Has use of PSMA PETCT revealed increased local failures than previously known after definitive prostate EBRT with biochemical failure?

Prior to the advent of PET imaging, the published rates of local recurrence (LR) after definitive RT vary widely in phase III trials from ≈ 1% (e.g., PCS IV) to ≈ 30% (e.g., PROG 9509). The heterogeneity is likely explained by several factors including (1) differences in baseline risk of local recurrence due to a patient’s NCCN risk stratum (Zelefsky et al., PMID 33587971), (2) differences in treatment intensity (i.e., RT dose as seen in RTOG 0126 or FLAME or the addition of ADT seen in RTOG 0815), (3) the trial definition of local failure (i.e., DRE vs MRI vs. PSMA vs. Biopsy), and (4) the length of follow-up over which patients are observed as local recurrences may present many years after EBRT. One additional factor important to contextualize these results is the consistency with which patients actually undergo the specified, pre-defined investigations to adjudicate a local recurrence in a prospective study. For example, when post-treatment biopsy is part of the definition, many patients may decline as has been noted in prior trials, which presents yet another challenge to understanding the true rate of local failure by a certain definition.

Regarding the absolute rate of PET/CT-specific local failure detection, a recent publication of NCT02761889 profiled 444 patients (4% LR, 23% FIR, 61% UIR, 12% HR) who underwent dose-escalated SBRT (45 Gy/5#) with or without ADT (36%). At a median follow-up of 58 months, 8% (n=37) were observed to have biochemical failure (BF). Of the patients with 68Ga-PSMA PET/CTs available at the time of BF (n=35), 71% had evidence of local relapse translating to a raw proportion of approximately 6% in the entire study population. 

We have several trials here at the NCI investigating local recurrence after primary radiotherapy, and we recently reported a patterns of failure study in a combined cohort of these trials – NCT01834001, NCT02190279, NCT03181867 (Salerno et al., PMID 35523680). All patients underwent MRI and biopsy, and two-thirds of patients underwent PSMA PET/CT imaging. One finding of relevance to the present discussion was that ≈ 10% of patients had LR (prostate/seminal vesicle), which was detectable on PSMA PET but not on mpMRI. A precise estimate of this rate, however, is limited by several factors including the small number of patients in this subgroup. Further, the sensitivity of mpMRI itself to detect a recurrence will be dependent not only on the MRI technique but also on the experience of the interpreting radiologist, and many centers may not employ mpMRI as a method to detect local recurrence for this reason. However, to the point of the question, these findings do characterize a subset of local failures which are MRI-negative, PSMA-positive, suggesting an increase in detection for PET/CT when MRI is the referent. A similar retrospective study from the combined experience of the NCI and JHU also found that PET/CT increased the detection of LR compared to the combination of CT/Tc99m bone scan (Mena et al., PMID 34916246).

The question of whether or not the advent of PET/CT restaging will increase the rate of detection of LR depends on what the referent is. It is almost certain however that PET/CT will increase rates in comparison to methods with lower sensitivity such as a post-treatment DRE or conventional imaging, as described. It seems unlikely, however, that PET/CTs could increase rates above those estimated by the systematic, planned application of delayed post-treatment biopsies, as this is considered the gold standard. In fact, it is possible that post-treatment biopsy can possibly misclassify a local recurrence if conducted too early, as it has been characterized that many patients may have delayed resolution of disease on post-treatment biopsy (Crook et al., PMID 10974448). Given this observation and the possibility of delayed resolution, it will be important to (1) profile treatment response on PET/CT after RT, (2) profile the natural history PSMA+/conventional imaging-negative findings for patients with BF, and (3) obtain tissue verification of these findings when intervention is planned based on this result. Several studies are ongoing here at the NCI to profile treatment response by PET/CT after treatment (NCT05155046), describe the natural history of PSMA+ findings for men with BF (NCT05588128), and diagnose local radio-recurrence (NCT01834001). Providers may be free to get in touch if you wish to refer a patient for one of these trials as they are open to all patients willing to travel to the NCI.

In summary, it is likely that the increased utilization of PET/CT for patients with biochemical failure will increase the detection of local recurrences. This will be important as management paradigms progress to include potentially curative salvage strategies with manageable toxicities such as salvage brachytherapy (Crook et al., PMID 34740768) or salvage focal SBRT (Patel et al., PMID 37442430) for patients with isolated LR. Further studies will be required to determine the optimal time-point to conduct imaging and the conditions under which positive in-gland findings after definitive radiotherapy merit salvage therapy. Providers may increase the rate of detection of isolated LR by initiating workup prior to the nadir + 2.0 definition upon multiple consecutive PSA rises, especially after RT without ADT; however, early PET-positive findings after treatment may not be representative of recurrence/persistence but instead delayed resolution. Biopsy-verification of LR will be important prior to salvage local therapy given the possibility of serious toxicity, and at this time such treatments may be best delivered at high-volume centers with significant expertise. 

Yes

Anecdotally, with > 300 institutional PSMA scans, yes. MpMRI is still good for prostate since renal excretion of PET tracer limits sensitivity.