Accuracy and relevance of clinical stage in diagnosis and prognosis of localized prostate cancer
Posted Nov 22 2010 12:00am
According to a new report from the University of California, San Francisco (to be published in Cancer this week), clinical staging errors may at least in part underlie the fact that the researchers were unable to demonstrate a clear link between clinical stage and risk for prostate cancer recurrence in patients initially diagnosed with localized prostate cancer.
Reese et al. published a report , earlier in the year, showing that clinical stage had minimal impact on the prognosis of a contemporary cohort of patients with localized prostate cancer.
What Reese and his colleagues have now been been able to show, for the very first time, is that initial clinical stages (cT1,2) assigned to patients newly diagnosed with localized prostate cancer often do not follow the staging criteria laid down by the American Joint Commission on Cancer (AJCC). The ramifications of this are not completely clear as yet, as we shall discuss below. The fact that physicians do not strictly follow the AJCC staging criteria, however, is not entirely surprising, so let’s see if we can understand why.
Jimmy Chen is sent to his urologist by his primary care doctor. He is 53 years of age and his recent PSA test result was 3.2 ng/ml (up from 2.8 ng/ml two years earlier). The urologist gives him a second PSA test and carries out a very thorough digital rectal exam (DRE). The new PSA result is 3.3 ng/ml, but the urologist says there is no indication of any palpable tumor or any other problem on DRE.
Now Jimmy’s family has a significant history of prostate cancer, so Jimmy and the urologist decide that despite the negative DRE an ultrasound-guided biopsy is a “must.”
When the urologist is carrying out the biopsy he thinks that, under ultrasound guidance, he can maybe see a small focus of cancer in the lower half of the left lobe of Jimmy’s prostate. The biopsy results come back from the pathologist, and they are not at all what the urologist was expecting. They show significant Gleason 3 + 3 = 6 cancer in both halves of the right lobe and (indeed) a small focus of Gleason 3 + 4 = 7 in the lower half of the left lobe. This convinces the urologist that he did indeed see the cancer in the left lobe while he was looking at the ultrasound images, so he tells Jimmy that his clinical stage is T2c after all, Jimmy clearly does have cancer in both lobes of his prostate, but it does seem to be confined to the prostate.
The problem is that T2c isn’t Jimmy’s correct clinical stage. If you follow the AJCC guidelines rigorously, Jimmy’s clinical stage is T1c. Why? Because the AJCC criteria very specifically state that T1c is the correct clinical stage for localized prostate cancer found at biopsy when the patient has only a positive PSA result but no other clear clinical sign of the disease (i.e., no abnormality felt on DRE or clearly visible on any form of imaging test).
In the new study by Reese et al., the research team was able to show that clinical staging errors occurred in 1,370/3,875 men (35.4 percent) initially diagnosed with localized disease based on data in the CaPSURE database. Among these 1,370 men, 55.1 percent were downstaged (in other words the clinical stage assigned was inappropriately low) and 44.9 percent, like Jimmy Chen above, were upstaged (i.e., they were assigned an inappropriately high clinical stage). One can create imaginary scenarios like Jimmy Chen’s to explain all of these.
This all leads to three further questions:
Are the current AJCC clinical staging guidelines sufficiently clear?
Are the current AJCC clinical staging guidelines actually appropriate any more (for localized disease)?
Does any of this actually make any difference to a patient’s prognosis and treatment?
Reese and his colleagues discuss all this at length in their paper. The bottom line would seem to be as follows:
The distinction between clinical stages T2a, T2b, and T2c probably don’t have any great impact on the prognosis or the treatment of men with localized disease any more, because physicians are already taking other data into account (e.g., number, location, and amount of cancer in positive biopsy cores) in making treatment decisions.
Most of the prognostic nomograms widely used today also take account of the number of positive biopsy cores, in addition to PSA, clinical stage, and Gleason score. So errors in clinical staging for localized disease do not seem to significantly impact prediction of risk for biochemical recurrence.
The distinction between clinical stage T2a and T2b is already of dubious value.
The AJCC staging criteria may need a major upgrade to take account of current clinical practice.
What is very clear is that:
There is confusion (or possibly just differences in perceptions) about exactly how to assign clinical stage for newly diagnosed patients with localized disease (at least among a proportion of the urology community).
This may not be too important in terms of how men actually get treated (or indeed the prediction of risk for later biochemical recurrence).
The value of DRE- and TRUS-based information in assigning a clinical stage to patients with localized disease may be very doubtful indeed.
Perhaps it would be better to determine clinical stage today exclusively on the basis of biopsy data.
It will be interesting to see how the AJCC responds to this paper by Reese and his colleagues when the AJCC prostate cancer staging guidelines next come up for revision.