A new prognostic tool, the nomogram, has demonstrated the ability to predict outcomes following treatment with lutetium-177 prostate-specific membrane antigen in patients with metastatic castration-resistant prostate cancer, and could potentially be used to inform future trial designs and clinical decisions.
A new prognostic tool, the nomogram, has demonstrated the ability to predict outcomes following treatment with lutetium-177 (177Lu) prostate-specific membrane antigen (PSMA) in patients with metastatic castration-resistant prostate cancer (mCRPC), and could potentially be used to inform future trial designs and clinical decisions, according to data from a retrospective study published in Lancet Oncology.1
The nomograms, which were developed by investigators at the UCLA Jonsson Comprehensive Cancer Center and those from 5 other institutions, included predictive variables, such as the time since initial diagnosis of prostate cancer, chemotherapy status, hemoglobin concentration at baseline, and [68Ga]Ga-PSMA-11 PET/CT parameters such as molecular imaging TNM classification and tumor burden.
Results from the study showed that the C-index of the overall survival (OS) model was 0.71 (95% CI, 0.69-0.73). Similar C-indices were observed with internal validation (0.71; 95% CI, 0.69-0.73) and external validation (0.72; 95% CI, 0.68-0.76). Moreover, the C-index for the prostate-specific antigen (PSA) progression-free survival (PFS) model was 0.70 (95% CI, 0.68-0.72). Again, similar C-indices were achieved with both internal (0.70; 95% CI, 0.68-0.72) and external (0.71; 95% CI, 0.68-0.74) validation. As such, both models were found to have acceptable calibration with predictions that correlated with observed outcomes.
“Until now, there has been no validated tool to adequately predict the response of patients with advanced metastatic prostate cancer to Lu-PSMA treatment,” said lead author Andrei Gafita, MD, a postdoctoral scholar in the Ahmanson Translational Theranostics Division of the department of molecular and medical pharmacology at the David Geffen School of Medicine, said in a press release.2 “Most important, this study suggests that screening with PSMA PET imaging could help to select patients who are most likely to benefit from this treatment.”
177Lu-PSMA is a radiolabeled small molecule inhibitor that binds with high affinity to PSMA and delivers β particle radiation. In June 2021, the FDA granted a breakthrough therapy designation to the agent based on results from the phase 3 VISION trial (NCT03511664), which demonstrated that when combined with standard-of-care treatment, 177Lu-PSMA resulted in a 40% reduction in the risk of death vs SOC alone in patients with progressive PSMA-positive mCRPC.3,4
Although many efforts have been made to understand tumor heterogeneity and to develop prognostic nomograms in patients with mCRPC who receive frontline or second-line treatment, no models exist to efficiently evaluate the prognosis for patients who are later in their disease course and are potentially eligible for treatment with 177Lu-PSMA. As such, investigators sought to develop nomograms as a way to predict outcomes with the therapy in this population.
For the multicenter, retrospective study, predictive nomograms were developed using data collected from 6 institutions across the United States, Germany, and Australia—all of whom had screened patient who had received treatment with 177Lu-PSMA from December 10, 2014, to July 19, 2019, as part of a phase 2 clinical trial (NCT03042312, ACTRN12615000912583).
To be eligible, patients had to have received an intravenous injection of 177Lu-PSMA, at a dose ranging from 6.0 GBq to 8.5 GBq once every 6 to 8 weeks for a maximum of 4 to 6 cycles, without experiencing disease progression or notable toxicity. Additionally, patients had to have a [68Ga]Ga-PSMA-11 PET/CT scan, clinical data, and information on survival outcomes available. Patients who received 18F-labeled PSMA-PET at baseline were excluded.
Overall, 270 were enrolled to the trial; these patients were assigned to a development cohort for model development (n = 196) or a validation cohort for model validation (n = 74). The primary outcomes for the nomograms were OS and PSA-PFS.
Data on 26 different pretherapeutic parameters were collected from patients, 18 of which were tested in models as putative predictors of outcome after treatment with 177Lu-PSMA. Some of the pretherapeutic parameters included information on demographics, initial diagnosis of prostate cancer, treatment history, baseline clinical status, baseline laboratory tests, and baseline [68Ga]Ga- PSMA-11 PET/CT tumor characteristics.
Among the 270 patients included, the median age across the cohorts was 71 years (range, 65-76), and the median time since initial diagnosis was 6.5 years (range, 4-12). Fifty-eight percent of patients had received treatment with 177Lu-PSMA as part of compassionate access programs, and 42% had received the therapy as part of a clinical trial. The therapy was administered for a median of 3 cycles per patient. Patients had received a median of 3 prior systemic treatments, with 95% having received treatment with second-generation antiandrogens and 81% having previously received chemotherapy.
At a median follow-up of 21.5 months, 83% of the 270 patients had died and 91% experienced PSA progression.
Additional data from the study indicated that the estimated 12-month OS rate was 54% (95% CI, 48%-60%), and the estimated 18-month OS rate was 34% (95% CI, 29%-40%). Predictors selected in this model included time since cancer diagnosis, chemotherapy status, baseline hemoglobin concentration, bone and liver involvement status, number of metastatic lesions, and tumor SUVmean.
Moreover, the estimated 3-month PSA-PFS rate was 64% (95% CI, 58%-70%); at 6 months, the estimated rate was 38% (95% CI, 32%-43%). Selected predictors for this model were time since cancer diagnosis, chemotherapy status, pelvic nodal status, bone and liver involvement status, and tumor SUVmean.
Additionally, patients in the development cohort, the validation cohort, and the complete set were stratified into 2 risk groups using the calculated optimal cutoff for the risk scores; this was 197 points for the OS nomogram and 178 points for the PSA-PFS nomogram.
In the development cohort, the median OS for low- vs high-risk patients were 19.1 months (95% CI, 17.1-21.1) and 8.4 months (95% CI, 7.4-9.4), respectively (P <.0001). The median PSA-PFS in this cohort was 9.4 months (95% CI, 6.6-12.1) in patients who were high risk vs 3.3 months (95% CI, 2.9-3.7) in low-risk patients (P < .0001).
In the validation cohort, patients who were low risk were found to have a significantly longer median OS than those who were high risk, at 24.9 months (95% CI, 16.8-27.3) vs 7.4 months (95% CI, 4.0-10.8), respectively (P <.0001); this was also true with regard to PSA-PFS, at 6.6 months (95% CI, 6.0-7.1) vs 2.5 months (95% CI, 1.2-3.8), respectively (P = .022).
In complete set, the median OS for low- and high-risk patients were 19.9 months (95% CI, 17.5-22.3) vs 8.2 months (95% CI, 7.2-9.1), respectively (P <.0001). The median PSA-PFS in these subsets were 8.8 months (95% CI, 7.3-10.3) vs 3.3 months (95% CI, 2.8-3.7), respectively (P <.0001).
“The findings from the research are encouraging and may provide the foundation for patient selection for LuPSMA therapy,” Gafita said. “Nevertheless, until its clinical validity is demonstrated in prospective trials, the tool should be used cautiously and should not replace the clinical judgement of treating physicians.”