Results of External-beam Radiotherapy Alone in Invasive Cancer of the Uterine Cervix: A Retrospective Analysis

Material and Methods We included 146 patients with invasive cancer of the cervix who were treated with EBRT to a total dose of 60–66 Gy between January 1996 and December 2001. None of these patients were suitable for intracavitary radiotherapy (ICRT) after a median dose of 46 Gy. A boost dose of 14–20 Gy was given after a gap of 2–4 weeks. Most patients belonged to stage IIIB ( n = 124). Results Follow-up of patients at risk ranged from 19 to 89 months (median 48 months). One hundred and thirty-six patients (93.2%) received EBRT to a dose of 66 Gy, and 10 patients (6.8%) received 60 Gy. Overall treatment time (OTT) ranged from 56 to 160 days (median 78 days). At completion of 46 Gy of EBRT, 63 patients achieved partial response and 83 patients had stable disease. Five-year overall survival, disease-free survival (DFS) and pelvic control were 15.1% (median 9 months), 11.6% (median 5 months) and 21.9% (median 6 months), respectively. Factors found to affect 5-year pelvic control in univariate analysis by Kaplan–Meier method were response to EBRT at 46 Gy (partial response 36.5% and stable disease 10.8%), age (≥50 years 28.8% and <50 years 13.6%) and OTT (<90 days 26.5% and ≥90 days 12.5%). For DFS and overall survival, response to EBRT was the only factor that was significant in univariate analysis. In multivariate analysis by Cox's proportional hazard model, response to EBRT was the only factor to influence pelvic control ( P = 0.007), DFS ( P = 0.01) and overall survival ( P < 0.001). Conclusions Overall outcome of patients in whom ICRT was not given remains less than satisfactory. Response to EBRT emerged as the most important factor to predict all clinical outcomes. To improve upon the dismal results of EBRT alone, we will have to decrease the OTT and consider concurrent chemo-radiation with cisplatin. Key words Cancer cervix external-beam radiotherapy intracavitary radiotherapy overall treatment time Introduction Intracavitary radiotherapy (ICRT) has become an integral component of radiotherapy in the treatment of carcinoma of the uterine cervix since its importance in achieving better pelvic control and overall survival was established by the Patterns of Care Studies carried out in the USA [1–5] . In locally advanced disease, external-beam radiotherapy (EBRT) is usually given initially to decrease the volume of disease, allowing an improved geometry for ICRT applications [6] . However, it may not be possible for all patients to have ICRT owing to gross residual disease or distortion of local anatomy after external radiation. This is especially true in developing countries, such as India, where most of the patients present at late stages. Some published studies have shown acceptable results with EBRT alone in those patients who are not suitable for ICRT in cancer of the cervix [6–10] . In this retrospective audit, we describe the results of EBRT alone in carcinoma of the cervix treated at our centre. Patients and Methods Patients Between January 1996 and December 2001, 1069 patients were treated with radical radiotherapy at our centre [11] . After a median EBRT dose of 46 Gy, ICRT was possible in 871 patients, and 52 patients either had progressive disease or were lost to follow-up. These patients were excluded from this analysis. One hundred and forty-six patients received further EBRT to a total dose of 60–66 Gy, and these patients were included in this study. Patients were staged according to the International Federation of Gynecologists and Oncologists (FIGO) staging system, after a work-up that included clinical examination, complete blood counts, kidney function tests, chest X-ray, intravenous pyelography, cystoscopy and rectosigmoidoscopy. Teletherapy EBRT was delivered using either a Cobalt-60 unit or a linear accelerator (6 MV Clinac). The upper border of the pelvic field was at the L5–S1 junction; the lower border was at the lower most part of the obturator foramen, which was modified according to the vaginal extent of disease. Lateral borders were kept 1.5 cm beyond the widest pelvic brim. When four-field box technique was used, the anterior and posterior borders of the lateral portals were kept at the anterior part of the pubic symphysis and S2–S3 junction respectively. Dose of EBRT was 46 Gy in 23 fractions. Dose was prescribed at the midplane on anterior–posterior/posterior–anterior (AP/PA) and isocentre on four-field box technique. Patients with antero-posterior separation less than 20 cm were treated with Cobalt-60 using AP/PA technique, and more than 20cm were treated with Linac mostly by AP/PA and a few by four-field box technique. When the patient was not suitable for ICRT after an EBRT dose of 46 Gy, further EBRT (14–20 Gy) was given in the same radiation portals after a gap ranging from 2 to 4 weeks (to reassess for suitability for ICRT). Many patients who had only poor response to EBRT with stable disease were treated only when they again became symptomatic of the residual disease and hence their overall treatment time (OTT) was unduly prolonged. Follow-up All the patients were advised to attend an assessment of response or complications 1 month after completing treatment, and thereafter every 2 months for 1 year, 3 months for the subsequent 3–4 years and then 6 monthly. PAP smear and biopsy were carried out at follow-up only if clinically indicated. Chest X-ray, Ultrasonography and Computed Tomography scans were carried out at the discretion of the attending physician. Any patient with symptoms suggestive of late complications were followed up closely or admitted to hospital for investigation. Acute and late toxicities were documented according to the Radiation Therapy Oncology Group (RTOG) criteria. Statistical Analysis Kaplan–Meier method was used to evaluate overall survival, disease-free survival (DFS) and pelvic control. Interval for all these outcomes was taken from the date of registration to the time of event. It was assumed that all patients lost to follow-up died from their disease regardless of status. Outcomes for various prognostic factors were also calculated by Kaplan–Meier method, and comparisons were made by Log-rank test. Cox's proportional hazard model was used to carry out multivariate analysis of all factors for outcomes. Incidence of late toxicities was calculated by Kaplan–Meier method. Results Follow-up ranged from 2 to 89 months (median 9 months; mean 16.5 months). Follow-up of living patients ranged from 19 to 89 months (median 48 months; mean 49.7 months). Sixteen patients were lost to follow-up. Demographic profiles are given in Table 1 . In stage I/IIA ( n = 3), ICRT was not given due to medical reasons. In stage IIB, two patients did not receive ICRT because of comorbid diseases making them unfit for anaesthesia. For the remaining patients in stage IIB ( n = 8), and all the patients in stages IIIB ( n = 124) and IVA ( n = 9), ICRT could not be carried out because of persistent gross disease. The decision to consider a patient suitable for ICRT mostly depended on the evaluating physician. The main reason for not doing ICRT was presence of large residual disease, which may be best described as any residual disease with extension beyond the medial half of the parametrium on either side or a size greater than 4 cm. Reasons for not doing ICRT are given in Table 2 . One hundred and thirty-six patients (93.2%) received an EBRT of 66 Gy, and 10 patients (6.8%) received 60 Gy. AP/PA technique was used in 140 patients (95.9%), and four-field box technique in six patients (4.1%). Cobalt-60 was used in 95 patients (65.1%) and 6 MV linear accelerator in 51 patients (34.9%). Partial response was seen in 63 patients (43.2%), and stable disease was observed in 83 patients (56.8%) at the end of EBRT (after 46 Gy). OTT ranged from 56 to 160 days (median 78 days; quartiles 67 [25%], 78 [50%] and 90 [75%]). Five-year overall survival, DFS and pelvic control were 15.1% (median 9 months), 11.6% (median 5 months) and 21.9% (median 6 months), respectively. In patients who were lost to follow-up ( n = 16), eight had evidence of disease at last visit. If we censor them at last visit according to their status, the 5-year DFS and overall survival will increase to 17.1% and 26%, respectively. Stage-wise results were as follows: 5-year overall survival, DFS and pelvic control were 0% each in stage I; 50%, 0% and 0% in stage IIA; 10%, 10% and 20% in stage IIB; 14.5%, 12.1% and 23.4% in stage IIIB; and 22.2%, 11.1% and 11.1% in stage IVA, respectively. The only patient in stage I was lost to follow up soon after treatment without having a post-treatment evaluation. One patient in stage IIA died of cardiac problems after radiation treatment without having an assessment of her disease status. Second patient in stage IIA had recurrent pelvic and para-aortic node disease, and she refused any kind of chemotherapy (surgery was not possible as she was medically unfit) but was alive at the time of data analysis. So the inferior results of stages I–IIA should be interpreted with caution. Inability to control the local disease was the main cause of failure, with 78.1% of patients having documented disease in the cervix and parametrium. The failure pattern is shown in Table 3 . EBRT was well tolerated. Five patients had grade 3 lower gastrointestinal toxicity and four patients developed grade 3 skin reactions in the perineum. Five-year actuarial late toxicity (RTOG) rate (all grades) was 6.2% in the rectum, 2.1% in the bladder, 5.4% in the small intestine and 50.7% in the subcutaneous tissue, respectively. Severe toxicity (≥grade 3) rate was 0.7% each in the bladder and rectum, 0% in the small intestine and 5.5% in the subcutaneous tissue respectively [12] . The prognostic factors evaluated for pelvic control, DFS and overall survival were age (<50 years vs ≥50 years), comorbid diseases (presence vs absence), pre-treatment haemoglobin (≤10 g% vs >10 g%), histology (squamous vs non-squamous), stage (I–IIB vs IIIB/IVA), bulk (<4 cm vs >4 cm), hydronephrosis (presence vs absence), treatment machine (cobalt vs linear accelerator), response at 46 Gy dose of EBRT (partial response vs stable disease), total dose (60 Gy vs 66 Gy) and OTT (<90 days vs ≥90 days). In univariate analysis by Kaplan–Meier method ( Table 4 ; Figs. 1–3 ), age, OTT and response to EBRT emerged as significant factors affecting pelvic control. For DFS and overall survival, response to EBRT was the only significant factor. In multivariate analysis by Cox's proportional hazard model, response to EBRT was the only factor to influence pelvic control ( P = 0.007; odds ratio: 2.172; 95% confidence interval: 1.234–3.821), DFS ( P = 0.01; odds ratio: 1.917; 95% confidence interval: 1.152–3.192) and overall survival ( P < 0.001; odds ratio: 1.987; 95% confidence interval: 1.377–2.867). Discussion The combination of EBRT and ICRT is the standard treatment for carcinoma of the uterine cervix. Addition of ICRT to EBRT has been found to significantly decrease local recurrence and cancer-related deaths in almost all patterns of care studies from North America [1–5] . However, persistent gross disease and distortion of anatomy after EBRT sometimes necessitate the entire treatment to be given by EBRT alone without ICRT. Koeck et al. [13] treated 55 patients with cancer of the cervix (all stages) with Cobalt-60 rotation therapy up to a dose of 80 Gy without ICRT. They found that their 3-year overall survival rates (67%) were comparable with the contemporary series using a combination of EBRT and ICRT 3 decades previously. However, these days, the technique used is almost obsolete, and in this study EBRT alone was given to patients who were otherwise suitable for ICRT. Castro et al. [14] in 1970 published the experience of MD Anderson hospital in 108 cases of cancer of the cervix (all stages) using EBRT alone. They concluded that when ICRT is not possible, an EBRT dose of 60–70 Gy within 7 weeks is necessary to attain reasonable pelvic control. The dose–response relationship was more evident in tumours with limited volume than advanced diseases. Within the past 2 decades, many trials have documented the use of EBRT alone in patients in whom ICRT was not possible. Ulmer and Frischbier [7] reported impressive 5-year overall survival of 75%, 30% and 13% in stages II, III and IV cancer of the cervix with EBRT alone, respectively, with acceptable complications. Akine et al. [8] treated 104 patients with cancer of the cervix with EBRT alone, and the 5-year survival rate was 17% overall and 36%, 17% and 5% for patients with stages II, III and IV, respectively. EBRT dose ranged from 60.8 Gy in 6 weeks to 80.5 Gy in 8 weeks. Major complications due to treatment were observed in five patients, with four requiring surgical intervention; two patients died. Authors concluded that doses up to 70 Gy may be optimum in these patients. In the study by Lei and He [9] , the 5-year survival rate of 97 patients with stage IIB cancer of the cervix was 56.7%, but late complications were slightly higher than that reported in other studies, with two patients developing rectovaginal fistula. The dose of EBRT in this study was 60 Gy in 6–8 weeks followed by a boost of 10 Gy by a reduced AP/PA portal or a perineal portal within a week. Montana et al. [6] evaluated the results of EBRT alone in 88 patients with stage III cancer of the cervix, and compared the results with those treated with a combination of EBRT and ICRT. Dose of EBRT was 65 Gy when it was used alone. They documented that, even though the 2-year DFS was better with the combination (61% vs 36%), by 10 years the results became similar (27% each). Incidence of late complications was also similar in both groups. Ferreira et al. [10] compared the results of 109 patients of stage IIIB cancer of the cervix treated with EBRT alone with those treated with EBRT and ICRT, and reported that the 10-year overall survival and pelvic control rate were significantly higher with the combination (22.5% vs 15.6% and 48.4% vs 31.2%, respectively). However, a subset analysis of patients who received higher doses of EBRT in an appropriate period of time (60 Gy/6–8 weeks or 70 Gy/7–9 weeks) showed that the difference in overall survival (22.5% vs 18.9%) and local control (48.4% vs 40%) became smaller and insignificant. The trials by Montana et al. [6] and Ferreira et al. [10] show that acceptable results can be achieved by using EBRT alone, and the results are even comparable with that of a combination of EBRT and ICRT. In our study, 5-year overall survival and pelvic control were 15.1% and 21.9%, respectively. Small numbers of patients in stages other than stage IIIB preclude us from making meaningful interpretations about the effect of stage in outcome. Our results with EBRT alone are not only inferior to the results obtained by Montana et al. [6] and Ferreira et al. [10] with EBRT alone, but also significantly inferior to the results we achieved with the combination of EBRT and ICRT ( n = 871; 5-year overall survival: 60.7%; DFS: 58.6% and pelvic control: 73.5%) [11] . Stage-wise distribution of patients in the combination group was as follows: stages I, 3.2%; IIA, 4.8%; IIB, 27.3%; IIIA, 0.7%; IIIB, 63.8% and IVA, 0.2%. Stage-wise comparison of EBRT alone and EBRT + ICRT is given in Table 5 . Except for stage IIIB, the small number of patients makes this comparison meaningless. Logsdon and Eifel [15] evaluated 30-year experience of MD Anderson hospital and documented that the 5-year disease-specific survival of stage IIIB patients treated with EBRT alone ( n = 266; 24%) is significantly lower than that of patients treated with the combination EBRT and ICRT ( n = 641; 45%). Our study, and the study by Logsdon and Eifel, reinforces the importance of ICRT in the management of cancer of the cervix documented earlier by the patterns of care studies. ICRT was attempted in every possible patient, and this is shown by the ICRT rate of 81.5% in all stages (871 out of 1069 patients) and 76.5% in stage IIIB (556 out of 729 patients) [11] . In the trial by Montana et al. [6] , the ICRT rate in stage III disease is only 56.6%, and Ferreira et al. [10] only managed ICRT in 46% of stage IIIB patients. This implies that the patients treated with EBRT alone in our study were having a higher tumour burden and worse prognosis than the above-mentioned trials. The condition of these patients was generally poor, and nutritional status due to advanced disease prevented us from treating them aggressively. Another important factor responsible for the unsatisfactory results is the prolongation of OTT. Treatment was completed within 6–9 weeks in the above-mentioned studies, whereas only 25% of our patients completed treatment within 9 weeks. Chemo-radiation is the current standard of care that may improve outcome considerably in this unfavourable situation. We are planning to use concurrent chemo-radiation with weekly cisplatin, which may be ideal for developing countries [16] . Intensity-modulated radiotherapy (IMRT) also has shown some promise in locally advanced cervical cancer, but its application to developing countries is limited in the present scenario [17] . Younger patients (<50 years) had significantly inferior pelvic control than older patients ( Fig. 1 ), and this fact has been well documented [5,18] . Patients treated within 3 months had better pelvic control ( Fig. 2 ) at 5 years (26.5%) than those patients who had further prolongation in OTT (12.5%). Petereit et al. [19] showed that survival decreased by 0.6% a day and pelvic control by 0.7% a day for each additional day of treatment beyond 55 days for all stages of cancer of the cervix. In our study, only six patients completed treatment within 55 days. For the 37 patients with OTT ≤ 67 days (first quartile of OTT), the rate of pelvic control, DFS and overall survival was 29.7%, 18.5%, and 21%, respectively, which was much better than their counterparts (19.3%, 9.8%, and 13.6%, respectively) but not statistically significant ( P = 0.1, P = 0.2 and P = 0.2, respectively). Response to EBRT was the only factor significant for pelvic control ( Fig. 3 ), DFS and overall survival on multivariate analysis, and the same findings have been observed by others [20,21] . In patients who have attained at least a partial response to EBRT dose of 46 Gy, we could attain reasonable 5-year pelvic control (36.5%) and overall survival (23.8%). This is the subgroup that also had a shorter OTT. Almost half of the patients had late subcutaneous fibrosis, of which 5.5% developed grade 3 toxicity. This was due to the use of AP/PA portals (96%) with cobalt machines (65%) without shrinking the fields for the boost. This toxicity is seldom seen nowadays, with the use of four-field box technique and higher energy linear accelerators. Incidence of late toxicity in the rectum, bladder and small intestine was acceptably low, but this should be interpreted with caution. Most of the patients must have succumbed to death before manifesting the late toxicity in these organs. Conclusions Results of EBRT alone for patients with locally advanced cancer of the cervix in whom ICRT is not possible remain less than satisfactory in our study. Subgroups of patients with younger age, poor response to EBRT and prolonged OTT have dismal results. Shortening of OTT and use of chemo-radiation with cisplatin-based chemotherapy may yield a better outcome. However, to obtain the best results, it is inevitable that ICRT should be used in all possible patients. References 1 G.E. Hanks D.F. Kerring S. Kramer Patterns of Care outcome studies: results of the national practice in cancer of the cervix Cancer 51 1983 959 967 2 R.M. Lanciano K. Martz L.R. Coia G.E. Hanks Tumor and treatment factors improving outcome in stage III-B cervix cancer Int J Radiat Oncol Biol Phys 20 1991 95 100 3 L. Coia M. Won R. Lanciano V. Marcial K. Martz G. Hanks The Patterns of Care outcome study for cancer of the uterine cervix: results of the Second National Practice Survey Cancer 66 1990 2451 2456 4 R. Komaki T. Brickner A. Hanlon J. Owen G. Hanks Long-term results of treatment of cervical carcinoma in the United States in 1973, 1978, and 1983: Patterns of Care Study (PCS) Int J Radiat Oncol Biol Phys 31 1995 973 982 5 R. Lanciano M. Won L. Coia G. Hanks Pretreatment and treatment factors associated with improved outcome in squamous cell carcinoma of the uterine cervix: a final report of the 1973 and 1978 Patterns of Care Studies Int J Radiat Oncol Biol Phys 20 1991 667 676 6 G.S. Montana W.C. Fowler M.A. Varia L.A. Walton Y. Mack L. Shemanski Carcinoma of the cervix, stage III: result of radiation therapy Cancer 57 1986 148 154 7 H.U. Ulmer H.J. Frischbier Treatment of advanced cancers of the cervix uteri with external irradiation alone Int J Radiat Oncol Biol Phys 9 1983 809 812 8 Y. Akine I. Hashida Y. Kajiura Carcinoma of the uterine cervix treated with external irradiation alone Int J Radiat Oncol Biol Phys 12 1986 1611 1616 9 Z-Z Lei FZ He External cobalt 60 irradiation for stage IIB carcinoma of the uterine cervix Int J Radiat Oncol Biol Phys 16 1989 339 341 10 P.R.F. Ferreira A. Braga-Filho A. Barletta L.A. Ilha Radiation therapy alone in stage III-B cancer of the uterine cervix: a 17-year experience in Southern Brazil Int J Radiat Oncol Biol Phys 45 1999 441 446 11 E.P. Saibishkumar F.D. Patel S.C. Sharma Results of radiotherapy alone in the treatment of carcinoma of uterine cervix: a retrospective analysis of 1069 patients Int J Gynecol Cancer 15 2005 890 897 12 Saibishkumar EP, Patel FD, Sharma SC. Evaluation of late toxicities of patients with carcinoma of the cervix treated with radical radiotherapy: an audit from India. Clin Oncol , in press. 13 G.P. Koeck L.E. Jacobson W.R. Hillsinger Results of cobalt 60 rotation therapy in carcinoma of the cervix Am J Roentgenol Radium Ther Nucl Med 96 1966 81 91 14 J. Castro P. Issa G. Fletcher Carcinoma of the cervix treated by external irradiation alone Radiology 95 1970 163 166 15 M.D. Logsdon P.J. Eifel FIGO IIIB squamous cell carcinoma of the cervix: an analysis of prognostic factors emphasizing the balance between external beam and intracavitary radiation therapy Int J Radiat Oncol Biol Phys 43 1999 763 775 16 J.A. Green J.M. Kirwan J.F. Tierney Survival and recurrence after concomitant chemotherapy and radiotherapy for cancer of the uterine cervix: a systematic review and meta analysis Lancet 358 2001 781 786 17 B.D. Kavanagh T.E. Schefter Q. Wu Clinical application of intensity-modulated radiotherapy for locally advanced cervical cancer Semin Radiat Oncol 12 2002 260 271 18 A.W. Fyles M. Pintilie P. Kirkbride W. Levin L.A. Manchul G.A. Rawlings Prognostic factors in patients with cervix cancer treated by radiation therapy: results of a multiple regression analysis Radiother Oncol 35 1995 107 117 19 D.G. Petereit J.N. Sarkaria R. Chappell The adverse effect of treatment prolongation in cervical carcinoma Int J Radiat Oncol Biol Phys 32 1995 1301 1307 20 J.H. Hong M.S. Chen F.J. Lin S.G. Tang Prognostic assessment of tumor regression after external irradiation for cervical cancer Int J Radiat Oncol Biol Phys 22 1991 913 917 21 S.W. Chen J.A. Liang S.N. Yang H.L. Ko F.J. Lin The adverse effect of treatment prolongation in cervical cancer by high-dose-rate intracavitary brachytherapy Radiother Oncol 67 2003 69 76