THE "NEW" PROSTATE CANCER INFOLINK

The population-based rate of Medicare patients getting treatment for prostate cancer dropped by 42 percent from 2007 to 2012 … but this drop did not occur in patients who had a high risk of dying from a cause other than prostate cancer within 10 years, who are arguably among those who least need immediate treatment for prostate cancer (and especially for low-risk forms of prostate cancer). These are key findings from a study by researchers at Michigan Medicine (formerly the University of Michigan Health System).

Based on the information in their report in Health Affairs, Borza et al. identified 67,000+ Medicare recipients (all being treated under fee-for-service plans as opposed to Medicare managed care plans) who were newly diagnosed with prostate cancer here in the US though the use of the Medicare claims database for the years 2007 through 2012. And we need to interpret these data with some caution because the vast majority of such men were therefore, by definition, > 65 years of age.

We also need to interpret these data with some caution because they are data from a time prior to the recent, more intense efforts to ensure thoughtful adoption of carefully structured forms of active surveillance in appropriate patients diagnosed with low-risk forms of prostate cancer. This is an effort — consistent with efforts to decrease over-diagnosis and over-treatment of prostate cancer — in which the Michigan Urological Surgery Improvement Collaborative (MUSIC) has been heavily involved in the state of Michigan.

Now the first thing we need to understand is exactly what Borza et al. mean by  a “population-based rate of Medicare patients getting treatment for prostate cancer.” They define this as the treatment rate per 1,000 men who are eligible for screening. So they are not saying that 42 percent fewer Medicare patients got treated from prostate cancer in 2007 as opposed to 2012. They are saying that the treatment rate per 1,000 Medicare patients who were eligible for screening dropped from 4.3 to 2.5 per 1,000 over the 6-year time period (which is a 42 percent decrease in that treatment rate).

Some other key facts are the following:

• Of the 67,023 patients diagnosed with incident prostate cancer whose data met the criteria for inclusion in this analysis,
  • 48,474 men (72.3 percent) received treatment within 12 months of diagnosis.
  • 10,643 men (15.9 percent) were ≥ 80 years of age.
  • 33,479 men (50.0 percent) had a > 58 percent probability of dying from something other than prostate cancer within the next 10 years .

Borza et al. — as one might expect — suggest that this apparently major drop in the population-based rates of men being treated from prostate cancer over the 6-year period from 2007 to 2012 can be laid squarely at the feet of the US Preventive Services Task Force (USPSTF) as a consequence of changes in the USPSTF guidelines in 2008 and 2011/2012. And it is quite certainly the case that the USPSTF guidelines on PSA screening have impacted the numbers of men getting PSA tests as a routine part of a “check up” conducted by primary care physicians for any man over 50. Whether that is or isn’t a good thing may take a while to actually know the answer to. If the numbers of men who either get metastatic prostate cancer or die of metastatic prostate cancer start to rise significantly at some point in the next 5 years or so, then we will know there is a problem, but it is almost certainly too early to make a judgement about that as yet.

What we are more concerned about in the paper by Borza et al. is their other two findings that, over the same time period, in the same group of Medicare patients:

• There had been only an 8 percent decrease in the population-based rate of treatment of men actually diagnosed with prostate cancer (from 718 to 659 per 1,000 diagnosed men)
• There had been no apparent change at all in the population-based rate of treatment among men diagnosed with prostate cancer who had the highest probability of dying from something other than their prostate cancer within the next 10 years (which seems to have been stable at about 500 per 1,000 diagnosed men).

Every major organizational guideline now suggests that men newly diagnosed with prostate cancer who have a life expectancy of 10 years or less should be monitored (at least on active surveillance but arguably on watchful waiting) as opposed to being given immediate treatment. And yet, according to these data, about half of these men were still getting immediate treatment in  2012.

Borza et al. suggest that two primary reasons for this are because “market forces continue to favor treatment” over active surveillance and watchful waiting and that improvements in technology have been widely perceived as offering much better forms of treatment than were available in the past. There is undoubtedly a lot of truth to this suggestion, but there may also be another important reason as well.

Many of the men > 65 years of age with a life expectancy of < 10 years are going to be > 75 years of age. (It was actually 25,918/67,023 or 38.7 percent of the men in this study.) These men have lived all their lives as recipients of an almost constant societal message that “cancer” is an inevitably deadly disease (which we are only now becoming aware isn’t necessarily the case). Tell most of these men that they have cancer, and their first reaction is almost inevitably going to be that “something” needs to be done about it promptly. When this unsurprising fact is combined with the “market forces” and technology improvements argument put forward by Borza et al., we may have a better appreciation for why half of these men with limited life expectancies are getting treatments (often radiotherapy) that most of them probably don’t need.

In addition to the “market forces” and technology improvements factors, Borza et al. lay out other well appreciated facts that don’t help the situation:

• We have no really good tools for the accurate prediction of overall and prostate cancer-specific life expectancy in prostate cancer patients (although the Memorial Sloan-Kettering Cancer Center’s Male Life Expectancy tool is a step in the right direction).
• There is no national consensus about the appropriate management of low-risk forms of prostate cancer among the treating community (which is particularly evident in the continued regional variation in patterns of care across the country).

We would add to that the fact that currently available tests still offer very little really sound, clinical information about which patients need treatment and which can be just monitored. We have better tests than we had in the past, but we still have no combination of tests that (a) excludes a combination of systematic and MRI/TRUS-guided fusion biopsy and that still (b) allows a specialist to tell his or her patient that, “There is a 90 percent probability that you will live for 15 years with a prostate cancer like this before any form of treatment would be needed.”

We are in the middle of a major paradigm shift in the management of prostate cancer. There is an increasing onus on specialists to really think hard about whether they are doing the very best that they can to appropriately manage and care for each patient they see. But at the same time there is an increasing onus on patients to really be able to understand the wrinkles associated with all the management options that are available to them. The decision that I, your sitemaster, might make in a specific situation today could well be very different to the one that might be made by any reader of the same age, and state of health, who was faced with an exactly similar diagnosis. There are no “perfect answers” regarding “the right” thing to do.

Quoted in a commentary on the ScienceDaily web site, Dr. Borza stated:

We know prostate cancer is a deadly disease in some men. We need better tools to identify which men should be screened and, among those diagnosed, which men should be treated aggressively. This is still a black box. It’s that uncertainty that leads to different approaches to treatment based on how different physicians view the risk. If we get better predicting who’s at highest risk, we can more accurately tailor screening and treatment.

And that’s very definitely the bottom line.

Editorial note: The “New” Prostate Cancer InfoLink thanks Dr. Tudor Borza of Michigan Medicine for promptly providing us with a full text PDF of his and his colleagues’ article  in Health Affairs.

Filed under: Diagnosis, Living with Prostate Cancer, Management, Risk, Treatment | Tagged: change, Diagnosis, Management, Medicare, risk, Treatment |