INTRODUCTION — Annually, approximately 11,150 American women are diagnosed with cervical cancer, and 3670 die from the disease . This represents 0.13 percent of all cancer deaths in women. Squamous cell carcinoma (SCC) accounts for approximately 80 percent of cervical cancers, adenocarcinoma for 15 percent, and adenosquamous carcinoma for 3 to 5 percent (show table 1). In addition, neuroendocrine or small cell carcinomas can infrequently originate in the cervix.
Management of women with nonmetastatic invasive cervical SCC depends upon the FIGO stage of disease at diagnosis (show table 2). This topic will discuss treatment of FIGO stage IB2 and bulky stage IIA cervical SCCs, and the treatment of women with locally advanced (stage IIB-IVA) disease.
Treatment of early stage IA, IB1, nonbulky IIA cervical SCCs, management of disseminated or recurrent disease, and management of cervical adenocarcinomas and small cell cancers are reviewed elsewhere, as is the epidemiology, clinical features, diagnosis, and staging of cervical cancer. (See “Management of invasive cervical cancer: early stage disease (FIGO IA, IB1, nonbulky IIA) and special circumstances” and see “Management of recurrent or disseminated cervical cancer” and see “Management of adenocarcinoma and neuroendocrine carcinoma of the cervix” and see “Epidemiology, clinical features, and diagnosis of invasive cervical cancer” and see “Staging of cervical cancer”).
BULKY STAGE IB AND IIA DISEASE — Women with bulky (ie, tumor >4 cm) stage IB (ie, stage IB2) and IIA cervical cancer (show table 2) have a higher local failure rate and worse survival than those with smaller volume disease. After surgery alone, the rate of relapse is as high as 30 percent [2,3]. In contrast to earlier stage disease, it is not possible to reliably predict which patients are at a low-enough risk of recurrence in order to tailor therapy accordingly. (See “Management of invasive cervical cancer: early stage disease (FIGO IA, IB1, nonbulky IIA) and special circumstances”, section on Indications for adjuvant therapy).
The optimal management of women with primary tumors measuring 4 cm in diameter is controversial. Proposed strategies include: Primary chemoradiotherapy Neoadjuvant (induction) chemotherapy, followed by radical hysterectomy and subsequent chemoradiotherapy, if indicated according to the pathologic findings Primary radical hysterectomy and lymphadenectomy followed by tailored RT with concomitant chemotherapy
Chemoradiotherapy — Primary radiotherapy (RT) has usually been the treatment of choice for women with bulky stage IB and IIA cervical cancer. In a review of 1352 patients with stage IB disease treated with RT alone at the MD Anderson Cancer Center and followed for a median of 12.2 years, rates of central and pelvic tumor control and disease-specific survival were excellent for tumors <5>8 cm, outcomes were less favorable (central and pelvic control and disease-specific survival rates were 69, 57, and 40 percent, respectively). Furthermore, for patients with tumors in the 5 to 7.9 cm category, outcomes were significantly better for those with exophytic as compared to endocervical morphology (central tumor control 91 versus 91 percent, disease-specific survival 76 versus 66 percent, respectively).
If primary RT is utilized, concomitant cisplatin during RT provides additional benefit over RT alone [5,6]. Much of the data supporting chemoradiotherapy over radiotherapy alone come from trials conducted in the setting of locally advanced disease (see “Concomitant chemoradiotherapy versus RT alone” below).
However, benefit for chemoradiotherapy over RT alone has also been demonstrated in women with earlier stage disease. As an example, a trial was sponsored by the Gynecologic Oncology Group (GOG) in which 369 women with bulky stage IB cervical cancer were randomly assigned to RT (both external beam and intracavitary) with or without concomitant cisplatin (40 mg/m2 weekly for up to six doses, maximum 70 mg weekly), and followed by hysterectomy . Compared to RT alone, chemoradiotherapy was associated with a significant 50 percent reduction in the risk of disease progression, and better three year survival (83 versus 74 percent).
Thus, when definitive RT is chosen, cisplatin-based chemoradiotherapy rather than RT alone is indicated.
Optimal timing and schedule of RT — Timely completion of RT is essential for good outcomes, whether chemotherapy is used or not [7-11]. The importance of time to complete RT to overall outcomes was illustrated in a series of 1224 women with cervical cancer treated with definitive RT for stage IB to III disease . Women who required over nine weeks to complete treatment had significantly higher rates of pelvic failure and poorer disease-specific survival at 10 years as compared to those whose treatment was administered over a shorter time period (show table 3).
Similar findings were noted in a patterns of care study involving 837 women undergoing RT for stages I to III cervical cancer . As compared to a total treatment time of 6 weeks or less, those treated over 10 weeks or more had significantly higher rates of four-year in-field recurrence (20 versus 4 percent respectively).
These data are retrospective rather than from prospective trials, and it is possible that longer treatment duration may be a surrogate for the presence of unfavorable tumor or patient characteristics. Nevertheless, in general, RT should be completed within 56 days, if at all possible. From a biologic standpoint, it is likely that similar considerations apply to patients undergoing chemoradiotherapy, as well.
Is there a role for hysterectomy after chemoradiotherapy? — Many of these bulky tumors extend laterally beyond the tumoricidal isodose curve of the brachytherapy application; further, they may contain hypoxic central areas, which do not respond well to RT. These observations provide the rationale for some to recommend an extrafascial hysterectomy following chemoradiotherapy, since approximately one-half of these specimens harbor residual disease, even if concomitant chemotherapy is administered [12,13].
While many studies find that pelvic recurrence rates are lower than expected (2 to 5 versus 15 to 20 percent) in women who have postradiotherapy hysterectomy, its impact on extrapelvic recurrence and survival is less well established [14-18]. The only randomized study comparing RT with and without extrafascial hysterectomy in 256 women with bulky IB disease showed a lower local recurrence rate in the surgery arm (15 versus 27 percent), but the difference was not statistically significant . Unfortunately, the study was hampered by the delivery of suboptimal doses in the RT alone arm (87 percent received 78 to 80 Gy) with 51 percent receiving RT over a protracted treatment period (>60 days). Furthermore, concurrent chemotherapy was not administered. Despite the difference in local disease control, survival was similar in both groups.
It has been suggested that the presence of residual local disease on cervical biopsies performed under anesthesia 8 to 10 weeks after the completion of chemoradiotherapy may serve to identify those women who may benefit from hysterectomy . Of the 21 women found to have positive biopsies, 13 were selected for salvage hysterectomy, of whom 5 achieved long-term complete remission. In contrast, all eight patients who were not selected for salvage surgery because of nonresectable disease or patient refusal died of progressive disease. This approach requires validation in controlled trials before it can be adopted as a method to select patients for hysterectomy.
Summary — When definitive RT is chosen as primary therapy for bulky cervical cancer, cisplatin-based chemoradiotherapy rather than RT alone is indicated. Every effort should be made to complete the RT within 56 days. There is no consensus as to the indications for postchemoradiotherapy hysterectomy. Some clinicians (including one of the authors [CH]) recommend this approach if the tumor siz
e is 8 cm, if there has been a poor response to RT, or if there is concern that the cancer is involving the noncervical portion of the uterus.
Neoadjuvant chemotherapy followed by surgery — Cervical SCC is a chemosensitive malignancy, particularly when cisplatin-containing regimens are used. Numerous nonrandomized studies suggest that neoadjuvant chemotherapy might improve the resectability of bulky lesions, and possibly, long-term survival [20-30].
In the only published prospective trial that directly compared surgery with or without neoadjuvant chemotherapy, 205 women with stage IB disease >2 cm in diameter were randomly assigned to surgery followed by adjuvant RT, or neoadjuvant chemotherapy (three courses of cisplatin 50 mg/m2, vincristine 1 mg/m2, and bleomycin 25 mg/m2 on days one to three, at 10 day intervals) followed by surgery and postoperative RT . Sixty-one patients in the study group and 56 in the control group had bulky stage IB tumors.
Neoadjuvant chemotherapy was associated with a 90 percent objective response rate, a higher likelihood of resectability with negative margins (100 versus 85 percent), and a significant decrease in the rate of pelvic failure, but only a trend towards better survival (82 versus 77 percent). The subgroup with bulky resectable disease (stage IB2) also had a significantly better long-term survival with neoadjuvant chemotherapy (80 versus 61 percent), postoperative adjuvant therapy was given as RT alone, an approach that is inferior to concomitant chemoradiotherapy.
Other trials (including GOG 141, a randomized trial of neoadjuvant cisplatin plus vincristine versus surgery alone which has not been published ) have failed to substantiate a significant benefit for induction chemotherapy compared to surgery alone [32-34], while others that purport to show a benefit have used RT alone as the control arm .
The importance of the timing and dose intensity of cisplatin-based neoadjuvant chemotherapy was addressed In a meta-analysis of individual patient data from 2964 women enrolled on 18 randomized trials . Chemotherapy cycle lengths of 14 days or less and cisplatin dose intensities >25 mg/m2 per week tended to show a survival advantage for NACT. Data from five trials and 872 patients indicated a highly significant 35 percent reduction in the risk of death with neoadjuvant chemotherapy followed by surgery compared with radical radiotherapy alone .
Summary — It cannot be concluded that the use of neoadjuvant chemotherapy followed by surgery gives superior long-term results in the era of modern chemoradiotherapy . Hopefully some answers will be provided by a randomized trial that is underway in the EORTC (EORTC 55994) comparing chemoradiotherapy versus neoadjuvant chemotherapy followed by surgery.
Primary surgery — The third management option for women with bulky stage lB or IIA cervical carcinoma is primary radical hysterectomy with complete lymphadenectomy.
Potential benefits of this approach include: Accurate pathological determination of the extent of the disease and subsequent individualized tailored adjuvant therapy. (See “Staging of cervical cancer”, section on surgical staging). Potential for the resection of bulky metastatic lymph nodes, which may improve prognosis [38,39] (see “Primary surgery versus chemoradiation” below). Radical hysterectomy allows for removal of the primary tumor and accurate pathological evaluation. In a large study comparing patients with stage IB1 and IB2 cervical cancers managed by primary radical hysterectomy, the prognosis of stage IB cervical cancer was best determined by lymphvascular space invasion and depth of invasion, not tumor size as staging criteria would suggest . These factors are best determined pathologically after radical hysterectomy. For the subset of patients for whom radical hysterectomy is the sole treatment required, treatment time is shortened and acute and late radiation sequelae are avoided. Primary radical hysterectomy will also avoid the difficulties of determining if there is viable tumor left after completion of primary chemoradiation. The potential for preservation of ovarian function in young women (although this is frequently not successful) and prevention of radiation-associated vaginal stenosis may be an important advantage of primary surgical management. (See “Ovarian failure due to anticancer drugs and radiation”, section on Ovarian transposition). A primary surgical approach is mandatory in the setting of acute or chronic pelvic inflammatory disease, an undiagnosed coexistent pelvic mass, or anatomic alterations that make optimal RT difficult . Furthermore, if patients are poorly compliant with RT or if expert RT is not available, primary radical hysterectomy should be performed.
Primary surgery versus chemoradiation Potential need for multimodal therapy — One of the main arguments against a primary surgical approach is the high potential for multimodal therapy since the majority of women will be found to have high-risk or intermediate-risk factors for recurrence after surgery; therefore, postoperative adjuvant RT or chemoradiotherapy will be recommended. Using the GOG intermediate and high risk criteria, only about 12 percent of women are able to avoid adjuvant chemoradiotherapy . Treatment-related morbidity is a concern with multimodality therapy, with some series demonstrating a substantial risk of complications, particularly urologic [43-46]. In one RTOG trial of 367 women with stage IB or IIA cervical cancer randomly assigned to pelvic or pelvic plus paraaortic RT, the estimated cumulative incidence of grade 4 and 5 complications was 11 percent in women who underwent RT after abdominal surgery, compared with 2 percent in the pelvic only RT arm .
On the other hand, other trials have found that with modern surgical and RT techniques, the rates of severe post-operative RT complications are acceptably low. As an example, GOG 92 randomly assigned 277 patients with intermediate risk factors to pelvic RT versus no further therapy following radical abdominal hysterectomy . The rates of grade 3 or 4 complications involving the gastrointestinal and urogenital tracts in the RT group were 2.3 and 3.1 percent, respectively, which was similar to the rates of grade 3 or 4 gastrointestinal tract and urological toxicity with radical surgery alone (0 and 1.4 percent, respectively). Similarly, in GOG 109, a randomized trial of adjuvant RT versus chemoradiation following radical hysterectomy in 243 patients with high risk factors, rates of posttreatment small bowel obstruction in the chemoradiation and RT alone groups were 3 and 2 percent, respectively . However, neither of these trials specifically enrolled patients with stage IB2 cancers.
The GOG attempted a randomized trial comparing radical hysterectomy followed by tailored chemoradiotherapy versus primary chemoradiotherapy, however the study closed due to poor accrual, reflecting the strong biases in primary treatment approach to this group of patients across the country. SEER data published in 2001 suggest that about 50 percent of women age 40 and younger with bulky 1B cervical cancer were treated primarily by radical surgery, as were 42 percent of women older than age 40 . Bulky lymphadenopathy — Another point of controversy is whether RT can control bulky lymphadenopathy, and whether there is therapeutic benefit to resecting bulky lymph nodes prior to RT. There are no data from randomized trials that address these issues.
The efficacy of RT in controlling nodal disease in nonsurgically treated patients was evaluated in a retrospective analysis of two French series involving 87 women with FIGO stages IB2 to IVA cervical cancer . The outcomes of 53 patients undergoing pelvic lymphadenectomy followed by RT (44 to 50 Gy to the pelvis with a 6 to 10 Gy boost to enlarged nodes, and LDR brachytherapy 20 to 25 Gy to the tumor volume ) were compared with those of 34 patients whose pelvic nodes were left intact, but who underwent paraaortic lymphadenectomy followed by
RT. Each group underwent completion surgery after RT. Patient characteristics were well balanced between groups and the majority were treated without chemotherapy. The number of patients with positive pelvic nodes was significantly higher among those who had nodal dissections upfront as compared with dissections postradiation (40 versus 18 percent), suggesting that pelvic RT could control some, but not all, pelvic nodal metastases. At least in theory, the addition of chemotherapy to RT might provide higher control rates, but would be less effective at controlling macroscopic disease >2 cm than for smaller tumor volume [52,53].
A survival benefit for primary surgical lymph node debulking was suggested in a study of 266 women with cervical cancer who underwent lymphadenectomy prior to RT . Resection of macroscopically involved nodes was associated with an overall survival of 50 percent, a value comparable to that of women with microscopic nodal involvement only, and significantly better than that of women with unresectable nodes (0 percent). Several other series also suggest that complete resection of enlarged pelvic or paraaortic lymph nodes improves the outcome of subsequent RT and chemoradiation [39,54,55].
Data on clinical outcomes are limited for newer RT techniques that spare normal tissues such as intensity modulated RT (IMRT [56-58]), and newer equipment that allows image-guided treatment delivery and may permit dose escalation to bulky nodes.
Summary — For women with stage IB2 or IIA cervical SCC, initial radical hysterectomy (followed by cisplatin-based chemoradiotherapy for those with intermediate or high-risk features) or definitive chemoradiotherapy are both reasonable treatment options.
Whether either approach is preferable is unknown. A primary surgical approach may be beneficial in premenopausal women where ovaries may be spared. Primary surgery is indicated in the setting of acute or chronic pelvic inflammatory disease, an undiagnosed, coexistent pelvic mass, or anatomic alterations that make optimal RT difficult . Furthermore, if patients are poorly compliant with RT or if expert RT is not available, primary radical hysterectomy should be performed.
Since the majority of women who undergo initial surgery for bulky IB or IIA disease will need postoperative chemoradiotherapy, many clinicians, including one of the authors (JFD), prefer definitive chemoradiotherapy.
The benefit of routinely resecting lymph nodes prior to RT in women who choose definitive chemoradiotherapy is also controversial. If lymphadenectomy is recommended, an extraperitoneal approach should be used.
LOCALLY ADVANCED (STAGE llB, III, IVA) DISEASE — Women with locally advanced cervical squamous cell cancer (greater than stage IIA disease, show table 2) are best treated with primary RT (external beam plus brachytherapy) and concomitant chemotherapy . Nodal involvement, particularly of paraaortic nodes, is the most important adverse prognostic factor, reducing survival by one-half.
Staging lymphadenectomy — The presence of lymph node metastases is the most important prognostic factor for patients with cervical cancer . While surgical evaluation should not change the stage determined by using the FIGO clinical staging system (see “Staging of cervical cancer”), knowledge of the extent of disease may allow for a more individualized therapeutic approach. As with bulky stage IB/IIA disease, the role of staging lymphadenectomy is controversial for women with stage IIB or higher cervical cancer: Arguments in favor of lymphadenectomy include the fact that to date, surgical staging is the most accurate method of determining lymph node involvement. Furthermore, as discussed above, there is a potential therapeutic survival benefit of resecting bulky lymph nodes prior to chemoradiation [38,39,54,55]. At least in theory, chemoradiotherapy is less effective for macroscopic disease >2 cm than it is for smaller volue disease [52,53]. Residual disease in pelvic lymph node after concomitant chemo-radiation for locally advanced cervical cancer has been found in 16 percent of 113 patients with locally advanced cervical cancer who underwent chemoradiotherapy followed by hysterectomy and pelvic lymphadenectomy . Arguments against lymphadenectomy include the delay in the institution of primary chemoradiotherapy, and the increased risk of morbidity (especially late bowel obstructions) with the combined modality approach. The latter risk can be reduced but not eliminated by using a laparoscopic or extraperitoneal approach to lymphadenectomy, and careful RT technique [61-63].
Only one randomized trial has explored this question. A Taiwanese trial randomly assigned 61 women with locally advanced cervical SCC to clinical versus surgical staging, and then, for those assigned to surgical staging, to either a laparoscopic or extraperitoneal approach to lymphadenectomy . Paraaortic nodal metastases were documented in 25 percent of surgically staged women, and study accrual was terminated prematurely when an interim analysis showed that the surgically staged group had a significantly worse progression-free survival and overall survival as compared to those undergoing clinical staging. However, in light of the existing retrospective data, treatment recommendations should not be solely based on this one prospective study, because of design limitations, very small patient numbers, and highly variable treatment regimens.
Many clinicians use the results of the pretreatment staging studies to select women for therapeutic lymphadenectomy. One of us (CH) offers preirradiation staging lymphadenectomy to medically fit patients if they have evidence of bulky nodes on CT scan or if a pretreatment PET scan shows uptake of FDG in the region of the infrarenal paraaortic or pelvic lymph nodes. If both the CT and PET scan are negative, we do not routinely perform staging lymphadenectomy. (See “Staging of cervical cancer”, section on PET scan).
Elective paraaortic RT — Elective paraaortic RT is an alternative to surgical staging for advanced cervical cancer. At least three randomized trials have explored prophylactic extended field RT for women with high-risk cervical cancer and no clinical or radiographic evidence of paraaortic metastases [46,64-66]. The EORTC randomly assigned 441 women at risk for paraaortic involvement (stage I and IIB disease with proximal vaginal and/or parametrial involvement and positive pelvic LNs either on lymphangiogram or at surgery, and stage III regardless of pelvic node status on lymphangiogram) to pelvic with or without paraaortic RT . No significant differences were found between the groups with regard to local control, distant metastases, or survival. However, the majority of patients had bulky stage II or stage III disease, with a high locoregional failure rate, possibly obscuring any potential benefit from paraaortic RT. In contrast, a survival benefit for paraaortic irradiation was suggested in a similarly designed Radiation Therapy Oncology Group (RTOG) study, in which 337 evaluable women with stage IIB disease were randomly assigned to receive or not receive 45 Gy to the paraaortic region in addition to standard pelvic RT . Although locoregional control rates were similar, the 10-year survival rate was significantly better for those undergoing extended field RT (55 versus 44 percent). Treatment-related toxicity was greater, particularly among those who underwent prior surgery for any reason. The relative benefit of extended field RT alone compared to concomitant chemoradiotherapy plus pelvic only RT was tested in RTOG 90-01, which randomly assigned 403 women with advanced cervical cancer confined to the pelvis (stages IIB through IVA or stage IB or IIA with a tumor diameter of at least 5 cm or involvement of pelvic nodes) to RT alone (45 Gy) followed by low dose rate intracavitary irradiation, or RT with concomitant cisplatin (75 mg/m2 on days 1 and 22) plus infusional 5-FU (1000 mg/m2 per day days one through four and 22 through 25 of RT) followed by low-dose-rate (LDR) i
ntracavitary brachytherapy concomitant with the third cycle of chemotherapy . The results were recently updated .
With a median follow-up of 6.6 years for the 228 surviving patients, women with stage IB-IIB disease treated with chemoradiotherapy had significantly superior overall and DFS rates compared to those receiving extended field RT alone . For those with stage III to IVA disease, pelvic RT plus chemotherapy was associated with significantly better DFS, however only a trend was noted toward better overall survival (p = 0.07). It is unclear whether the addition of extended field RT to concomitant chemotherapy may improve upon these results.
Concomitant chemoradiotherapy versus RT alone — The benefit of cisplatin-based concomitant chemoradiotherapy compared to RT alone for locally advanced cervical cancer has been shown in three of four randomized trials and in a meta-analysis [64,65,67-70]:
Randomized trials — The benefit of concomitant chemoradiotherapy over RT alone has now been confirmed in multiple trials, all of which study different populations, and slightly different chemotherapy regimens. The following represents the range of findings: A GOG trial randomly assigned 368 women with locally advanced disease (stage IIB, III, or IVA, show table 4) to RT concurrent with hydroxyurea (HU, 80 mg/kg twice weekly during RT) versus the same RT dose RT plus concomitant cisplatin (50 mg/m2 on day 1 and 29) and 5-FU (1000 mg/m2 by continuous infusion on days 2 to 5, and 31 to 33) . Adverse effects were predominantly hematologic or gastrointestinal in both regimens, and severe or life-threatening leukopenia was more common in the HU group (24 versus 4 percent). Cisplatin-based chemoradiotherapy was associated with a significant improvement in both PFS and overall survival. The relative risk of death in the cisplatin/5-FU group compared to the HU group was 0.74. A second GOG study randomly assigned 526 women with stage IIB, III, or IVA cervical cancer to RT with one of three different chemotherapy regimens: HU alone (3 g/m2 twice weekly for six weeks), cisplatin alone (40 mg/m2 weekly during RT), or cisplatin (50 mg/m2 on days 1 and 29) plus infusional 5-FU (1000 mg/m2 per day on days 1 to 4) and HU (2 g/m2 twice weekly for three weeks) . Two-year survival rates were significantly better with both cisplatin alone and cisplatin/5-FU/HU (66 and 67 percent, respectively) than with HU alone (50 percent). Benefit for chemoradiotherapy was also shown in the previously described RTOG 90-01 trial, where 403 women with bulky (> 5 cm) stage IB/IIA or stage IIB to IVA cervical cancer were randomly assigned to extended field RT versus pelvic RT with concomitant chemotherapy with cisplatin and 5-FU . Chemoradiotherapy was associated with a significantly better five-year overall survival (73 versus 58 percent) and DFS (67 versus 40 percent), and benefits were maintained with longer follow-up .
In contrast to these results, two other randomized trials, one from the National Cancer Institute of Canada (NCIC), and the other from Taiwan  have not shown a significant survival benefit for chemoradiotherapy versus RT alone in women with locoregionally advanced disease: In the larger of the two, from the NCIC, 259 women with stage IB-IVA cervical SCC were randomly assigned to RT alone (external beam plus brachytherapy) or with concomitant cisplatin (40 mg/m2 weekly), failed to show a survival benefit from chemoradiotherapy . The difference in five-year survival between the chemoradiotherapy and RT groups did not reach the level of statistical significance (62 versus 58 percent, respectively). The Taiwanese trial randomly assigned 122 women with bulky IIB or IIIb cervical cancer to RT with or without concurrent multiagent chemotherapy (cisplatin/vinblastine/bleomycin administered on days 1 to 4, and 22 to 25 of the RT course, followed by two additional cycles) . At a median 47 month follow-up, the chemoradiotherapy did not have significantly better three-year disease-free (52 versus 53 percent) or overall survival (62 versus 65 percent, respectively).
Meta-analysis — A Cochrane meta-analysis included trials that randomly assigned patients with FIGO stage IB to IVA disease to chemoradiotherapy versus RT with or without surgery . Compared to RT alone, combined chemoradiotherapy was associated with a significant 31 percent reduction in the risk of death, a 34 percent improvement in PFS, and a significant decrease in both local and distant recurrence rates. Although there was some statistical heterogeneity, the greatest beneficial effect was noted in those trials that included a higher proportion of patients with stage IB and II disease.
Patients with paraaortic metastases — The benefit of adding chemotherapy to radiation therapy (RT) in patients with paraaortic nodal metastases (an indication of distant rather than regional spread) is unclear [12,72-74]. All three of the randomized trials discussed previously specifically excluded such women from eligibility. One series using extended field RT alone reported a 50 percent five-year survival rate among women with clinical stage IB and IIA (show table 4) cervical SCC who were identified at staging laparotomy as having histologically confirmed paraaortic nodal metastases . Despite the use of extended field RT, 14 of 31 patients developed a pelvic recurrence, 12 of which were combined local and distant.
At least two cooperative group studies have evaluated the benefit of chemoradiotherapy in combination with extended field RT in women with positive paraaortic nodes [73,75]: An RTOG trial included 30 patients with clinical stage I to IV disease (17 stage I or II) and positive paraaortic nodes who received twice daily (hyperfractionated) extended field RT plus endocavitary brachytherapy and two to three cycles of concomitant chemotherapy (cisplatin 75 mg/m2 on days 1 and 22, and infusional 5-FU 1000 mg/m2 daily, days 1 to 4, and 22 to 25) . Paraaortic RT was delivered at 1.2 Gy per fraction twice daily in an attempt to safely escalate the dose (median dose 48 Gy, range 7.2-60 Gy). The two and four-year overall survival rates were 46 and 29 percent, and locoregional failure rates at one and two years were 40 and 50 percent, respectively . However, there were unacceptably high rates of acute and late > grade 3 gastrointestinal toxicity (50 and 34 percent, respectively). This trial illustrates the difficulties encountered with delivering RT doses at levels expected to control gross nodal disease in the abdomen. Similar oncologic outcomes were noted in a GOG study investigating treatment outcomes with RT delivered with standard fractionation RT and concurrent cisplatin and 5-FU in 95 women with positive paraaortic nodes, 54 of whom had clinical stage I or II disease . The three-year overall and progression-free survival (PFS) rates for the entire group were 39 and 34 percent, respectively. Survival rates for those with stage I and II disease were 50 and 39 percent, respectively. The dose to the paraaortic nodes was lower than the dose in the RTOG study (45 Gy delivered daily at 1.5 Gy per fraction) and, coupled with single fractionation, resulted in lower rates of gastrointestinal toxicity. Nineteen percent of patients experienced > grade 3 acute GI toxicity with a 14 percent actuarial rate of late morbidity noted at four years. Unfortunately, precise data regarding paraaortic failure was not obtained, and it is unknown whether these lower doses also resulted in persistent disease in the paraaortic nodes.
As noted above, a reduction in toxicity may be achieved using newer RT techniques such as IMRT. At our institutions, we give concomitant chemotherapy with extended field RT as toxicity permits, and many institutions have adopted weekly cisplatin at 40 mg/m2 in this setting due to the more tolerable toxicity profile.
Are there any patients who can be treated with RT alone? — It is unclear whether there are subgroups of women with locally advanced cervical cancer who are adequately treated with RT alone. In
addition to the NCIC randomized trial that failed to show better outcomes with chemoradiotherapy as compared with RT alone, a lack of benefit for chemoradiotherapy was also suggested in a retrospective series of 65 women with node-negative cervical cancer (stage IB2 in 11, IIB in 37, and IIIB in 17) who were staged by FDG-PET . Fifteen received RT alone while 50 underwent concomitant chemoradiotherapy. There appeared to be no overall nor cause-specific survival advantage to the addition of chemotherapy, and pelvic versus distant recurrences were also not significantly different between the groups.
Despite these data, RT alone is not standard of care for any subset of locally advanced cervical cancer at present. Further research is needed to determine which women, if any, are adequately treated with RT alone as compared to chemoradiotherapy.
Summary — Chemoradiotherapy has become the preferred approach to care in the United States and most Western countries for women with locally advanced (greater than stage IIA, show table 2) cervical SCC. The available data support a 30 to 50 percent reduction in the risk of death from cervical cancer for women with locally advanced disease undergoing RT and concomitant cisplatin-based chemotherapy compared to RT alone.
As in patients with earlier stage disease, radiation should be administered within 56 days, if possible. (See “Optimal timing and schedule of RT” above).
We offer preirradiation staging lymphadenectomy to medically fit patients, particularly if they have bulky pelvic or infrarenal lymph nodes on radiographic studies or a PET scan in the absence of more distant metastases. In most cases, a laparoscopic or open retroperitoneal approach to lymphadenectomy is used, unless oophoropexy or harvest of ovarian tissue is needed, in which case, the peritoneum must be entered.
Although not conclusively established, there appears to be comparable benefit from either weekly low dose cisplatin (40 mg/m2 weekly, or a total cumulative dose of 200 mg/m2) or two cycles of cisplatin (50 to 75 mg/m2 on day one) plus 5-FU (1000 mg/m2 per day by continuous infusion daily for four days). We prefer weekly cisplatin because of its more favorable toxicity profile.
IMPACT OF ANEMIA AND ITS CORRECTION DURING TREATMENT — Anemia is common among women with cervical cancer. The etiology is often multifactorial (tumor bleeding, poor nutritional status, anemia of chronic disease, treatment-related myelosuppression). On average, hemoglobin levels generally decrease by approximately 0.4 g/dL per week during concomitant cisplatin-based chemoradiotherapy .
Observational series suggest that low hemoglobin levels during RT for locally advanced cervical cancer are an independent predictor of inferior survival and local control [78-88], although this dogma has more recently been challenged . The impact of pretreatment hemoglobin levels is even less clear, with most (but not all ) reports noting a significant prognostic impact of pretreatment hemoglobin levels in univariate but not multivariate analyses [81,82,85-87,90].
The mechanisms underlying the association between adverse outcomes and anemia are incompletely understood, but may be in part linked to tumor hypoxia [85,91-95]. It is thought that tumor hypoxia induces resistance to RT and some chemotherapy drugs, which are dependent on adequate tissue oxygenation for cytotoxicity . Others suggest that hypoxic tumors behave more aggressively overall, with higher rates of locoregional and systemic failure, regardless of whether they are treated with RT or surgery [97,98].
The optimal hemoglobin level to achieve tumor oxygenation for gynecologic tumors has not been systematically studied, but is thought to be between 12 and 14 g/dL . It is postulated (but not proven) that correction of anemia to these levels may enhance the radiosensitivity of solid tumors, possibly by modifying the hypoxic environment .
The most provocative retrospective series correlating anemia and its correction to outcomes in cervical cancer was a multiinstitutional review of 605 Canadian women who underwent RT alone (89 percent) or concurrent chemoradiotherapy (11 percent) . Five-year survival rates differed significantly according to the average weekly nadir hemoglobin level during treatment: 74, 52, and 45 percent for levels >12, 11-12, and <11 g/dL, respectively. There were also significant differences in local and distant recurrence rates. In multivariate analysis, anemia was the most important negative prognostic factor for outcome after tumor stage, with the magnitude of incremental survival for nonanemic as compared to anemic patients exceeding the gain achieved by concomitant chemoradiation strategies. Correction of anemia by transfusion (to a hemoglobin level 12 g/dL) was associated with treatment outcomes that were similar to those of patients with normal hemoglobin levels.
Although these and other reports suggest value for maintaining hemoglobin levels during RT (via red cell transfusion or the use of recombinant human erythropoietin) [78,79,81,90,101,102], the benefit of this approach has not been prospectively validated in a randomized trial incorporating multivariate statistical analysis.
Only one prospective randomized trial addressing the impact of anemia correction on outcomes was completed in the early 1960s at the Princess Margaret Hospital [79,101]. Women with locally advanced cervical cancer were randomly assigned to transfusion (with the goal of maintaining hemoglobin above 12 g/dL) or no transfusion (unless the hemoglobin level dropped below 10 g/dL) during RT [79,101]. As patients were randomized at diagnosis, only a subset of patients enrolled to the transfusion group (38 of 66) were ever transfused (show table 4), while 25 of the 66 patients in the control group were transfused when their hemoglobin levels dropped below 10 g/dL. There was no survival difference between the two randomly assigned groups. Subgroup analysis of the patients who were transfused during therapy found a significant difference in local relapse rate in favor of the transfused patients in the experimental arm compared with the anemic (and transfused) patients in the control group (show table 4) .
This study was criticized because it was stopped prematurely after recruiting only 132 patients, there were different thresholds for transfusion in both groups, statistical analysis was limited to univariate analysis, and there was no stratification by tumor volume or stage, known important prognostic factors . Thus, these results, although widely quoted, cannot be considered as proof that maintaining hemoglobin levels through red cell transfusion improves outcome in anemic patients.
An alternative approach to correcting anemia in cancer patients is through erythropoiesis stimulating agents (ESAs) such as epoetin and darbepoetin. When given with oral iron, ESAs gradually increase hemoglobin levels by 0.5 g/dL per week. Numerous reports document that administration of ESAs is an effective alternative to red cell transfusion to sustain or raise hemoglobin levels during RT [90,102-104], and some show that outcomes are better in patients who have higher hemoglobin levels at the end of treatment . Based upon these data, the use of ESAs to maintain hemoglobin levels above 12 g/dL has been considered by many to represent standard practice for anemic patients undergoing chemoradiotherapy for locally advanced cervical cancer.
However, whether this improves outcomes is still open to question. There are no randomized trials in women with uterine cervix cancer showing that correction of anemia with ESAs favorably impacts survival and/or local control. Such a trial was initiated by the Gynecologic Oncology Group (GOG 0191), in which women with locally advanced cervical cancer undergoing chemoradiotherapy with weekly cisplatin were randomly assigned to hemoglobin maintenance at 10 g/dL versus aggressive intervention to raise hemoglobin levels to 12 to 13 g/dL (through transfusion and epoetin). However, the st
udy was closed prematurely because of an increase in the number of thromboembolic events in the epoetin arm (although the absolute number was fairly low, 10 of 58 enrolled patients, versus 5 of 55 controls, odds ratio 2.07 ). The study’s primary goal, the impact of ESAs on cancer progression and survival, could not be assessed due to small sample size.
The decision to terminate accrual was made in the context of other reports documenting an increase in thromboembolic events among cancer patients who receive ESAs, including those treated for cervical cancer [90,103,106-108]. Although the data are limited, the risk appears to be highest in patients whose target hemoglobin levels are above 12 g/dL. (See “Role of erythropoiesis-stimulating agents in the treatment of anemia in patients with cancer”, section on Target hemoglobin levels).
There are reasons other than the use of ESAs for these patients to have thromboemboli, including release of procoagulants by tumor cells, comorbid predisposing conditions, and the use of cytotoxic drugs during chemoradiotherapy [109-112].
Nevertheless, these reports, in conjunction with randomized trials suggesting that survival and local control may be worse in patients receiving ESAs with chemotherapy or chemoradiotherapy in the setting of advanced breast or head and neck cancer have raised considerable concern as to this practice, particularly when hemoglobin levels above 12 g/dL are being targeted. The randomized trials that have suggested inferior local control and survival all delivered ESAs to achieve hemoglobin levels above normal levels. (See “Role of erythropoiesis-stimulating agents in the treatment of anemia in patients with cancer”, section on Effect on disease control and survival).
Published guidelines on the use of ESAs from the American Society of Clinical Oncology (ASCO), the American Society of Hematology (ASH), and the National Comprehensive Cancer Network (NCCN) all suggest that the use of ESAs be limited to those patients whose hemoglobin level prior to initiating therapy is 10 g/dL (show table 5) [113,114]. They recommend that the dose be titrated to maintain a hemoglobin level of approximately 12 g/dL.
As noted above, anemia is typically multifactorial in patients with cervical cancer. Treatable causes of anemia (ie, hemolysis, deficiency of B12, folate, iron) should be excluded or treated, if present. At least some data suggest that correction of anemia during chemoradiotherapy might be accomplished successfully through parenteral iron supplementation alone, without the use of ESAs, although this report did not provide information on the iron status of these patients . (See “Approach to the adult patient with anemia”).
Summary — In accordance with published guidelines, we suggest the use of epoetin or darbepoetin during chemoradiotherapy to maintain hemoglobin levels of 11 to 12 g/dL. Use of erythropoiesis-stimulating agents to maintain higher hemoglobin levels may be associated with increased toxicity and is not recommended.
RECURRENT OR DISSEMINATED DISEASE — Management of women with recurrent or disseminated cervical cancer is discussed separately. (See “Management of recurrent or disseminated cervical cancer”).
POSTTREATMENT FOLLOW-UP — Surveillance after primary therapy for cervical cancer is uniformly recommended, although its effectiveness is not well studied [116,117]. The main goal of surveillance is early detection of recurrent disease so that patients can be offered potentially curative salvage therapy. (See “Management of recurrent or disseminated cervical cancer”, section on Surgery and/or radiation therapy).
Recommendations of expert groups — The optimal surveillance strategy has not been established. Guidelines from the National Comprehensive Cancer Network (NCCN) suggest the following : Clinical evaluation every three months for one year, every four months for one year, every six months for three years, and then annually . Annual chest x-ray Other radiographic studies, including CT or PET scan, only as clinically indicated.
Clinical evaluation consists of a review of systems and physical examination with particular attention to the supraclavicular and inguinal lymph nodes, as well as rectovaginal and abdominal examinations. Cervical cytology is obtained at each visit, since many vaginal recurrences are asymptomatic, but have abnormal cytology smears . However, the value of cervical cytology in asymptomatic women has been questioned by others [119,120].
Efficacy — Few studies have specifically addressed the efficacy of routine surveillance for asymptomatic women compared to symptom-based reassessment. Retrospective series suggest that early detection of a localized disease recurrence in asymptomatic women may provide a survival benefit, but the best way to identify these women is unclear [119,121,122].
One of these reports attempted to devise an optimal surveillance program based upon an outcomes analysis of 1096 women with stage 1B cervical cancer . A posttreatment surveillance strategy similar to that suggested by the NCCN was followed. Of the 133 women who recurred, only 19 were asymptomatic when diagnosed. All asymptomatic pelvic recurrences were diagnosed by pelvic examination, and routine cervical cytology smears did not detect a single case.
The median survival from initial diagnosis was significantly longer among women diagnosed while asymptomatic (83 versus 31 months, respectively). The authors suggested that the follow-up strategy be modified to three annual clinical visits yearly for the first three years, twice yearly visits in years four and five, annual Pap tests, and an annual chest x-ray.
Two other posttreatment surveillance tests have been studied: serial assay of the serum tumor marker squamous cell cancer tumor antigen (SCC antigen) and PET scanning.
SCC antigen — The value of serial assays of the serum tumor marker squamous cell carcinoma antigen (SCC antigen) has not been established. Elevated levels can precede clinical detection of recurrence by 4 to 16 months [123-127]. However, in a review of 225 patients, SCC antigen was insensitive for the detection of early recurrence, and survival was no better among women whose disease recurrence was diagnosed by elevations in SCC antigen compared to those diagnosed by other means .
PET scans — Initial staging of cervical cancer is a Medicare-reimbursable indication for PET scanning. FDG-PET is a sensitive and specific method of detecting lymph node metastases in newly diagnosed cervical cancer, and most institutions perform a PET scan during initial staging as an adjunct to conventional imaging. (See “Staging of cervical cancer”, section on PET scans).
The value of PET scanning following treatment for cervical cancer is unsettled. Results using PET or fused PET/CT for diagnosis of recurrent disease are variable. Several studies report sensitivity and specificity rates of 85 percent or higher in this setting [127-130,130,131]. Others suggest limited sensitivity for lesions smaller than 1.5 cm , and lower specificity rates [133,134]. Posttreatment FDG uptake is also reported to be a significant prognostic factor for outcomes after chemoradiotherapy .
The value of PET as a component of routine posttreatment surveillance has not been systematically studied. The limitations of such an approach were noted in a report of 121 women who had undergone whole body PET at least six months after successfully completing treatment for primary cervical cancer; 76 were diagnosed with a recurrence, 20 of whom were asymptomatic at the time of diagnosis . PET scan was performed for a variety of reasons (tumor marker elevation, abnormal physical examination or imaging study, or abnormal histology on smear); 30 women received the test for no reason other than as a component of routine posttreatment surveillance.
Although PET detected disease in 17 of 20 asymptomatic patients with recurrence (85 percent sensitivity), all had undergone the PET for workup of an abnormal
finding. Six of the 30 patients who underwent a PET scan as a component of routine surveillance had a positive PET; all six were determined to be false positive studies.
These data suggest that PET might be useful in the diagnostic workup of abnormal findings on physical examination or Pap smears, but they are insufficient to prove value from including PET in the routine posttreatment surveillance strategy.
Summary — In view of the lack of prospective data to guide posttreatment surveillance, it is not surprising that clinical practice varies. There is consensus on the need to perform clinical evaluation (a review of systems, physical examination with particular attention to the supraclavicular and inguinal lymph nodes as well as rectovaginal and abdominal examinations, cervical cytology) every three months for one year, every four months for one year, every six months for three years, and then annually, as per NCCN guidelines . The majority of clinicians also recommend an annual chest x-ray.
One of the authors (JFD) also routinely performs CT of the abdomen and pelvis at three months after therapy, every six months for the first three years, then annually in year four and five. Others order radiographic studies only as clinically indicated by symptoms or findings on examination.
HORMONE REPLACEMENT THERAPY — As a result of cancer therapy, many women have reduced or absent ovarian hormone production, which causes hot flashes and vaginal dryness . There are scant data regarding use of hormone replacement therapy (HRT) after treatment of cervical cancer [137-139]. Patients who have not undergone hysterectomy should have a discussion with their doctors regarding the risk to benefit ratio of HRT. If HRT is given, estrogen should be used in combination with progesterone, as residual functional endometrium may persist after curative RT for cervical cancer . (See “Postmenopausal hormone therapy: Benefits and risks”).
PROGNOSIS — The major prognostic factors affecting survival are stage, nodal status, tumor volume, depth of cervical stromal invasion, lymphovascular space invasion (LVI), and, to a lesser extent, histologic type and grade. A summary of survival rates by stage of disease according to the latest FIGO data is presented in the table (show table 6) .
Disease stage is the most important prognostic factor, followed by lymph node status. After radical hysterectomy and lymphadenectomy, women with stage IB or IIA disease who have negative pelvic lymph nodes have a five-year survival of 88 to 96 percent, compared to 64 to 74 percent for those with similar stage disease and pelvic nodal metastasis [3,141,142]. Outcomes are worse for women with involved paraaortic nodes [3,12,72-74]. Among patients who have undergone surgical staging or lymphadenectomy, the number of involved lymph nodes also influences prognosis. As an example, in one report, five- year survival rates for patients with one, two, three to four, and five or more positive lymph nodes were 62, 36, 20 and 0 percent, respectively .
The importance of LVI as an independent risk factor is controversial.
HPV subtype — The presence of persistent human papillomavirus (HPV) and specific HPV subtypes may also impact on prognosis after surgery or RT [144-147]. In one population-based series of 399 women diagnosed with stage IB to IV cervical SCC over an 11 year period, 86 were HPV 18-positive while 210 were HPV 16-positive . Compared to HPV 16, HPV 18 positivity was associated with a significantly higher overall and cause-specific mortality. This effect was most pronounced for FIGO stage IB and IIA tumors (hazard ratio [HR] 3.1).
These findings were supported by another study that examined cancer-specific survival in 255 women with cervical cancer and HPV 18/45 . These women were significantly more likely to die of their disease than women with other HPV types or HPV 16/31/52 (HRs 2.1 and 2.4, respectively). The prognosis was even worse in smokers. (See “Cervical intraepithelial neoplasia: Etiology; diagnosis; and natural history”).
Smoking — In contrast to the well-described association between cigarette smoking and cervical cancer incidence, there are conflicting data on whether smoking worsens prognosis. An adverse influence of smoking on survival following treatment of locally advanced cervical cancer has been noted in some studies [149,150], but not others [151,152].(See “Epidemiology, clinical features, and diagnosis of invasive cervical cancer”, section on Epidemiology).
Smoking may also increase the likelihood of late treatment-related complications. In a retrospective review of 3,489 women treated with radiation therapy for cervical cancer at the MD Anderson Cancer Center, heavy smoking was independently correlated with the development of small bowel complications .
Women should be encouraged to quit smoking if at all possible. (See “Patient information: Smoking cessation”).
INFORMATION FOR PATIENTS — Educational materials on this topic are available for patients. (See “Patient information: Treatment of early stage cervical cancer” and see “Patient information: Treatment of early stage cervical cancer for women who desire future pregnancy”). We encourage you to print or e-mail these topics, or to refer patients to our public web site www.patients.uptodate.com, which includes these and other topics.
SUMMARY AND RECOMMENDATIONS
Bulky stage IB and IIA disease — Women with bulky stage IB or IIA disease can be treated by chemoradiotherapy, or initial surgery followed by chemoradiotherapy, if indicated. We suggest a primary surgical approach in the following circumstances (Grade 2C): (see “Primary surgery” above) Acute or chronic pelvic inflammatory disease An undiagnosed coexistent pelvic mass Anatomic alterations that make optimal RT difficult, or if expert RT is unavailable
For women who do not fall into one of these categories, there is a lack of consensus as to the best approach. One of us (CH) prefers initial surgery for women with potentially resectable tumors (particularly in premenopausal women where ovaries can be spared), while the other (JFD) prefers chemoradiotherapy, largely due to the greater toxicity that accompanies surgery followed by chemoradiotherapy in this setting. If radiotherapy is chosen, it should be administered within 56 days, if possible. (See “Chemoradiotherapy” above).
For women who present with bulky lymphadenopathy, there is also controversy as to the best approach. We suggest lymphadenectomy, as long as an extraperitoneal approach is used (Grade 2C). (See “Staging lymphadenectomy” above).
For women who undergo surgery, if postoperative therapy is indicated because of high or intermediate risk factors, we recommend chemoradiotherapy rather than RT or chemotherapy alone (Grade 1A). We also recommend concomitant cisplatin-based chemotherapy rather than RT alone for women who undergo definitive RT instead of surgery (Grade 1A). (See “Chemoradiotherapy” above).
Until additional data become available, we suggest the use of recombinant epoetin or darbepoetin during chemoradiotherapy for anemic patients (hemoglobin 10 g/dL), aiming to maintain the hemoglobin level 11 to 12 gm/dL (Grade 2C). The use of erythropoiesis-stimulating agents to maintain higher hemoglobin levels may be associated with increased toxicity and is not recommended (Grade 1B). (See “Impact of anemia and its correction during treatment” above).
Locally advanced disease — For women with stages IIB, III, and IVA cervical SCC, primary chemoradiotherapy is recommended over RT alone (Grade 1A). The optimal management of patients with bulky lymphadenopathy on preirradiation staging CT or PET scan is controversial. One of us (CH) prefers lymphadenectomy (using an extraperitoneal approach) for medically fit patients who have lymphadenopathy or radiotracer uptake in the region of the pelvic or infrarenal paraaortic lymph nodes in the absence of more distant metastases; the other (JFD) prefers chemora
diotherapy with extended field RT and newer RT techniques such as intensity modulated radiation therapy (IMRT), if available. (See “Staging lymphadenectomy” above).
Although the optimal regimen has not been established, we suggest weekly cisplatin (40 mg/m2 to a cumulative dose of at least 200 mg/m2) during RT (Grade 2C). (See “Concomitant chemoradiotherapy versus RT alone” above).
Until additional data become available, we suggest the use of epoetin or darbepoetin during chemoradiotherapy to maintain hemoglobin levels of 11 to 12 g/dL (Grade 2C). The use of erythropoiesis-stimulating agents to maintain higher hemoglobin levels may be associated with increased toxicity and is not recommended (Grade 1B). (See “Impact of anemia and its correction during treatment” above).
INTRODUCTION — Annually, approximately 11,150 American women are diagnosed with cervical cancer, and 3670 die from the disease . This represents 0.13 percent of all cancer deaths in women. Squamous cell carcinoma (SCC) accounts for approximately 80 percent of cervical cancers, adenocarcinoma for 15 percent, and adenosquamous carcinoma for 3 to 5 percent (show table 1). In addition, neuroendocrine or small cell carcinomas can infrequently originate in the cervix.