1992, The British Journal of Radiology, 65, 792-798
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder By D. J. Cole, K. R. Durrant, *J. T. Roberts, *P. J. T. K. Dawes, tH. Yosef and §J. W. Hopewell Department of Radiotherapy and Oncology, Churchill Hospital, Oxford, "Regional Radiotherapy Centre, Newcastle General Hospital, Newcastle upon Tyne, tBeatson Oncology Centre, Belvidere Hospital, Glasgow and §C.R.C. Normal Tissue Research Group, Research Institute (University of Oxford), Churchill Hospital, Oxford, UK
{Received 14 October 1991 and in revised form 24 February 1992, accepted 26 March 1992) Keywords: Bladder cancer, Radiotherapy, Accelerated fractionation
Abstract. 24 patients with muscle invasive carcinoma of the bladder were treated in a pilot study of twice daily fractionation at radiation doses of 1.8-2.0 Gy per fraction to total doses of 54-64 Gy to the bladder and 39.6-44 Gy to the whole pelvis. The treatment aim was to give 32 fractions in 22 days. The interfraction interval was a minimum of 6 h. The principle objective was to record acute and late tolerance, but local control and survival data is also presented. Acute radiation morbidity was scored according to the RTOG system. Grade 2 large bowel effects were seen in 52% of patients, Grade 3 effects in 26% and there was one Grade 4 and one Grade 5 effect. The mean duration of effect was 4.5 weeks although the more severe reactions were also more protracted. Grade 2 urinary effects occurred in 30% and Grade 3 in 17% of patients. The mean duration of effect was 7.2 weeks. There were no Grade 4 or 5 acute urinary effects. Late radiation morbidity was scored according to the EORTC/RTOG system and was assessable in 16 cases who survived more than 6 months. There were two cases (12%) of Grade 1 bowel toxicity, two cases of Grade 1 and three of Grade 2 urinary toxicity. There were no cases of late skin effects. Actuarial analysis at 2 years shows a local control probability of 56% and survival probability of 35%.
The 5-year survival of patients with muscle invasive transitional cell carcinoma of the bladder treated by radical radiotherapy or combined pre-operative radiotherapy and elective radical cystectomy is 40% for T2 tumours (Duncan & Quilty, 1986) and 29-38% for T3 tumours (Bloom et al, 1982). Following radical radiotherapy, complete response in the bladder occurs in up to 56%, but one third of these responders subsequently fail locally (Shearer et al, 1988). Survival in patients treated by irradiation has not been improved by adjuvant methotrexate (Shearer et al, 1988) or cisplatin (Wallace et al, 1991) chemotherapy. The short latency times to local recurrence after radiotherapy in invasive transitional cell carcinoma of the bladder indicate that the disease is biologically aggressive (Duncan & Quilty, 1986). Moderate or poor histological differentiation is reported in 66% of T2 and 89% of T3 tumours (Duncan & Quilty, 1986), Cell kinetic studies demonstrate rapid proliferation of tumour clonogens with a potential doubling time of 3 days (Trott & Kummermehf, 1985; Sheehan & Newman, 1991). It has been demonstrated that acutely responding normal tissue undergoes accelerated clonogen repopulation starting at between 14 and 28 days following the start of a conventionally fractionated course of irradiation (Morris & Hopewell, 1986). It has also been suggested that rapidly proliferating tumours undergo considerable clonogen repopulation during a lengthy course of irradiation (Trott & Kummermehr, 1985) and that this phenomenon is an important cause 792
of local failure in treatments involving conventional fractionation schedules (Fowler, 1986). A reduction in the overall treatment time for radiotherapy may reduce the capacity for tumour clonogens to repopulate and thereby improve local control in invasive bladder cancer. However, this cannot be achieved by an increase in the dose per fraction if late responding tissues are to be spared injury (Quilty et al, 1985) and therefore multiple fractions per day are necessary if the required dose is to be delivered in a shorter time. The fractions of radiation must be given at least 6 h apart to allow for the repair of sublethal damage. However, even with this interfraction interval, repair may not be complete (Hopewell and van den Aardweg, 1988, 1989; Turesson and Thames, 1989) and a small decrease in dose per fraction may be required to compensate for this. This paper presents the results of a pilot study of accelerated fractionation in invasive bladder cancer which was carried out to investigate the potential toxicity of such a schedule. All patients have had a minimum of 12 months follow-up. Materials and methods
Patients with invasive carcinoma of the bladder with no evidence of nodal or distant metastasis who were suitable for radical irradiation with curative intent were selected for this study. Eastern Co-operative Group (ECOG) performance status 0-2 and normal (or only slightly elevated) serum creatinine were required for The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder
entry, but patients over the age of 75 years were generally excluded. Later in the study, those with a history of inflammatory bowel disease, previous lower abdominal or pelvic surgery were excluded. Patients underwent cystoscopy, biopsy and usually endoscopic resection of tumour, and bimanual examination of the pelvis under anaesthetic. Haemoglobin, serum creatinine and liver function tests were measured. Computed tomography of the abdomen and pelvis was performed to exclude nodal and metastatic disease. 24 patients were entered into this pilot study. There were 18 males and six females. The mean ( + SD) age at entry was 66.4 + 6.3 years. There were 21 cases of transitional cell carcinoma (TCC), two cases of squamous carcinoma (SCC) and in one case the histological type was not stated. There were six T2, 14 T3 and four T4 cases. In 19 cases the grade was reported, three were Grade 2 (moderately differentiated) and 16 were Grade 3 (poorly differentiated). Radiotherapy was planned using computed tomography localization and treatment was given on linear accelerator machines of energy 6-8 MV. The whole pelvis was treated during the first week of treatment (10 fractions), the bladder alone was treated during the second week (10 fractions) and treatment to the whole pelvis was completed during the third or early fourth week in 10-12 fractions. The pelvic volumes were treated with three or four field techniques in all but three cases in which an anterior and posterior pair of fields were used. The bladder volumes were all treated with three or four field plans. Two treatments were given each day, 5 days per week and these treatments were given not less than 6 h apart. The intention was to deliver a radical dose of radiation in 30-32 fractions in 15-16 treatment days (treating 5 days per week) with an overall time of 19-22 days. Three centres contributed patients to the study and differences in local practice accounted for the slight differences in dose prescription.
Of the 24 patients entered into the study, the first 13 were treated at 2.0 Gy per fraction and the remaining 11 received 1.8 Gy per fraction. The dose per fraction (and consequently the total dose) was reduced because of concern regarding acute toxicity. The mean (+ SD) dose delivered was 58.85 + 2.2 Gy with a range of 54-64 Gy. The mean number of fractions was 30.8 ±1.1 (range 29-32). The mean overall time was 22.7 + 5.3 days (range 19-44). A protracted skin reaction led to an overall treatment time of 44 days in one patient. Otherwise the maximum treatment time was 29 days. The mean target volume for the whole pelvis was 3030 + 961 cm3 (range 1755-5486) and for the bladder 925 + 313 cm3 (range 427-1595). Response was scored at follow-up cystoscopy which was performed at 6-monthly intervals after radiotherapy. Patients with local failure were considered for salvage cystectomy. Normal tissue toxicity was recorded using the RTOG criteria for early effects and RTOG/EORTC criteria for late effects (see Tables I(a) and (b)). Early effects were scored weekly during treatment and then at 4, 6 and 12 weeks after starting treatment. Late effects were scored at 6 months and then annually. Results
Acute toxicity Acute large bowel toxicity was assessable in 23 patients and is shown in a histogram in Fig. 1. It can be seen that the majority of patients (78%) encountered Grade 2 toxicity (diarrhoea requiring medication) or Grade 3 toxicity (requiring hospital admission or intravenousfluids).The mean duration (in weeks) of toxicity for each grade is stated on the histogram. There was one case of Grade 4 large bowel toxicity. This patient completed irradiation to a dose of 57.6 Gy in 32 fractions in 24 days without problems and then 1 week later developed severe diarrhoea, anorexia,
Table I. (a) Radiaton Therapy and Oncology Group (RTOG) acute radiation morbidity scoring criteria Tissue
Grade 0
Grade 1
Grade 2
Grade 3
Grade 4
Skin
No change
Erythema, dry desquamation
Confluent moist desquamation
Small bowel
No change
Nausea
Requiring i.v. fluids
Ulceration, haemorrhage and necrosis Requiring surgery
Large bowel
No change
Loose stool
Painful erythema, patchy moist desquamation Nausea/vomiting requiring antiemetics Diarrhoea requiring medication
Genitourinary
No change
Frequency x 2 baseline
Frequency > hourly, dysuria
Frequency < hourly, dysuria requiring opiate
Requiring i.v. fluids
Obstruction fistula, perforation requiring surgery Haematuria requiring transfusion, ulceration, necrosis
i.v. = intravenous. Vol. ^65, No. 777
793
D. J. Cole ef*al 100 90
a3 0 O)
o © TO CO
a> oh. a>
Q.
0
1
2
3
4
5
70 60 50 40 30 20 10 0
vomiting and weight loss. She was admitted to hospital with intestinal obstruction and at laparotomy found to have an ileus. A defunctioning colostomy was performed and she made a slow but complete recovery over 8 weeks. 18 months later she developed abdominal carcinomatosis and underwent another laparotomy at which time no sequelae attributable to the radiation were found. This patient had a history of appendectomy followed by ureteric surgery for reflux uropathy about 30 years previously. There was one case of Grade 5 large bowel toxicity. In this case there was an ileo-vesical fistula, probably due to malignancy prior to the start of radiation. Treatment was completed to 60 Gy in 30 fractions in 27 days and 2 weeks later total colectomy was performed for perforated diverticulitis and peritonitis from which the patient succumbed within a month. The acute urinary toxicity is shown in Fig. 2. It is notable that Grade 3 toxicity (frequency and nocturia
3
6
1mi
15 wks
0 1 2 3 4 5 RTOG grade acute urinary toxicity
RTOG acute large bowel toxicity Figure 1. Histogram showing the distribution of acute large bowel toxicity following accelerated fractionation. The number above the columns indicate the mean time to resolution of symptoms after completion of therapy.
n=23
80
Figure 2. Histogram showing the distribution of acute urinary toxicity following accelerated fractionation. The numbers above the columns indicate the mean time to resolution of symptoms after completion of radiotherapy.
< hourly and/or pain requiring strong analgesics) was unusual (17% of cases), but that it was relatively prolonged with a mean duration of 15 weeks. In one of these cases, severe dysuria and frequency persisted for 9 months but was eventually relieved with an antispasmodic agent. The mean duration of acute urinary effects was 5.4 weeks compared to 3.4 weeks for acute large bowel effects and this difference was significant (p = 0.045, Mest). There was one case of Grade 3 skin reaction (confluent moist desquamation). An interruption in treatment was required and the final dose was 58 Gy in 29 fractions in 44 days. Otherwise there were three cases of Grade 2 toxicity and' seven cases of Grade 1 toxicity. In all cases the acute skin reaction had healed 3 weeks after the completion of treatment. The remaining 12 cases showed no toxicity.
Table I. (b) RTOG/European Organisation for Research into the Treatment of Cancer (EORTC) late radiation morbidity scoring scheme Tissue
Grade 0
Grade 1
Grade 2
Grade 3
Skin
None
Mild atrophy, mild alopecia, ± pigmentation
Bowel
None
Genitourinary
None
Mild diarrhoea. Slight rectal bleeding Microscopic haematuria
Moderate Gross atrophy and t'g'ectasis* t'g'ectasis*, complete alopecia Diarrhoea > Obstruction or 5/day, requiring bleeding fistula surgery Frequency, Severe frequency, intermittent frequent haematuria. macroscopic Contracted haematuria bladder
Grade 4 Ulceration
Necrosis, perforation Necrosis, Contracted bladder( < 100 ml). Severe haemorrhage
*telangiectasis Any fetal toxicity is Grade 5. 794
The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder 100-1
a 3
n = 16
o
CO
u Q.
0 RTOG/EORTC late bowel effects Figure 3. Histogram showing the distribution of late bowel toxicity following accelerated fractionation.
1
2
RTOG/EORTC
3 late
urinary
4
5 toxicity
Figure 4. Histogram showing the distribution of late urinary toxicity following accelerated fractionation.
There was only one case of Grade 4 small bowel toxicity. This was the case described above who also had Grade 4 large bowel toxicity. Otherwise, there were three cases of Grade 2 toxicity, 10 Grade 1 and nine Grade 0. The influence of age, dose, dose per fraction, overall treatment time, pelvis target volume and bladder target volume on severity and duration of acute bowel and urinary toxicity was examined using the Mann-Whitney test for ordinal data, and x2 f° r comparing the two groups treated with 1.8 Gy and 2.0 Gy per fraction. No significant associations were found.
not evaluable for local control: three died within 6 months from metastatic disease, one died shortly after treatment following colectomy for perforated divericulitis and follow-up information is insufficient in another four. Seven patients had persistent or recurrent disease in the bladder and this was manifest in all cases within 12 months from the start of treatment. Nine patients had durable local control after 12-51 months from treatment. The actuarial local control at 24 months was 56%. Distant control was assessable in 21 patients. Insufficient follow-up information was available in the other three. There were eight distant failures and six of these died within a year. Late toxicity Actuarial survival in all 24 patients at 24 months was Late effects were assessable in 16 patients. The distribution of late bowel effects is shown in Fig. 3. Clinical 35%. There were 14 deaths, eight from metastatic symptoms were recorded in two patients (12% of the disease, four from local failure (two of whom also had group) and in both these cases they were Grade 1. metastatic disease), there was one toxic death, one death Figure 4 shows the distribution of late urinary effects. from other causes and in two cases the cause of death Grade 2 effects occurred in 13% and Grade 1 effects in was unknown. The probabilities associated with local 20% of patients. No late skin effects were seen. There control and survival are illustrated graphically in Fig. 5. was one patient who developed a non-invasive recurrence which was cysto-diathermied and he proceeded to Discussion The feasibility of reducing the overall time in radical cystectomy because of suspicion of local relapse with muscle invasive disease. The procedure was complicated radiotherapy by multiple daily treatments critically by severe fibrosis around the bladder, but the patho- depends on the tolerance of the associated acutely logical specimen showed no evidence of invasive responding normal tissues. This "tolerance" is not easily tumour. This patient subsequently developed lower defined and there is considerable variation among pelvic sepsis due to a chronic discharging perineal sinus patients and radiotherapists as to what degree of toxiand has undergone three surgical procedures in an city is acceptable. In this small pilot series the incidence of Grade 3 or attempt to control the problem. This late effect has been categorized as Grade 4 although clearly the morbidity greater acute bowel effects was 35%. One patient arose from a combination of radiotherapy and surgery. developed intestinal obstruction secondary to an ileus Despite these problems the patient has maintained an requiring laparotomy and defunctioning colostomy (Grade 4) and another died after colectomy for perforECOG performance status of 1. ated diverticulitis (Grade 5). These two patients had prior abdominal surgery and inflammatory bowel Outcome Local control was evaluable at 6 months or later after disease, respectively and such patients were subsethe start of treatment in 16 patients. Eight patients were quently excluded from entry to our study. Nevertheless Vol. 65, No. 777
795
D. J. Cole e( al 100 Local control Survival
0
6
12
18
Months from start of
24
30
36
radiotherapy
Figure 5. Graph of actuarial local control and survival at 24 months in 24 patients with muscle invasive carcinoma of the bladder treated with twice daily irradiation.
acute radiation reactions requiring operative intervention or those proving fatal are not acceptable unless they are very rare events. A substantial incidence of Grade 3 reactions are acceptable to some patients and radiotherapists provided there is evidence for improved local control and preferably survival. Grade 3 acute urinary effects occurred in 17% of our patients and in these cases this persisted for a mean of 15 weeks. It is often difficult to separate the symptoms attributable to persistent tumour from those due to radiation effect, but nevertheless it is likely that a significant proportion of this morbidity was radiation related. There have been other reports of serious reactions to accelerated radiotherapy in bladder cancer. In one case a female aged 77 years received 60 Gy in 30 fractions in 19 days and suffered a perforated sigmoid colon 2 months after the completion of treatment from which she died. Post-mortem confirmed that she had suffered a treatment related death. In another case, a patient developed life-threatening intestinal obstruction after receiving 44 Gy (24 Gy to the whole pelvis) in 22 fractions in 3 weeks. Spontaneous recovery occurred with conservative management. Review of the computed tomography scans showed adhesions of small bowel to the bladder (Horwich, 1991). The acute toxicity initially seen in the present study and that encountered by others led to a reduction in dose per fraction from 2.0 Gy to 1.8 Gy, although the case of Grade 4 acute bowel toxicity was treated at 1.8 Gy per fraction. This reduction in dose per fraction is supported by the results of recent animal studies which have suggested that sublethal damage is not completely repaired in acutely responding normal tissue even with a 6 hour interfraction interval (van den Aardweg & Hopewell, 1992). It has not been possible to demonstrate a reduction in acute radiation morbidity following this reduction in dose per fraction, although this may be due to the small number of patients in each treatment group. 796
It is likely that the acute bowel toxicity observed is largely explained by the inclusion of most of the true pelvis in the target volume up to 39.6-44.0 Gy in order prophylactically to treat the regional lymph nodes, although treatment to the larger volume was interrupted for 1 week whilst the bladder boost was delivered. The argument for treating the lymph nodes was based on a series of cases who received pre-operative radiation and cystectomy (Wallace and Bloom, 1976). The incidence of pathologically positive lymph nodes was lower than in a matched group of historical controls treated by surgery alone. However, no survival benefit attributable to nodal downstaging was identified. More recently, a retrospective analysis of 731 patients (Symonds et al, 1990), of whom 395 underwent irradiation to the primary and regional lymph nodes, and 336 historical controls matched for stage and total dose were treated to the bladder only, showed no survival difference. Moreover serious complications occurred in 9.6% of the group receiving nodal irradiation versus 4.3% in the group treated to the bladder only. A Phase I study of acute toxicity following a similar accelerated fractionation regimen in invasive bladder cancer found a lower incidence of acute bowel toxicity (Horwich, 1991). Maximum acute bowel toxicity was noted at the completion of treatment, at which time 30% of the group had Grade 2 effects and there were no Grade 3 effects, although one patient had had a Grade 3 effect requiring a rest from treatment for 1 week after an accumulated dose of 40 Gy. In this study the maximum bladder toxicity was also seen at the end of treatment, with 10% Grade 2 effects and 20% Grade 3 effects. The Grade 3 bladder toxicity had improved after 4, 3, 5 and 4 weeks, respectively, after onset. Most of the patients in this study were treated to a small volume encompassing the bladder only, thereby excluding a significant volume of small and large bowel. The importance of clonogen repopulation as a cause of therapeutic failure in clinical radiotherapy has still to be demonstrated in randomized controlled trial. The possible benefits of shortening the overall treatment time could be counterbalanced by recent evidence suggesting that longer overall treatment times may lead to greater cell cycle synchrony and therefore increased radiosensitivity (Hopewell & van den Aardweg, 1991). This work indicates that longer treatment times {i.e. 5-6 weeks) may be associated with more severe acute responses for skin than short courses {i.e. 2-4 weeks). The rate of clonogen doubling is in the region of 3-6 days for most invasive transitional cell tumours of the bladder (Trott & Kummermehr, 1985; Sheehan & Newman, 1991). There is evidence in normal tissues that fractionated radiation stimulates an increased rate of clonogen proliferation among surviving cells and that in pig skin this effect occurs at between 14 and 28 days after the start of radiation (Morris & Hopewell, 1986). It is not known whether a similar pattern for the kinetics of repopulation holds for tumours. A method of comparing the radiobiological effect of different fractionation regimens has been proposed The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder Table II. Extrapolated response doses for a variety of treatment schedules Regimeni
Extrapolated response doses (Gy)
Dose (Gy)
Fractions Days
Tumours
Late responding tissue
"Oxford 60 64 54
regimens' 30 40 44 32 18 44
53.5 56.5 51.7
99.6 106.2 108
"Manchester regimens" 50 20 26 55 22 30
50.5 54.9
91.6 100.8
"Accelerated regimens" 57.6 32 24 61.2 34 25 64.8 36 26
56.9 60.7 64.5
92.2 97.9 103.7
(Fowler, 1990). This method uses the linear quadratic formula to obtain an extrapolated response dose (ERD) for a given alpha/beta ratio. It incorporates a time factor which may be used for tumours but omitted for late responding tissues, and uses the potential doubling time as a measure of proliferative capacity in the tumour and is expressed as follows: ERD = Total dose (1 + d/[a/P\-0.693la(T/Tpot) (1)
probability of tumour control than a daily regimen of 64 Gy in 32 fractions. It should be emphasized that this model relies on estimates of alpha/beta ratios and potential doubling times for which there is little conclusive evidence at the present time and more data is awaited. It should be stressed that fractionation regimens should be developed in the light of previous clinical data although in the future it may be possible to tailor regimens to the kinetics of individual tumours. Based on our clinical experience and bearing in mind these theoretical considerations our next step is to investigate a twice daily regimen over 25-26 days at 1.8 Gy per fraction to a total dose of 61.2—64.8 Gy. The patients will not be treated over the weekend. We will treat the bladder only and there will be 6 h between the twice daily fractions. If this regimen proves to be adequately tolerated, then it is hoped to test its efficacy compared to conventional radiation schedules in a randomized fn'a/. \n summary, l\\e \oca\ corvlro\ and survwa\ \n \h\s small pilot series are at least comparable to other published series and this provides a strong stimulus to continuing the investigation of twice daily fractionation schedules in bladder cancer. References BLOOM, H. J. G., HENDRY, W. F., WALLACE, D. M. & SKEET,
R. G., 1982. Treatment of T3 bladder cancer: controlled trial of pre-operative radiotherapy and radical cystectomy versus radical radiotherapy. British Journal of Urology, 54, 136-151. DUNCAN & QUILTY, P. M., 1986. The results of a series of 963
The following assumptions are made: a — 0.3 (i.e. 49% survival at 2 Gy); potential doubling time (rpot = 5 days); constant proliferation throughout 39 days; d — dose per fraction; a/j5 = alpha/beta ratio for the tissue in question; T = overall time. Using this method, different fractionation regimens may be compared. The extrapolated response dose is expressed in grays for a given alpha/beta ratio (e.g. 10 Gy for tumours and 3 Gy for late responding tissue). The extrapolated response doses for a variety of treatment schedules are shown in Table II. This model predicts that the regimen previously used in Oxford (54 Gy in 18 fractions in 44 days) is associated with the greatest chance of late toxicity, closely followed by the conventional daily regimen of 64 Gy in 32 fractions in 44 days. The accelerated regimens are predicted to have lower late toxicity. The extrapolated dose-response values for tumours are substantially higher for the accelerated regimens than the conventional "Oxford" regimens for a given extrapolated dose-response value for late responding tissue, thus predicting a higher tumour control probability for acceleration. It is also notable that the schedule used in the second half of our study (57.6 Gy in 32 fractions in 24 days) has an extrapolated response dose for tumours that is almost indentical to the conventional daily regimen of 64 Gy in 32 fractions in 44 days. According to this model, therefore, 57.6 Gy in 32 fractions in 24 days does not offer a higher Vol. 65, No. 777
patients with transitional cell carcinoma of the urinary bladder primarily treated by radical megavoltage X-ray therapy. Radiotherapy and Oncology, 7, 299-310. FOWLER, J. F., 1986. Potential for increasing the differential response between tumours and normal tissues: can proliferation be used? International Journal of Radiation Oncology, Biology and Physics, 12, 641-645. FOWLER, J. F., 1990. Fractionation in radiotherapy. Proceedings of 32nd Annual ASTRO Meeting. International Journal of Radiation Oncology, Biology and Physics, 19, Suppl. 1, p. 113. HOPEWELL, J. W . & VAN DEN AARDWEG, G . J. M . J., 1988.
Current concepts of dose fractionation in radiotherapy. Normal tissue tolerance. British Jounral of Radiology (Suppl), 22, 88-94. HOPEWELL, J. W . & VAN DEN AARDWEG, G . J. M . J., 1989.
Kinetics of recovery from sublethal radiation damage. British Journal of Radiology, 62, 1009-1010. HOPEWELL, J. W . & VAN DEN AARDWEG, G . J. M . J., 1991.
Studies of dose-fractionation on early and late responses in pig skin: a reappraisal of the importance of overall treatment time and its effects on radiosensitization and incomplete repair. International Journal of Radiation Oncology, Biology and Physics, 21, 1441-1450. HORWICH, A., 1991. Treatment of muscle invasive (T2/T3) disease: the role of radiotherapy. In Urological Oncology Dilemmas and Developments, ed. by A. R. Alderson, R. T. D. Oliver, I. W. F. Hanham and H. J. G. Bloom (John Wiley & Sons, Chichester), chap. 13, pp. 137-144. MORRIS, G. M. & HOPEWELL, J. W., 1986. Changes in the cell
kinetics of pig epidermis after repeated daily doses of X-rays.
797
D. J. Cole et al In Radiation Damage to Skin, Supplement to British Journal of Radiology (BIR, London), pp. 34-38.
fractionation and hyperfraction? Radiotherapy and Oncology, 3, 1-9.
QUILTY, P. M , DUNCAN, W. & KERR, G. R., 1985. Results of a
TURESSON, I. & THAMES, H. D., 1989. Repair capacity and
randomised study to evaluate influence of dose on morbidity in radiotherapy for bladder cancer. Clinical Radiology, 36, 615-618. SHEARER, R. J., CHILVERS, C. E. D., BLOOM, H. J. G., BLISS, J. M., HORWICH, A. & BABIKER, A. (for THE CO-OPERATIVE
UROLOGICAL CANCER GROUP, 1988). A prospective trial:
preliminary report. British Journal of Urology, 62, 558-564. SHEEHAN, T. & NEWMAN, H., 1991. Proliferation characteristics
of human bladder tumours in vivo determined by bromodeoxyuridine incorporation. British Journal of Cancer, 64, Suppl. XV, 6 (abstract). SYMONDS, R. P., DAVIDSON, R. P., HABESHAW, T., ROBERTSON, A. G., SNEE, M. P. & WATSON, E. R., 1990. The significance
of dose, nodal irradiation and other factors upon the outcome after radiotherapy for bladder cancer (abstract). British Journal of Cancer, 61, 167. TROTT, K.-R. & KUMMERMEHR, J., 1985. What is known about proliferation rates to choose between accelerated
798
kinetics of human skin during fractionated radiotherapy: erythema, desquamation and telangiectasia after 3 and 5 years follow-up. Radiotherapy and Oncology, 15, 169-188. VAN DEN AARDWEG, G. J. M. J. & HOPEWELL, J. W., 1992.
The kinetics of repaif for radiation induced sublethal damage in pig epidermis: an interpretation based on fast and slow components of repair. Radiotherapy and Oncology, 23, 94-104. WALLACE, D. M. & BLOOM, H. J. G., 1976. The management
of deeply infiltrating (T3) bladder carcinoma; controlled trial of radical radiotherapy versus pre-operative radiotherapy and radical cystectomy (first report). British Journal of Urology, 48, 587-594. WALLACE, D. M. A., RAGHAVAN, D., KELLY, K. A., SANDEMAN, T. F., CONN, I. G., TERIANA, N., DUNN, J.,
BOULAS, J. & LATIEF, T., 1991. Neo-adjuvant (pre-emptive)
cisplatin therapy in invasive transitional cell carcinoma of the bladder. British Journal of Urology, 67, 608-615.
The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder By D. J. Cole, K. R. Durrant, *J. T. Roberts, *P. J. T. K. Dawes, tH. Yosef and §J. W. Hopewell Department of Radiotherapy and Oncology, Churchill Hospital, Oxford, "Regional Radiotherapy Centre, Newcastle General Hospital, Newcastle upon Tyne, tBeatson Oncology Centre, Belvidere Hospital, Glasgow and §C.R.C. Normal Tissue Research Group, Research Institute (University of Oxford), Churchill Hospital, Oxford, UK
{Received 14 October 1991 and in revised form 24 February 1992, accepted 26 March 1992) Keywords: Bladder cancer, Radiotherapy, Accelerated fractionation
Abstract. 24 patients with muscle invasive carcinoma of the bladder were treated in a pilot study of twice daily fractionation at radiation doses of 1.8-2.0 Gy per fraction to total doses of 54-64 Gy to the bladder and 39.6-44 Gy to the whole pelvis. The treatment aim was to give 32 fractions in 22 days. The interfraction interval was a minimum of 6 h. The principle objective was to record acute and late tolerance, but local control and survival data is also presented. Acute radiation morbidity was scored according to the RTOG system. Grade 2 large bowel effects were seen in 52% of patients, Grade 3 effects in 26% and there was one Grade 4 and one Grade 5 effect. The mean duration of effect was 4.5 weeks although the more severe reactions were also more protracted. Grade 2 urinary effects occurred in 30% and Grade 3 in 17% of patients. The mean duration of effect was 7.2 weeks. There were no Grade 4 or 5 acute urinary effects. Late radiation morbidity was scored according to the EORTC/RTOG system and was assessable in 16 cases who survived more than 6 months. There were two cases (12%) of Grade 1 bowel toxicity, two cases of Grade 1 and three of Grade 2 urinary toxicity. There were no cases of late skin effects. Actuarial analysis at 2 years shows a local control probability of 56% and survival probability of 35%.
The 5-year survival of patients with muscle invasive transitional cell carcinoma of the bladder treated by radical radiotherapy or combined pre-operative radiotherapy and elective radical cystectomy is 40% for T2 tumours (Duncan & Quilty, 1986) and 29-38% for T3 tumours (Bloom et al, 1982). Following radical radiotherapy, complete response in the bladder occurs in up to 56%, but one third of these responders subsequently fail locally (Shearer et al, 1988). Survival in patients treated by irradiation has not been improved by adjuvant methotrexate (Shearer et al, 1988) or cisplatin (Wallace et al, 1991) chemotherapy. The short latency times to local recurrence after radiotherapy in invasive transitional cell carcinoma of the bladder indicate that the disease is biologically aggressive (Duncan & Quilty, 1986). Moderate or poor histological differentiation is reported in 66% of T2 and 89% of T3 tumours (Duncan & Quilty, 1986), Cell kinetic studies demonstrate rapid proliferation of tumour clonogens with a potential doubling time of 3 days (Trott & Kummermehf, 1985; Sheehan & Newman, 1991). It has been demonstrated that acutely responding normal tissue undergoes accelerated clonogen repopulation starting at between 14 and 28 days following the start of a conventionally fractionated course of irradiation (Morris & Hopewell, 1986). It has also been suggested that rapidly proliferating tumours undergo considerable clonogen repopulation during a lengthy course of irradiation (Trott & Kummermehr, 1985) and that this phenomenon is an important cause 792
of local failure in treatments involving conventional fractionation schedules (Fowler, 1986). A reduction in the overall treatment time for radiotherapy may reduce the capacity for tumour clonogens to repopulate and thereby improve local control in invasive bladder cancer. However, this cannot be achieved by an increase in the dose per fraction if late responding tissues are to be spared injury (Quilty et al, 1985) and therefore multiple fractions per day are necessary if the required dose is to be delivered in a shorter time. The fractions of radiation must be given at least 6 h apart to allow for the repair of sublethal damage. However, even with this interfraction interval, repair may not be complete (Hopewell and van den Aardweg, 1988, 1989; Turesson and Thames, 1989) and a small decrease in dose per fraction may be required to compensate for this. This paper presents the results of a pilot study of accelerated fractionation in invasive bladder cancer which was carried out to investigate the potential toxicity of such a schedule. All patients have had a minimum of 12 months follow-up. Materials and methods
Patients with invasive carcinoma of the bladder with no evidence of nodal or distant metastasis who were suitable for radical irradiation with curative intent were selected for this study. Eastern Co-operative Group (ECOG) performance status 0-2 and normal (or only slightly elevated) serum creatinine were required for The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder
entry, but patients over the age of 75 years were generally excluded. Later in the study, those with a history of inflammatory bowel disease, previous lower abdominal or pelvic surgery were excluded. Patients underwent cystoscopy, biopsy and usually endoscopic resection of tumour, and bimanual examination of the pelvis under anaesthetic. Haemoglobin, serum creatinine and liver function tests were measured. Computed tomography of the abdomen and pelvis was performed to exclude nodal and metastatic disease. 24 patients were entered into this pilot study. There were 18 males and six females. The mean ( + SD) age at entry was 66.4 + 6.3 years. There were 21 cases of transitional cell carcinoma (TCC), two cases of squamous carcinoma (SCC) and in one case the histological type was not stated. There were six T2, 14 T3 and four T4 cases. In 19 cases the grade was reported, three were Grade 2 (moderately differentiated) and 16 were Grade 3 (poorly differentiated). Radiotherapy was planned using computed tomography localization and treatment was given on linear accelerator machines of energy 6-8 MV. The whole pelvis was treated during the first week of treatment (10 fractions), the bladder alone was treated during the second week (10 fractions) and treatment to the whole pelvis was completed during the third or early fourth week in 10-12 fractions. The pelvic volumes were treated with three or four field techniques in all but three cases in which an anterior and posterior pair of fields were used. The bladder volumes were all treated with three or four field plans. Two treatments were given each day, 5 days per week and these treatments were given not less than 6 h apart. The intention was to deliver a radical dose of radiation in 30-32 fractions in 15-16 treatment days (treating 5 days per week) with an overall time of 19-22 days. Three centres contributed patients to the study and differences in local practice accounted for the slight differences in dose prescription.
Of the 24 patients entered into the study, the first 13 were treated at 2.0 Gy per fraction and the remaining 11 received 1.8 Gy per fraction. The dose per fraction (and consequently the total dose) was reduced because of concern regarding acute toxicity. The mean (+ SD) dose delivered was 58.85 + 2.2 Gy with a range of 54-64 Gy. The mean number of fractions was 30.8 ±1.1 (range 29-32). The mean overall time was 22.7 + 5.3 days (range 19-44). A protracted skin reaction led to an overall treatment time of 44 days in one patient. Otherwise the maximum treatment time was 29 days. The mean target volume for the whole pelvis was 3030 + 961 cm3 (range 1755-5486) and for the bladder 925 + 313 cm3 (range 427-1595). Response was scored at follow-up cystoscopy which was performed at 6-monthly intervals after radiotherapy. Patients with local failure were considered for salvage cystectomy. Normal tissue toxicity was recorded using the RTOG criteria for early effects and RTOG/EORTC criteria for late effects (see Tables I(a) and (b)). Early effects were scored weekly during treatment and then at 4, 6 and 12 weeks after starting treatment. Late effects were scored at 6 months and then annually. Results
Acute toxicity Acute large bowel toxicity was assessable in 23 patients and is shown in a histogram in Fig. 1. It can be seen that the majority of patients (78%) encountered Grade 2 toxicity (diarrhoea requiring medication) or Grade 3 toxicity (requiring hospital admission or intravenousfluids).The mean duration (in weeks) of toxicity for each grade is stated on the histogram. There was one case of Grade 4 large bowel toxicity. This patient completed irradiation to a dose of 57.6 Gy in 32 fractions in 24 days without problems and then 1 week later developed severe diarrhoea, anorexia,
Table I. (a) Radiaton Therapy and Oncology Group (RTOG) acute radiation morbidity scoring criteria Tissue
Grade 0
Grade 1
Grade 2
Grade 3
Grade 4
Skin
No change
Erythema, dry desquamation
Confluent moist desquamation
Small bowel
No change
Nausea
Requiring i.v. fluids
Ulceration, haemorrhage and necrosis Requiring surgery
Large bowel
No change
Loose stool
Painful erythema, patchy moist desquamation Nausea/vomiting requiring antiemetics Diarrhoea requiring medication
Genitourinary
No change
Frequency x 2 baseline
Frequency > hourly, dysuria
Frequency < hourly, dysuria requiring opiate
Requiring i.v. fluids
Obstruction fistula, perforation requiring surgery Haematuria requiring transfusion, ulceration, necrosis
i.v. = intravenous. Vol. ^65, No. 777
793
D. J. Cole ef*al 100 90
a3 0 O)
o © TO CO
a> oh. a>
Q.
0
1
2
3
4
5
70 60 50 40 30 20 10 0
vomiting and weight loss. She was admitted to hospital with intestinal obstruction and at laparotomy found to have an ileus. A defunctioning colostomy was performed and she made a slow but complete recovery over 8 weeks. 18 months later she developed abdominal carcinomatosis and underwent another laparotomy at which time no sequelae attributable to the radiation were found. This patient had a history of appendectomy followed by ureteric surgery for reflux uropathy about 30 years previously. There was one case of Grade 5 large bowel toxicity. In this case there was an ileo-vesical fistula, probably due to malignancy prior to the start of radiation. Treatment was completed to 60 Gy in 30 fractions in 27 days and 2 weeks later total colectomy was performed for perforated diverticulitis and peritonitis from which the patient succumbed within a month. The acute urinary toxicity is shown in Fig. 2. It is notable that Grade 3 toxicity (frequency and nocturia
3
6
1mi
15 wks
0 1 2 3 4 5 RTOG grade acute urinary toxicity
RTOG acute large bowel toxicity Figure 1. Histogram showing the distribution of acute large bowel toxicity following accelerated fractionation. The number above the columns indicate the mean time to resolution of symptoms after completion of therapy.
n=23
80
Figure 2. Histogram showing the distribution of acute urinary toxicity following accelerated fractionation. The numbers above the columns indicate the mean time to resolution of symptoms after completion of radiotherapy.
< hourly and/or pain requiring strong analgesics) was unusual (17% of cases), but that it was relatively prolonged with a mean duration of 15 weeks. In one of these cases, severe dysuria and frequency persisted for 9 months but was eventually relieved with an antispasmodic agent. The mean duration of acute urinary effects was 5.4 weeks compared to 3.4 weeks for acute large bowel effects and this difference was significant (p = 0.045, Mest). There was one case of Grade 3 skin reaction (confluent moist desquamation). An interruption in treatment was required and the final dose was 58 Gy in 29 fractions in 44 days. Otherwise there were three cases of Grade 2 toxicity and' seven cases of Grade 1 toxicity. In all cases the acute skin reaction had healed 3 weeks after the completion of treatment. The remaining 12 cases showed no toxicity.
Table I. (b) RTOG/European Organisation for Research into the Treatment of Cancer (EORTC) late radiation morbidity scoring scheme Tissue
Grade 0
Grade 1
Grade 2
Grade 3
Skin
None
Mild atrophy, mild alopecia, ± pigmentation
Bowel
None
Genitourinary
None
Mild diarrhoea. Slight rectal bleeding Microscopic haematuria
Moderate Gross atrophy and t'g'ectasis* t'g'ectasis*, complete alopecia Diarrhoea > Obstruction or 5/day, requiring bleeding fistula surgery Frequency, Severe frequency, intermittent frequent haematuria. macroscopic Contracted haematuria bladder
Grade 4 Ulceration
Necrosis, perforation Necrosis, Contracted bladder( < 100 ml). Severe haemorrhage
*telangiectasis Any fetal toxicity is Grade 5. 794
The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder 100-1
a 3
n = 16
o
CO
u Q.
0 RTOG/EORTC late bowel effects Figure 3. Histogram showing the distribution of late bowel toxicity following accelerated fractionation.
1
2
RTOG/EORTC
3 late
urinary
4
5 toxicity
Figure 4. Histogram showing the distribution of late urinary toxicity following accelerated fractionation.
There was only one case of Grade 4 small bowel toxicity. This was the case described above who also had Grade 4 large bowel toxicity. Otherwise, there were three cases of Grade 2 toxicity, 10 Grade 1 and nine Grade 0. The influence of age, dose, dose per fraction, overall treatment time, pelvis target volume and bladder target volume on severity and duration of acute bowel and urinary toxicity was examined using the Mann-Whitney test for ordinal data, and x2 f° r comparing the two groups treated with 1.8 Gy and 2.0 Gy per fraction. No significant associations were found.
not evaluable for local control: three died within 6 months from metastatic disease, one died shortly after treatment following colectomy for perforated divericulitis and follow-up information is insufficient in another four. Seven patients had persistent or recurrent disease in the bladder and this was manifest in all cases within 12 months from the start of treatment. Nine patients had durable local control after 12-51 months from treatment. The actuarial local control at 24 months was 56%. Distant control was assessable in 21 patients. Insufficient follow-up information was available in the other three. There were eight distant failures and six of these died within a year. Late toxicity Actuarial survival in all 24 patients at 24 months was Late effects were assessable in 16 patients. The distribution of late bowel effects is shown in Fig. 3. Clinical 35%. There were 14 deaths, eight from metastatic symptoms were recorded in two patients (12% of the disease, four from local failure (two of whom also had group) and in both these cases they were Grade 1. metastatic disease), there was one toxic death, one death Figure 4 shows the distribution of late urinary effects. from other causes and in two cases the cause of death Grade 2 effects occurred in 13% and Grade 1 effects in was unknown. The probabilities associated with local 20% of patients. No late skin effects were seen. There control and survival are illustrated graphically in Fig. 5. was one patient who developed a non-invasive recurrence which was cysto-diathermied and he proceeded to Discussion The feasibility of reducing the overall time in radical cystectomy because of suspicion of local relapse with muscle invasive disease. The procedure was complicated radiotherapy by multiple daily treatments critically by severe fibrosis around the bladder, but the patho- depends on the tolerance of the associated acutely logical specimen showed no evidence of invasive responding normal tissues. This "tolerance" is not easily tumour. This patient subsequently developed lower defined and there is considerable variation among pelvic sepsis due to a chronic discharging perineal sinus patients and radiotherapists as to what degree of toxiand has undergone three surgical procedures in an city is acceptable. In this small pilot series the incidence of Grade 3 or attempt to control the problem. This late effect has been categorized as Grade 4 although clearly the morbidity greater acute bowel effects was 35%. One patient arose from a combination of radiotherapy and surgery. developed intestinal obstruction secondary to an ileus Despite these problems the patient has maintained an requiring laparotomy and defunctioning colostomy (Grade 4) and another died after colectomy for perforECOG performance status of 1. ated diverticulitis (Grade 5). These two patients had prior abdominal surgery and inflammatory bowel Outcome Local control was evaluable at 6 months or later after disease, respectively and such patients were subsethe start of treatment in 16 patients. Eight patients were quently excluded from entry to our study. Nevertheless Vol. 65, No. 777
795
D. J. Cole e( al 100 Local control Survival
0
6
12
18
Months from start of
24
30
36
radiotherapy
Figure 5. Graph of actuarial local control and survival at 24 months in 24 patients with muscle invasive carcinoma of the bladder treated with twice daily irradiation.
acute radiation reactions requiring operative intervention or those proving fatal are not acceptable unless they are very rare events. A substantial incidence of Grade 3 reactions are acceptable to some patients and radiotherapists provided there is evidence for improved local control and preferably survival. Grade 3 acute urinary effects occurred in 17% of our patients and in these cases this persisted for a mean of 15 weeks. It is often difficult to separate the symptoms attributable to persistent tumour from those due to radiation effect, but nevertheless it is likely that a significant proportion of this morbidity was radiation related. There have been other reports of serious reactions to accelerated radiotherapy in bladder cancer. In one case a female aged 77 years received 60 Gy in 30 fractions in 19 days and suffered a perforated sigmoid colon 2 months after the completion of treatment from which she died. Post-mortem confirmed that she had suffered a treatment related death. In another case, a patient developed life-threatening intestinal obstruction after receiving 44 Gy (24 Gy to the whole pelvis) in 22 fractions in 3 weeks. Spontaneous recovery occurred with conservative management. Review of the computed tomography scans showed adhesions of small bowel to the bladder (Horwich, 1991). The acute toxicity initially seen in the present study and that encountered by others led to a reduction in dose per fraction from 2.0 Gy to 1.8 Gy, although the case of Grade 4 acute bowel toxicity was treated at 1.8 Gy per fraction. This reduction in dose per fraction is supported by the results of recent animal studies which have suggested that sublethal damage is not completely repaired in acutely responding normal tissue even with a 6 hour interfraction interval (van den Aardweg & Hopewell, 1992). It has not been possible to demonstrate a reduction in acute radiation morbidity following this reduction in dose per fraction, although this may be due to the small number of patients in each treatment group. 796
It is likely that the acute bowel toxicity observed is largely explained by the inclusion of most of the true pelvis in the target volume up to 39.6-44.0 Gy in order prophylactically to treat the regional lymph nodes, although treatment to the larger volume was interrupted for 1 week whilst the bladder boost was delivered. The argument for treating the lymph nodes was based on a series of cases who received pre-operative radiation and cystectomy (Wallace and Bloom, 1976). The incidence of pathologically positive lymph nodes was lower than in a matched group of historical controls treated by surgery alone. However, no survival benefit attributable to nodal downstaging was identified. More recently, a retrospective analysis of 731 patients (Symonds et al, 1990), of whom 395 underwent irradiation to the primary and regional lymph nodes, and 336 historical controls matched for stage and total dose were treated to the bladder only, showed no survival difference. Moreover serious complications occurred in 9.6% of the group receiving nodal irradiation versus 4.3% in the group treated to the bladder only. A Phase I study of acute toxicity following a similar accelerated fractionation regimen in invasive bladder cancer found a lower incidence of acute bowel toxicity (Horwich, 1991). Maximum acute bowel toxicity was noted at the completion of treatment, at which time 30% of the group had Grade 2 effects and there were no Grade 3 effects, although one patient had had a Grade 3 effect requiring a rest from treatment for 1 week after an accumulated dose of 40 Gy. In this study the maximum bladder toxicity was also seen at the end of treatment, with 10% Grade 2 effects and 20% Grade 3 effects. The Grade 3 bladder toxicity had improved after 4, 3, 5 and 4 weeks, respectively, after onset. Most of the patients in this study were treated to a small volume encompassing the bladder only, thereby excluding a significant volume of small and large bowel. The importance of clonogen repopulation as a cause of therapeutic failure in clinical radiotherapy has still to be demonstrated in randomized controlled trial. The possible benefits of shortening the overall treatment time could be counterbalanced by recent evidence suggesting that longer overall treatment times may lead to greater cell cycle synchrony and therefore increased radiosensitivity (Hopewell & van den Aardweg, 1991). This work indicates that longer treatment times {i.e. 5-6 weeks) may be associated with more severe acute responses for skin than short courses {i.e. 2-4 weeks). The rate of clonogen doubling is in the region of 3-6 days for most invasive transitional cell tumours of the bladder (Trott & Kummermehr, 1985; Sheehan & Newman, 1991). There is evidence in normal tissues that fractionated radiation stimulates an increased rate of clonogen proliferation among surviving cells and that in pig skin this effect occurs at between 14 and 28 days after the start of radiation (Morris & Hopewell, 1986). It is not known whether a similar pattern for the kinetics of repopulation holds for tumours. A method of comparing the radiobiological effect of different fractionation regimens has been proposed The British Journal of Radiology, September 1992
A pilot study of accelerated fractionation in the radiotherapy of invasive carcinoma of the bladder Table II. Extrapolated response doses for a variety of treatment schedules Regimeni
Extrapolated response doses (Gy)
Dose (Gy)
Fractions Days
Tumours
Late responding tissue
"Oxford 60 64 54
regimens' 30 40 44 32 18 44
53.5 56.5 51.7
99.6 106.2 108
"Manchester regimens" 50 20 26 55 22 30
50.5 54.9
91.6 100.8
"Accelerated regimens" 57.6 32 24 61.2 34 25 64.8 36 26
56.9 60.7 64.5
92.2 97.9 103.7
(Fowler, 1990). This method uses the linear quadratic formula to obtain an extrapolated response dose (ERD) for a given alpha/beta ratio. It incorporates a time factor which may be used for tumours but omitted for late responding tissues, and uses the potential doubling time as a measure of proliferative capacity in the tumour and is expressed as follows: ERD = Total dose (1 + d/[a/P\-0.693la(T/Tpot) (1)
probability of tumour control than a daily regimen of 64 Gy in 32 fractions. It should be emphasized that this model relies on estimates of alpha/beta ratios and potential doubling times for which there is little conclusive evidence at the present time and more data is awaited. It should be stressed that fractionation regimens should be developed in the light of previous clinical data although in the future it may be possible to tailor regimens to the kinetics of individual tumours. Based on our clinical experience and bearing in mind these theoretical considerations our next step is to investigate a twice daily regimen over 25-26 days at 1.8 Gy per fraction to a total dose of 61.2—64.8 Gy. The patients will not be treated over the weekend. We will treat the bladder only and there will be 6 h between the twice daily fractions. If this regimen proves to be adequately tolerated, then it is hoped to test its efficacy compared to conventional radiation schedules in a randomized fn'a/. \n summary, l\\e \oca\ corvlro\ and survwa\ \n \h\s small pilot series are at least comparable to other published series and this provides a strong stimulus to continuing the investigation of twice daily fractionation schedules in bladder cancer. References BLOOM, H. J. G., HENDRY, W. F., WALLACE, D. M. & SKEET,
R. G., 1982. Treatment of T3 bladder cancer: controlled trial of pre-operative radiotherapy and radical cystectomy versus radical radiotherapy. British Journal of Urology, 54, 136-151. DUNCAN & QUILTY, P. M., 1986. The results of a series of 963
The following assumptions are made: a — 0.3 (i.e. 49% survival at 2 Gy); potential doubling time (rpot = 5 days); constant proliferation throughout 39 days; d — dose per fraction; a/j5 = alpha/beta ratio for the tissue in question; T = overall time. Using this method, different fractionation regimens may be compared. The extrapolated response dose is expressed in grays for a given alpha/beta ratio (e.g. 10 Gy for tumours and 3 Gy for late responding tissue). The extrapolated response doses for a variety of treatment schedules are shown in Table II. This model predicts that the regimen previously used in Oxford (54 Gy in 18 fractions in 44 days) is associated with the greatest chance of late toxicity, closely followed by the conventional daily regimen of 64 Gy in 32 fractions in 44 days. The accelerated regimens are predicted to have lower late toxicity. The extrapolated dose-response values for tumours are substantially higher for the accelerated regimens than the conventional "Oxford" regimens for a given extrapolated dose-response value for late responding tissue, thus predicting a higher tumour control probability for acceleration. It is also notable that the schedule used in the second half of our study (57.6 Gy in 32 fractions in 24 days) has an extrapolated response dose for tumours that is almost indentical to the conventional daily regimen of 64 Gy in 32 fractions in 44 days. According to this model, therefore, 57.6 Gy in 32 fractions in 24 days does not offer a higher Vol. 65, No. 777
patients with transitional cell carcinoma of the urinary bladder primarily treated by radical megavoltage X-ray therapy. Radiotherapy and Oncology, 7, 299-310. FOWLER, J. F., 1986. Potential for increasing the differential response between tumours and normal tissues: can proliferation be used? International Journal of Radiation Oncology, Biology and Physics, 12, 641-645. FOWLER, J. F., 1990. Fractionation in radiotherapy. Proceedings of 32nd Annual ASTRO Meeting. International Journal of Radiation Oncology, Biology and Physics, 19, Suppl. 1, p. 113. HOPEWELL, J. W . & VAN DEN AARDWEG, G . J. M . J., 1988.
Current concepts of dose fractionation in radiotherapy. Normal tissue tolerance. British Jounral of Radiology (Suppl), 22, 88-94. HOPEWELL, J. W . & VAN DEN AARDWEG, G . J. M . J., 1989.
Kinetics of recovery from sublethal radiation damage. British Journal of Radiology, 62, 1009-1010. HOPEWELL, J. W . & VAN DEN AARDWEG, G . J. M . J., 1991.
Studies of dose-fractionation on early and late responses in pig skin: a reappraisal of the importance of overall treatment time and its effects on radiosensitization and incomplete repair. International Journal of Radiation Oncology, Biology and Physics, 21, 1441-1450. HORWICH, A., 1991. Treatment of muscle invasive (T2/T3) disease: the role of radiotherapy. In Urological Oncology Dilemmas and Developments, ed. by A. R. Alderson, R. T. D. Oliver, I. W. F. Hanham and H. J. G. Bloom (John Wiley & Sons, Chichester), chap. 13, pp. 137-144. MORRIS, G. M. & HOPEWELL, J. W., 1986. Changes in the cell
kinetics of pig epidermis after repeated daily doses of X-rays.
797
D. J. Cole et al In Radiation Damage to Skin, Supplement to British Journal of Radiology (BIR, London), pp. 34-38.
fractionation and hyperfraction? Radiotherapy and Oncology, 3, 1-9.
QUILTY, P. M , DUNCAN, W. & KERR, G. R., 1985. Results of a
TURESSON, I. & THAMES, H. D., 1989. Repair capacity and
randomised study to evaluate influence of dose on morbidity in radiotherapy for bladder cancer. Clinical Radiology, 36, 615-618. SHEARER, R. J., CHILVERS, C. E. D., BLOOM, H. J. G., BLISS, J. M., HORWICH, A. & BABIKER, A. (for THE CO-OPERATIVE
UROLOGICAL CANCER GROUP, 1988). A prospective trial:
preliminary report. British Journal of Urology, 62, 558-564. SHEEHAN, T. & NEWMAN, H., 1991. Proliferation characteristics
of human bladder tumours in vivo determined by bromodeoxyuridine incorporation. British Journal of Cancer, 64, Suppl. XV, 6 (abstract). SYMONDS, R. P., DAVIDSON, R. P., HABESHAW, T., ROBERTSON, A. G., SNEE, M. P. & WATSON, E. R., 1990. The significance
of dose, nodal irradiation and other factors upon the outcome after radiotherapy for bladder cancer (abstract). British Journal of Cancer, 61, 167. TROTT, K.-R. & KUMMERMEHR, J., 1985. What is known about proliferation rates to choose between accelerated
798
kinetics of human skin during fractionated radiotherapy: erythema, desquamation and telangiectasia after 3 and 5 years follow-up. Radiotherapy and Oncology, 15, 169-188. VAN DEN AARDWEG, G. J. M. J. & HOPEWELL, J. W., 1992.
The kinetics of repaif for radiation induced sublethal damage in pig epidermis: an interpretation based on fast and slow components of repair. Radiotherapy and Oncology, 23, 94-104. WALLACE, D. M. & BLOOM, H. J. G., 1976. The management
of deeply infiltrating (T3) bladder carcinoma; controlled trial of radical radiotherapy versus pre-operative radiotherapy and radical cystectomy (first report). British Journal of Urology, 48, 587-594. WALLACE, D. M. A., RAGHAVAN, D., KELLY, K. A., SANDEMAN, T. F., CONN, I. G., TERIANA, N., DUNN, J.,
BOULAS, J. & LATIEF, T., 1991. Neo-adjuvant (pre-emptive)
cisplatin therapy in invasive transitional cell carcinoma of the bladder. British Journal of Urology, 67, 608-615.
The British Journal of Radiology, September 1992
