Download How Much Tissue Sampling Is Required When Unsuspected

How Much Tissue Sampling Is Required When
Unsuspected Minimal Prostate Carcinoma Is Identified
on Transurethral Resection?
Kiril Trpkov, MD, FRCPC; Jenny Thompson, BSc; Andrew Kulaga, MD, FRCPC; Asli Yilmaz, MD, FRCPC
● Context.—When minimal prostate cancer is detected in
the initial transurethral resection of the prostate (TURP)
sample, it is uncertain how extensively the remaining tissue
should be sampled for accurate grading and staging.
Objective.—To identify whether additional partial or
complete sampling is required to accurately evaluate TURP
samples with minimal cancer (stage T1a).
Design.—We prospectively examined all TURP samples
in our institution during 1 year. All specimens were sampled randomly in 6 cassettes. When minimal cancer was
found, we performed additional partial sampling (1 block
per 5 g of remaining tissue), followed by complete submission of all remaining tissue. All samples were evaluated
separately to identify possible changes in Gleason score
and tumor volume. We performed a cost analysis for the
additional tissue sampling.
Results.—Of 747 TURP samples evaluated on the initial
6 cassettes, 125 (16.7%) contained prostate cancer. Minimal cancer involving less than 5% of sampled tissue was
found in the initial submission in 26 (3.5%) patients. Additional partial examination required 3.5 blocks per case
(median; range, 1–23), while complete processing required
an additional 5.5 blocks per case (median; range, 2–25).
Initial Gleason scores and tumor volumes were not
changed in any of the studied cases after evaluating the
additional partial and complete samples. In our laboratory,
we calculated a cost of $4336 per year for the additional
sampling of TURPs with minimal cancer ($1681 for partial
and $2655 for complete sampling).
Conclusions.—When minimal cancer was found in the
first 6 cassettes of transurethral resections, additional partial and complete sampling did not change the initial Gleason scores and tumor volumes.
(Arch Pathol Lab Med. 2008;132:1313–1316)
T
However, controversy exists as to how much additional
sampling is required to ensure an accurate tumor volume
estimate and to ascertain the Gleason score in the resected
specimen when unsuspected cancer is identified in a
TURP sample.4–11
Patients with incidental tumors found on histology involving 5% or less of resected tissue (stage T1a) require
no further invasive treatment, particularly in patients older than age 60 years.12 The reported clinical disease progression rate of T1a tumors is low and varies from 8% to
27%.13–18 Older studies, before the prostate-specific antigen
era, show that only 4% of patients with T1a tumors will
have disease progression in 4 years, while in 16% to 25%
the tumors will progress in 8 to 10 years after the
TURP.13,14,19 In contrast, patients with T1b tumors, which
involve more than 5% of the resected tissue, are considered at much higher risk for disease progression (33% in
1 study)14 and usually require additional treatment.20
Therefore, accurate substaging of T1 disease is important
and mandates different patient care. Studies performed
prior to the prostate-specific antigen era showed that 90%
to 100% of the incidental tumors of both stages T1a and
T1b will be detected if 5 to 8 blocks or 12 g of tissue are
submitted and indicated that it is not necessary to put
through the whole tissue.5–7 However, some early studies
advocated complete tissue submission in all cases.4
In this study, we questioned whether there is a need for
additional sampling when incidental minimal carcinoma
ransurethral resection of the prostate (TURP) is still a
common urologic procedure that is primarily used in
current practice for the surgical management of benign
prostatic hyperplasia. The quantity of tissue chips received
in the pathology laboratory for examination varies. Recommendations by the College of American Pathologists
(CAP) require submission of specimens weighing 12 g or
less in their entirety, usually in 6 to 8 cassettes.1–2 For specimens greater than 12 g, the initial 12 g should be submitted, and 1 cassette for every additional 5 g may be
submitted. The CAP Cancer Committee recently added a
recommendation that ‘‘if an unsuspected carcinoma is
found in the tissue submitted and it involves 5% or less
of the tissue examined, the remaining tissue is generally
submitted for microscopic examination.’’ 1 Most recently,
this approach was seconded by a pathology expert group
for the Association of Anatomic and Surgical Pathology.3
Accepted for publication February 8, 2008.
From the Department of Pathology and Laboratory Medicine, Anatomical Pathology, Rockyview General Hospital, Calgary Laboratory
Services and University of Calgary, Calgary, Alberta.
The authors have no relevant financial interest in the products or
companies described in this article.
Reprints: Kiril Trpkov, MD, FRCPC, Department of Pathology and
Laboratory Medicine, Anatomical Pathology, Rockyview General Hospital, Calgary Laboratory Services and University of Calgary, 7007 14th
St SW, Calgary, Alberta, Canada T2V 1P9 (e-mail: [email protected].
ca).
Arch Pathol Lab Med—Vol 132, August 2008
Sampling of Minimal Prostate Cancer on TURP—Trpkov et al
1313
involving 5% or less of the tissue is detected in the initial
specimen in contemporary practice. This prospective
study sought to investigate whether additional partial or
complete submission of the remaining tissue chips will
change the initial Gleason score and tumor volume, that
is, change the stage from T1a to T1b. Lastly, since the cost
of the additional sampling is also an issue, we performed
a cost analysis for the submission of additional partial and
complete TURP samples in our laboratory.
MATERIALS AND METHODS
Three pathologists prospectively examined all TURP specimens in our centralized urologic pathology practice during 1 year
(June 1, 2006 to May 31, 2007). All specimens were routinely sampled by pathologists’ assistants according to our departmental
protocol. Of 747 TURP specimens evaluated during the study
period by the initial 6-cassette submission, 125 (16.7%) exhibited
prostate cancer. Tumor volume was assessed as a percentage of
tissue involved by marking the areas of involvement and then
combining these foci to estimate the percentage of the submitted
tissue involved by cancer. More than 5% tumor involvement in
the initial sample was found in 99 (79.2%) of the 125 specimens
containing cancer. Some of these patients with nonminimal cancer had known or suspected cancer and TURP was performed
due to retention or obstruction symptoms. These cases represented stage T1b tumors, and additional sampling was not performed, according to the current CAP recommendations and previous studies.1,2,5–7,10 In 26 TURP specimens the initial random
sample of 6 cassettes contained minimal carcinoma involving 5%
or less of the sampled tissue. Patients with unsuspected cancer
typically presented clinically with symptoms of retention, bladder stones, recurrent urinary tract infections, or gross hematuria
or had undergone unsuccessful medical management of benign
prostatic hyperplasia. Prostate-specific antigen testing was not
routinely performed in these older patients because they were not
suspected to have cancer prior to the procedure. In this scenario,
the previous routine practice among the pathologists in our institution varied from complete sampling to variable partial sampling, estimated either on the whole tissue or on the remaining
tissue. In all 26 cases, an additional partial and complete sampling was performed in a systematic fashion. Partial sampling
entailed an additional random sample of 1 block for every 5 g of
remaining tissue, which was followed by additional complete
sampling that included submission of all residual tissue. To identify possible changes in the Gleason score and the tumor volume,
these parameters were separately evaluated in the initial and in
the additional partial and complete samples. Pathologists involved in the study recorded the information for each case prospectively on score sheets, which were collected and tabulated.
Cost analysis, based on the technical and the pathologist fees,
was performed to calculate the cost of the additional TURP sampling in our laboratory. Pathologists’ time for slide review of
TURP specimens was estimated at 1 minute per slide. The cost
for 1 minute of pathologist time was $3.12, based on the hourly
rate in our institution. The technical component of the cost was
estimated at $15 per slide and included pathologist assistants’
time for grossing and histotechnologists’ time for embedding,
cutting, and staining, as well as the cost of the reagents.
RESULTS
The study group of 26 TURP specimens represented
20.8% of cases containing cancer and 3.5% of all TURP
cases processed in our hospital during 1 year. Patient demographics, documented clinical history on the requisition, and TURP sampling data are presented in Table 1.
All patients with minimal cancer included in this study
were older (median age, 68 years); only 2 patients were
younger than 60 years. All patients with minimal cancer
were clinically unsuspected to have cancer prior to the
1314 Arch Pathol Lab Med—Vol 132, August 2008
Table 1. Demographics and Sampling Variables for
Patients With Minimal Cancer on Transurethral
Resection of the Prostate
Variable
No. of patients
26
Age, y
Median/mean (range)
68/70 (56–85)
Clinical history of patients
No history
Benign prostatic hyperplasia
Microinvasive bladder cancer
Bladder stones
Tissue weight, g
Median/mean (range)
13
11
1
1
24.8/33.5 (10.2–125)
Blocks/case, No.
Median/mean/mode (range)
Weight/block, g
Median/mean (range)
16/18.7/10 (10–45)
1.7/1.7 (1.0–3.3)
Table 2. Volume of Cancer and Gleason Scores in
the Initial Transurethral Resection of the Prostate
(TURP) Samples and in the Additional Partial and
Complete TURP Samples
Initial
Samples
Additional
Partial
Samples
Additional
Complete
Samples
No. of blocks
Median/mean
(range)
6/6
(6)
3.5/4.9
(1–23)
5.5/8
(2–25)
Volume of carcinoma, %
0
1–5
⬎5
0
26
0
13
13
0
9
17
0
Gleason score
5
6
7
3
22
1
2
10
1
1
15
1
Variable
procedure. The median number of cassettes for complete
sampling was 16, and all cases required 10 or more cassettes for complete tissue processing. In 16 cases (62%), 15
or more blocks were sampled. After submitting the initial
6 blocks, partial examination of the specimen required 3.5
blocks (median; range, 1–23) (Table 2). An additional 5.5
blocks (median; range, 2–25) were necessary for complete
processing of the remaining tissue. Seven (27%) of 26 cases
showed no additional tumor in the remaining tissue. Additional tumors were identified in 19 (73%) of 26 cases: in
both additional partial and complete samples in 11 cases,
only in the additional partial samples in 2 cases, and only
in the additional complete samples in 6 cases. When cancer foci were found in the additional partial and complete
samples, they were always found in less than 5% of the
respective samples. Thus, after the additional partial and
complete tissue evaluation, the Gleason scores and the percentage of initially estimated tumor volumes remained
unchanged in all cases, that is, none of the cases were
changed from stage T1a to T1b. Gleason scores of 5, 6,
and 7 were reported in the initial specimens in 3 (12%),
22 (84%), and 1 (4%) patients, respectively. In 1 patient,
Gleason score 7 (4 ⫹ 3) was found in less than 5% of the
Sampling of Minimal Prostate Cancer on TURP—Trpkov et al
Table 3. Cost Analysis for Tissue Samples in
Transurethral Resection of the Prostate Specimens
With Minimal Cancer
Sample
Initial sample
Additional partial sample
Initial ⫹ partial
Additional complete sample
Initial ⫹ partial ⫹ complete
Partial ⫹ complete
Median
Cost/Sample, $
Total Cost, $
72.16
52.46
124.62
78.72
203.34
131.18
2046.72
1681.25
3727.97
2655.21
6383.18
4336.46
tissue in all 3 samples (initial, additional partial, and additional complete sample). This tumor was found incidentally in an 81-year-old patient with benign prostatic hyperplasia and a history of colon cancer. Although some
may not regard this tumor as stage T1a, the current TNM
system does not stipulate that Gleason score should alter
the clinical stage in a TURP specimen. Median costs per
case for the additional partial and complete samples were
$52.46 and $78.72, respectively (Table 3). In our laboratory,
we calculated a total cost of $4336.46 per year for the additional sampling of TURP specimens with minimal cancer ($1681.25 for the additional partial and $2655.21 for
the additional complete sampling). This cost analysis reflects our practice scenario; in institutions where grossing
of TURP specimens is done by a pathologist, this additional cost should also be considered.
COMMENT
Current clinical classification for the primary prostate
tumor (T), according to the TNM staging system of the
American Joint Committee on Cancer, defines tumors of
stage T1 as clinically inapparent tumors, neither palpable
nor visible by imaging.21 Stage T1 is subdivided into T1a,
which includes incidental tumors found on histology in
5% or less of resected tissue, and T1b, which includes
tumors identified in more than 5% of the resected tissue.
The T1a and T1b substages of the TNM staging system
correspond to stages A1 (ⱕ5% TURP chips positive) and
A2 (⬎5% TURP chips positive) of the earlier WhitmoreJewett system. This distinction is relevant because patients with T1a tumors do not require additional treatment and have a significantly better prognosis than patients with T1b tumors, who usually require additional
treatment.13–20
The most recent CAP recommendations indicate that if
unsuspected minimal (T1a) carcinoma is detected in the
examined tissue, all remaining tissue should be submitted
for microscopic examination.1 In this study, subsequent
partial and complete sampling had no additional yield beyond the initial 6 blocks in establishing correctly the final
T1 substage and the Gleason score. In all studied cases,
regardless of the patient age and weight of the specimen,
Gleason score and overall tumor percentage remained unchanged. Although tumor was found in the additional
partial and complete samples in 73% (19/26) of cases, additionally detected tumor volumes did not alter the stage
from T1a to T1b in any of the cases. To ensure safe practice
and for quality assurance, if minimal (T1a) cancer is found
in the initial specimen, a reasonable sampling approach of
the residual tissue may include 1 cassette per 5 g of the
remaining chips. According to our study, this approach
Arch Pathol Lab Med—Vol 132, August 2008
should be adequate, regardless of the patient age, weight
of the submitted tissue, or the Gleason score found in the
initial random specimen.
One early study found that complete submission of all
TURP specimens made a clinically significant difference
for accurate staging.4 However, this study used an older
classification system and examined 2 different patient
populations during different periods. In contrast, several
subsequent studies found that the odds of correctly staging a tumor after random partial examination are significantly high.5–7,10 Vollmer6 determined that after submission of 10 blocks, 97% of tumors would be discovered.
Rohr7 found that if 8 blocks were submitted, 90% and
100% of A1 and A2 carcinomas, respectively, would be
detected, and concluded that complete sampling of TURP
specimens would not alter the original diagnosis and was
unnecessary. He demonstrated by statistical analysis that
99.93% and 100% of A2 carcinomas would be detected by
submission of 5 and 8 blocks, respectively. Murphy et al5
found that sampling 12 g of the randomly selected chips
detected almost 90% of all incidental carcinomas and determined that all clinically significant prostatic carcinomas
(T1b tumors) would be detected if only 6 g of chips were
examined microscopically. Similarly, our study did not
identify a significant tumor volume variation in the additional partial and complete samples, which indicates
that tumor foci are distributed uniformly in the sampled
chips. McDowell et al10 recommended complete submission of remaining tissue only in cases where T1a tumor,
but not T1b tumor, is detected in the initial 8 blocks. This
approach was justified based on 1 case upstage (of 34
studied), the discordance between the initial and the subsequent Gleason scores, and the low incidence of unsuspected T1a tumors requiring more than 9 blocks for complete sampling. Humphrey and Walther9 and Humphrey11
suggested that partial sampling may be sufficient for patients older than 60 years, while in younger patients, in
whom tumors of stage T1a may progress, they recommended complete specimen sampling.
In our study, we initially submitted 6 cassettes (median,
1.7 g per block) for all TURP specimens, which would
approximate 10 to 12 g of tissue. This approach provided
a high probability that most T1a and all T1b cancers
would be diagnosed in the initial examination. Our aim
was to investigate the yield of additional tissue sampling
only when T1a disease is encountered. We have restricted
our study to stage T1a, and not T1b, tumors because previous studies have shown that downstage would not occur
in patients with T1b tumors if additional tissue is sampled.5–7,10 Despite the small chance of tumor upstaging or
changing the Gleason score, these were not documented
in any of our patients. Therefore, additional examination
of the remaining tissue in this study was not contributory
when T1a tumor was detected in the initial 6 cassettes. We
have not included the initial Gleason scores in stratifying
stage T1 disease. However, of 26 cases only 1 showed
Gleason score 7, which underscores the rarity of high
Gleason grade cancers in TURP specimens containing
minimal cancers.
To our knowledge, no contemporary study has considered the cost implications of conducting partial and complete sampling, although they have been addressed in
some previous studies.5,6,8,22 Although some authors claim
that the cost of complete sampling should not be a concern
when minimal carcinoma is incidentally detected, our
Sampling of Minimal Prostate Cancer on TURP—Trpkov et al
1315
study found that complete sampling is unnecessary when
unsuspected minimal cancer is found on TURP. We calculated that a median of $78.72 per case and $2655.21 per
year would be saved in our laboratory if only partial sampling was implemented in this instance. Although the cost
analysis for our laboratory may not be identical with other
centers, the time tradeoff and the general cost for complete
sampling are not justified. A limitation of this study is the
small sample size, as only 26 specimens were identified
during 1 year containing unsuspected carcinomas of stage
T1a. However, our study was prospective and was conducted in a centralized urologic pathology setting. The
low incidence of 3.5% of cases with minimal cancer (stage
T1a) in TURP specimens in this study is comparable to
the estimated incidence of 4.7% in TURPs performed for
presumed benign prostatic hyperplasia.10 The small sample size of this study requires confirmation by studies performed in other institutions before changing the current
CAP practice guidelines.
In summary, this study demonstrates that when minimal carcinoma was incidentally detected in the initial 6
cassettes of submitted TURP tissue, significant carcinomas
that had not been adequately sampled initially were not
identified upon further investigation. The initial random
sample was sufficient to accurately identify the Gleason
score and the tumor volume in the entire resected tissue.
For quality assurance, partial sampling of 1 block for 5 g
of the residual tissue should be considered a reasonable
option. If only partial additional sampling was performed
in TURP specimens upon finding less than 5% of tissue
involved by cancer in the initial sample, cost savings in
our institution would have been $78.72 per case and a total
of $2655.21 per year.
References
1. CAP Cancer Committee. Reporting on Cancer Specimens: Case Summaries
and Background Documentation. Northfield, Ill: College of American Pathologists; 2005:genitourinary, prostate 15–16.
2. CAP Cancer Committee. Reporting on Cancer Specimens: Protocols and
1316 Arch Pathol Lab Med—Vol 132, August 2008
Case Summaries. Northfield, Ill: College of American Pathologists; 2000:genitourinary, prostate 9.
3. Epstein JI, Srigley J, Grignon D, Humphrey P. Recommendations for the
reporting of prostate carcinoma. Hum Pathol. 2007;38:1305–1309.
4. Newman AJ Jr, Graham MA, Carlton CE Jr, Lieman S. Incidental carcinoma
of the prostate at the time of transurethral resection: importance of evaluating
every chip. J Urol. 1982;128:948–950.
5. Murphy WM, Dean PJ, Brasfield JA, Tatum L. Incidental carcinoma of the
prostate: how much sampling is adequate? Am J Surg Pathol. 1986;10:170–174.
6. Vollmer RT. Prostate cancer and chip specimens: complete versus partial
sampling. Hum Pathol. 1986;17:285–290.
7. Rohr LR. Incidental adenocarcinoma in transurethral resection of the prostate: partial versus complete microscopic examination. Am J Surg Pathol. 1987;
11:53–58.
8. Vollmer RT. Tissue sampling to detect incidental carcinoma of prostate. In:
Damjanov I, Cohen AH, Mills SE, Young RH. Progress in Reproductive and Urinary Tract Pathology. Vol 1. New York, NY: Field & Wood Medical Publishers;
1990:91–100.
9. Humphrey PA, Walther PJ. Adenocarcinoma of the prostate, part I: tissue
sampling considerations. Am J Clin Pathol. 1993;99:746–759.
10. McDowell PR, Fox WM, Epstein JI. Is submission of remaining tissue necessary when incidental carcinoma of the prostate is found on transurethral resection? Hum Pathol. 1994;25:493–497.
11. Humphrey PA. Prostate tissue handling and sampling: transurethral prostatectomy (TURP). In: Humphrey PA, ed. Prostate Pathology. Chicago, Ill: American Society of Clinical Pathology; 2005:40–44.
12. Lowe BA. Management of stage T1a prostate cancer. Semin Urol Oncol.
1996;14:178–182.
13. Blute ML, Zincke H, Farrow GM. Long-term follow-up of young patients
with stage A adenocarcinoma of the prostate. J Urol. 1986;136:840–843.
14. Epstein JI, Paull G, Eggleston JC, et al. Prognosis of untreated stage A1
prostatic carcinoma: a study of 94 cases with extended follow-up. J Urol. 1986;
136:837–839.
15. Lowe BA, Listrom MB. Incidental carcinomas of the prostate: an analysis
of the predictors of progression. J Urol. 1988;140:1341–1347.
16. Thompson IM, Zeidman EJ. Extended follow-up of stage A1 carcinomas of
the prostate. Urology. 1989;33:455–458.
17. Roy CR, Horne D, Raife M, et al. Incidental carcinoma of the prostate:
long-term follow-up. Urology. 1990;36:210–213.
18. Cheng L, Neumann RM, Blute ML, et al. Long-term follow-up of stage T1a
prostate cancer [letter]. J Natl Cancer Inst. 1998;90:1105–1107.
19. Cantrell BB, DeKlerk DP, Eggleston JC, et al. Pathological factors that can
influence prognosis in stage A prostatic cancer: the influence of extent versus
grade. J Urol. 1981;125:516–520.
20. van Andel G, Vleeming R, Kurth K, et al. Incidental carcinoma of the
prostate. Semin Surg Oncol. 1995;11:36–45.
21. Sobin LH, Wittekind C, for the International Union Against Cancer [UICC],
eds. TNM Classification of Malignant Tumours. 6th ed. New York, NY: Wiley-Liss;
2002:184–187.
22. Eble JN, Tejada E. Cost implications of sampling strategies for prostatic
transurethral resection specimens: analysis of 549 cases [abstract]. Am J Clin
Pathol. 1986;85:382–383.
Sampling of Minimal Prostate Cancer on TURP—Trpkov et al