|Mitchell Morris, Patricia J. Eifel, Jiandong Lu, Perry W. Grigsby, Charles Levenback, Randy E. Stevens, Marvin Rotman, David|
|Abramson Cancer Center of the University of Pennsylvania|
| Last Modified: November 1, 2001
Reviewers: Kenneth Blank, MD
The results of this trial will be reported in the April 15, 1999 issue ofthe New England Journal of Medicine. Because of the implications of thisarticle for public health, the article has been released early and ispublished in full on the Internet.
Materials and Methods
Women of any age with squamous cell carcinoma, adenosquamous oradenocarcinoma of the cervix and International Federation of Gynecology andObstetrics (FIGO) stage IIB through IVA were eligible for enrollment.Patients with FIGO stage IB or IIA cervical cancer were eligible if thetumor measured 5cm or more in greatest dimension or had spread to thepelvic lymph nodes. Exclusion criteria included a performance status lessthen sixty, para-aortic lymph node metastasis, prior chemotherapy orradiotherapy, or a transperitoneal staging procedure.
All women underwent complete staging, including cystoscopy, proctoscopy andeither lymphangiogram or retroperitoneal lymph node biopsy. Eligiblepatients were randomized to receive either external beam radiotherapy tothe pelvic and para-aortic areas (termed extended field radiotherapy) andintracavitary radiation or external beam radiotherapy to the pelvis andintracavitary radiation concurrent with chemotherapy.
External beam radiotherapy was delivered at a dose of 180cGy per day to atotal dose of 4500cGy using either anterior-posterior fields, or a fourfield technique (anterior, posterior and two lateral fields). The upperborder of the field was set at L4-5 for pelvic radiotherapy and L1-2 forextended field radiotherapy. Intracavitary radiation was delivered in twosessions, the first before or during external beam radiotherapy and thesecond within two weeks after completion of external beam radiotherapy. Thetotal treatment time was kept under eight weeks whenever possible. At least8500cGy was delivered to point A (a reference point located two centimeterssuperior and two centimeters lateral to the cervical os).
Patients randomized to the chemotherapy arm received three cycles ofchemotherapy beginning on day one of external beam radiotherapy with atleast one cycle administered during intracavitary radiotherapy.Chemotherapy consisted of two drugs, cisplatin and 5-fluorouracil.Cisplatin was dosed over four hours at 75mg per square meter of bodysurface area and 5-fluorouracil was administered over 96 consecutive hours,one gram each day.
Between 1990 and 1997, 403 patients were enrolled in the study, fifteen ofwhom were ineligible mostly due to lack of adequate evaluation of thepara-aortic lymph nodes. 193 patients were assigned to receive radiotherapyalone and 195 combined modality therapy. The two groups had similar ratesof treatment protocol violations. Minor deviations occurred inapproximately 80% of patients in both groups and major, "but acceptable" ,deviations occurred in 11 and 9 percent of patients in the combinedmodality therapy group and radiotherapy group, respectively. In each studyarm the median total duration of treatment was 59 days and the median doseto point A was 8900cGy.
In the combined modality group, four patients did not receive anychemotherapy but were included in this group on the intent to treatanalysis. In total, 159 (81%) and 133 (68%) of patients received two andthree cycles of chemotherapy. The primary reason for not receiving thescheduled chemotherapy was refusal to continue (17 patients).
Estimated rates of survival at five years revealed a significant differencefavoring the group of patients receiving combined chemotherapy andradiotherapy. Specifically, Kaplan-Meier survival plots estimate that 73%of patients receiving combined treatment were alive at five years comparedto 58% in the radiotherapy alone group. Disease -free survival at fiveyears was also significantly higher in those receiving combined treatment(67% vs. 40%).
Administration of chemotherapy impacted on both local and distant diseasecontrol. The rates of distant and local relapse in the combined therapygroup (14% and 10%) were significantly lower than those rates in theradiotherapy alone group ( 33% and 35%).
The incidence of moderate and severe side effects was higher in patientsreceiving combined modality treatment, but the majority of these eitherresolved spontaneously or responded rapidly to medical treatment. Lateeffects were similar in both groups.
Previous studies examining the role of chemotherapy in addition toradiotherapy in the treatment of high risk cervical cancer have beeninconclusive. This study is unique in utilizing higher doses of bothradiotherapy and chemotherapy compared to earlier studies, and one cycle ofchemotherapy was administered during intracavitary radiation, whichdelivers 25% of the central radiation dose. In addition, an emphasis wasplaced on completing treatment within eight weeks- a known good prognosticfactor.
Delivering radiotherapy to the para-aortic lymph nodes has beencontroversial. A randomized prospective study by the RTOG published inJournal of the American Medical Association in 1995, found a survivaladvantage to including the para-aortics in the radiation field compared topelvic radiotherapy alone. The trial reported here provides evidence thatwhen no disease is evident in the para-aortic lymph nodes (either onlymphangiogram or biopsy) radiotherapy limited to the pelvis combined withchemotherapy provided better disease control than pelvic and para-aorticradiotherapy. Whether combining extended field radiotherapy with chemotherapywould improve on the results of the combined arm reported in this trial isunknown. However, a pilot study of hyperfractionated extended fieldradiotherapy with the same chemotherapy used in this trial was associatedwith severe acute toxicity. The authors conclude those patients with highrisk cervical cancer should be treated with pelvic radiation concurrentwith cisplatin and 5-fluorouracil.