Carcinoma of the prostate: Results of radical radiotherapy (1970–1985)

Inr J Radialron Oncology Biol. Phys., Vol. 26. pp. 203-2 Printed in the U.S.A. All rights reserved.

IO Copy&i

0360-3016/93 $6.00 + 03 Q 1993 Pergamon Press Ltd.

??Clinical Original Contribution CARCINOMA OF THE PROSTATE: RESULTS OF RADICAL RADIOTHERAPY (1970-1985) WILLIAM DUNCAN, M.B.CH.B., PADRAIG WARDE, M.B., CHARLES N. CATTON, M.D., ALASTAIR J. MUNRO, M.B.CH.B., ROY LAIUER, M.D., TEHANY GADALLA, M.Sc. AND MARY K. GOSPODAROWICZ, M.D. Departmentof Radiation Oncology, PrincessMargaretHospital, Toronto, Ontario,M4X 1K9 Canada Purpose: To determine the outcome and prognostic factors in patients with localized carcinoma of the prostate treated with external beam radiation therapy. Methods and Materials:A retrospective review of 999 patients with histologically confirmed adenocarciaoma of the prostate treated mdically with megavoltage irradiation at the Princess Margaret Hospital between 1970 and 1985. Prognostic factors were analyzed using recursive partitionning method. Results: Overall survival at 5 and 10 years were 69.8% and 40.1% for the whole group. The cause-specificsurvival rates were 78.9%and 53.58, respectively.Tbe cause-specificsurvivalrates were significantlydifferentat 10 years by T stage, Tl being 79.0%,T2 66.08, T3 55%and T4 22%.The overall clinical local control rates was 77% in the first 5 years following treatment. There was no statistically significant difference in the local control rates of Tl and T2 stage disease at 5 years, the combined rate being 88%. Significant differences were observed between other stages, being 76% for T3 and 55% for T4. At 10 years the control rate for Tl tumours was maintained for Tl stage disease (92%) but was significantly reduced for other stages, T2 75%, T3 67% and for T4 37%. In the whole group 33.5% of patients had distant metastases in the first 5 years. The distant relapse rates at 10 years were significantly different by T stage, being 20% for Tl, 33% for T2, 55% for T3 and 87% for T4. Multivariate analysis demonstrated that only T stage and histological grade were independent prognostic covariates for causespecliic survival. Age was the only other independent variate in terms of overall survival. The late radiation related morbidity was 2.3% overall; 1.3% afiectlng rectum and recta-sigmoid and 1.0% arising in the bladder. Conclusion: In terms of survival the results of radiotherapy of intracapsular disease were excellent, but they were less satisfactory in patients with direct extracapsuhu extension. The assessment of local control was difficult and may have reflected more the lack of local disease progression rather than true local tumor control. The treatment was well tolerated and there were few serious late complications.

METHODS

INTRODUCTION Radiotherapy has an established role in the curative management of the patients with carcinoma of the prostate ( 1, 4). Overall survival is considered to be similar for patients with intracapsular disease following either definitive radiotherapy or radical prostatectomy (9). Radiotherapy is claimed to be the treatment of choice for patients with locally advanced disease in the absence of metastases. This paper records the results of treating 999 patients with prostate cancer, stages T 1-T4 NxMo who were managed by radical radiotherapy at Princess Margaret Hospital (PMH), Toronto. A multivariate analysis of prognostic factors based on this experience is presented.

AND

MATERIALS

The records of 3293 consecutive patients registered at the PMH between January 1970 and December 1985 with the diagnosis of carcinoma of the prostate have been reviewed. Most patients (68.5%) were referred for palliative radiotherapy. A group of 1117 patients received palliative radiotherapy to painful metastasis; 645 patients were given palliative treatment to the primary tumor. A group of 493 patients (24.9%) were seen in consultation only, since they were considered unsuitable for either radical or palliative radiotherapy. They were given supportive care or managed by androgen ablative therapy. The increasing referral of patients suitable for radical

Reprint requests to: Mary K. Gospodarowicz, M.D., Princess Margaret Hospital, 500 Sherbourne St., Toronto, Ontario M4X 1K9 Canada.

Accepted for publication 3 September 1992.

203

204

I. J. Radiation Oncology 0 Biology0 Physics

treatment over this 16 year period is illustrated in Figure 1. A total of 1038 patients received a radical course of radiotherapy to the primary prostate cancer with or without irradiation of the regional lymphatic nodes. Radical radiation therapy was defined arbitrarily in the context of this review as the prescription of a target absorbed dose of at least 50 gray in 20 daily fractions or a biologically equivalent dose given in larger number of daily fractions. Thirty-nine patients meeting these criteria were excluded from this retrospective review for the following reasons. It was found that 24 patients had evidence of metastatic disease although given a “radical” dose of irradiation to the prostate gland. Fourteen patients had primary tumors of the prostate other than adenocarcinoma. All details of treatment were lost on one patient who was recorded as having received radical radiotherapy (Table 1). The subsequent analysis is based therefore on a series of 999 patients with histologically confirmed adenocarcinema of the prostate who received a radical course of radiotherapy as defined above. The histological diagnosis had been established by a needle biopsy in 427 patients (42.7%) and by TURP in 541 patients (54.1%). The histological grade (well, moderate or poorly differentiated) was established in 9 18 patients. The Gleason scoring system which is currently used, was not used routinely during the whole of this period and so could not be included in the analysis of prognostic factors. It should be noted that 2 17 patients (22%) had received a variety of treatments for cancer of the prostate prior to referral for definitive radiotherapy (Table 2). The majority (195) of patients had received some form of hormonal manipulation. The results of the sub-group of 782 patients treated radically who had not received previous treatment are given separately but interestingly do not differ significantly from the whole series.

All Cases /\ / ’

Radically Treated

‘80

‘75

_

Volume 26, Number 2, 1993

Table 1. Carcinoma of the prostate Patients treated radically Exclusions: Patients with metastatic disease Histology not adenocarcinoma All details lost

1038 24 14 1

Note: Patients selected for radical radiotherapy at PMH: 19701985.

Routine pretreatment assessment included a full clinical history and physical examination in all patients. The details of pre-treatment assessment are given in Table 3. Blood count, serum urea, creatinine, alkaline phosphatase and prostatic acid phosphatase estimations were performed in 93% of patients. Prostate specific antigen assay was not available in the era subject to this study. Radiological investigations included a chest x-ray, IVP, bone scan, lymphogram, and CT scan of the pelvis. Lymphograms were performed in 591 patients (59.1%) and were reported as positive in 174 patients (29.5%), however no pathological confirmation of nodal involvement was obtained. Staging lymph node dissection was performed prior to referral in 134 patients (13.5%). The nodal status was disregarded in the analysis and all patients were analyzed by their T-category. Intravenous pyelograms (IVP) were performed in 838 patients (83.9%), but in only 57 (5.7%) was there evidence of obstruction of the ureters or hydronephrosis. The bone scan was performed in 87 1 patients (87.2%) and was reported to be positive or equivocal in 123 patients ( 12.3%). This group of patients was given radical radiotherapy because the result of the bone scan was considered to have been a “false positive” report. Patients were retrospectively staged according to the TNM classification (UICC 1978). Radiotherapy techniques and doses All patients were treated with megavoltage irradiation. Telecobalt radiation was used in 108 patients and in 89 1 patients high energy photons (18 MV) from a linear accelerator or betatron were employed. Most patients were treated using one of two fractionation regimes. In the earlier years of the study a schedule of 50 Gy in 20 fractions (28 days) was normally used. The whole pelvis Table 2. Carcinoma of the prostate: Treatment to radical radiotherapy

‘85

Year

Fig. 1. Referral of patients with prostate cancer to Department of Radiation Oncology, Princess Margaret Hospital, Toronto ( 1970- 1985) illustrating increase in proportion suitable for radical treatment.

39 999

Patients analysed in this report

Hormone medication Hormones and orchiectomy Orchiectomy Hormone and lymphadenectomy Hormone and prostatectomy Orchiectomy and prostatectomy Orchiectomy and lymphadenectomy Prostatectomy Prostatectomy and lymphadenectomy Previous treatment

prior 118 32 27 7 6 3 2 17 5 217

Carcinoma of the prostate 0 W. DUNCANet al. Table 3. Carcinoma of the prostate: Pre-treatment investigations Investigation

Patients

%

XR chest Alkaline phosphatase Acid phosphatase Bone scan IVU Cystoscopy and EUA Lymphangiogram CT pelvis Skeletal survey Pelvic lymph node sampling

754 944 911 870 838 754 592 304 240 134

96 94 91 87 84 75 59 30 24 14

205

ularly in the Department of Radiation Oncology at PMH. Occasionally follow-up data was conducted by the family doctor or referring urologist. Additional follow-up data were obtained, where necessary, from these physicians, by correspondence with the patients and from death certificates. At each visit the patients had a history and physical examination which included a rectal digital examination as well as estimation of the prostatic acid and alkaline phosphatases. X-rays and bone scans were performed only if indicated clinically. Prostatic biopsies were not routinely performed after radiotherapy. Relapse and survival analysis

by a four field box technique to a dose of 35 Gy in 15 daily fractions with an immediate boost to the prostatic bed of 15 Gy in 5 daily fractions. The boost dose was given by parallel opposed field AP/PA, lateral opposed field or by a 360’ rotation technique. In the later years patients were treated by 60 Gy in 30 daily fractions (42 days). In a group of patients (23.6% of the whole series) with intra-capsular disease and no evidence of regional lymph node spread the prostate alone was treated by a 3 or 4 field technique. The whole pelvis was treated to a dose of 40 Gy in 20 daily fractions. The dose to the prostate was boosted by 20 Gy in 10 daily fractions to a smaller field, usually 10 X 10 cm. by a four field box technique, immediately after the pelvic treatment. The absorbed dose was prescribed at the midline for parallel opposed fields or to the isocenter of the two treatment volumes where multiple fields or rotation techniques were used. Most patients (82%) who received irradiation to the whole pelvis were treated by parallel opposed field. In the remainder a four field box technique was used. All fields were treated daily. The pelvic fields extended from the inferior level of the ischium to the sacral promontory and 1.0 cm lateral to the pelvic side wall. Laterally the fields normally extended from mid symphysis to the mid rectum. The distribution of techniques employed is given in Table 4.

was treated

Follow-up assessment The follow-up protocol required that patients be reviewed every 4 months for the first 2 years and then 6 monthly or yearly thereafter. Most patients were seen regTable 4. Carcinoma of the prostate: Radiotherapy techniques Pelvis Prostate

AP:PA

4 field

None

Total

2 field AP:PA 2 field lateral 3 field 4 field potation Total

164 194 9 5 243 615

0 0 9 139 0 148

32 0 67 120 17 236

196 194 85 264 260 999

Treatment failure was assessed by either clinical or radiological evidence of progressive disease. Failure was as-

sessed occasionally on the evidence of a rising prostatic acid phosphatase. Local failure was defined as the clinical or radiological evidence of residual or recurrent disease in the pelvis within the irradiated volume or a biopsy of the prostate positive for adenocarcinoma more than 12 months after the course of radiation therapy. Survival curves have been calculated using the actuarial method. All survival rates and disease free intervals were calculated from the date of start of radiotherapy. In calculating the cause-specific survival only patients with known cause of death other than prostate cancer were censored. Patients with active disease where the exact cause of death was not known were considered to have died of prostate cancer. Patients lost to follow-up were deemed to have died of prostate cancer. Multivariate analysis of prognostic factors Analysis has been made of a large number of factors that may be thought to influence survival following radical radiotherapy. The method of recursive partitioning has been used to determine the clinical, radiological and pathological features that significantly affect survival. The analysis is reported for overall survival and for cause-specific survival. Treatment related morbidity Morbidity was assessed retrospectively and scored according to the RTOG/EORTC system. Radiation related morbidity was scored as acute for effects occurring within the first 3 months after completion of treatment. Late morbidity was defined as clinical features which presented more than 3 months after treatment or persisted for more than 6 months after treatment. It is acknowledged that the documentation of acute and late reactions may be incomplete since there was no standard protocol, for assessing morbidity during this period. The review probably underestimates the overall morbidity, particularly the incidence of late radiation related may be incomplete since there was no standard protocol assessing morbidity during this period. The review probably underestimates overall morbidity, particularly the incidence of late radiation re-

I. J. Radiation Oncology 0 Biology 0 Physics

206

lated complications. However, it is considered that all the severe complications have been recorded accurately. Descriptive statisticsof series The median age of patients was 65 years, the range being 26-92 years. The distribution of T stage and histological grade of the tumours is shown in Table 5. Just over half (53.4%) of patients presented with clinical evidence of extracapsular disease. Only 7% had Tl stage cancer and 34.1% of patients had T2 stage disease. Fifty-four patients (5.4%) had inadequate information to allow retrospective TNM staging to be undertaken with confidence. They are recorded as TX in the Tables. The majority of patients (74%) had well or moderately differentiated tumors; 18% of patients presented with a poorly differentiated histology. Patients with locally advanced disease (stage T3/T4) had a significantly higher incidence of poorly differentiated cancer (23.4%) compared with patients with stage Tl/T2 disease (12.2%) (p < 0.0005). RESULTS

Overall and cause-speciJicsurvivalat 5 and 10 years The actuarial overall survival rates for the whole group were 69.8% and 40.1% at 5 and 10 years respectively. The cause-specific survival was 78.9% and 53.6% at 5 and 10 years (Table 6). The corresponding data for the group of 782 patients who had not received any treatment (other than TURP) before referral for radical radiotherapy is given in Table 7. The overall survival was 73% and 4 1.9% at 5 and 10 years; the cause-specific survival being 79.8% and 5 1.9%, respectively. These differences are not significant from those of the whole series. The overall survival of patients with Tl and T2 disease was similar, 82.0% and 54.5% at 5 and 10 years, respectively. These survival figures are significantly higher than for patients with T3 stage disease which was 69.7% at 5 years and 42.5% at 10 years. Patients with T4 disease had a significantly poor overall survival than patients with T3 disease which was 50.1% at 5 years and only 15.8% at 10 years. Comparison of these data with the cause-specific survival clearly shows that almost all the observed mortality was stated as being due to prostate cancer. At 10 years Table 5. Carcinoma of the prostate: Distribution and histological grade

by T stage

Volume 26, Number2, 1993 Table 6. Carcinoma of the prostate: Radical radiotherapy survival by T stage Overall survival (%) T stage

5 years

10 years

5 years

10 years

1 2 3 4 Total

83.8 81.8 69.7 50.1 69.8

57.8 53.6 42.5 15.8 40.1

93.0 91.8 78.5 56.8 78.9

79.0 66.0 54.8 22.0 53.6

the cause-specific survival was significantly different by T stage. Patients with T 1 disease had a 10 year cause-specificsurvival at 79%, compared with T2-66%, T3-55% and T4-22%. Analysis was also performed on the 592 patients who had bi-pedal lymphograms performed. Overall survival at 5 and 10 years was similar by T stage for the groups of patients with lymphograms reported as positive or negative. This result suggests that a lymphographic determination of nodal involvement, in the absence of histologic confirmation, did not provide prognostic information and therefore outcome was analyzed by T stage alone. Transuretheral resection of the prostate (TURP) was found to have a significant influence on cause-specific survival for patients with stage T3 and T4 tumors but not for patients with Tl or T2 stage disease. However, when these results are corrected for histological grade of tumor the differences in survival are no longer observed. In this series TURP was more often performed in patients with locally advanced, stage T3 or T4 poorly differentiated tumors associated with a significantly poorer prognosis. A detailed analysis of the effect of TURP on survival is the subject of another paper (2). The influence on survival of the delay from diagnosis to the start of treatment was also investigated. Three subgroups were analyzed by interval between diagnosis and the start of radiotherapy. No significant difference was seen in the groups who were treated less than 6 months from the alleged date of the diagnosis, between 6 and 12 months and more than 12 months of diagnosis. Overall 560 patients (56%) in the series remain alive and apparently free of prostate cancer with a median survival of 4.7 years.

Table 7. Carcinoma of the prostate: Radical radiotherapy Overall survival (%)

Differentiation T stage

Well

Moderate

Poor

Not stated

Total

1 2 3 4 x Total

34 128 92 35 17 306

31 149 181 61 14 436

4 43 60 53 16 176

1 21 29 23 7 81

70 341 362 172 54 999

Cause specific (%)

Cause specific (%)

T stage

5 years

10 years

5 years

10 years

1 2 3 4 Total

87.7 82.9 69.9 54.9 73.0

54.3 55.7 43.3 12.1 41.9

95.2 89.1 78.1 56.7 79.8

66.5 66.1 53.5 13.3 51.9

Note: Survival by T stage of previously untreated patients.

207

Carcinoma of the prostate 0 W. DUNCANet al.

Local and distant relapse The distribution of local and distant relapse by T stage is given in Table 8. Documented evidence of failure to control the primary tumor was present in 236 patients (23.6%) of whom 137 (6 1.2%) also developed distant metastasis with 5 years from treatment. In 184 patients ( 18.4%) there was evidence of distant metastases without any record of local failure. The local tumor control rates at 5 years was 92% for patients with Tl disease, 86% for T2 disease, 76% for T3 and 55% for patients with T4 stage disease. At 10 years the local control rate of 92% was maintained for T 1 stage tumors. There was a significant reduction in local control between 5 and 10 years in patients with stage T2, T3 and T4 disease for whom the local tumor control rates fell to 75%, 57% and 37% at 10 years, respectively. Distant metastases were evident overall in 33.5% of patients within the first 5 years and in 5 1.6% of patients within 10 years. The differences between 5 year distant relapse rates for patients with stage T2 and T3 tumors ( 16.6%) and between T3 and T4 tumors (12.8%) were statistically significant. (T2/T3 - p = 0.0001; T3/T4 - p = 0.0001) The first site of distant failure was bone in most cases (85.2%). The other sites of initial relapse were lymph nodes in 2 1 (20%) patients and liver in nine (0.9%) patients. In 10 ( 1.O%) patients there was only evidence of raised acid phosphatase when the primary tumor was apparently controlled on assessment by rectal digital examination. Radiation related morbidity Almost all patients developed some acute reaction during or just after the course of irradiation. They were usually mild and self-limiting. Severe acute gastrointestinal morbidity (RTOG Grade 3 and 4) was recorded in 25 patients (2.5%). In 13 patients (1.3%) there was documentation of severe acute bladder symptoms. No patient had both severe bladder and bowel symptoms during or immediately after radiotherapy. No patient died of an acute radiation related reaction. Documentation of serious late morbidity was also uncommon. (Table 9) Nine patients (0.9%) were observed to have Grade 3 or 4 late radiation related complications in the bladder. One patient was assessed to have died of a serious late bladder reaction (RTOG grade 5 compliTable 8. Carcinoma of the prostate: Radical radiotherapy Local relapse (%)

Distant relapse (%)

T stage

S years

10 years

5 years

1 2 3 4 Total

8.1 14.5 23.6 45.3 23.0

8.1 25.4 42.5 63.1 43.0

12.1 19.5 36.1 48.9 33.5

Note: Disease control by T stage.

10

years

20.1 32.5 55.0 86.8 51.6

Table 9. Carcinoma of the prostate: Radical radiotherapy Bladder Bowel 0 1 2 3 4 5 X Total

0 773 68 34 4 3 10 0 883

1

2

5 46 4 14 1 14 10000 0 3000 0000 0 0 11 77

Note: Late complications

3

4

5

X

Total

5 2 1

1 0 0

1 0 0

1 9

0 1

0 1

0 0 1 0 1 0 15 17

831 88 51 5 7 1 16 999

(RTOG scores).

Severe late bowel complications were observed in 13 patients ( 1.3%) one of whom died directly as a result of these complications. The crude overall serious late radiation related morbidity was 2.3%.

cation).

Multivariate analysis (recursive partitioning) The decision tree for all 999 patients treated by radical irradiation is illustrated in Figure 2. It should be noted that the p-values for all partitions are < 0.03. It can be seen that the first partition was determined by T stage. Patients with T4 stage disease did significantly less well than those with earlier stage disease. Patients with T4 stage disease and poorly differentiated tumors had a very low probability of survival and formed a “poor” prognosis group. Patients with T4 stage disease and moderate or well differentiated cancers had a “fair” prognosis, similar to that of patients of 70 years of age or over with earlier stage disease and a group of patients less than 70 years with poorly differentiated T3 stage cancers. A group of patients associated with a “moderate” prognosis comprised patients under 70 years of age with either moderate or well differentiated T3 stage cancers, poorly differentiated Tl and T2 tumors and those aged 60-69 years with moderate or well differentiated tumors. A “good” prognosis group consisted of patients less than 60 years of age with moderately or well differentiated Tl and T2 cancers. The distribution of these prognostic groups together with their overall survival and median survival rates are given in Table 10 (Fig. 3). Overall survival was significantly related only to age, T stage, and histologic differentiation in this series of patients (Fig. 3). TURP did not adversely affect prognosis as in independent variate. This was also confirmed for the group of 782 previously untreated patients. Their survival rates by prognostic group are shown in Table 10. Cause-specific survival in this series of patients was influenced significantly only by the T stage and histological grade of the tumor as independent variates. DISCUSSION The results of this study are in agreement with those of most other major series ( 1,4,5,8, 12, 13) and consistent

208

I. J. Radiation Oncology 0 Biology0 Physics

Volume 26, Number 2, 1993 Note:

P < 0.03 for all Partitions

Fig. 2. Decision tree of 999 patients after radiotherapy for prostate cancer (all stages) with survival as the endpoint. The p value -C 0.03 for all partitions.

with an earlier review of results reported from PMH (3). During the period of this review increasing numbers of patients with prostate cancer were referred for radical radiation therapy, indicating the widespread recognition of its role in the management of these patients. Radiotherapeutic resources were limited and these patients were treated using relatively simple techniques and an overall treatment time of 4 weeks. The modal target absorbed dose was 50 gray in 20 daily fractions. Current techniques is now more sophisticated and the standard treatment at PMH delivers 65 gray in 35 daily fractions in patients with locally advanced disease. One may expect that local tumor control rates may be improved without any increase in morbidity as a consequence of these changes that have proved advantageous in many centres in North America (4, 8, 12, 13). Table 10. Median survival and survival rates at 5 and 10 years for oroanostic arouos Survival rates at Prognostic group

Patients

5 years %

10 years %

Median survival years

1 2 3 4

79 474 389 53

91 77 61 34

83 49 23 5

14.1 9.7 6.7 2.7

Excellent overall survival rates are recorded for patients with Tl and T2 stage prostate cancers at 5 and 10 years. The survival is close to that expected of a “normal” male population of similar age distribution. Survival is less good for patients with stage T3 disease being 70% at 5 years in this series. This analysis confirms that most of the mortality observed in this group is due to prostate cancer rather than intercurrent illness. Consideration of offering adjuvant management by androgen deprivation would seem appropriate in this group of patients. Patients with T4 disease, in the absence of overt metastasis, have been shown to have a very poor prognosis after radical radiotherapy alone. In this group of patients hormonal management would currently seem to be the preferred management, with irradiation of the prostate reserved for patients with progressive local features that are unresponsive to androgen deprivation. Local tumor control of prostate cancer is difficult to assess with great certainty after radiation therapy. The determination of the PSA levels post XRT helps to identify early failure, it does not indicate the site of failure and thus by itself cannot be an indicator of local recurrence. Also a normal PSA level cannot be used as a definitive indicator of local control, since a microscopic residual disease may be present with a normal PSA as with any other normal biochemical marker. However, our assessment of “local control” may be an indication rather of

Carcinoma of the prostate 0 W. DUNCAN

T STAGE

3

112

209

et ai.

4

Diffsrrntiation

Prognottlc groups best to worst Group 1, 79 patients

Group 3. 309 patbnts

Group 2. 474 patients

Group4. 53 pstkmts

Fig. 3. Chart of four prognostic groups for survival determined by the three independent variates obtained from multi-variate analysis. lack of evidence of local disease progression. Some patients will have been treated by androgen deprivation because of distant relapse. Lasting assessment of local control may therefore have been influenced by secondary management. It is interesting that in spite of the difficulties of both staging and the assessment of local control, a highly significant correlation was observed between T stage and local control (Table 10). The observations show a high probability of local tumor control for Tl (92%) and T2 (75%) stages of prostatic cancer assessed up to 10 years after radiotherapy. When the primary tumor is considered clinically to be extracapsular local control of the primary was less good. In patients with T3 stage disease, 23% had evidence of local relapse in the prostate in the 5 years following irradiation. It may be important to note that most of these patients received only 50 Gy in 20 daily treatments. This relapse rate is consistent with the data of Hanks et al. (7) and may be reduced to perhaps 17% by increasing the total absorbed dose to 70.0 gray or more but at the cost of a higher incidence of local radiation related morbidity. Currently all patients with stage T3 disease receive at least 65.0 gray in 35 daily fractions to the prostate. In these patients it would seem appropriate to evaluate the possibility of improving the results of radiotherapy by combining “cytoreductive” or adjuvant hormonal therapy. In patients with T4 stage disease 63% had evidence of local progression of the primary cancer within 10 years of x-ray therapy. In many of these patients management by androgen deprivation may be the treatment of choice, particularly when the probability of distant relapse is also so very high (55% at 10 years). Patients with stage T4 disease have only a 50% probability of achieving local control and in view of their high metastatic potential are best managed by androgen deprivation therapy. A multivariate analysis has demonstrated that three in-

dependent clinico-pathological factors are of importance in determining overall survival following radical radiotherapy. These are T stage and histological grade and age. Patients may be segregated into four groups with highly significantly different prognoses. Prognosis may be estimated with much greater accuracy using these three variates than by using T stage alone. Other factors, including a history of previous TURP (6, 10) did not significantly affect prognosis. When causespecific survival is considered only T stage and histological grade are independent prognostic variates. These factors should influence decision making in the choice of treatment offered to patients and also the design of innovative regimes of combined management. The results reported in this paper were obtained using simple radiotherapeutic techniques, and compared with many series, a relatively low total dose was given in a short overall treatment time. The treatment was well tolerated. There were few acute complications of any severity. The overall incidence of serious late complications was estimated to be 2.3% which is considered acceptable and similar to other large series (11). These late effects were observed almost equally in the bladder and recta-sigmoid colon. Two patients (0.2%) were considered to have died as a result of radiation related complications, one with bladder morbidity and the other recta-sigmoid ulceration. While the results of radiotherapy of intracapsular disease are excellent, they are less satisfactory in patients with direct extra-capsular extension. Improvements in the management of patients with locally advanced disease should be sought. Randomised trials are currently in progress to investigate the possible advantages of combining endocrine therapy with radical pelvic irradiation and also to evaluate the efficacy of androgen deprivation therapy alone for patients with locally advanced adenocarcinoma of the prostate.

210

I. J. Radiation Oncology 0 Biology 0 Physics

Volume 26, Number 2, 1993

REFERENCES 1. Bagshaw, M. A.; Ray, G. R.; Cox, R. S. Radiotherapy of prostatic carcinoma: Long or short term efficacy. Urol. (Suppl.) 25: 17-23; 1985. 2. Duncan, W.; Gospodarowicz, M. K.; Lakier, R.; Catton, C. N.; Munro, A. J.; Rawlings, G. A.; Thomas, G. M.; Warde, P.; Elkahim, T.; Simm, J. The influence of TURP on prognosis of adenocarcinoma of the prostate treated by radical radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. (In press). 3. Garrett, P. G.; Herman, J. G.; Rawlings, G. A.; Hawkins, N. V.; Gospodarowicz, M. K.; Keen, C. W.; Rider, W. D. Radical external beam radiation therapy for prostate carcinoma. J. Can. Assoc. Radiol. 35: 139-143;1984. 4. Hanks, G. E.; Diamond, J. J.; Krall, J. M.; Martz, K. L.; Kramer, S. A ten year follow-up of 682 patients treated for prostatic cancer in the United States. Int. J. Radiat. Oncol. Biol. Phys. 13:499-505;1987. 5. Hanks, G. E.; Krall, J. M.; Martz, K. L.; Diamond, J. J.; Kramer, S. The outcome of treatment of 38 patients with T-l (UICC) prostate cancer treated with external beam irradiation. Int. J. Radiat. Oncol. Biol. Phys. 14:243248;1988. 6. Hanks, G. E.; Leibel, S.; Kramer, S. The dissemination of cancer by transurethral resection of locally advanced prostate cancer. J. Urol. 129:309-311;1983. 7. Hanks, G. E.; Martz, K. L.; Diamond, J. J. The effect of dose on local control of prostatic cancer. Int. J. Radiat. Oncol. Biol. Phys. 15: 1299-l 305; 1988.

8. McGowan, D. G.; Hanson, J. The National Canadian Study of carcinoma of the prostate treated by external beam radiation (1920-1978). J. Can. Assoc. Radiol. 36:2 16222;1985. 9. National Institute of Health Consensus Development Conference. The management of clinically localised prostate cancer. J. Urol. 138:1369-1375;1987. 10. Pilepich, M. V.; Kmll, J. M.; Hanks, G. E.; Sause, W. T.; Baerwald, H.; Russ, H. H.; Perez, C. A.; Zinniger, M.; Martz, K. L. Correlation of pre-treatment transurethral resection and prognosis of patients with stage C carcinoma of the prostate treated with definitive radiotherapy-RTOG experience. Int. J. Radiat. Oncol. Biol. Phys. 13: 195-199;1987. 11. Pilepich, M. V.; Krall, J. M.; Sause, W. T.; Johnson, R. J.; Russ, H. H.; Hanks, G. E.; Perez, C. A.; Zinninger, M.; Martz, K. L.; Gardner, P. Correlation of radiotherapeutic parameters and treatment related morbidity in carcinoma of the prostate-Analysis of RTOG study 75-06. Int. J. Radiat. Oncol. Biol. Phys. 13:351-357;1987. 12. Rosen, E.; Cassady, J. R.; Connolly, J.; Chaffey, J. T. Radiotherapy for prostate carcinoma: The JCRT experience (1968-1978)-B factors related to tumour control and complications. Int. J. Radiat. Oncol. Biol. Phys. 11:725739;1985. 13. Zagars, G. K.; Van Eschenbach, A. C.; Johnson, D. E.; Owsald, M. J. The role of radiation therapy in Stages A2 and B adenocarcinoma of the prostate. Int. J. Radiat. Oncol. Biol. Phys. 14:701-709;1988.