18F-DCFPyL PET/CT Imaging in Patients with Suspected Recurrence of Prostate Cancer (CONDOR) Journal Club - Christopher Wallis & Zachary Klaassen

June 28, 2021

In this UroToday Journal Club, Christopher Wallis and Zachary Klaassen discuss the Clinical Cancer Research publication on the results from the CONDOR Phase 3, multicenter study. In particular, conventional imaging performs poorly when PSA levels are low, less than 2 ng/mL, and the use of novel PET radiotracers may address this issue and provide more sensitive imaging. CONDOR was a prospective study designed to determine the performance of radiotracer 18F-DCFPyL-PET/CT in patients with biochemical recurrence and uninformative standard imaging. Drs. Klaassen and Wallis review the study design and findings. The Performance of 18F-DCFPyL-PET/CT achieved the study's primary endpoint, demonstrating disease localization in the setting of negative standard imaging and providing clinically meaningful and actionable information. These data further support the utility of 18F-DCFPyL-PET/CT to localize disease in men with recurrent prostate cancer. The ability of 18F-DCFPyL PET/CT to localize the extent of recurrent disease offers physicians the opportunity to adjust and tailor their management planning and potentially improve outcomes in men with biochemical recurrence.


Christopher J.D. Wallis, MD, Ph.D., Instructor in Urology, Vanderbilt University Medical Center, Nashville, Tennessee

Zachary Klaassen, MD, MSc, Urologic Oncologist, Assistant Professor Surgery/Urology at the Medical College of Georgia at Augusta University, Georgia Cancer Center

Read the Full Video Transcript

Christopher Wallis: Hello, and thank you for joining us for this UroToday Journal Club. Today, we are discussing a recently published paper from the CONDOR trial, examining the diagnostic performance of 18F-DCFPyL PET/CT in men with biochemically recurrent prostate cancer, results from the CONDOR Phase 3 Multicenter Trial. I'm Chris Wallis, a Fellow in Urologic Oncology at Vanderbilt. And with me is Zach Klaassen, an Assistant Professor in Urology at the Medical College of Georgia. Here is the citation for the CONDOR trial led by Dr. Michael Morris and recently published in Clinical Cancer Research. Most UroToday viewers will know that biochemical recurrence is a relatively common undertaking after local therapy for prostate cancer, regardless of whether men are treated with radical prostatectomy or radiotherapy. Between 20% and 50% of men will experience biochemical recurrence within 10 years following their initial treatment. And many, if not most, of these men, will have biochemical evidence of disease with PSA recurrence in the absence of radiographic evidence of metastases using conventional imaging with CT and bone scan.

In particular, conventional imaging performs poorly when PSA levels are low, less than 2 ng/mL, and the use of novel PET radiotracers may address this issue and provide more sensitive imaging.

So one such radiotracer of interest is 18F-DCFPyL. This is a urea-based small molecule that binds to the extracellular domain of PSMA, and you can see its molecular structure here in the top right, as well as its PSMA binding site here in the extracellular domain. And the use of this radiotracer was initially demonstrated in OSPREY, which was the initial staging of high-risk prostate cancer, as well as the detection of suspected recurrent lesions seen on conventional imaging. And that paper was very recently published in the Journal of Urology. For the CONDOR study, looking at biochemically recurrent disease, the authors included men aged 18 years and older with biochemically recurrent prostate cancer defined as a PSA greater than 0.2 and rising following surgery or a PSA that is at least 2 ng/mL above nadir for those who had undergone radiotherapy. Obviously, to meet this definition, patients had to have had local prior therapy, and further, they had to have a negative or equivocal standard of care imaging. And for the purposes of CONDOR, our standard of care imaging is defined as MRI, bone scan, CT scan, Axium PET/CT, or a choline PET/CT.

Of note, patients were excluded if they had prior high-energy gamma-emitting radioisotopes within five half-lives prior to use of the experimental agent, as well as if they had ADT within three months of imaging, investigational treatments for prostate cancer within 60 days of imaging, or any other ongoing systemic therapies for prostate cancer.

This was a multicenter open-label phase 3 single-arm study, and 18F-DCFPyL was administered through intravenous bolus one to two hours before a PET/CT was performed. Patients emptied their bladder, and then a PET scan was performed, as well as a non-contrast CT scan. These images were then uploaded for central review, and interpretation was performed by three independent blinded reviewers. We will talk more about the comparison later, but the SOT evaluation was performed by two blinded reviewers as a separate group from the interpretation of the PET/CT scans.

So the primary outcome of interest was driven by a discussion with the FDA and comprised of what was deemed a composite standard of truth. This was a hierarchical tiered assessment based on first evaluable histopathology, which could be from a prostatectomy, salvage lymphadenectomy, or biopsy. Absent this, correlative follow-up imaging using MRI, CT scan, Axium scan, or choline CT could also be used.

And in the absence of this, it confirmed PSA response up to nine months, following targeted radiotherapy of PET-positive sites, using a PSA50 criterion as the third component of this composite standard of truth. The primary endpoint was a correct localization rate. This is essentially a positive predictive value while considering anatomic co-localization. And so the study defined this as the proportion of patients with a one-to-one correspondence between identified lesions on their PET/CT and the composite standard of truth. So we calculated this the same way we would a positive predictive value, 100 times the true positive over true positive plus false positive. Secondary endpoints included the proportion of patients who had a change in their intended management as a result of the scan being performed, as well as safety. And exploratory endpoints examined the detection rate and positive predictive value of PET/CT on the basis of regions and the function of the PSA level at the time of a PET/CT scan.

So this here highlights the study schema, and again, we've discussed this, but we will go through it. So before PET/CT was performed, the patient's treating physician performed a clinical management questionnaire indicating what treatment would be recommended on the basis of the available data at that time. Subsequently, they underwent whole-body PET/CT. Following this, clinicians completed another clinical management questionnaire, which was used to assess whether there was a change in management and for patients who had at least one lesion detected on their PET/CT scan. And this was then compared to the composite standard of truth.

In order to design the study, the authors relied on a prior meta-analysis, which suggested that approximately three-quarters of patients would have a positive PSMA scan when accrued among the patient population, as the CONDOR study was designed for. The authors then assumed that they would have 60% positive scans, and 30% would have a confirmatory standard of truth. Thus, 18F-DCFPyL is expected to detect recurrent disease in 18% of the overall accrued patients. And they hypothesized the conventional imaging would detect less than 5% of these lesions. To allow for power to detect this difference, 81 positive scans or 134 valuable patients would be required. And thus, they increased the sample size to 192 patients to allow for a 30% loss rate.

They defined the primary outcome, as we discussed, as the correct localization rate. And they said that this would be successful if the lower limit of the 95% confidence interval exceeded 20% for two out of three readers. The secondary outcome was the direct proportion and confidence intervals, as we previously discussed. In terms of exploratory endpoints, the detection rate was defined as the proportion of positive scans among all scans performed and positive predictive value defined in a conventional way. Further, as they used three readers for all of these PET/CTS, they assessed inter-and intra-reader reliability. And so first they performed central reader inter-reader agreement, and they assessed this using Fleiss's generalized kappa. They then assessed for agreement between the centralized review, as well as local readers performed in the individual patient's treatment center, and compared this using Cohen's pairwise kappa.

Finally, among their central readers, they performed intra-reader agreement in which a 20% subsample of studies were presented to the same reader at least 28 days apart. And for each individual, their intra-reader agreement was compared again using Cohen's pairwise kappa. At this point in time, I will now hand it over to Dr. Klaassen who will walk us through the results of the CONDOR trial.

Zachary Klaassen: Thanks Chris. So there were 270 patients that were consented. There were nine screening failures, and then subsequently 208 patients underwent 18F-DCFPyL PET/CT scan. 132 patients with the lesion had follow-up by SOT, as Chris previously explained. This included 31 patients with pathology verification, a 100 with correlative imaging, and just one patient with response to external beam radiotherapy.

This is the table looking at baseline characteristics in this trial, a pretty standard subset of patients with advanced prostate cancer. And you can see here, the median age was 68. There were 71 months from a prostate cancer diagnosis. 49.5% percent of patients had a radical prostatectomy only, 14.9 had radiotherapy only, and 35.6% had radiotherapy plus radical prostatectomy. 27.9% of patients had at least one prior systemic therapy. In this trial, the majority of patients had less than Gleason 8 disease, at about three-quarters of the patients. The median PSA at the time of trial enrollment was 0.8. And you can see here the breakdown between the PSA group of 68.8% of patients having a PSA less than 2, 31.2% of patients with a PSA greater than 2.

This table looks at the disease detection and correct localization rate across three independent readers. I'll turn your attention to the box at the bottom of the table, which looks at the correct localization rate amongst these readers. Reader one was 85.6%, reader two 87%, and reader three 84.8%. So a high level of correct localization rate with very comparable results between these three readers.

This figure looks at the correct localization rate by baseline PSA levels. At a PSA <0.5 it was 73.3%, for PSA <0.5 to <1.0 was 75%, for 1.0 to <2 was 83.3%, for 2.0 to <5.0 PSA was 91.3%. And for  >5.0 was 96.4%.

This figure looks at the detection rate by baseline PSA levels. So lower than we saw in the previous figure. PSA <0.5, 36.2%, 0.5 to <1.0 at 51.4%, 1.0 to <2.0 was 66.7%, 2.0 to < 5.0 was 84.8%, and >5.0 was 96.7.

This figure looks at the positive predictive value by anatomic region, which was highest for detection of the prostate. M0 at 79.5%, 70.9% for pelvic lymph nodes or N1 disease, and 67.4% for extra-pelvic or M1 disease.

This table looks at the positive predictive value for the extra-pelvic region, which is lower than the previous figure in terms of lymph nodes, which were M1a 61.5%, bone or M1b at 62.5%, and 28.6% for visceral soft tissue or M1c disease.

This figure looks at a change in planned medical management. The dark boxes show patients with a positive 18F-DCFPyL scan, and the lighter boxes show patients with a negative 18F-DCFPyL scan. And so looking at this figure in more detail, there were 131 patients that had a change in management, which included 78.6% of patients with a positive scan. In looking at more specific details, this included 58% that changed from salvage local therapy to systemic therapy, which comprised 78.6% of patients with a positive scan. There were 43 patients with noncurative systemic therapy that were changed to salvage local therapy with roughly even distribution between positive and negative scans. And probably the most important comment from this figure is that 48, sorry, 49% of patients switched from observation to initiating therapy with nearly universal positive imaging in this subgroup.

So several discussion points from the CONDOR trial. This population had a PSA of <2.0 in 68.8% of patients, <1.0 in 52.5% of patients, and <0.5 in 34.2% of patients, thus, providing prospective evidence of diagnostic accuracy to reliably detect prostate cancer recurrence at PSA levels where we know that conventional imaging is suboptimal. The performance of 18F-DCFPyL PET/CT scan at a correct localization rate of more than 73%, which is comparable to previously published studies looking at 68 gallium-PSMA in men with biochemical recurrence that had a positive predictive value of 0.84 by histological comparison and a positive predictive value of 0.92 by the composite reference standard.

The ability of 18F-DCFPyL PET/CT to localize the extent of recurrent disease offers physicians the opportunity to adjust and tailor their management planning and potentially improve outcomes in men with biochemical recurrence.

So in conclusion, the CONDOR trial met its primary endpoint of a high correct localization rate and demonstrated that information provided by 18F-DCFPyL PET/CT was associated with frequent changes in management plans.

This study shows that PSMA-targeted PET radiopharmaceuticals labeled with 18F can offer an alternative to 68 gallium agents. And this data further supports 18F-DCFPyL PET/CT as an imaging tool to detect recurrent prostate cancer, even in patients with low PSA levels. Thank you for your attention to this UroToday Journal Club discussing the CONDOR trial.