Main

All colorectal carcinomas arising in the setting of Hereditary Nonpolyposis Colorectal Cancer (HNPCC) demonstrate high-level microsatellite instability (MSI-H), as do 15–20% of sporadic colorectal cancers.1, 2, 3 The phenomenon reflects accumulated errors in DNA microsatellite repeat sequences following inactivation of a DNA mismatch repair enzyme via mutation (in the hereditary setting) or hypermethylation of the promoter region (in sporadic cancers).2, 3 Whether hereditary or sporadic, MSI-H tumors characteristically involve the right side of the colon and show medullary, mucinous or signet ring histology.4, 5, 6, 7 Host lymphoid response in the form of tumor infiltrating lymphocytes or peritumoral Crohn's-like infiltrate is usually prominent.7, 8 In most studies, patients with MSI-H colorectal cancer have better prognosis than do those with microsatellite stable carcinoma.9, 10, 11, 12, 13, 14, 15, 16

Mucinous carcinoma is a histologic subtype constituting 11–15% of colorectal cancer.17, 18, 19, 20, 21 Mucinous histology has been shown to be an independent adverse prognostic factor in some studies,19, 21, 22, 23 but not in others.17, 18, 20, 24, 25 The current consensus of the College of American Pathologists (CAP) and American Joint Committee on Cancer is that mucinous differentiation is not proven as a statistically significant prognostic factor independent of histologic grade.26, 27 The CAP has recommended that ‘tumor type (like mucinous carcinoma) should be correlated with outcome after adjustment for MSI status in statistically robust studies with multivariate analysis, in order to definitively determine its prognostic significance’.27 We studied associations between MSI status, clinicopathologic features and survival in a consecutive series of 248 resected mucinous carcinomas.

Methods

The study population comprised 248 patients who had a mucinous carcinoma of the colorectum resected during the years 1984–2001 at the Mayo Clinic. Age, gender, tumor size and site were obtained from the pathology report. Tumors in the cecum, ascending colon and transverse colon were classified as right-sided and those in the descending colon, sigmoid and rectum were left-sided. Clinical charts were reviewed to obtain information about adjuvant therapy and the presence of underlying conditions such as inflammatory bowel disease, familial adenomatous polyposis and hereditary nonpolyposis colorectal cancer. Survival data were obtained from the Mayo tumor registry; clinical follow-up ranged from 1 month to 16 years (mean 5.8 years). The study was approved by the Institutional Review Board.

Histologic parameters were evaluated on hematoxylin–eosin-stained slides without knowledge of MSI status. Two to five slides were examined in each case (average 3). By definition, all tumors had mucin constituting more than 50% of tumor volume. The presence of additional patterns (medullary, signet ring cell) was noted. Tumors were grades 1–4 depending on the degree of gland formation: >75%, grade 1, 50–75%, grade 2, 25–50%, grade 3, and <25%, grade 4. The growth pattern was recorded as pushing when the advancing front of the tumor was sharply defined, and infiltrating if there was irregular infiltration of tumor cells at the tumor–stroma interface. Tumor-infiltrating lymphocytes (TIL) were counted in 10 high-power fields and expressed as average count per high-power field. Crohn's-like reaction was recorded as positive or negative; it was regarded as positive only when there were two or more lymphoid aggregates per glass slide. The presence of vascular invasion was noted. Depth of tumor invasion and lymph node status were assessed and used in conjunction with clinical information to determine tumor stage according to the Astler–Coller modification of Dukes' staging.

Mouse monoclonal antibodies against hMLH1, hMSH2 and hMSH6 were used for immunohistochemistry. Formalin-fixed paraffin-embedded sections were rehydrated using xylene and graded alcohol washes. After blocking the endogenous peroxidase, heat-induced antigen retrieval was performed with EDTA buffer, pH 8.0 in a steamer for 5 min, and the sections were incubated with primary antibody (hMLH1 1:200, Pharmingen, hMSH2 1:50, Oncogene Sciences and hMSH6 1:500, Transduction Laboratories) for 60 min. Following washing with PBS, sections were incubated with the biotinylated goat anti-mouse secondary antibody (1:100) for 30 min. Sections were incubated with avidin–biotin complex (ABC elite kit, Vector laboratories, Burlingame, CA, USA) for 20 min, developed with diacetyl benzidine (DAB) for 5–10 min and counterstained with hematoxylin. The adjacent normal colonic mucosa and stromal cells with intact mismatch repair function served as the internal control. Immunohistochemistry results were classified as positive or negative. Any nuclear staining was considered a positive result. Tumors with no nuclear staining for hMLH1, hMLH2 and/or hMLH6 were classified as MSI-H. All other cases were regarded as microsatellite stable (MSS).

The relationship between MSI and various clinicopathologic features was tested by χ2-test for discrete variables, Mantel–Hanzel test for categorical variables and rank sum test for continuous variables. In the survival analysis, the starting point for the survival time was the date of surgery. Survival curves were calculated using the Kaplan–Meier method and statistical significance between curves was tested by the log–rank test. Univariate analysis was performed using multivariate Cox proportional hazards model; the hazard ratio and its 95% confidence interval (CI) were assessed for each factor. P-values of <0.05 were considered statistically significant.

Results

In all, 72 (29.3%) mucinous carcinomas were designated MSI-H on the basis of loss of staining for hMLH1 (n=69) or hMSH2 (n=3). Staining for hMSH6 was lost in all the three cases lacking hMLH2, but not in any others. Four mucinous carcinomas arose in the setting of inflammatory bowel disease. None of the 248 patients carried a diagnosis of a hereditary syndrome as judged by the review of patient charts.

MSI-H cancers were more likely to be right-sided, low-stage tumors with Crohn's-like reaction, numerous TILs and mixed histologic patterns (Table 1). There was no correlation between MSI status and sex, tumor size, grade, growth pattern or vascular invasion. MSI-H tumors tended to occur at an older age, but the correlation was not statistically significant. There was no significant difference in the number of patients receiving adjuvant therapy for MSI-H cancer vs MSS cancer: 22.2% of MSI-H stage B cancers received adjuvant therapy compared to 32.8% MSS stage B cancers (P=0.2), and 68.4% of MSI-H stage C cancers received adjuvant therapy compared to 75.3% of MSS stage C cancers (P=0.4). Adjuvant therapy did not significantly affect survival for localized or metastatic disease (data not shown).

Table 1 Association of clinicopathologic features with MSI in mucinous colorectal cancer
Full size table

Univariate analysis identified low stage, low grade, right-sided location, absence of vascular invasion, high TIL count, presence of Crohn's-like reaction and MSI-H status as having favorable influence on age-independent survival (Table 2 and Figure 1). The survival benefit of MSI-H status was not independent of stage. In multivariate analysis, older age, higher grade, higher stage and absence of Crohn's-like infiltrate were independent predictors of poor outcome (Table 3). Neither TIL count nor MSI status were significant prognostic factors in multivariate analysis.

Table 2 Impact of factors on survival in colorectal mucinous carcinoma, adjusted for age at surgery, as estimated by the Cox model
Full size table
Figure 1
figure 1

Kaplan–Meier curves demonstrating survival by microsatellite instability status: all patients.

Full size image
Table 3 Multivariate Cox proportional hazards model predicting death in mucinous colorectal cancer
Full size table

Discussion

We found that about one-third of mucinous carcinomas (29.3%) are MSI-H. The reported prevalence for all colon cancers is 15–20%. Our result could be a slight underestimation, because we used immunohistochemical staining for mismatch repair enzymes as a marker for MSI. The correlation between immunohistochemistry and microsatellite analysis by PCR is essentially perfect for MSI caused by methylation of the hMLH1 promotor region, but occasional hereditary mutations eliminate mismatch repair function while retaining stainable protein.28, 29, 30 HNPCC is estimated to account for 1–2% of colorectal cancer.2 Our three patients with loss of hMSH2 staining almost certainly carried germline mutations,4, 31 so our yield of 3/248 makes it unlikely that our approach missed many hereditary cancers. It is also conceivable that derangement of other mismatch repair proteins such as PMS2 resulted in microsatellite instability not detected by our stains, but those too are rare events.32

The clinicopathologic profile of MSI-H mucinous carcinoma is similar to that of MSI-H carcinoma in general: right-sided location, low stage at presentation, many TILs and a prominent Crohn's-like infiltrate. We found no correlation between MSI status and growth pattern in mucinous carcinoma. Our findings differ from those of those of the next-largest study of mucinous carcinoma; Messerini et al33 reviewed 50 mucinous carcinomas and found a positive correlation between MSI-H and expanding growth pattern, but no correlation with tumor site, lymphocytes or Crohn's-like infiltrate.

Our patients with MSI-H mucinous carcinoma had significantly better age-adjusted survival than did those with MSS mucinous cancer (P=0.0176). Comparisons within stage, however, failed to show differences in survival for same-stage tumors. Multivariate analysis bore out this observation, demonstrating that MSS was not an independent predictor of prognosis. Instead, patient age, tumor grade, tumor stage and Crohn's-like infiltrate proved to be independent variables affecting outcome. MSI-H mucinous carcinoma was much less likely than MSS mucinous carcinoma to present with lymph node metastases or distant metastases (P<0.001), accounting for its favorable prognosis overall.

The relative inability of MSI-H tumors to develop nodal or distant metastases has been noted in several studies.9, 34 It has been hypothesized that the absence of chromosomal instability and a lower incidence of p53 mutations in MSI-H tumors may be responsible for the delayed progression to metastatic disease.15 For our patients with nodal or distant metastases at presentation, the prognosis for MSI-H mucinous carcinoma was not different from microsatellite stable mucinous carcinoma. This is in contrast to several large studies that have shown a survival benefit for patients with lymph node-positive MSI-H cancers, but those studies did not divide the tumors based on histologic type, and can say nothing about mucinous carcinoma in particular.12, 13, 14 It has also been suggested that adjuvant chemotherapy confers a strongly favorable effect on outcome in node-positive MSI-H colorectal cancer.13, 14 Our patients were equally likely to have received adjuvant therapy regardless of microsatellite instability status, and therapy status did not appear to influence survival for either MSI-H or microsatellite stable cancers. It may be that mucinous carcinoma is biologically distinct from conventional adenocarcinoma in terms of response to adjuvant therapy, but our retrospective study cannot answer that question.

A prominent host lymphoid response has been a consistent finding in MSI-H carcinoma,5, 6, 7, 8, 35, 36 and our study documented an association between MSI-H mucinous carcinoma and both the Crohn's-like infiltrate and increased numbers of TILs. Both features have been correlated with favorable outcome.37, 38, 39, 40, 41, 42 In our series, high numbers of TILs were associated with better survival in univariate but not in multivariate analysis. The Crohn's-like reaction, in contrast, did retain statistical significance in multivariate analysis as a marker of favorable prognosis.

MSI-H mucinous carcinomas are right-sided, low-stage tumors with Crohn's-like reaction and TILs. The prognosis for MSI-H mucinous carcinoma is better than for microsatellite stable mucinous carcinoma, but the difference is related to lower stage at presentation and prominent Crohn's-like infiltrate. These two parameters can be evaluated with routine pathologic techniques. While the assessment of MSI status may add value in the setting of suspected HNPCC, it is not necessary to determine MSI status solely to stratify outcome for mucinous carcinoma.