Retroperitoneal and Lateral Pelvic Lymphadenectomy Mapped by Lymphoscintigraphy for Rectal Adenocarcinoma Staging

Transcription

1 Retroperitoneal and Lateral Pelvic Lymphadenectomy Mapped by Lymphoscintigraphy for Rectal Adenocarcinoma Staging Claudio Almeida Quadros 1,*, Ademar Lopes 2 and Iguaracyra Araújo 3 1 Colorectal Division, Aristides Maltez Cancer Hospital, Salvador, Bahia, 2 Pelvic Surgery Department, A.C. Camargo Cancer Hospital, São Paulo and 3 Pathology Division, Aristides Maltez Cancer Hospital, Salvador, Bahia, Brazil *For reprints and all correspondence: Claudio Almeida Quadros, Hospital Aristides Maltez, Serviço de Proctologia, Av. D. João VI, 332, Brotas, , Salvador, Bahia, Brazil. caquadros@gmail.com Received February 1, 2010; accepted March 28, 2010 Jpn J Clin Oncol 2010;40(8) doi: /jjco/hyq060 Advance Access Publication 10 May 2010 Background: The good prognosis of retroperitoneal and lateral pelvic lymphadenectomy has raised the question of whether total mesorectal excision is suitable for adequate staging of rectal adenocarcinoma patients. The aims of this study were to determine the accuracy of dye and probe detection of metastatic retroperitoneal and/or lateral pelvic nodes and to define the upstaging impact of retroperitoneal and lateral pelvic lymphadenectomy in rectal adenocarcinoma patients. Methods: Ninety-seven rectal adenocarcinoma patients were submitted to total mesorectal excision and retroperitoneal and lateral pelvic lymphadenectomy. Lymphoscintigraphy using technetium-99 m-phytate and patent blue was performed to detect blue and/or radioactive retroperitoneal and/or lateral pelvic nodes which were examined histopathologically and immunohistochemically with a step-sectioning technique. Results: Mesorectal mean node count was 11.5 and retroperitoneal and/or lateral pelvic node was Retroperitoneal and lateral pelvic lymphadenectomy identified metastases in 17.5%, upstaging 8.2%. Variables related to metastatic retroperitoneal and/or lateral pelvic nodes were the following: Stage III in total mesorectal excision specimens (P, 0.04), pt3/ pt4 tumors (P ¼ 0.047), high levels of carcinoembryonic antigen (P ¼ 0.014) and large tumors (P ¼ 0.03). Marker migration to retroperitoneal and/or lateral pelvic nodes occurred in 37.1%, upstaging 11.1%. The markers accuracy in the detection of metastatic retroperitoneal and/or lateral pelvic nodes was 100%. Conclusions: Retroperitoneal and lateral pelvic lymphadenectomy detected an important rate of metastatic retroperitoneal and/or lateral pelvic nodes (RLPN), resulting in upstaging. When markers migrated, they were able to detect RLPN metastases. The use of markers should be improved in the identification of RLPN metastases for selective indication of retroperitoneal and lateral pelvic lymphadenectomy. Key words: rectal cancer retroperitoneal and lateral pelvic lymphadenectomy lateral node dissection lymph node excision rectal adenocarcinoma staging INTRODUCTION There is controversy concerning the best surgical approach for the treatment of extraperitoneal rectal adenocarcinoma. Most Western surgeons defend total mesorectal excision (TME) (1), combined with chemoradiotherapy, in patients with T3/T4 tumors or mesorectal lymph node metastasis (2). Japanese surgeons defend TME combined with a radical extended retroperitoneal and lateral pelvic lymphadenectomy (RLPL), with a more restricted indication of post-operative radiotherapy (3). RLPL is the standard surgical procedure in Japan indicated in low rectal cancer patients with T3/T4 tumors, which show a greater risk of metastasizing to the retroperitoneal and/or lateral pelvic nodes (RLPN) (3). RLPL, as a complement to the surgical treatment of rectal cancer, was initially performed in the West and was replaced # The Author (2010). Published by Oxford University Press. All rights reserved.

2 Jpn J Clin Oncol 2010;40(8) 747 by adjuvant chemoradiotherapy (4). The reasons that made Western surgeons abandon RLPL were related to the high morbidity of the procedure, the concept that metastases to RLPN may represent systemic disease not amenable to surgical cure, and the low incidence of RLPN metastases (4). Japanese surgeons refined RLPL with pelvic autonomic nerve preservation, decreasing operating time, blood loss and genitourinary dysfunction (4). In Japan, metastases to RLPN are considered regional node involvement, not systemic disease, and RLPL is considered a curative procedure (3). Currently, the results of the treatment options for rectal adenocarcinoma in the West or in Japan have improved to the point that they are similar. The best results obtained in the West in the treatment of patients with T3/T4 rectal adenocarcinomas with preoperative chemoradiotherapy followed by TME reveal a 5-year pelvic recurrence of 6% and survival of 76% (5). Japanese results for the treatment of patients with T3/T4 rectal adenocarcinomas using TME plus RLPL indicate a 8.3% pelvic recurrence and a 77.0% 5-year survival (4,6). Even with similar treatment results obtained in the West and Japan, metastasis to RLPN is a pattern of pelvic dissemination that has not been addressed by surgeons in the West, who consider these patients as TNM Stage IV (7). Pelvic radiotherapy, delivered as an adjuvant treatment in T3/T4 or N1 patients (2), is the only treatment option for patients who may have increased risk of RLPN metastases, with no assurance of efficacy. Rectal cancer patients with RLPN metastases benefit from RLPL, as seen by a 5-year survival of 39.8% and pelvic recurrence rate of 26.1% (4). Despite increased recurrence, these rectal cancer patients show a 5-year survival rate similar to that in treated rectal cancer patients with TNM Stage III in the West, i.e. 41.5% (8), and they have a better 5-year survival compared with Stage IV patients, i.e. 30% (9). This suggests that RLPN metastases should not be considered systemic tumor dissemination and that RLPL should be offered to these patients (4). The Japanese protocol of submitting T3/T4 low rectal cancer patients to RLPL shows a 20.1% incidence of RLPN metastases (3). Assuming that only patients with RLPN metastases benefit from RLPL, a great number of Japanese patients are submitted to extensive lymphadenectomy without necessity. Surgeons should study the ways of selectively indicating RLPL for patients with RLPN metastases. This could be done with the use of lymphoscintigraphic markers that identify metastatic RLPN in a direct drainage pathway of the tumor s lymphatic basin. The aims of this study were to evaluate the accuracy of lymphoscintigraphy with technetium-99 m-phytate and patent blue in the detection of metastases in RLPN and the eventual upstaging impact of performing RLPL. Low rectal adenocarcinoma patients without distant metastasis were submitted to the standard TME procedure, and their RLPN were analyzed for the detection of blue or technetium-positive nodes. RLPL was then performed, and the nodes were assessed by histopathological and immunohistochemical examinations. PATIENTS AND METHODS Local and national ethics committees approved this trial, and written informed consent was obtained from all patients who were aware that they would be submitted to RLPL. The completion of the study protocol was supported by the finding of no additional morbidity/mortality related to RLPL in the first 30 published cases (10), with definitely no prohibitive genitourinary dysfunction related to RLPL. Between January 2004 and August 2008, 97 patients with histologically confirmed primary invasive rectal adenocarcinoma were prospectively enrolled in the Division of Colorectal Tumors of Aristides Maltez Hospital. The patients included in the study were treated with curative intent, where they were submitted to TME, lymphatic mapping with technetium-99 m-phytate and patent blue and RLPL with pelvic autonomic nerve preservation (Fig. 1), by the same surgeon (first author). All surgical specimens were obtained by a single pathologist (third author). Preoperative chemoradiotherapy was indicated in patients with fixed tumors on digital rectal assessment, and also in patients with T4 tumors or positive pelvic lymph nodes suggested by computed tomography. Preoperative treatment consisted of a 5040 cgy dose of radiotherapy delivered in 180 cgy per day fractions, combined with a chemotherapy regimen of daily 5-fluorouracil (375 mg/m 2 of body surface) and leucovorin (30 mg/m 2 of body surface), delivered in a 5-day cycle, one in the first and other in the third week of radiotherapy. Patients with pt3/pt4 tumors or nodal metastases who did not receive preoperative treatment underwent post-operative chemoradiotherapy following the same protocol. Four additional cycles of chemotherapy were given, one per month, after surgery in patients submitted to preoperative chemoradiotherapy or after post-operative chemoradiotherapy, using the above-mentioned protocol of 5-fluorouracil and leucovorin. Figure 1. Pelvis appearance after total mesorectal excision and retroperitoneal and lateral pelvic lymphadenectomy with pelvic autonomic nerve preservation. The clamp identifies the left branches of the hypogastric and pelvic nerves. A color version of this figure is available as supplementary data at

3 748 RLPL mapped by lymphoscintigraphy Details concerning the definition of RLPN and the RLPL procedure were described in a previous publication (10). All patients had low rectal tumors located at or below the peritoneal reflection (11). The extraperitoneal rectum was divided into three anatomic regions (12). The upper rectum was considered the proximal 1/3 starting at the peritoneal reflection (12). The lower rectum was considered the lower 1/3 including the anal canal (12). The middle rectum was taken as the intermediate 1/3 rectal segment between the upper and lower rectum (12). After anesthetic induction, a solution containing 22.2 MBq (0.6 mci) of technetium-99 m-phytate, diluted in 2.0 ml saline, was injected around the tumor using a tuberculin syringe, followed by injection of 1.0 ml of a 2.5% solution of patent blue dye. In patients with upper rectal tumors, the radiotracer and dye were injected subserosally, after laparotomy, in a circumferential manner around the tumor. In lower and middle rectal tumors, subserosal injection was precluded by the mesorectum and deep pelvic rectal tumor location, and therefore, the radiotracer and dye were injected into the submucosal and muscular layer through a transanal approach with a rectoscope. After TME, lymphatic mapping was performed using a handheld gamma probe (Europrobe w, CaTe-detector) for identification of radioactive RLPN. RLPN were considered radioactive if they had counts at least three times greater than the background counts, away from the site of the technetium-99 m-phytate injection. Blue nodes were identified by visual assessment. All identified radioactive and/or blue nodes were sent separately for pathological inspection. All the lymph nodes identified in the surgical specimens were submitted to routine pathological analysis which consisted of the examination of one lymph node section stained by hematoxylin eosin. If these first section slides of the radioactive and/or blue RPLN were negative for metastases, enhanced pathologic analysis was applied, including stepsectioning and immunohistochemistry. Step-sectioning consisted of cutting formalin-fixed and paraffin-embedded tissues at 4 mm, creating three distinct levels with two sections at each level. One section of each level was stained with hematoxylin eosin, and the other was used for immunohistochemistry. The first level consisted of the two slides of the lymph node sectioned in its longitudinal axis, submitted to routine pathological examination. The next 40 mm of the node were discarded, and a second level was obtained; 40 mm were again discarded, and a third level of sections was obtained. A micrometastasis was defined as a tumor focus within a single node that measured,0.2 cm in its greatest dimension or nodal tumor that was only detectable by immunohistochemistry (12). Upstaging was only considered if enhanced pathological examination detected RPLN metastases or micrometastases. Metastases to RLPN were considered as regional nodal dissemination following the Japanese Society for Cancer of the Colon and Rectum classification, and patients were considered Stage III (11). Immunohistochemical examination was performed using the streptavidin biotin peroxidase method (LSAB-dako, Carpinteria, CA, USA). A primary monoclonal antibody against cytokeratin (clone AE1/AE3) was used at a dilution of 1:100. Before the application of the monoclonal antibody, the sections were pretreated in citrate buffer at ph 6 in a 968C steam bath for 35 min. In four patients submitted to preoperative chemoradiotherapy, it was not possible to determine the histological grade because of tissue alterations related to radiotherapy. These patients were excluded from the statistical analysis related to the variable histological grade. Despite that preoperative tumor biopsy revealed invasive rectal adenocarcinoma, the surgical specimen of one patient after preoperative chemoradiotherapy showed in situ tumor. This patient was classified as Stage 0 (ptisn0m0) and was excluded from the statistical analysis related to the variable TME stage. The accuracy of lymphatic mapping in the detection of metastases to RLPN in rectal cancer patients was evaluated in those in whom the method was effective. The method was considered effective when at least one radioactive and/or blue RLPN was detected. The method s effectiveness could be evaluated by the identification rate which was the number of lymphatic mapping procedures where at least one radioactive and/or blue RLPN was identified, divided by the total number of procedures in which the method was attempted. The following statistical measures were used. True positives (TP) were defined as the cases where radioactive and/or blue RLPN had metastatic cells, whether or not metastatic cells were found in other RLPN. True negatives (TN) were the cases where no metastatic cells were found in the radioactive and/or blue RLPN or in all other RLPN. False negatives (FN) were defined as cases where the radioactive and/or blue RLPN did not contain metastatic cells but there were metastatic cells in other RPLN. In this study, there were no false positives (FP) since the radioactive and/or blue nodes were also RLPN. The equations of the variables are given as follows: sensitivity, TP/(TP þ FN); false negatives, FN/ (FN þ TP); negative predictor value, TN/(FN þ TN); and accuracy, (TP þ TN)/(TP þ FP þ FN þ TN). Comparison of the categorical variables was performed with the chi-squared test (x 2 ) and Fisher s exact test. Continuous variables were compared using the Mann Whitney test and Student s t-test. In all statistical tests, the a error was set at 5%. Logistic regression was used for multivariate statistical analysis using the backward Wald method. RESULTS Females were predominant, comprising 61% of the group studied. Lower rectal tumors represented 48% of patients, middle rectal 37% and upper rectal 14%. Preoperative chemoradiotherapy was performed in 47.4% of the group. Mean age was 55.1 years (SD ¼ 15.7), with a median of

4 Jpn J Clin Oncol 2010;40(8) years. Mean tumor size was 4.4 cm, with a median of 4.0 cm (ranging from 0.8 to 15.8 cm). Mean mesorectal node count was 11.5 (1119/97) and mean RLPN count was 11.7 (1136/97), totaling 2255 lymph nodes examined. When staging was performed with only the surgical specimen obtained with the TME procedure, 1 patient (1.0%) was classified as Stage 0 (ptisn0m0), 6 Stage I (6.2%), 61 Stage II (62.9%) and 29 Stage III (29.9%). The inclusion of RLPN upstaged eight patients (8.2%) from Stage II to III, as in these eight patients there were metastases to RLPN but no metastases to mesorectal nodes. RLPL identified 41 metastatic nodes in 17 patients (17.5%), with an average of 2.4 metastatic RLPN per patient. The metastatic nodes were in the following anatomic Table 1. Incidence of metastases to retroperitoneal and/or lateral pelvic nodes (RLPN) in rectal adenocarcinoma patients according to clinical, pathological and treatment-related category variables Variable Category Metastatic RLPN P value No [n (%)] Yes [n (%)] Gender Male 33 (86.8) 5 (13.2) 0.42* Female 47 (79.7) 12 (20.3) Rectal tumor site Lower 40 (85.1) 7 (14.9) 0.67** Mid 28 (77.8) 8 (22.2) Upper 12 (85.7) 2 (14.3) Stage in the TME I and II 59 (88.1) 8 (11.9),0.04* specimen III 20 (69.0) 9 (31.0) Histological grade 1 28 (87.5) 4 (12.5) 0.25** 2 47 (79.7) 12 (20.3) 3 1 (50.0) 1 (50.0) Lymphovascular No 64 (86.5) 10 (13.5) 0.11** invasion Yes 16 (69.6) 7 (30.4) Tumor necrosis No 68 (84.0) 13 (16.0) 0.47** Yes 12 (75.0) 4 (25.0) Ulceration No 45 (81.8) 10 (18.2) 1.0* Yes 35 (83.3) 7 (16.7) Perineural invasion No 77 (84.6) 14 (15.4) 0.065** Yes 3 (50.0) 3 (50.0) Surgery APR 42 (85.7) 7 (14.3) 0.52** AR 35 (77.8) 10 (22.2) Pelv. exent. 3 (100.0) pt Classification T1 and T2 30 (93.8) 2 (6.3) 0.047* (TNM) T3 and T4 49 (76.6) 15 (23.4) Preoperative No 42 (82.4) 9 (17.6) 1.0* Chemoradiotherapy Yes 38 (82.6) 8 (17.4) APR, abdominoperineal rectal resection; AR, anterior rectal resection; Pelv. exent., total pelvic exenteration; TNM, tumor, nodes, metastasis; TME, total mesorectal excision. *Chi-squared test (x 2 ). **Fisher s exact test. locations: 24 along the iliac arteries, 8 along the obturator vessels and nerve, 5 around the aortic bifurcation and 4 around the aorta and inferior vena cava (para-aortic nodes). Of these 17 patients with metastases to RLPN, 8 did not have metastases to mesorectal nodes and 9 did. Four patients with metastases to mesorectal nodes had already been submitted to preoperative chemoradiotherapy. Thus, 12 patients, 12.4% of the group studied, would not have had additional pelvic radiotherapy as an adjuvant treatment to metastatic RLPN as they had been previously submitted to radiotherapy or would not have had the treatment indicated as they did not have metastatic mesorectal nodes. Univariate statistical analysis demonstrated that the variables related to metastatic RLPN were the following: Stage III in TME specimen (P, 0.04); pt3/pt4 tumors (P ¼ 0.047); high levels of carcinoembryonic antigen (average of 30.6 ng/ml, median of 9.9 ng/ml) (P ¼ 0.014); and large tumors (mean size: 5.5 cm cm) (P ¼ 0.03), as shown in Tables 1 and 2. Multivariate analysis did not identify any independent variable related to metastatic RLPN. The variable showing the P value closest to the significant level, albeit non-significant, was pt3/pt4 tumors with P ¼ Radioactive and/or blue RLPN were found in 36 patients, yielding an identification rate of 37.1%. There were 68 radioactive and/or blue RLPN, with an average of 1.8 nodes detected per patient (ranging 1 6). Technetium identified 68 nodes and blue dye 22. All RLPN identified with blue dye were also radioactive. The accuracy of radioactive and/or blue nodes as a predictor of metastatic RLPN involvement was 100%, with a sensitivity of 100%, negative predictor value of 100% and zero FN (Table 3). The method permitted an upstaging of 11.1%, where micrometastases were responsible for an upstaging of 5.5% (Table 3). Univariate statistical analysis did not identify variables related to the inability of technetium and/or blue dye to migrate to RLPN (Tables 4 and 5). Table 2. Incidence of metastases to RLPN in rectal adenocarcinoma patients according to median and mean continuous variables Variable RLPN metastases Median Mean (SD) P value Age (years) Yes (20.5) 0.74* No (15.3) Time between preop. Yes (56.1) 0.12* chemoradiotherapy and surgery (days) No (78.5) Preop. CEA (ng/ml) Yes (53.0) 0.014* No (33.0) Tumor size (cm) Yes (3.2) 0.03** No (1.8) Preop., preoperative; CEA, carcinoembryogenic antigen; SD, standard deviation. *Mann Whitney s test. **Students t-test.

5 750 RLPL mapped by lymphoscintigraphy Table 3. Identification rate and accuracy of lymphatic mapping using technetium-99 m-phytate and patent blue in rectal adenocarcinoma patients for detection of metastases in radioactive and/or blue RLPN Lower rectum Mid-rectum Upper rectum Total Patients [n (%)] 47 (48) 36 (37) 14 (14) 97 (100) Identification rate (%) 38.3 (18 of 47) 33.3 (12 of 36) 42.9 (6 of 14) 37.1 (36 of 97) Accuracy (%) 100 (18 of 18) 100 (12 of 12) 100 (6 of 6) 100 (36 of 36) Sensibility (%) 100 (3 of 3) 100 (3 of 3) 100 (1 of 1) 100 (7 of 7) Negative predictive value (%) 100 (15 of 15) 100 (9 of 9) 100 (5 of 5) 100 (29 of 29) False negatives Upstaging rate (%) 11.1 (2 of 18) 8.3 (1 of 12) 16.7 (1 of 6) 11.1 (4 of 36) Micrometastases* upstaging rate (%) 5.5 (1 of 18) 8.3 (1 of 12) (2 of 36) *Upstaging exclusively by micrometastases, defined as metastases,0.2 cm or identified with immunohistochemistry. DISCUSSION Satisfactory results with regard to the surgical outcomes of these patients submitted to RLPL with pelvic autonomic nerve preservation permitted the completion of this study protocol. A later paper will report the details related to morbidity/mortality, quality of life and survival of the patients enrolled in this study. Considering only metastases 0.2 cm, RLPL provided an unquestionable upstaging rate of 6.2%. These patients would have been mistakenly classified as Stage II and would not have had the benefit of post-operative chemoradiotherapy. When micrometastases were included, the upstaging rate increased to 8.2%. The inclusion of micrometastases in the upstaging rate achieved with RLPL may be considered important, as there is evidence that RLPN micrometastases have prognostic value and are related to recurrence (13). Shimoyama et al. (13) performed RLPL in 66 rectal adenocarcinoma patients, and identified 9 patients (13.6%) with metastatic RLPN using conventional hematoxylin eosin pathological examination. In the 57 patients without metastases to RLPN, two more histopathological sections were examined, one with hematoxylin eosin and the other with immunohistochemistry using antibodies against cytokeratins and CAM 5.7 (13). This enhanced pathological examination detected micrometastases in 11 patients (19.3%) (13). The 10-year recurrence rate in patients without metastases to RLPN (23.4%) was significantly lower than that obtained in patients with micrometastases to RLPN (45.5%; P ¼ 0.048); the recurrence rate of patients with metastases to RLPN was 58.3% (13). The 10-year survival rate of patients without metastases to RLPN was 80.4%, whereas it was 54.4% in patients with micrometastases to RLPN (P ¼ 0.01) and 38.9% in patients with metastases to RLPN (13). The present study demonstrated that 12.4% of the patients would not have had their metastatic RLPN submitted to any regional treatment, as 8.2% would have been erroneously considered Stage II with TME and 4.2% had already been treated with preoperative chemoradiotherapy. In fact, preoperative chemoradiotherapy was not related to a lower rate of RLPN metastases. Metastatic nodes would have been left behind in 17.5% of the patients if RLPL had not been performed, with a mean of 2.4 not resected metastatic nodes per patient. Metastatic RLPN would have persisted after a surgery considered curative, and not performing the indicated adjuvant chemoradiotherapy would have certainly led to pelvic or systemic recurrence in these patients. In reviewing studies that identified variables present in rectal adenocarcinoma patients who had metastatic RLPN, the only association identified in this study and not found in other publications was high levels of carcinoembryonic antigen. The association of metastatic mesorectal nodes with metastatic RLPN was also revealed by Ueno et al. (14), Ueno et al. (15) and Kusters et al. (6). The Japanese Society for Cancer of the Colon and Rectum (3) and Sugihara et al. (16) also associated pt3/pt4 tumors with metastatic RLPN. In matters of tumor size, Sugihara et al. (16) also reported an association of large rectal tumors with RPLN metastases, considering that patients with tumors greater than 4.0 cm have an increased incidence of metastases to RLPN. These studies are important in defining characteristics that could interfere in the treatment plans of patients with an increased risk of having metastases to RLPN. Comparing this study s final results with the preliminary results published previously (10), they were similar in terms of incidence of metastatic RLPN (17.5% and 20.0%, respectively) and upstaging with RLPL (8.2% and 6.5%). This study s final data show that chemoradiotherapy does not interfere with marker migration to RLPN and with the detection of RLPN metastasis. As this study included in its statistical accuracy measures only patients in whom lymphatic mapping was effective, accuracy results differed from those obtained in the preliminary results published previously (10). This change in accuracy measurement was learned from colorectal sentinel lymph node studies that consider lymphatic mapping

6 Jpn J Clin Oncol 2010;40(8) 751 Table 4. Incidence of RLPN identified or not with technetium-99 m-phytate and/or patent blue in rectal adenocarcinoma patients according to clinical, pathological and treatment-related category variables Variable Category Radioactive and/or blue RLPN P value No [n (%)] Yes [n (%)] Gender Male 25 (65.8) 13 (34.2) 0.63* Female 36 (61.0) 23 (39.0) Rectal tumor site Lower 29 (61.7) 18 (38.3) 0.80* Mid 24 (66.7) 12 (33.3) Upper 8 (57.1) 6 (42.9) Stage in the TME specimen I 5 (83.3) 1 (16.7) 0.69** II 38 (62.3) 23 (37.7) III 18 (62.1) 11 (37.9) Histological grade 1 19 (59.4) 13 (40.6) 0.41** 2 39 (66.1) 20 (33.9) 3 1 (50) 1 (50) Lymphovascular invasion No 45 (60.8) 29 (39.2) 0.47* Yes 16 (69.6) 7 (30.4) Tumor necrosis No 52 (64.2) 29 (35.8) 0.58* Yes 9 (56.3) 7 (43.8) Ulceration No 32 (58.2) 23 (41.8) 0.30* Yes 29 (69.0) 13 (31.0) Perineural invasion No 56 (61.5) 35 (38.5) 0.41** Yes 5 (83.3) 1 (16.7) Surgery RAP 30 (61.2) 19 (38.8) 0.48** RAR 30 (66.7) 15 (33.3) Pelv. exent. 1 (33.3) 2 (66.7) pt classification (TNM) T1 and T2 20 (62.5) 12 (37.5) 1.0* T3 and T4 41 (64.1) 23 (35.9) Preoperative No 28 (54.9) 23 (45.1) 0.097* Chemoradiotherapy Yes 33 (71.7) 13 (37.1) *Chi-squared test (x 2 ). **Fisher s exact test. effective only when the markers migrate, identifying at least one sentinel lymph node (12). The lack of marker migration affects lymphoscintigraphy s effectiveness which is evaluated by the method s identification rate and not its accuracy (12). The accuracy of a lymphatic mapping procedure can only be evaluated when it is effective (12). If the preliminary results had been evaluated using these parameters, the identification rate for RLPN radioactive and/or blue nodes would not have changed (23.3%) (10),andaccuracyresultswould have been the same as that obtained in this study s final results, with an accuracy of 100% and zero FN. With regard to studies written in English, Funahashi et al. (17) published the only paper on lymphatic mapping in rectal adenocarcinoma patients, aimed at the detection of mesorectal, retroperitoneal and lateral pelvic lymphatic basins in those at higher risk of metastases. Technetium-99 m-colloid, 111 MBq (3 mci), was injected by rectoscopy in 43 patients, with a migration rate to retroperitoneal and pelvic lymphatic basins of 39.5%, similar to that obtained in this study (17). As only the lymphatic basins were evaluated, not single lymph nodes, a comparison of accuracy measures cannot be made (17). The accuracy of technetium-99 m-phytate and patent blue in detecting metastatic RLPN in rectal adenocarcinoma patients, demonstrated in this study, is encouraging. It demonstrates that enhanced histopathological examination of retroperitoneal and pelvic radioactive and/or blue nodes might have the ability to indicate a more extensive lymphadenectomy. Although, in order to determine the precise accuracy of lymphoscintigraphy in the detection of

7 752 RLPL mapped by lymphoscintigraphy Table 5. Incidence of RLPN identified or not with technetium-99 m-phytate and/or patent blue in rectal adenocarcinoma patients according to median and mean continuous variables Variable metastases to RLPN, all retrieved RLPN should be subjected to the same level of pathological assessment. In this study, because of budget limitations, ultrastaging was only performed on blue and/or radioactive RLPN. In this context, the FN rate may not be accurate. In fact, the same level of pathological assessment should be performed on all mesorectal nodes in order to determine the true level of upstaging following additional RLPL. Further lymphoscintigraphy studies aimed at selective indication of RLPL should address this topic, where all patients lymph nodes would be evaluated using the same histopathological and immunohistochemical step-sectioning technique. The low migration rate of the markers used in RLPN lymphoscintigraphy suggests the need for improvement of the mapping technique and marker concentrations and volumes, as well as the selection of other markers that may migrate better to RLPN. By improving the marker migration rate to RLPN, lymphoscintigraphy may become a tool for selective indication of RLPL in low rectal adenocarcinoma patients. Blue and/or radioactive RLPN could be evaluated using a pathological frozen section procedure, and if the high accuracy parameters are maintained, the method could be useful for intra-operative definition of RLPL. CONCLUSIONS Radioactive and/or blue RLPN Median Mean (SD) P value* Age (years) Yes (17.0) 0.74 No (15.9) Time between preop. Yes (30.8) 0.24 chemoradiotherapy and surgery (days) No (85.95) Preop. CEA (ng/ml) Yes (43.0) 0.90 No (34.1) Tumor size (cm) Yes (1.8) 0.15 No (2.4) *Mann Whitney s test. RLPL detected an important rate of metastatic RLPN, resulting in upstaging even in patients submitted to preoperative chemoradiotherapy. This study s univariate statistical analysis demonstrated that the variables related to metastatic RLPN were metastases to mesorectal nodes, pt3/pt4 tumors, high levels of carcinoembryonic antigen and large tumors. When the lymphoscintigraphic markers technetium- 99 m-phytate and patent blue migrated to RLPN, they were able to detect RPLN metastases with an excellent accuracy. The low migration rate of the markers used in RLPN lymphoscintigraphy suggests that the method should be improved for selective indication of RLPL. Acknowledgements It was the PhD thesis of Claudio Almeida Quadros at the School of Medicine of the University of São Paulo, Brazil. Dr Fernanda Fahel provided assistance in Nuclear Medicine, Dr Kleber Pimentel performed the statistical analysis, and Dr A. Leyva performed the English editing. Funding This study was supported by the Liga Bahiana Contra o Câncer (LBCC). Conflict of interest statement None declared. References 1. Heald RJ, Husband EM, Ryall RD. The mesorectum in rectal cancer surgery the clue to pelvic recurrence? Br J Surg 1982;69: National Institutes of Health Consensus Conference. Adjuvant therapy for patients with colon and rectal cancer. JAMA 1990;264: Japanese Society for Cancer of the Colon and Rectum. General Rules for Clinical and Pathological Studies on Cancer of the Colon, Rectum and Anus, 7th edn. Tokyo: Kanehara & Co Yano H, Moran BJ. The incidence of lateral pelvic side-wall nodal involvement in low rectal cancer may be similar in Japan and the West. Br J Surg 2008;95: Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. NEnglJMed 2004;351: Kusters M, van de Velde CJ, Beets-Tan RG, Akasu T, Fujida S, Yamamoto S, et al. Patterns of local recurrence in low rectal cancer: a single-center experience. Ann Surgical Oncol 2009;16: Greene FL, Page DL, Fleming ID, Fritz A, Balch CM, Haller DG, et al., editors. UICC/AJCC Cancer Staging Manual, 6th edn. Berlin, Heidelberg, New York: Springer Jessup JM, Stewart AK, Menck HR. The National Cancer Data Base Report on patterns of care for adenocarcinoma of the rectum, Cancer 1998;83: Ries LAG, Kosary CL, Hankey BF, Miller BA, Edwards BK. SEER Cancer Statistics Review Bethesda: National Cancer Institute Quadros CA, Lopes A, Araújo I, Fahel F, Bacellar MS, Dias CS. Retroperitoneal and lateral pelvic lymphadenectomy mapped by lymphoscintigraphy and blue dye for rectal adenocarcinoma staging: preliminary results. Ann Surg Oncol 2006;13: Japanese Society for Cancer of the Colon and Rectum. Japanese Classification of Colorectal Carcinoma, 2nd English edn. Tokyo: Kanehara & Co Quadros CA, Lopes A, Araujo I, Fregnani JH, Fahel F. Upstaging benefits and accuracy of sentinel lymph node mapping in colorectal adenocarcinoma nodal staging. J Surg Oncol 2008;98: Shimoyama M, Yamazaki T, Suda T, Hatakeyama K. Prognostic significance of lateral lymph node micrometastases in lower rectal cancer: an immunohistochemical study with CAM5.2. Dis Colon Rectum 2003;46:333 9.

8 Jpn J Clin Oncol 2010;40(8) Ueno H, Yamauchi C, Hase K, Ichikura T, Mochizuki H. Clinicopathological study of intrapelvic cancer spread to the iliac area in lower rectal adenocarcinoma by serial sectioning. Br J Surg 1999;86: Ueno M, Oya M, Azekura K, Yamaguchi T, Muto T. Incidence and prognostic significance of lateral lymph node metastasis in patients with advanced low rectal cancer. Br J Surg 2005;92: Sugihara K, Kabayashi H, Kato T, et al. Indication and benefit of pelvic sidewall dissection for rectal cancer. Dis Colon Rectum 2006;49: Funahashi K, Koike J, Shimada M, Okamoto K, Goto T, Teramoto T. A preliminary study of the draining lymph node basin in advanced lower rectal cancer using a radioactive tracer. Dis Colon Rectum 2006;49(Suppl 10):S53 8.