ProPSMA-PET

Lancet. 2020 Apr 11;395(10231):1208-1216. doi: 10.1016/S0140-6736(20)30314-7. Epub 2020 Mar 22.

Prostate-specific Membrane Antigen PET-CT in Patients With High-Risk Prostate Cancer Before Curative-Intent Surgery or Radiotherapy (proPSMA): A Prospective, Randomised, Multicentre Study

Hofman et al

https://pubmed.ncbi.nlm.nih.gov/32209449/

Clinical Question

In men with high risk, biopsy proven prostate cancer, is PSMA PET scan more accurate in detecting nodal or metastatic disease compared to standard staging with WBBS and CT?

Participants: men with high risk, biopsy proven prostate cancer Intervention: PMSA PET scan as first line imaging Control: bone scan and CT as first line imaging Outcome: accuracy of first line investigation in detecting nodal or metastatic prostate cancer

Bottom Line

This is a multicentre, two arm Australian trial involving 302 patients across 10 sites. It demonstrated PSMA PET scan to be more accurate than traditional prostate cancer staging imaging modalities (WBBS and CT) in the detection of nodal and distant metastatic disease.

Major Points

Prostate cancer is a common malignancy worldwide and many treatment options are available. Decision making about treatments requires timely and accurate staging information. This paper examines staging tools for men with high risk prostate cancer. In comparison to WBBS and CT (conventional staging tools), PET PSMA has higher specificity and sensitivity in the detection of nodal or metastatic disease. Using CT PET PSMA for staging leads to a higher rate of change in patient management compared with conventional staging. In comparison to WBBS and CT, PET PSMA has a lower radiation dose.

Guidelines

American Urology Association guidelines: recommend WBBS and CT for high risk prostate cancer (Gl >7/ISUP 4 or PSA >20). European Urology Association: recommend to “perform metastatic screening including at least cross-sectional abdominopelvic imaging and a bone-scan” for high risk prostate cancer (Gl >7/ISUP 4, PSA >20 or cT2c [tumour involving both lobes on DRE]). Guidelines acknowledge the potential for PSMA PET but currently do not recommend it as first line due to insufficient evidence. Note that this study classes high risk as ISUP 3 and above, whereas guidelines class high risk as ISUP 4 and above

Design

• Trial type: multicentre, two-arm, randomised trial
• N: 302
  • Experimental arm: 150
  • Standard: 152
• Setting: 10 Hospitals in Australia
• Enrollment: March 2017 to Nov 2018
• Mean follow-up: 6 months +/- 30 days
• Primary outcome: Accuracy of first line imaging for identifying either pelvic nodal or distant metastatic disease

Population

Inclusion Criteria

• 18 years or older with
• histopathologically-confirmed prostate cancer and were being considered for radical prostatectomy or radiotherapy with curative intent
• All patient had high risk features including at least one of
• prostate-specific antigen (PSA) concentration of 20 ng/mL or more within 12 weeks before randomisation
• International Society of Uropathology (ISUP) grade group 3–5
• clinical stage T3 or worse

Exclusion Criteria

• Imaging done for staging within 8 weeks before randomisation, with the exception of MRI of the prostate before biopsy

Baseline Characteristics

• Outlined in inclusion criteria
• Nil significant between group differences

Interventions

• PMSA PET scan as first line imaging compared against control of conventional imaging
• cross over to receive the alternate image, unless 3 or more distant metastases had been observed

Outcomes

Comparisons are intensive therapy vs. standard therapy.

Primary Outcomes

• Accuracy of first line imaging for identifying pelvic nodal disease or distant metastatic disease = comparative accuracy of PSMA PET versus conventional imaging (CT + WBBS)
  • PSMA 27% absolute greater AUC for accuracy (95% CI 23–31, p<0·0001)
  • 85% versus 38% sensitivity for PSMA versus conventional imaging
  • 98% versus 91% specificity for PSMA versus conventional imaging

Secondary Outcomes

• Change in patient management (pre-staging imaging compared to post-staging imaging). Stratified as high (change in management intent or modality), medium (change in modality delivery), low (management plan was not altered).
  • Higher rate of “high” or “medium” effect management change with first-line PSMA (28% versus 15%, p = 0.008)
  • First-line PSMA PET = 14% directed from curative to palliative intent treatment
• Incremental accuracy of second-line imaging to change stage of regional lymph node or distant metastatic disease
  • Higher rate of “high” or “medium” effect management change with second-line PSMA (27% versus 5%)
  • Second-line PSMA 17% absolute greater AUC for accuracy (95% CI 23–31, p<0·0001)
• Effective radiation exposure (mSv)
  • 8.4 mSv versus 19.2 mSv for PSMA PET versus conventional imaging (p < 0.001)

Subgroup Analysis

• Sensitivity analysis (lesions rated as equivocal were considered positive rather than negative)
  • No clinically or significantly significant difference in primary outcome with PSMA 28% absolute greater AUC for accuracy
• Pelvic nodal disease
  • PSMA 32% absolute greater AUC for accuracy
• Distant metastatic disease
  • PSMA 22% absolute greater AUC for accuracy
• First-line PSMA
  • 20% pelvic nodal disease
  • 9% abdominal nodal disease
  • 10% bone metastases
• Equivocal findings
  • Greater with conventional imaging (23% versus 7%, p < 0.001)

Adverse Events

No adverse events reported for PSMA PET

Criticisms

Below are the Cochrane Risk of Bias tool elements for this study:

• Risk of bias from randomisation: Low

• Risk of bias due to deviation from the intended intervention (effect of assignment to intervention): Low, however there was no ability to blind interventions.

• Risk of bias due to deviation from the intended intervention (effect of adhering to intervention): Low. More patients in PSMA PET group did not receive 2nd line imaging because due to finding significant metastatic disease burden. However this was part of the protocol, and did not affect the primary outcome.

• Risk of bias in measurement of the outcome: Low/Some concern.
  • Was the method of measuring the outcome appropriate? Yes, although to some extent imaging results acted as their own standard which may skew the results to appear more accurate.
  • Having a multicentre trial would have reduced the risk of having differences in measuring/ascertaining the outcome
  • Radiologists knew the results of the 1st line imaging, possible risk of bias in reporting the 2nd line imaging.
  • Not all patients received a histological diagnosis to confirm the accuracy of the imaging (i.e. received radiotherapy despite having possible pelvis lymph node disease)
  • Assessors identified that their ‘soft’ criteria for diagnosis were not conventional and might be subject to investigator bias.
  • Not so much a problem for the primary outcome, but due to the cross-over nature of the study, but downstream effects on management could not be measured.

• Risk of bias in selection of the reported result: Low. Result was about diagnostic accuracy of the 1st line imaging, This was part of the plan, and I don’t think the assessors picked this from one of multiple possible outcomes.

Funding

Movember and the Prostate Cancer Foundation of Australia. Both not profit organisations.

Some involved members of the study ‘declaration of interests. Mostly in regards to grants from various foundations including Movember, Prostate Cancer foundation. Some declared interests from pharmaceutical companies. Unlikely to influence results given this was a study on diagnostic accuracy, but if that influences the diagnosis and treatment received, possibly reflects a conflict.

However, no author declared any competing interests.

Funder had not role in study design, data collection, data analysis, data interpretation or writing of the report according to authors.

Further Reading

• A follow-up paper regarding the economics of PSMA-PET is expected.