Prostate cancer grade migration: an age-based analysis

About a month ago, Jani et al. published a structured analysis of SEER data from over 450,000 patients collected between 1974 and 2003 focused on the relationship between prostate cancer grade at the time of diagnosis and patient age.

Although it is clearly not possible that all these patients were treated surgically, the authors misleadingly refer to this as an analysis of the “pathologic” grade of the cancers. That cannot be the case because a pathologic grade is only available for surgically treated patients. Therefore we assume that the authors are actually referring to the biopsy-based, “clinical” Gleason grade.

The SEER data were used to categorize the patients by age at diagnosis (40-49, 50-59, 60-69, 70-79, and 80+ years), by grade (well differentiated, moderately differentiated, and poorly differentiated disease), and by 5-year interval (1974-78, 1979-83, 1984-88, …, 1999-2003).

According to the authors there were four key findings from this study:

  • There was a small but statistically non-significant change in distribution of age at diagnosis (from older to younger).
  • There was a statistically significant (P < 0.001) grade migration over the study period, principally from well- to moderately differentiated disease
  • This grade migration occurred across all age groups.
  • Trends in 5-year cause-specific survival time were similar for all age groups, but well-differentiated disease appeared to converge with moderately differentiated disease in later years.

Based on these results, the authors claim the following conclusions: “An overall grade migration has occurred in prostate cancer, primarily observed as a shift from [well-differentiated] to [moderately differentiated] disease; this may weaken grade-based prognostic categorizations. This migration occurred independent of patient age, reinforcing that the grade migration is likely due to changes in pathologic interpretation rather than to screening-related changes in disease characteristics.” [Italic type added for emphasis.]

In a subsequent “Beyond the Abstract” comment just published on the UroToday web site, the authors add the following: “Our study has limitations, most of which relate to limitations in the SEER data itself (unavailability of PSA, Gleason score, and hormone therapy information); however, within stated limitations, our study further supports the concept that grade migration has occurred due to changes in interpretation criteria as opposed to screening-related changes in the natural history of prostate cancer. ”

Given this limitation, The “New” Prostate Cancer InfoLink is utterly unable to know what to make of the secondary conclusion of this study (see italic type above). There has been a vast change in our diagnostic capabilities regarding prostate cancer since 1974. And there have been highly time-dependent issues regarding the number of men who were being tested. In 1974 the majority of men diagnosed with prostate cancer would have been diagnosed because they had symptomatic conditions that lead to a diagnosis (back pain, urinary tract probelms, you name it). Today, the majority of newly diagnosed men are diagnosed because of a screening program of some sort (i.e., because of an elevated PSA level and/or an abnormal physical exam). Almost no one was “screened” for prostate cancer until at the earliest about 1988-90. Furthermore, because of the vast increase in number of prostate biopsies since the early 1990s, there has been at least some improvement in the ability of pathology laboratories to apply the Gleason grading system with consistency (even though there is still variation in how any two pathologists may “read” a particular set of biopsy slides).

Last but not least, most men aren’t having a DRE and a PSA every year today. It is more likely that most men may now be having such a test on an irregular basis, especially if they have a couple of negative results in their mid to early 50s or 60s.

Given the limitations of the study noted by the authors and the additional issues itemized above, The “New” Prostate Cancer InfoLink seriously questions whether the authors’ secondary conclusion to this paper is justified. We do not dispute the findings of the study. However, our conclusion based on these data would be less categorical, perhaps more like the following:

“Changes in patterns of prostate cancer screening and diagnosis have led to less significant distinctions between the pathologic use of the terms ‘moderately differentiated’ and ‘well-differentiated’ disease across all age groups. The precise reasons for this apparent ‘blending’ of well- and moderately differentiated disease are undetermined, but they may have implications leading to reassessment of the current definitions of ‘well differentiated,’ ‘moderately differentiated,’ and ‘poorly differentiated’ disease.” 

It is clearly in the interests of patients and their physicians to be able to apply the available data with accuracy in order to categorize patients into low, moderate and high risk groups when it comes to making decisions about treatment. If we really can’t differentiate well any more between low and moderate grades of disease based on pathologic assessment of biopsy specimens, then let’s stop trying to do this!

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: