|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2007.12.7704 on March 24 2008 © 2008 American Society of Clinical Oncology. Prognostic Implications of the Distribution of Lymph Node Metastases in Rectal Cancer After Neoadjuvant Chemoradiotherapy
From the Departments of Surgery, Pathology, Radiation Oncology, Nuclear Medicine, Radiology, and Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, NY Corresponding author: José G. Guillem, MD, MPH, Colorectal Service, Memorial Sloan-Kettering Cancer Center, 1275 York Ave, Room C-1077, New York, NY 10021; e-mail: guillemj{at}mskcc.org
Purpose After preoperative chemoradiotherapy of rectal cancer, the number of retrievable and metastatic lymph nodes is decreased. The current TNM classification is based on number and not location of lymph node metastases and may understage disease after chemoradiotherapy. The aim of this study was to examine the prognostic significance of location of involved lymph nodes in rectal cancer patients after preoperative chemoradiotherapy. Patients and Methods We prospectively examined whole-mount specimens from 121 patients with uT3-4 and/or N+ rectal cancer who received preoperative chemoradiotherapy followed by resection. Location of involved lymph nodes was compared with median number of lymph nodes involved as well as presence of distant metastasis at presentation. Results Lymph node metastases were detected in 37 patients (31%). Thirteen patients with lymph node involvement along major supplying vessels (proximal lymph node metastases) had a significantly higher rate of distant metastatic disease at time of surgery than patients without proximal lymph node involvement (P < .001); median number of lymph nodes involved was two for patients with proximal lymph node metastases and 1.5 for patients with mesorectal lymph node involvement alone. Conclusion Our data suggest that, after preoperative chemoradiotherapy, proximal lymph node involvement is associated with a high incidence of metastatic disease at time of surgery. Because the median number of involved lymph nodes is low after preoperative chemoradiotherapy, the TNM staging system may not provide an accurate assessment of metastatic disease. Therefore, the ypTNM staging system should incorporate distribution as well as number of lymph node metastases after preoperative chemoradiotherapy for rectal cancer.
In 1931, Ernest Miles identified the upward zone of lymphatic spread of rectal cancer along the superior hemorrhoidal vessels and the inferior mesenteric vein as being the most constant and, therefore, most important route of rectal cancer spread.1 Subsequently, Grinnell2 and Dukes and Bussey3 reported that both the number of lymph node metastases and their location affect prognosis in patients with rectal cancer. In the early 1980s, lower rates of 5-year survival were reported for patients with lymph node involvement along the trunk of the supplying vessels than for patients with peripheral/pararectal lymph node metastases alone.4,5 Despite these observations, there is still significant controversy as to whether number or distribution of lymph node metastases provides the basis for a more accurate prognosis. Although different staging classifications for colorectal carcinoma exist, the TNM staging system remains the standard.6,7 Since the emergence of preoperative chemoradiotherapy as the optimal treatment approach for cT3-4 and/or N+M0 rectal cancer,8 staging classifications based on number of involved lymph nodes may be inadequate. In fact, several studies have reported a decrease in the number of lymph nodes retrieved, as well as fewer lymph node metastases, after preoperative chemoradiotherapy.9-11 Observation of the distribution rather than the number of lymph node metastases may, therefore, provide a more accurate prognosis after preoperative chemoradiotherapy. However, the distribution of lymph node metastases in rectal cancer after preoperative chemoradiotherapy has not been prospectively examined. The aims of this study were to quantify and map lymph nodes in resected specimens of rectal cancer patients after preoperative chemoradiotherapy and to correlate the distribution of lymph node metastases with the presence of metastatic disease at time of surgery.
Patient Population We prospectively examined the operative specimens of 121 patients with uT3-4 and/or N+ rectal adenocarcinoma who received preoperative chemoradiotherapy at our institution from September 1999 through December 2005. Patients with distant metastatic disease (n = 17) were included if they received preoperative chemoradiotherapy according to the study protocol and then underwent resection of their primary cancer. This study was approved by the Institutional Review Board at Memorial Sloan-Kettering Cancer Center (MSKCC). Informed consent was obtained from all participants. Table 1 lists the clinicopathologic characteristics of the patient population.
Clinical Assessment Before commencement of preoperative chemoradiotherapy, the operating surgeon comprehensively assessed each rectal cancer in terms of its size, distance from the anal verge, location (anterior, posterior, lateral, or circumferential), mobility (fixed or tethered), morphology (exophytic, ulcerated, or sessile), and involved bowel circumference (evaluated as a percentage of the entire bowel circumference), as we have described in a previous study.12
Preoperative Chemoradiotherapy All patients received preoperative fluorouracil (FU)–based chemotherapy (bolus infusion, n = 44, 36%; continuous infusion, n = 66, 55%; unknown, n = 11, 9%). The most common protocol for bolus infusion chemotherapy was FU (325 mg/m2/d) with leucovorin (20 mg/m2/d) administered for two cycles of 5 consecutive days on week 1 (days 1 through 5) and week 5 (days 29 through 33) of radiation therapy. Leucovorin was administered by intravenous (IV) infusion for 30 minutes and was followed 20 minutes later with the IV bolus FU. The most common protocol for continuous infusion chemotherapy was FU (225 mg/m2/d) for a 6-week continuous cycle. Five patients in the continuous infusion chemotherapy group received cetuximab in addition to FU (400 mg/m2 IV delivered over a period of 120 minutes on day 1, followed by weekly doses of 250 mg/m2 IV delivered over a period of 60 minutes; after completion of chemoradiotherapy, cetuximab was administered alone weekly for an additional 4 weeks before surgery). Median time from completion of chemoradiotherapy to surgery was 46 days (range, 19 to 100 days).
Imaging FDG-PET was performed with the GE Advance (General Electric Medical Systems, Milwaukee, WI) whole-body PET scanner, no earlier than 45 minutes after injection of 10 to 15 mCi of [18F]FDG. CT of the abdomen and pelvis was performed either at MSKCC using single-slice helical (HiLite Advantage; GE Healthcare, Milwaukee, WI) or multislice helical scanners (Lightspeed, Lightspeed Ultra, and Lightspeed 16; GE Healthcare) or at an outside institution using a variety of scanners. MSKCC scans were performed with the patient in a prone position, with rectal air insufflation after glucagon 1.0 mg for bowel relaxation. IV and oral contrast was administered to all patients. Endorectal ultrasound (ERUS) was performed before the start of chemoradiotherapy in 113 patients (93.4%) using a Brüel & Kjaer 2102 Hawk ultrasound machine (Brüel & Kjaer, Naerum, Denmark) equipped with a rotating endosonic probe and a 10-mHz transducer. ERUS stage was determined according to the techniques previously reported by Hildebrandt et al14 and Beynon et al.15 Circular or oval structures measuring more than 3 mm were considered to be malignant lymph nodes. Lymph nodes measuring less than 3 mm with central hyperechogenicity were considered benign. Twenty-three patients (20.4%) were staged as uN0, 79 patients (70.0%) were staged as uN1, and eight patients (7.1%) were staged as uN2; in three patients (2.7%), lymph node status could not be assessed (uNX).
Surgical Techniques
Pathologic Assessment
Statistical Analysis The difference between rates of distant metastatic disease relative to the presence and distribution of regional lymph node metastases was evaluated using Fisher's exact test (Table 4). The association of tumor and disease characteristics with the presence of proximal lymph node involvement and the presence of distant metastatic disease was examined using Fisher's exact test for categoric variables and an exact version of the Mantel-Haenszel test for trending ordinal variables. Pathologic tumor stage was collapsed into the following three categories: T0 and Tis; T1 and T2; and T3 and T4 (Table 5).
A total of 2,156 lymph nodes were identified in 121 patients (average, 17.8 nodes per patient). Lymph nodes with metastatic involvement were detected in 37 patients (31%). In patients with lymph node metastases, the median number of involved lymph nodes was two (range, one to 16 nodes). Although the majority of lymph nodes (58%) were found along the major supplying vessels (superior rectal and inferior mesenteric arteries) and in the pericolonic region, the majority of lymph node metastases (77%) were found in the mesorectum (Table 2).
Of the 37 patients with lymph node metastases, 13 (35%) had proximal lymph node involvement, with a median number of two involved nodes (range, one to 16 nodes). In seven patients, metastases were identified only in proximal lymph nodes (Table 3).
The percentage of patients with American Joint Committee on Cancer stage IV disease was 46% for patients with proximal and mesorectal lymph node metastases and 25% for patients with mesorectal lymph node metastases only. There was a significant difference in the number of patients with metastatic disease who showed proximal lymph node involvement versus those who showed no proximal lymph node involvement (P < .001; Table 4). However, pairwise comparison demonstrated no significant difference in the number of patients with metastatic disease showing proximal lymph node involvement versus those showing only mesorectal lymph node involvement (P = .27). The location of lymph node metastases in relation to the radiation field was examined. Lymph nodes with metastatic involvement in the apical, mid, and pericolonic regions were scored as being outside the radiation field because of their location proximal to the L5/S1 junction (Fig 1). Overall, 18% of lymph node metastases (19 of 108 metastases) were located above the superior radiation field border (Table 2). This proximal location of lymph node metastases was noted in 10 (77%) of 13 patients with proximal lymph node involvement and in 8% (10 of 121 patients) of the entire study population. We compared the accuracy of ERUS, CT, and PET in detecting proximal lymph node involvement. ERUS and PET failed to detect proximal lymph node metastases in 13 patients. CT accurately identified proximal lymph node involvement in two (15%) of 13 patients. In general, upper (> 6 cm from the anal verge), larger (> 2 cm), and circumferential (67% to 100%) rectal cancers were associated with greater likelihood of proximal lymph node involvement; however, this relationship was not statistically significant (Table 5).
In our study, lymph node metastases were noted in 37 (31%) of 121 patients with uT3-4 rectal cancer after preoperative chemoradiotherapy. Approximately one third of these patients had proximal lymph node involvement (13 of 37 patients). Because patients with proximal lymph node involvement had a significantly higher rate of distant metastatic disease at time of surgery than patients without proximal lymph node involvement (46% v 9%, respectively; P = .003), a finding of proximal lymph node metastasis is likely associated with a poor prognosis. Given that we could not demonstrate a significant difference between the rates of distant metastatic disease in patients with mesorectal lymph node involvement alone compared with patients with proximal lymph node involvement (25% v 46%, respectively; P = .27), this finding most likely reflects the limited number of patients in our study who presented with metastatic disease. Studies examining the prognostic significance of proximal lymph node involvement in patients not receiving preoperative radiation or chemotherapy demonstrate a lower 5-year survival rate for both rectal5 and colon cancer patients.4,17 Although these studies also underscore the prognostic significance of proximal lymph node involvement, Hermanek and Altendorf4 further reported that colorectal cancer patients with proximal lymph node involvement usually have at least four or more metastatic nodes. The prognostic significance of greater than three involved lymph nodes is incorporated into the most recent edition of the TNM classification system,7 in which patients with one to three positive lymph nodes are staged as IIIA or IIIB and patients with more than three positive lymph nodes are staged as IIIC. Because our data demonstrated a median number of two involved lymph nodes, we question the utility of the ypTNM classification system for rectal cancer patients after preoperative chemoradiotherapy. This is of clinical importance because it has been recommended that patients with stage IIIC disease receive intensified postoperative chemotherapy.18 Our results suggest that rectal cancer patients with proximal lymph node involvement after preoperative chemoradiotherapy should receive intensified postoperative chemotherapy regardless of the number of lymph nodes involved. Because proximal lymph node involvement seems to be an important prognostic factor, our results emphasize the need for thorough pathologic assessment and classification of lymph nodes by location in rectal cancer patients after preoperative chemoradiotherapy. The high incidence of metastatic disease identified at time of surgery in patients with proximal lymph node involvement after preoperative chemoradiotherapy should be considered when choosing neoadjuvant therapy. Such patients, who otherwise do not require preoperative rectal cancer shrinkage to enhance the possibility of sphincter preservation or negative circumferential resection margins, may be best treated with neoadjuvant chemotherapy first, followed by restaging. We anticipate that, with improved pretherapeutic imaging, patients with proximal lymph node involvement who otherwise do not require preoperative rectal cancer shrinkage to enhance the possibility of sphincter preservation or negative circumferential resection margins may be best treated with neoadjuvant chemotherapy alone. We emphasize the benefits of accurately identifying proximal lymph node involvement before beginning therapy because, in this subset, 77% of patients (10 of 13 patients) are likely to have lymph node metastases located above the standard pelvic radiation fields. Given the significance of proximal lymph node involvement in rectal cancer, accurate pretreatment assessment is essential. In our study, although upper (> 6 cm from the anal verge), larger (> 2 cm), and circumferential (67% to 100%) rectal cancers were associated with a greater likelihood of proximal lymph node involvement, this relationship was not statistically significant. Thus, it seems that clinical assessment cannot identify patients at higher risk of proximal lymph node involvement. Similarly, none of the imaging modalities used during initial staging, including FDG-PET scan, CT scan, and ERUS, were able to accurately identify proximal lymph node involvement. Although CT scan detected 15% of patients with proximal lymph node involvement, PET and ERUS detected none. The sensitivity of PET and CT for detecting lymph node involvement in our study was lower than that reported by others (ranging from 21% to 29%19,20 and 25% to 55%20,21 in other studies). This is not surprising because other reports have included both mesorectal and proximal lymph nodes, whereas we focused on proximal lymph nodes alone. Although ERUS seems to have a higher sensitivity in staging mesorectal lymph nodes (66% to 88%)22 than either PET or CT, most proximal lymph node metastases cannot be visualized on ERUS. There are some limitations to this study, including the relatively small number (n = 37) of patients with lymph node metastases, which limited our ability to compare overall and recurrence-free 5-year survival rates for patients with or without proximal lymph node involvement. Additionally, considerable improvements in imaging methods have been made during the 5-year period of our data collection. For example, dedicated PET was used exclusively for this study, and comparison was made to dedicated CT. Dedicated PET has now been virtually completely replaced by PET/CT, and it is likely that the dedicated PET/CT would have performed better than dedicated PET or CT alone. Also, recent studies have demonstrated that magnetic resonance imaging (MRI) has sensitivity, specificity, and accuracy in preoperative staging of rectal cancer similar to that of ERUS, and MRI easily images proximal lymph nodes. Thus, MRI may be able to identify patients with proximal lymph node involvement preoperatively.23-25 In conclusion, our data, based on a prospective, comprehensive pathologic analysis of resected specimens, suggest that proximal lymph node involvement is associated with a high incidence of metastatic disease at time of surgery after neoadjuvant chemoradiotherapy. Because the median number of involved lymph nodes is only two, the current TNM staging classification (which distinguishes between one and three and > three lymph node metastases) may not be useful in patients after preoperative chemoradiotherapy. Rectal cancer staging after preoperative chemoradiotherapy should incorporate the location of lymph node metastases. Therefore, comprehensive pathologic analysis of the resected specimen is essential in identifying patients with proximal lymph node metastases; patients with proximal lymph node involvement should be staged as IIIC. Our results also suggest that patients with proximal lymph node involvement should be treated with neoadjuvant chemotherapy, followed by restaging. However, neither clinical assessment nor imaging tools, including CT, PET, and ERUS, have the ability to accurately identify proximal lymph node involvement. Further studies are needed to evaluate the impact of proximal lymph node involvement on survival, to improve the accuracy of imaging modalities, and to establish treatment options that will provide the best outcome for this group of patients.
The author(s) indicated no potential conflicts of interest.
Conception and design: Jinru Shia, Leyo Ruo, José G. Guillem Administrative support: W. Douglas Wong, José G. Guillem Provision of study materials or patients: Timothy Akhurst, Marc J. Gollub, Michelle S. Ginsberg, Steven Larson, José G. Guillem Collection and assembly of data: Tobias Leibold, Jinru Shia, Leyo Ruo, Elyn Riedel, José G. Guillem Data analysis and interpretation: Tobias Leibold, Jinru Shia, Bruce D. Minsky, Timothy Akhurst, Marc J. Gollub, Michelle S. Ginsberg, Steven Larson, Elyn Riedel, José G. Guillem Manuscript writing: Tobias Leibold, Bruce D. Minsky, José G. Guillem Final approval of manuscript: Tobias Leibold, Jinru Shia, Leyo Ruo, Bruce D. Minsky, Timothy Akhurst, Marc J. Gollub, Michelle S. Ginsberg, Steven Larson, Elyn Riedel, W. Douglas Wong, José G. Guillem
published online ahead of print at www.jco.org on March 24, 2008. Supported in part by National Cancer Institute Grant No. R01 82534-01 (J.G.G). Authors’ disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Miles WE: The pathology of the spread of cancer of the rectum, and its bearing upon the surgery of the cancerous rectum. Surg Gynecol Obstet 52:350-359, 1931 2. Grinnell RS: The spread of carcinoma of the colon and rectum. Cancer 3:641-652, 1950[Medline] 3. Dukes CE, Bussey HJ: The spread of rectal cancer and its effect on prognosis. Br J Cancer 12:309-320, 1958[Medline] 4. Hermanek P, Altendorf A: Classification of colorectal carcinomas with regional lymphatic metastases. Pathol Res Pract 173:1-11, 1981[Medline] 5. Hojo K, Koyama Y, Moriya Y: Lymphatic spread and its prognostic value in patients with rectal cancer. Am J Surg 144:350-354, 1982[CrossRef][Medline] 6. Japanese Society for Cancer of the Colon and Rectum: Japanese Classification of Colorectal Carcinoma (ed 5). Tokyo, Japan, Kanehara, 1997 7. Greene FL, Page DL, Fleming ID, et al (eds): AJCC Cancer Staging Manual (ed 6). New York, NY, Springer-Verlag, 2002 8. Sauer R, Becker H, Hohenberger W, et al: Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med 351:1731-1740, 2004 9. Wichmann MW, Muller C, Meyer G, et al: Effect of preoperative chemoradiotherapy on lymph node retrieval after resection of rectal cancer. Arch Surg 137:206-210, 2002 10. Sermier A, Gervaz P, Egger JF, et al: Lymph node retrieval in abdominoperineal surgical specimen is radiation time-dependent. World J Surg Oncol 4:29, 2006[CrossRef][Medline] 11. Baxter NN, Morris AM, Rothenberger DA, et al: Impact of preoperative radiation for rectal cancer on subsequent lymph node evaluation: A population-based analysis. Int J Radiat Oncol Biol Phys 61:426-431, 2005[CrossRef][Medline] 12. Guillem JG, Chessin DB, Shia J, et al: Clinical examination following preoperative chemoradiation for rectal cancer is not a reliable surrogate end point. J Clin Oncol 23:3475-3479, 2005 13. Grann A, Feng C, Wong D, et al: Preoperative combined modality therapy for clinically resectable uT3 rectal adenocarcinoma. Int J Radiat Oncol Biol Phys 49:987-995, 2001[CrossRef][Medline] 14. Hildebrandt U, Feifel G, Schwarz HP, et al: Endorectal ultrasound: Instrumentation and clinical aspects. Int J Colorectal Dis 1:203-207, 1986[CrossRef][Medline] 15. Beynon J, Roe AM, Foy DM, et al: Preoperative staging of local invasion in rectal cancer using endoluminal ultrasound. J R Soc Med 80:23-24, 1987[Abstract] 16. Guillem JG, Chessin DB, Shia J, et al: A prospective pathologic analysis using whole-mount sections of rectal cancer following preoperative combined modality therapy: Implications for sphincter preservation. Ann Surg 245:88-93, 2007[CrossRef][Medline] 17. Shida H, Ban K, Matsumoto M, et al: Prognostic significance of location of lymph node metastases in colorectal cancer. Dis Colon Rectum 35:1046-1050, 1992[CrossRef][Medline] 18. Fietkau R, Barten M, Klautke G, et al: Postoperative chemotherapy may not be necessary for patients with ypN0-category after neoadjuvant chemoradiotherapy of rectal cancer. Dis Colon Rectum 49:1284-1292, 2006[CrossRef][Medline] 19. Kantorova I, Lipska L, Belohlavek O, et al: Routine (18)F-FDG PET preoperative staging of colorectal cancer: Comparison with conventional staging and its impact on treatment decision making. J Nucl Med 44:1784-1788, 2003 20. Llamas-Elvira JM, Rodriguez-Fernandez A, Gutierrez-Sainz J, et al: Fluorine-18 fluorodeoxyglucose PET in the preoperative staging of colorectal cancer. Eur J Nucl Med Mol Imaging 34:859-867, 2007[CrossRef][Medline] 21. Bipat S, Glas AS, Slors FJ, et al: Rectal cancer: Local staging and assessment of lymph node involvement with endoluminal US, CT, and MR imaging—A meta-analysis. Radiology 232:773-783, 2004 22. Schaffzin DM, Wong WD: Endorectal ultrasound in the preoperative evaluation of rectal cancer. Clin Colorectal Cancer 4:124-132, 2004[Medline] 23. Peschaud F, Cuenod CA, Benoist S, et al: Accuracy of magnetic resonance imaging in rectal cancer depends on location of the tumor. Dis Colon Rectum 48:1603-1609, 2005[CrossRef][Medline] 24. Chun HK, Choi D, Kim MJ, et al: Preoperative staging of rectal cancer: Comparison of 3-T high-field MRI and endorectal sonography. AJR Am J Roentgenol 187:1557-1562, 2006 25. Kim CK, Kim SH, Chun HK, et al: Preoperative staging of rectal cancer: Accuracy of 3-Tesla magnetic resonance imaging. Eur Radiol 16:972-980, 2006[CrossRef][Medline] Submitted May 23, 2007; accepted January 4, 2008.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2008 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
