Cancers of the female genital tract are a major public health problem worldwide. Cancer data from the National Cancer Institute, Centers for Disease Control and Prevention, North American Association of Central Cancer Registries, and National Center for Health Statistics are compiled by the American Cancer Society (ACS), which estimates that there will be approximately 11,150 new cases of invasive cervical cancer resulting in approximately 3670 deaths in U.S. for the year 2007. The probability of developing invasive cervical cancer in women from birth to death has been reported as 1 in 138 in the U.S.1 Based on these figures, cervical cancer currently ranks as the third most common cancer of the female genital tract after endometrial and ovarian cancer and is the third most common cause of mortality from gynecologic cancer. Worldwide, cervical cancer is not only the most common cause of female genital cancer, with an incidence of almost 493,100 and mortality of approximately 273,000; it also is the third most common cancer diagnosed after breast and colorectal cancer in the female population. It primarily affects the women of developing nations, who bear up to 85% of the brunt of the disease.2 Over the past few decades, efficient screening with the Papanicolaou (Pap) test, which detects cervical cytopathology and premalignant precursors, has contributed remarkably to the reduction in cervical cancer burden in developed countries. Physical examination, along with a few procedures, and radiologic studies, such as conization, cystoscopy, intravenous pyelography, barium enema, chest X-ray, and skeletal X-ray, are used by the International Federation of Gynecology and Obstetrics (FIGO) 1995 classification system to stage cervical cancer. Diagnostic studies, such as lymphangiography, ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI), may be used in an attempt to plan therapy. It should be noted that the FIGO staging does not incorporate lymph node involvement, in contrast to the tumor, lymph node, metastases staging for most other cancers.

The therapeutic modalities used to cure cervical cancer are dependant on the tumor stage at the time of presentation and include primary treatment with surgery, radiation therapy, or chemoradiation. The most important and widely studied prognostic variables for cervical cancer are lymph node spread, tumor size, involvement of parametrial tissue, depth of invasion, and lymphovascular invasion.3–6 The incidence of patients with lymph node involvement increased with parametrial extension (P<.01). The corrected 5-year survival rates were 94% for patients with stage IB disease, 95% for patients with stage IIA disease, and 78% for patients with stage IIB disease.3 In 1 study that was conducted on 628 patients who had cervical cancer between stage IA1 and stage IIB, the investigators reported that, with parametrial extension, lymph node metastasis rates increased significantly from 12% to 32% for patients with stage IB disease (P<.005), from 9% to 70% for patients with stage IIA disease (P<.001), and from 29% to 58% for patients with stage IIB disease (P<.001).4 Similarly, it was reported that lymph node status, parametrial invasion, lymphovascular space invasion, and a histology of pure adenocarcinoma were important histopathologic prognostic factors for patients with early-stage cervical carcinoma who underwent radical surgery.5 Although lymphovascular invasion does not alter FIGO staging, if the lymphovascular invasion is >3 mm and < 5 mm, then the risk of recurrence is 15.7% compared with a risk of only 3.1% if the invasion is < 3 mm.6

Positron emission tomography (PET) is an in vivo, functional imaging modality that depends on the metabolic behavior of the viable cells. Malignant cells have a unique biochemical feature of increased glucose metabolism, and this factor is the basis of using PET in cancer detection and prognostication. The radiolabeled glucose analogue, 18F-fluoro-2-deoxy-D-glucose (FDG), is the most commonly used radiotracer for PET. The role of FDG-PET in oncology is mainly in initial staging, detecting distant metastasis, and stratifying patients for treatment.7 However, certain physiologic processes can interfere with the optimal evaluation of these lesions. Particularly, in women with cervical cancer, physiologic excretion of FDG through the urinary tract can cause intense activity in the urinary bladder, which may result in falsely upstaging the disease. Therefore, several experimental techniques have been used to avoid interference by urinary bladder radiotracer activity. The easiest technique is to ask the patient to empty the urinary bladder just before imaging and to start scanning from the pelvis. The other reported techniques are the use of hydration (1 L normal saline intravenously) with diuretics (furosemide) and continuous bladder irrigation to dilute and remove the radioactive urine. FDG-PET interpretation also is challenging because of physiologic hormone-dependent changes in the ovaries and endometrium and a variety of benign pathologies, such as serous and mucinous cystadenomas, corpus luteal cysts, dermoid cysts, endometriosis, inflammation, and urinary diversions.8, 9 It also has been demonstrated that FDG uptake varies during various phases of the menstrual cycle.9, 10 It was observed that normal endometrial uptake was higher during the ovulatory and menstrual phases than during the proliferative and secretary phases.10 Clinicians should be aware of these normal variations in FDG uptake in the pelvis prior to interpretation. A combined PET-CT scan provides the added advantage of anatomic and structural characterization of the lesion over the functional image derived from a PET scan in the same setting and provides greater specificity than PET and CT alone.

Many studies have evaluated the role of FDG-PET in the primary staging of cervical cancer. A recent meta-analysis by Havrilesky et al. demonstrated a combined, pooled sensitivity and specificity of 84% and 95%, respectively, for aortic lymph node metastases and 79% and 99%, respectively, for pelvic lymph node metastases.11 It has been established that there is an increase in the accumulation of FDG over time; delayed imaging with FDG-PET can be helpful in extracting more information. The advantages of dual time-point imaging have been demonstrated in patients with cancers of lung, breast, and head and neck. Likewise, the superiority of dual time-point FDG-PET imaging over conventional imaging has been demonstrated in patients with cervical cancer, who had an increase in accuracy from 70% to 90%.12–14 It also has been established that FDG-PET plays an important role in detecting recurrence in many tumors, including cervical cancer.7, 15 The meta-analysis by Havrilesky et al. demonstrated a combined, pooled sensitivity of 96% and specificity of 81% with FDG-PET.11 There is ample evidence in the literature of the superiority of PET over conventional imaging modalities in detecting recurrent cervical cancer.

FDG-PET imaging has emerged as a useful tool for predicting survival and monitoring therapy in patients with cervical cancer.7, 15 Singh et al. investigated patients with FIGO clinical stage IIIB cervical cancer and observed that patient outcome in terms of cause-specific survival was highly dependent on the degree of lymph node involvement by tumor cells as demonstrated on FDG-PET studies.16 The reported 3-year cause-specific survival rate was 73% for patients without lymph node metastasis, 58% for patients with pelvic lymph node metastasis only, 29% for patients with pelvic and para-aortic lymph node metastases, and 0% for patients with pelvic, para-aortic, and supraclavicular lymph node metastasis. In consonance, a retrospective study of 152 patients with cervical cancer established similar results.17 The 5-year survival rate in that study reportedly declined with FDG uptake from 80% in patients with no abnormal FDG uptake, to 32% in patients with persistent abnormal uptake, and to 0% for patients with new sites of abnormal FDG uptake. Similarly, in another retrospective study that involved 76 patients with newly diagnosed cervical cancer, both pretreatment and posttreatment FDG-PET studies were obtained.18 In that study, the reported 2-year progression-free survival rate was 86% for patients with no abnormal FDG uptake, 40% for patients with persistent abnormal uptake, and 0% for patients with new sites of abnormal FDG uptake. Grigsby et al. evaluated pretreatment lymph node size, irradiation dose, and failure patterns in patients with cervical cancer using both PET scans and CT scans, using PET to score lymph nodes positive or negative and using CT to determine lymph node size.19 The results indicated that positive lymph nodes of any size at diagnosis were the most significant predictor for developing distant metastases. The study identified distant metastases as the most common site of treatment failure. Yoshida et al. considered patients with advanced uterine cervical cancer to evaluate the role of FDG-PET in monitoring treatment with neoadjuvant chemotherapy.20 Tumor volume and standardized uptake values (SUVs) were decreased significantly by chemotherapy, and FDG-PET was superior to MRI for these measurements.

FDG uptake by the primary tumor in patients with lymphoma, head and neck cancer, lung cancer, and esophageal cancer has been correlated with patient outcomes.21, 22 Xue et al. evaluated the prognostic significance of FDG uptake in primary cervical cancer and reported that FDG uptake was predictive of disease-free survival in patients who received radiation therapy. In that study, the 5-year disease-free survival rate was 71% for patients with an SUV < 10.2 and 52% for patients with an SUV ≥10.2. Recently, Kidd et al. investigated the associations between the maximal SUV (SUVmax) of the primary cervical tumor at initial diagnosis and local tumor control, recurrence rates, risk of distant metastasis, and overall survival in patients with cervical cancer using FDG-PET studies. Their report is published in this issue of Cancer.24 The authors demonstrated that the SUVmax at the primary tumor site was an independent predictive biomarker of lymph node status, persistent disease after treatment, pelvic recurrence, and overall survival outcome in patients with cervical cancer. However, the primary tumor SUVmax was not related to patient-specific factors, such as histology, tumor stage, or tumor volume. This finding is contrary to many other studies, which demonstrated a good correlation between tumor volume and SUVmax. We also observed that the SUVmax was related directly to primary tumor size in patients with breast cancer.25 We have demonstrated that the SUVmax increases as tumor size increases, and vice versa. Kidd et al. demonstrated that the SUVmax is a stronger predictor of overall survival than tumor volume or lymph node status. In addition, those authors also demonstrated that the SUVmax of the primary tumor at initial diagnosis is predictive of lymph node involvement. For breast cancer, we also observed an association between primary tumor size, SUV, and the presence of lymph node involvement.26 These data are very important from the standpoint of managing cervical cancer, because, until recently, initial tumor staging and lymph node involvement were considered predictors of outcome.

Because of the high sensitivity of PET-CT studies in performing noninvasive studies, there always has been a search for the development of other PET tracers. To overcome the limitations of FDG, tracers other than FDG also have been evaluated in patients with cervical cancer. Hockel et al. reported that patients with hypoxic, low-apoptotic cervical cancers had a very high probability of lymphatic spread and an extremely poor outcome.27 Dehdashti et al. evaluated pretreatment Cu-60 diacetyl-bis-N(4)-methylthiosemicarbazone (Cu-ATSM) PET in predicting responsiveness to subsequent therapy in patients with cervical cancer.28 Tumor uptake of Cu-ATSM was related inversely to progression-free survival and overall survival. In addition, the frequency of locoregional lymph node metastasis was greater in patients who had hypoxic tumors. Lapela et al. evaluated 11C-methionine PET in 14 patients (8 patients with endometrial cancer and 6 patients with cervical cancer).29 Those authors reported that the mean SUV of the carcinomas was 8.4, whereas the mean SUV for normal endometrium was only 4.6. We believe that these radiotracers may be helpful in staging and detecting recurrent cervical cancer but are less likely to play any role in prognostication for these patients.

In conclusion, FDG-PET/PET-CT studies may play a potential role in the staging and restaging of patients with cervical cancer. Recent research has demonstrated another important aspect of FDG-PET studies in predicting the prognosis for patients according to their baseline SUVmax. Thus, more aggressive treatment strategies can be developed for these high-risk patients, who may be identified by pretreatment PET studies. There is a very strong argument for entering these patients into randomized trials and comparing, for instance, more advanced therapeutic modalities in patients with cervical cancer who have high baseline SUVmax with other conventional treatments. Currently, further data in this regard from other institutes with larger numbers of patients will help us incorporate PET/PET-CT as a primary imaging modality in the routine work-up and management of patients with cervical cancer.


  1. Top of page