Featured16 November 2026PSMA: PET and radioligand therapy

Total tumour volume on PSMA PET: predicting survival with a single number

The latest analysis of the PROMISE PET Registry, covering 13,893 patients (Karpinski et al., J Nucl Med, November 2026), reported that among the quantifiable PSMA PET metrics, PSMA-positive total tumour volume (PSMA TTV) best predicted overall survival. This article summarises the study, what the measurement is, and how it relates to the same group's earlier risk tools.

511 keV × 2PSMAExtracellularIntracellularto scale · 5 nmagent symbol not to scale
Left: prostate, prostate cancer; the orange focus marks the site of disease (schematic); small foci are metastases; the blocks on either side are PET detectors. Right: the agent binds PSMA on the cell membrane. This panel is to scale (membrane ≈ 4 nm; extracellular domain (homodimer) ≈ 10.4 nm wide and ≈ 7.7 nm tall, measured from PDB 1Z8L coordinates); the agent symbol is not to scale. Prostate cancer · PSMA (prostate-specific membrane antigen). Schematic; derived from the item's title and summary. Target and mechanism from a sourced table. UniProtKB Q04609 (FOLH1): subcellular location, cell membrane.

Why a single number?

Reading PSMA PET with the PROMISE criteria is a strong prognostic tool for overall survival; the same group's PROMISE risk nomograms were equal or superior to clinical risk scores 2,3. Nomograms, however, need several variables. This study asked whether a single quantifiable PET metric could do the job instead 1.

What did the study do?

  • Karpinski et al., J Nucl Med (November 2026) [1]: International PROMISE PET Registry; 13,893 patients with histologically proven prostate cancer who had PSMA PET between December 2012 and April 2024. Patients were split 2:1 into development (9,271) and validation (4,622) cohorts. PSMA TTV and whole-body SUV were measured with predefined SUV thresholds; PSMA total lesion uptake (PSMA TL = TTV × SUV) and TTV-adjusted PSA (PSA × TTV) were calculated. Accuracy was compared with Harrell's C-index.

Findings

  • C-index in the validation cohort: PSMA TTV 0.78 (95% CI 0.77-0.80); superior to SUV (0.62) and PSA at PET (0.71), similar to PSMA TL (0.78) and PSA × TTV (0.77) 1.
  • Overall survival by quartile (whole cohort): Median overall survival was not reached for quartile 1 (0 mL) and quartile 2 (>0-2.1 mL); it was 11.4 years in quartile 3 (>2.1-16.2 mL) and 4.4 years in quartile 4 (>16.17 mL) 1.

Context: the group's risk tools and response assessment

  • PPP nomograms, Lancet Oncol 2024 [2]: 2,414 patients from four German centres; the quantitative nomogram included tumour volume and tumour mean SUV among its variables. Validation C-indices were 0.80 (internal) and 0.77 (external).
  • PPP3, Lancet Oncol 2026 [3]: 11,154 registry patients from 35 sites; C-indices 0.83 and 0.84 for the visual and quantitative nomograms; a simplified risk table for 3-, 5- and 7-year survival.
  • RECIP 1.0, J Nucl Med 2022 [4]: Interim PET response during Lu-177 PSMA therapy also rests on total PSMA-positive tumour volume (PSMA-VOL): partial response is a ≥30% decrease with no new lesions, progression a ≥20% increase with new lesions.

Practical points

  • Prognosis, not treatment selection: The study tested prediction of overall survival; it did not show that TTV tells which patient benefits from which treatment. The quartile limits (2.1 and 16.2 mL) are descriptive values from this registry, not validated decision thresholds 1.
  • The measurement depends on the method: TTV was obtained by segmentation with predefined SUV thresholds 1. Before using the same number in another centre, the segmentation method should be confirmed to be the same (comment).
  • In the report, volume says more than SUV alone: In this registry the ability of whole-body SUV to separate survival (C 0.62) was clearly lower than that of volume (0.78) 1. The authors expect automated TTV measurement to allow fast and reproducible prediction 1.
  • Limit of the design: Retrospective registry data; patients were imaged at all disease stages and with different PSMA agents. A subgroup analysis by stage was done 1.

Test yourself

  1. In the PROMISE registry analysis by Karpinski et al. (J Nucl Med 2026), which metric had the lowest C-index in the validation cohort?

  2. In the same study, which group had a median overall survival of 4.4 years?

  3. Under RECIP 1.0, what defines partial response on interim PET during Lu-177 PSMA therapy?

References

  1. Karpinski MJ, Civan C, Hoberück S, et al. PSMA total tumor volume: a simple quantifiable biomarker for overall survival prediction—results from the international PROMISE PET Registry. J Nucl Med. 2026 (online 12 Nov). doi:10.2967/jnumed.126.273371 PMID 43020616
  2. Karpinski MJ, Hüsing J, Claassen K, et al. Combining PSMA-PET and PROMISE to re-define disease stage and risk in patients with prostate cancer: a multicentre retrospective study. Lancet Oncol. 2024;25(9):1188-1201. doi:10.1016/S1470-2045(24)00326-7 PMID 39089299
  3. Karpinski MJ, Civan C, Rauscher I, et al. New prostate cancer risk groups by PSMA-PET (PPP3): an international, retrospective, registry-based cohort study. Lancet Oncol. 2026;27(4):480-490. doi:10.1016/S1470-2045(26)00016-1 PMID 41819113
  4. Gafita A, Rauscher I, Weber M, et al. Novel framework for treatment response evaluation using PSMA PET/CT in patients with metastatic castration-resistant prostate cancer (RECIP 1.0): an international multicenter study. J Nucl Med. 2022;63(11):1651-1658. doi:10.2967/jnumed.121.263072 PMID 35422442
Citing this piece

If you use this piece in a presentation, paper or thesis, add this line to your reference list (Vancouver style). "Copy" puts it on your clipboard; to cite the studies it reviews, use the numbered references above.

Nuclear Medicine Explained. Total tumour volume on PSMA PET: predicting survival with a single number [Internet]. Last updated 16 November 2026. Available from: https://nuclearmedicineexplained.com/en/updates/psma-total-tumour-volume

Published after physician editor approval · last updated 16 November 2026

Revision history
  1. 16 November 2026: First draft. Details verified on PubMed.
  2. 16 November 2026: Article and three "Test yourself" questions published after physician editor approval.

Prepared by Claude from primary sources and published after approval by the physician editor. It summarises what the evidence says; it is not a treatment recommendation.