Abstract
Objective
To identify factors which influence the total and negative lymph node counts in colorectal cancer resection specimens, independent of pathologists and surgeons.
Design
We utilized multivariate negative binomial regression. Covariates included age, sex, body mass index, family history of colorectal carcinoma, year of diagnosis, hospital setting, tumor location, resected colorectal length (specimen length), tumor size, circumferential growth, TNM stage, lymphocytic reactions and other pathologic features, and tumor molecular features (microsatellite instability, CpG island methylator phenotype, LINE-1 methylation, and BRAF, KRAS and PIK3CA mutations).
Setting
U.S. nationwide prospective cohort studies.
Patients
Rectal and colon cancer patients (N=918).
Main Outcome Measures
The negative and total node counts (continuous).
Results
Specimen length, tumor size, ascending colon location, T3N0M0 stage and year of diagnosis were positively associated with the negative node count (all p<0.002). KRAS mutation might also be positively associated with the negative node count (p=0.032; borderline significance considering multiple hypothesis testing). Among node-negative (stage I-II) cases, specimen length, tumor size, ascending colon, and T3N0M0 stage remained significantly associated with the node count (all p<0.005), and PIK3CA and KRAS mutations might also be positively associated with the node count (p=0.027 and p=0.049, respectively, with borderline significance).
Conclusions
This molecular pathological epidemiology study shows that specimen length, tumor size, tumor location, TNM stage and year of diagnosis are operator-independent predictors of the node count. These crucial variables should be examined in any future evaluation of adequacy of lymph node harvest and nodal staging towards personalized medicine for colorectal cancer patients.
Keywords: colon cancer, quality, assessment, surgery, pathology, practice
Introduction
The presence of lymph node metastasis has important implications in prognosis and treatment for colorectal cancer patients.1 Observational studies indicate that the number of lymph nodes assessed by pathologic examination (in particular, the negative node count) is associated with longer survival of colorectal cancer patients.2 Thus, along with disease stage and tumor molecular features, the node count is often used for treatment decision making by oncologists. However, the optimal number of lymph nodes that must be assessed remains controversial.2-5 Although the average number of lymph nodes evaluated for colorectal cancer has increased in the past decade, it remains uncertain how the node count is influenced by demographic, clinical and tumor molecular factors.6,7
The number of recovered lymph nodes may be influenced by not only surgeons and pathologists, but also factors independent of surgeons and pathologists. Those “operator-independent” factors include tumor location, disease stage, tumor size, host immune response,8 and tumor molecular features such as microsatellite instability (MSI) and the CpG island methylator phenotype (CIMP).9 Molecular features of colorectal cancer, and host immune response have been associated with the node count.9,10 Previous studies have examined the relationship between the recovered node count and various demographic and clinical features in population-based studies11-16 and in hospital-based studies,17-27 but all of those studies11-26,28,29 lacked comprehensive data on specimen length, tumor size, host immune reaction to tumor, and tumor molecular features. Beyond surgeon and pathologist (i.e., operator) factors, it is important to identify patient-specific node count predictors (i.e., clinical, pathologic or tumor molecular factors), in order to assess adequacy of lymph node examination for each patient. To accomplish this aim, a comprehensive database of a large number of colorectal cancer cases with clinical, specimen, pathologic and molecular annotations is needed.
We therefore conducted this molecular pathological epidemiology (MPE8,30,31) study, utilizing a database of 918 colorectal cancer cases in two prospective cohort studies. Considering overall distribution of the node count, we utilized negative binomial regression analysis to identify factors associated with the negative and total node counts. Since we utilized the U.S. nationwide cohort database with clinical, specimen, pathologic, and tumor molecular variables (including MSI, CIMP, and KRAS, BRAF and PIK3CA mutations), we could assess each node count predictor, independent of operator (surgeon and pathologist) factors.
Materials and Methods
Study population
We utilized databases of two prospective cohort studies; the Nurses’ Health Study (N = 121,701 women followed since 1976) and the Health Professionals Follow-up Study (N = 51,529 men followed since 1986).32,33 Every 2 years, participants have been sent follow-up questionnaires to update information on potential risk factors and to identify newly diagnosed cancer in themselves and their first degree relatives. For non-responders, we searched the National Death Index to discover deaths and ascertain causes of death and diagnosis of cancer. Study physicians reviewed medical records including pathology reports, and recorded tumor location and pathological TNM (tumor-node-metastasis) stage, the positive and negative node counts,9 tumor size, circumferential growth along the bowel wall, and resected colorectal length (specimen length). We collected paraffin-embedded tissue blocks from hospitals where patients underwent tumor resections.33 We collected diagnostic biopsy specimens for rectal cancer patients who received preoperative treatment, in order to avoid artifacts or bias introduced by treatment. Based on availability of data on the node count and tumor molecular features, we included a total of 918 colorectal cancer cases diagnosed up to 2006. Hospitals where our participants underwent colorectal resections were distributed throughout the U.S. (Figure 1). Informed consent was obtained from all study subjects. This study was approved by the Human Subjects Committees at Brigham and Women's Hospital and the Harvard School of Public Health.
Figure 1.
Number of cases analyzed in this study by State. Our patients were distributed throughout the U.S., and our results would not have been influenced by any particular surgeon or pathologist, thus increasing generalizability of our findings.
Histopathologic evaluations
Tissue sections from all colorectal cancer cases were evaluated by a pathologist (S.O.). Tumor differentiation was categorized as poor vs. well-moderate (≤50% vs. >50% glandular areas). The presence and extent of mucinous and/or signet ring cell component were recorded. Lymphocytic reaction patterns, i.e., peritumoral lymphocytic reaction, and tumor infiltrating lymphocytes (TIL), were examined as previously described.34 A subset of cases (N > 100) were reviewed by another pathologist (T.M.) and concordance was as follows: κ = 0.72 for tumor differentiation; Spearman r = 0.87 for mucin (%); Spearman r = 0.65 for signet ring cells (%); Spearman r = 0.65 for the summation score of peritumoral reaction and TIL.
Pyrosequencing of KRAS, BRAF and PIK3CA, and microsatellite instability (MSI) analysis
DNA was extracted from tumor and PCR and Pyrosequencing targeted for KRAS (codons 12 and 13),35BRAF (codon 600),36 and PIK3CA (exons 9 and 20) were performed.37 MSI status was determined using D2S123, D5S346, D17S250, BAT25, BAT26, BAT40, D18S55, D18S56, D18S67 and D18S487.38,39 MSI-high was defined as the presence of instability in ≥30% of the markers, microsatellite stability (MSS)/MSI-low as 0-29% unstable markers.
Analyses of CpG island methylation and LINE-1 methylation
Sodium bisulfite treatment on DNA and real-time PCR (MethyLight) assays were validated and performed.40 We quantified promoter methylation in 8 CIMP-specific markers (CACNA1G, CDKN2A, CRABP1, IGF2, MLH1, NEUROG1, RUNX3 and SOCS1).38,41,42 CIMP-high was defined as ≥6/8 methylated markers, CIMP-low/0 as 0 to 5 methylated markers, according to the previously established criteria.38 To accurately quantify relatively high LINE-1 methylation levels, we utilized Pyrosequencing.43,44
Statistical analysis
We used SAS software (Version 9.1.3, SAS Institute, Cary, NC). All p values were two-sided. Because of multiple hypothesis testing, a p value for significance was adjusted conservatively by Bonferroni correction to 0.0023 (= 0.05/22). Chi-square test was used to assess an association between categorical variables, and ANOVA was used to compare continuous variables across categories.
We adopted multivariate negative binomial regression analysis to assess predictors of the node count. This was because the marginal distribution of the total or negative node count fitted the Gamma-Poisson-like distribution (Figure 1), and then, over-dispersion occurred with Poisson generalized linear models. The process of estimation was based on a negative binomial distribution that can be conceptualized as a mixture of a Poisson distribution and a Gamma distribution.45 Variables initially included in a model were; sex; age (continuous); pre-diagnosis body mass index (continuous); family history of colorectal cancer in any first-degree relative; year of diagnosis (continuous); hospital setting (academic vs. non-academic); tumor size (continuous); specimen length (continuous); circumferential growth (100% complete vs. incomplete); tumor location and TNM stage (categorized as in table 1); tumor differentiation (poor vs. well-moderate); mucinous component (%, continuous); signet ring cells (%, continuous); peritumoral lymphocytic reaction and TIL (absent/minimal vs. mild vs. moderate vs. marked; ordinal); MSI (high vs. low/MSS); CIMP (high vs. low/0); LINE-1 methylation (continuous); and KRAS, BRAF and PIK3CA (mutant vs. wild-type). A backward elimination with a threshold of p=0.1 was performed to select variables in the final model, except for TNM stage and tumor location for which all categories were forced into the model. For cases with missing data in a covariate, we carried out two separate analyses; the first analysis included all patients with creation of a categorical indicator for missing responses (MIM: missing indicator method); the second analysis included all patients with missing responses imputed (MI: multiple imputation). The MI procedure in SAS was used to perform 20 imputations of all variables with missing cases by using regression method. The results from regression analysis were then appropriately combined by using the MIANALYZE procedure.
Table 1.
Features of colorectal cancer according to quartiles of the negative node count and total node count
| Clinical or molecular feature | All cases | No. of negative lymph nodes |
P value | Total No. of recovered lymph nodes |
P value | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 0-4 (quartile 1) | 5-8 (quartile 2) | 9-14 (quartile 3) | ≥15 (quartile 4) | 0-6 (quartile 1) | 7-10 (quartile 2) | 11-15 (quartile 3) | ≥16 (quartile 4) | ||||
| Total N | 918 | 237 | 226 | 233 | 222 | 247 | 240 | 195 | 236 | ||
| Sex | 0.60 | 0.32 | |||||||||
| Male (HPFS) | 455 (50%) | 121 (51%) | 118 (52%) | 108 (46%) | 108 (49%) | 134 (54%) | 116 (48%) | 89 (46%) | 116 (49%) | ||
| Female (NHS) | 463 (50%) | 116 (49%) | 108 (48%) | 125 (54%) | 114 (51%) | 113 (46%) | 124 (52%) | 106 (54%) | 120 (51%) | ||
| Age (mean ± SD) | 67.6 ± 8.4 | 67.3 ± 7.9 | 67.1 ± 8.2 | 67.8 ± 9.0 | 68.2 ± 8.5 | 0.16 | 67.4 ± 8.0 | 67.5 ± 8.2 | 67.8 ± 9.4 | 67.8 ± 8.2 | 0.50 |
| Body mass index | 0.36 | 0.58 | |||||||||
| <25 kg/m2 | 362 (40%) | 91 (39%) | 79 (37%) | 102 (45%) | 90 (41%) | 91 (38%) | 93 (40%) | 85 (45%) | 93 (40%) | ||
| 25-29.9 kg/m2 | 378 (42%) | 108 (46%) | 97 (45%) | 87 (39%) | 86 (39%) | 109 (45%) | 94 (41%) | 78 (41%) | 97 (42%) | ||
| ≥30 kg/m2 | 154 (17%) | 34 (15%) | 40 (19%) | 37 (16%) | 43 (20%) | 41 (17%) | 45 (19%) | 25 (13%) | 43 (18%) | ||
| Family history of colorectal cancer in first degree relative(s) | 0.076 | 0.069 | |||||||||
| (-) | 718 (80%) | 193 (83%) | 161 (74%) | 183 (81%) | 181 (83%) | 194 (81%) | 174 (75%) | 154 (82%) | 196 (84%) | ||
| (+) | 177 (20%) | 40 (17%) | 56 (26%) | 43 (19%) | 38 (17%) | 47 (20%) | 59 (25%) | 34 (18%) | 37 (16%) | ||
| Hospital setting | 0.14 | 0.20 | |||||||||
| Non-academic | 836 (91%) | 220 (93%) | 201 (89%) | 218 (94%) | 197 (89%) | 226 (92%) | 223 (93%) | 180 (92%) | 207 (88%) | ||
| Academic | 82 (8.9%) | 17 (7.2%) | 25 (11%) | 15 (6.4%) | 25 (11%) | 21 (8.5%) | 17 (7.1%) | 15 (7.7%) | 29 (12%) | ||
| Year of diagnosis | 0.0024 | 0.083 | |||||||||
| Prior to 1995 | 351 (38%) | 99 (42%) | 91 (40%) | 87 (37%) | 74 (33%) | 94 (38%) | 103 (43%) | 74 (38%) | 80 (34%) | ||
| 1995 to 1999 | 337 (37%) | 89 (38%) | 96 (42%) | 75 (32%) | 77 (35%) | 102 (41%) | 85 (35%) | 66 (34%) | 84 (36%) | ||
| 2000 to 2004 | 230 (25%) | 49 (21%) | 39 (17%) | 71 (30%) | 71 (32%) | 51 (21%) | 52 (22%) | 55 (28%) | 72 (31%) | ||
| Resected colorectal length (specimen length) | <0.0001 | <0.0001 | |||||||||
| <20 cm | 369 (43%) | 114 (56%) | 103 (49%) | 93 (42%) | 59 (27%) | 121 (55%) | 112 (50%) | 70 (39%) | 66 (29%) | ||
| 20-49 cm | 451 (53%) | 85 (42%) | 100 (48%) | 122 (55%) | 144 (67%) | 93 (42%) | 107 (48%) | 102 (57%) | 149 (66%) | ||
| ≥50 cm | 31 (3.6%) | 4 (2.0%) | 7 (3.3%) | 7 (3.2%) | 13 (6.0%) | 8 (3.6%) | 6 (2.7%) | 6 (3.4%) | 11 (4.9%) | ||
| Tumor size (cm) | 4.4 ± 2.0 | 3.9 ± 1.8 | 4.1 ± 1.8 | 4.7 ± 2.2 | 5.0 ± 2.1 | <0.0001 | 3.6 ± 1.7 | 4.4 ± 2.1 | 4.6 ± 1.9 | 5.0 ± 2.0 | <0.0001 |
| Tumor location | <0.0001 | <0.0001 | |||||||||
| Cecum | 162 (18%) | 30 (13%) | 41 (18%) | 40 (17%) | 51 (23%) | 35 (14%) | 43 (18%) | 31 (16%) | 53 (22%) | ||
| Ascending colon | 183 (20%) | 28 (12%) | 32 (14%) | 53 (23%) | 70 (32%) | 21 (8.5%) | 41 (17%) | 46 (24%) | 75 (32%) | ||
| Transverse colon | 98 (11%) | 23 (9.8%) | 20 (8.9%) | 30 (13%) | 25 (11%) | 22 (8.9%) | 27 (11%) | 24 (13%) | 25 (32%) | ||
| Descending colon | 64 (7.0%) | 18 (7.6%) | 22 (9.8%) | 13(5.6%) | 11 (5.0%) | 23 (9.3%) | 19 (8.0%) | 12 (6.3%) | 10 (4.2%) | ||
| Sigmoid colon | 228 (25%) | 73 (31%) | 63 (28%) | 55 (24%) | 37 (17%) | 77 (31%) | 60 (25%) | 48 (25%) | 43 (18%) | ||
| Rectum | 178 (20%) | 64 (27%) | 47 (21%) | 40 (17%) | 27 (12%) | 69 (28%) | 48 (20%) | 31 (16%) | 30 (13%) | ||
| Circumferential growth | 0.011 | 0.0002 | |||||||||
| Incomplete | 729 (79%) | 204 (86%) | 182 (81%) | 177 (76%) | 166 (75%) | 219 (89%) | 188 (78%) | 150 (77%) | 172 (73%) | ||
| Complete (100%) | 189 (21%) | 33 (14%) | 44 (19%) | 56 (24%) | 56 (25%) | 28 (11%) | 52 (22%) | 45 (23%) | 64 (27%) | ||
| TNM stage | <0.0001 | <0.0001 | |||||||||
| T1N0M0 | 75 (8.7%) | 32 (15%) | 19 (9.1%) | 13 (5.9%) | 11 (5.1%) | 42 (19%) | 15 (6.5%) | 9 (5.0%) | 9 (3.9%) | ||
| T2N0M0 | 144 (17%) | 36 (16%) | 36 (17%) | 39 (18%) | 33 (15%) | 56 (25%) | 30 (13%) | 29 (16%) | 29 (13%) | ||
| T3N0M0 | 255 (30%) | 28 (13%) | 62 (30%) | 71 (32%) | 94 (44%) | 47 (21%) | 74 (32%) | 52 (29%) | 82 (36%) | ||
| T4N0M0 | 21 (2.4%) | 5 (2.3%) | 3 (1.4%) | 7 (3.2%) | 6 (2.8%) | 7 (3.2%) | 4 (1.7%) | 4 (2.2%) | 6 (2.6%) | ||
| T1-2N1M0 | 40 (4.6%) | 13 (5.9%) | 9 (4.3%) | 15 (6.8%) | 3 (1.4%) | 15 (6.8%) | 8 (3.5%) | 14 (7.7%) | 3 (1.3%) | ||
| T3-4xN1M0 | 140 (16%) | 24 (11%) | 38 (18%) | 41 (19%) | 37 (17%) | 25 (11%) | 37 (16%) | 37 (20%) | 41 (18%) | ||
| T(any)N2M0 | 78 (9.0%) | 34 (16%) | 16 (7.7%) | 12 (5.5%) | 16 (7.4%) | 5 (2.3%) | 26 (11%) | 19 (10%) | 28 (12%) | ||
| T(any)N(any)M1 | 110 (13%) | 47 (21%) | 26 (12%) | 22 (10%) | 15 (7.0%) | 25 (11%) | 36 (16%) | 18 (9.9%) | 31 (13%) | ||
| Tumor differentiation | 0.46 | 0.010 | |||||||||
| Well-moderate | 838 (91%) | 222 (94%) | 205 (91%) | 212 (91%) | 199 (90%) | 238 (96%) | 215 (90%) | 176 (90%) | 209 (89%) | ||
| Poor | 80 (8.7%) | 15 (6.3%) | 21 (9.3%) | 21 (9.0%) | 23 (10%) | 9 (3.6%) | 25 (10%) | 19 (9.7%) | 27 (11%) | ||
| Mucinous component | 0.0031 | 0.012 | |||||||||
| 0% | 585 (64%) | 165 (70%) | 156 (69%) | 140 (60%) | 124 (56%) | 174 (70%) | 157 (65%) | 124 (64%) | 130 (55%) | ||
| 1-49% | 214 (23%) | 48 (20%) | 53 (23%) | 54 (23%) | 59 (27%) | 54 (22%) | 52 (22%) | 42 (22%) | 66 (28%) | ||
| ≥50% | 119 (13%) | 24 (10%) | 17 (7.5%) | 39 (17%) | 39 (18%) | 19 (7.7%) | 31 (13%) | 29 (15%) | 40 (17%) | ||
| Signet ring cell component | 0.23 | 0.82 | |||||||||
| 0% | 845 (92%) | 219 (92%) | 206 (91%) | 214 (92%) | 206 (93%) | 231 (94%) | 218 | 181 (93%) | 215 (91%) | ||
| 1-49% | 60 (6.5%) | 15 (6.3%) | 19 (8.4%) | 12 (5.2%) | 14 (6.3%) | 14 (5.7%) | 19 (7.9%) | 11 (5.6%) | 16 (6.8%) | ||
| ≥50% | 13 (1.4%) | 3 (1.3%) | 1 (0.4%) | 7 (3.0%) | 2 (0.9%) | 2 (0.8%) | 3 (1.3%) | 3 (1.5%) | 5 (2.1%) | ||
| Peritumoral lymphocytic reaction | 0.0022 | 0.035 | |||||||||
| Absent/minimal | 107 (12%) | 35 (15%) | 23 (11%) | 22 (9.8%) | 27 (13%) | 29 (12%) | 27 (12%) | 18 (9.5%) | 33 (14%) | ||
| Mild | 672 (76%) | 180 (79%) | 173 (80%) | 172 (77%) | 147 (69%) | 193 (81%) | 177 (78%) | 143 (75%) | 159 (70%) | ||
| Moderate/marked | 104 (12%) | 14 (6.1%) | 21 (9.7%) | 30 (13%) | 39 (18%) | 17 (7.1%) | 23 (10%) | 29 (15%) | 35 (15%) | ||
| Tumor infiltrating lymphocytes | 0.0012 | 0.018 | |||||||||
| Absent/minimal | 669 (76%) | 189 (82%) | 173 (79%) | 162 (72%) | 145 (68%) | 195 (81%) | 174 (76%) | 142 (75%) | 158 (70%) | ||
| Mild | 125 (14%) | 31 (13%) | 25 (11%) | 36 (16%) | 33 (15%) | 34 (14%) | 29 (13%) | 27 (14%) | 35 (15%) | ||
| Moderate/marked | 91 (10%) | 10 (4.4%) | 20 (9.2%) | 26 (12%) | 35 (16%) | 11 (4.6%) | 25 (11%) | 21 (11%) | 34 (15%) | ||
| MSI status | <0.0001 | <0.0001 | |||||||||
| MSI-low/MSS | 742 (85%) | 212 (93%) | 194 (89%) | 187 (83%) | 149 (73%) | 221 (93%) | 192 (83%) | 165 (87%) | 164 (75%) | ||
| MSI-high | 134 (15%) | 16 (7.0%) | 25 (11%) | 37 (17%) | 56 (27%) | 16 (6.8%) | 39 (17%) | 25 (13%) | 54 (25%) | ||
| CIMP status | <0.0001 | <0.0001 | |||||||||
| CIMP-low/0 | 739 (84%) | 209 (92%) | 192 (87%) | 183 (81%) | 155 (73%) | 219 (92%) | 194 (84%) | 159 (83%) | 167 (74%) | ||
| CIMP-high | 146 (17%) | 18 (7.9%) | 28 (13%) | 44 (19%) | 56 (27%) | 19 (8.0%) | 37 (16%) | 32 (17%) | 58 (26%) | ||
| LINE-1 methylation (mean ± SD) | 62.0 ± 9.4 | 61.1 ± 10.5 | 61.2 ± 9.4 | 62.4 ± 8.8 | 63.5 ± 8.7 | 0.0032 | 62.0 ± 10.1 | 61.2 ± 9.8 | 61.4 ± 8.5 | 63.4 ± 8.7 | 0.12 |
| KRAS mutation | 0.40 | 0.25 | |||||||||
| (-) | 560 (63%) | 146 (64%) | 141 (63%) | 151 (67%) | 122 (59%) | 158 (66%) | 149 (64%) | 126 (66%) | 127 (58%) | ||
| (+) | 322 (37%) | 82 (36%) | 82 (37%) | 74 (33%) | 84 (41%) | 81 (34%) | 84 (36%) | 65 (34%) | 92 (42%) | ||
| BRAF mutation | 0.0056 | 0.0037 | |||||||||
| (-) | 760 (86%) | 207 (90%) | 199 (90%) | 189 (85%) | 165 (80%) | 222 (92%) | 197 (85%) | 164 (87%) | 177 (80%) | ||
| 121 (14%) | 23 (10%) | 23 (10%) | 33 (15%) | 42 (20%) | 19 (7.9%) | 34 (15%) | 25 (13%) | 43 (20%) | |||
| PIK3CA mutation | 0.19 | 0.38 | |||||||||
| (-) | 663 (83%) | 182 (87%) | 170 (85%) | 165 (81%) | 146 (79%) | 182 (85%) | 178 (83%) | 148 (85%) | 155 (79%) | ||
| 136 (17%) | 28 (13%) | 31 (15%) | 39 (19%) | 38 (21%) | 33 (15%) | 36 (17%) | 26 (15%) | 41 (21%) | |||
(%) indicates the proportion of cases with a specific clinical, pathologic or molecular feature, among a given quartile category of the negative or total node count. P values were calculated by ANOVA (analysis of variance) for age, tumor size and LINE-1 methylation, and by chi-square test for all other variables. Because of multiple hypothesis testing, a p value for significance was adjusted by Bonferroni correction to 0.0023. CIMP, CpG island methylator phenotype; HPFS, Health Professionals Follow-up Study; MSI, microsatellite instability; MSS, microsatellite stable; NHS, Nurses’ Health Study; SD, standard deviation.
Results
Lymph node count in colorectal cancer resection
The total node count showed a skewed (Gamma-Poisson-like) distribution (Figure 1): range, 0 to 54; mean, 12.0; median, 10; interquartile rage, 6 to 16. The negative node count also showed a skewed (Gamma-Poisson-like) distribution: range, 0 to 54; mean, 10.5; median, 8; interquartile rage, 4 to 15.
Node count and clinical, pathologic and molecular features
Table 1 shows clinical, pathologic and molecular features of colorectal cancers according to quartiles of the negative or total node count. Both of the negative and total node counts were positively associated with specimen length, tumor size, ascending colon location, T3N0M0 stage, microsatellite instability (MSI), and the CpG island methylator phenotype (CIMP) (all p < 0.0001).
Multivariate (negative binomial regression) analysis for the node count
In a multivariate model, factors independently associated with the negative node count included specimen length, tumor location, TNM stage, year of diagnosis, and tumor size (all p < 0.002) (Table 2). In Table 2, for example, specimens with rectal cancer in average yielded the negative node count of approximately 2/3 (0.67) of that in specimens with ascending colon cancer, after controlling for the effects of other variables. Interestingly, KRAS mutation appeared to be a predictor of the negative node count (p = 0.032), although multiple hypothesis testing should be considered, and this finding needs to be confirmed by an independent dataset.
Table 2.
Multivariate negative binomial regression analysis to predict the negative lymph node count in colorectal cancer resections
| Variable in the final model to predict the negative node count | Adjusted fold-change in mean negative node count by a given variable (95% CI) | P value (class level-test) | P value (overall test) |
|---|---|---|---|
| Specimen length (for 10 cm increase as a unit) | 1.11 (1.06-1.16) | <0.0001 | |
| Tumor location (vs. ascending colon as a referent) | <0.0001 | ||
| Cecum | 0.90 (0.76-1.05) | 0.19 | |
| Transverse colon | 0.79 (0.65-0.95) | 0.012 | |
| Descending colon | 0.57 (0.45-0.71) | <0.0001 | |
| Sigmoid colon | 0.72 (0.62-0.83) | <0.0001 | |
| Rectum | 0.67 (0.57-0.78) | <0.0001 | |
| TNM stage (vs. T3N0M0 as a referent) | <0.0001 | ||
| T1N0M0 | 0.73 (0.60-0.89) | 0.0022 | |
| T2N0M0 | 0.86 (0.74-1.01) | 0.063 | |
| T4N0M0 | 0.75 (0.54-1.04) | 0.082 | |
| T1-2N1M0 | 0.69 (0.53-0.89) | 0.0049 | |
| T3-4N1M0 | 0.87 (0.75-1.01) | 0.067 | |
| T(any)N2M0 | 0.69 (0.57-0.84) | 0.0002 | |
| T(any)N(any)M1 | 0.59 (0.50-0.69) | <0.0001 | |
| Year of diagnosis (for 5 year increase as a unit) | 1.10 (1.04-1.16) | 0.0010 | |
| Tumor size (for 5 cm increase as a unit) | 1.24 (1.08-1.42) | 0.0019 | |
| Age at diagnosis (for 10 year increase as a unit) | 0.92 (0.86-0.98) | 0.0092 | |
| Circumferential growth (complete vs. incomplete) | 1.17 (1.03-1.33) | 0.019 | |
| KRAS mutation (vs. wild-type) | 1.13 (1.01-1.25) | 0.032 |
Variables (potential predictors) initially included in a regression model were: the variables listed in the table, sex, body mass index, family history of colorectal cancer, hospital setting, tumor differentiation, mucinous component, signet ring cells, peritumoral lymphocytic reaction, tumor infiltrating lymphocytes, CpG island methylator phenotype, microsatellite instability, LINE-1 methylation, and BRAF and PIK3CA mutations. A backward elimination with a threshold of p=0.1 was used to select variables in the final model. The adjusted fold-change in mean node count by a given variable represents the exponential of adjusted β coefficient derived by the final model. Because of multiple hypothesis testing, a p value for significance was adjusted by Bonferroni correction to 0.0023. CI, confidence interval.
Factors independently associated with the total node count included specimen length, tumor location, TNM stage, and tumor size, after controlling for effects of other variables (all p < 0.0001) (Table 3). KRAS mutation appeared to a predictor of the total node count (p = 0.0088), although multiple hypothesis testing should be considered.
Table 3.
Multivariate negative binomial regression analysis to predict the total number of recovered nodes in colorectal cancer resections
| Variable in the final model to predict the total node count | Adjusted fold-change in mean total node count by a given variable (95% CI) | P value (class level-test) | P value (overall test) |
|---|---|---|---|
| Specimen length (for 10 cm increase as a unit) | 1.09 (1.05-1.13) | <0.0001 | |
| Tumor location (vs. ascending colon as a referent) | <0.0001 | ||
| Cecum | 0.86 (0.75-0.99) | 0.039 | |
| Transverse colon | 0.82 (0.69-0.97) | 0.019 | |
| Descending colon | 0.61 (0.50-0.74) | <0.0001 | |
| Sigmoid colon | 0.73 (0.64-0.84) | <0.0001 | |
| Rectum | 0.71 (0.61-0.82) | <0.0001 | |
| TNM stage (vs. T3N0M0 as a referent) | <0.0001 | ||
| T1N0M0 | 0.74 (0.62-0.90) | 0.0022 | |
| T2N0M0 | 0.86 (0.75-0.99) | 0.030 | |
| T4N0M0 | 0.78 (0.58-1.03) | 0.083 | |
| T1-2N1M0 | 0.84 (0.67-1.05) | 0.12 | |
| T3-4N1M0 | 0.99 (0.87-1.13) | 0.92 | |
| T(any)N2M0 | 1.21 (1.03-1.43) | 0.019 | |
| T(any)N(any)M1 | 0.94 (0.81-1.08) | 0.38 | |
| Tumor size (for 5 cm increase as a unit) | 1.29 (1.14-1.46) | <0.0001 | |
| Year of diagnosis (for 5 year increase as a unit) | 1.07 (1.02-1.13) | 0.0044 | |
| KRAS mutation (vs. wild-type) | 1.13 (1.03-1.24) | 0.0088 | |
| Age at diagnosis (for 10 year increase as a unit) | 0.94 (0.89-0.99) | 0.028 | |
| Circumferential growth (complete vs. incomplete) | 1.14 (1.01-1.27) | 0.028 | |
| Peritumoral lymphocytic reaction (for a unit increase in severerity) | 1.08 (0.99-1.17) | 0.091 |
Variables (potential predictors) initially included in a regression model were: the variables listed in the table, sex, body mass index, family history of colorectal cancer, hospital setting, tumor differentiation, mucinous component, signet ring cells, tumor infiltrating lymphocytes, CpG island methylator phenotype, microsatellite instability, LINE-1 methylation, and BRAF and PIK3CA mutations. A backward elimination with a threshold of p=0.1 was used to select variables in the final model. The adjusted fold-change in mean node count by a given variable represents the exponential of adjusted β coefficient derived by the final model. Because of multiple hypothesis testing, a p value for significance was adjusted by Bonferroni correction to 0.0023. CI, confidence interval.
Node count in node-negative (stage I-II) colorectal cancer resections
In stage I-II patients, ascending colon tumor location, tumor size, and specimen length were positively associated with the node count, after controlling for effects of other variables (all p < 0.002) (Table 4). PIK3CA (p = 0.027) and KRAS (p = 0.049) mutations also appeared to predict the node count, although multiple hypothesis testing should be considered.
Table 4.
Multivariate negative binomial regression analysis to predict the node count in stage I-II colorectal cancer resections
| Variable in the final model to predict the total node count | Adjusted fold-change in mean total node count by a given variable (95% CI) | P value (class level-test) | P value (overall test) |
|---|---|---|---|
| Tumor location (vs. ascending colon as a referent) | <0.0001 | ||
| Cecum | 0.91 (0.75-1.10) | 0.33 | |
| Transverse colon | 0.77 (0.61-0.96) | 0.021 | |
| Descending colon | 0.66 (0.52-0.85) | 0.0013 | |
| Sigmoid colon | 0.64 (0.53-0.76) | <0.0001 | |
| Rectum | 0.66 (0.54-0.80) | <0.0001 | |
| Tumor size (for 5 cm increase as a unit) | 1.29 (1.10-1.52) | 0.0017 | |
| Specimen length (for 10 cm increase as a unit) | 1.09 (1.03-1.15) | 0.0019 | |
| TNM stage (vs. T3N0M0 as a referent) | 0.0090 | ||
| T1N0M0 | 0.75 (0.62-0.91) | 0.0031 | |
| T2N0M0 | 0.84 (0.72-0.97) | 0.016 | |
| T4N0M0 | 0.81 (0.60-1.09) | 0.17 | |
| PIK3CA mutation (vs. wild-type) | 1.21 (1.02-1.43) | 0.027 | |
| KRAS mutation (vs. wild-type) | 1.15 (1.00-1.31) | 0.049 |
Variables (potential predictors) initially included in a regression model were: the variables listed in the table, sex, age, body mass index, family history of colorectal cancer, year of diagnosis, hospital setting, circumferential growth, tumor differentiation, mucinous component, signet ring cells, peritumoral lymphocytic reaction, tumor infiltrating lymphocytes, CpG island methylator phenotype, microsatellite instability, LINE-1 methylation, and BRAF mutation. A backward elimination with a threshold of p=0.1 was used to select variables in the final model. The adjusted fold-change in mean node count by a given variable represents the exponential of adjusted β coefficient derived by the final model. Because of multiple hypothesis testing, a p value for significance was adjusted by Bonferroni correction to 0.0023. CI, confidence interval.
Discussion
We conducted this study to identify clinical, pathologic and tumor molecular variables which predict the node count in colorectal cancer resections, independent of human operator factors (i.e., surgeons and pathologists). We found that specimen length, tumor size, T3N0M0 stage, ascending colon tumor location, and year of diagnosis were positively associated with the negative node count. Interestingly, KRAS mutation might also be positively associated with the negative node count, which needs to be confirmed by an independent dataset. These results indicate that those operator (surgeon and pathologist)-independent variables influence the node count in colorectal resection, and should be examined in any future study that assesses adequacy of lymph node harvesting and staging.
Comprehensive assessment of clinical, pathological and molecular features is important in cancer research.46-49 Previous studies have reported that the recovered node count is positively associated with specimen length,14,18,19,21-24 proximal colon cancer,13,18,19,25,28 larger tumor size,11,14,18,20 number of vascular pedicles,50 and higher disease stage.17,18 However, in those studies,11,13,14,17-24,28 there were no data on host immune response to tumor, or tumor molecular features, despite possible influence of immune reaction and tumor molecular variables on the node count.9 Previous studies that examined the relationship between MSI and the node count lacked specimen length and molecular variables besides MSI.51,52 In contrast to all of those previous studies that examined potential predictors of the node count,11,13,14,17-24,28 we have utilized the U.S. nationwide cohort database with well-annotated clinical, specimen, pathologic and tumor molecular data including MSI, CIMP, and KRAS, BRAF and PIK3CA mutations, all of which are potential predictors of the node count.
With regard to the influence of medical care quality or socioeconomic status on the lymph node count,12,53 academic hospital status5,14,25 and the degree of practicing experience of surgeons11,18 and pathologists11,13,17,18,54 have been associated with the node count (reviewed by Storli et al.55). However, all but three14,18,24 of those previous studies on medical care quality or socioeconomic status5,11,13,16,17,53,54 lacked data on specimen length.
There are advantages in utilizing the database of the two U.S. nationwide prospective cohort studies, to assess operator-independent predictors of the node count. First, cohort participants who developed cancer were treated at hospitals throughout the U.S. (Figure 1), and more representative colorectal cancers in the general U.S. population than patients in one to a few hospitals. Second, because of our study design, any particular surgeon or pathologist could not have influenced our results, increasing generalizability of our findings. Third, our rich molecular pathological epidemiology8,30,31 database enabled us to simultaneously assess a number of variables, and to adjust for potential confounding.
One weakness of our current study is that participants of our cohort studies are U.S. health professionals and predominantly non-Hispanic Caucasians, thus constituting a rather homogeneous group of people, and lacked other occupational and ethnic groups. One of the primary reasons of selecting health professionals as subjects in the cohort studies was that those people have good understanding of various diseases as well as the value of the cohort studies, which increase reliability and compliance in questionnaire-based follow-up and data collection. Second, the exclusion of a subset of cancer cases without available tumor tissue, which might potentially cause bias. Nonetheless, tumor specimen procurement rate has been 60-70% of attempts, and our previous study has shown that there is no substantial demographic or clinical difference between cases with and without tumor tissue analyzed.32
Our ultimate goal is to determine how many nodes are enough for an optimal care in each individual case for personalized medicine. To achieve this goal, we need to assemble an adequate database with prospective follow-up to record detailed outcome data, preferably in a trial setting. At this point, we do not have enough data to recommend a specific number of nodes to be examined for an optimal patient care. Nonetheless, our unique dataset, which has provided strong evidence for effects of specimen length, tumor size, tumor location and TNM stage on the node count, will likely serve as a guide for such future trials.
In conclusion, our study has shown that specimen length, tumor size and location, and TNM stage are predictors for the lymph node count in colorectal cancer resections, independent of operator (surgeon and pathologist) factors. In addition, some tumor molecular features such as KRAS mutation might influence the node count, which needs to be confirmed by an independent dataset. Our data suggest that those clinical, pathologic, specimen and molecular variables should be examined as crucial variables in any future evaluation of adequacy of lymph node examination for colorectal cancer patients.
Figure 2.
Lymph node counts in 918 colorectal cancers in our two U.S. nationwide prospective cohort studies. (A) Distribution of the negative node count. (B) Distribution of the total node count. Both negative and total node counts approximately follow a Gamma-Poisson-like distribution. (C) Correlation between the negative node count and specimen length. (D) Correlation between the negative node count and tumor size.
Acknowledgments
We would like to thank the participants and staff of the Nurses’ Health Study and the Health Professionals Follow-Up Study, for their valuable contributions as well as the following state cancer registries for their help: AL, AZ, AR, CA, CO, CT, DE, FL, GA, ID, IL, IN, IA, KY, LA, ME, MD, MA, MI, NE, NH, NJ, NY, NC, ND, OH, OK, OR, PA, RI, SC, TN, TX, VA, WA, WY.
Funding: This work was supported by U.S. National Institute of Health [P01 CA87969 (to S.E. Hankinson), P01 CA55075 (to W.C. Willett), P50 CA127003 (to C.S.F.), R01 CA151993 (to S.O.), and R01 CA137178 (to A.T.C.)]; the Bennett Family Fund for Targeted Therapies Research; and the Entertainment Industry Foundation through National Colorectal Cancer Research Alliance. The content is solely the responsibility of the authors and does not necessarily represent the official views of NCI or NIH. Funding agencies did not have any role in the design of the study; the collection, analysis, or interpretation of the data; the decision to submit the manuscript for publication; or the writing of the manuscript.
Abbreviations
- ANOVA
analysis of variance
- BMI
body mass index
- CI
confidence interval
- CIMP
CpG island methylator phenotype
- HPFS
Health Professionals Follow-up Study
- MI
multiple imputation
- MIM
missing indicator method
- MPE
molecular pathological epidemiology
- MSI
microsatellite instability
- MSS
microsatellite stable
- NHS
Nurses’ Health Study
- TIL
tumor infiltrating lymphocytes
- TNM
tumor-node-metastasis
Footnotes
TM, NT, AK, and KN contributed equally
Conflict of interest: None to declare.
Author Contributions: Study concept and design: Ogino. Acquisition of data: Morikawa, Tanaka, Kuchiba, Nosho, Yamauchi, Hornick, Swanson, Chan, Meyerhardt, Fuchs and Ogino. Analysis and interpretation of data: Morikawa, Tanaka, Kuchiba, Nosho, Chan, Meyerhardt, Huttenhower, Schrag, Fuchs and Ogino. Drafting of the manuscript: Morikawa, Tanaka, Kuchiba, Nosho and Ogino. Critical revision of the manuscript for important intellectual content: Morikawa, Tanaka, Kuchiba, Nosho, Yamauchi, Hornick, Swanson, Chan, Meyerhardt, Huttenhower, Schrag, Fuchs and Ogino. Obtained funding: Chan, Fuchs and Ogino. Administrative, technical, and material support: Morikawa, Fuchs and Ogino. Study supervision: Fuchs and Ogino.
N.T. and S.O. had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
References
- 1.Compton CC. Optimal pathologic staging: defining stage II disease. Clin Cancer Res. 2007;13(22 Pt 2):6862s–6870s. doi: 10.1158/1078-0432.CCR-07-1398. [DOI] [PubMed] [Google Scholar]
- 2.Chang GJ, Rodriguez-Bigas MA, Skibber JM, Moyer VA. Lymph node evaluation and survival after curative resection of colon cancer: systematic review. J Natl Cancer Inst. 2007;99(6):433–441. doi: 10.1093/jnci/djk092. [DOI] [PubMed] [Google Scholar]
- 3.Hammond ME, Fitzgibbons PL, Compton CC, et al. College of American Pathologists Conference XXXV: solid tumor prognostic factors-which, how and so what? Summary document and recommendations for implementation. Cancer Committee and Conference Participants. Arch Pathol Lab Med. 2000;124(7):958–965. doi: 10.5858/2000-124-0958-COAPCX. [DOI] [PubMed] [Google Scholar]
- 4.Nelson H, Petrelli N, Carlin A, et al. Guidelines 2000 for colon and rectal cancer surgery. J Natl Cancer Inst. 2001;93(8):583–596. doi: 10.1093/jnci/93.8.583. [DOI] [PubMed] [Google Scholar]
- 5.Bilimoria KY, Bentrem DJ, Stewart AK, et al. Lymph node evaluation as a colon cancer quality measure: a national hospital report card. J Natl Cancer Inst. 2008;100(18):1310–1317. doi: 10.1093/jnci/djn293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Parsons HM, Tuttle TM, Kuntz KM, Begun JW, McGovern PM, Virnig BA. Association between lymph node evaluation for colon cancer and node positivity over the past 20 years. JAMA. 2011;306(10):1089–1097. doi: 10.1001/jama.2011.1285. [DOI] [PubMed] [Google Scholar]
- 7.Porter GA, Urquhart R, Bu J, Johnson P, Rayson D, Grunfeld E. Improving Nodal Harvest in Colorectal Cancer: So What? Ann Surg Oncol. 2012 doi: 10.1245/s10434-011-2073-9. in press (published online: Oct 4, 2011; doi: 10.1245/s10434-011-2073-9) [DOI] [PubMed] [Google Scholar]
- 8.Ogino S, Galon J, Fuchs CS, Dranoff G. Cancer immunology--analysis of host and tumor factors for personalized medicine. Nat Rev Clin Oncol. 2011;8(12):711–719. doi: 10.1038/nrclinonc.2011.122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ogino S, Nosho K, Irahara N, et al. Negative lymph node count is associated with survival of colorectal cancer patients, independent of tumoral molecular alterations and lymphocytic reaction. Am J Gastroenterol. 2010;105(2):420–433. doi: 10.1038/ajg.2009.578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.George S, Primrose J, Talbot R, et al. Will Rogers revisited: prospective observational study of survival of 3592 patients with colorectal cancer according to number of nodes examined by pathologists. Br J Cancer. 2006;95(7):841–847. doi: 10.1038/sj.bjc.6603352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Morris EJ, Maughan NJ, Forman D, Quirke P. Identifying stage III colorectal cancer patients: the influence of the patient, surgeon, and pathologist. J Clin Oncol. 2007;25(18):2573–2579. doi: 10.1200/JCO.2007.11.0445. [DOI] [PubMed] [Google Scholar]
- 12.Wong SL, Ji H, Hollenbeck BK, Morris AM, Baser O, Birkmeyer JD. Hospital lymph node examination rates and survival after resection for colon cancer. Jama. 2007;298(18):2149–2154. doi: 10.1001/jama.298.18.2149. [DOI] [PubMed] [Google Scholar]
- 13.Lemmens VE, van Lijnschoten I, Janssen-Heijnen ML, Rutten HJ, Verheij CD, Coebergh JW. Pathology practice patterns affect lymph node evaluation and outcome of colon cancer: a population-based study. Ann Oncol. 2006;17(12):1803–1809. doi: 10.1093/annonc/mdl312. [DOI] [PubMed] [Google Scholar]
- 14.Wright FC, Law CH, Last L, et al. Lymph node retrieval and assessment in stage II colorectal cancer: a population-based study. Ann Surg Oncol. 2003;10(8):903–909. doi: 10.1245/aso.2003.01.012. [DOI] [PubMed] [Google Scholar]
- 15.Grote T, Hughes AH, Rimmer CC, Less DA, Abernethy AP. Targeting Lymph Node Retrieval and Assessment in Stage II Colon Cancer: A Quality Outcome Community-Based Cancer Center Study. J Oncol Pract. 2008;4(2):55–58. doi: 10.1200/JOP.0822001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cone MM, Shoop KM, Rea JD, Lu KC, Herzig DO. Ethnicity Influences Lymph Node Resection in Colon Cancer. J Gastrointest Surg. 2010;14(11):1752–1757. doi: 10.1007/s11605-010-1296-6. [DOI] [PubMed] [Google Scholar]
- 17.Evans MD, Barton K, Rees A, Stamatakis JD, Karandikar SS. The impact of surgeon and pathologist on lymph node retrieval in colorectal cancer and its impact on survival for patients with Dukes’ stage B disease. Colorectal Dis. 2008;10(2):157–164. doi: 10.1111/j.1463-1318.2007.01225.x. [DOI] [PubMed] [Google Scholar]
- 18.Valsecchi ME, Leighton J, Jr., Tester W. Modifiable factors that influence colon cancer lymph node sampling and examination. Clin Colorectal Cancer. 2010;9(3):162–167. doi: 10.3816/CCC.2010.n.022. [DOI] [PubMed] [Google Scholar]
- 19.Shen SS, Haupt BX, Ro JY, Zhu J, Bailey HR, Schwartz MR. Number of lymph nodes examined and associated clinicopathologic factors in colorectal carcinoma. Arch Pathol Lab Med. 2009;133(5):781–786. doi: 10.5858/133.5.781. [DOI] [PubMed] [Google Scholar]
- 20.Thorn CC, Woodcock NP, Scott N, Verbeke C, Scott SB, Ambrose NS. What factors affect lymph node yield in surgery for rectal cancer? Colorectal Dis. 2004;6(5):356–361. doi: 10.1111/j.1463-1318.2004.00670.x. [DOI] [PubMed] [Google Scholar]
- 21.Norwood MG, Sutton AJ, West K, Sharpe DP, Hemingway D, Kelly MJ. Lymph node retrieval in colorectal cancer resection specimens: national standards are achievable, and low numbers are associated with reduced survival. Colorectal Dis. 2010;12(4):304–309. doi: 10.1111/j.1463-1318.2009.01788.x. [DOI] [PubMed] [Google Scholar]
- 22.Gelos M, Gelhaus J, Mehnert P, et al. Factors influencing lymph node harvest in colorectal surgery. Int J Colorectal Dis. 2008;23(1):53–59. doi: 10.1007/s00384-007-0378-8. [DOI] [PubMed] [Google Scholar]
- 23.Neufeld D, Bugyev N, Grankin M, et al. Specimen length as a perioperative surrogate marker for adequate lymphadenectomy in colon cancer: the surgeon's role. Int Surg. 2007;92(3):155–160. [PubMed] [Google Scholar]
- 24.Stocchi L, Fazio VW, Lavery I, Hammel J. Individual Surgeon, Pathologist, and Other Factors Affecting Lymph Node Harvest in Stage II Colon Carcinoma. Is a Minimum of 12 Examined Lymph Nodes Sufficient? Ann Surg Oncol. 2011;18(2):405–412. doi: 10.1245/s10434-010-1308-5. [DOI] [PubMed] [Google Scholar]
- 25.Senthil M, Trisal V, Paz IB, Lai LL. Prediction of the adequacy of lymph node retrieval in colon cancer by hospital type. Arch Surg. 2010;145(9):840–843. doi: 10.1001/archsurg.2010.182. [DOI] [PubMed] [Google Scholar]
- 26.Lan YT, Lin JK, Jiang JK, Chang SC, Liang WY, Yang SH. Significance of lymph node retrieval from the terminal ileum for patients with cecal and ascending colonic cancers. Ann Surg Oncol. 2011;18(1):146–152. doi: 10.1245/s10434-010-1270-2. [DOI] [PubMed] [Google Scholar]
- 27.Lagoudianakis E, Pappas A, Koronakis N, et al. Lymph node harvesting in colorectal carcinoma specimens. Tumori. 2011;97(1):74–78. doi: 10.1177/030089161109700114. [DOI] [PubMed] [Google Scholar]
- 28.Chou JF, Row D, Gonen M, Liu YH, Schrag D, Weiser MR. Clinical and pathologic factors that predict lymph node yield from surgical specimens in colorectal cancer: a population-based study. Cancer. 2010;116(11):2560–2570. doi: 10.1002/cncr.25032. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Tekkis PP, Smith JJ, Heriot AG, Darzi AW, Thompson MR, Stamatakis JD. A national study on lymph node retrieval in resectional surgery for colorectal cancer. Dis Colon Rectum. 2006;49(11):1673–1683. doi: 10.1007/s10350-006-0691-2. [DOI] [PubMed] [Google Scholar]
- 30.Ogino S, Stampfer M. Lifestyle factors and microsatellite instability in colorectal cancer: the evolving field of molecular pathological epidemiology. J Natl Cancer Inst. 2010;102(6):365–367. doi: 10.1093/jnci/djq031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Ogino S, Chan AT, Fuchs CS, Giovannucci E. Molecular pathological epidemiology of colorectal neoplasia: an emerging transdisciplinary and interdisciplinary field. Gut. 2011;60(3):397–411. doi: 10.1136/gut.2010.217182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Chan AT, Ogino S, Fuchs CS. Aspirin and the Risk of Colorectal Cancer in Relation to the Expression of COX-2. New Engl J Med. 2007;356(21):2131–2142. doi: 10.1056/NEJMoa067208. [DOI] [PubMed] [Google Scholar]
- 33.Morikawa T, Kuchiba A, Yamauchi M, et al. Association of CTNNB1 (beta-catenin) alterations, body mass index, and physical activity with survival in patients with colorectal cancer. JAMA. 2011;305(16):1685–1694. doi: 10.1001/jama.2011.513. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ogino S, Nosho K, Irahara N, et al. Lymphocytic reaction to colorectal cancer is associated with longer survival, independent of lymph node count, microsatellite instability, and CpG island methylator phenotype. Clin Cancer Res. 2009;15(20):6412–6420. doi: 10.1158/1078-0432.CCR-09-1438. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ogino S, Kawasaki T, Brahmandam M, et al. Sensitive sequencing method for KRAS mutation detection by Pyrosequencing. J Mol Diagn. 2005;7(3):413–421. doi: 10.1016/S1525-1578(10)60571-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Ogino S, Kawasaki T, Kirkner GJ, Loda M, Fuchs CS. CpG island methylator phenotype-low (CIMP-low) in colorectal cancer: possible associations with male sex and KRAS mutations. J Mol Diagn. 2006;8:582–588. doi: 10.2353/jmoldx.2006.060082. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Nosho K, Kawasaki T, Ohnishi M, et al. PIK3CA mutation in colorectal cancer: relationship with genetic and epigenetic alterations. Neoplasia. 2008;10(6):534–541. doi: 10.1593/neo.08336. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Nosho K, Irahara N, Shima K, et al. Comprehensive biostatistical analysis of CpG island methylator phenotype in colorectal cancer using a large population-based sample. PLoS ONE. 2008;3(11):e3698. doi: 10.1371/journal.pone.0003698. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Ogino S, Nosho K, Kirkner GJ, et al. CpG island methylator phenotype, microsatellite instability, BRAF mutation and clinical outcome in colon cancer. Gut. 2009;58:90–96. doi: 10.1136/gut.2008.155473. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Ogino S, kawasaki T, Brahmandam M, et al. Precision and performance characteristics of bisulfite conversion and real-time PCR (MethyLight) for quantitative DNA methylation analysis. J Mol Diagn. 2006;8(2):209–217. doi: 10.2353/jmoldx.2006.050135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Weisenberger DJ, Siegmund KD, Campan M, et al. CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer. Nat Genet. 2006;38(7):787–793. doi: 10.1038/ng1834. [DOI] [PubMed] [Google Scholar]
- 42.Ogino S, Cantor M, Kawasaki T, et al. CpG island methylator phenotype (CIMP) of colorectal cancer is best characterised by quantitative DNA methylation analysis and prospective cohort studies. Gut. 2006;55(7):1000–1006. doi: 10.1136/gut.2005.082933. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Ogino S, Nosho K, Kirkner GJ, et al. A cohort study of tumoral LINE-1 hypomethylation and prognosis in colon cancer. J Natl Cancer Inst. 2008;100:1734–1738. doi: 10.1093/jnci/djn359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Irahara N, Nosho K, Baba Y, et al. Precision of Pyrosequencing assay to measure LINE-1 methylation in colon cancer, normal colonic mucosa and peripheral blood cells. J Mol Diagn. 2010;12(2):177–183. doi: 10.2353/jmoldx.2010.090106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Agresti A. Categorical Data Analysis. 2nd edition Wiley Interscience; New York, NY: 2002. [Google Scholar]
- 46.Hayanga AJ, Mukherjee D, Chang D, et al. Teaching hospital status and operative mortality in the United States: tipping point in the volume-outcome relationship following colon resections? Arch Surg. 2010;145(4):346–350. doi: 10.1001/archsurg.2010.24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Nicolini A, Ferrari P, Duffy MJ, et al. Intensive risk-adjusted follow-up with the CEA, TPA, CA19.9, and CA72.4 tumor marker panel and abdominal ultrasonography to diagnose operable colorectal cancer recurrences: effect on survival. Arch Surg. 2010;145(12):1177–1183. doi: 10.1001/archsurg.2010.251. [DOI] [PubMed] [Google Scholar]
- 48.Cone MM, Herzig DO, Diggs BS, et al. Dramatic decreases in mortality from laparoscopic colon resections based on data from the Nationwide Inpatient Sample. Arch Surg. 2011;146(5):594–599. doi: 10.1001/archsurg.2011.79. [DOI] [PubMed] [Google Scholar]
- 49.Piessen G, Muscari F, Rivkine E, et al. Prevalence of and Risk Factors for Morbidity After Elective Left Colectomy: Cancer vs Noncomplicated Diverticular Disease. Arch Surg. 2011;146(10):1149–1155. doi: 10.1001/archsurg.2011.231. [DOI] [PubMed] [Google Scholar]
- 50.Nash GM, Row D, Weiss A, et al. A predictive model for lymph node yield in colon cancer resection specimens. Ann Surg. 2011;253(2):318–322. doi: 10.1097/SLA.0b013e318204e637. [DOI] [PubMed] [Google Scholar]
- 51.Eveno C, Nemeth J, Soliman H, et al. Association between a high number of isolated lymph nodes in T1 to T4 N0M0 colorectal cancer and the microsatellite instability phenotype. Arch Surg. 2010;145(1):12–17. doi: 10.1001/archsurg.2009.224. [DOI] [PubMed] [Google Scholar]
- 52.Belt EJ, Te Velde EA, Krijgsman O, et al. High Lymph Node Yield is Related to Microsatellite Instability in Colon Cancer. Ann Surg Oncol. 2012 doi: 10.1245/s10434-011-2091-7. in press (published online: Oct 12, 2011; doi: 10.1245/s10434-011-2091-7) [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.McBride RB, Lebwohl B, Hershman DL, Neugut AI. Impact of socioeconomic status on extent of lymph node dissection for colon cancer. Cancer Epidemiol Biomarkers Prev. 2010;19(3):738–745. doi: 10.1158/1055-9965.EPI-09-1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Ostadi MA, Harnish JL, Stegienko S, Urbach DR. Factors affecting the number of lymph nodes retrieved in colorectal cancer specimens. Surg Endosc. 2007;21(12):2142–2146. doi: 10.1007/s00464-007-9414-6. [DOI] [PubMed] [Google Scholar]
- 55.Storli K, Lindboe CF, Kristoffersen C, Kleiven K, Sondenaa K. Lymph node harvest in colon cancer specimens depends on tumour factors, patients and doctors, but foremost on specimen handling. APMIS. 2011;119(2):127–134. doi: 10.1111/j.1600-0463.2010.02702.x. [DOI] [PubMed] [Google Scholar]