Radiotherapeutic strategies for soft tissue sarcomas in adults]

Soft tissue sarcomas account for under 1% of all cancers in adults. Most soft tissue tumours are benign, only about 1% being characterized clinically and histomorphologically as malignant neoplasms. Since these tumours are often treated by excisional biopsies without any further diagnostic imaging, precise planning of postoperative irradiation therapy is often difficult to achieve. Therefore, all soft tissue tumours with a short history should be regarded as malignant until the contrary has been proven. In general, the mode of tumour resection should be postponed until CT or MR diagnosis is complete and a biopsy of the tumour has been carried out. Whereas benign lesions need only marginal tumour resection, malignant lesions require considerable safety margins. In soft tissue sarcomas of the extremities, which account for 50-60% of all sarcomas, a wide local resection followed by postoperative irradiation with about 66 Gy can guarantee local control rates above 80% and preserve the function of the limb. Radical surgery alone can achieve the same local control, but without the high level of functional integrity. In the case of marginally resectable tumours, preoperative irradiation can induce partial tumour remission and thus allow definitive limb-sparing tumour resection. Recently, multimodal and neoadjuvant therapeutic strategies have been developed. The efficacy of these experimental strategies is not yet proven.     

Pion irradiation at Paul Scherrer Institute. Results of dynamic treatment of unresectable soft tissue sarcoma

Since November 1981, when pion irradiation was introduced for deep seated tumors at the Swiss Institute for Nuclear Research (SIN; now Paul Scherrer Institute, PSI) a dynamic, three-dimensional spot scan application technique has been in use. To exploit this technique a special planning system for optimisation of the dose distribution has been designed. From November 1981 to December 1988 a total of 406 patients have been treated with pions. From April 1983 to October 1987 a total of 35 patients were prospectively treated for unresectable soft tissue sarcomas in a phase I/II-study. In 32/35 patients, tumor sites were retroperitoneal, pelvic or in the groin or thigh. 27 patients received a high, curative total dose of 30 to 36 Gy. After a median follow-up time of 19 months (13 to 68) the actuarial five-year rate of local tumor control for these 27 patients was 64%; the actuarial five-year survival rate of the 20 patients treated without metastases was 58%. Late reactions appeared in 5/27 patients: 2/8 patients with extremity/groin sarcomas (1/2 caused by biopsy) and 3/19 patients with retroperitoneal/pelvic sarcomas (one a skin reaction after Actinomycin-D, one a small bowel reaction after 36 Gy, a dose no longer given). Dynamic spot scan pion irradiation proves to be a successful treatment technique for unresectable sarcomas with a high rate of tumor control and a very low rate of severe late reactions.     

Carcinoma of the colon: irradiation by delayed split whole-abdominal technique

Locally advanced and recurrent colon cancer was treated by irradiating first the pelvis and two hours later the upper abdomen. Curative treatment consisted of 4,000-5,000 rad (40-50 Gy) in 200-rad (2-Gy) fractions to the pelvis and 3,000 rad (30 Gy) in 150-rad (1.5-Gy) fractions to the upper abdomen. Palliative treatment consisted of 2,100 rad (21 Gy) in 300-rad (3-Gy) fractions to the liver, 3,000 rad (30 Gy) in 150-rad (1.5-Gy) fractions to the upper abdomen, and 4,000-5,000 rad (40-50 Gy) in 200-rad (2-Gy) fractions to the pelvis. Treatment was tolerated well, and acute toxicity was limited. Seven of the 11 patients treated curatively remain free of abdominal disease after 10-35 months; median survival among 9 patients treated palliatively was 9 months.     

Magnetic resonance imaging of soft tissue expanders used in the management of musculoskeletal sarcomas

A soft tissue expander is surgically inserted into the body to displace radiosensitive organs from the treatment field in a small number of patients receiving radiotherapy for musculoskeletal sarcoma. MRI is routinely used to monitor the response to the radiotherapy, local recurrence and complications of treatment. This study retrospectively reviews MRI of soft tissue expanders in seven patients with musculoskeletal sarcomas; six arising in the pelvis and one in the retroperitoneum. In the absence of an appropriate clinical history, the soft tissue expander may be mistaken for a pathological fluid collection such as abscess, post-operative seroma or even recurrent tumour. MRI of the soft tissue expanders and potential errors in image interpretation are illustrated.     

Radiotherapy of carcinoma of the glottis in the initial stage

The control of early glottic cancer is equally satisfactory with either surgical resection or radiation therapy; this last method gives the patient good functional results. During the period from 1/1978 to 12/1985, 73 patients with early glottic carcinoma (T1 N0 M0) were treated in the Institute of Radiotherapy L. Galvani, University of Bologna; 45 were stage T1a (tumour limited to one vocal cord) and 28 were stage T1b (tumour of both vocal cords or involving anterior commissure); radiation treatment utilised a 60Co machine and 5 x 5 cm fields; the median dose was 67.2 Gy (range 50-76) with conventional fractionation. Ten patients had local recurrence; the median time of recurrence was 13.4 months; 9/10 were treated by surgery and 2/10 died, so the overall control by radiotherapy with surgery in reserve was 100% in T1a tumours and 90.6% in T1b ones. The 5-years disease free survival rate was 93.1% in T1a tumours and 69% in T1b; lesions involving anterior commissure had the worst prognosis, independent of the dose and time-dose factor (3/10 recurrences in the group treated with TDF less than 110 and 4/18 recurrences in the group with TDF more than 110).     

Imaging features of retroperitoneal and pelvic schwannomas

AIM: To describe the imaging features of retroperitoneal and pelvic schwannomas. MATERIALS AND METHODS: The presenting cross-sectional imaging for 18 sequential patients with retroperitoneal or pelvic schwannomas was reviewed retrospectively. Note was made of tumour diameter, position, homogeneity, margin, shape, calcification and invasion into adjacent structures. Where MRI had been performed, T1 and T2 signal intensity relative to skeletal muscle, and the degree and pattern of enhancement with gadolinium, were also assessed. RESULTS: Imaging from 13 patients was available for review. The mean tumour diameter was 8.7 cm (range 4 to 15 cm); 9 schwannomas were located in the pelvis and 4 in the retroperitoneum; 12 cases showed smooth, regular margins and 1 case irregular, invasive margins. The tumours were homogeneous in 5 cases and heterogeneous with cystic change in 8; in 2 cases there was smooth expansion of a sacral nerve root exit foramen, and in 1 there was bony destruction of the sacrum and extension of tumour into the spinal canal. In 5 cases MRI was performed; on T1-weighted images all tumours were isointense; on T2-weighted images 4 tumours were hyperintense and 1 was isointense to skeletal muscle. In all cases the diagnosis was confirmed by core biopsy. CONCLUSION: Retroperitoneal and pelvic schwannomas typically form large, well-circumscribed masses in the retroperitoneum or presacral area, and frequently undergo cystic degeneration. They can occasionally cause bony changes in the spine, but otherwise do not invade or obstruct adjacent structures. Although they are rare, it is important for the radiologist to recognize the typical appearance of schwannomas because they can be mistaken for malignant tumours.     

Soft tissue sarcomas after radiation treatment for breast cancer. Three case studies and review of literature.

AIMS: By means of 3 cases with infield soft tissue carcinomas after radiotherapy for breast cancer, symptoms and therapy are described. Consequences for treatment planning and patient's information before radiotherapy for breast cancer are discussed. PATIENTS: Three of 1,025 patients with breast cancer irradiated from 1984 to 1997 suffered from infield secondary soft tissue sarcomas. The latency periods were 61, 49 and 59 months. Two patients had been treated with breast-conserving therapy (computerized planning, 50 Gy to reference point, 5 times 2 Gy/week, 5-MV photons), 1 patient received a local boost dose of 15 Gy (10-MeV electrons), patient 3 radiotherapy of the thoracic wall and regional lymph nodes after mastectomy using 12-MeV electrons (thoracic wall) and 5-MV photons (lymph node areas) to 50 Gy, 5 times 2 Gy/week. No adjuvant chemotherapy was given. All sarcomas were very extensive, all patients died from local progression and/or distant failure after 17, 13 and 12 months. RESULTS: The incidence of spontaneous sarcomas of the breast is about 0.06%, after operation and radiotherapy 0.09 to 0.45%. No correlations to radiotherapy technique and no risk factors were found. Radiation dose could play a role, but there are very sparse data about this. CONCLUSIONS: Secondary soft tissue sarcomas are very rare, but familiar complications of radiotherapy. Only early diagnosis leads to a chance for cure. Because of unclear correlations to the treatment parameters and rareness of this event, in our opinion no regular information to the patient receiving radiotherapy for breast cancer is mandatory.     

Xenografts of five human leiomyosarcomas: radiation response after 60cobalt- and d(14)+Be neutron single doses

Five permanently established xenograft lines of human soft tissue sarcomas were irradiated with single doses of 5.8 MeV d(14)+Be neutrons and of 60Co rays, respectively, at several dose levels to generate dose response relationships. The tumors were clamped ten minutes prior to and during irradiation to induce uniform hypoxia. All tumours were previously characterized by means of histomorphology, tumour doubling times (DT's), DNA-index and enzyme pattern of the lactate dehydrogenase (LDH) and glucose-6-phosphate dehydrogenase (GPD). According to these criteria, three out of five leiomyosarcomas were identical referring to the biopsy of origin, whereas two had changed in successive passages. For the different tumour lines, specific growth delays ranged from 0 to 8.7 after 5.3 Gy neutrons and from 0 to 11.4 after 16 Gy60Co, respectively. In terms of radiosensitivity for different single doses and irradiation qualities, a highly significant overall correlation (rs = 0.82 +/- 0.06) was found for the ranking of the tumours with respect to the growth delay and specific growth delay endpoints. No correlation was found between tumour doubling times and the relative biological effectiveness (RBE). In general, calculated RBE-values decreased with increasing effect level. For the five tumour lines, RBE-values ranged from 1.6 to 12.7 and 2.0 to 4.4 at specific growth delays of 0.5 and 2.0, respectively, under acutely hypoxic conditions. These results indicate a potential advantage for neutrons in a subgroup of human soft tissue sarcomas compared with sparsely ionising irradiation.     

Neutron and neutron boost irradiation of soft tissue sarcomas

The results of neutron and neutron boost irradiation of 199 patients with soft tissue sarcomas treated between 1978 and 1983 are presented. The median follow-up period is 42 months. The recurrence free survival rates by last review are 93% for patients with T1 tumours (n = 14), 87% for T2 tumours (n = 84) and 73% for T3 tumours (n = 101). The actuarial survival rates at six years are 77% for T1, 63% for T2 and 34% for T3 tumours (p = 0.018). The actuarial survival rate for the group of patients irradiated after surgery without clinical evidence of residual tumour is 63.8% compared with 30.9% for the group of patients with measurable tumour volume at the beginning of radiotherapy (p = 0.002). The survival rates according to grading are 52% for patients with G1 tumours (n = 44), 54% for G2 tumours (n = 130) and 36% for G3 tumours (n = 25). The morbidity rate of 22% after full neutron irradiation was reduced to 15% by the introduction of a neutron boost. At the present time, the results of this modified treatment are not inferior to a full neutron course. The effectiveness of neutron or neutron boost irradiation in the postoperative treatment of soft tissue sarcomas will be evaluated in a forthcoming EORTC trial.     

Nine pheochromocytomas in the same patient. Final mapping with ultrasound and angiography

A 37-year-old man presented with hypertension and elevated urine catecholamine. Ultrasound scanning revealed a solid tumour of the right adrenal gland and two solid tumours in the retroperitoneum. The findings were confirmed with computed tomography and abdominal angiography. At surgery only the tumour of the right adrenal gland was removed. The histopathologic diagnosis was pheochromocytoma. Postoperatively the symptoms and biochemistry were unchanged and the patient was referred for further treatment. At ultrasonography and abdominal aortography 6 remaining tumours were demonstrated. Surgery was performed and 8 pheochromocytomas were extirpated (3 were closely spaced small tumours in a conglomerate corresponding to one of the visualized tumour sites). On histopathologic examination no signs of invasive growth were found. The patient recovered completely. The blood pressure was still normal 2 1/2 years later. Angiography and non-invasive examination of the entire abdomen and pelvis should be routine when pheochromocytomas are searched for.     

Simultaneous radiotherapy and chemotherapy with cisplatin and 5-fluorouracil in advanced head and neck tumors]

Between 1.1. 1987 and 31.12. 1988 47 patients with locally advanced head- and neck tumours (UICC-stage II: one patient, stage III: 16 patients, stage IV: 30 patients) were treated with simultaneous radio-chemotherapy. Localisations were hypopharynx (23 = 49%), oropharynx (16 = 34%) and other sites (8 = 17%). Radiation therapy consisted of 60 Gy (primary region, N+) or 50 Gy (N0) in daily fractions of 2 Gy each over six weeks. During week 1 and 5 of radiotherapy two courses of chemotherapy with 5-FU (600 mg/m2/day, i.v. continuous infusion for five days) and cis-platin (25 mg/m2/day i.v. bolus for five days) were administered. 43/47 patients (91%) responded to therapy. 34/47 (72%) patients achieved a complete remission, 9/47 (19%) a partial remission, and 4/47 (9%) no change. With a minimum follow-up of 26 months 17/47 patients (36%) are alive and NED, 4/47 (9%) are alive with tumour. 18 patients (38%) died of cancer, eight patients (17%) died of second tumours, intercurrent diseases and of unknown reasons. Actuarial four-year survival is 45% (stage III: 56%, stage IV: 42%), four-year NED survival is 35% (stage III: 68%, stage IV: 18%). There were no loco-regional recurrences after additional surgical treatment of the primary and the neck following complete remission. In contrast after RCT alone and complete remission local recurrences in 6/21 patients (29%) and regional recurrences in 5/24 patients (21%) occurred. We conclude that simultaneous RCT is a new very effective treatment modality of locally advanced head and neck tumours producing superior loco-regional control compared to conventional management.     

Ergebnisse der bestrahlung inoperabler nichtkleinzelliger bronchialkarzinome im stadium III mit 25 Gy in fünf fraktionen

BACKGROUND: Patients with advanced Stage III inoperable non-small cell lung cancer who were not suitable for irradiation with curative doses, were treated at the Department of Radiotherapy of the University Hospital of Dresden with 25 Gy in 5 fractions over 1 to 2 weeks. Survival of these patients was compared in this retrospective study to the survival of patients treated during the same period with 60 Gy in 30 fractions. PATIENTS AND METHOD: Between 1985 and 1994 298 patients were treated for a histologically or cytologically proven non-small cell lung carcinoma with 60 Gy in 30 fractions (n = 80), with 40 Gy in 20 fractions (n = 26) or with 25 Gy in 5 fractions (n = 192). Overall survival was determined using actuarial methods. Prognostic parameters were analyzed using uni- and multivariate tests. RESULTS: Median overall survival for all patients was 6 months (95% confidence interval 5; 7). In univariate analysis, survival of the patients treated with 60 Gy was significantly better than survival in the other groups. Median survival was 11 months (9; 13) after 60 Gy, 6 months (4; 8) after 40 Gy and 5 months (4; 6) after 25 Gy. In multivariate analysis the treatment schedule lost its significant influence on outcome of the therapy. The most important prognostic parameter was the performance status of the patients. CONCLUSIONS: When stratified for performance status as the most important prognostic parameter the survival time after hypofractionated irradiation to 25 Gy given in 5 fractions in 1 to 2 weeks was not significantly different from the results after conventional fractionation to 60 Gy. Hypofractionated radiation schedules are often more convenient for the patient, economical, and have been shown to be effective in symptom control. Thus, in clear palliative situations hypofractionated treatment with 25 Gy in 5 fractions or a comparable schedule appears to be a reasonable therapeutic option.     

Hyperthermia promotes the incidence of tumours following X-irradiation of the rat cervical cord region.

The cervical region of the rat, including the spinal cord (cervical 5-thoracic 2) was irradiated with single doses of 15-32 Gy 250 kV X-rays. Hyperthermia, at temperatures of 42-, 43- and 44 +/- 0.1 degrees C for 30 min was applied to the cervical vertebral column and immediate adjacent tissues for 5-10 min or 7 h after X-irradiation. Over a period of 18-21 months, animals were followed up to monitor neurological complications occurring as a result of damage to the spinal cord (Sminia et al. 1991). We also noted the development of neoplasms either inside or outside the cervical region. The data on tumour incidence were analysed retrospectively using the actuarial method. Although hyperthermia alone was not carcinogenic, it led to a significant increase of radiation-induced tumours. This increase of radiation carcinogenesis was observed both with hyperthermia applied 5-10 min after X-rays and with an interval of 7 h between X-rays and heat. Cancer induction was highest after the lower radiation doses (16 Gy) combined with high heat doses (30 min 44 degrees C). The latent period for induction of tumours by X-rays was 472 +/- 19 days (mean +/- SEM; n = 24). Latency was significantly shortened by hyperthermia to 404 +/- 34 days (n = 22) if applied 5-10 min after X-rays and to 348 +/- 6 days (n = 33) with an interval of 7 h. Histology revealed that 86% (38/44) of the examined tumours found inside the volume treated with hyperthermia and irradiation were sarcomas. The percentage of animals with a tumour outside the treated volume was almost the same for all treatment groups. Most of these tumours were of the mammary gland type.     

Fluorodeoxyglucose positron emission tomography of soft tissue tumours: is a non-invasive determination of biological activity possible?

Since musculoskeletal tumours comprise a large heterogeneous group of entities with different biological behaviour, clinical diagnosis of such lesions can be very difficult. The aim of this prospective study was to assess the usefulness of 2-[F-18]-fluoro-2-deoxy-d-glucose (FDG) positron emission tomography (PET) in the non-invasive evaluation of soft tissue tumours. One hundred and two patients with suspected soft tissue neoplasms were investigated by FDG-PET. The uptake of FDG was evaluated semiquantitatively by determining the tumour to background ratio (TBR). All patients underwent biopsy, resulting in the histological detection of 39 high-grade sarcomas, 16 intermediate-grade sarcomas, 11 low-grade sarcomas, 25 benign tumours, 10 tumour-like lesions such as spontaneous myositis ossificans (n = 6) and one non-Hodgkin lymphoma. All lesions except for two lipomas disclosed an increased FDG uptake. Sarcomas showed significantly higher TBR values than latent or active benign lesions (P<0.001) and aggressive benign lesions (P<0.05). Using a TBR cut-off level of 3.0 for malignancy, sensitivity of FDG-PET was 97.0%, specificity 65.7% and accuracy 86. 3%. From our data there are three main conclusions: (1) Except for patients with pseudotumoral myositis ossificans, lesions with a TBR >3 were sarcomas (91.7%) or aggressive benign tumours (8.3%). (2) Tumours with a TBR <1.5 were latent or active benign lesions, exclusively. (3) The group with intermediate TBR values (<3 and >1. 5) comprised primarily latent or active benign lesions, but also four aggressive benign tumours and two low-grade sarcomas. Our data suggest that FDG-PET represents a useful tool for the evaluation of the biological activity of soft tissue neoplasms.     

Hyperthermia Promotes the Incidence of Tumours Following X-irradiation of the Rat Cervical Cord Region

Summary

The cervical region of the rat, including the spinal cord (cervical 5-thoracic 2) was irradiated with single doses of 15–32 Gy 250 kV X-rays. Hyperthermia, at temperatures of 42-, 43- and 44 ± 0·1°C for 30 min was applied to the cervical vertebral column and immediate adjacent tissues for 5–10 min or 7 h after X-irradiation. Over a period of 18–21 months, animals were followed up to monitor neurological complications occurring as a result of damage to the spinal cord (Sminia et al. 1991). We also noted the development of neoplasms either inside or outside the cervical region. The data on tumour incidence were analysed retrospectively using the actuarial method. Although hyperthermia alone was not carcinogenic, it led to a significant increase of radiation-induced tumours. This increase of radiation carcinogenesis was observed both with hyperthermia applied 5–10 min after X-rays and with an interval of 7 h between X-rays and heat. Cancer induction was highest after the lower radiation doses (16 Gy) combined with high heat doses (30 min 44°C). The latent period for induction of tumours by X-rays was 472 ± 19 days (mean ± SEM; n = 24). Latency was significantly shortened by hyperthermia to 404 ± 34 days (n = 22) if applied 5–10 min after X-rays and to 348 ± 6 days (n = 33) with an interval of 7 h. Histology revealed that 86% (38/44) of the examined tumours found inside the volume treated with hyperthermia and irradiation were sarcomas. The percentage of animals with a tumour outside the treated volume was almost the same for all treatment groups. Most of these tumours were of the mammary gland type.     

Cisplatin as a radiosensitizer in the treatment of solid tumours--a clinical pilot study

A variety of experimental tumour models have shown that cisplatin in combination with ionizing radiation enhances tumour regression. To evaluate the feasibility, efficacy and toxicity of a combined regimen, 25 patients with different solid tumours were treated with cisplatin and photon irradiation. A local tumour control was achieved in 22 cases. Thirteen patients with squamous cell carcinoma of the lung (T1-3, N1-2, M0) received cisplatin (20 mg/m2 per day) on 5 consecutive days with NaCl hyperhydratation (0,9% NaCl, 2400 ml/24 hours as continuous infusion) in the first and last week of radiotherapy (5 X 2 Gy/week, 56 Gy target volume dose). 12 of 13 patients have shown local tumour control. Increased clinically evident side effects were not observed. From the underlying data the conclusion is drawn that the experimental results are reproducable under clinical conditions and that improved local tumour response rates can be achieved.     

Improved trigeminal and facial nerve tolerance following fractionated stereotactic radiotherapy for large acoustic neuromas

The purpose of this study was to demonstrate improved cranial nerve tolerance following fractionated stereotactic radiotherapy for large acoustic neuromas, defined as tumours with pons-petrous distance (A) > 1 cm and midporous transverse diameter (A + Y) > 2 cm. Of 28 patients with acoustic neuromas treated with fractionated stereotactic radiotherapy, 19 had large tumours at high risk for radiosurgery-induced cranial neuropathy. Six patients received 36 Gy in six, weekly, fractions and 13 patients received 30 Gy in six, weekly, fractions. 15 patients had evaluable trigeminal nerve function and 16 had evaluable facial nerve function. At a median follow-up of 4.5 years, tumour shrinkage was seen in 10 patients and tumour size was stable in nine. None of the patients developed any evidence of transient, or permanent, trigeminal or facial neuropathy at any time during their follow-up period. Fractionated stereotactic radiotherapy may offer a superior therapeutic ratio to single fraction stereotactic radiosurgery in the management of large acoustic neuromas, as evidenced by the absence of post-treatment trigeminal and facial neuropathy.     

Incidence of tumours in the cervical region of the rat after treatment with radiation and hyperthermia

The incidence of tumours in the irradiated cervical region in female Wistar (WU) rats after retreatment of part of the volume with hyperthermia was examined retrospectively. The cervical spinal cord (cervical 5-thoracic 2) was irradiated with a single dose of 15, 18 or 20 Gy. Ninety days thereafter, the cervical region was heated by means of a microwave applicator at a maximum temperature of 43 degrees C for 50-90 min measured at the vertebral column. Over a period of 18 months after treatment, animals were regularly observed. Neurological complications and the development of neoplasms were noted. From the 354 animals included in the study, 82 animals developed a tumour. Hyperthermia alone was not carcinogenic, but enhanced the carcinogenesis induced by radiation. The percentage of animals that developed a tumour inside the volume treated with hyperthermia 90 days after irradiation was significantly higher relative to radiation alone (33 +/- 5 per cent versus 4 +/- 2 per cent, P less than 0.001). The duration of the latent period before appearance of these tumours was not affected (355 +/- 18 days versus 425 +/- 54 days). No significant differences in the percentage of animals that developed a tumour at another site were observed between different treatment groups. Histology revealed that 88 per cent (14/16) of the examined tumours found inside the treated volume after hyperthermia and irradiation were soft tissue rhabdomyosarcomas. Outside the treated volume, most tumours were tumours of the mammary gland.     

Results of a randomised study to evaluate influence of dose on morbidity in radiotherapy for bladder cancer

Ninety-four patients with invasive transitional-cell carcinoma of the bladder were randomised to receive three different doses of megavoltage X-rays. Doses of 50.0 Gy, 52.5 Gy and 57.5 Gy were given in 20 daily fractions. Patients were stratified into two groups by tumour size: less than 5.0 cm and 5.0-7.0 cm in diameter. The large tumours were more advanced (higher T stage) and included a larger number of patients with less well differentiated tumours. Comparisons of morbidity are made in each dose group defined by tumour size. Early and late morbidity was carefully assessed during follow-up and was found to be unacceptably high in the group who received 57.5 Gy in 20 daily fractions.     

The potential therapeutic gain of radiation-associated gene therapy with the suicide gene cytosine deaminase

PURPOSE: To estimate the concentration of 5-fluorocytosine (5-FC), necessary for conversion to 5-fluorouracil (5-FU) in tumours transduced with the gene cytosine deaminase (CD), to achieve clinically significant radiosensitization to radiotherapy. MATERIALS AND METHODS: Starting with a tumour control probability (TCP) of 37% from radiotherapy of 66 Gy in 2 Gy fractions, estimates were made of increase in TCP expected from sensitizer enhancement ratios (SER) of 1.1, 1.2, etc. SER values for 5-FU were obtained from a literature review. Clinical toxicity of 5-FC is also reviewed. RESULTS: 5-FU has been reported to be an effective radiosensitizer if maintained for several days after each irradiation at concentrations of 0.6-0.9 microg/ml in surrounding medium. 5-FC is well tolerated by patients at concentrations of 25-100 microg/ml (average 60 microg/ml) for 6 weeks in standard antifungal treatment. Sufficient 5-FU should be available if conversion efficiency from 5-FC is 1-3%. SER values of 1.1 to 1.2 should be achievable with daily 2 Gy fractions. In vitro and xenograft experiments are reviewed and they do not contradict the conclusions. CONCLUSIONS: Increases in tumour control of 20 to 40% can be expected, which should be detectable in a 2-arm randomized trial of 260 (for 20%) or 60 (for 40%) patients.