Lymph Node Recovery From Colorectal Resection Specimens Removed for Adenocarcinoma

Transcription

1 ANATOMIC PATHOLOGY Original Article Lymph Node Recovery From Colorectal Resection Specimens Removed for Adenocarcinoma Trends Over Time and a Recommendation for a Minimum Number of Lymph Nodes to be Recovered NEAL S. GOLDSTEIN, MD, 1 WELDON SANFORD, MD,' MARY COFFEY, PHD, 2 AND LESTER J. LAYFIELD, MD 3 Recovery of pericolorectal lymph nodes from colectomy specimens has long been part of colorectal cancer staging. Recently, adjuvant therapy has been added for high stage carcinomas, providing further impetus for performing careful lymph node dissections. Pericolorectal lymph nodes were examined to determine if there has been a change over time in the number of lymph nodes recovered and proportion of specimens with pericolonic lymph node metastases from colorectal carcinoma resection specimens. Also, the authors attempted to establish a recommendation for a minimum number of lymph nodes that should be recovered before a colon can be considered free of metastases. Slides and reports of the first 20 consecutive pi 3 colorectal carcinoma resections in each year from 19 to 199 at William Beaumont Hospital that did not have known metastases at the time of surgery were reviewed (70 specimens total). The mean number of lymph nodes recovered per spec- The prognostic importance of pericolorectal lymph node metastases in colorectal carcinoma has been recognized for at least 90 years. Charles Mayo noted its importance in a 1904 address to the Oregon State Medical Association, and Cuthbart Dukes incorporated perirectal lymph node metastases into his staging systems. 1-7 Recently, the TNM staging system has replaced the original and modified Dukes' staging systems. It also uses the presence of pericolorectal lymph node metastases as a component of overall tumor stage. 8 The presence of one lymph node metastasis de- From ihe Department of''anatomic Pathology. William Beaumont Hospital. Royal Oak. Michigan: 2 The William Beaumont Hospital Research Institute. Royal Oak, Michigan: and } Duke University Medical Center, Durham. North Carolina. Manuscript received December 1, 199; revision accepted February 14, Address reprint requests to Dr. Goldstein: Department of Anatomic Pathology. William Beaumont Hospital, 31 West Thirteen Mile Road. Royal Oak. MI imen and incidence of detected lymph node metastases increased over the 41-year period, with the greatest increase occurring during The greatest proportion of patients with lymph node metastases detected occurred in the 17 to 20 lymph nodes recovered per specimen group. Specimens with more than 20 lymph nodes did not have a higher proportion of lymph node metastases detected compared to specimens with 17 to 20 lymph nodes. Approximately 20% of the specimens with metastases had more than 17 lymph nodes recovered. These results suggest that pathologists should retrieve all the lymph nodes that can be recovered, but at least 17 lymph nodes should be recovered to insure accurate documentation of nodal metastases when present. (Key words: Colorectal; Adenocarcinoma; Metastases; Lymph nodes) Am J Clin Pathol 1996;106: creases overall survival in comparison with absence of metastases in patients with similar tumors. 3-9 " 14 Prior to the previous decade, therapy for patients with curable colorectal carcinomas was limited to surgical resection of the tumor. The main impetus to identify pericolorectal lymph nodes in these specimens was to fulfill the requisites of Dukes-based staging systems in determining prognosis. The use of adjuvant therapy in subsets of patients with colorectal carcinoma has increased the importance of accurate documentation of nodal metastases The Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute showed a decline in the rates of regional disease (metastatic carcinoma to pericolonic lymph nodes) in the late 1980s, presumably attributable to improvements in early cancer detection procedures. 18 However, another cause of the decreased incidence of regional (pericolorectal lymph node) metastases might be that pathologists in recent years are not performing lymph node dissections with the same care as in prior years. 209

2 210 ANATOMIC PATHOLOGY Article The Working Party Report to the World Congresses of Gastroenterology, Sydney, 1990, recommended that a minimum of 12 lymph nodes be recovered before the patient is considered free of metastases. 19 It has been our experience that lymph node metastases may be found when more than 12 uninvolved lymph nodes are recovered. Had the recovery stopped at only 12 lymph nodes, a metastasis may have been missed. We evaluated the number of pericolonic lymph nodes recovered and pericolonic lymph node metastases detected to study if there has been a change over time in the number of lymph nodes recovered and proportion of specimens with pericolonic lymph node metastases from colon and rectal carcinoma resection specimens. We also estimated the minimum number of lymph nodes that should be recovered before a patient can be considered free of metastases by analysis using the proportion of patients with metastases and the number of lymph nodes recovered from each resection specimen. MATERIALS AND METHODS The first 20 consecutive, pathologic stage T3 colon or rectal carcinomas in each year from that did not have known liver or distant metastases at the time of surgery at William Beaumont Hospital were reviewed. Stage T3 colorectal carcinomas are defined as those that extend through the muscularis propria but do not penetrate the serosa or extend into adjacent organs. Only pt3 cases were included because they have a higher rate of metastases than do carcinomas penetrating less deeply. 1 Cases were identified with the help of the computer files of the William Beaumont Hospital Tumor Registry. Review began in January of each year. Cases with known liver or other distant metastases at the time of surgery were excluded, because the knowledge of distant metastasis could have influenced the extent of surgery performed or the pathologist's diligence in dissecting lymph nodes. Patients who received palliative-type partial resection and tumor-bypass procedures were also excluded. All slides from each case were required to be available for review; incomplete cases were excluded. All slides from all the cases were examined by one of two authors for confirmation of tumor stage T3 status (NSG and WS). The number of lymph nodes recovered and number of metastases detected in pericolonic lymph nodes were counted from the slides. Metastases to pericolonic adipose tissue outside of lymph nodes were not counted as lymph node metastases. Lymph node architectural structures were required to be present to be counted as a lymph node metastasis. Pathology reports from each case were reviewed for specimen length, tumor location, and size. The pathology report was not used for determination of the number of lymph nodes recovered nor the number of lymph nodes with metastases, except for the information regarding the number of sections from the same lymph node that were submitted in a single block. In cases with matted and fused lymph nodes, an estimate of how many lymph nodes were incorporated into a mass was made by counting the number of lymph node capsule silhouettes around the periphery of the mass. The linear-by-linear association test was used for statistical analysis (StatXact User Manual, CYTEL Software, Cambridge, MA). The analysis of this study was planned with the advice of a statistician (MC). Formal statistical inference was not thought to be valid for determining the optimum number of lymph nodes necessary to detect at least one lymph node metastasis for a patient with metastatic disease. Lymph Node Recovery RESULTS Seven hundred fifty stage pt3 colon carcinomas were used from thefilesof William Beaumont Hospital during the period During the period , there were fewer than 20 cases per year with slides available for review. After 1961, we used the first 20 cases per year. There was no apparent change in the distribution of tumor location over time. The sigmoid colon was most common site (29.33%), followed by the ascending colon (22.8%), and the descending colon (14.8%). Mean tumor size and mean specimen length did not appreciably change during the 41 years of the study. The mean tumor size was 4.9 cm (range cm). The mean number and median number of lymph nodes detected per specimen over the 41 years were 9.8 and 8.0, respectively (range 0-62). Seventy percent had between 1 and 12 lymph nodes recovered from each specimen; 13% had 13 to 16 lymph nodes, and 14% had at least 17 lymph nodes (Fig. 1). In 3% of specimens, no lymph nodes were recovered. More lymph nodes were recovered in later time intervals than in earlier years of the study (Table 1). The mean number of lymph nodes recovered per specimen was less than 7 during the period The mean increased to nine lymph nodes per specimen during the period 1963 to The mean was 11.6 lymph nodes per specimen during , 13.6 during , and increased to 19.4 during the period. The maximum number of lymph nodes recovered per specimen also increased with time. However, there A.J.C.P.-,ugusl 1996

3 GOLDSTEIN ET AL. 211 Lymph Node Recovery FlG. 1. Distribution of 70 specimens according to number of lymph nodes recovered. % of Cases Number of Lymph Nodes Recovered Per Specimen >20 was at least one specimen in which no lymph nodes or only one was recovered during all time periods. Lymph Node Metastases Fewer than 3% of specimens prior to 1963 had detectable metastases (Table 1). The proportion of specimens with detectable metastases remained almost constant at approximately 0% during the next 10 periods covering the time from 1963 to During the time period , the proportion rose to 70%. The relationship between the number of lymph nodes recovered per specimen and the proportion of specimens with detectable metastases was examined with a test for linear-by-linear association. This test excluded specimens in which no lymph nodes were recovered and used the actual number of lymph nodes recovered rather than grouping the number of lymph nodes into categories. The proportion of specimens with metastases was strongly and positively related to the number of lymph nodes recovered (P <.0001). Recovery of a Minimum Number of Lymph Node Figures 2A and 2B display the percentage of cases with detectable metastases based on the number of lymph nodes recovered from the specimen. Figure 2A divides the specimens into the groupings, 1-,6-11, 12-20, and >20 lymph nodes recovered per specimen. These lymph node groupings are identical to those used by Hermanek and colleagues 20 and later published in the TNM 1993 Supplement. 21 There was a marked increase in the percentage of patients with at least one lymph node metastasis when 12 to 20 lymph nodes were recovered per specimen (73% of patients had metastasis) compared to specimens in which 6 to 11 lymph nodes were recovered (36% of patients had metastasis). There was no major change in the percentage of patients with a lymph node TABLE 1. LYMPH NODES RECOVERED FROM COLORECTAL CANCER SPECIMENS No. of Lymph Nodes Recovered Time Period No. of Specimens Mean Range Lowest Quartile Highest Quartile % of Specimens with Lymph Node Metastases Vol. 106-No. 2

4 212 ANATOMIC PATHOLOGY Original Article 100% 80%? % a. 40% 20% A Number of Lymph Nodes Recovered 100% 80% % CD O 20% 0% B Number of Lymph Nodes Recovered metastasis when greater than 20 lymph nodes were recovered per specimen compared to patients in whom 12 to 20 lymph nodes were recovered per specimen. Figure 2B divides the specimens into narrower lymph node groupings, 1-4,-8,9-12, 13-16, 17-20, and >20. The percentage of patients with a lymph node metastasis increased markedly with the number of lymph nodes recovered between the second group (-8 lymph nodes) and the fifth group (17-20 lymph nodes). Eighty-seven percent of specimens with 17 to 20 lymph nodes recovered had detectable metastases. The percentage of specimens with detectable metastases was 67% with 13 to 16 lymph nodes recovered, and only 0% with 9 to 12 lymph nodes recovered. The 17 to 20 lymph node group and the >20 lymph node group were similar in the percentage of specimens containing detectable metastases. Figure 3 shows the relation between cumulative percentage of all cases with detectable metastases and the number of lymph nodes that were recovered from these specimens. Fifty-five percent of all specimens with metastases had 12 or fewer lymph nodes recovered. Twentyfive percent of all specimens with metastases had more than 16 lymph nodes recovered from the specimen and >20 >20 FIG. 2. A, The percentage of specimens with lymph node metastases increased when at least 12 lymph nodes were recovered per case compared to the situation with 11 or fewer lymph nodes recovered per case. There was no appreciable change between the 12 to 20 and >20 lymph node groups. Hermanek 20 found a similar result in the comparison between the 12 to 20 and >20 lymph node group. Compare these results with Figure 2B, which analyzes the same cases separated into narrower intervals of lymph node recovery. B, The percentage of specimens with lymph node metastases was greater in the 17 to 20 group than in the 9 to 12, and 13 to 16 lymph node groups. 12% of specimens with metastases had greater than 20 lymph nodes recovered. One specimen had 2 lymph node metastases out of the 62 lymph nodes recovered. DISCUSSION The prognostic significance of lymph node metastases from colorectal carcinoma has long been recognized. 722 " 26 It is a well studied topic. 1 ' Nevertheless, we thought it appropriate to reexamine the topic, because several recent factors may have influenced the methods and results of lymph node dissections. We found that the number of lymph nodes recovered from colorectal carcinoma resection specimens increased over the past 41 years at William Beaumont Hospital, especially since We believe that availability of adjuvant therapy for certain patients with pericolonic lymph node metastases and the consequent need for this information motivated the increased recovery of lymph nodes. There was also a marked increase in the proportion of cases with documented metastases: 70% of specimens during had detectable metastases, A.J.C.P.-August 1996

5 GOLDSTEIN ET AL. 213 Lymph Node Recovery FIG. 3. Approximately 4% of specimens with lymph node metastases had more than 12 lymph nodes recovered from pericolonic adipose tissue. Twenty-five percent of specimens had more than 16 lymph nodes recovered. S % T i! I I I ' I I I I I I I I i I i ' I r 2 I 4! 6 I 8 10 I 12 : ' I 22 I 24 I 26 I 28 I 30 I 32 I 34 i 36 I 38 Uo Number of Lymph Nodes Recovered Per Specimen compared to only 3% during the period This frequency of 70% pericolonic lymph node metastases is greater than that found in most other studies. However, it is similar to that found in most studies showing a mean recovery of >20 lymph nodes per specimen. 6 - I0 ' ' Our results show a strong positive association (P <.0001) between the proportion of specimens with metastases detected and the number of lymph nodes recovered. There is a strong tendency for the proportion of specimens with metastases to increase with the number of lymph nodes recovered. This increase may slow or stop after 17 lymph nodes are recovered; however, there were only 93 specimens in our study with more than 17 lymph nodes recovered. Despite the increased frequency of metastases detected, tumor size and specimen length changed slightly over the 41 years of the study. This same relationship between number of lymph nodes examined and frequency of metastatic disease was described by Gilchrist in The most recently published SEER data show a decrease in colorectal regional metastases from 198 to 1990.' 8 It is not known how much of this decrease is due to the recovery of fewer lymph nodes per specimen. These data suggest that there has been no decrease in the number of lymph nodes recovered and no decrease in the frequency of metastatic disease. If these data are representative of other pathology laboratories, it would suggest that changes in the incidence of cases with regional metastases are attributable to factors other than pathologist performance. It should be pointed out that the SEER data were influenced by colorectal cancer surgery performed at William Beaumont Hospital, because William Beaumont Hospital contributes data to the SEER program. The influence was probably negligible. We have attempted to identify the minimum number of lymph nodes examined before a patient can be considered free of metastases. Multiple factors complicate the determination of a universally valid minimum number. One difficulty stems from the published variation (2.% to 48%) in number of pericolonic lymph nodes occurring in different parts of the colon and rectum Another difficulty stems from the variation in number of lymph nodes recovered regardless of the dissection technique USed.' ' 27 - Despite of these difficulties, Jass addressed this question in a recent editorial and concluded that 10 lymph nodes were the minimum that should be recovered from a colorectal resection specimen. 63 He based this conclusion on data provided by Newland and colleagues. 8,64 In these studies, lymph node metastasis-negative patients with colon carcinoma extending beyond the lamina muscularis propria and no serosal penetration had the same outcome as patients with carcinoma confined to the muscularis propria. A median of 10 lymph nodes was recovered from each specimen. He states, if a lymph node metastasis had been missed, it probably would have occurred in the cases with carcinoma extending beyond the muscularis propria. Because all the cases had the same outcome, recovery of additional lymph nodes would not have changed the prognosis. Our objection to this line of reasoning is that lymph node dissections at our institution are not performed only for prognostic purposes, because the identification of a single lymph node metastasis in colon carcinoma usually mandates adjuvant chemotherapy therapy for the patient. Therefore, the goal of lymph node dissections has moved beyond representative sampling for prognosis and toward the identification of any lymph node metastasis. Hernanz and colleagues 34 recommended that a minimum of six lymph nodes be examined from each colon specimen. However, we believe that the statistical method on which their recommendation is based has significant flaws, leading to an underestimate of the number of lymph nodes that should be examined. They used the proportion of specimens with any positive lymph nodes as an estimate of the proportion of individual lymph nodes that have metastases and they assumed a binomial distribution for the possible number of metastatic lymph nodes within a specimen. Examination of our data indicates larger variability for this number than would occur with a binomial distribution. In our data, cases with 6 or fewer lymph nodes recovered account for only 2% of cases with metastases. Vol. 106-No. 2

6 214 ANATOMIC PATHOLOGY Original Article Scott and Grace 30 examined 103 colon carcinomas using manual and clearing techniques for the lymph node dissection. They found that 49% of cases had metastases, which is similar to our overall 4.6% incidence of metastases. They stated that examining a minimum of 13 lymph nodes per case would have detected 90% of their positive cases, and recommended this should be the minimum number of lymph nodes recovered per case. Figure 3 displays our data presented in an identical format as the Scott and Grace study. This figure shows that approximately % of the specimens with a metastasis had 13 or fewer lymph nodes recovered. Almost 80% of specimens with a metastasis had 17 or fewer lymph nodes recovered. One case, that had 2 lymph node metastases, had 62 lymph nodes recovered. We do not have an explanation for the "right shift" of our graph compared to their results. On average, their study identified more lymph nodes per specimen than ours (18. versus 9.8). However, this is not true when one considers traditional dissection techniques, for which their mean number was 6.2. Our study had more than 7 times as many cases as their study. The larger sample size of our study means that our results were affected by a smaller magnitude of chance variation than their results. Hermanek 20 grouped his cases into specimens with 1-, 6-11, 12-20, and >20 lymph nodes recovered. He found no apparent difference in the percentage of patients with lymph node metastases between specimens with 12 to 20 lymph nodes recovered and >20 lymph nodes recovered. Using Hermanek's lymph node grouping, we also found a marked increase in the proportion of specimens with metastases between the second and third group and no apparent difference in the proportion of specimens with metastases between the third and fourth group (Fig. 2A). Using arbitrarily chosen smaller intervals of 4 lymph nodes, we obtained more specific results than using the Hermanek grouping (Fig. 2B). The highest frequency of lymph node metastases was in the 17 to 20 lymph node group. Specimens with >20 lymph nodes had a slightly smaller frequency of metastases. The Working Party Report to the World Congress of Gastroenterology, Sydney, 1990, recommended that a minimum of 12 lymph nodes should be recovered from a colon with carcinoma before it is considered free of metastases, based on the Hermanek and Scott and Grace data. This number was adopted by the American Joint Committee on Cancer and TNM Committee of the International Union Against Cancer as the minimum desirable number of lymph nodes recovered from pericolonic adipose tissue. 8 2I ' 6 With all the factors that can influence pericolonic lymph node recovery and this study design, it is impossible to statistically determine the minimum number of lymph nodes needed to insure optimal documentation of nodal metastases. The probability that the remaining unexamined lymph nodes contain one or more metastases cannot be determined. Unlike other organs such as the prostate, where studies used statistics to define a minimal number of transurethral resection tissue chips that need to be examined to exclude the possibility of there being an adenocarcinoma present, we do not know the metastasis status of every lymph node in the colon. 66 " 68 If more lymph nodes were examined in the 407 specimens of our study that were devoid of metastases, additional specimens may have been identified as having metastases. We have presented our data in a manner that respects these limitations, and purposely used only descriptive rather than inferential statistics for most of the comparisons. We believe that our data supports the conclusion that there is no minimum number of lymph nodes that should be retrieved from a colon or rectum. Our response to the question of how many lymph nodes should be recovered from a colon or rectum specimen containing carcinoma is: all of them. However, this is not a practical answer to the legitimate question of how many lymph nodes a pathologist should strive to retrieve once all the obvious, and sometimes few, lymph nodes have easily been removed. We believe that pathologists should strive to retrieve at least 17 lymph nodes from colorectal resection specimens. This number is based on Figures 2B and 3. We provide this number with the provision that it reflects a large variation in the number of lymph nodes that were retrieved, as there were a substantial number of specimens with more than 17 or < lymph nodes retrieved. It is also our opinion that the recommendation made by the American Joint Committee on Cancer and TNM Committee of the International Union Against Cancer on the minimal number of lymph nodes to be retrieved is too low. We believe that the recommended minimum number of lymph nodes that should be retrieved from a colon or rectum should be changed from 12 to 17. We did not measure lymph node size in this study. Others have shown that small lymph node size does not correlate with the presence or absence of metastases. '-3.ll Metastases to small ] ymph nodes comprise between 9% and 78% of all lymph node metastases. 33 ' 0 In summary, we have shown that the mean number of lymph nodes recovered and the proportion of cases with pericolonic lymph node metastases from colorectal can- A.J.C.P.-August 1996

7 GOLDSTEIN ET AL. 21 Lymph Node Recovery cer resection specimens obtained from William Beaumont Hospital has increased, especially during the period Pathologists should strive to recover as many lymph nodes as possible from colorectal resection specimens. In general, this number should be at least 17 lymph nodes from each specimen. This can usually be achieved by careful, thorough manual dissection. Acknowledgments The authors thank Drs. J. Neill, J. Watts, D. Bostwick, and R. Goldstein for their suggestions and manuscript reviews. REFERENCES 1. Dukes CE. The spread of cancer of the rectum. Br J Surg 1929; 17: Dukes CE. The classification of cancer of the rectum. J Pathol Bacteriol 1932;3: Astler VB, Coller FA. The prognostic significance of direct extension of carcinoma of the colon and rectum. Ann Surg 194; 139: Dukes CE. Cancer of the rectum: An analysis of cases. J Pathol Bactehot 1940; 0: Shephard JM, Jones JSP. Adenocarcinoma of the large bowel. Br J Cancer 1971;2: Gabriel WB, Dukes CE, Bussey HJR. Lymphatic spread in cancer of the rectum. Br J Surg 193;23: MayoCH. Cancer of the large bowel. Med Sentinel 1904; 12: Colon and Rectum. In: American Joint Committee on Cancer Manual for Staging of Cancer, ed 4. Philadelphia: J.B. Lippincott, 1992, pp Dukes CE, Bussey HJR. The spread of rectal cancer and its effect on prognosis. Br J Cancer 198; 12: Gilchrist RK, David VC. A consideration of pathological factors influencing five year survival in radical resection of the large bowel and rectum for carcinoma. Ann Surg 1947; 126: Grinnell RS. The grading and prognosis of carcinoma of the colon and rectum. Ann Surg 1939; 109: Grinnell RS. Results in the treatment of carcinoma of the colon and rectum. Surg Gynecol Obstet 193;96: Spratt J, Spjut HJ. Prevalence and prognosis of individual clinical and pathological variables associated with colorectal carcinoma. Cancer 1967;20: Steele G Jr, Murphy GP, Lawrence W Jr, Lenhard RE Jr, eds. Colorectal cancer. In: American Society Textbook of Clinical Oncology, ed 2. The American Cancer Society, 199, pp Steele G Jr. Augenlicht LH, Begg CB, et al. National Institutes of Health Consensus Development Conference: Adjuvant therapy for patients wit colon and rectal cancer. JAMA 1990:264: van Triest B, van Groeningen CJ, Pinedo HM. Current chemotherapeutic possibilities in the treatment of colorectal cancer. Eur J Cancer 199:31 A: Haller DG. An overview of adjuvant therapy for colorectal cancer. Eur J Cancer 199:31 A: Chu KC, Tarone RE, Chow WH, Hankey BF, Gloeckler Ries LA. Temporal patterns in colorectal cancer incidence, survival, and mortality from 190 through J Natl Cancer Inst 1994; 86: Fielding LP, Arsenault PA, Chapuis PH, et al. Working party report to the world congress of gastroenterology, Sydney J Gastroenterol Hepatol : Hermanek P. Onkologische Chirurgie/Pathologisch-anatomische Sicht. LangenhecksArch ChirSuppI 1991: Hermanek P. Henson DE, Hutter RVP, Sobin LH, eds. TNM Supplement 1993: A Commentary on Uniform Use. Berlin: Springer-Verlag, 1993, pp Smith JW. The operative treatment of carcinoma of the rectum. BMJ 1911;1: Hayes JM. The involvement of the lymph glands in carcinoma of the large intestine. Minn Med : McVay JR. Involvement of the lymph nodes in carcinoma of the rectum. Ann Surg 1922;76: Miller RT. Cancer of the colon. Ann Surg 1923;78: Miles WE. The spread of cancer of the rectum. Lancet 192:208: Gilchrist RK, David VC. Lymphatic spread of carcinoma of the rectum. Ann Surg 1938:108: Blenkinsopp WP, Stewart-Brown S, Blesovsky L. Kearney G. Fielding LP. Histopathology reporting in large bowel cancer../ Clin Pathol 1981:34: Cawthorn SJ, Gibbs NM, Marks CG. Clearance technique for the detection of lymph nodes in colorectal cancer. Br J Sure; 1986;73: Scott KWM. Grace RH. Detection of lymph node metastases in colorectal carcinoma before and after fat clearance. Br J Surg 1989:76: Davidson BR, Sams VR, Styles J, Deane C, Boulos PB. Detection of occult nodal metastases in patients with colorectal carcinoma. Cancer 1990;6: Hyder JW, Talbott TM, Maycroft TC. A critical review of chemical lymph node clearance and staging of colon and rectal cancer at Ferguson Hospital, Dis Colon Rectum 1990:33: Herrera L, Villarreal JR. Incidence of metastases from rectal adenocarcinoma in small lymph nodes detected by a clearing technique. Dis Colon Rectum 1992;3: Hernanz F, Revuelta S, Redondo C, et al. Colorectal adenocarcinoma: Quality of the assessment of lymph node metastases. Dis Colon Rectum 1994:37: Wilking N, Herrera L. Petrelli NJ, Mittelman A. Pelvic and perineal recurrencesafterabdominoperineal resection for adenocarcinoma of the rectum. Am J Surg 198:10: Dukes CE. The surgical pathology of rectal cancer. Proc Roval Soc Med 1944;37: Herrera-Ornelas L, Justiniano J, Castillo NB. et al. Metastases in small lymph nodes from colon cancer. Arch Surg 1987:122: Wood WQ, Wlike DPD. Carcinoma of the rectum: An anatomicopathological study. Edin Med J 1933;40: PickrenJW. Nodal clearance and detection. JAMA 197:231: JassJR, Northover JMA. Fat clearance method versus manual dissection of lymph nodes in specimens of rectal cancer. Int J Colored Dis 1986;1: Grinnell RS. Lymphatic metastases of carcinoma of the colon and rectum. Ann Surg 190; 131: Glover RP, Waugh JM. The retrograde lymphatic spread of carcinoma of the "rectosigmoid region". Surg Gvnecol Obstet 1946;82: Greenson JK, Isenhart CE, Rice R., et al. Identification of occult micrometastases in pericolonic lymph nodes of Dukes' B colorectal cancer patients using monoclonal antibodies against cytokeratin and CC49. Cancer 1994;73: Nicholson AG, Marks CG, Cook MG. Effect on lymph node status of triple leveling and immunohistochemistry with CAM.2 on node negative colorectal carcinomas. Gut 1994;3: Morikawa E, Yasutomi M. Shindou K, et al. Distribution of metastatic lymph nodes in colorectal cancer by the modified clearing method. Dis Colon Rectum 1994;37: Haboubi NY. Schofield PF, Clark P, Kaftan SM. The importance of combining xylene clearance and immunohistochemistry in the accurate staging of colorectal carcinoma. J R Soc Med 1992;8: Jeffers MD, O'Dowd GM, Mulcahy H, et al. The prognostic significance of immunohistochemically detected lymph node mi- Vol. 106-No. 2

8 216 ANATOMIC PATHOLOGY Original Article crometastases in colorectal carcinoma. J Pathol 1994; 172: Sheffield JP, Talbot IC. ACP Broadsheet 132; Gross examination of the large intestine. J Clin Pathol 1992;4: Morodomi T, Isomoto H, Shirouzu K, et al. An index for estimating the probability of lymph node metastasis in rectal cancers: Lymph node metastasis and the histopathology of actively invasive regions of cancer. Cancer 1989;63: Hida J, Mori N, Kubo R, et al. Metastases from carcinoma of the colon and rectum detected in small lymph nodes by the clearing method. J Am Coll Surg 1994; 178: Coller FA, Kay EB, Maclntyre RS. Regional lymphatic metastasis of carcinoma of the rectum. Surgery 1940;8: Hojo K, Koyama Y, Moriya Y. Lymphatic spread and its prognostic value in patients with rectal cancer. Am J Surg 1982; 144: Grinnell RS. The lymphatic and venous spread of carcinoma of the rectum. Ann Surg 1942; 116: Grinnell RS. The spread of carcinoma of the colon and rectum. Cancer 190;3: Peltier LF. The search for lymph node metastases in cancer of the rectum. Surgery 191; 30: Hatta M. Effect of lymphatic vessel invasion and stromal response on lymph node metastasis of colorectal cancer: By modified clearing method. MedJKinki Univ 1987; 12: Bognel C, Rekacewicz C. Mankarios H, et al. Prognostic value of neural invasion in rectal carcinoma: A multivariant analysis on 339 patients with curative resection. Eur J Cancer 199; 31 A: Newland RC, Dent OF, Lyttle MNB, Chapuis PH, Bokey EL. Pathologic determinants of survival associated with colorectal cancer with lymph node metastases. Cancer 1994; 73: Hermanek P. Colorectal carcinoma: histopathological diagnosis and staging. Bailliere's Clin Gastroenterol 1989;3: Craig WM, MacCarty WC. Involvement of the lymph glands in cancer of the caecum. Ann Surg 1923;77: Coller FA, Kay EB, Maclntyre RS. Regional lymphatic metastases of carcinoma of the colon. Ann Surg 1941; 144: Izumoto G. Studies on the lymph node metastases of the colorectal cancer by the modified clearing method. Med J Kinki Univ 1984;9: Jass JR. Prognostic factors in rectal cancer. Eur J Cancer l99;31a: Newland RC, Chapuis PH, Smyth EJ. The prognostic value of substaging colorectal carcinoma: A prospective study of 1,117 cases with standardized pathology. Cancer 1987;: Beahrs OH, Henson DE, Hutter RVP, Kennedy BJ, eds. Handbook for Staging of Cancer, ed 4. Philadelphia: J.B. Lippincott, 1993, pp Murphy WM, Dean PJ, Brasfield JA, Tatum L. Incidental carcinoma of the prostate: How much sampling is adequate? Am J Surg Pathol 1986; 10: Rohr LR. Incidental adenocarcinoma in transurethral resection specimens of the prostate. Partial versus complete microscopic examination. Am J Surg Pathol 1987; 11: Vollmer RT. Prostate cancer and chip specimens: Complete versus partial sampling. Hum Pathol 1986; 17: A.J.C.R-August 1996