Staging and Prognosis
Suspected early disease
When a diagnosis of invasive cervical cancer has been established histologically, an evaluation of all pelvic organs should be performed to determine whether the tumor is confined to the cervix or has extended to the adjacent vagina, parametrium, endometrial cavity, bladder, ureters, or rectum. According to the International Federation of Gynecology and Obstetrics (FIGO) guidelines for clinical staging (Table 1), diagnostic studies may include intravenous urography, cystoscopic examination of the bladder and urethra, a proctosigmoidoscopic study, a barium enema, and in the case of early-stage disease, a colposcopic study of the vagina and the vaginal fornices. Colposcopic findings may be used for assigning a stage to the tumor (eg, FIGO stage IIA), but the results must be confirmed by biopsy.
A pelvic examination must be performed as part of the staging process, and the procedure is best done with the patient completely relaxed by general anesthesia. In up to 20% of patients, the initial clinical classification of the disease has proved to be incorrect at the time of pelvic examination. Such an examination can reveal a more advanced stage of the disease than was originally found; additional biopsies (if indicated) or fractional curettage can be performed as well as colposcopy, cystoscopy, and proctosigmoidoscopy.
Suspected advanced disease
When studies detect ureteral obstruction, a tumor is classified as a stage IIIB lesion, regardless of the size of the primary lesion. Ureteral obstruction, either hydronephrosis or nonfunction of the kidneys, is well established as an indicator of poor prognosis, as recognized in the FIGO classification.
In women with bulky or advanced-stage tumors, the bladder mucosa also should be inspected cystoscopically for possible bullous edema, which indicates lymphatic obstruction within the bladder wall. Evidence of tumor in the bladder must be confirmed by biopsy before the lesion can be classified as stage IVA. Rectal mucosal lesions also require a biopsy via proctosigmoidoscopy, because they can be related to an inflammatory process rather than to the cervical tumor.
Surgical experience from pelvic lymphadenectomy has confirmed an error rate of 15% to 25% in the clinical staging of patients with stage IB or II lesions. In 10% to 30% of cases involving stage II/III tumors, in addition to positive findings of occult pelvic lymph nodes, other metastases may be found in the para-aortic nodes. Unfortunately, pelvic examinations and clinical staging as defined by FIGO cannot detect such metastases.
Consequently, there is a growing body of literature that shows the superiority of cross-sectional imaging (computed tomography [CT] and magnetic resonance imaging [MRI]) over clinical staging in delineating the extent of disease in patients with cervical cancer. As stated previously, official FIGO guidelines do not incorporate the use of either CT or MRI findings into the staging of cervical cancer. However, as knowledge of prognostic factors and the value of cross-sectional imaging has accumulated, its use in treatment planning has increased without changing the official FIGO clinical staging guidelines. Similarly, although the benefits of laparoscopic extraperitoneal surgical staging have also been reported in this setting, this approach has not been incorporated into the FIGO staging system.
In a study (Tsai CS et al: Int J Radiat Oncol Biol Phys 2010), 129 patients with cervical cancer at stages IB to IV who had pretreatment MRI were randomized to an FDG-PET scan (n = 66) or to no additional study (n = 63). Seven patients (11%) were found to have extrapelvic metastasis on PET scans (six para-aortic nodes and one omental). The 4-year overall survival was 79% and 85%, respectively (P = .65), and the disease-free survival was 75% and 77%, respectively (P = .64).
The value of CT scanning in the pretreatment evaluation of patients with cervical cancer is in the assessment of advanced disease (stage IIB and greater) and in the detection and biopsy of suspected lymph node metastasis. The treatment plan for patients with locally advanced disease must be modified if upper abdominal tumor masses and/or distant metastasis is discovered. The soft tissue contrast resolution of CT scanning does not allow for consistent tumor visualization at the primary cervical site and, therefore, neither tumor size nor early parametrial invasion can be evaluated reliably. However, T2-weighted MRI allows consistent tumor visualization and has been reported to be more than 90% accurate in determining tumor size to within 5 mm of measurements of surgical specimens. Nevertheless, a study by the American College of Radiology Imaging Network in 208 patients with invasive cervical cancer evaluated with CT or MRI before radical hysterectomy showed that MRI was superior to CT and clinical examination in evaluating uterine body involvement and in measuring tumor size, but neither method was accurate in evaluating the cervical stromal depth of tumor invasion.
Recent reports show the value of positron emission tomography (PET) in the pretreatment evaluation, treatment planning, and post-therapy assessment of response in patients with higher-risk invasive carcinoma of the cervix. In a series of 120 patients with newly diagnosed cervical cancer, FIGO stage higher than IB, PET/CT scanning demonstrated a 94% positive predictive value (PPV) and 100% negative predictive value (NPV) in detecting positive para-aortic lymph nodes. There was 100% sensitivity and 99% specificity. When scanning for the presence of distant metastasis, PET/CT scanning had a 63% PPV and 100% NPV, with 100% sensitivity and 94% specificity. In a prospective study of 103 patients who were successfully treated initially with concurrent chemotherapy and radiation therapy, surveillance FDG-PET (fluorodeoxyglucose-PET) detected asymptomatic recurrent disease earlier, which may be potentially amenable to salvage therapy.
Noteworthy, in 60 patients with stages IA2–IIA cervical cancer up to 4 cm with MRI-negative nodes, preoperative FDG-PET scanning detected 1 para-aortic node metastasis, but only 1 of 10 pelvic node metastases, which led to the conclusion that PET scanning is of little value in evaluating patients with stages IA2–IIA cervical cancer up to 4 cm. A second study showed that 3 of 38 patients with no para-aortic uptake on FDG-PET/CT imaging had histologically proven para-aortic node involvement.
Clinical staging of cervical carcinoma, although widely used, is not without controversy. When compared with surgical staging performed by large cooperative groups, clinical staging is frequently inaccurate in predicting locoregional tumor spread. For many cooperative groups, including the Gynecologic Oncology Group (GOG), surgical staging may be required for patients who are entering prospective randomized clinical protocols.
The most common method used to stage patients with advanced disease is extraperitoneal sampling of the pelvic and para-aortic lymph nodes. This approach minimizes the risk of subsequent radiation injury to the small bowel due to surgical adhesions and, in patients with advanced disease, allows for individualized treatment planning. Another approach is sentinel lymph node detection in the pelvis, which is still an active area of research.
Pros and cons of surgical staging
The advantage of surgical staging is that patients with microscopic disease in the para-aortic lymph nodes can be treated with extended-field radiation therapy (EFRT) and, possibly, chemotherapy, and potentially benefit in terms of long-term survival. The controversy regarding surgical staging stems from the fact that a small number of patients will actually benefit from the procedure; the majority of patients who undergo it will be found not to have metastatic disease and will receive the same treatment as planned prior to surgical staging; if they are found to have metastatic disease, they will be unlikely to benefit from EFRT. Because of this controversy, the GOG considers surgical staging to be optional for patients with advanced-stage cervical cancer.
Workup for advanced disease
The standard workup of a patient with advanced cervical carcinoma who is not considered a candidate for radical surgery includes an abdominopelvic CT scan with both intravenous and gastrointestinal oral contrast. If there is evidence of para-aortic lymph node metastases, the patient should undergo fine-needle aspiration (FNA) of these enlarged lymph nodes. If FNA confirms that there is para-aortic lymph node metastasis, treatment should be individualized, and EFRT should be considered part of the primary treatment regimen.
If the scalene lymph nodes are negative on clinical examination and the patient is known to have positive metastatic disease to the para-aortic lymph nodes, consideration can be given to performing a scalene lymph node biopsy; the incidence of positive scalene nodes when para-aortic lymph nodes are known to be positive ranges from 0% to 17%. The rationale for doing a biopsy of the scalene nodes is that if there is disease outside the radiation therapy field, chemotherapy may be appropriate.
If the result of FNA is negative, or if the abdominopelvic CT scan does not demonstrate enlarged para-aortic lymph nodes, the patient can be considered for surgical staging.
An American Society of Clinical Oncology (ASCO) 2007 report compared the accuracy of PET vs PET/CT for detecting lymph node metastasis in cervical cancer. In a series of 86 patients with stages IB–IVA cervical cancer, a total of 688 lymph node regions were evaluated. PET/CT was more sensitive than PET for detecting small (< 5 mm) lymph node metastases. Recent meta-analyses performed by Choi et al showed that PET or PET/CT had an overall higher diagnostic performance (82% sensitivity and 95% specificity) in detecting metastatic lymph nodes in patients with cervical cancer than CT (50% sensitivity and 92% specificity) and MRI (56% sensitivity and 91% specificity).
The introduction of minimal-access surgery has allowed surgeons to accurately stage patients via the laparoscope before initiation of radiation therapy. Laparoscopic para-aortic lymph node dissection has been shown to be feasible in gynecologic malignancies and has been proposed as a diagnostic tool for determining the presence of lymph node metastasis. More recently, robotic surgery has also been used in surgical staging before radiation treatment. However, the safety and efficacy of both laparoscopic and robotic surgical staging are areas of ongoing investigation.
Workup for early-stage disease
For patients who have early-stage disease for which surgery is contemplated, only a minimal diagnostic workup is indicated before surgery. At most institutions, this would include two-view chest radiographs. Patients who have stage IA cervical carcinoma (microinvasive carcinoma) do not require preoperative CT scanning before hysterectomy. For patients with a small stage IB carcinoma of the cervix, a CT scan of the abdomen and pelvis has a low yield and is unlikely to change the treatment plan.
The most important determinant of prognosis remains clinical stage, which is defined by tumor volume and extent of disease spread. The overall 5-year survival rate ranges from 95% to 100% for patients with stage IA cancer and from 75% to 90% for those with stage IB disease. Patients with stage IV disease have a 5% or less chance of surviving 5 years after diagnosis.
Patients with early disease
For patients with early invasive carcinoma (stage IB), the size of the lesion, percentage of cervical stromal invasion, histology, tumor grade, and lymphovascular space involvement are important local factors that predict prognosis. In general, good prognostic signs are lesions that are 2 cm or less in diameter, superficially invasive, and well-differentiated with no lymphovascular space involvement. In a study of 1,067 patients treated with surgery, HPV-16 was detected in 63.8% and HPV-18, in 16.5% of samples. With a median follow-up of 77 months, HPV was not found to be a significant prognostic factor.
For patients who have undergone radical hysterectomy for early cervical carcinoma, poor prognostic factors, in addition to the local factors previously mentioned, include positive vaginal or parametrial margins and metastasis to the pelvic lymph nodes. For patients with stage IB disease and positive pelvic nodes, the 5-year survival rate drops from approximately 75% to 85%, to 50%.
Patients with advanced disease
For patients with advanced-stage disease (stages IIB–IV), the primary determinants of prognosis are histology and size of the primary lesion. Survival is significantly longer for patients with small stage IIB cervical carcinomas and minimal parametrial involvement than for patients with large bulky tumors and bilateral parametrial involvement. Disease extension beyond the pelvis to the para-aortic nodes is associated with a significant decrease in overall survival rate. With regard to histology, a better prognosis is associated with a large-cell nonkeratinizing squamous cell cancer of the cervix, as opposed to a poorly differentiated adenocarcinoma. In patients with stage IIIB cervical cancer restricted to the pelvis, it was shown that the presence of hydronephrosis at presentation is a significant prognostic factor associated with a poor performance status and a worse survival.
Other prognostic factors
Other factors that may predict outcome include the patient's general medical and nutritional status. Patients who are anemic may respond poorly to radiation therapy, as compared with those with normal hemoglobin levels. Patients with significant alterations in their immune system may not respond as well; this result is becoming increasingly apparent with regard to patients who are HIV-seropositive.
A retrospective review (Grogan et al: Cancer 1999) of 605 patients from seven institutions in Canada treated with irradiation for cervical cancer described average weekly nadir hemoglobin levels as significant prognostic factors for survival, second only in importance to tumor stage. Interestingly, Winter et al reported that hemoglobin levels during treatment were independent predictors of treatment outcome through a retrospective study of 494 patients from two consecutive prospective GOG trials. The pretreatment level was not a significant predictor of outcome in the multivariate regression model. Hemoglobin levels in the last part of treatment were the most predictive of disease recurrence and survival. However, erythropoietin(Drug information on erythropoietin) should not be given outside a clinical trial, because thrombosis is a significant complication and cause/effect has not been proved.
PET scanning is being used to determine response and outcome after therapy. In 152 patients with cervical cancer treated with irradiation alone or in combination with chemotherapy, Grigsby et al reported cause-specific survival of 80% in 114 patients without post-therapy PET abnormalities, 32% in 20 patients with persistent abnormal scans, and no survivors in 18 patients who developed new sites of abnormal uptake.