Department of Pathology, SGPGIMS, Lucknow, India.

On-line Seminars and Tutorials

Prognostic Factors PredictingOutcome in Colorectal Cancers

The 5-year crude survival rate after curative resection for colorectalcancer ranges between 40 and 60% in most studies.¹ Local recurrencesand/or regional lymph node metastases occur in over 90% of failure cases.²The prognosis of colorectal cancers is affected by a number of factorswhich can be broadly classified into those related to the tumor stage,clinical variables, pathologic features, and oncogenetic, molecular, andimmunologic variables. The stage and factors directly or indirectly affectingthe stage of the primary tumor remain the most important prognostic factors.

Factors related to the tumorstage

Tumor stage. Starting with the Dukes’classification, the categories of all major staging systems remain powerfulpredictors of prognosis. The 5-year survival rates are ~90% for Dukes’A, 50-65% for Dukes’ B and 15-25% for Dukes’ C.³ The factorsnot utilized in the staging scheme used assume independent importance,specially in stages B and C. Thus, the number of lymph nodes involved becomeadditional factors for Dukes' and Astler-Coller systems.

Adjacent organ involvement. A significantdecrease in survival is noted in colonic carcinoma patients with adjacentorgan involvement (B3 of modified Astler-Coller),⁴ if the involvementis verified histopathologically.

Residual tumor. The chances of residualtumor and local recurrence are reported to increase if the distance betweentumor and resection end is less than 1 mm. The lateral (radial) marginis important in rectal carcinomas if the tumor is located below the levelof peritoneal reflection and its involvement significantly increases thechances of local failure.⁵ Some staging systems therefore classifyall patients with residual tumor as stage D.

Nodal and distant metastases. Althoughthe nodal involvement is a part of all staging systems, the emphasis onlocation and extent of nodal involvement is not equally addressed. Somestudies report involvement of apical node as a particularly ominous featurewhile in others it appears to be less important than the number of involvednodes in multivariate analyses.^6,7 The survival is markedlyreduced for patients with distant metastases.

Clinical Prognostic Variables

Age. Many studies have reportedpoorer prognosis in patients of colorectal cancer who are less than 40years old. Majority of these present with higher stages and are attributedto delayed diagnosis, higher histological grade and greater numbers ofmucinous and signet ring cell tumors in the young.⁸ However,when stage-adjusted survival is analysed, there is no significant differencein relative prognosis for the younger age group.⁹

Gender. Many studies report morefavorable prognosis for females as with other malignancies.¹⁰

Obstruction and Perforation. Obstructionand perforation significantly worsen the survival and some studies claimthem as important determinants of prognosis independent of Dukes’ stage.¹¹There are essentially no cures once free perforation into the peritonealcavity occurs.¹²

Location of the primary tumor. Thefive year survival for rectal cancers at or below the peritoneal reflectionis lower than for those above the reflection and for colon cancers.¹³There are conflicting reports about the differences in prognosis betweenright and left colon primaries.

Primary tumor configuration. Ingeneral, polypoid and exophytic cancers have a better prognosis than flatand ulcerated lesions. The former show a lower frequency of bowel wallpenetration and nodal metastasis, presenting at a comparatively lower stages.¹²

Duration of symptoms. Asymptomaticpatients have a better 5-year survival (71%) than symptomatic colorectalcancer patients (49%),¹⁴ possibly because they are detectedat an earlier stage using screening methods. Patients with symptoms formore than six months have a higher 5-year survival rate.¹⁵ Thismay reflect the slow growth rate of tumors in these patients. However,duration of symptoms have no effect on survival in stage-controlled multivariateanalysis.¹⁶

Rectal bleeding. Association ofhemorrhage or rectal bleeding with colorectal cancer is associated witha better prognosis possibly because it leads to earlier diagnosis and intervention.The significance of rectal bleeding is not seen in multivariate analysis.¹⁶

Peri-operative blood transfusion.Peri-operative blood transfusion is reported to be associated with increasedrecurrence rate and may be associated with reduced overall survival.¹⁷The factors which necessitate transfusions may be actually responsible.

Pathologic Prognostic factors

Pathological stage and other factorslike resection margins determined on pathological examination, are themost important prognostic determinants.

Histopathological tumor type. Themajority of colorectal cancers are adenocarcinomas. Some subtypes, likemucinous or colloid carcinoma (defined in the recent TNM classificationas tumors with mucinous areas comprising more than 50% of tumor), signetring cell carcinoma, and undifferentiated carcinoma, have a worse prognosisthan typical adenocarcinoma.¹⁰

Histopathological tumor grade (degree of differentiation).In some multivariate analyses, the microscopic grade is independently predictiveof survival, higher grades being associated with increasing wall penetrationand more chances of nodal and distant metastasis, and worse prognosis.¹⁰

Lymphatic vessel invasion. Lymphaticvessel invasion has been found to be an independent prognostic factor byproportional hazard analysis.¹⁸ The incidence of lymphatic vesselinvasion increases with stage and grade, and presence and extent of lymphnode metastasis.

Blood vessel (venous) invasion.The invasion of extramural veins (in pericolonic or perirectal fat) isassociated with a significantly worse prognosis,¹⁹ and has beenrecognised as an independent prognostic marker in some studies with multivariateanalysis.¹⁶

Perineurial invasion. The presenceof perineurial invasion is associated with higher rate of recurrence andreduced survival, and has been recognised as an independent prognosticfactor.²⁰

Peritumoral lymphocytic infiltration.Marked (Crohn’s disease like) infiltration in and around the tumor is associatedwith a better prognosis, to the point that this feature has been incorporatedin the staging scheme of Jass and colleagues.²¹

Eosinophilic and Mast cell infiltration.The presence of four or more mast cells per 30 oil immersion fields hasbeen found to correlate with a lower overall survival and was an independentprognostic factor.²² The eosinophilic infiltration remains controversialwith claims for improved and reduced survival in different studies.

Pattern of lymph node reaction.Sinus histiocytosis and paracortical immunoblastic activity in the lymphnodes draining the tumor have been reported to correlate with improvedsurvival.²³

Other factors

DNA ploidy and other genetic and karyotypicmarkers. Several studies have shown a correlation between aneuploidyand risk of recurrence and survival, specially in rectal tumors. In general,is more aneuploidy is observed with increasing stage and its prognosticvalue independent of stage remains controversial.²⁴ The relationshipof S-phase fraction with survival is also controversial.²⁵ Allelicloss of chromosome 18q has been shown to have a strong negative prognosticsignificance in colorectal carcinomas.²⁶ It has been suggestedthat DNA index provides better prognostic information than ploidy.²⁷

Oncogene and Anti-oncogene expression.K-ras mutations at certain sites and over-expression of the ras p21 proteinare more common in patients with recurrent disease.²⁸ High expressionof p53 correlates with reduced survival and has been shown to be an independentpredictor in some studies.²⁹ Mutations in DCC (Deleted in ColonCarcinoma) gene,³⁰ nm23 gene,³¹ and disordered allelotype³²have also been reported to correlate with prognosis.

Tumor related antigens and other immunologicalfactors. Higher serum levels of CEA (Carcinoembryonic Antigen)have been correlated with increased risk for recurrence and poor survival,particularly in higher stages, and in colon cancer.³³ Colorectalcarcinomas expressing mucin-associated antigens (sialosyl-Tn and sialylLewis^x) have been found to run a more aggressive clinical course.^34,35Tumors with strong HLA-DR expression show better survival within same stages.³⁶Higher expression of growth factors and receptors like epidermal growthfactor receptor have been shown to have prognostic value.³⁷

Additional factors. Many other factorshave been described as having a correlation with poor prognosis. Some ofthese include expression of sucrase-isomaltase,³⁸ autocrine-motilityfactor receptor³⁹ and proliferating cell nuclear antigen,⁴⁰reduced collagen type IV in tumor matrix,⁴¹ irregularity ofnuclear shape on morphometry,⁴² reduced ornithine decarboxylaseactivity,⁴³ expression of cathepsin B,⁴⁴ expressionof E-cadherin and alpha-catenin,⁴⁵ tumor angiogensis and intratumoralmicrovessel density,⁴⁶ overexpression of CD44,⁴⁷and telomerase activity.⁴⁸ More data are needed for substantiationof importance of these and other factors before they are added to the listof established prognostic factors for colorectal cancer.

References

Patel SC, Tovee EB, Langer B. Twenty-five years of experience with radicalsurgical treatment of carcinoma of the extraperitoneal rectum. Surgery1977;82:460.

Russel AH, Pelton J, Reheis CE, Wisbeck WM, Tong D, Dawson LE. Adenocarcinomaof the colon: An autopsy study with implications for new therapeutic strategies.Cancer 1985;56:1446.

Rich T, Gunderson LL, Lew R, Galdibini JJ, Cohen AM, Donaldson G. Patternsof recurrence of rectal cancer after potentially curative surgery. Cancer1983;52:1317.

Nathanson SD, Tilley BC, Schultz L, Smith RF. Perioperative allogeneicblood transfusions: Survival in patients with resected carcinomas of thecolon and rectum. Arch Surg 1985; 120:734.

Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinomadue to inadequate surgical resection: Histopathological study of lateraltumor spread and surgical excision. Lancet 1986;1:996.

Newland RC, Dent OF, Lyttle MN, Chapuis PH, Bokey EL. Pathologic determinantsof survival associated with colorectal cancer with lymphnode metastases.A multivariate analysis of 579 patients. Cancer 1994;73:2076.

Wolmark N, Fisher B, Wieand HS. The prognostic value of the modificationsof the Dukes’ C class of colorectal cancer. Ann Surg 1986;203:115.

Rao BN, Pratt CB, Fleming ID, Dilawari RA, Green AA, Austin BA. Colon carcinomain children and adolescents: A review of 30 cases. Cancer 1985;55:1322.

Umpleby HC, Williamson RCN. Carcinoma of the large bowel in the first fourdecades. Br J Surg 1984;71:272.

Griffin MP, Bergstralh EJ, Coffey RJ, Beart RW Jr, Melton LJ III. Predictorsof survival after curative resection of carcinoma of the colon and rectum.Cancer 1987;60:2318.

Wolmark N, Wieand HS, Rockett HE et al. The prognostic significance oftumor and location and bowel obstruction in Dukes’ B and C colorectal cancer:Findings from the NSABP clinical trials. Ann Surg 1983;198:743.

Steinberg SM, Barkin JS, Kaplan RS, Stablein DM. Prognostic indicatorsof colon tumors: The Gastrointestinal Tumor Study Group experience. Cancer1986;57:1866.

Freedman LS, Macaskill P, Smith AN. Multivariate analysis of prognosticfactors for operable rectal cancer. Lancet 1984;2:733.

Beahrs OH, Sanfelippo PM. Factors in the prognosis of colon and rectalcancer. Cancer 1971;28:213.

Pescatori M, Maria G, Beltrani B, Mattana C. Site, emergency, and durationof symptoms in the prognosis of colorectal cancer. Dis Colon Rectum 1982;25:33.

Chapuis PH, Dent OF, Fisher R et al. A multivariate analysis of clinicaland pathological variables in prognosis after resection of large bowelcancer. Br J Surg 1985;72:698.

Tartter PI. The association of perioperative blood transfusion with colorectalcancer recurrence. Ann Surg 1992;216:633.

Minsky BD, Mies C, Rich TA, Recht A. Lymphatic vessel invasion is an independentprognostic factor for survival in colorectal cancer. Int J Rad Oncol BiolPhys 1989;17:311.

Minsky BD, Mies C, Recht A, et al. Resectable adenocarcinoma of the rectosigmoidand rectum. II. The influence of blood vessel invasion. Cancer 1988;61:1417.

Bognel C, Rekacewicz C, Mankarios H, et al. Prognostic value of neuralinvasion in rectal carcinoma: A multivariate analysis on 339 patients withcurative resection. Eur J Cancer 1995;31A:894.

Jass JR, Love SB, Northover JM. A new prognostic classification of rectalcancer. Lancet 1987;1:1303.

Fisher ER, Pale SM, Rockette H, Jones J, Kaplan R, Fisher B. Prognosticsignificance of eosinophils and mast cells in rectal cancer: Findings fromthe National Surgical Adjuvant Breast Bowel Project (Protocol R-01). HumanPath 1989;20:159.

Patt DJ, Brynes RK, Vardiman JW, Coppleson LW. Mesocolic lymphnode histologyis an important prognostic indicator for patients with carcinoma of thesigmoid colon. An imunomorphologic study. Cancer 1975;35:1388.

Visscher DW, Zarbo RJ, Ma CK, Sakr WA, Crissman JD. Flow cytometric DNAand clinicopathologic analysis of Dukes’ A & B colonic adenocarcinomas.A retrospective study. Mod Pathol 1990;3:709.

Bauer KD, Lincoln ST, Vera-Roman JM, et al. Prognostic implications ofproliferative activity and DNA aneuploidy in colonic adenocarcinomas. LabInvest 1987;57:329.

Jen J, Kim H, Piantadosi S, et al. Allelic loss of chromosome 18q and prognosisin colorectal cancer. N Engl J Med 1994;331:213.

Tomada H, Kakeji T. Prognostic significance of flow cytometric analysisof DNA content in colorectal cancer. J Surg Oncol 1993;53:144.

Benhattar J, Losi L, Chaubert P, Givel JC, Costa J. Prognostic significanceof K-ras mutations in colorectal carcinoma. Gastroenterology 1993;104:1044.

Zeng ZS, Sarkis AS, Zhang ZF, et al. p53 nuclear overexpression. An independentpredictor of survival in lymph node-positive colorectal patients. J ClinOncol 1994;12:2043.

Saito M, Yamaguchi A, Goi T, et al. Expression of DCC protein in colorectaltumors and its relationship to tumor progression and metastasis. Oncology1999;56:134.

Tannapfel AK, Katalinic A. Expression of nm23-H1 predicts lymph node involvementin colorectal carcinoma. Dis Colon Rectum 1995;38:651.

Arai T, Akiyama Y, Yamamura A, et al. Allelotype analysis of early colorectalcancers with lymph node metastasis. Int J Cancer 1998;79:418.

Webb A, Scott-Mackie P, Cunningham D, et al. The prognostic value of CEA,bHCG, AFP, CA125, CA19-9, and C-erb B-2, b HCGimmunohistochemistry in advanced colorectal cancer. Ann Oncol 1995;6:581.

Itzkowitz SH, Bloom EJ, Kokal WA, Modin G, Hakomori S, Kim YS. Sialosyl-Tn.A novel mucin antigen associated with prognosis in colorectal cancer patients.Cancer 1990;66:1960.

Nakamori S, Kameyama M, Imaoka S, et al. Increased expression of sialylLewis^x antigen correlates with poor survival in patients withcolorectal carcinomas. Clinicopathological and imunohistochemical study.Cancer Res 1993;53:3632.

Anderson SN, Rognum TO, Lund E, Meling GI, Hauge S. Strong HLA-DR expressionin large bowel carcinomas is associated with good prognosis. Br J Cancer1993;68:80.

Steele RJ, Kelly P, Ellul B, Eremin O. Epidermal growth factor receptorexpression in colorectal cancer. Br J Surg 1990;77:1352.

Jessup JM, Lavin. Sucrase-isomaltase is an independent prognostic markerfor colorectal carcinoma. Dis Colon Rectum 1995;38:1257.

Nakamori S, Watanabe H, Kameyama M, et al. Expression of autocrine motilityfactor receptor in colorectal cancer as a predictor for disease recurrence.Cancer 1994; 74:1855.

Al-Sheneber IF, Shibata HR. Prognostic significance of proliferating cellnuclear antigen expression in colorectal cancer. Cancer 1993;71:1954.

Havenith MG, Arends JW, Simon R, Volovics A, Wiggers T, Bosman FT. TypeIV collagen immunoreactivity in in colorectal cancer: prognostic valueof basement membrane deposition. Cancer 1988;62:2207.

Mitmaker B, Begin LR, Gordon PH. Nuclear shape as a prognostic discriminantin colorectal carcinoma. Dis Colon Rectum. 1991;34:249.

Matsubara NH, Gilmour SK. Association between high levels of ornithinedecarboxylase activity and favorable prognosis in human colorectal carcinoma.Clin Cancer Res 1995;1:665.

Kos J, Nielsen HJ, Krasovec M, et al. Prognostic value of cathepsin B andcarcinoembryonic antigen in sera of patients with colorectal cancer. ClinCancer Res 1998;4:1511.

Gofuku J, Shizaki H, Tsujinaka T, et al. Expression of E-cadherin and alpha-cateninin patients with colorectal carcinoma. Correlation with cancer invasionand metastasis. Am J Clin Pathol 1999;111:29.

Vermeulen PB, Van den Enyden GG, Huger P, et al. Prospective study of intratumoralmicrovessel density, p53 expression and survival in colorectal cancer.Br J Cancer 1999;79:316.

Bhatavadekar JM, Patel DD, Chikhlikar PR, et al. Overexpression of CD44:a useful independent predictor of prognosis in patients with colorectalcarcinoma. Ann Surg Oncol 1998;5:495.

Okayasu I, Mitomi H, Yamashita K, et al. Telomerase activity significantlycorrelates with cell differentiation, proliferation and lymph node metastasisin colorectal carcinomas. J Cancer Res Clin Oncol 1998;124:444.

Pathology Home | Back to On-line Tutorials