Can active surveillance of prostate cancer be conducted without biopsies?
While patients on active surveillance undergo periodic prostate-specific antigen (PSA) and tumor burden assessments, which typically involve periodic prostate biopsies, "these are increasingly being augmented, and, in some very careful circumstances, replaced by MRI and/or biomarker studies," said Matthew Cooperberg, MD, MPH, of the Helen Diller Family Comprehensive Cancer Center at University of California San Francisco (UCSF). "I say very carefully because there are concerns that biopsies are being replaced by MRI and biomarkers too frequently, and maybe ahead of the evidence."
"The guidelines clearly say active surveillance is based on PSA and tumor biopsy," Cooperberg told 51˶. "There are parts of the world, the U.K. for example, where they believe MRI is an adequate replacement for prostate biopsy. I and many others do not think MRI in 2020, based on the current PI-RADS system, is anywhere close to ready to replace biopsies on a routine basis. The accuracy is just not there using the PI-RADS system, and the false negative rate for high grade disease -- nearly 25% -- is too high. MRI is a great augment to biopsy, but not a replacement."
According to an , recent data from the multicenter Canary Prostate Cancer Active Surveillance (PASS) Study cohort indicate that systematic biopsy should be performed on patients with negative magnetic resonance imaging and included in the management in patients with positive magnetic resonance imaging.
This study, led by Michael Liss of the University of Texas Health Science Center in San Antonio, included 361 patients who underwent 395 prostate MRIs with median follow-up of 4.1 years (IQR 2.0–7.6). The MRIs led to reclassification in 27% of cases. Positive predictive value for detecting GG ≥2 cancer was 31% (95% CI 26%-37%), while the negative predictive value was 83% (95% CI 76%-90%), "suggesting that a negative MRI will still miss a substantial proportion of patients with GG≥2 disease."
"In addition, systematic biopsies detected a similar number of unique GG≥2 cancer as targeted MRI cores," Liss and his colleagues wrote. "Thus, if the goal of surveillance biopsy is to identify higher-grade disease, both systematic and targeted biopsies should be obtained for men with a region of interest identified on MRI."
They also found that that while PI-RADS 5 lesions were significantly associated with upgrading or reclassification when compared to PI-RADS 1 and 2, models including PI-RADS scores were only minimally improved over models that contain clinical variables alone.
According to another recent study, published in , multiparametric (mp) MRI can improve the detection of clinically significant prostate cancer, but by itself it can't replace confirmatory or surveillance biopsies.
The study, led by Carissa Chu, MD, of UCSF, and co-authored by Cooperberg, included 344 men on active surveillance who had at least one mpMRI scan and biopsy after their cancer diagnosis. The men had 408 mpMRI scans during a median 71 months on active surveillance. The median time between prostate biopsies was 16.5 months.
The overall negative predictive value for mpMRI was 79.5% and ranged from 74.4% to 84.6% for all active surveillance biopsies up to the fourth surveillance biopsy. In men with PSA density ≥0.15 ng/ml/cm3, the overall negative predictive value for mpMRI was 65.5% and ranged from 57.1% to 73.3% across serial mpMRI scans.
"These findings support the hypothesis that mpMRI is helpful but insufficient to rule out pathological reclassification, especially at confirmatory biopsy or in the presence of other risk factors," wrote Chu and her colleagues.
"We are very interested in tailoring the intensity of the surveillance protocol," Cooperberg emphasized. "Biopsies are uncomfortable, they have risks of infection, and there are costs associated with them. And patients certainly don't want to sign up for 20 biopsies over 20 years."
Cooperberg and colleagues recently identified several clinical parameters that can predict disease progression and can be used to identify patients on active surveillance who can be followed less intensively. These include maximum percent positive cores, history of any negative biopsy after diagnosis, time since diagnosis, body mass index, prostate size, prostate-specific antigen at diagnosis, and prostate-specific antigen kinetics.
The group determined that a prediction model based on these parameters, and tested on 850 men in the PASS cohort, could potentially be a less invasive way of assessing disease change, and thus avoid multiple biopsies.
As for the role biomarkers should play in active surveillance protocols, Cooperberg said tests such as Decipher, Prolaris, or Oncotype DX Prostate can provide important prognostic information.
Earlier this year the American Society of Clinical Oncology published in which an expert panel recommended that proprietary tests may be offered in situations in which the assay result, when considered as a whole with routine clinical factors, is likely to affect management. However, they did not recommend routine ordering of molecular biomarkers.
The panel reached a similar conclusion for use of MRI and genomics in men with newly diagnosed cancer eligible for active surveillance: "only in situations in which the result, when considered with routine clinical factors, is likely to affect management" are such assessments clearly worthwhile.
"If a biomarker test is high, that will usually drive us to do a more aggressive schedule of re-biopsy," said Cooperberg. "Can we say to a patient now, 'Your biomarker tests were low so you can now defer the interval to your next biopsy for a longer period of time'? That makes some clinical sense, but the data really do not exist to support that, yet."
As far as ultimately replacing the biopsy, Cooperberg suggested that with better studies and longer term follow-up, physicians can probably think about ways of using MRI and biomarkers to increase biopsy intervals.
"But, we're not quite there yet as far as replacing biopsy all together," said Cooperberg. "We need next generation imaging, or we need better studies about the existing biomarkers to get around concerns about tumor heterogeneity, and other factors that just decrease confidence in the current generation of markers. Eventually, the holy grail is to have a liquid test that will be less expensive, and avoid the heterogeneity concerns. But, we are still years away from that degree of evidence."
Disclosures
Cooperberg disclosed relationships with AbbVie, Astellas, Bayer, Dendreon, Janssen, and Merck. Co-authors disclosed various patents and grant support from the NIH, the Canary Foundation, Myriad Genetics, and Decipher Biosciences.