SAN FRANCISCO––The proliferation marker Ki-67 has been known for decades and is widely used in breast cancer prognosis. Now data suggest that it has prognostic value in prostate cancer as well, according to a study presented here at the Genitourinary Cancers Symposium.
When investigators incorporated Ki-67 into a multivariate model with Gleason score, prostate specific antigen (PSA), and clinical stage, the ability to identify patients at risk of cancer-specific death following radical prostatectomy improved significantly compared with the same model without Ki-67. Based on these data and data from two previous studies, some experts are saying that the marker is now ready for regular clinical use in prostate cancer patients.
“When we are looking for ways to stratify patients with prostate cancer, we're looking for new markers of disease aggression,” said the researcher who presented the data, Matthew K. Tollefson, MD, a fellow in the Urologic Oncology Department at the Mayo Clinic. “The argument I was trying to make is that we have this marker already: It is cheap, it has been available, it is generic, and it provides probably the best prognostic ability over and above what we already use.”
That view was echoed by Nicholas J. Vogelzang, MD, Chair and Medical Director of the Developmental Therapeutics Committee and a member of the Genitourinary Committee for US Oncology Research, who was not involved in the study. “I thought [the presentation] was fascinating. For all this fancy gene microarray and heatmaps, here we've got this little old fashioned Ki-67 that says ‘If it grows fast it is bad, if it doesn't grow fast it's good.’”
The study included 1,149 patients who had undergone a biopsy and radical prostatectomy at the Mayo Clinic between 1996 and 2005. Mayo pathologists have routinely tested prostate biopsy samples with the anti-Ki-67 antibody, MIB-1, since 1996. Thus, the pathology report includes a quantitative assessment of the percentage of cancer cell nuclei that stain positive for the marker.
The median age of men in the study population was 64, and their mean PSA score was 5.6. More than half of the men (65%) had a Gleason score of 6 on biopsy and most (85%) had clinical Stage T2 or T3 tumors. Additionally, 21% had a pathologic T3 tumor and 3.7% of patients had node-positive disease.
Ki-67 positivity was strongly associated with pathologic stage, Gleason score, tumor volume, and nodal status. Men whose tumors had 5% or more of their cells positive for Ki-67 were significantly more likely to have biochemical failure within 10 years of surgery, based on a Kaplan-Meier analysis, than men whose tumors had fewer than 5% of cells Ki-67 positive. They were also more likely to suffer systemic progression or cancer-specific death during the 10-year follow-up.
“Systemic progression was extremely unlikely in patients with a proliferation index less than 5%,” Dr. Tollefson said.
The hazard ratio for the risk of cancer-specific death for men with a Ki-67 proliferation index above 5% was 2.78 compared with individuals with a lower Ki-67 proliferation index score. Furthermore, men whose Ki-67 proliferation index was above 15% had an even higher risk of cancer-specific death, with a hazard ratio of 21.5, relative to those below 5%.
The predictive value of Ki-67 remained significant after controlling for clinical and pathological features in a multivariate analysis. “In fact, in this data set the Ki-67 index greater than 5% was a better predictor of outcome than was preoperative PSA, clinical stage, or Gleason score,” Dr. Tollefson said.
Incorporation of the Ki-67 proliferation index into a risk-prediction nomogram significantly improved the model's prognostic accuracy for biochemical failure, systemic progression, or cancer-specific death than the model did without Ki-67 inclusion. This was particularly true for systemic progression and cancer-related death.
Consistent with Retrospective Studies
The newly reported study is consistent with two previous retrospective studies, one of which tested Ki-67's prognostic value in prostate cancer patients treated with radiation and the other in those who chose active surveillance. Dr. Tollefson told OT that given the total body of data, Ki-67 should be incorporated into contemporary practice, especially because pathologists already have standardized protocols for its use in other malignancies.
Dr. Vogelzang said, “I don't get it in the routine prostate cancer pathology report, but I will probably start asking for it.”
He added that he expects that more researchers will examine the marker's value in prostate cancer. “I am sure we will get a validation or refutation—though I suspect a validation—of the value of Ki-67 fairly soon in a retrospective analysis,” he said, noting that other large prostate cancer collections are available for study.
Of course, Dr. Tollefson's study does not tell a clinician how a patient with a high Ki-67 proliferation index should be treated. “That is something we will have to sort out with further studies,” Dr. Tollefson said. “But I would argue that if you have somebody who otherwise would be a candidate for active surveillance—maybe has a small Gleason 6 tumor with relatively low PSA—but has a very high Ki-67 index, then a clinician may say this is an aggressive tumor based on its Ki-67 expression. Maybe this patient is not the best candidate for active surveillance and maybe we should at least look earlier for curative options.”
‘Progress in Multidisciplinary Management’
This year's Genitourinary Cancers Symposium had the subtitle of Progress in Multidisciplinary Management and was co-sponsored by the American Society of Clinical Oncology, the American Society for Radiation Oncology, and the Society of Urologic Oncology.