Use of Positron Emission Tomography in Cervical Cancer
By David M. Gershenson, MD
Because of progress related to pap smear screening programs over the past half-century, invasive cervical cancer has become a rather uncommon malignancy in the United States, with fewer than 13,000 new cases annually. However, cervical cancer remains one of the most common malignancies among women worldwide. Unlike the surgical staging systems of endometrial, ovarian, and vulvar cancers, the staging system for cervical cancer is clinical. Therefore, staging for cervical cancer uses only physical examination and commonly available diagnostic studies. In fact, information on the lymph node status—pelvic, para-aortic, and supraclavicular—assessed by surgery or imaging studies is not considered in this staging system. Nevertheless, we have known for quite some time that lymph node status is 1 of the most important prognostic factors in cervical cancer. Lymph node spread is a predominant mode of spread in this disease and generally follows an orderly pattern, with involvement of pelvic nodes initially and subsequent dissemination to para-aortic and then supraclavicular nodes. Importantly, failure to detect lymph node spread may result in failure to treat this site altogether or failure to completely eradicate tumor because of inadequate doses of irradiation.
Over the past 2 decades or so, several different imaging techniques have been used to assess lymph node involvement in cervical cancer patients. These have included lymphangiography, computed tomography (CT), and magnetic resonance imaging (MRI). Lymphangiography was used extensively at M.D. Anderson Cancer Center for several decades with excellent results. Some studies indicated that this technique was superior to CT in detecting lymph node spread from cervical cancer. However, this technique requires greater expertise in performing and interpreting; thus, it has fallen out of favor even at our institution. More recent studies have indicated that lymphangiography, CT, and MRI are equivalent in terms of detection of lymph node involvement.1 However, the sensitivity of both techniques is unacceptably low. Hence, the emergence of pretreatment surgical staging for cervical cancer. At several centers, patients with newly diagnosed cervical cancer undergo extraperitoneal lymphadenectomy prior to definitive chemoradiation for advanced cervical cancer. In my view, the precise role of surgical staging for cervical cancer remains unclear. We currently do not perform this procedure on every new cervical patient but rather reserve it for patients with enlarged metastatic lymph nodes detected by imaging studies. However, the benefits of debulking grossly positive retroperitoneal lymph nodes in cervical cancer also remains controversial and unproven.
Over the past few years, the use of positron emission tomography (PET) scanning has been studied in patients with invasive cervical cancer with promising findings. This technique uses a derivative of glucose—fluorodeoxyglucose (FDG)—that is useful in imaging of solid tumors because of the high glycolytic rate of many malignancies. In 1999, Rose and colleagues reported on the use of PET scanning prior to surgical staging in 32 patients with stages IIB, IIIB, and IVA cervical cancer.2 FDG was taken up by 91% of the cervical cancers. Six of 8 patients with positive para-aortic lymph nodes had PET scan evidence of para-aortic nodal metastasis. One of the 2 false-negatives had only 1 microscopic focus of metastasis. In the para-aortic nodes, PET scanning had a sensitivity of 75%, a specificity of 92%, a positive predictive value (PPV) of 75%, and a negative predictive value of 92%. FDG para-aortic nodal uptake conferred a relative risk of 9.0 for para-aortic nodal metastasis. All 11 of 17 patients with metastasis to pelvic lymph nodes were predicted by PET scanning (P < .0001); 5 of these patients had abnormalities on CT scans.
In a study from Germany, 35 patients with stages IB or II cervical cancer underwent FDG-PET scanning and MRI prior to radical hysterectomy and pelvic lymphadenectomy.3 Histologic examination revealed lymph node metastasis in 11 of the 35 patients. This included 3 of 21 patients (14%) with stage IB disease and 8 of 14 (57%) with stage II disease. Nodal staging resulted in sensitivities of 0.91 with PET and 0.73 with MRI and specificities of 1.00 with PET and 0.83 with MRI. The PPV of PET was 1.00 and that for MRI was 0.67. The metastatic involvement of lymph node sites was identified at PET with a PPV of 0.90; at MRI, 0.64 (P < .05).
In a recently reported study from the Mallinckrodt Institute of Radiology in St. Louis, Grigsby and associates retrospectively compared the results of CT lymph node staging and whole-body FDG-PET in 101 consecutive patients with cervical cancer.4 Patients were treated with standard irradiation and chemotherapy, as clinically indicated, and subsequently followed at 3-month intervals. No pretreatment surgical staging was performed in this group of patients. CT demonstrated abnormally enlarged pelvic lymph nodes in 20 (20%) and para-aortic lymph nodes in 7 (7%) of the 101 patients. PET demonstrated abnormal FDG uptake in pelvic lymph nodes in 67 (67%), in para-aortic lymph nodes in 21 (21%), and in supraclavicular lymph nodes in 8 (8%). The 2-year progression-free survival, based solely on para-aortic lymph node status, was 64% in CT-negative and PET-negative patients, 18% in CT-negative and PET-positive patients, and 14% in CT-positive and PET-positive patients (P < .0001). A multivariate analysis demonstrated that the most significant prognostic factor for progression-free survival was the presence of positive para-aortic lymph nodes as detected by PET imaging (P = .025).
In summary, there is mounting evidence that PET scanning is currently the best technique for detection of lymph node spread from cervical cancer. Further, in the study by Grigsby et al, positive PET appears to be a major predictor of disease progression. While larger confirmatory studies will be necessary, I believe that PET will rapidly move into our armamentarium of diagnostic studies in patients with cervical cancer. Technology is already available to combine CT and PET so that the anatomic landmarks for sites of disease will be more precise. To what degree the introduction of CT/PET will influence the frequency of pretreatment surgical staging remains undetermined.
1. Scheidler J, et al. JAMA. 1997;278:1096-1101.
2. Rose PG, et al. J Clin Oncol. 1999;17:41-45.
3. Reinhardt MJ, et al. Radiology. 2001;218:776-782.
4. Grigsby PW, et al. J Clin Oncol. 2001;19:3745-3749.