The diagnostic accuracy of CT and MRI in the staging of pelvic lymph nodes in patients with prostate cancer: a meta-analysis

Aim

To compare the diagnostic accuracy of computed tomography (CT) and magnetic resonance imaging (MRI) in the diagnosis of lymph node metastases in prostate cancer.

Methods

After a comprehensive literature search, studies were included that allowed construction of contingency tables for detection of lymph node metastases using CT or MRI. In addition, a summary receiver-operating characteristic (ROC) analysis was performed.

Results

A total of 24 studies were included. For CT, pooled sensitivity was 0.42 (0.26–0.56 95% CI) and pooled specificity was 0.82 (0.8–0.83 95% CI). For MRI, the pooled sensitivity was 0.39 (0.22–0.56 95% CI) and pooled specificity was 0.82 (0.79–0.83 95% CI). The differences in performance of CT and MRI were not statistically significant.

Conclusion

CT and MRI demonstrate an equally poor performance in the detection of lymph node metastases from prostate cancer. Reliance on either CT or MRI will misrepresent the patient's true status regarding nodal metastases, and thus misdirect the therapeutic strategies offered to the patient.

Introduction

Prostate cancer is the most common cancer in men and the second most common cause of death among malignancies afflicting men. In circumstances where metastatic disease does not appear to be present and the probability of organ-confined prostate cancer is high, the use of local therapy, such as radical prostatectomy (RP) or various methods of radiation therapy (RT), is associated with a significant likelihood of cure. However, once prostate cancer spreads to the lymphatic tissue, the patient's status is changed to one of systemic disease and the opportunity for cure with a local therapy is either markedly diminished or no longer present. Currently, the recommended strategy to assess risk regarding lymph node involvement involves the use of predictive models employing nomograms, algorithms, and/or neural nets such as those of Partin, Narayan, Cagiannos, and Bluestein.1, 2, 3

These risk assessment tools use inputs such as prostate-specific antigen (PSA), Gleason score, and clinical stage, and employing the methodology of combined variable analysis are able to provide statistically more significant information.4, 5

Patients found to have a low risk for lymph node metastases (<5%) usually receive curative treatment and most often do not undergo further radiological imaging or diagnostic lymph node dissection, whereas patients with higher risk should be referred for additional staging. A number of nodal staging methods are used in patients having an intermediate or high risk for lymph node involvement. The most reliable method available to document nodal involvement is pelvic lymph node dissection (PLND). However, this is both invasive, expensive, and may be associated with significant morbidity. Computed tomography (CT) and magnetic resonance imaging (MRI) are non-invasive procedures that are commonly used for nodal staging. A finding of lymph node metastases with either test can be verified by means of a PLND. Alternatively, a fine-needle aspiration biopsy (FNAB) is often used to provide a decisive answer in the case of a positive imaging result. Frequently, however, the lymph node is hard to reach because of its anatomic position. In addition, FNAB is not a highly sensitive staging procedure and a high false-negative rate of 40% has been reported by Jager et al.⁶ When the result is indecisive or negative using FNAB, an excisional biopsy or a PLND may be performed.

The medical literature has consistently demonstrated that the specificity of CT and MRI in the detection of lymph node metastases is very high. For example, Wolf et al.⁷ reported a specificity of 97% for CT and MRI in finding lymph node metastases in prostate cancer. In contrast, however, in their report they found a sensitivity of only 36%. They concluded that nodal imaging studies should only be recommended for patients having a probability of 45% or higher for lymph node metastases.⁷

The current literature shows a broad range in the diagnostic performance for both CT and MRI. Methodological, as well as patient group, characteristics appear to cause bias and an over- or underestimation of the diagnostic performance of these tests.⁸ A meta-analysis of the diagnostic accuracy of presurgical CT or MRI and the criteria used for staging pelvic lymph nodes was undertaken to evaluate the value of these studies in the staging of men with prostate cancer. In addition, the effect of patient group characteristics and methodological characteristics were investigated on the staging performance of CT and MRI for the diagnosis of lymph node metastases in prostate cancer.