New data from ProtecT highlight pitfalls in active surveillance of prostate cancer
The current protocol for risk stratification of prostate cancer at diagnosis is inaccurate, according to new data from the ProtecT (Prostate Testing for Cancer and Treatment) trial presented at the 15th Urological Association of Asia Congress (UAA 2017) held recently in Hong Kong.
In the ProtecT trial, 1,643 men with prostate-specific antigen (PSA)–detected localized prostate cancer were randomized to receive active monitoring (n=545), radical prostatectomy (n=553), or external-beam radiotherapy (n=545). Patients in the active monitoring arm were followed up with PSA testing and offered radical treatment on suggestion of disease progression or patient/clinician anxiety. The primary endpoint of prostate cancer mortality was ≤1.2 percent and did not differ significantly between groups (p=0.48). The secondary endpoint of all-cause mortality was approximately 10 percent and did not differ significantly between groups (p=0.87) either. [N Engl J Med 2016;375:1415-1424]
“ProtecT was criticized for having a largely low-risk population because 77 percent of patients had a Gleason score of 6 at baseline,” said investigator Professor Freddie Hamdy of the University of Oxford, UK. “This is a misconception. As shown in data recently submitted for publication, 40 percent of patients in the study had intermediate-risk disease and 2 percent had high-risk disease.”
Of the 17 prostate cancer–specific deaths that occurred in the study, eight were in the active monitoring group, five were in the surgery group, and four were in the radiotherapy group. Eight of the men who died of prostate cancer (three in the active monitoring group, three in the surgery group, and two in the radiotherapy group) had disease features for which active surveillance would be recommended by most guidelines.
“Interestingly, a majority of patients who died of prostate cancer had opted for or were randomized to receive active monitoring rather than active treatment,” said Hamdy. “In men who received surgery, none with a pure pathological grade of 6 died of prostate cancer. In men who did not progress to receive surgery, there was a significant number of Gleason 3+4 and 4+3 disease.”
“When we looked at PSA and progression, we found no difference between surgery and radiotherapy in patients with PSA levels <10 ng/mL. In patients with PSA levels >10 ng/mL, however, we found a strange pattern of higher progression rate with surgery than radiotherapy,” he continued. “The same pattern was found for Gleason score of 3+4 and ≥4+3.”
“What we have learnt from ProtecT is that active surveillance can be a valid option in men with low-risk and intermediate-risk disease. However, risk stratification at diagnosis is inaccurate, and genomic diversity of prostate cancer is our Achilles heel,” Hamdy suggested. [J Clin Invest 2013;123:4918-4922; Nat Genet 2015;47:367-372] “Accurately distinguishing lethal disease from nonlethal disease at baseline is our next challenge, which will produce a major paradigm shift.”
“Active surveillance is a by-product of irrational and inappropriate diagnosis of histological prostate cancer in men who will never suffer consequences from the disease if left untreated in their lifetime. It should therefore be replaced by reduced detection of insignificant prostate cancer, using innovative diagnostic pathways including prebiopsy imaging, liquid biopsies, and risk calculation prior to diagnosis,” he concluded. “Men diagnosed with prostate cancer should be better stratified at baseline using ‘omics’ to appropriately tailor treatment and interventions. Monitoring needs to improve to keep men in a genuine window of curability for intervention when necessary.”