Clinicopathologic features and prognostic value of DNA mismatch protein expression patterns in stage (cid:0)/(cid:0) colorectal cancer.

MSI CRCs were associated with better prognosis and limited predictive value for adjuvant chemotherapy. However, whether the same is true in Northeastern China is still unclear. The aim of the present study was to evaluate the association of clinicopathologic features and MMR/MSI status determined with immunohistochemistry analysis in Northeast China patients with stage (cid:0)/(cid:0) CRCs. Particularly, we sought to detect the relationship between MMR/MSI status and ecacy of oxaliplatin and uoropyrimidine based adjuvant chemotherapy.


Conclusions
MMR protein appeared distinct associations with tumor staging, serum CEA level and tumor size. And MMR protein was an independent prognostic marker in patients with stage CRC, whereas dMMR CRC patients did not seem to bene t from oxaliplatin combined with uorouracil-based adjuvant chemotherapy.

Take Home Messages
In this work, IHC was used to analyze the MMR protein status of surgical specimens obtained from CRC patients, and the clinicopathological characteristics and prognosis were compared between dMMR and pMMR CRCs.
In addition, our results suggest that the de ciency of MMR protein in tissue of color rectal cancer was associated with a better prognosis in stage patients, but not to bene t from oxaliplatin/5-FU based adjuvant chemotherapy.
This study is of reference for the chemotherapy treatment of dMMR CRCs .

Background
Colorectal cancer (CRC) is the third most common cancer and the third leading cause of cancer-related death in the United States, with approximately 78,500 new cases in males and 67,100 new cases in females, and it is estimated that 27,640 males and 23,380 females died from CRC in 2019 [1].
Currently, it was wildly accepted that chromosomal instability (CIN) and microsatellite instability (MSI) were the major pathways for CRC development [2]. CIN carcinomas, the most commonly pathway, characterized by gene mutations, which leading to chromosomal aneuploidy, de ciency of heterozygosity and chromosomal rearrangements. However, MSI tumors were usually caused by the de ciency of mismatch repair (MMR) genes including MSH1, PMS2, MSH2 and MSH6 with the morbidity of approximately 15% of all sporadic CRCs [3]. Compared with CIN CRCs, dMMR/MSI colorectal malignancy tend to be proximal and histologically exhibit poor differentiation, a mucinous cell type and a better prognosis [4 5]. Therefore, it's crucial to identify the MMR/MSI status of CRCs.
Although there were currently three approaches for detection including immunohistochemistry (IHC), polymerase chain reaction (PCR) based methods and next generation sequencing (NGS), the sensitivity and speci city among the three methods were with high concordance rate (92-97%) [6].
And the sensitivity to detect dMMR/MSI carcinomas with IHC is nearly 95% with the speci city of almost 100% in most reports. [7 8].
It is well known that dMMR/MSI is associated with a better prognosis in colorectal malignancy. However, whether MSI CRCs can bene t from postoperative adjuvant chemotherapy remains controversial. Multiple previous studies had demonstrated that 5-FU-based adjuvant chemotherapy did not prolong 5-year overall survival (OS) in dMMR/MSI colorectal carcinomas [9 10]. Recently, several researches considered that the addiction of oxaliplatin to uoropyrimidine adjuvant chemotherapy can signi cantly improve OS and DFS compared to uoropyrimidine alone treatment in dMMR/MSI CRCs [11 12]. Therefore, further study is needed to verify the predictive signi cance of MMR/MSI status for adjuvant chemotherapy, especially in Northeast China. After all, few relevant reports were conducted here.
The aim of the present research was to evaluate the association of clinicopathologic features and MMR/MSI status determined with immunohistochemistry analysis in Northeast China patients with stage / CRCs. Particularly, we sought to detect the relationship between MMR/MSI status and e cacy of oxaliplatin and uoropyrimidine based adjuvant chemotherapy.

Materials And Method Patients and materials
The approval has been obtained by the ethics committee of Jilin University Second hospital before the study proceeded, and we conducted the research complying with Helsinki Declaration of World Medicine Association. The detection of the four MMR proteins with immunohistochemistry has been routinely performed since 2016 in our institution, therefore, colorectal carcinoma patients diagnosed from December 1 st , 2016 to December 1 st , 2018 were enrolled, eventually, 476 patients were considered eligible after rigorous screening. The clinical data of the eligible CRCs (such as gender, age etc.) was obtained from medical records. The pathologic data (such as T category, pathologic N category, differentiation, tumor pathologic type and vascular invasion) was collected from pathological examination result, and IHC was used to evaluated MMR status in CRC patients. The overall survival (OS) of the CRCs was de ned as the time from radical surgical operation to death. There are two main approaches to obtain the follow-up information, outpatient clinic system and telephone questionnaire respectively. Inspired by Lee' article, we de ned the proximal colon as cecum, ascending colon or transverse colon [13]. Pathological stage (TNM) depended on depth of in ltration, lymphatic metastasis and distant metastasis in accordance with the American Joint Committee on Cancer (AJCC) cancer staging manual. Adjuvant chemotherapy was performed postoperatively was determined by the TNM stage and the patient's willingness with the regimen of oxaliplatin combined with capecitabine.

IHC
The immunohistochemistry analyses for surgical specimens obtained from CRC patients who underwent radical resection were performed in Department of Pathology, Second A liated Hospital of Jilin University. The pre-programmed autostainer (YZB/USA 2016-2012) was used to perform IHC on para nembedded tissues. Here are the brief steps. 2-μm-thinck formalin-xed and para n-embedded tumor tissue sections were heating for two hours at 70 •C.
After that, we used a PT link machine(California, America) to perform HIER(heat induced epitope retrieval) after the sections were depara nized and rehydrated. The slides were incubated after adding primary antibodies for PMS2 (clone: ZA-0542; 1:1; Wuxi), MLH1(clone: ZM-0154; 1:1; Wuxi), MSH2(clone: ZA-0622; 1:1; Wuxi), MSH6(clone: G24072; 1:1; Ventana). Finally, hematoxylin was used to counterstain the slides, and then sealed with neutral balsam. In all the runs, the negative and positive controls were existed. The tissue sections omitting the primary antibodies were regarded as negative controls, and the tissues which were known to express the proteins were positive controls.

Evaluation of IHC
Two pathologists (Z.Y. Wang and D.W. Huang) were responsible for reviewing the IHC strains of colorectal carcinoma samples. The MMR protein status was considered of expression loss when nuclear straining of carcinoma cells was absent whereas the surrounding stromal cells performed positive nuclear straining. Therefore, pro cient mismatch repair (pMMR) is considered only if all the four MMR proteins expression in tumor tissues, while nuclear staining absence at least one MMR proteins was regarded as de cient mismatch repair (dMMR) (Figure 1).

The inclusion and exclusion criteria
The following inclusion criteria were: 1) Patients were in the 18-75 age ranges. 2) The pathological diagnosis was colorectal carcinoma. 3) TNM / . 4) Radical surgery was applied. The exclusion criteria were showed as follows: 1) History of malignant carcinoma. 2) Poor physical condition(such as severe liver, respiratory tract, cardiovascular or kidney disease). 3) Underwent preoperative neoadjuvant therapy, which was considered likely to affect MMR protein expression [14]. 4) Clinicopathological data cannot be collected accurately.

Statistical analysis
Statistical analyses were performed using SPSS for MAC, version 26.0 (IBM Corporation). Continuous variables were performed using Mann-Whitney U test or t test, while χ2 test or Fisher exact test was used for comparing categorical data. For multivariate analyses, Cox proportional hazards model was applied, and survival curves were created by the Kaplan-Meier method. The distinction was considered statistically signi cant if P values were less than 0.05.

Result
Initially, we enrolled a total of 670 CRC patients who underwent radical surgery between December 1 st , 2016 and December 1 st in the Second A liated Hospital of Jilin University. After that, 194 CRCs were excluded according to the exclusion criteria, of which 39 patients with history of malignant carcinoma, 14 cases in poor physical conditions in poor physical conditions, 28 CRCs undergoing preoperative neoadjuvant therapy, and 113 patients' clinicopathological data cannot be collected accurately ( Figure   2).

MMR protein expression in CRCs
Among the 476 patients, 63(13.2%) CRCs were identi ed as defective expression of MMR protein. The loss of PMS2 was observed in 37 of 476(7.9%) CRCs and was the most frequent, followed by 31(6.6%) with the loos of MLH1, 29(6.2%) with the loss of MSH6, 29(6.2%) with the loss of both MSH1 and PMS2, 16(3.4%) with the isolated loss of MSH2, 11(2.4%) with the loss of both MSH2/MSH6, and 3(0.1%) CRCs with MLH1-/PMS2-/MSH6-. Patients with at least one loss of MLH1 and PMS2 but no MSH2 or MSH6 de ciency were assigned to group type 1, while patients with at least one loss of MSH2 and MSH6 but no MLH1 or PMS2 de ciency were assigned to group type 2.

MMR expression and patient characteristics
The clinicopathological features between pMMR CRCs and dMMR CRCs were shown in Table 1. There was a signi cant difference in tumor size, tumor site, CEA level, pN status, TNM stage, differentiation, and pathological type between the two groups (P<0.05 for all comparisons). Moderate mucinous carcinoma, stage , pN0, and proximal colon cancer were more common in dMMR CRCs compared with pMMR CRCs.
The CEA level was signi cantly higher in pMMR group (4.05 vs. 2.75, P=0.004), whereas bigger tumor size was observed in dMMR group (5.8 vs. 4.7, P= 0.000). No difference in gender, age, BMI, pT status, and vascular invasion was witnessed between patients with dMMR CRC and pMMR CRC. Among the 63 dMMR CRCs, we identi ed 53 type1 and type 2 dMMR CRCs. The clinicopathological features of 29 type 1 and 24 type 2 dMMR CRCs are presented in Table 2. The MSH2/MSH6 protein expression loss was associated with gender, and which was more common in male CRC patients (P=0.011). However, there was no signi cant differences between the two groups among the other clinicopathological features (P>0.05).

MMR expression and clinical outcomes
In the present research, 89 CRCs patients had died (dMMR CRCs, N=82; pMMR CRCs, N=7). To further analyze the association of MMR status and prognosis in patients suffering from colorectal carcinoma, Kaplan-Meier analyses were performed (Figure 3). The survivorship analysis (Kaplan-Meier) showed 86% OS rate in dMMR group and a 68% OS rate in pMMR group after 5 years (P=.004, Kaplan-Meier log-rank). In stage CRC patients, the estimated OS rate for patients with loss of MMR protein was 89% and patients without de ciency was 74% after 5 years, which indicated that the loss expression of MMR protein preformed a more favorable prognosis in stage colorectal carcinoma patients (P=.014, Kaplan-Meier log-rank). However, OS did not differ from the two groups in patients with stage colorectal carcinomas (P=.353, Kaplan-Meier log-rank).
To determine whether dMMR was independent prognostic factor associated with CRC clinical outcomes, Cox proportional hazard model was used to performed univariate and multivariate analysis (  The predictive value of MMR protein for e cacy of chemotherapy Further assessment was performed to analysis the effect of POC in both pMMR and dMMR CRC patients using Kaplan-Meier analyses (Figure 4). Among the 413 pMMR CRCs, the overall 5 years survival rates of patients with POC and without POC were 78.1% and 57.2% respectively(P=.026, Kaplan-Meier log-rank). In the subgroup of 192 stage CRCs, POC didn't seem to make any sense to promote a better prognosis(P=.254, Kaplan-Meier log-rank), whereas did in subgroup of 221 stage CRCs(P=.000, Kaplan-Meier log-rank). However, among 63 dMMR CRCs, POC did not improve the outcome of patients with either stage or .

Discussion
The present study elicited three main ndings. First, in Northeast China, dMMR CRCs were commonly located in the proximal colon, had poorly differentiation histology with mucinous features, and appeared to have bigger tumor size, lower CEA level, and earlier TNM stage. Second, dMMR was an independent prognostic marker with a favorable impact on survival in stage CRCs.
Finally, postoperative adjuvant chemotherapy, based on oxaliplatin capecitabine, cannot prolong the overall 5 years survival in patients with dMMR CRC.
The morbidity of dMMR in the present study was 13.2%, which was similar with Western countries [15]. Several researches comparing the clinicopathological features between dMMR and pMMR CRCs have been published over recent decades. patients with all pathological stage, which may account for the difference. The tumor size of dMMR CRC was considered bigger in the present study, which was coincident with the report of Batur's [17]. The serum CEA level was signi cant lower in dMMR CRC in our study.
As far as we know, no literature has reported the present nding. This result can probably be explained by the fact that serum CEA level was associated with carcinoma pathological stage [18], and the rate of stage CRC was higher in the dMMR group in the present research. In addition, male patients were more frequently seen in type 2 CRCs among 53 dMMR colorectal cancer patients, which may imply a potential ethnic difference in the molecular pathogenesis of CRC.
For the past few years, a great quantity of studies had found that the de ciency of MMR protein for CRC was associated with a favor prognosis [19][20][21]. However, researches evaluating the association between MMR protein and prognosis in CRC patients with stage were quite rare. The prognostic impact of dMMR/MSI was studied in 1254 patients with stage /III colon cancer who participated in the PETACC-3 trial [22]. Among the 1254 patients, 190 patients had dMMR tumors, and of which, and 104 patients of which were diagnosed with stage colorectal cancer. The study revealed that dMMR/MSI was associated with better OS (HR 0.47, 95% CI: 0.31-0.72, p < 0.001), however, the prognostic effect was mainly driven by the bene ts seen in stage disease since no signi cant difference was witnessed in OS between stage CRCs with dMMR and pMMR. The result was almost concurrent with the ndings in the present study.
Why is dMMR/MSI irrelevant to prognosis in stage CRCs? It is considered that the prognostic bene ts from dMMR rely on the immunological reaction associated with dMMR/MSI tumors, which can to increase host anti-tumor immunity to suppress tumor metastasis [23]. However, with disease progression and tumor metastasis, mechanisms of immune evasion develop that enable dMMR/MSI tumors to evade immune surveillance with loss of a prognostic advantage [24].
It's generally believed that uorouracil (5-FU)-based adjuvant chemotherapy does not improve 5-year OS in patients with dMMR tumors, and lack of bene t seems to be similar in both stage and dMMR cancers [9 25] In the present study, IHC was used to detect the MMR status. IHC directly evaluates the MMR protein presence/absence in the tumor cells while PCR-based tests use a set of primers to check for PCR products size differences between normal and tumor tissues. These two approaches are sensitive and speci c with high concordance rate (97%)[6]. Therefore, IHC detection for CRC tumors to evaluate the MMR status was acceptable if PCR-based test was not available. The chemotherapy regimens of oxaliplatin and capecitabine was performed in patients after radical surgery. Capecitabine, an oral uorouracil, was designed to preferentially produce 5-FU at tumor sites. Studies showed that the regimen has fewer side effects with the same e ciency compared with the traditional oxaliplatin combined with 5-FU regimen, and patient compliance was good due to oral route of administration.
The present study had several limitations. First, MMR gene analyses were not performed in the present research, and that may be meaningful for the mechanism. Second, confounding factors were remained in the retrospective study. Third, the number of patients included was still small because the detection of MMR protein using IHC has been routinely performed after December 2016. But 467 patients enrolled in the present study were also acceptable. Fours, there was no differences in the adjuvant chemotherapy regimen among CRC patients. Therefore, the effect of different adjuvant chemotherapy regimens for dMMR CRC patients should be put forward through further study. In addition, it was necessary to carry out a prospective and multi-center study with a large sample size in the future.

Conclusion
In the present study, we found that expression of MMR protein appeared distinct associations with tumor staging, serum CEA level and tumor size. The expression of MMR protein was an independent prognostic marker in patients with stage CRC, whereas dMMR CRC patients did not seem to bene t from oxaliplatin combined with uorouracil-based adjuvant chemotherapy.

Declarations
Availability of data and materials The datasets generated and analyzed during the current study available from the corresponding author on reasonable request. AS and MW conceived the study design. ZZ and DL acquired the data for the study. YG, YY, and RQ analyzed and interpreted the data. ZW read the pathological section. AS drafted the manuscript. YY and ZZ revised the manuscript critically. The authors read and approved the nal manuscript.

Ethics declarations
Ethics approval and consent to participate The institutional review board of The Second Hospital of Jilin University had approved the research. And the ethics approval number was 2021111. The patient informed consent was not necessary because of the retrospective design, which had been con rmed by the local ethic committee.

Consent for publication
Not applicable.

Competing interests
The authors declare that they have no competing interests

Competing interests
The authors declare that they have no competing interests