while SUVmax and SUVmean did not differ significantly. That leaves the study identifying a measurement difference, rather than establishing that either tracer is better.
Paper data and sources
Original title: Comparative analysis of treatment response to PSMA radioligand therapy using Ga-68 PSMA-11 and F-18 DCFPyL PET imaging
Authors: Hong C, Moradi Tuchayi A, Abtahi M et al.
Journal/Repository: Frontiers in oncology
Status: Peer-reviewed
First online: 2026-10-19
DOI: Not available
Original paper · Full text