The current report offers a revised and updated commentary on the paper by Abd et al., first referred to yesterday on this blog. We have now had the opportunity to read the entire paper. While it does not change the general tenor of our prior comments, it is useful to make sure one has all relevant information on a topic this important.
Back in the 1990s, it became customary to carry out prostate cancer biopsies under transrectal ultrasound (TRUS) guidance using the so-called “sextant” approach, under which six preplanned biopsy cores were taken from each of six specific areas of the prostate. In addition, the urologist might also take one or two additional cores from specific areas of the prostate that looked as though they might be cancerous on the ultrasound.
By the early years of the 21st century, it had become more common to take eight preplanned biopsy cores, and today most urologists would probably tell you that 10 or 12 biopsy cores taken from preplanned areas of the prostate (usually all in the so-called peripheral zone) was the “right” number for an initial biopsy. However, as far as we are aware, there has never been any good evidence from randomized clinical trials to support this gradual “creep” in the number of biopsy cores. Indeed, in 2009, de la Rosette et al. did in fact publish data from a small, randomized clinical trial showing that there was little to no difference between the rates of positive biopsies of men who were given 8-core or 12-core biopsies between September 2004 and September 2007.
Abd et al. wanted to see if data available to them in the database at the U.S. Veterans Administration Medical Center in Atlanta, Georgia would give results that substantiated and expanded upon the data developed by prior research, including that be de la Rosette et al. Their database included the biopsy results from > 1,500 men of > 40 years of age collected between January 1991 and May 2007. From January 1991 though March 2004 the patients were all receiving 8-core biopsies, and from April 2004 though May 2007 they all received 12-core biopsies.
The protocols for the 8-core and the 12-core biopsies were targeted such that all cores would come from the the peripheral (outer) zone of the prostate. The data available allowed the researchers to compare rates of cancer detection based on the two protocols, the characteristics of the cancers, and the sensitivity of the two protocols in selected subgroups of men.
We should be clear however, that this was a retrospective analysis of data collected according to certain standardized protocols and not a “trial.” Patients were not randomized to an 8-core or a 12-core biopsy, nor was it possible to pre-emptively stratify patients to ensure that the patients in the two arms of the study had similar overall characteristics.
Here is what they were able to show for the study population as a whole:
- The total number of men biopsied was 1,546 (600 or 38.8 percent were given an 8-core biopsy and 946 or 61.2 percent were given a 12-core biopsy).
- Their average (median) age at time of biopsy was 62 years (range, 41 to 87 years).
- 461. percent (712/1,546) of the patients were African-American.
- Most patients (1,251/1,546 or 81.2 percent) had a PSA of < 10 ng/ml and 1,037/1,546 (67.1 percent) had a PSA between 2 and 10 ng/ml.
There were two statistically significant differences between the group of men getting 8-core as opposed to the group given 12-core biopsies:
- Men who got 8-core biopsies were more likely to have a PSA > 10 ng/ml than men getting the 12-core biopsies (24.0 vs. 15.8 percent).
- Men who got 12-core biopsies were more likely to have a PSA velocity > 0.75 ng/ml/yr than men getting the 8-core biopsies (59.4 vs. 42.3 percent).
And here are the major results of the analysis:
- The rate of positive biopsies using the 8-core protocol was 51.2 percent (307/600).
- The rate of positive biopsies using the 12-core protocol was 49.2 percent (465/946).
- There was no significant difference between the rate of positive biopsies using the two different protocols.
- Advanced age and high body mass index were significantly associated with higher likelihood of a positive biopsy result.
- Larger prostate volumes were associated with lower probability of a positive biopsy result.
There is an enormous amount of additional detail provided in this study. For example, African-American patients were more likely to have a positive biopsy (412/712 or 57.9 percent) than non-African Americans (360/834 or 43.2 percent).
The authors conclude that “there was no evidence that 12-core biopsy improved the likelihood of prostate cancer diagnosis compared with 8-core biopsy” and that “the results of this cohort from a US veteran population suggest that targeting the peripheral zone is more important than the absolute number of biopsy cores.” They go on to note, however, that in, for example, men with very large prostates (and some other specific clinical characteristics), more than eight biopsy cores cores may be advisable.
Should we expect every urologist in America to revert to 8-core biopsies tomorrow? Probably not. Should someone else repeat this study to see if they get a similar result? Probably. (It wouldn’t actually take that long if three or four large centers were to collaborate.)
The other thing that is interesting about this study is that the rate of positive biopsies was of the order of the 50 percent. Historically, positive biopsy rates of 20 to 40 percent in “screening” populations has been considered to be closer to the norm. Does this say something about the age and body mass index of veterans presenting for biopsy in Atlanta? Or does it tell us that the Atlanta VA Medical Center is better than average at deciding which patients really are appropriate for a biopsy? The authors themselves address this issue, suggesting that the high proportion of African Americans in their study population at 46.2 percent, combined with the previous exposure to Agent Orange among members of the US veterans community within the 40- to 80-year-old age range in 2001 to 2007 may have been factors that account for this high rate of positive biopsies.
In conclusion, the authors emphasize that their data applies only to patients who were getting an initial biopsy.