The Importance of the Depth of Invasion in Stage T1 Bladder Carcinoma: A Prospective Cohort Study

The Importance of the Depth of Invasion in Stage T1 Bladder Carcinoma: A Prospective Cohort Study

oOn6347/97/1573-0800803.00/0 Vol. 157, 800-804,March 1997 Printed in U.S.A. IkE JOURNALOF UROLOCY Copyright 0 1997 by AMERICAN UROLOCICAL AS.WCUTIO...

820KB Sizes 0 Downloads 56 Views

oOn6347/97/1573-0800803.00/0

Vol. 157, 800-804,March 1997 Printed in U.S.A.

IkE JOURNALOF UROLOCY Copyright 0 1997 by

AMERICAN UROLOCICAL AS.WCUTION, h C .

THE IMPORTANCE OF THE DEPTH OF INVASION IN STAGE T1 BLADDER CARCINOMA: A PROSPECTIVE COHORT STUDY STEN HOLMANG, HANS HEDELIN, CLAES ANDERSTROM, ERIK HOLMBERG AND SONNY L. JOHANSSON From the Oncology Center and Department of Urology, Sahlgrenska University Hospital, Giiteborg and Department of Urology, Karnsjukhuset, Skovde, Sweden, and Department of Pathology and Microbiology, and Eppley Institute for Research in &ncer and Allied Diseases, University of Nebraska Medical Center, Omaha, Nebraska

ABSTRACT

Purpose: We studied the depth of invasion in the lamina propria in all patients with primary stage T1 bladder cancer in a geographical region and related the findings to the long-term prognosis. Materials and Methods: All 121 primary stage T1 tumors diagnosed in western Sweden between 1987 and 1988 were analyzed with respect to the depth of invasion in relation to the lamina muscularis mucosae. All clinical records were reviewed in 1994 and 1995. Results: More than 90% of the histopathological specimens could be separated into superficially (pTla) or deeply (pTlb) invasive stage T1 tumors. Grade 3 tumors were significantly more common among patients with stage pTlb disease (79 versus 40%, p <0.001). Patients with stage pTlb grade 3 cancer had a higher progression rate (58 versus 36%, p >0.05) and an almost doubled risk of dying of bladder carcinoma compared to those with stage pTla grade 3 disease (45 versus 23%, p >0.05). Carcinoma in situ at the primary operation was associated with an impaired prognosis in patients with grade 3 tumors regardless of the depth of invasion in the lamina propria. Conclusions: The prognosis is poor in patients with deep lamina propria invasion (stage pTlb) treated with transurethral resection alone. Patients treated with radical cystectomy had excellent survival regardless of the depth of invasion in the lamina propria. Radiotherapy was associated with poor survival. KEYWORDS:bladder neoplasms, neoplasm metastasis

It has been known for a decade that the lamina propria of noma in situ was diagnosed. Patients unfit for cystectomy the bladder contains a thin smooth muscle layer.l.2 This were treated with radiotherapy. tunica muscularis mucosae varies in its development within All primary bladder tumors in western Sweden were proa given bladder and may be formed as a continuous (which is spectively registered between February 1987 and January rare) or interrupted layer, or it may consist of only thin wisps 1989 as a consequence of the bladder cancer treatment proof smooth muscle fibers, which is the most common appear- gram. The surgeon completed a form with the patient name, It can be recognized approximately a third to halfway date of birth, gender, clinical stage and microscopy findings. between the epithelium and the detrusor muscle at the same A total of 759 tumors was registered. The clinical records level where rather large arteries can be ~ e e n . l - 2It. ~is possi- were reviewed between 1990 and 1992 and again (a minible to separate tumors into those that show invasion above mum of 5 years after first diagnosis) between 1994 and 1995 this landmark (stage pTla) and those that show invasion to by 1 of us (S. H.). Several patients were erroneously regisor below the level of the muscularis mucosae (stage pTlb, see tered, such as those with previous bladder or upper tract figure).*5 It has been suggested that patients with stage tumor or no histopathological documentation, and 735 repTlb tumors have as poor a prognosis as those with detrusor mained for histopathological review. The slides of all primary muscle i n ~ a s i o nWe . ~ compared the long-term outcome for a tumors were reviewed by 1 of us (S. L. J.).The tumors were cohort of patients whose primary tumor was stage pTla or graded according to the World Health Organization system. pTlb. Bladder tumor could not be confirmed histopathologically in 22 patients and 713 tumors remained, of which 701 were PATIENTS AND METHODS bladder carcinomas and 12 were inverted papillomas. Lamina Propria invasion (stage pT1) was detected in 130 of In 1986 a comprehensive bladder cancer treatment prothe 701 carcinomas. Of the 130 patients 8 had a more adgram was to and accepted by surgeons and urologists in Sweden.6 me recommended treatment of vanced clinical stage at diagnosis (T3, T4 or metastatic disstage T1 carcinoma was transurethral resection. Intravesical ease). One patient had a pure adenocarcinoma with deep therapy was infrequently used in Sweden but rec- lamina Propria invasion and was excluded from this study. ommended for recurrent tumors. Bacillus Ca]met&-Guerin ‘‘I’he Primary tumors were subclassified by 1 of us (S. L. J., therapy was used only occasionally &fore 1990. Radical cys- who was unaware of the final outcome of the patients) into 2 groups: 1) tUmOrS showing invasion above the level of the *tomy was the treatmentof choice in cases with an@olpphatic invasion or large tumors and when extensive carci- muscularis mucOsa (stage pTla) or the large arteries present at that level, and 2) tumors showing invasion (stage pTlb) to or beyond the level of the muscularis mucosa. Accepted for publication September 27, 1996. All but 4 of the 121 patients have been followed to death or Su ported by the Oncology Center for Western Sweden and the G(lteLrgMedical Society. for at least 5 years. These 4 patients were alive in 1994 but 800

DEPTH OF INVASION IN STAGE T1 BLADDER CANCER

80 1

A, stage pTla grade 2 tumor with superficial lamina propria invasion. H & E, reduced from X 140. B , sta e pTlb grade 3 tumor with deep lamina propria invasion. Tumor reaches level of muscularis mucosa (arrowheads). d , detrusor muscle. H E, reduced from X56.

k

have not attended regular clinic visits, mainly due to old age. Recurrence was defined as a new tumor diagnosed a t least 3 months after the primary operation. Disease progression was defined as the development of detrusor muscle invasion or metastatic disease. The chi-square test was used for comparison among groups. Logistic regression was used for calculation of odds ratios with 95% confidence intervals regarding progression rates among groups. RESULTS

Deep invasion into the lamina propria (stage pTlb) was identified in 48 patients (40%) and superficial invasion (stage pTla) was noted in 65 (54%). The depth of invasion could not be evaluated in 8 cases (stage pT1 x, 6%).The distribution of the various tumors is shown in table 1. Mean and median patient age was 73.1 and 73 years, respectively (range 48 to 97) and 70% were men. There was no difference in mean age or male-to-female ratio between stages pTla and pTlb tumors. Only 3 patients were initially treated with intravesical therapy and 7 with radiotherapy. Another 7 patients underwent cystectomy after diagnostic or therapeutic transurethral resection of bladder tumors (table 2). Patients with stage pTlb cancer had a significantly greater progression rate than those with stage pTla disease (table 3). o f the patients with stage pTlb grade 3 disease (5 treated with early cystectomy were excluded from the study) 58% had progression in stage, compared to 36% of those with stage pTla grade 3 disease (1treated with early cystectomy was excluded from analysis, the difference was not significant). There were too few patients with stage pTlb grade 2 cancer for a meaningful comparison. The difference in pro-

TABLE1. Relationshio between DTI subcategorv and grade ~~

~

Total No.

No. Grade (%)

Stage 1

2

3

RS.

34 (52) 26 (40) 65 10 (21) 38 (79)* 48 48 (40) 68 (56) 121 DTlt 5- ( 4 -,) * Significantly different from stage pTla (p <0.001). t Includes 8 stage T1 tumors in which the depth of lamina pmpna invasion pTla pTlb

5 (8) 0

gression rates between grades 2 and 3 tumors was significant (p <0.05). The crude survival 5 years after the initial diagnosis was less (42%) in patients with stage pTlb grade 3 compared to stage pTla grade 3 disease (58%, table 4). The crude as well as corrected survival curves were not significantly different (log rank test). At the most recent review of the cohort in 1994,45%of the patients with stage pTlb grade 3 cancer had died of urothelial disease compared to 23% of those with stage pTla grade 3 tumor (table 5). The percentage of patients with stage pT1 grade 3 disease and associated carcinoma in situ is shown in table 6. The rate of progression was greater in both groups when carcinoma in situ was found a t the primary operation. Vascular invasion in the primary tumor was identified in 8 of 48 patients (17%) with stage pTlb and 4 of 65 (6%) with stage pTla disease. Of these patients 9 died of bladder carcinoma or had metastatic disease within 6 years. Three patients were alive without evidence of disease after 43 to 77 months (2 had undergone transurethral resection of bladder tumors only and 1underwent transurethral resection of bladder tumors and radiotherapy). There was no difference in the rate of progression among 81 patients with single primary tumors compared to 40 with multiple primaries. Disease recurred in 80% of the patients, with no difference between stages pTla and pTlb (those treated with early cystectomy were excluded from analysis as were patients who died within 3 months after diagnosis). A total of 11 patients (median age 72.5 years) underwent full dose radiotherapy (60 Gy. or more) after transurethral resection of bladder tumors was done for diagnosis or for a recurrent stage Ta or T1 tumor. Seven of these patients had progression to deep muscle invasion and died of the disease within 47 months. Of 2 patients with a contracted bladder (capacity 100 ml.) 1died of intestinal perforation with peritonitis and l of anastomosis insufficiency after urinary diversion. One patient had severe hemorrhagic cystitis during radiotherapy, which necessitated cessation of oral anticoagulant therapy. The patient died of a pulmonary embolism 6 weeks later. Only 1of the 11patients treated with radiotherapy was tumor-free with normal voiding until death from intercurrent disease after 32 months. A total of 11 patients (medianage 69 years) underwent cystectomy after transurethral resection of bladder tumom

802

DEPTH OF INVASION IN STAGE T1 BLADDER CANCER TABLE 2. Trecrtment of 121 patients with primary stage pT1 disease stage

pTla: Grade 1 Grade 2 Grade 3 pTlb Grade 2 Grade 3

No. Pts.

9% Cystectomy

% Radiotherapy (60Gy. or more)

8 Intravesical Therapy Initially

lklapd

Initially

Delayed

5 34 26

0 4

0 18 12

0 0 4

12 12

0 0 4

0 15 4

10 38

0

0

5

5

10 11

10 8

0 13

13

65 48

1 4

14 4

1 10

11 8

1 10

9 10

6

10

pTla pTlb

0

INtidy

0

pT1* 121 2 11 6 10 Includes 8 stage T1 tumm in which the depth of lamina propria invasion could not be chsified (stage pTlx).

stage

Grade2

5% Alive

NoJI’otaI No. (5%)

Grade 3

pTla 9/34 (26) 9/25 (36) 18/59 (31) pTlb 4/10 (40) 19/33(58) 23/43 (53). 30/61(49)$ 43/109 (39) pTlt 13/48 (27) Seven patients treated with early cyatecbmy and 5 with grade 1tumor were excluded h m analyeis. Significantly different from stage pTla, p <0.05. odds ratio 2.6 (1.2-5.9). t Includes 7 stage T1 tumors in which the depth of lamina propria invasion muld not be classified (stage pTlx). $ Sienifieantly different from stage pT1 grade 2, p <0.05, odds ratio 2.7 (1.2-6.0).

Table 4. Ovemll survival stage

8 survival

No. Pts. 1Yr.

3 Yrs.

5 Yrs.

5 34 26

80 91 92

60 71 77

60

10 38

80 95

50 63

40 42

pTla pTlb

65 48

91 92

72 60

54 42

pT1*

121

91

68

49

pTla: Grade 1 Grade 2 Grade 3 pTlb Grade 2 Grade 3

0

TABLE6. Status at followup in 1994

TABLE 3. Relationship between grade or stage of tumor and progression of disease No. RagremioIlrrotal No. (%)

Delayed

stase pTla: Grade1 Grade2 Grade 3 oTlb Grade 2 Grade3 pTla pTlb

No. Pta.

% Dead

Alive With Disease*or

NoEvidence Of Disease*

unknownstatus*

26

20 32 38

10 38 65

5

34

48

Disease

Other Cause

20 9 16

20 24 23

40 35 23

20 34

10 5

40

30 16

34 31

12

23

6

45

44$

31 19

pT14 121 33 9 32 26 * Minimum followup 5 years. t Four patients not examined later than 43 months after diagnosis but alive after 60 months. $ Significantly different from stage pTla, p ~ 0 . 0 5 . 5 Includes 8 stage T1 tumors in which the depth of lamina propria invasion could not be classified (stage pT1x 1.

50 58

* Includes 8 stage T1 tumors in which the depth of lamina propria invasion could not be classified (stagepTlx).

performed for diagnosis or for a recurrent stage Ta or T1 tumor. Of these patients. 10 were alive without evidence of disease or had died of intercurrent disease at followup in 1994.Only 1 patient died of metastatic disease. Upper tract tumors were diagnosed in 5 patients aRer the bladder tumor diagnosis. One patient was alive 6 years after nephrourekecbmy and 4 died of metastatic disease. These 4 patienta as well as 5 (among the cohort of 121)who died of treatment related complicationswere considered to have died of the dieease. Four patients. died of complications of radiotherapy; 3 of the cases were discussed, and 1 died of multiple intestinal perforations with fistula formation and osteitis of the symphysis. One patient died of pulmonary embolism 4 weeks after cystectomy.

TABLE6. Relationship between initial carcinoma in situ and progression StagdGrade

No. Pta.

No. Ca In Situ h&Fssiodoa No. (%) Initial None

Total No. (96)

Tld3 25 6 / 9 (56) 4/16 (25) 9 (36) 33 10114 (71) 9/19 (47) l 9 J 5 8 J TlW3 TotalTY3 61; 15/24 (63) 15/37 (41) 30(49) Seven patienta treated with early cystectomy were excluded from analysis. * Includes 3 stage T1 grade 3 tumors in which the depth of lamina propria invasion could not be classified (stage pTlx).

series was only 57 years, which indicated a heavy selection of patients for such treatment. The positive outcome after cystectomy was confirmed in our study, in which only 1 of 11 patients died of urothelial carcinoma. External beam radiotherapy was used mainly in patients considered unfit for cystectomy. This treatment was associated with high rates of local failure, contracted bladders and intestinal complications. The overall survival was poor. Full dose radiotherapy does not seem to have been of benefit to our patients and seems to have caused more harm than good. Birch and Harland also raised the issue of whether radiotherapy really improves the prognosis of patients with stage T1 grade 3 DISCUSSION disease.7 The treatment of patienta with stage T1 grade 3 tumors A conservative approach with transurethral resection and has been a subject of discussion and controversy for years.’ intravesical immunotherapy is another a l t e r n a t i ~ e .This ~ Stockle et al showed that early cystectomy is a treatment treatment appears reasonable, particularly in elderly paoption with excellent results measured as 5-year survival.* It tients in whom cystectomy is associated with significant morshould be noted that the mean age at cystectomy in their tality. Mean age a t diagnosis was as old as 73 years among

DEPTH OF INVASION IN STAGE T1 BLADDER CANCER

our patients with stage T1 tumors, which may be explained by the inclusion of all cases in the region. We found, as stated previously, that tumor grade at presentation is an important prognostic factor in stage T1 bladder cancers,1° and grade 2 tumors have a better prognosis than grade 3 disease. Furthermore, our results indicate that patients with stage pTlb grade 2 tumors have as poor a prognosis as those with stage pTla grade 3 disease. The influence of depth of lamina propria invasion on prognosis has been studied previously.4-5. These authors found lower 5-year survival and higher progression rates in stage pTlb compared to pTla disease. Our results support these observations. The differences in survival and progression rates between the groups in the previous studies4.6.11 were much greater and reached statistical significance despite having fewer patients (36 to 110) than in our study. An explanation may be that these reports come from single institutions in contrast to our report, which encompasses all patients in a geographical area treated by a large number of urologists and surgeons. There presently is no consensus as to whether patients with stage pTla bladder cancer should be treated differently than those with stage pTlb disease. The fact that the prognosis is worse in patients with stage pTlb disease does not necessarily mean that cystectomy is indicated in such cases. A repeat resection after 2 to 6 weeks has been recommended when the primary tumor was stage pT1 grade 3.7~12One report showed a surprisingly high rate of tumor at the repeat resection but did not separate the tumors into deeply or superficially invasive.12 Tentatively, if the fist histopathological report showed stage pTlb, it seemed to be more important to repeat resection than in cases of superficial invasion. We previously showed that vascular invasion in the lamina propria is associated with a poor prognosis,13 a result that was confirmed in our present study. In our previous report 10 of 99 patients with lamina propria invasion (10%)had lymphatic invasion and 7 of the 10 died of bladder carcinoma.ls "he difference in prognosis with and without lymphatic invasion was statistically significant (p 4 . 0 1 ) . Patients with lymphatic invasion (in both studies) treated with radiotherapy or cystectomy did not have improved survival compared to those treated with transurethral resection of bladder tumors only. We agree with Lamm, who suggested that these patients may be treated just as well with transurethral resection of bladder tumors and intravesical bacillus CalmetteGuerin.14 Patients with stage pT1 grade 3 and carcinoma in situ at diagnosis had an increased risk of progressive disease, which confirms the findings of Vicente et al.16 Furthermore, our study shows that patients with carcinoma in situ have an increased risk of progression regardless of whether they have stages pTla grade 3 or pTlb grade 3 disease. This finding underscores the importance of a thorough search for carcinoma in situ in patients with stage pT1 grade 3 cancer. Although intra-observer and interobserver variation is a well-known phenomenon in histopathological evaluations, it has been found to be low in determining the level of muscuh i s mucosa invasion and to decrease with further pract i ~ e .We ~ . found, ~ as did other investigators, that the level of muscularis mucosa invasion can be identified in the majority of the specimens after transurethral resection of bladder tumors. The fact that patients with stage pTlb tumors have a worse prognosis than those with stage pTla tumors in our study as well as in previous studies underscores the importance of stating this finding in the pathology report. Further btudies are needed to determine the optimal therapy for patients with stage T l b tumors.

803

CONCLUSIONS

Grade 3 tumors are significantly more common among patients with stage T1 disease with invasion deeper than the muscularis layer of the lamina propria than among those with superficial invasion. Patients with deeply invasive stage pT1 grade 3 tumors have a greater progression rate and lower 5-year survival rate than those with superficially invasive stage pT1 grade 3 tumors. consequently, the histopathological report should include information regarding the depth of lamina propria invasion. The clinical implication may be that early repeat resection, adjuvant intravesical therapy or early cystectomy should be used more frequently in patients with deep invasion than in those with superficial lamina propria invasion. Radiotherapy is associated with a poor survival and may even be harmful. Our colleagues at the departments of urology, oncology, surgery and pathology in western Sweden provided access to patient records and histopathological material. Karin Karlsson, Department of Pathology, Sahlgrenska Hospital, and Ulla-Britt Wallgren, Oncology Center, provided valuable assistance. REFERENCES

1. Diuon, J. S. and Gosling, J. A: Histology and fine structure of the muscularis mucosae of the human urinary bladder. J. Anat., 138: 265,1983. 2. Ro,J. Y.,Ayala, A. G. and El-Naggar, A.: Muscularis mucosa of urinary bladder. Importance for staging and treatment. Amer. J. Surg. Path., 11: 668, 1987. 3. Keep, J. C., Piehl, M., Miller, A. and Oyasu, R.: Invasive carcinomas of the urinary bladder. Evaluation of tunica muscularis mucosal involvement. h e r . J. Clin. Path., 91: 575, 1989. 4. Younes, M., Sussman, J. and True, L. D.: The usefulness of the level of the muscularis mucosae in the staging of invasive transitional cell carcinoma of the urinary bladder. Cancer, 86: 543,1990. 5. Hasui, Y., Osada, Y., Kitada, S. and Nishi, S.: SigNficance of invasion to the muscularis mucosae on the progression of superficial bladder cancer. Urology, 43:782, 1994. 6. Onkologiskt Centrum i Gteborg. Urinbllsecancer. Diagnostik och behandling. Edited by C. Anderstrom, S. L. Johansson and S. Nilsson. Med. Fak.Gteborg: Goteborgs Universitet, 1986. 7. Birch, B. R. P. and Harland, S. J.: The pTlG3 bladder tumour. Brit. J . Urol., &4: 109, 1989. 8. Stockle, M., Alken, P., Engelmann, U., Jacobi, G. H., Riedmiller, H. and Hohenfellner, R.: Radical cystaxtomy-otten too late. Eur. Urol., 1 3 361, 1987. 9. Pansadoro, V., Emiliozzi, P., Defidio, L., Donadio, D., Florio, A, Maurelli, S., Lauretti, S. and Sternberg, C. N.: Bacillus Calmette-Guerin in the treatment of stage T1 grade 3 transitional cell carcinoma of the bladder: long-term results. J. Urol., 164: 2054,1995. 10. Kaubisch, S., Lum, B. L., Reese, J., Freiha, F. and Torti, F. M.: Stage T1 bladder cancer: grade is the primary determinant for risk of muscle invasion. J . Urol., 148: 28, 1991. 11. Angulo, J . C., Lopez,J. I., Grignon, D. J. and Sanchez-Chapado, M.: Muscularis mucosa differentiates two populations with different prognosis in stage T1 bladder cancer. Urology, 45:47, 1995. 12. Kliin,R., Loy, V. and Huland, H.: Residual tumor discovered in routine second transurethral resection in patients with stage T1 transitional cell carcinoma of the bladder. J. Urol.. 146: 316, 1991. 13. Anderstrijm, C., Johansson, S. and Nilsson, S.: The sigruficance of lamina propria invasion on the prognosis of patients with bladder tumors. J. Urol., 124: 23, 1980. 14. Lamm, D. L.: Prophylaxis for recurrent transitional cell carcinoma. Urology, suppl. 5,37: 21, 1991. 15. Vicente, J., Laguna, M. P., D u e , D., Algaba, F. and Chechile, G.: Carcinoma in situ as a prognostic fador for G3pT1 bladder tumours. Brit. J. Urol., 6& 380,1991.

804

DEPTH OF INVASION IN STAGE T 1 BLADDER CANCER EDITORIAL COMMENT

Responding to the need to subclassify patients into populations that will have tumors of similar biological potential, the authors addressed the depth of lamina propria invasion of transitional cell carcinomas. They concluded that depth of lamina propria invasion can be reliably determined in biopsies and transurethral resection specimens, and can be expressed as stages pTla and pTlb, prognosis for patients with stages pTla and pTlb grade 3 tumors is significantly different if there is no associated carcinoma in situ and no treatment other than transurethral resection of bladder tumors, and there is no value of substaging if patients are treated with radical cystectomy. Interestingly, there was no statistical difference based on substaging in the number of patients who were alive with no evidence of disease but only differences in the causes of death. Based on this information the authors recommend that depth of lamina propria invasion should be included in (? every) pathology report. The study is well done with appropriate pathology expertise but before we accept the recommendations, particularly in a n environment of limited resources, it seems reasonable to ask a few questions. Can the various levels of lamina propria be reliably identified in transurethral resections of bladder tumors and biopsies in everyday clinical practice rather than in reviews performed by individuals whose attention can be concentrated on large numbers of cases reviewed at 1 time? How often is the distinction necessary, that is what percent of patients will have stage pT1 grade 3 transitional cell carcinoma in the initial biopsies and transurethral resection specimens? What specific treatment decisions for individual patients will be made based on this information? Can treatment decisions affecting these patients be made using appropriate pathological information provided as needed rather than in general? All who have examined the anatomy of the lamina propria agree that a continuous layer of smooth muscles separating the connective tissue into 2 distinctive compartments, such as is seen in the colon, essentially does not exist in the human bladder. Instead, one finds scattered fascicles of smooth muscle loosely associated with relatively thick walled blood vessels occurring a t various degrees of development and appearing at various levels. Tissues from transurethral resection of bladder tumors and bladder biopsies tend to ball up in solution so that the mucosa is often well oriented but the underlying tissue is not. Under these circumstances tangential sections are common. To require the practicing pathologist to base pathological staging on a variably developed, poorly demarcated landmark in a maloriented specimen seems to me to invite observational inaccuracies. We might be better served to recognize that transitional cell carcinomas spatially associated with the detrusor muscle in histological sections, whether or not actual invasion can be demonstrated, are likely to act as if they were muscle invasive tumors and should perhaps be included among the stage pT2 lesions for the purposes of treatment decisions. From the tables and the text of this article one might conclude that the actual number of patients who might benefit from substaging is small. Starting with a registry of 759 individuals, we must subtract those without appropriately documented stage pT1 transitional cell

carcinomas (638) plus those with grades 1 and 2 tumors (too few for statistical comparison), plus those with associated carcinoma in situ (not different in prognosis). This leaves roughly 45 patients (6%),of whom the disease in 8 (18%)could not be substaged pathologically. Of course, an observation that is unlikely to have been made by chance (statistically significant using the chi-square test) is not necessarily clinically important to an individual patient. As far as counseling is concerned, one wonders what impact a 22% difference in survival or progression for groups of patients might mean to the individual. Considering that no statistically significant impact on survival based on substaging could be documented for patients having associated carcinoma in situ and vascular invasion or for those treated with cystectomy, the value of substaging seems to be in those not treated, a n unusual event presently, a t least in the United States. Even if we could assume that these figures would hold true for patients with stage pT1 tumors treated with topical therapy and that some sort of reliable pathological evaluation could be achieved, the patient with a stage pTla grade 3 transitional cell carcinoma would have to accept a 36% risk of progression and a 23% risk of death from disease as the best outcome. In recent years our collective attention has turned to the identification of a n ever increasing number of putative prognostic factors, many of which involve pathological assessment. Substaging stage pT1 transitional cell carcinoma is only 1of a list that might include markers of genetic activity, stromal response, tumor volume, vascular density and many more, in addition to the more standard grading and staging. At the same time, the resources available to provide this information are diminishing. Is there a more efficient way to use o u r resources than to require a long list of pathological information on every case? I believe so. Any information contained in a tissue section is preserved indefinitely. It can be retrieved when needed for specific purposes, investigational or for individual patient care. Pathological information necessary for only a small subgroup of patients is particularly suited to this approach. William M. Murphy Department of Pathology University of Florida College of Medicine Gaineswille, Florida REPLY BY AUTHORS According to substaging of pT1 tumors deep and superficial lamina propria invasion does not require additional sections, special stains or special equipment. It is unlikely that all pathologists can subgroup as many as 90% of the pT1 tumors, a s in our study. However, determining the depth of tumor invasion into the lamina propria is relatively simple and it requires little effort, and most pathologists can readily become proficient with this technique. All previous publications (references 4 , 5 and 11in article) as well as our study have arrived a t the same conclusion: the depth of invasion into the lamina propria has an impact on prognosis. Therefore, we believe that this information should be communicated to the urologist.