Brief Reports
Copyright ©2005 Baishideng Publishing Group Inc. All rights reserved.
World J Gastroenterol. Apr 14, 2005; 11(14): 2179-2183
Published online Apr 14, 2005. doi: 10.3748/wjg.v11.i14.2179
Relationships of tumor inflammatory infiltration and necrosis with microsatellite instability in colorectal cancers
Jing-Fang Gao, Gunnar Arbman, Tabasum Imran Wadhra, Hong Zhang, Xiao-Feng Sun
Jing-Fang Gao, Tabasum Imran Wadhra, Xiao-Feng Sun, Department of Oncology, Institute of Biomedicine and Surgery, University of Linköping, Linköping, Sweden
Gunnar Arbman, Department of Surgery, Vrinnevi Hospital, Norrköping, Sweden
Hong Zhang, Department of Dermatology, Institute of Biomedicine and Surgery, University of Linköping, Linköping, Sweden
Jing-Fang Gao, Department of Oncology, Institute of Biomedicine and Surgery, University of Linköping, S-581 85 Linköping, Sweden
Author contributions: All authors contributed equally to the work.
Supported by grants from the Swedish Cancer Foundation and the Health Research Council in the South-East of Sweden
Correspondence to: Xiao-Feng Sun, Associate Professor, MD, PhD, Department of Oncology, Institute of Biomedicine and Surgery, University of Linköping, S-581 85 Linköping, Sweden.
Telephone: +46-13-222066 Fax: +46-13-222846
Received: September 16, 2004
Revised: September 17, 2004
Accepted: October 8, 2004
Published online: April 14, 2005


AIM: The relationships between microsatellite instability (MSI) and survival in colorectal cancer patients are not consistent. The favorable survival of patient with MSI has been suggested to be related to pronounced inflammatory infiltration; however, the reason for non-association of MSI with survival is unclear. Our aims were to investigate the associations of inflammatory infiltration and tumor necrosis (TN) with microsatellite status and clinicopathological factors in colorectal cancer patients in whom MSI was not related to survival.

METHODS: Three hundred and one colorectal adenocar-cinomas were evaluated for inflammatory infiltration and 300 for TN under light microscope.

RESULTS: Low infiltration at invasive margin (χ2 = 3.94, P = 0.047) and in whole tumor stroma (χ2 = 3.89, P = 0.049) was associated with MSI, but TN was not (χ2=0.10, P = 0.75). Low infiltration was related to advanced stage (χ2 = 8.67, P = 0.03), poorer differentiation (χ2 = 8.84, P = 0.03), DNA non-diploid (χ2 = 10.04, P = 0.002), higher S-phase fraction (χ2 = 11.30, P = 0.004), positive p53 expression (χ2 = 7.94, P = 0.01), and worse survival (P = 0.03 for both univariate and multivariate analyses). Abundant TN was related to advanced stage (χ2 = 17.74, P = 0.001) and worse survival (P = 0.02 for univariate, and P = 0.05 for multivariate analysis).

CONCLUSION: The result that high inflammatory infiltration was not related to MSI might help explain the non-association of MSI with survival in colorectal cancer patients.

Key Words: Inflammatory infiltration, Necrosis, Microsatellite instability, Prognosis, Colorectal cancer


Colorectal cancer arises through at least two distinct genetic pathways in its carcinogenesis: microsatellite instability (MSI) and chromosomal instability[1]. MSI refers to genome-wide alteration in repetitive DNA sequence caused by deficiencies in DNA mismatch repair machinery[2], which accounts for about 10-15% of sporadic colorectal cancers and nearly all hereditary non-polyposis colorectal cancers[1]. Studies have shown that colorectal cancers with MSI are likely to be characterized by more frequent right-sided location, poor differentiation, mucinous/signet-ring cell carcinoma, and intense peri- and intra-tumoral inflammatory reaction[1,3,4]. Patients with MSI tumors appear to have a favorable prognosis compared with those with microsatellite stability (MSS) tumors[1,5-8]. The favorable prognosis associated with MSI has been suggested to be related to an enhanced inflammatory infiltration in the tumors[1,5] although the mechanism behind this phenomenon is unclear.

However, several studies including our previous two studies have reported the lack of association between MSI and survival either in entire group of colorectal cancer patients[9-14] or in subgroup with stage II colorectal cancers[15,16]. The reason for the non-association remains unclear. To our knowledge, no one has studied the association of inflammatory infiltration with microsatellite status in the patients in whom MSI is not related to survival. A recent study showed that necrosis in tumor was related to MSI-H in colorectal cancers[17]. Tumor necrosis (TN) is a common feature of solid tumors associated with a poor clinical outcome due to rapid tumor growth without sufficient blood supply[18]. However, the association of TN with microsatellite status has not been well studied. Therefore, it is interesting to evaluate the relationship of inflammatory infiltration and TN with MSI in colorectal cancers in which MSI was not related to survival. Meanwhile, we analyzed the relationship of inflammatory infiltration and TN with clinicopathological and other variables.


Three hundred and one primary colorectal adenocarcinomas were studied for inflammatory infiltration and 300 for TN. The patients were diagnosed at the Department of Pathology, Linköping Hospital, Linköping, and Vrinnevi Hospital, Norrköping, Sweden, between 1975 and 2001. The patient’s gender, age, tumor location, and Dukes’ stage were confirmed from surgical and/or pathological records. Tumor growth pattern and the grade of differentiation were scored by two pathologists. The mean age of the patients was 71 years (ranging from 34 to 94 years). Tumors from the ascending and transverse colon were regarded as proximal tumors, whereas tumors from descending and sigmoid colon, and the rectum were considered distal. The tumor growth pattern was divided into expansive or infiltrative type based on pattern of growth and invasiveness. Differentiation was graded as well, moderately, poorly differentiated, and mucinous/signet-ring cell carcinoma. The data on microsatellite status[13,14], DNA ploidy, S-phase fraction (SPF)[19], and p53 expression[20] were taken from previous studies carried out at our laboratory. Microsatellite status was determined by a microsatellite analysis using the Bat26 marker; 25 cases were MSI, and 152 were MSS. DNA ploidy and SPF were measured by flow cytometry; 107 cases were DNA diploid, and 118 were non-diploid; 52 were <5% SPF, 65 were 5-10%, and 72 were >10%. p53 expression was identified with immunohistochemistry by using CM1 antibody; 117 cases were p53 negative (completely negative cases plus the cases with <5% of p53-stained tumor cells), and 114 were positive (the cases having ≥5% of p53-stained tumor cells). No information was available about patients’ age in two cases, tumor site in 7, Dukes’ stage in 8, growth pattern in 21, grade of differentiation in 1, microsatellite status in 125, DNA ploidy in 75, SPF in 111, and p53 expression in 69 cases. Among 301 patients, 10 MSS and one MSI cases had received adjuvant preoperative radiotherapy, one MSS case had palliative radiotherapy, two MSS had adjuvant chemotherapy, and one MSS had palliative chemotherapy. No information was available for two cases, and the rest did not receive any radiotherapy or chemotherapy. The patients were followed up until the end of October 2001, and 125 died of colorectal cancer.

Histopathological evaluation

Three to ten sections from different parts of the tumor were examined at low magnification (×10) under light microscope by two of the authors (of whom one is a pathologist) independently in a blinded fashion without knowing the clinicopathological and other data of the patients. After the first run of the scoring, approximately 20% of the cases with disagreed score were reread independently by the two authors. Finally, about 4% of the cases with discrepant scoring were discussed under a dual-headed microscope to reach agreement on the scoring. Inflammatory infiltration and TN in the margins of the sections were not included in order to avoid artifacts. Infiltrating inflammatory cells were identified as small mononuclear cells in the stroma of tumor. The distributions of the infiltration were classified into two groups by localization: (a) those presented along the invasive margin of the tumor; (b) those distributed in entire tumor. The degree of infiltration was classified as absent, sparse, moderate, and intense according to the density of inflammatory cell[21]. Necrosis was scored as absent; <10%; 10-30%, and >30% based on the percentage of necrosis in whole tumor area[18]. Since the distributions of inflammatory infiltration and NT were often heterogeneous, the entire sections were examined to assess tumor areas including high and low inflammatory infiltration and NT. If higher infiltration or NT was more than one-third of the section, it was taken into account for scoring.

Statistical analysis

The relationships of inflammatory infiltration and TN with survival were tested using Cox’s Proportional Hazard Model. Survival curves were calculated using the Kaplan-Meier method. The relationships of inflammatory infiltration and TN with other variables were tested by using the χ2 test or Fisher exact test. Two-sided P values of less than 5% were considered statistically significant.

Inflammatory infiltration in relation to microsatellite status, clinicopathological and other variables

Among the 301 tumors studied, inflammatory cells at the invasive margin were absent in one case, sparse in 105, moderate in 131, and intense in 64 (Figure 1A); inflammatory cells in the tumors were absent in 17, sparse in 135, moderate in 110, and intense in 39 cases (Figure 1B). Absent, sparse, and moderate were combined as a low-level group, and intense as a high-level group, according to the similarities of the clinicopathological features.

Figure 1
Figure 1 High inflammatory infiltration at the tumor invasive margin (A) and in the tumor (B), and TN (C). Hematoxylin and eosin-stained sections of colorectal cancer (×10).

The relationship of inflammatory infiltration at the invasive margin or in the tumor with microsatellite status is present in Table 1. Low inflammatory infiltration either at the invasive margin or in the tumor was lightly associated with MSI (χ2 = 3.94, P = 0.047 and χ2 = 3.89, P = 0.049).

Table 1 Relationship of tumor inflammatory infiltration and necrosis with microsatellite status.
Variable CategoryMicrosatellite stability (%)Microsatellite instability (%)P
Infiltration at the invasive margin0.047
Low120 (79)24 (96)
High31 (21)1 (4)
Infiltration in the tumor0.049
Low71 (46)17 (68)
High81 (54)8 (32)
Little76 (51)13 (54)
Abundant74 (49)11 (46)

The relationship of inflammatory infiltration at the invasive margin with clinicopathological and other variables is summarized in Table 2. The degree of inflammatory infiltration was lower in tumors with Dukes’ stage D (χ2 = 8.67, P = 0.03), DNA non-diploid (χ2 = 10.04, P = 0.002), higher SPF (χ2 = 11.30, P = 0.004), p53 positive expression (χ2 = 7.94, P = 0.01), as well as poor differentiation and mucinous/signet-ring cell carcinoma (χ2 = 8.84, P = 0.03). Moreover, the patients with low inflammatory infiltration had worse survival than those with high infiltration (P = 0.03; Figure 2A), even after adjustment for patients’ gender, Dukes’ stage and differentiation (P = 0.03, data not shown). We did not find any relationship between inflammatory infiltration and patients’ gender, tumor location, or growth pattern (P>0.05).

Table 2 Relationship of tumor inflammatory infiltration at the invasive margin and TN with clinicopathological and other variables.
VariableCategoryInflammatory infiltration at the invasive margin
nHigh (%)PnAbundantP
Male15827 (17)15885 (54)
Female14337 (26)14268 (48)
Age (yr)0.680.12
≤7013831 (22)13877 (56)
>7016133 (21)16075 (47)
Tumor location0.550.75
Proximal10925 (23)10854 (50)
Distal18537 (20)18596 (52)
Dukes stage0.030.001
A4711 (23)4515 (33)
B10226 (25)10345 (44)
C9423 (24)9453 (56)
D503 (6)5036 (72)
Gowth pattern0.650.38
Expansive12825 (20)12864 (50)
Infiltration15233 (22)15284 (55)
Well2310 (43)247 (29)
Moderately19642 (21)195116 (60)
Poorly447 (16)4322 (51)
Mucinous+signet-ring cell375 (14)378 (22)
DNA ploidy0.0020.14
Diploid10732 (30)10753 (49)
Non-diploid11815 (13)11870 (59)
S-phase fraction0.0040.24
<5%5219 (37)5226 (50)
5-10%6516 (25)6531 (48)
>10%728 (11)7244 (61)
p53 expression0.010.07
Negative11733 (28)11757 (49)
Positive11415 (13)11469 (61)
Figure 2
Figure 2 Tumor inflammatory infiltration at the invasive margin (A) and necrosis (B) in relation to survival in patients with colorectal cancer.

There was no significant relationship of inflammatory infiltration in the tumor with the above clinicopathological and other variables (P>0.05, data not shown).

TN in relation to MSS, and clinicopathological and other variables

Among the 300 tumors studied, necrosis was absent in 76 cases, 54 had <10%, 93 had 10-30%, and 77 had >30% of necrosis. The cases with <10% of necrosis were graded as little necrosis and the remainder as abundant necrosis(Figure 1C), based on the similarities of the clinicopathological features.

We did not find a relationship between TN and MSI (χ2 = 0.10, P = 0.75; Table 1). As shown in Table 2, the frequency of TN was increased from Dukes’ stages A to D (χ2 = 17.74, P = 0.001). TN was more frequent in moderately/poorly differentiated tumors, but was the lowest in mucinous/signet-ring cell carcinomas (χ2 = 22.98, P<0.0001). Patients with abundant TN had worse survival than those with little TN in univariate analysis (P = 0.02; Figure 2B). The survival significance was borderline after adjustment for Dukes’ stage and differentiation (P = 0.05, data not shown). There were no associations of TN with other factors including gender, age, tumor location, growth pattern, DNA ploidy, SPF, and p53 expression (P>0.05).

Considering the relationship of inflammatory infiltration with TN, tumors with high infiltration tended to have less necrosis although the relationship did not reach statistical significance (χ2 = 3.62, P = 0.057).


MSI has been shown to be a favorable prognostic factor in colorectal cancer patients, even independent of tumor stage, or radiotherapy and chemotherapy[1,5-8,22,23]. The reason for this evidence has been proposed to be partly due to a stronger host response of immune system in MSI tumors than in MSS ones[1,5]. However, several other studies have reported the lack of association between MSI and survival, either in entire group of colorectal cancer patients[9,10,12] or in the subgroup with stage II colorectal cancers[15,16]. Such non-association was also shown in our previous two studies[13,14] and another Swedish study[11]. The National Cancer Institute Workshop concluded that MSI has not yet been shown conclusively to be an independent predictor of prognosis in colorectal cancer patients[2]. The reason for the non-associ-ation between MSI and survival is unclear. Subgroup vari-ation within populations is unlikely an explanation for this evidence because the similar findings have been reported by various groups among different populations. The relationship between the degree of inflammatory infiltration and microsatellite status in the patients without association of MSI with survival has not been studied. Based on our previous findings of the non-association between MSI and survival[13,14], we observed that a low level of inflammatory infiltration either at the invasive margin or in the tumor was related to MSI phenotype in the present study. Although the statistically significant differences were borderline (P = 0.047 and P = 0.049), the results provided at least an indication that pronounced inflammatory infiltration did not always accompany MSI tumors as proposed[1,3,4].

It is generally accepted that the immune system represents a specific host response to tumors. The survival advantage of pronounced infiltration of all the inflammatory cells, or various subsets of inflammatory cells such as lymphocytes and macrophages, around or within colorectal tumors has been demonstrated[21,24-26]. Anti-tumor effects of infiltrating inflammatory cells may be mediated by cytokine secretion induced by the response of inflammatory cells to tumor stimulation. The expression of two cytokines interleukin-4 (IL-4) and TN factor-α (TNF-α) in colon cancer has been found to be associated with better survival[27]. In the present study, we confirmed the previous findings in which abundant inflammatory infiltration at the invasive margin predicted a favorable prognosis in colorectal cancer patients[6,21,25]. Moreover, we found that the high level of inflammatory infiltration at the invasive margin was associated with earlier Dukes’ stage, better differentiation, DNA diploid, lower SPF, and negative p53 expression. However, the degree of inflammatory infiltration in the tumor was not related to clinicopathological or other factors studied. These results indicated that the degree of inflammatory infiltration at the invasive margin, compared with that in the tumor, was more effective against the tumor.

Clinical results regarding the relationship of microsatellite status in colorectal cancer patients with adjuvant chemotherapy are fairly inconsistent. Some patients with MSI tumor had no survival benefit from chemotherapy[12,22,23] but others had[8].Reviewing all eight previous studies, that showed the non-association between MSI and survival in colorectal cancer patients, two studies were carried out on patients without chemotherapy[9,15], and one study on patients with chemotherapy, in which no survival benefit of chemotherapy was observed in MSI patients although the benefit was found in total group of the patients or microsatellite status patients[12]. Four studies did not provide information about chemotherapy[10,11,13,16]. In our previous study in 438 colorectal cancers with microsatellite status[14], 34 (8%) had received radiotherapy, 15 (3%) chemotherapy, and 7 (2%) both treatments. No information was available for 59 (13%), and the rest, 316 (73%), did not receive any adjuvant or palliative treatments. Statistical analysis showed that neither radiotherapy (P = 0.23) nor chemotherapy (P = 0.84) improved survival of the patients (unpublished data). Taken together, it seems that chemotherapy and radiotherapy were unlikely the reasons behind the non-association between MSI and survival in colorectal cancer patients. Obviously, further studies are needed to evaluate the effect of chemotherapy and radiotherapy on MSI patients.

TN is a common feature of solid tumors and caused by ischemia due to rapid tumor growth. The degree of TN reflected the level of intra-tumor hypoxia, and increased hypoxia has been associated with high metastatic potential. TN has been previously demonstrated as an indicator of poor prognosis in several types of cancers, including colorectal cancer[18]. In the present study, abundant TN was related to poor survival, advanced Dukes’ stage, and moderate/poor differentiation. Unexpectedly, the lowest frequency of TN was seen in mucinous carcinomas/signet-ring cell carcinoma that had more aggressive behaviors. This was probably due to mucinous carcinomas that had a substantial amount of mucin and were less likely to have gland formation. In addition, mucinous carcinomas usually grow slowly, which requires less blood supply, and were therefore lack of TN. Recently, Greenson et al[17], reported that the absence of TN was related to MSI-H in colorectal cancers. However, we could not prove this in the present study. The varying results may be due to the different characteristics of the patients and pathological features of the tumors. Besides, it may partly depend on the classification of MSI: they divided MSS into MSI-H, and microsatellite status which included MSI-low cases. Therefore, they had a lower frequency of MSI (9.8%) than ours (14%).

In conclusion, the lack of association of pronounced inflammatory infiltration with MSI might be one of the explanations for non-association between MSI and survival in colorectal cancers.


Science Editor Li WZ Language Editor Elsevier HK

1.  Gryfe R, Gallinger S. Microsatellite instability, mismatch repair deficiency, and colorectal cancer. Surgery. 2001;130:17-20.  [PubMed]  [DOI]
2.  Boland CR, Thibodeau SN, Hamilton SR, Sidransky D, Eshleman JR, Burt RW, Meltzer SJ, Rodriguez-Bigas MA, Fodde R, Ranzani GN. A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res. 1998;58:5248-5257.  [PubMed]  [DOI]
3.  Smyrk TC, Watson P, Kaul K, Lynch HT. Tumor-infiltrating lymphocytes are a marker for microsatellite instability in colorectal carcinoma. Cancer. 2001;91:2417-2422.  [PubMed]  [DOI]
4.  Alexander J, Watanabe T, Wu TT, Rashid A, Li S, Hamilton SR. Histopathological identification of colon cancer with microsatellite instability. Am J Pathol. 2001;158:527-535.  [PubMed]  [DOI]
5.  Gryfe R, Kim H, Hsieh ET, Aronson MD, Holowaty EJ, Bull SB, Redston M, Gallinger S. Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med. 2000;342:69-77.  [PubMed]  [DOI]
6.  Gafà R, Maestri I, Matteuzzi M, Santini A, Ferretti S, Cavazzini L, Lanza G. Sporadic colorectal adenocarcinomas with high-frequency microsatellite instability. Cancer. 2000;89:2025-2037.  [PubMed]  [DOI]
7.  Samowitz WS, Curtin K, Ma KN, Schaffer D, Coleman LW, Leppert M, Slattery ML. Microsatellite instability in sporadic colon cancer is associated with an improved prognosis at the population level. Cancer Epidemiol Biomarkers Prev. 2001;10:917-923.  [PubMed]  [DOI]
8.  Elsaleh H, Joseph D, Grieu F, Zeps N, Spry N, Iacopetta B. Association of tumour site and sex with survival benefit from adjuvant chemotherapy in colorectal cancer. Lancet. 2000;355:1745-1750.  [PubMed]  [DOI]
9.  Feeley KM, Fullard JF, Heneghan MA, Smith T, Maher M, Murphy RP, O'Gorman TA. Microsatellite instability in sporadic colorectal carcinoma is not an indicator of prognosis. J Pathol. 1999;188:14-17.  [PubMed]  [DOI]
10.  Ko JM, Cheung MH, Kwan MW, Wong CM, Lau KW, Tang CM, Lung ML. Genomic instability and alterations in Apc, Mcc and Dcc in Hong Kong patients with colorectal carcinoma. Int J Cancer. 1999;84:404-409.  [PubMed]  [DOI]
11.  Salahshor S, Kressner U, Fischer H, Lindmark G, Glimelius B, Påhlman L, Lindblom A. Microsatellite instability in sporadic colorectal cancer is not an independent prognostic factor. Br J Cancer. 1999;81:190-193.  [PubMed]  [DOI]
12.  Carethers JM, Smith EJ, Behling CA, Nguyen L, Tajima A, Doctolero RT, Cabrera BL, Goel A, Arnold CA, Miyai K. Use of 5-fluorouracil and survival in patients with microsatellite-unstable colorectal cancer. Gastroenterology. 2004;126:394-401.  [PubMed]  [DOI]
13.  Evertson S, Wallin A, Arbman G, Rütten S, Emterling A, Zhang H, Sun XF. Microsatellite instability and MBD4 mutation in unselected colorectal cancer. Anticancer Res. 2003;23:3569-3574.  [PubMed]  [DOI]
14.  Emterling A, Wallin A, Arbman G, Sun XF. Clinicopathological significance of microsatellite instability and mutated RIZ in colorectal cancer. Ann Oncol. 2004;15:242-246.  [PubMed]  [DOI]
15.  Curran B, Lenehan K, Mulcahy H, Tighe O, Bennett MA, Kay EW, O'Donoghue DP, Leader M, Croke DT. Replication error phenotype, clinicopathological variables, and patient outcome in Dukes' B stage II (T3,N0,M0) colorectal cancer. Gut. 2000;46:200-204.  [PubMed]  [DOI]
16.  Wang C, van Rijnsoever M, Grieu F, Bydder S, Elsaleh H, Joseph D, Harvey J, Iacopetta B. Prognostic significance of microsatellite instability and Ki-ras mutation type in stage II colorectal cancer. Oncology. 2003;64:259-265.  [PubMed]  [DOI]
17.  Greenson JK, Bonner JD, Ben-Yzhak O, Cohen HI, Miselevich I, Resnick MB, Trougouboff P, Tomsho LD, Kim E, Low M. Phenotype of microsatellite unstable colorectal carcinomas: Well-differentiated and focally mucinous tumors and the absence of dirty necrosis correlate with microsatellite instability. Am J Surg Pathol. 2003;27:563-570.  [PubMed]  [DOI]
18.  Swinson DE, Jones JL, Richardson D, Cox G, Edwards JG, O'Byrne KJ. Tumour necrosis is an independent prognostic marker in non-small cell lung cancer: correlation with biological variables. Lung Cancer. 2002;37:235-240.  [PubMed]  [DOI]
19.  Sun XF, Carstensen JM, Stål O, Zhang H, Nilsson E, Sjödahl R, Nordenskjöld B. Prognostic significance of p53 expression in relation to DNA ploidy in colorectal adenocarcinoma. Virchows Arch A Pathol Anat Histopathol. 1993;423:443-448.  [PubMed]  [DOI]
20.  Sun XF, Carstensen JM, Zhang H, Stål O, Wingren S, Hatschek T, Nordenskjöld B. Prognostic significance of cytoplasmic p53 oncoprotein in colorectal adenocarcinoma. Lancet. 1992;340:1369-1373.  [PubMed]  [DOI]
21.  Nagtegaal ID, Marijnen CA, Kranenbarg EK, Mulder-Stapel A, Hermans J, van de Velde CJ, van Krieken JH. Local and distant recurrences in rectal cancer patients are predicted by the nonspecific immune response; specific immune response has only a systemic effect--a histopathological and immunohistochemical study. BMC Cancer. 2001;1:7.  [PubMed]  [DOI]
22.  Colombino M, Cossu A, Manca A, Dedola MF, Giordano M, Scintu F, Curci A, Avallone A, Comella G, Amoruso M. Prevalence and prognostic role of microsatellite instability in patients with rectal carcinoma. Ann Oncol. 2002;13:1447-1453.  [PubMed]  [DOI]
23.  Halling KC, French AJ, McDonnell SK, Burgart LJ, Schaid DJ, Peterson BJ, Moon-Tasson L, Mahoney MR, Sargent DJ, O'Connell MJ. Microsatellite instability and 8p allelic imbalance in stage B2 and C colorectal cancers. J Natl Cancer Inst. 1999;91:1295-1303.  [PubMed]  [DOI]
24.  Ropponen KM, Eskelinen MJ, Lipponen PK, Alhava E, Kosma VM. Prognostic value of tumour-infiltrating lymphocytes (TILs) in colorectal cancer. J Pathol. 1997;182:318-324.  [PubMed]  [DOI]
25.  Adachi Y, Mori M, Kuroiwa S, Sugimachi K, Enjoji M. Histopathologic evaluation of survival time in patients with colorectal carcinoma. J Surg Oncol. 1989;42:219-224.  [PubMed]  [DOI]
26.  Funada Y, Noguchi T, Kikuchi R, Takeno S, Uchida Y, Gabbert HE. Prognostic significance of CD8+ T cell and macrophage peritumoral infiltration in colorectal cancer. Oncol Rep. 2003;10:309-313.  [PubMed]  [DOI]
27.  Barth RJ, Camp BJ, Martuscello TA, Dain BJ, Memoli VA. The cytokine microenvironment of human colon carcinoma. Lymphocyte expression of tumor necrosis factor-alpha and interleukin-4 predicts improved survival. Cancer. 1996;78:1168-1178.  [PubMed]  [DOI]