Comparing the clinical performance and cost efficacy of [Ga]Ga-PSMA-11 and [F]PSMA-1007 in the diagnosis of recurrent prostate cancer: a Markov chain decision analysis.

PURPOSE

Amongst others, [Ga]Ga-PSMA-11 and [F]PSMA-1007 are available for the detection of recurrent prostate cancer (rPC). There are currently limited data comparing the performance of these two radioligands with respect to clinical outcomes or their cost efficacy, which this study aims to address.

METHODS

Two hundred and forty-four patients undergoing PSMA PET/CT for rPC were retrospectively analysed for this study (one hundred and twenty two with each radiopharmaceutical) to generate rates of PET positivity, negativity and unclear findings. Patients underwent follow-up to determine the rate of additional examinations and to confirm PET findings. A Markov chain decision analysis was implemented to model clinical decision-making processes and to analyse clinical performance of the two tracers. We determine their clinical cost efficacies using cost data from several countries where both radiotracers are in routine use.

RESULTS

The PET positivity rate was non-significantly higher for [F]PSMA-1007 compared to [Ga]Ga-PSMA-11 (91.8% vs. 86.9%, p = 0.68), whereas the rate of uncertain findings was significantly greater (17.2% vs. 8.25%, p = 0.02). The probability of a true positive finding was higher for [Ga]Ga-PSMA-11 (0.90, 95% CI 0.70-0.98) vs. [F]PSMA-1007 (0.81, 95% CI 0.66-0.91). A significantly (p < 0.0001) higher PPV for [Ga]Ga-PSMA-11 (0.99, 95% CI 0.99-1.0 vs. 0.86) was found compared to [F]PSMA-1007 (0.86, 95% CI 0.82-1.00). Intervention efficacy analysis favoured [Ga]Ga-PSMA-11, where the number needed to image (to achieve a true positive finding) was 10.58 and the number needed to image to harm (to achieve a false positive finding) was - 8.08. A cost efficacy analysis favours [Ga]Ga-PSMA-11 in three of the four jurisdictions analysed where health economic data was available (Switzerland, Israel, Australia) and [F]PSMA-1007 in one jurisdiction (Denmark).

CONCLUSION

The analysis reveals a non-significantly higher PET positivity rate for [F]PSMA-1007, but finds significantly greater rates of uncertain findings and false positive findings when compared to [Ga]Ga-PSMA-11. We find differences in the two tracers in terms of clinical performance and cost efficacy. The method presented herein is generalisable and can be used with clinical or cost data for other countries or tracers.