Tomography for Preoperative Lymph-Node Staging in Intermediate-Risk and High-Risk Prostate Cancer: Comparison with Clinical Staging Nomograms

Transcription

1 european urology 54 (2008) available at journal homepage: Prostate cancer 11 C-Choline Positron Emission Tomography/Computerized Tomography for Preoperative Lymph-Node Staging in Intermediate-Risk and High-Risk Prostate Cancer: Comparison with Clinical Staging Nomograms Riccardo Schiavina a, *, Vincenzo Scattoni b, Paolo Castellucci c, Maria Picchio d, Barbara Corti e, Alberto Briganti b, Alessandro Franceschelli a, Francesco Sanguedolce a, Alessandro Bertaccini a, Moshen Farsad c, Giampiero Giovacchini d, Stefano Fanti c, Walter Franco Grigioni e, Ferruccio Fazio d, Francesco Montorsi b, Patrizio Rigatti b, Giuseppe Martorana a a Department of Urology, University of Bologna, S. Orsola-Malpighi Hospital, Italy b Department of Urology, University Vita-Salute, Scientific Institute San Raffaele, Milan, Italy c Department of Nuclear Medicine, University of Bologna, S. Orsola-Malpighi Hospital, Bologna, Italy d Department of Nuclear Medicine, University Milano-Bicocca, Scientific Institute San Raffaele, IBFM-CNR, Milan, Italy e F. Addarii Institute of Oncology, University of Bologna, S. Orsola-Malpighi Hospital, Bologna, Italy Article info Article history: Accepted April 10, 2008 Published online ahead of print on April 18, 2008 Keywords: 11 C-Choline-PET/TC Imaging Lymph node metastasis Nomogram Pelvic lymph node dissection Prostate cancer Radical prostatectomy Staging Abstract Background: Conventional imaging (CI) techniques are inadequate for lymph node (LN) staging in prostate cancer (PCa). Objectives: To assess the accuracy of 11 C-Choline positron emission tomography/computerized tomography (PET/CT) for LN staging in intermediate-risk and high-risk PCa and to compare it with two currently used nomograms. Design, Setting, and Participants: From January 2007 to September 2007, 57 PCa patients at intermediate risk (n = 27) or high risk (n = 30) were enrolled at two academic centres. All patients underwent preoperative PET/CT and radical prostatectomy with extended pelvic LN dissection (PLND). Risk of LN metastasis (LNM) was assessed using available nomograms. Measurements: Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and number of correctly recognized cases for LNM detection at PET/CT were assessed. The accuracy of PET/CT for LNM detection was compared with the accuracy of nomograms for LNM prediction by using receiver operating characteristic (ROC) analysis. Results and Limitations: Fifteenpatients (26%) hadlnms, anda totalof41lnmswere identified. On a patient analysis, sensitivity, specificity, PPV, NPV, and number of correctly recognized cases at PET/CT were 60.0%, 97.6%, 90.0%, 87.2%, and 87.7% while, on node analysis, these numbers were 41.4%, 99.8%, 94.4%, 97.2%, and 97.1%. The mean diameter (in mm) of the metastatic deposit of true-positive LNs was significantly higher than that of false-negative LNs (9.2 vs 4.2; p = 0.001). PET/ CT showed higher specificity and accuracy than the nomograms; however, in pairwise comparison, the areas under the curve (AUCs) were not statistically different (all p values >0.05). Conclusions: In patients with intermediate-risk and high-risk PCa, 11 C-Choline PET/CT has quite a low sensitivity for LNM detection but performed better than clinical nomograms, with equal sensitivity and better specificity. # 2008 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Urology, University of Bologna, S. Orsola-Malpighi Hospital, Via Palagi 9, Bologna, Italy. Tel ; Fax: address: rschiavina@yahoo.it (R. Schiavina) /$ see back matter # 2008 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 european urology 54 (2008) Introduction In patients with newly diagnosed prostate cancer (PCa), the knowledge of lymph node (LN) status is of key importance for appropriate treatment planning. The incidence of lymph node metastasis (LNMs) has dramatically decreased in the prostate-specific antigen (PSA) era, but it may still range from 1 to 26%, depending on the patient populations, the extent of pelvic LN dissection (PLND), and the quality of histopathological analysis [1,2]. Conventional imaging (CI) techniques are still inadequate, and PLND remains the only method for an accurate staging [3]. Several nomograms have been developed to quantify the risk of LNMs and consequently the need for PLND, but a probability cut-off that separates patients at high risk from those at low risk has not yet been determined [1,4,5]. Positron emission tomography with the tracer fluorine-18 fluorodeoxyglucose ( 18 FDG-PET) has proven to be inaccurate in PCa staging, and alternative radiopharmaceuticals (basically radiolabelled Choline and Acetate) have been tested. The first report on 11 C-Choline for the LN staging was encouraging; however, as poorer results with 18 F- Choline have been reported, further studies are needed [6 10]. The aim of the present study was to evaluate the accuracy of 11 C-Choline positron emission tomography/computerized tomography (PET/CT) for preoperative LN staging in intermediate-risk and high-risk prostate cancer (PCa) and to compare it with two currently used nomograms. 2. Methods 2.1. Patients From January 2007 to September 2007, patients with biopsyproven PCa who were suitable for surgical treatment were prospectively evaluated at our centres. Two inclusion criteria were used: (1) intermediate-risk PCa, defined as ct1 2 and PSA between 10 and 20 ng/ml and/or Gleason score (Gs) = 7 or (2) high-risk PCa, defined as PSA 20 and/or Gs = 8 10 and/or ct3 [11]. Two exclusion criteria were used: (1) a positive bone scan (performed in cases of PSA > 20) or suspected extra-nodal disease at PET/CT or (2) hormonal therapy at the moment of PET/CT investigation or before surgery. Fifty-seven patients underwent PET/CT preoperatively, and thereafter a radical prostatectomy (RP) with extended PLND (eplnd) was performed. Forty-five patients were enrolled at the S. Orsola- Malpighi Hospital in Bologna, and 12 were enrolled at the S. Raffaele Institute in Milan. Clinical stage was assessed with digital rectal examination and transrectal ultrasound (TRUS). Two nomograms for predicting pathological stage (ie, the Briganti nomogram and the Kattan nomogram), based on preoperative PSA, biopsy Gs, and clinical stage, were used to assess the risk of LNMs in each patient [4,5]. The study was performed in accordance with the Helsinki declaration and national regulations. All patients provided informed consent for anonymous publication of data Radiopharmaceuticals and imaging protocol PET/CT was performed an average of 19.5 d 4 SD (median: 20; range 12 28) after TRUS biopsy. PET/CT studies were acquired with two integrated PET/CT systems, either Discovery LS or Discovery ST (General Electric Medical Systems, Waukesha, Wisconsin, USA). Once the scan range from neck to pelvis was defined, a CT scan was performed. PET data of the whole-body tracer distribution were then acquired (4 min per bed) in 2-D mode (from pelvis to neck), starting 5 min after injection of about 370 MBq of 11 C-Choline. Attenuation correction was performed using CT images. CT and PET images were matched and fused into transaxial, coronal, and sagittal images. Maximum standardized uptake value (SUVmax) was computed after normalization to the decay-corrected injected dose and body weight [12]. Using the same protocol in each study centre, a group of experienced nuclear medicine physicians evaluated the images in consensus to localize the sites of pathological 11 C-Choline uptake Surgical technique Surgical treatment was performed an average of 21.2 d 8.5 SD (median: 23; range 3 35) after PET/CT. Patients underwent RP with eplnd either with an open approach (38 patients) or with a transperitoneal laparoscopic approach (7 patients) at the Department of Urology in Bologna and with an open approach (12 patients) at the Department of Urology in Milan. Extended PLND was performed up to and including the bifurcation of the common iliac artery, the fibrofatty tissue along the external iliac vessels (the distal limit being the deep circumflex vein and femoral canal), the internal iliac vessels, and the obturator fossa. The lateral limit consisted of the pelvic sidewall, and the medial limit was the perivesical fat. Common iliac and presacral nodes were removed only in the case of positive PET/CT in these locations. All eplnd specimens were prospectively mapped according to their anatomic location Histological examination Formalin-fixed tissue samples were routinely processed. LNs were microscopically examined for metastatic disease by an experienced uropathologist at each study centre. When necessary an immunohistochemical analysis to confirm the presence of neoplastic cells was carried out. The total number of LNs and the number of metastatic LNs were assessed according to their location. For each metastatic LN the diameter (expressed in mm) of the metastatic deposits was recorded. Pathological tumor stage, Gleason grade and score of the prostatectomy specimen were determined according to the 2002 TNM classification [13]. Disease positivity was defined as the presence of any metastatic deposits of PCa in the LNs

3 394 european urology 54 (2008) examined. Histopathological analysis of LNs was used as the reference standard Image reconstruction and data analysis The diagnosis of tumor-positive LNs on PET/CT images (test positivity) was based, in both study centres, on the visual assessment of the focal increased 11 C-Choline uptake on PET images, the location of which corresponded to LN chains on CT images; these included the external iliac, the internal iliac, the common iliac, the obturator fossa nodes bilaterally, and the presacral nodes. At PET/CT imaging, LNs with increased tracer uptake were considered positive for metastatic spread, even when they were <10 mm in short-axis diameter; conversely, LNs with no detectable tracer uptake were not considered pathological at PET/CT, even when they were >10 mm. The SUVmax was measured in each LN with increased tracer uptake and in each prostate. Positive matching was defined as correspondence between disease positivity and test positivity Statistical analysis Means SD, medians, range, and frequencies were used as descriptive statistics. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and number of correctly recognized cases at PET/CT in the detection of LNMs were calculated for a patient analysis and for a node analysis. Univariate and multivariate logistic regression was applied to search for relevant predictive factors for LNM detection at PET/CT (eg, possible factors influencing PET/CT sensitivity) considering age, PSA, clinical and pathological Gs, clinical and pathological tumor stage, diameter of the metastatic deposit in the LNs, location of LNMs, and number of LNMs removed as independent variables. The odds ratios (OR) together with their 95% confidence intervals (CI) were also reported. Mann-Whitney tests, Kruskal-Wallis tests, and Spearman rank correlation were used as appropriate. For each patient, the risk of LNM as predicted by the two nomograms was assessed; thereafter, the mean and the median risk of LNM of the whole population and of the two risk groups were compared to the observed prevalence of LNM. The accuracy of PET/CT for detecting LNM (patient analysis) was compared to the accuracy of the two nomograms for predicting LNM using receiver operating characteristic (ROC) analysis; the areas under the ROC curves (AUC) were compared using the Matel-Haentzel test. Analyses were run on the SPSS statistical package (version 13.0 for Windows). Two-tailed p < 0.05 were considered significant. 3. Results Patient characteristics are presented in Table 1. Fifteen of 57 patients (26%) had histologically proven LNMs, of whom 6 of 27 (22%) were in the intermediate-risk group and 9 of 30 (30%) were in the high-risk group. Table 2 shows the characteristics of the 15 patients with LNMs. Three of 15 patients (20%) had LNMs only in the region of the standard PLND (defined as PLND limited to external iliac and obturator regions), 9 of 15 (60%) had LNMs inside as well as outside the region of the standard PLND, and 3 of 15 (20%) had LNMs only outside the region of the standard PLND. Twelve LNMs were located in the external iliac region, 9 in the obturator fossa, 15 in the internal iliac region, 3 in the common iliac region, and 2 in the presacral region. Twenty out of 41 LNMs (48%) were outside the region of standard PLND. Table 3 shows the detection rate of PET/CT for LNMs according to the diameter of the metastatic deposit. The mean diameter of the metastatic deposits of true positive LNs was significantly higher ( mm, median: 8.0 range: 3 28) than the mean diameter of the metastatic deposit of false negative LNs ( mm, median: 3.0 range: ; p = 0.001). Fig. 1 shows an example of intraprostatic and presacral 11 C-Choline-PET/CT focal uptake (in patient no. 4). Table 4 reports the numbers of Table 1 Characteristics of the total patient population (57 patients) Age (yr) Mean ( SD) 65 ( 6.1) Median (range) 67 (46 75) PSA (ng/ml) Mean ( SD) 16.5 ( 13.5) Median (range) 12.0 ( ) No. clinical Gs (%) (39%) 7 27 (47%) (14%) No. clinical stage (%) T1c 5 (9%) T2 36 (63%) T3a 16 (28%) No. pathological Gs (%) (19%) 7 38 (66%) (15%) No. pathological stage (%) T2 21 (37%) T3a 15 (26%) T3b 20 (35%) T4 1 (2%) Total no. of LNs removed 892 No. LNs removed/pts Mean ( SD) 15.6 ( 5.5) Median (range) 14.0 (9 32) No. of N1 patients (%) 15/57 (26%) Bologna 12/45 (26%) Milan 3/12 (25%) Total no. of positive LNs 41 PSA: prostate-specific antigen; Gs: Gleason score; LNs: lymph nodes; SD: standard deviation.

4 Table 2 Characteristics of the patients with lymph node metastases (LNMs) (15 patients) Patient No. Age (yr) PSA (ng/ml) ct cgs pt pgs No. of LNM Size of metastatic deposit (mm) PET (SUVmax) Positive LN site T3a T3b , 8.8, 7.7, 7.2, neg r ext iliac, l com iliac, l ext iliac, l obt, l int iliac T2c T3a r int iliac T2c T3a , 4.3, neg, neg l int iliac, pres, l ext iliac, r ext iliac T2c T3b , neg pres, r obt T3a T3b , 14, 14 r int iliac, l ext iliac, l com iliac T3a T3b , neg, neg r obt, r int iliac, l ext iliac T3a T3a , 5.9, 4.7, neg r obt, r com iliac, r ext iliac, r int iliac T2c T3b , neg, neg r obt, r obt, r ext iliac T2c T3b , neg r int iliac, l int iliac T2c T3b neg l ext iliac, l int iliac, r int iliac T3a T3b neg r int iliac T3a T3b neg l ext iliac, l int iliac, l int iliac, l obt, r int iliac, l obt T3a T3a neg l obt T2b T3b neg r int iliac, l ext iliac T2b T2c neg l ext iliac Mean SD Median (Range) 66.0 (61 75) 11.0 (4 70) 7.0 (6 9) 7.0 (6 10) (1 6) 5.0 ( ) PSA: prostate-specific antigen; ct: clinical tumor stage; cgs: clinical Gleason score; pt: pathological tumor stage; pgs: pathological Gleason score; PET: positron emission tomography; SUVmax: maximum standardized uptake value; r: right; l: left; int: internal; ext: external; com: common; obt: obturator. european urology 54 (2008)

5 396 european urology 54 (2008) Fig. 1 Example of right intraprostatic (yellow arrow) and presacral (red arrow) 11 C-Choline positron emission tomography/ computerized tomography (PET/CT) focal uptake in patient with PCa Gs 5 + 5, pt3b and a presacral lymph node metastasis (LNM;diameter of metastatic deposit: 10 mm). Another LNM in the right obturator region (diameter of metastatic deposit: 1,2 mm) was not visualized. Table 3 Detection rate by positron emission tomography/computerized tomography (PET/CT) according to the diameter of the metastatic deposit in the (lymph node) LN Diameter of metastatic deposit (mm) (No. of LNs) Detection rate of PET/CT (n = 6) 0 (0%) (n = 10) 3 (30%) (n = 16) 7 (43%) >10 (n = 9) 7 (77%) Total (n = 41) 17 (41%) correctly recognized cases at PET/CT for the detection of LNMs based on an analysis of each patient and an analysis of each LN. At patient analysis, 8 of 12 patients (66.6%) and 1 of 3 patients (33.3%) with LNM were correctly identified by PET/CT at the S. Orsola-Malpighi Hospital in Bologna and at the S. Raffaele Institute in Milan respectively. Globally, the sensitivity of PET/CT for LNM detection was 60.0% for patient analysis and 41.4% for node analysis. Logistic regression analysis was performed for the 892 removed LNs to search for relevant predictive factors that could influence the sensitivity of PET/ CT. At univariate analysis, the only factors that reached significance were the number of LNMs ( p = 0.001, odds ratio (OR) = 2.042, 95% confidence interval (CI): ) and the diameter of metastatic deposits ( p = 0.001, OR = 1.440, 95% CI: ); both the variables were confirmed to be significant independent predictive factors in a backward stepwise multivariate analysis (for number of LNM: p = 0.001, OR = 1.815, 95% CI: ; for diameter of metastatic deposits: p = 0.001, OR = 1.280, 95% CI: ), suggesting that LN detection at PET/CT could be influenced by the number of LNMs and the diameter of metastatic deposits. Table 5 shows the observed prevalence of LNM in the patient population compared with the rate of LNM predicted with two nomograms: the median risk of LNMs using the Briganti nomogram was very Table 4 Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and number of correctly recognized cases with positron emission tomography/computerized tomography (PET/CT) in the detection of lymph node metastases (LNMs) Sensitivity Specificity PPV NPV No. correctly recognized cases Patient analysis (n = 57) 60.0% (9 of 15) 97.6% (41 of 42) 90.0% (9 of 10) 87.2% (42 of 48) 87.7% (50 of 57) Node analysis (n = 892) 41.4% (17 of 41) 99.8% (850 of 851) 94.4% (17 of 18) 97.2% (851 of 875) 97.1% (867 of 892)

6 european urology 54 (2008) Table 5 Observed prevalence of lymph node metastasis (LNM) in the patient populations compared with the mean (W SD) and the median (range) rate of LNM predicted with two nomograms Observed prevalence of LNMs Rate of LNMs predicted with Briganti nomogram Rate of LNMs predicted with Kattan nomogram Whole population (n = 57) 26.3 (15/57) (5 80) 6.0 (2 69) Patients at intermediate risk (n = 26) 22.2 (6/26) (5 24) 4.0 (2 10) Patients at high risk (n = 31) 29.0 (9/31) (5 80) 12.5 (3 69) SD: standard deviation. Table 6 Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of positron emission tomography/computerized tomography (PET/CT) in the detection of lymph node metastases (LNMs) and of the nomograms in the prediction of LNMs Sensitivity (95% CI) Specificity (95% CI) PPV NPV AUC (95% CI) PET/CT 60.0% ( ) 97.6% ( ) 90.0% 87.2% ( ) Briganti nomogram ( * cut-off > 25%) 60.0% ( ) 73.8% ( ) 45.0% 83.8% ( ) Kattan nomogram ( * cut-off > 9%) 60.0% ( ) 64.3% ( ) 37.5% 81.8% ( ) CI: confidence interval; AUC: area under the curve analysis. * Cut-off value corresponding to the highest level of accuracy (minimal false-negative and false-positive results) for each nomogram. similar to the observed prevalence of LNMs in the whole population as well as in the different risk groups, while the Kattan nomogram underestimated the prevalence of LNMs. Table 6 shows the accuracy of PET/CT in detecting LNM in comparison with the accuracy of nomograms in predicting LNM; for each nomogram a cutoff value corresponding to the highest level of accuracy (minimal false-negative and false-positive results) was utilized. PET/CT showed better specificity, PPV, NPV, and accuracy than the nomograms; Fig. 2 Receiver operating characteristic (ROC) curves of positron emission tomography/computerized tomography (PET/CT) for detecting lymph node metastasis (LNM) and of the two nomograms for predicting LNM. however, in pairwise comparison, the AUCs were not statistically different (all p values >0.05; Fig. 2). In two patients, focal 11 C-Choline uptake was noted at the inguinal level, but this finding was interpreted as an inflammatory response because of the location. There were no positive bone scans among the 15 patients who underwent this exam preoperatively. 4. Discussion The optimal therapy for PCa with LNMs is still under discussion, but surgery combined with hormonal therapy may provide an excellent cancer-specific survival rate [14 16]. Preoperative cross-sectional modalities like CT and magnetic resonance imaging (MRI) are very limited in the nodal staging of newly diagnosed PCa, with a sensitivity generally of less than 30% [3,17]. Recent studies have indicated the suitability of radioisotope guided sentinel LN (SLN) dissection for PCa staging, with a sensitivity of 97% and an NPV of 98%; nevertheless this technique can be technically demanding, expensive, and time-consuming for the surgeons as well as for pathologists [18,19]. PET with radiolabelled choline (both 11 C-Choline and 18 F-Choline) and with 11 C-Acetate has emerged as a promising new technique for PCa imaging [6,7,9,10,12,20 23]. In particular, 11 C-Choline PET has been shown to present higher sensitivity than CI [7,10,20,23]. 18 F-labeled analogue of 11 C-Choline was

7 398 european urology 54 (2008) proposed to circumvent the principal disadvantage of 11 C-Choline which is the short half-life of 11 C (20 min), which limits its use to on-site cyclotron centres [6]. However, 18 F-Choline seems to present a higher secretion into the urinary system, which represents a relevant disadvantage in the imaging of PCa. 11 C-Acetate (which seems very similar to 11 C- Choline) has been proposed as an alternative tracer, but few studies are yet available [6]. De Jong et al had excellent results with 11 C-Choline PET in the preoperative LN staging of 67 newly diagnosed PCa, with a sensitivity of 80% (patient analysis) [7]. However, it must be noted that the mean preoperative PSA of patients with LNMs was 123 ng/ml, which may not represent the typical presentation of high-risk PCa in the PSA era. Recently Hacker et al noted a lower sensitivity of 18 F-Choline PET/CT scans (10%) compared to SLN dissection (80%) in 20 patients with intermediate- and high-risk, clinically localized PCa and concluded that 18 F-Choline PET/CT is not useful for occult LNM [8]. Of note, the mean diameter of LNM in the study of Hacker et al was 3.8 mm, which is less than the spatial resolution of the PET scanner. Scattoni et al evaluated the accuracy of 11 C-Choline PET/CT in the diagnosis of LNM in 25 patients with biochemical failure after surgery and treated with pelvic or retroperitoneal LN dissection [12]. In a per patient analysis, the sensitivity was 100%, while it decreased to 64% in a per lesion analysis. In the present study 11 C-Choline PET/CT is highly specific (97.6%) for LNMs, but it has a fairly low sensitivity (60%) and NPV (87%) (patient analysis). Certainly the limited spatial resolution of PET/CT scanners (4 5 mm) may hamper the recognition of micrometastasis; however, it must be noted that 10 of 25 metastatic deposits >5 mm were not visualized by PET/CT, and thus the metabolic state of the tumor may play a major role in some of these cases; for these reasons, a negative 11 C-Choline PET/CT is not sufficient to rule out a PLND. Another factor that may influence the sensitivity of PET/CT is the number of LNMs in the single patient independently of the Gs, the PSA, or the tumor stage. The only falsepositive node was located in the external iliac region in a node-negative patient, and the histopathology revealed an inflammatory response. Finally the high NPV, specificity, and accuracy of the node analyses (Table 6) are probably due to the high proportion of true-negative results. Nowadays it is well accepted that eplnd can identify LNMs that would not otherwise be detected by limited or standard PLND [2 24]. Our results confirm that the sampling of the obturator and external iliac regions may be inadequate; it is worth noting that PET/CT visualized LNMs outside the field of a standard PLND in four patients and was useful in the detection of presacral/common iliac metastatic nodes in three. PLND still remains the gold standard for LN staging in PCa, but it is time-consuming and adds to the cost and to the morbidity of RP [25]. In the absence of accurate imaging techniques for LN staging, many nomograms and algorithms have been developed to evaluate the risk of LNMs and ultimately the need for PLND [1,4,5,26]. On the basis of these preoperative tools, most institutions do not perform PLND in lowrisk PCa (defined by ct1c, Gs 6, PSA < 10 or absence of biopsy cores containing Gleason grade 4) because the incidence of LNMs is said to be only 1 5% [1,5,27]. Moreover, it has been demonstrated that biopsy information (either the number or the percentage of positive biopsy core) may improve the ability of other clinical variables to predict the rate of LNM [26]. Nevertheless, some authors have reported an incidence of about 10% LNMs even in a low-risk group based on meticulous PLND, and PLND should probably be discussed on an individual basis even in this group [2,4,24,28 30]. By contrast, PLND should always be recommended in intermediate-risk or high-risk PCa in which the likelihood of LNM is 20 25% and 30 60%, respectively [27,29,30]. The principal limitation of the preoperative nomograms is that they were developed on the basis of standard PLND and may underestimate the actual risk of LNM [1,5]. Briganti et al. developed the first nomogram based on eplnd that does not have this limitation [4]. In fact, in our study the median rate of LNMs predicted with the Briganti nomogram was comparable to the observed prevalence of LNMs in the patient populations (Table 5) while the Kattan nomogram underestimated the risks of LNM; this is certainly due to the differing extent of PLND utilized in the construction of the two nomograms. In a direct comparison of accuracy, the AUC of PET/ CT was higher than that obtained with the nomograms, even if this difference was not statistically significant. Most importantly, PET/CT had a higher specificity than clinical nomograms; in fact in patients with intermediate-risk and high-risk PCa, nomograms may merely reinforce the desirability of further LN assessment, and the correct cut-off to define the low-risk population has not yet been well defined. At the present time, the use of 11 C-Choline-PET/ CT imaging for preoperative staging should be limited to research protocols because of the high costs, and routine clinical use cannot be recommended. Furthermore, current PET/CT scanners

8 european urology 54 (2008) may be not suitable for the detection of micrometastasis because of their limited spatial resolution. Moreover, the short half-life of 11 C limits the use of 11 C-Choline to only those PET centres that have on-site cyclotrons. At this moment, large prospective studies including a cost-effectiveness analysis are needed to assess the real impact of 11 C-Choline PET/CT in PCa staging. In terms of study limitations, the study was not designed to directly compare PET/CT with CI (ie, contrast enhanced CT or MRI). Furthermore, the median number of LNs removed in our study is lower than those reported by other authors. Finally, both of the nomograms were developed and validated on populations consisting mainly of lowrisk PCa, and this may limit the comparison with PET/CT; however, the performance characteristics of Briganti s nomogram showed close agreement between predicted probabilities and observed rate of LNM among patients with intermediate and high risk of LNM, and this demonstrates the excellent predictive performance of the nomogram in highrisk patients despite the lack of a specific validation in this subset of patients. 5. Conclusions In patients with intermediate-risk and high-risk PCa, 11 C-Choline PET/CT performed better than clinical nomograms for LNM detection, with equal sensitivity and better specificity. The sensitivity of 11 C-Choline- PET/CT was quite low at patient analysis (60%) but the diagnostic accuracy of this technique seems to be better than that of the current imaging techniques as reported in the literature. Author contributions: Riccardo Schiavina had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Schiavina, Martorana, Scattoni, Franceschelli. Acquisition of data: Schiavina, Castellucci, Farsad, Giovacchini, Picchio, Fanti, Corti, Grigioni. Analysis and interpretation of data: Schiavina, Franceschelli. Drafting of the manuscript: Schiavina, Briganti, Picchio. Critical revision of the manuscript for important intellectual content: Schiavina. Statistical analysis: Schiavina, Sanguedolce. Obtaining funding: Schiavina, Martorana, Bertaccini. Administrative, technical, or material support: Schiavina. Supervision: Schiavina, Martorana, Fazio, Rigatti, Montorsi. Other (specify): none. Financial disclosures: I certify that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: none. Funding/Support and role of the sponsor: SAMUR Association (Association for Advanced Studies of Urological Diseases) and the del Monte di Bologna e Ravenna Foundation provided funding for collection of data and analysis. References [1] Makarov DV, Trock BJ, Humphreys EB, et al. Updated nomogram to predict pathologic stage of prostate cancer given prostate-specific antigen level, clinical stage, and biopsy Gleason score (Partin tables) based on cases from 2000 to Urology 2007;69: [2] Heidenreich A, Varga Z, Von Knobloch R. Extended pelvic lymphadenectomy in patients undergoing radical prostatectomy: high incidence of lymph node metastasis. J Urol 2002;167: [3] Sadeq A, Ian D, Himu L. Baseline staging of newly diagnosed prostate cancer: a summary of the literature. J Urol 2004;171: [4] Briganti A, Chun FK-H, Salonia A, et al. Validation of a nomogram predicting the probability of lymph node invasion among patients undergoing radical prostatectomy and an extended pelvic lymphadenectomy. Eur Urol 2006;49: [5] Cagiannos A, Karakiewicz P, Eastham JA, et al. A preoperative nomogram identifying decreased risk of positive pelvic lymph nodes in patients with prostate cancer. J Urol 2003;170: [6] Powles T, Murray I, Brock C, Oliver T, Avril N. Molecular positron emission tomography and PET/CT imaging in urological malignancies. Eur Urol 2007;51: [7] de Jong IJ, Pruim J, Elsinga PH, Vaalburg W, Mensink HJ. Preoperative staging of pelvic lymph nodes in prostate cancer by (11)C-choline PET. J Nucl Med 2003;44: [8] Häcker A, Jeschke S, Leeb K, et al. Detection of pelvic lymph node metastases in patients with clinically localized prostate cancer: comparison of [18F]fluorocholine positron emission tomography-computerized tomography and laparoscopic radioisotope guided sentinel lymph node dissection. J Urol 2006;176: [9] Oyama N. Editorial comment on: Detection of lymphnode metastases with integrated [ 11 C]choline PET/CT in patients with PSA failure after radical prostatectomy: results confirmed by open pelvic-retroperitoneal lymphadenectomy. Eur Urol 2007;52:429. [10] Picchio M, Messa M, Landoni C, et al. Value of [11C]choline-positron emission tomography for re-staging prostate cancer: a comparison with [18F]fluorodeoxyglucosepositron emission tomography. J Urol 2003;169: [11] Yossepowitch O, Eggener SE, Bianco FJ, et al. Radical prostatectomy for clinically localized, high-risk prostate

9 400 european urology 54 (2008) cancer: critical analysis of risk assessment methods. J Urol 2007;178: [12] Scattoni V, Picchio M, Suardi N, et al. Detection of lymphnode metastases with integrated [11C]choline PET/CT in patients with PSA failure after radical retropubic prostatectomy: results confirmed by open pelvic-retroperitoneal lymphadenectomy. Eur Urol 2007;52: [13] Sobin LH, Wittekind C, editors. TNM Classification of Malignant Tumours. Ed. 6. New York: Wiley-Liss; [14] Cheng L, Zincke H, Blute ML, et al. Risk of prostate carcinoma death in patients with lymph node metastasis. Cancer 2001;91: [15] Boorjian SA, Thompson RH, Siddiqui S, et al. Long-term outcome after radical prostatectomy for patients with lymph node positive prostate cancer in the prostate specific antigen era. J Urol 2007;178: [16] Allaf ME, Palapattu GS, Trock BJ, Carter HB, Walsh PC. Anatomical extent of lymph node dissection: impact on men with clinically localized prostate cancer. J Urol 2004;172: [17] Hricak H, Choyke PL, Eberhardt SC, Leibel SA, Scardino PT. Imaging prostate cancer: a multidisciplinary perspective. Radiology 2007;243: [18] Weckermann D, Dorn R, Trefz M, Wagner T, Wawroschek F, Harzmann R. Sentinel lymph node dissection for prostate cancer: experience with more than 1000 patients. J Urol 2007;177: [19] Weckermann D, Dorn R, Holl G, Wagner T, Harzmann R. Limitations of radioguided surgery in high-risk prostate cancer. Eur Urol 2007;51: [20] de Jong IJ, Pruim J, Elsinga PH, Vaalburg W, Mensink HJA. Visualization of prostate cancer with 11 C-choline positron emission tomography. Eur Urol 2002;142: [21] Farsad M, Schiavina R, Castellucci P, et al. Detection and localization of prostate cancer: correlation of (11)C-choline PET/CT with histopathologic step-section analysis. J Nucl Med 2005;46: [22] Martorana G, Schiavina R, Corti B, et al. 11C-choline positron emission tomography/ computerized tomography for tumor localization of primary prostate cancer in comparison with 12-core biopsy. J Urol 2006;176: [23] Testa C, Schiavina R, Lodi R, et al. Prostate cancer: sextant localization with MR imaging, MR spectroscopy, and 11Ccholine PET/CT. Radiology 2007;244: [24] Bader P, Burkhard FC, Markwalder R, Studer UE. Is a limited lymph node dissection an adequate staging procedure for prostate cancer? J Urol 2002;168: [25] Briganti A, Chun FK-H, Salonia A, et al. Complications and other surgical outcomes associated with extended pelvic lymphadenectomy in men with localized prostate cancer. Eur Urol 2006;50: [26] Briganti A, Karakiewicz PI, Chun FK-H, et al. Percentage of positive biopsy cores can improve the ability to predict lymph node invasion in patients undergoing radical prostatectomy and extended pelvic lymph node dissection. Eur Urol 2007;51: [27] Conrad S, Graefen M, Pichlmeier U, et al. Prospective validation of an algorithm with systematic sextant biopsy to predict pelvic lymph node metastasis in patients with clinically localized prostatic carcinoma. J Urol 2002; 167: [28] Schumacher MC, Burkhard FC, Thalmann GN, et al. Is pelvic lymph node dissection necessary in patients with a serum PSA < 10 ng/ml undergoing radical prostatectomy for prostate cancer?. Eur Urol 2006;50: [29] Weckermann D, Goppelt M, Dorn R, et al. Incidence of positive pelvic lymph nodes in patients with prostate cancer, a prostate-specific antigen (PSA) level of 10 ng/ml and biopsy Gleason score of 6, and their influence on PSA progression-free survival after radical prostatectomy. BJU Int 2006;97: [30] Heidenreich A, Ohlmann CH, Polyakov S. Anatomical extent of pelvic lymphadenectomy in patients undergoing radical prostatectomy. Eur Urol 2007;52: Editorial Comment on: 11 C-Choline-Positron Emission Tomography/Computerized Tomography for Preoperative Lymph-Node Staging in Intermediate-Risk and High-Risk Prostate Cancer: Comparison with Clinical Staging Nomograms Felix K.-H. Chun Department of Urology, University Hospital Hamburg- Eppendorf, University of Hamburg, Hamburg, Germany chun@uke.uni-hamburg.de The preoperative lymph node staging of prostate cancer patients is of key importance for adequate treatment selection or (adjuvant) treatment modification. In absence of an accurate diagnostic tool, this risk stratification is done by the use of clinical experience, categorized decision aids, or individualized nomograms [1]. In the field of lymph node risk stratification, Conrad et al, Cagiannos et al, and Briganti et al have pioneered this approach by developing models based on clinical variables to predict the probability of a lymph node invasion [2 4]. Despite their clinical usefulness and acceptance within the urologic community, novel imaging techniques such as 11Choline PET/CT (11Choline positron emission tomography/computerized tomography) raise substantial hopes for accurate preoperative staging in patients with prostate cancer [5]. The current paper by Schiavina et al [6] contributes substantially to this exciting new field of prostate cancer. The authors investigated 11Choline PET/CT for lymph node staging. Furthermore, they

10 european urology 54 (2008) compared their findings to two established nomograms for preoperative lymph node prediction. The strength of the study stems from its prospective study design and the extent of lymph node dissection. The mean number of nodes removed was 16, which is important, since in the evaluation of 11Choline PET/CT the proportions of true negative/positive and false negative/positive patients are directly related to the number of nodes removed. Therefore, the more nodes that are being removed the better the comparability with 11Choline PET/CT. Furthermore, the authors performed per patient as well as per lymph node analyses resulting in high sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). However, smaller lymph node metastases demonstrated a higher rate of false-negative results at PET/CT. Finally, in the last step of their analyses, the head-to-head comparison of the novel technique versus the two established nomograms of Cagiannos et al. and Briganti et al exhibited higher accuracy. However, this difference did not reach statistical significance [6]. Taken together, Schiavina et al [6] present promising results within this emerging field of molecular imaging. Their findings may motivate us to perform larger-scale prospective trials to define the role of 11Choline PET/CT. However from a clinical viewpoint, one may question the benefit of 11Choline PET/CT in an intermediate- to highrisk patient cohort in which extended pelvic lymph node dissection (PLND) may have sufficed to stage the individual most accurately [7]. Finally, potential impact on the health care systems in terms of costeffectiveness of molecular imaging techniques needs also a mention. References [1] Chun FK-H, Karakiewicz PI, Briganti A, et al. Prostate cancer nomograms: an update. Eur Urol 2006;50: (discussion 926). [2] Conrad S, Graefen M, Pichlmeier U, Henke RP, Erbersdobler A, Hammerer PG, et al. Prospective validation of an algorithm with systematic sextant biopsy to predict pelvic lymph node metastasis in patients with clinically localized prostatic carcinoma. J Urol 2002;167: [3] Cagiannos I, Karakiewicz P, Eastham JA, Ohori M, Rabbani F, Gerigk C, et al. A preoperative nomogram identifying decreased risk of positive pelvic lymph nodes in patients with prostate cancer. J Urol 2003; 170: [4] Briganti A, Karakiewicz PI, Chun FK-H, et al. Percentage of positive biopsy cores can improve the ability to predict lymph node invasion in patients undergoing radical prostatectomy and extended pelvic lymph node dissection. Eur Urol 2007;51: [5] Powles T, Murray I, Brock C, Oliver T, Avril N. Molecular positron emission tomography and PET/CT imaging in urological malignancies. Eur Urol 2007;51: [6] Schiavina R, Scattoni V, Castellucci P, et al. 11 C-Choline-positron emission tomography/computerized tomography for preoperative lymph-node staging in intermediate-risk and high-risk prostate cancer: comparison with clinical staging nomograms. Eur Urol 2008;54: [7] Heidenreich A, Aus G, Bolla M, et al. EAU guidelines on prostate cancer. Eur Urol 2008;53: DOI: /j.eururo DOI of original article: /j.eururo