- © 2007 by American Society of Clinical Oncology
Multimodality Therapy for Locally Advanced Cervical Carcinoma: State of the Art and Future Directions
- From the Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, Chao Family Comprehensive Cancer Center, University of California-Irvine Medical Center, Orange, CA; and the Department of Radiation Oncology, Fred Hutchinson Cancer Research, University of Washington, Seattle, WA
- Address reprint requests to Bradley J. Monk, MD, FACS, FACOG, Division of Gynecologic Oncology, Department of Obstetrics/Gynecology, Chao Family Comprehensive Cancer Center, University of California-Irvine Medical Center, Building 56, Rm 262, 101 The City Dr, Orange, CA 92868; e-mail: bjmonk{at}uci.edu
Abstract
Globally, cervical cancer is the second most common cause of cancer-related mortality among women causing approximately 234,000 deaths annually among developing countries and killing 40,000 in developed nations. Most of these deaths occur in women with bulky or locally advanced cervical cancer, International Federation of Gynecology and Obstetrics (FIGO) stages IIB through IVA, when lesions are not amenable to high cure rates with surgery or radiation (RT). The standard prescription for RT used to treat locally advanced cervical cancer has been dictated by common practice and patterns of care studies. In contrast, the addition of concomitant chemotherapy to RT has been studied in a number of randomized prospective trials, which are discussed in detail. When added to RT, cisplatin reduces the relative risk of death from cervical carcinoma by approximately 50% by decreasing local/pelvic failure and distant metastases. In 1999, weekly intravenous cisplatin at 40 mg/m2 for 6 weeks in combination with RT was established as a new standard for the treatment of locally advanced cervical carcinoma. More recently, this recommendation has been expanded to include women with FIGO stage IB2 lesions as well as those with bulky stage IIA cancers. This monograph reviews the state of the art in treating locally advanced cervical cancer with combined chemotherapy and RT and discusses clinical and pathologic prognostic factors that impact cure. Quality of life during and after multimodality therapy is considered as well as ongoing clinical trials and future directions.
INTRODUCTION
Cervical cancer is second only to breast cancer in being the most common cancer among women worldwide. Globally, cervical cancer is also the second most common cause of cancer-related mortality causing approximately 234,000 deaths annually among developing countries yet only killing 40,000 women in developed nations.1 The discrepancy in cervical carcinoma–related mortality between developing and developed countries is a direct result of poor screening (Pap testing) in low resource settings, and it is hoped that widespread vaccination against the human papillomavirus, which is associated with 99.7% of invasive cervical cancers, will dramatically reduce the morbidity and mortality of this highly preventable cancer.2 In the United States, the majority of cervical carcinoma patients are diagnosed with early-stage disease. Among the 13,458 staged patients with cervical carcinoma registered by the Surveillance, Epidemiology, and End Results program between 1973 and 1987, 71% were diagnosed with International Federation of Gynecology and Obstetrics (FIGO) stage I-IIA tumors.3 Most of these women with early lesions are cured with surgery or radiation (RT) alone. However, patients with more advanced lesions are at greater risk of recurrence and account for the majority of cervical cancer deaths.4
The standard prescription for RT used to treat bulky (stage IB2) or locally advanced cervical cancer generally considered to be FIGO stages IIA through IVA (Fig 1) has been dictated by common practice and patterns of care studies.5-8 In contrast, the addition of concomitant chemotherapy to RT has been studied in a number of randomized prospective trials discussed in detail herein. When added to RT, cisplatin reduces the relative risk of death from cervical carcinoma by approximately 50% decreasing local/pelvic failure and distant metastases. In 1999, weekly intravenous cisplatin at 40 mg/m2 for 6 weeks in combination with RT was established as a new standard for the treatment of locally advanced cervical carcinoma.9-11 This article reviews the state of the art in treating locally advanced cervical cancer with combined chemotherapy and RT and discusses clinical and pathologic prognostic factors that impact cure. Quality of life during and after therapy is considered as well as ongoing clinical trials and future directions.
STAGING OF LOCALLY ADVANCED CERVICAL CARCINOMA
Staging describes the degree or severity of an individual's cancer based on the extent of the original (primary) tumor and/or the extent of spread throughout the body. Staging is important for four reasons: it helps clinicians plan a patient's treatment; it is used to estimate a patient's prognosis; it helps researchers and health care providers exchange information about patients; and finally, it provides a common language for evaluating the results of clinical trials and comparing the results of different studies. The TNM system is the most commonly used cancer staging system and has been accepted by the International Union Against Cancer and the American Joint Committee on Cancer. However, gynecologic cancers have traditionally been staged according to the FIGO system, which has its origin from the League of Nations first published in 1920.12 In 1954, FIGO was entrusted with completing what has been called the Annual Report on the Results of Treatment in Gynecological Cancer.12A With it has come the responsibility for overseeing the staging of gynecological cancers and has formed the foundation for staging cervical tumors. Initially, the FIGO system of classification was based on clinical examination, and all staging of gynecologic cancers has transitioned to surgical staging except for cervical cancer, which has remained clinically oriented. Since carcinoma of the uterine cervix grows locally and may extend in continuity to the paracervical (parametrial) tissues and to the pelvic organs and regional lymph nodes, only later metastasizing to distant structures, the essence of clinical staging of cervical cancer is careful physical examination (Fig 1). The following examinations are permitted by FIGO in the staging of cervical cancer: palpation, inspection, colposcopy, endocervical curettage, hysteroscopy, cystoscopy, proctoscopy, intravenous urography, and x-ray examination of the lungs and skeleton. Suspected bladder or rectal involvement should be confirmed by biopsy and histologic evidence. Conization or amputation of the cervix is regarded as a clinical examination. Findings of optional examinations (eg, laparoscopy, ultrasound, computed tomography [CT] scanning, magnetic resonance imaging [MRI], and positron emission tomography [PET] scanning) are of value for planning therapy, but because these are not generally available globally and the interpretation of results is variable, the findings of such studies should not be the basis for changing the clinical staging. Fine needle aspiration of scan-detected suspicious lymph nodes is recommended and may be helpful in treatment planning, but is also not used in the FIGO staging system, even though nodal spread is perhaps the most important factor in determining prognosis and treatment. When there is doubt as to which stage a particular cancer should be allocated, the earlier stage is mandatory.4,12,13 While the traditional staging work-up for these patients includes excretory urography, barium enema, examination under anesthesia, cystoscopy, and proctoscopy, there is now good evidence that CT scanning with intravenous contrast and office examination and biopsy are sufficient, with exam under anesthesia, cystoscopy and proctoscopy reserved for those few patients in whom clinical or imaging data suggest a higher risk of involvement.14 Indeed, patterns of care studies,15 as well as prospective clinical trials,16 have shown these invasive procedures to be declining with increased use of more sensitive and sophisticated imaging techniques to detect the extent of primary cervical cancers. In evaluating the pelvic extent of disease, MRI appears to be superior to CT and clinical examination.17,18
PET
Functional imaging methods using radioactive tracers that accumulate in abnormal tissue, such as PET, appear more sensitive in detecting metastases from solid tumors than imaging techniques that simply identify anatomic deviations.19,20 With PET, the most commonly used radioisotope tracer is [18F]flurodeoxyglucose (FDG), a glucose analog that is preferentially taken up by and retained within malignant cells. Depending on the area or organ under study, baseline glucose metabolism may be low, further establishing the difference between normal background tissue and tumor. Thus, compared with structural imaging techniques, FDG-PET has the potential to be a more accurate technique for diagnosis, staging, and treatment decisions in oncology. In a review of 15 studies published between 1966 and 2003 investigating the role of PET to assess the extent of disease among women with cervical cancer, the pooled sensitivity and specificity of PET to detect aortic node metastasis was 0.84 (95% CI, 0.68 to 0.94) and 0.95 (95% CI, 0.89 to 0.98), respectively. The pooled sensitivity and specificity for the detection of pelvic node metastasis was 0.79 (95% CI, 0.65 to 0.90) and 0.99 (95% CI, 0.96 to 0.99), respectively. This compared favorably to the ability of MRI to detect pelvic node metastasis, which only had a pooled sensitivity of 0.72 (95% CI, 0.53 to 0.87) and a pooled specificity of 0.96 (95% CI, 0.92 to 0.98).21 The inferiority of MRI compared with PET in detecting lymphatic spread has been confirmed by a recent prospective trial.22 Because of the superiority of PET to MRI and/or CT in detecting nodal metastasis, the United States Centers for Medicare and Medicaid Services has issued a national coverage determination paying for FDG-PET imaging for the detection of pretreatment metastases (ie, staging) in newly diagnosed cervical cancer subsequent to conventional imaging that is negative for extrapelvic metastasis.23
Surgical Staging
The most accurate method of detecting lymphatic metastasis is clearly surgical excision of the relevant nodal chains.24,25 This approach has two theoretical advantages over imaging assessments of the retroperitoneum. First, it is much more accurate in detecting nodal spread and only limited by the extent of the resection and the accuracy of the histological assessment of the resected lymph nodes. Second, the removal of grossly involved pelvic or aortic nodes might increase cure by rendering otherwise resistant bulky disease either on the pelvic sidewall or in the aortic area sensitive to the affects of chemotherapy and RT after bulk reducing surgery.26 This theoretical advantage of bulk-reducing surgery has been more thoroughly studied before the new era of multimodality therapy when RT was used alone but is probably still relevant today.
Surgical staging may be accomplished using either a transperitoneal approach or a retroperitoneal approach and may be performed using either a laparotomy or laparoscopy. The laparotomy retroperitoneal approach appears to be associated with a reduced risk of bowel obstruction compared with the transperitoneal laparotomy procedure.25 Laparoscopic surgical staging is feasible,27 but one randomized trial suggested that laparotomy allows removal of more pelvic nodal tissue although the clinical significance of this is unknown.28 Laparoscopy appears less morbid than open surgery.27,29
The trend over the past decade has been to do less surgical staging with the advent of more accurate imaging modalities, such as MRI and PET.30 Indeed, the Gynecologic Oncology Group (GOG) stopped mandating negative surgical staging of the aortic nodes as entry criteria into its clinical trials of chemotherapy and RT for locally advanced cervical cancer after the publication of GOG protocol 120.31 Interestingly, because the detection of otherwise occult aortic nodal metastasis through surgical staging allows extension of the pelvic radiation port to cover the periaortic area and thereby potentially improve cure rates, patients undergoing pelvic radiation alone who are thought to have negative aortic nodes by imaging should theoretically have a worse outcome than those found to have negative nodes after surgical staging. This has been suggested by comparing the results of GOG trials before and after the transition from surgical staging to tomographic imaging (CT/MRI) staging. Stage III patients, in whom the risk of aortic nodal spread is highest, treated on GOG protocol 120, where surgical staging was required, had a significantly better outcome than those treated in a similar fashion (once per week cisplatin during pelvic RT) in another GOG trial (protocol 165)32 where surgical staging was optional and only performed on a minority of patients (Fig 2). 32 Importantly, this was before the widespread use of PET, but suggests the increased accuracy of surgical staging compared with CT or MRI. However, cross-study comparisons are difficult to interpret, and one must weigh the costs and operative morbidity of surgical staging in choosing the appropriate staging technique.
RT FOR LOCALLY ADVANCED CERVICAL CARCINOMA: STATE OF THE ART
External-Beam RT
Cervical cancer found to be metastatic beyond the pelvis is beyond the scope of this review as it is treated using either extended-field radiation with chemotherapy or palliative chemotherapy alone emphasizing the importance of accurate pretreatment staging.33,34 Although women with bulky stage I lesions (> 4 cm, FIGO stage IB2) or bulky (> 4 cm) FIGO stage IIA lesions can be successfully treated with either neoadjuvant chemotherapy followed by either surgery or chemotherapy with RT or radical hysterectomy with pelvic and aortic lymphadenectomy followed by tailored postoperative therapy, the world standard is shifting toward including these two groups of patients into those more broadly entitled locally advanced cervical cancer, which has heretofore only included those patients with FIGO stages IIB, IIIB and IVA cancers. This is because the chance of bulky stage IB2 or bulky IIA tumors (defined as > 4 cm in diameter) being associated with surgical and pathologic risk factors that increase the risk of recurrence is so high that chemotherapy with RT after neoadjuvant chemotherapy or after radical surgery is frequently required based on carefully done prospective cooperative group trials.35-38 This makes the cost and morbidity prohibitive of three treatment modalities (surgery, RT, and chemotherapy) without clear evidence of therapeutic benefit compared with two treatment modalities (chemotherapy and RT). For example, a review of consecutive cases of women with stage IB2 lesions initially treated with surgery at a single institution found that 52% required at least postoperative RT and another 36% required postoperative RT and chemotherapy. This study showed that if a primary surgical approach is undertaken for stage IB2 lesions, only 12% will have been adequately treated with surgery alone.39 Indeed, the GOG now includes those with stage IB2 and stage IIA lesions more than 4 cm in their clinical trials of locally advanced cervical cancer. Although women with FIGO stage IIIA lesions involving the lower third of the vagina without sidewall involvement or hydronephrosis are considered by some to have locally advanced cervical cancer, the alteration of the pelvic RT port to include the entire vagina and frequently the vulva as well as the increased risk of groin metastasis causes most experts to exclude this small group of patients from reviews of locally advanced cervical cancer.
Although primary pelvic exenteration may be considered for stage IVA disease not extending to the pelvic sidewall, particularly if a vesicovaginal or rectovaginal fistula is present, standard primary treatment for stage IB2, IIA (> 4 cm), IIIB, and IVA cervical cancer without evidence of spread beyond the pelvis is chemotherapy (discussed in the Concurrent Chemoradiation section) and pelvic external-beam radiation and intracavitary brachytherapy. Those with nodal spread to the common iliac lymph nodes are treated with extended-field radiation similar to those with biopsy-proven aortic node metastasis because the risk of occult periaortic spread is so great. The fractionation and schedule of pelvic radiation in this setting has not been rigorously tested in randomized trials and is the result of common practice and patterns of care studies. Patients are generally treated with approximately 50 Gy using the four-field box technique using the anterior-posterior and lateral ports shown in Figures 3 and 4. A typical prescription would be to use 25 daily fractions of 1.8 Gy Monday through Friday. Patients are generally treated with bladder distention and bowel exclusion devices (eg, belly boards) to reduce the risk of intestinal and urinary toxicity. The importance of brachytherapy cannot be overemphasized because it delivers much higher RT doses to the primary tumor than the shrinking field technique.
All patients with stage IIB, IIIB, and IVA disease usually receive a parametrial boost. Patients with stage IB2 and IIA disease do not always receive a parametrial boost but may at the discretion of the treating oncologist. The involved parametrium at presentation is usually treated with an additional 5.4 to 9.0 Gy in three to five fractions of 1.8 Gy per fraction given anterior-posterior/posterior-anterior daily after the completion of the whole pelvic radiation between the two implants if two are planned or immediately after the single implant if only one implant is used.
Brachytherapy
Since Margaret Cleaves of New York City first introduced the use of radium to treat cervical cancer in 1903, intracavitary brachytherapy has gained steady acceptance as the mainstay to achieving local control in the treatment of locally advanced cervical cancer.40 Intracavitary brachytherapy has been profoundly impacted by the development of systems that have attempted to combine empiricism with a more practical scientific and systematic approach. A dosimetric system refers to a set of rules specific to a particular applicator type, radioactive isotope, and distribution of the radioactive sources in an applicator to deliver a defined dose of RT to a designated treatment volume. Within each system, the dose, timing, and administration of the radioactive sources are outlined ensuring consistent implementation of a particular prescription of brachytherapy in treating cervical lesions. Over the next half century after Dr Cleaves’ report, history saw an evolution of systems in Europe named after the cities in which they were developed—Stockholm, Paris, and Manchester.4 More recently, the Fletcher system was created at M.D. Anderson Cancer Care Center (Houston, TX) and an article describing its development was published in 1953.41 Although the initial work at M.D. Anderson was done before computerized dosimetry, which became available in the 1960s, Fletcher aspired to improve survival and reduce toxicity by using formulas and measurements on orthogonal x-rays to calculate the energy absorbed at various points in the pelvis rather than the more empiric arbitrary applications of milligrams of radium for a prescribed number of hours as reported in earlier systems, which were based purely on clinical experience. Two such points, A and B, have been widely used in calculating and communicating the dose of RT in the treatment of cervical cancer (Fig 5).
Today, brachytherapy is prescribed based on the results of patterns of care studies42 and working groups.43 Basically, afterloading applicators and computerized dosimetry are used to generate isodose curves treating the cervical tumor, pelvic wall (nodes), bladder, and rectum to prescribed doses. The bladder and rectum are treated to tolerance trying not to exceed approximately 75 Gy and 70 Gy respectively, using measurements in Figure 6. Vaginal packing and the largest vaginal ovoids possible reduce bladder and rectal toxicity. When combined with external-beam RT as described earlier, a typical intracavitary brachytherapy application would deliver an additional 35 to 40 Gy to point A in one or two insertions using low-dose rate (LDR) sources equivalent to radium, such as cesium.
A number of controversies currently exist in the use of brachytherapy to treat locally advanced cervical cancer. First, image-based dosimetery using MRI is becoming increasingly common but requires prospective study and is not widely available.44,45 Second, although most agree that interstitial brachytherapy should be considered for patients with disease that cannot be optimally encompassed by intracavitary brachytherapy,43,46 at least one study has shown it to be inferior to intracavitary techniques.47 Finally, the use of high-dose rate (HDR) brachytherapy is becoming increasingly common since it allows outpatient brachytherapy applications without the risks of general anesthesia. Randomized prospective and retrospective studies have suggested overall statistically equivalent local control, overall survival, and complication rates between HDR and LDR.48 However, LDR may be preferable for large, bulky tumors. Retrospective studies of HDR and concurrent chemotherapy are limited but have demonstrated toxicity rates similar to those with LDR.49 Thus, selected patients with locally advanced cervical carcinoma who have an adequate response to external-beam RT and concomitant chemotherapy may be treated with HDR brachytherapy. Indeed the GOG has allowed HDR as part of their studies since protocol 165.32 One common prescription that includes careful monitoring of the dose to the healthy tissues includes five HDR insertions beginning during the last week of external-beam RT (after day 28). HDR techniques use a number of methods to calculate the healthy tissue tolerance doses, such as the linear quadratic model.50 Using the linear quadratic model, an LDR dose to the late responding tissues, such as rectum of 70 Gy, corresponds to a Gy3 of 120. Gy3 is a theoretical dose that is a biologically effective dose to the late responding tissues. A total dose of 75 Gy, such as the tolerance dose of the bladder, corresponds to a Gy3 of 130. These Gy3 terms are helpful in calculating the total doses to the healthy tissues when adjusting for the dose-rate effect. In order to stay below an LDR equivalent of 70 Gy to the rectum (120 Gy3) for five HDR insertions, including a 45 Gy contribution from the external-beam RT, the rectum should receive less than 4.1 Gy for each HDR fraction of 6 Gy (68% of the prescribed dose to point A). The dose to the bladder should be less than 4.6 Gy per HDR fraction of 6 Gy (77% of the prescribed dose to point A). As in LDR brachytherapy, every attempt should be made to deliver tumoricidal doses, even if the late responding tissues receive a slightly higher dose.50-52
CONCURRENT CHEMORADIATION
RT alone fails to control the progression of cervical cancer in 35% to 90% of women with locally advanced disease. Concurrent chemoradiation has been employed in the treatment of many cancers in an attempt to improve local control and eradicate distant metastases and has been successfully integrated into the therapeutic program of not only cervical carcinomas but also those of the head and neck and anal canal. Mechanisms of drug-radiation interaction leading to enhanced radiation kill may include modification of the slope of the dose-response curve, inhibition of sublethal damage repair, inhibition of recovery from potentially lethal damage, alterations in cellular kinetics, decrements in tumor volume leading to improved blood supply and tissue oxygenation, and increased radiosensitivity.
PIVOTAL TRIALS OF THE GOG, RADIATION THERAPY ONCOLOGY GROUP AND SOUTHWEST ONCOLOGY GROUP
Early investigations by the GOG in the field of multimodality therapy with chemotherapy and radiation included studies of hydroxyurea, hypoxic cell sensitizers (eg, the nitroimidazoles), and immune modulators (eg, bacille Calmette-Guerin and Corynebacterium parvum),53-63 with only hydroxyurea demonstrating sufficient activity to merit continued analysis against cisplatin-based regimens during the 1990s. Five phase III trials of concurrent chemoradiation have demonstrated a reduction in the risk of recurrence by up to 50% in patients with locally confined bulky or advanced-stage cervical cancer, regional spread, or high-risk features after hysterectomy (Tables 1 and 2). 31,37,65-68 Three studies compared radiotherapy alone with radiotherapy plus cisplatin-based chemotherapy, one of which addressed the prescription of adjuvant therapy after radical surgery for early-stage tumors and is therefore not pertinent to this discussion (Intergroup study: GOG protocol 109/Southwest Oncology Group [SWOG] protocol 87-97/Radiation Therapy Oncology Group [RTOG] protocol 91-12).39 Excluding patients with nodal involvement by CT scan, GOG protocol 123 evaluated the benefit of preoperative chemoradiation therapy (once per week cisplatin 40 mg/m2, maximal weekly dose of 70 mg) versus RT alone in patients with locally advanced disease confined to the cervix (ie, stage IB2).65 All patients underwent adjuvant hysterectomy. In this landmark study, the rates of both progression-free survival (P < .001) and overall survival (P = .008) were significantly higher in the combined therapy group at 4 years. Patients receiving radiosensitizing chemotherapy experienced higher frequencies of grade 3 and grade 4 adverse hematologic effects and adverse gastrointestinal effects.
Morris et al66 reported the results from RTOG protocol 90-01. In this study, the effects of pelvic radiation plus concurrent cisplatin and fluorouracil (FU) were compared with pelvic radiation plus extended-field RT. To our knowledge, this was the only trial to include chemotherapy during LDR brachytherapy. Eligibility requirements for this study differed from the previous GOG studies with the inclusion of patients with FIGO stage IB2-IIA tumors. The estimated 5-year survival rates were 73% versus 58%, respectively, for patients treated with chemoradiation therapy versus RT alone. A significant difference in disease-free survival was also seen in favor of the chemotherapy arm. The addition of chemotherapy to RT was effective in reducing both the frequency of local recurrences and distant metastases, with the latter observation refuting those detractors who claim that the benefit conferred by radiosensitizing chemotherapy is strictly a function of increasing the relative dose intensity of the radiation that can be delivered to the pelvis. These results have been sustained in an update of RTOG protocol 90-01 with 8 years of follow-up (Table 2).67
Two additional phase III trials have confirmed the superiority of cisplatin-based chemoradiation for the treatment of locally advanced cervical cancer. Whitney et al68 published the results of concurrent cisplatin plus FU and pelvic RT versus hydroxyurea plus pelvic RT in women with FIGO stage IIB-IVA disease who had undergone surgical staging and were found to have negative common iliac and aortocaval lymph nodes (GOG protocol 85). Among 368 eligible patients, the median follow-up time among survivors was 8.7 years. Disease progression occurred in 43% of patients randomly assigned to cisplatin plus FU versus 53% of patients randomly assigned to hydroxyurea. Progression-free survival was significantly better among patients treated with the combined chemotherapy regimen (P = .033), with 3-year survival rates of 67% (cisplatin and FU arm) versus 57% (hydroxyurea).
Rose et al31 reported the results from the three-arm GOG trial of pelvic RT plus concurrent: single-agent cisplatin versus cisplatin plus FU plus hydroxyurea versus hydroxurea alone (protocol 120). All patients had FIGO stage IIB-IVA cervical cancer with surgically confirmed negative common iliac and aortocaval lymph nodes. The median duration of follow-up was 35 months for 526 women included in the final analysis. Significant improvements in progression-free and overall survival were observed in patients randomly assigned to either cisplatin-containing arm. Effectively, the results from GOG protocol 85 and GOG protocol 120 were critical in supplanting hydroxyurea as the radiosensitizer of choice.
The combination of cisplatin plus FU was tested in three of the trials (RTOG protocol 90-01, GOG protocol 85, and GOG protocol 109/SWOG protocol 8797/RTOG protocol 91-12),39,66,68 with higher doses of cisplatin being used in the RTOG and Intergroup study (70 mg/m2) than in GOG protocol 85 (50 mg/m2). Because the progression-free survival at 6 years in GOG protocol 85 was shorter than in the other two studies (60% 6-year progression-free survival in GOG 85 v 67% 6-year progression-free survival in RTOG 90-01 and 80% 6-year progression-free survival in the Intergroup study) some experts have recommended using the higher dose of cisplatin if the combination is to be employed for radiosensitization. It should be noted, however, that in the long-term results of RTOG 90-01 that appeared 5 years after the original publication, progression-free survival at 8 years was down to 61%67; furthermore, the relatively higher progression-free survival observed in the Intergroup study clearly reflected the relatively lower-risk population under study (ie, not all tumors were locally advanced, and chemoradiation was used exclusively as an adjuvant).
In GOG protocol 120, the frequency of grade 3 to 4 leukopenia was significantly higher in the three-drug combination versus either single-drug regimen.31 Because treatment with cisplatin alone was equally effective and less toxic than treatment with the three-drug combination, attention was drawn toward using the single agent. To date, there have been no direct randomized comparisons of single-agent cisplatin to combined cisplatin and FU in this disease setting. A randomized comparison of once per week cisplatin versus protracted venous infusion (PVI) of FU was closed prematurely when a planned interim futility analysis indicated that PVI-FU/RT would not result in an improvement in progression-free survival compared with once per week cisplatin during RT (GOG protocol 165).32 Returning to GOG protocol 85 and GOG protocol 120, it should be noted that the enrollment criteria were the same (ie, stage IIB-IVA with negative para-aortic nodes); importantly, the progression-free survival at 6 years for the cisplatin plus FU arm of GOG-85 was equivalent to the progression-free survival at 4 years of the cisplatin alone arm in GOG-120 (ie, 60% across studies).31,68 Because the combination of cisplatin plus FU results in added toxicity, once per week single-agent cisplatin administered at 40 mg/m2 has emerged as the standard radiosensitizer in locally advanced cervical cancer. At present, radiosensitizing chemotherapy is recommended during that part of the treatment program when external-beam pelvic RT is administered. The GOG has studied concurrent chemotherapy in the setting of pelvic and extended-field RT (protocol 125; Table 1).69 There is very little experience with delivery of chemotherapy during brachytherapy, and most recommend against this practice, especially when HDR techniques are utilized as concurrent chemotherapy could theoretically cause a disproportionate increase in the effect of large radiation fractions on healthy tissues.
These pivotal phase III trials not only identified a significant survival advantage associated with the addition of concurrent chemotherapy, but were noteworthy in that the degree of benefit achieved with chemotherapy was remarkably similar for each of the four trials that studied chemoradiation for primary therapy. The results changed the standard of care for the treatment of locally advanced cervical cancer and formed the basis for the 1999 National Cancer Institute clinical announcement (practice alert) in our subspecialty.10
HEMOGLOBIN LEVELS: GOG PROTOCOL 191 AND THE SWOG ERYTHROPOIETIN TRIAL
Hypoxia occurs across a wide range of experimental and human malignancies. Through oxygen depravation, solid tumors become resistant to the cytotoxic activity of chemotherapeutic agents and DNA injury resulting from ionizing radiation. Anemia, which leads to hypoxia, may also reduce tumor sensitivity to RT and chemotherapy through indirect mechanisms involving proteomic and genomic changes. A vicious cycle then ensues, with increased basement membrane degradation and metastatic potential, loss of apoptosis and chaotic angiogenesis, all of which further increases tumor resistance.
Winter et al70 performed a retrospective review of 494 patients treated on two consecutive prospective GOG trials which included 278 FIGO stage II and 216 FIGO stage III/IV cervical tumors. Controlling for age, race, performance status, disease stage, tumor size, cell type, and duration of RT, mean hemoglobin levels during treatment were predictive of disease progression (P < .0001). When the 6-week treatment course was divided into 2-week periods (early, middle, and late), the analysis revealed hemoglobin values during the late period were the most predictive of disease recurrence and survival (P = .0289). The prognostic significance of pretreatment tumor oxygenation status is further underscored by studies that have shown that hypoxia (defined as oxygen tension ≤ 10 mmHg) is associated with lower overall and disease-free survival, greater recurrence, and less locoregional control in cervical carcinoma.
Early evidence from experimental and clinical studies suggests that the administration of recombinant human erythropoietin may enhance the effectiveness of RT and chemotherapy by increasing hemoglobin levels and ameliorating anemia in patients with disease-related or treatment-related anemia. In 2004, Lavey et al71 reported the SWOG erythropoietin (EPO) study in which 53 patients with stage IIB-IVA tumors and a hemoglobin between 8.0 and 12.5 g/dL received recombinant human erythropoietin three times per week and oral iron 2 weeks before and during chemoradiation. Seven patients developed deep venous thrombosis (Table 3). The 2-year progression-free survival and overall survival rates were 43% and 51%, respectively, with survival being significantly associated with the hemoglobin level at the completion of chemotherapy but not at baseline.
The GOG tested this hypothesis in a randomized manner through protocol 191, enrolling patients with stage IIB-IVA tumors and hemoglobin < 14 g/dL.72 Chemoradiation with and without EPO (40,000 units subcutaneous injection weekly) comprised the treatment arms. This study was closed early because of potential concerns for thromboembolic events associated with EPO (Table 2); however, the thromboembolic rate was not statistically higher with EPO use among the 109 patients accrued before closure. The exact risk of deep venous thrombosis with EPO during chemotherapy and RT remains undetermined, and whether maintaining a hemoglobin ≥ 12 g/dL improves local control, time to progression, or survival is also unknown.73-75 In summary, the retrospective nature of most studies, and the lack of definitive data from randomized trials, make the benefits of EPO or blood transfusions in this setting uncertain at the present time.
GOG PROTOCOL 219 AND OTHER RADIATION SENSITIZERS
Other agents studied outside of the GOG in the setting of concurrent chemoradiation included mitomycin, epirubicin, bleomycin, and vincristine (Table 2).71,76-88 Finally, the GOG is currently investigating the potential benefits of radiosensitizing chemotherapy using the combination of cisplatin (75 mg/m2 every 14 days) and tirapazimine (SRI International, Menlo Park, CA) in a randomized trial compared with once per week cisplatin (40 mg/m2) in women with locally advanced cervical cancer receiving pelvic RT (GOG protocol 219). Tirapazimine is a bioreductively activated, hypoxia-selective antitumor agent of the benzotriazine series that increases cisplatin cytotoxicity both in vitro and in vivo. Two dosing schedules of tirapazimine are being employed: 290 mg/m2 given on the days of cisplatin therapy, and 220 mg/m2 given on days 8, 10, and 12 of each 14-day cycle.89
CURE RATES IN LIEU OF PROGNOSTIC CLINICAL AND PATHOLOGIC FACTORS
The most powerful predictor of survival among patients treated for locally advanced cervical cancer is the extent of disease expressed as FIGO stage (discussed earlier). FIGO stage is based on clinical assessments and acts as a surrogate for more sensitive assessments of disease spread, such as surgery, MRI, and FDG-PET. When disease is confined to the pelvis after surgical staging and patients are treated with once per week cisplatin and pelvic RT, as in GOG protocol 120, the 4-year progression-free survival and overall survival rates for stage II patients are 64.2% and 68.1%, respectively, and 51.4% and 55.4% for stage III patients, respectively. The survival is much less for stage III patients when imaging (without FDG-PET) rather than surgery is used to assess aortic spread as in GOG protocol 165 with 4-year progression-free survival and overall survival being 37.7% and 42.7%, respectively (Fig 2).33
In 1991, Stehman and colleagues90 from the GOG conducted an analysis of prognostic variables from three GOG trials conducted between 1977 and 1985 (protocols 04, 24, and 56 discussed earlier). Multivariate analysis showed patient age, performance status, aortic lymph node status, tumor size, and pelvic node status to be significantly associated with time to progression. When modeling for survival, all these factors, as well as clinical stage and bilateral extension within the pelvis, were also significant.90 These findings have been confirmed after analyzing more contemporary GOG trials where once per week cisplatin was used with pelvic RT.33 Interestingly, in this multivariate analysis, stage, tumor grade, race, and age were all independently predictive of time to progression and overall survival in the new era of multimodality therapy (for all, P < .05). This study suggested that African American women can achieve the same, or better, level of survival compared with white women when treated according to rigorous protocols, and that other minorities including Asians and Hispanics have the best prognosis compared with other ethnic groups, apparently independent of other prognostic factors. This pooled GOG analysis also showed that the association between age and survival appears to be nonlinear, with women ages 51 to 60 years having the best prognosis (progression-free survival: hazard ratio, 0.57; P = .027; and overall survival: hazard ratio, 0.56; P = .031), compared with those ≤ 40 years. Finally, the GOG has also shown that women who smoke while on therapy have a worse outcome than those that do not smoke during chemotherapy and RT.91 Although squamous tumors and adenocarcinomas are the most frequent histologic subtypes of cervical cancers, the prognostic significance of these different cell types is not clear. However, the preponderance of data suggest that they are prognostically equivalent in the era of chemoradiation therapy.33 Indeed, the GOG does not separate these two groups in their trials of locally advanced cancer nor in their trials of advanced or recurrent cervical cancer. In contrast, neuroendocrine (small tumors) cancers have a worse prognosis stage for stage.92
Clearly, tumors that do not respond to initial therapy have a worse outcome. Besides clinical and radiographic evidence of persistent disease after primary therapy, FDG-PET can be used to assess the response to treatment. One study suggested that the 5-year survival was only 32% if persistent (in the irradiated region) abnormal FDG uptake in the cervix or lymph nodes was seen on average 3 months after therapy for locally advanced cervical cancer.93
The impact of overall treatment time (elapsed time from beginning RT until the last implant) on prognosis in locally advanced cervical cancer has been very controversial. Most studies have suggested that protracted radiation allows tumor regrowth and worsens prognosis in many solid tumors. In cervical cancer, prolonged RT may be frequently confounded by adverse tumor characteristics, and overall treatment time may be a proxy variable for other predictors of poor prognosis. However, women treated on GOG protocol 165 (where treatment was supposed to be completed in < 8 weeks) whose RT lasted longer than 8 weeks, demonstrated a worse progression-free survival (Fig 7) and overall survival.33 One might speculate that subtle variations in treatment time, such as those due to bad weather, holidays, or unfavorable tumor geometry, might impact prognosis very little but every reasonable effort should be made to treat patients on time.
QUALITY OF LIFE DURING MULTIMODALITY THERAPY AND SURVIVORSHIP
None of the prospective clinical trials discussed herein evaluated the role of therapy on quality of life nor the impact of treatment on survivorship; although, the impact of radical radiation on psychosocial, bladder, rectal, and sexual function has been well-documented.94-96 However, questionnaires have been developed and validated for this purpose.97 In one retrospective cohort, severe (grade 3 to 4) late toxicities in the rectum, bladder, small intestine, and subcutaneous tissue were 12.3% and 1.1%, 11.2% and 1.2%, 9.2% and 0.2%, and 23.1% and 1.2%, respectively. Vaginal adhesions were seen in 29.6% of patients and vaginal stenosis in 33.9% of patients.98 This underscores the need for prospective evaluation of quality of life and survivorship in protocols studying locally advanced cervical cancer.
In addition to late toxicities, physician-reported (probably lower than actual patient reported) acute combined grade 3 to 4 rectal and bladder toxicities of once per week cisplatin and pelvic RT treated on GOG studies are approximately 25% and 5%, respectively.33
As more and more acute and chronic treatment-related psychosocial and physical distress and dysfunction are identified, methods are desperately needed to reduce these adverse effects and toxicities. One prospective randomized trial demonstrated that daily amifostine could reduce the intestinal and bladder toxicity associated with pelvic RT.99 The GOG tried to do a study of vaginal length, elasticity, lubrication, and sexual function in women with stage IB2 cervical carcinoma (protocol 8003) but this study was closed in late 2005 due to insufficient accrual. The GOG is currently planning a prospective evaluation of palifermin (keratinocyte growth factor) to prevent skin toxicity.99 Although this is being primarily investigated in vulvar lesions, it should be relevant to FIGO stage IIIA cervical cancers where the vulva is frequently included in the irradiated field.
FUTURE DIRECTIONS
With the decreasing incidence of cervical cancer in developing countries, international collaboration is critical to the future of large prospective clinical and translational studies of locally advanced cervical cancer. Areas of research need to include improved assessments of extent of disease (staging); improved RT techniques, including better methods to deliver brachytherapy; improved chemotherapeutic agents to be administered with RT, including hypoxic cell sensitizers and targeted therapies100; a better understanding of the molecular mechanisms of cervical carcinogenesis and tumor progression allowing the identification of molecular markers of RT and chemotherapy resistance, as well as the identification of molecular markers of metastasis; and a better understanding of quality of life during therapy and survivorship after treatment allowing development of methods to improve quality of life, acute and chronic treatment-related toxicity, and survivorship. Some of these areas are discussed herein, such as better staging techniques, image-based dosimetry, and novel chemotherapy regimens used with RT, such as tirapazamine.
Intensity-modulated RT offers technical advantages over conventional external-beam RT that might prove clinically advantageous in the management of locally advanced cervical cancer. Intensity-modulated RT provides an opportunity to improve the therapeutic ratio by allowing a selective combination of healthy tissue dose reduction and/or concomitant integrated boost dose to the tumor. The clinical and biologic rationale for intensity-modulated RT in this setting is well-documented, and pertinent technical considerations, such as the delineation of relevant clinical and planning target, have been reviewed.101,102 Preliminary clinical observations of toxicity and tumor response have shown that intensity-modulated RT with concurrent cisplatin is feasible although further study is needed before it can be widely recommended.102 Alterations in anatomy due to organ motion and physiologic changes in the position of the bladder and intestine make this an investigational therapy at the current time.
CONCLUSION
If the financial and moral barriers to widespread vaccination against the human papillomavirus can be overcome, the worldwide incidence of cervical cancer will continue to decline over the next few decades, and ultimately, cervical carcinoma will become eradicated. In the interim, screening will attempt to identify precancerous lesions amenable to ablation or cancers in their earliest stages when cure rates are more than 90%. Other more locally advanced lesions are treated with multimodality therapy where cisplatin combined with external-beam RT and brachytherapy can achieve cure in more than 65% of women with stage II lesions and 55% of women with stage III lesions when the cancer is confined to the pelvis. Methods to improve quality of life and survivorship are desperately needed as more and more women with locally advanced disease are rendered disease free.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The author(s) indicated no potential conflicts of interest.
AUTHOR CONTRIBUTIONS
Conception and design: Bradley J. Monk
Manuscript writing: Bradley J. Monk, Krishnansu S. Tewari, Wui-Jin Koh
Final approval of manuscript: Bradley J. Monk
Footnotes
-
Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received January 18, 2007.
- Accepted April 20, 2007.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.
- 7.
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 12A.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.
- 52.↵
- 53.↵
- 54.
- 55.
- 56.
- 57.
- 58.
- 59.
- 60.
- 61.
- 62.
- 63.↵
- 64.
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.
- 75.↵
- 76.↵
- 77.
- 78.
- 79.
- 80.
- 81.
- 82.
- 83.
- 84.
- 85.
- 86.
- 87.
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.
- 96.↵
- 97.↵
- 98.↵
- 99.↵
- 100.↵
- 101.↵
- 102.↵