It has long been recognized that many patients with locally advanced carcinoma of the cervix harbor occult paraaortic metastases. A randomized study demonstrated that elective paraaortic irradiation improved survival and reduced distant metastases. More recently, concomitant chemotherapy with pelvic irradiation has improved survival among patients with locally advanced carcinoma of the cervix. This has led to a reexamination of the role of extended-field irradiation. An important issue is the toxicity of concomitant chemotherapy and extended-field radiotherapy. The authors report a retrospective analysis of their experience with extended-field radiotherapy and high–dose-rate brachytherapy with or without concomitant chemotherapy.
The authors treated 54 women with biopsy-confirmed carcinoma of the cervix using extended-field radiotherapy and high–dose-rate brachytherapy with or without concomitant chemotherapy. The histology was squamous cell carcinoma in 49 patients (91%) and nonsquamous cell carcinoma in 5 patients (9%). The median size of the primary tumor was 7 cm (range, 3–10 cm). Each patient received 45 grays (Gy) of external beam radiotherapy to the pelvis and the paraaortic region, followed by a parametrial boost (9 Gy) in the patients with disease extension to the parametrium or the pelvic side wall(s). Each patient also underwent two applications of high–dose-rate brachytherapy, 1 week apart. The median dose delivered to Point A from each application was 9 Gy. Forty-four of the 54 patients (81%) received concomitant chemotherapy (cisplatin, 20 mg/m2/day for 5 days) during the first and the fourth weeks of external beam radiotherapy, and once after the second high-dose rate application. Chemotherapy was not assigned randomly.
One of the 10 patients (10%) treated without chemotherapy experienced acute toxicity, whereas 41 of 44 patients (93%) who received chemotherapy suffered from acute toxicity, including hematologic toxicity, gastrointestinal toxicity, and deep venous thrombosis. During a median follow-up period of 28 months (range, 12–70 months), 6 of the 54 patients have died (11%). The actuarial rate of local control at 3 years is 100% among the patients treated without chemotherapy, compared with 85% among those receiving chemotherapy. No one failed in the paraaortic region. The actuarial rates of freedom from distant metastases are 90% and 95% among the patients treated without and with chemotherapy, respectively. The actuarial incidence of late toxicity is 10% among the patients treated without chemotherapy and 6% among those receiving chemotherapy.
Many patients with locally advanced carcinoma of the cervix harbor occult paraaortic metastases.1–8 In 1995, Rotman et al.9 reported the results of a prospective, randomized study, which demonstrated that elective paraaortic irradiation improved survival and decreased distant metastases. More recently, cisplatin-based chemotherapy administered concurrently with pelvic irradiation improved survival among patients with locally advanced carcinoma of the cervix.10–15 The use of concomitant chemotherapy with irradiation has led to a reexamination of the role of extended-field irradiation.16 However, it is not clear whether concomitant chemotherapy and extended-field radiotherapy are superior to concomitant chemotherapy and pelvic radiotherapy. An important consideration is the toxicity of concomitant chemotherapy and extended-field radiotherapy. Some investigators have reported their experience using concomitant chemotherapy and extended-field radiotherapy,17, 18 but they did not employ high–dose-rate brachytherapy. High–dose-rate brachytherapy is gaining rapidly in popularity in the treatment of cervical carcinoma because it has many advantages over low–dose-rate brachytherapy, for both patients and health care workers.19 A prospective, randomized trial20 established that high–dose-rate brachytherapy was at least as safe and effective as traditional low–dose-rate brachytherapy, and Sood et al. have confirmed this.21 In the current study, the authors report their experience with extended-field radiotherapy and high–dose-rate brachytherapy with and without concomitant chemotherapy.
MATERIALS AND METHODS
Between 1995 and 2001, 54 women with biopsy-confirmed carcinoma of the uterine cervix were treated by extended-field radiotherapy and high–dose rate brachytherapy with or without concomitant chemotherapy. Their median age was 50 years (range, 33–79 years). The histology was squamous cell carcinoma in 49 patients (91%) and nonsquamous cell carcinoma in 5 patients (9%). All patients were staged according to the staging system of the International Federation of Gynecology and Obstetrics (FIGO). Their workup usually included a medical history and physical examination, a complete blood count, evaluation of electrolyte concentration, liver function tests, a chest X-ray, a bimanual examination under general anesthesia, a cystourethroscopy, and a rectosigmoidoscopy. Each of the 54 patients had a pretreatment hemoglobin level of at least 10 g/dL (median, 11; range, 10–16 g/dL). The median size of the primary tumor was 7 cm (range, 3–10 cm). Ten patients (19%) had FIGO Stage IB disease, 1 patient (2%) had Stage IIA disease, 23 patients (43%) had Stage IIB disease, and 20 patients (37%) had Stage IIIB disease(Table 1). Seven patients (13%) underwent exploratory laparotomy. All of these patients had Stage I disease clinically and were believed to be suitable for curative surgery, but the extent of disease encountered at laparotomy was greater than anticipated (lymphadenopathy in four patients, unresectable disease in three patients) and surgery was, therefore, abandoned.
Table 1. Characteristics of the Patients Treated with Extended-Field Radiotherapy and High–Dose-Rate Brachytherapy without or with Concomitant Cisplatin
Details of our treatment technique have been published elsewhere.21 Each patient received external beam radiation therapy to the pelvis and the paraaortic lymph nodes using four fields (anterior, posterior, and two laterals). All patients received 45 grays (Gy) to the pelvis and the paraaortic lymph nodes in 1.8-Gy fractions delivered once a day. The superior border of the field was located at the lumbar 1–2 disc space. The inferior border extended at least 4 cm inferior to the cervix (marked by radioopaque markers) or to the bottom of the obturator foramen, whichever was more inferior. For the AP-PA fields, the lateral borders extended 5 cm to each side of the midline in the paraaortic region and, in the pelvis, 2 cm lateral to the widest true pelvic dimension. For the lateral fields in the paraaortic region, the anterior border was 3 cm anterior to the anterior borders of the bodies of the vertebrae and the posterior border was located at the midvertebrae bodies. In the pelvis, the anterior border included the symphysis pubis, whereas the posterior border was located at the sacral 2–3 junction or, in the case of primary tumors measuring 4 cm or larger or Stage IIIB disease, it included the entire sacrum. An attempt was made to shield as much of the bowel as feasible using custom-designed shielding blocks. Contour-based treatment plans were used at the central axis, midpelvis, and 5 cm superior to the central axis. The dose was prescribed to the clinical target volume and up to 10% variation was considered acceptable. For patients with disease extension to the parametrium or the pelvic side wall(s), extended-field radiotherapy was followed by a parametrial boost (using AP/PA fields with midline shielding) to a median dose of 9 Gy (range, 5.4–14.40 Gy).
After external beam radiation therapy, each patient also underwent two applications of high–dose-rate brachytherapy, 1 week apart, with a Microselectron-HDR (Nucletron, Columbia, MD). The details have been published elsewhere, including a detailed discussion of the rectal and bladder doses.21 Of the total 108 applications of high–dose-rate brachytherapy, 107 consisted of tandem and colpostat insertions and 1 consisted of a tandem with interstitial needles in the parametrium. Briefly, dosimetry was performed before each application using two orthogonal radiographs with “dummy” sources in the applicators. The isodoses were plotted and the doses to Point A and the bladder were calculated according to recommendations from the International Commission on Radiation Units and Measurements.22 The rectal dose was calculated at multiple rectal points using a flexible radioopaque rectal marker. Of the total 108 applications, a 9-Gy dose was delivered to Point A for 77 applications (71%) and a 9.4-Gy dose was delivered for 26 applications (24%). A dose of 8 Gy was delivered for three applications whereas doses of 9.6 and 10 Gy were delivered for one application each. The total rectal dose from both applications of high–dose-rate brachytherapy did not exceed 11.2 Gy in any patient, whereas the total bladder dose did not exceed 15.8 Gy.21
Forty-four (81%) of the 54 patients received concomitant chemotherapy, whereas 10 (19%) either refused chemotherapy or were deemed medically ineligible after workup. Chemotherapy (cisplatin 20 mg/m2/day for 5 days) was administered during the first and the fourth weeks of external beam radiation therapy, as well as once following the second application of high–dose-rate brachytherapy. Colony-stimulating factors were generally not administered to these patients. However, blood transfusions were administered when necessary to maintain the hemoglobin level above 10 g/dL. The characteristics of the patients who did or did not receive chemotherapy are compared in Table 1.
The patients were followed by a radiation oncologist and a gynecologic oncologist every 3 months during the first 2 years, and every 6 months thereafter. Toxicity was graded according to the National Institutes of Health/National Cancer Institute Common Toxicity Criteria.
Of the 54 patients, 6 have died at this writing, 7–30 months after irradiation. All of them had experienced pelvic or distant recurrences. The survivors have been followed for a median of 28 months (range, 12–70 months), 54 months for the patients treated without chemotherapy and 22 months for those who received chemotherapy. Nine of the 10 patients treated without chemotherapy are alive compared with 39 of the 44 patients who received chemotherapy. Figure 1 shows the observed survival rates of the patients treated without/with chemotherapy.
None of the patients died during or within 1 month after treatment. However, 42 of the 54 patients (78%) suffered from Grade 2–4 acute toxicity. Among the 10 patients treated without chemotherapy, only 1 (10%) developed acute toxicity. Of the 44 patients receiving chemotherapy, 34 (77.5%) suffered from Grade 3–4 hematologic toxicity (Table 2), 11 (25%) from Grade 3 gastrointestinal toxicity (nausea, emesis, dehydration), and 3 (7%) developed deep venous thrombosis that necessitated anticoagulant treatment) (Table 3).
Table 2. Acute Hematologic Toxicity after Chemotherapya
Worst toxicity recorded, some patients developed more than one type of toxicity.
Table 3. Acute Nonhematologic Toxicity after Chemotherapya
All were Grade 3. There was no Grade 4 toxicity.
Deep venous thrombosis
The planned three cycles of chemotherapy were given to 41 (93%) of the 44 patients. The most severe hematologic toxicity developed after the third cycle of chemotherapy. For two patients (4.5%), the third cycle of chemotherapy was omitted by the physician (due to severe hematologic toxicity) and it was refused by one additional patient (2%). Administration of chemotherapy was delayed a median of 26 days (range, 20–30 days) for five patients (11%). The reasons for the delay included toxicity in four patients and a bone fracture unrelated to the treatment in the fifth patient. Twenty-five (57%) of the 44 patients who received chemotherapy required at least one blood transfusion during their treatment to maintain their hemoglobin levels above 10 g/dL. Blood transfusion was not required for any of the 10 patients treated without chemotherapy.
Four of the 54 patients had local recurrence in the pelvis. There have been no disease recurrences in the paraaortic region. Among the 10 patients treated with radiotherapy alone, there have not been any disease recurrences in the pelvis. Among the 44 patients receiving chemotherapy, 4 have had disease recurrence in the pelvis (2 had Stage IB and 2 had Stage IIIB disease). The actuarial rate of local control at 2 years was 100% among the patients treated without chemotherapy and 92% among those receiving chemotherapy (Fig. 2).
Two of the 54 patients developed disease recurrence at distant sites, including 1 (with Stage IIIB disease) who had received chemotherapy and 1 (with Stage IIB disease) who had not. The actuarial rate of no distant metastases was 90% among the patients treated without chemotherapy and 95% among those receiving chemotherapy (Fig. 3).
Two of the 54 patients developed Grade 2 or worse late toxicity. Among the 10 patients treated without chemotherapy, 1 developed Grade 2 rectal bleeding 18 months after irradiation. She was managed conservatively and the bleeding stopped after 6 months. Among the 44 patients receiving chemotherapy, 1 developed Grade 3 intestinal obstruction due to adhesions and required a laparotomy 21 months after radiation (she was one of the seven patients who had undergone a laparotomy before irradiation). The actuarial incidence of late toxicity was 10% among the patients treated without chemotherapy and 6% among those receiving chemotherapy (Fig. 4).
For many years, the standard treatment for patients with locally advanced carcinoma of the cervix was external beam irradiation to the pelvis and low–dose-rate brachytherapy. The standard of care has changed in recent years, thanks to several large, prospective, randomized clinical trials. It has been long recognized that many patients with locally advanced carcinoma of the cervix harbor occult paraaortic metastases.1–7 In their randomized trial, Haie et al.8 reported in 1988 that prophylactic paraaortic irradiation significantly reduced the incidence of paraaortic and distant metastases among patients whose pelvic disease was controlled. In 1995, in a prospective, randomized study, Rotman et al.9 reported that paraaortic irradiation improved the overall survival and reduced distant metastases. Pioneers in the field found that extended-field irradiation was associated with greater toxicity, especially gastrointestinal toxicity, than pelvic irradiation.8, 9 However, that may not be so with modern techniques.15–17 Following the publication of the Rotman et al. study,9 our group began administering extended-field irradiation to patients with locally advanced carcinoma of the cervix. More recently, randomized studies supported the value of cisplatin-based chemotherapy for patients with locally advanced carcinoma of the cervix.10–15 The use of concomitant chemotherapy with pelvic irradiation in locally advanced carcinoma of the cervix has led to a reexamination of the role of extended-field irradiation.
This is the first report of extended-field radiotherapy with concomitant chemotherapy and high–dose-rate brachytherapy in patients with cervical carcinoma. The preliminary results are encouraging (Figs. 1–3), with an actuarial survival rate of 80%, a local control rate of 85%, and a metastasis-free rate of 95%, albeit with limited follow-up. The incidence of late complications is also low (Fig. 4). The results are similar among the 10 patients treated without concomitant chemotherapy. However, the number of patients is much too small to draw any conclusions. Furthermore, Morris et al.16 conducted a prospective, randomized trial that compared extended-field radiotherapy without chemotherapy versus pelvic radiotherapy plus concomitant chemotherapy. They demonstrated convincingly that the survival rate was better among the latter group of patients. The use of brachytherapy in the Morris et al. study,16 however, was applied differently compared with the current study.
As expected, the authors observed a dramatic difference in the incidence of serious acute toxicity between the patients who did and did not receive chemotherapy. The majority of patients (91%) receiving chemotherapy suffered from serious (albeit nonfatal) acute toxicity, compared with only 1 in 10 treated without chemotherapy. The acute toxicity among the patients in the current study who received chemotherapy appears to be more severe than that observed by other investigators, who reported Grade 3–4 acute toxicity in fewer than 50% of their patients whether they administered concomitant chemotherapy and pelvic radiotherapy10–16 or concomitant chemotherapy and extended-field radiotherapy.17, 18 Table 4 summarizes the acute toxicities reported in these studies. Acute hematologic toxicity was observed in 2–37% of the patients and acute gastrointestinal toxicity was observed in 4–18%. We also observed a 7% incidence of deep venous thrombosis, which was not reported by the others.
Table 4. Reported Complication Rates in Patients Treated with Concomitant Chemotherapy and Irradiation
The dose of cisplatin among these studies ranged from 40 to 75 mg/m2/week. It is possible, therefore, that the greater acute toxicity among the patients in the current study was due to the higher dose of cisplatin used (100 mg/m2). In the only study that used high–dose-rate brachytherapy with pelvic irradiation and cisplatin,23 the incidence of late complications was high (28%), but details of the the rectal and bladder doses were not reported. Among our patients, the total rectal dose from high–dose-rate brachytherapy did not exceed 11.2 Gy, nor did the bladder dose exceed 15.8 Gy.
In the current study, a regimen of extended-field radiotherapy with concomitant cisplatin and high–dose-rate brachytherapy produced substantial acute nonfatal toxicity, but its long-term toxicity was low and the preliminary tumor control excellent, albeit with limited follow-up. Whether these results are truly superior to those of pelvic radiotherapy with concomitant chemotherapy, or the other regimens of extended-field radiotherapy with concomitant chemotherapy, can only be determined by prospective, randomized trials.