BERKELEY, CA (UroToday.com) - In the current era, prostate cancer (PCa) diagnoses at lower stages may have reduced the risk of occult lymph node metastasis for every PSA screening and Gleason score (GS). General agreement has been reached on the necessity of omitting the limited pelvic lymph node dissection (PLND) due to the high rates of false negative findings, arguing for extended PLND whenever it is indicated.[1, 2] Indeed ePLNDs might be necessary to detect occult lymph node metastases that would not otherwise be detected by limited PLND. PLND should be based on anatomical definitions, rather than on the number of lymph nodes removed, because PCa nodal metastases do not follow a predefined pathway of metastatic spread.[3]
Thus it has been established that ePLND should always include removal of lymph nodes along the hypogastric artery as well as obturator, external iliac, and hypogastric nodes. Several studies showed that the rate of lymph node invasion (LNI) was significantly associated to the extent of PLND, with an increased probability in proportion of the number of nodes examined, although some patients might have a great number of LN removed even in the presence of incomplete anatomical pelvic dissection.
Despite the low frequency of positive LNI, it has been necessary to create an accurate and predictive nomogram model for LNI detection. In a recent report, Briganti et al. updated a new pre-operative model (nomogram) to predict the presence of LNI at ePLND. The novel model included percentage of positive cores as a predictor of LNI, in addition to established covariates such as PSA, clinical stage, and biopsy Gleason score. This resulted in an improved LNI predictive accuracy in the internal validation setting. However, external validation data of this novel model are still lacking. To address this void, we set to formally validate the Briganti et al.[4] updated nomogram, using multi-centric data of prostate cancer patients treated with radical prostatectomy and ePLND.
Our results showed that the highest discrimination (AUC) in the external validation setting was achieved by the percentage of the positive cores (72%), which further stresses the important role of this covariate as a predictor of LNI. The discrimination of the multivariable model (nomogram) was 79%. The predictions of the model were virtually perfect for LNI-predicted values from 0 to 20%. The nomogram slightly overestimated the rate of LNI for LNI-predicted values above 20%.
We systematically tested the performance characteristics of various cut-offs for discriminating between patients with and without LNI at ePLND, and quantified the number of avoidable PLNDs vs. the number of potentially missed patients with LNI. In our population, 358 (41.6%) patients were classified below the 5% cut-off. Accordingly, a PLND would not be recommended in those patients by the cut-off proposed by Briganti et al.[4] Avoidance of PLND in those 358 individuals would result in missing LNI in 8 patients (7.9% of all patients with histologically confirmed LNI, and 2.2% of all patients with avoidable PLND). This cut-off would miss < 13% of patients with LNI, as was in the original publication by Briganti et al.[4] This is in line with the recommendations given by the NCCN guidelines, which suggest that a cut-off that allows the avoidance of roughly 50% of PLNDs prior to RP, at the cost of missing < 13% of patients with LNI, is considered acceptable. This corresponds to the 5% cut-off even in the external validation setting, thus confirming it as a possible threshold risk for ePLND indications.
The present study has several strengths. First of all, we included 860 consecutive patients (101 with LNI) enrolled in 2 tertiary-referred centers: the multi-centricity and consecutive enrollment of our large sample allow us to consider our population as representative of other cohorts and adequate for an effective external validation. Moreover, 12% of patients demonstrated a preoperative low-risk profile, and 70% an intermediate, according to the D’Amico classification: therefore, we externally validated the predictive nomograms on almost one third of men with low/intermediate risk PCa. Furthermore, the biopsy technique and scheme, with a median of 10 cores removed -- with pathologic analyses performed by the same referral uro-pathologists who have subsequently analyzed both lymph nodal and prostatic specimens -- has improved internal consistency of our population.
Despite these strengths, our study has several limitations. In particular, our external validation is lessened by the extension of nodal dissection (median number of nodes removed: 15 vs.19 from Briganti et al. as well by the extension of biopsy sampling (median number of total cores taken: 10 vs.16). As regards the number of nodes, only 110 men (12.8%) presented less than 10 nodes after ePLND, while the remaining 750 men (87.2%) presented a minimal number of removed nodes above 10, as reported in the update nomogram cohort of Briganti.[4] However, the optimal extent of PLND, defined as the ideal number of removed and examined LN that are required to optimize the yield of loco-regional LN staging, remains controversial. Several patients included in our analysis underwent RP and limited PLND before a precise guideline about ePLND was shared by the urologic community, and inclusion criteria of the two referral centers were based on a specific anatomical region. Thus we did not use cut-off point for the number of removed nodes, because the nomogram should be used preoperatively, when the final number of removed lymph nodes is unknown.
Moreover, our series included cases from different surgeons from both referral institutions, which adds variability to the surgical technique and yet reflects a realistic pattern likely to be found in most institutions. With regard to the extension of biopsy sampling of our population, 10 cores is what is recommended by the British Prostate Testing for Cancer and Treatment Study (LE: 2a), while more than 12 cores is not significantly more conclusive (LE: 1a), as reported in the EAU guidelines.
Even in the presence of these limitations, our study represents the first external validation of Briganti’s updated nomogram, in a large contemporary series, on men treated with RP and ePLND. This nomogram is based on readily and routinely available clinical parameters such as PSA, clinical stage, biopsy Gleason sum, and percentage of positive cores. Clinical stage, biopsy Gleason grade, and the percentage of positive cores involved with PCA are the most reliable predictors of LNI and indicate the need for EPLND. The updated nomogram, predicting lymph node invasion in patients with prostate cancer undergoing extended pelvic lymph node dissection, has been externally validated, demonstrating excellent accuracy and calibration characteristics and a general applicability for predicting the presence of LNI.
References:
- Heidenreich A, Bellmunt J, Bolla M, et al. EAU guidelines on prostate cancer. Part 1: screening, diagnosis, and treatment of clinically localised disease. Eur Urol 2011; 59:61–71.
- Briganti A, Blute ML, Eastham JH, et al. Pelvic lymph node dissection in prostate cancer. Eur Urol 2009; 55:1251–65.
- Mattei A, Fuechsel FG, Bhatta Dhar N, et al. The template of the primary lymphatic landing sites of the prostate should be revisited: Results of a multimodality mapping study. Eur Urol 2008; 53:118e125).
- Briganti A, Larcher A, Abdollah F, et al. Updated Nomogram Predicting Lymph Node Invasionin Patients with Prostate Cancer Undergoing Extended Pelvic Lymph Node Dissection: The Essential Importance of Percentage of Positive Cores. Eur Urol 2012; 61: 480-487.
Written by:
Mauro Gacci and Michele Lanciotti as part of Beyond the Abstract on UroToday.com. This initiative offers a method of publishing for the professional urology community. Authors are given an opportunity to expand on the circumstances, limitations etc... of their research by referencing the published abstract.
Department of Urology, University of Florence, Careggi Hospital, Florence, Italy
More Information about Beyond the Abstract