MR Lymphography in prostate cancer: Clinical Potentials

R.A.M. Heesakkers

Promotor: J.O. Barentsz
Copromotor: G. J. Jager and Y. Hoogeveen
Radboud University Nijmegen
January 25, 2012


In the current treatment of prostate cancer, the presence of pelvic lymph node metastases have a significant impact on the treatment and prognosis. In this thesis we evaluate the current diagnostic modalities and focus on the value of MRL in determine the lymph node status. In chapter 2 the diagnostic accuracy of CT and MRI in detecting lymph node metastases is evaluated in a meta-analysis. This study included 24 papers from 1980 to 2003. From the meta-analysis it was concluded that the performance of CT and MRI in detecting malignant lymph nodes was equally low. The sensitivity of CT and MRI is 42% and 39%, respectively. Therefore, in the assessment of lymph node involvement CT or MRI should not be used unless the a-priori risk of having lymph node metastases is high(>40%). If the risk is lower, both modalities will misrepresent the lymph node status in many patients and misdirect therapeutic strategies. In chapter 3 the results of the multicentre evaluating study with 375 patients are presented. All patients are assessed by most recent CT multi detector scanners (MDCT) and MRL, and underwent PLND or fine-needle aspiration biopsy. Fifty of 61 patients with lymph node metastases are correctly diagnosed. 80% have metastases in normal size lymph nodes. The sensitivity of MDCT is 34% and the sensitivity of MRL is 82%. The negative predictive value is 88% for MDCT and 96% for MRL. Most important is that the post-test-probability of having a negative result after MRL is <4%. For this reason MRL can rule out lymph node metastases and it is justified to omit a PLND. With the improvement of MRI technology, we also evaluated the feasibility of Ferumoxtran-10 enhanced imaging when using a higher fi eld strength (3 Tesla) in chapter 4. Three readers evaluated all images for total image quality, lymph node border delineation, muscle-fat contrast, and vessel-fat contrast. The significantly better muscle-fat contrast, vessel-fat contrast, lymph node border delineation, and total image quality could be attributed to the use of a 3T scanner. MRL can be performed at high magnetic fi eld strengths and results in improved image quality, which may lead to improved detection of small positive lymph nodes. The extent of the PLND is widely discussed. With the generally used, limited PLND, nodes outside the obturator fossa will be missed. Recent studies report that over 30% of positive nodes are exclusively outside the limited dissection area. When performing the limited PLND the lymph node stage can be underestimated and therefore an unnecessary prostatectomy will be performed. An extended PLND however, will detect more lymph nodes, but increases the risk of complications. Chapter 5 describes the feasibility of MRL to detect lymph nodes outside the limited dissection area. 296 patients were included. In 41% of these patients, nodal metastases were exclusively found outside the area of routine pelvic lymphadenectomy. These positive nodes were detected with MRL. In these cases a limited PLND would not have found metastatic lymph nodes and patients would have been under-staged. In chapter 6 a decision analytic model is applied to determine whether the addition of MRL to the diagnostic workup of patients with intermediate or high probability of lymph node metastases is cost effective from a health care perspective. In 63% of 1000 Monte Carlo simulations the PLND strategy is dominated by the MRL strategy. MRL was cost saving and resulted in better patient outcome for patients with prostate cancer. The probability of MRL being inferior is less than 3%. Finally, in the general discussion (chapter 7) the individual studies regarding the diagnostic performance of MRL are put into a broader perspective. It can be concluded that MRL is a valuable addition to the staging process of patients with prostate cancer. In conclusion, the evaluation of pelvic lymph nodes in patients with prostate cancer with Ferumoxtran-10 enhanced MR imaging is very promising. The negative predictive value is high enough to rule out lymph node metastases. Therefore it is justifi ed to omit a PLND. In 41% of all patients, metastatic lymph nodes outside the limited surgical fi eld are found by MRL. Therefore MRL can be used as guidance for PLND and IMRT. MRL is feasible when using 3T scanners. The use of this higher fi eld strength may improve the detection of lymph nodes. Pre-surgical CT should only be used in patients with a very high risk of having lymph node involvement (>40%). Lastly, MRL is cost effective and results in a better patient outcome for patients with prostate cancer.

A pdf file of this publication is available for personal use. Enter your e-mail address in the box below and press the button. You will receive an e-mail message with a link to the pdf file.