The Prostate Lab www.prostatelab.com
"Focused on the Prostate since 1996"
Minimal Residual Cancer
(From the American Journal of Surgical Pathology (Vol 21, No 12, 1997, December)
To the Editor:
The recent article entitled "increasing incidence of minimal
residual cancer in radical prostatectomy specimens'' by
DiGuiseppe et al. (1) should be of interest to all pathologists
who sign out prostate biopsy specimens. This article documents a
rising incidence of "minimal residual cancer'' in radical
prostatectomy specimens at the authors' institution. Evaluation
of 3,038 consecutive radical prostatectomy specimens obtained
between 1988 and 1995 showed 84 cases with minimal or no residual
cancer. (Less than 1 % of cases had minimal tumor during the
first 2 years of the study, whereas 3% to 4% had minimal tumor
during the last few years.)
The article attributes this increasing incidence to more
aggressive screening for early prostate cancer and "better''
criteria for diagnosing tiny cancers on needle biopsy specimens.
The authors conclude that very extensive examination of the
prostatectomy specimens (with complete submission of tissue,
repeat ectioning, and perhaps flipping of the blocks and
sectioning from the opposite side) should be performed in order
to identify tiny foci of "residual" cancer. What the
article does not adequately address is whether carcinoma is being
over diagnosed and whether too many prostatectomies are being
performed as a result.
It must strike an unbiased observer as somewhat ironic that so much effort should go into confirming the presencc of a tiny focus of' carcinoma in the resected prostate while relatively little effort is put into confirming that a clinically significant tumor is present before radical prostatectomy. (Even if a small amount of "cancer" is subsequently found, it begs the question as to whether such a tiny focus is the same as that originally examined via biopsy.)
Of course, it is not an easy task to predict which tumors are going to be clinically significant. A recent article from the same authors' institution reported that very limited tumor on needle biopsy (<3 mm on one core) was not predictive of clinically favorable tumor, because 33% had "moderate" tumor and 8% had "advanced" tumor in the radical prostatectomy specimens (3). However, if 60% with <3 mm "cancer" on needle biopsy have clinically insignificant or minimal tumor, it would seem that a small amount of "cancer" on needle biopsy indicates at least a reasonable possibility that clinically insignificant tumor may be present. Moreover, a needle biopsy with <1 mm "cancer" must have an even greater chance of representing clinically insignificant tumor.
Another recent article describing the "vanishing cancer phenomenon" indicates that, of 12 prostatectomy specimens with minimal or no cancer, the average length of cancer in the previous needle biopsy was 0. 15 cm (4). It would be informative if the authors of the present article would provide the sizes of cancer foci in the needle biopsies for those cases with minimal or no residual cancer at prostatectomy. One suspects that many of these are quite small, as one of the authors advocates diagnostic criteria for "limited adenocarcinoma of the prostate" and gives examples with as few as two glands (2).
Perhaps, rather than advocating laborious procedures for
preventing litigation when unnecessary prostatectomies are
performed, efforts may be better spent repeating biopsies to
confirm clinically significant tumor before performing radical
prostatectomy. There is a need to rush to surgery when a patient
is in imminent danger, but most patients with a questionable
prostate biopsy are not onl the verge of dying of their disease.
Michael Z. Gilcrease, MD, Ph.D.
Division of Anatomic Pathology
Department of Pathology
University of Texas Southwestern Medical Center
Dallas, Texas, USA
Authors' Reply:
Dr. Gilcrease brings up two distinct issues in his letter to the
editor. Our study evaluated the increasing incidence of minimal
residual cancers in radical prostatectomy specimens and how' the
pathologist should process and analyze such cases. Dr. Gilcrease
states that in retrospect the prostates in many of these cases
need not have come out as these were "insignificant
cancers." We agree with Dr. Gilcrease. However, when a
pathologist is presented with such a specimen, his or her first
responsibility is to arrive at an accurate diagnosis and find
cancer if present. When very limited cancer is present, the
pathologist should use expressions such as "minute foci of
carcinoma" to convey to the urologist that the patient is
cured. It also informs the urologist that this patient's prostate
cancer was relatively "insignificant" and sensitizes
the urologist to the issue that not all prostate cancers are
aggressive tumors requiring immediate surgery.
The second issue, which was not the focus of our manuscript, is
whether we can preoperatively predict which tumors are
"insignificant" so that patients may be offered the
option of expectant therapy. As Dr. Gilcrease noted, we have
devoted much work to this issue. In our earlier paper, we
proposed one of the first algorithms to predict
"insignificant carcinoma" based on both biopsy findings
and serum prostate-specific antigen (PSA) measurements (2). In a
more recent work, our initial findings have been validated (I).
However, it is important for pathologists to recognize that the
prediction of insignificant tumor cannot be based solely on the
biopsy findings. In contrast to the assertion of Dr. Gilcrease,
we have found that very limited cancer on needle biopsy does not
equate with very limited cancer within the radical prostatectomy
specimen. As discussed in our JAMA paper, when <3 mm of cancer
was present on one core (Gleason score <7), 33% of patients
had moderate tumor and 8% had advanced tumor within the radical
prostatectomy specimen (2). In a different data set analyzing 113
matched sextant biopsies and corresponding radical prostatectomy
specimens, with similar limited findings on needle biopsy, 32%
showed extraprostatic spread of tumor in the radical
prostatectomy specimen. In cases with 1 mm of cancer on one core
with no Gleason pattern 4 or 5, 45% of' patients showed
extraprostatic spread of tumor (4). Only when limited cancer on
needle biopsy is combined with relatively low serum PSA
measurements can there be more accurate prediction of'
"insignificant cancer".
Even with these preoperative findings, prediction is not perfect
and some patients who are predicted to have "insignificant
cancer" will have more advanced disease (1.2). Repeat
biopsies have been proposed by others as well as Dr. Gilcrease to
identify those patients who are more likely to have insignificant
cancers who might be followed expectantly. However we have found
that in men with limited cancer (<3 mm) on one core and serum
PSA of <10, repeat sextant biopsy does not provide accurate
information on tumor extent: despite no cancer on repeat biopsy,
38% still had moderate amounts of tumor within the radical
prostatectomy specimen (3). These included cases with tumor
volumes of >0.5 mi.Gleason score 7, or extraprostatic spread.
These findings demonstrate that one can only have limited
confidence in repeat biopsies to determine whether truly
insignificant cancer is present. Whether modifications to the
sextant biopsy procedure, such as by increasing the number of
biopsies or different placement of the biopsies, will improve
predictability remains to be seen.
In summary, we agree with Dr. Gilcrease that this is an
increasing problem with 20% to 25% of our radical prostatectomies
performed for nonpalpable prostate cancer diagnosed by needle
biopsy (stage T1c disease) containing "insignificant
cancers." Although we must deal with these radical
prostatectomy specimens in terms of finding the tumor, at the
same time we must continue to look for improved methods to
identify preoperatively men more likely to have minimal tumors
who may be given the option of expectant therapy.
Joseph A. DiGiuseppe, M.D., Ph.D.
Jurgita Sauvageot, B.S.
Jonathan I. Epstein, M.D.
Departments of Pathology and Urology
The Johns Hopkins Hospital
Baltimore, Maryland, USA