Association of miR-21 and miR-335 to microsatellite instability and prognosis in stage III colorectal cancer

Abstract

BACKGROUND:

MicroRNAs (miRs) are frequently altered in colorectal cancer (CRC) and can be used as prognostic factors.

OBJECTIVE:

To confirm in stage III CRC patients a reported miR signature that was associated to the presence of metastatic disease. To correlate miR expression with microsatellite instability (MSI) and mutations in RAS and BRAF.

METHODS:

miR-21, miR-135a, miR-206, miR-335 and miR-Let-7a expression was analyzed by RT-qPCR in 150 patients out of the 329 patients used to analyze MSI and RAS and BRAF mutations. Association with disease free survival (DFS) and overall survival (OS) was analyzed. Data was confirmed by a multivariate analysis.

RESULTS:

MiR-21 high expression ( $p=$ 0.034) and miR-335 low expression ( $p=$ 0.0061) were significantly associated with MSI-H. A positive trend ( $p=$ 0.0624) between miR-135a high expression and RAS mutations was found. Lower miR-21 expression levels are associated with DFS (HR $=$ 2.654, 95% CI: 1.066–6.605, $p=$ 0.036) and a trend with OS (HR $=$ 2.419, 95% CI: 0.749–7.815, $p=$ 0.140). MiR-21 high expression significantly improves DFS of the poor prognosis group (T4 or N2) ( $p=$ 0.03).

CONCLUSIONS:

Association of increased expression of miR-21 and better prognosis in the poor prognostic group may be of interest and could be explored in future prospective clinical trials.

Keywords

Colorectal cancer survival microRNAs expression microsatellite instability prognostic biomarkers stage III

1. Introduction

The CRC represents an important health problem, being the third cause of death in developed countries [1]. Around 75% of patients are diagnosed with localized disease and almost half of them are in stage III. 5-year disease-free survival of colon cancer stage III patients, treated only with surgery, varies widely depending on T-N sub-stages, from 79.6% for T1N1a to 13.9% for T4N2b. Adjuvant chemotherapy is generally recommended for all stage III subgroups, but the magnitude of benefit increases with the higher risk of relapse [2]. Apart from TN sub-stage, few data derived from clinical or molecular studies have provided robust enough results to be implemented in clinical practice. In this scenario, most patients with low risk stage III are overtreated with the adjuvant chemotherapy treatment while patients in the very high risk group will relapse after the current recommended oxaliplatin-based adjuvant therapy. Molecular alterations such as high microsatellite instability (MSI-H), RAS or BRAF mutations, correlate with histopathological characteristics and clinical different colorectal cancer types and they have prognostic impact in advanced disease [3, 4]. However, none of these molecular factors have a great prognostic impact in stage III, and therapeutic options are not driven by them.

It is essential in clinical practice to identify prognostic and predictive factors that permit identification of patients with increased risk of relapse, those who benefit from chemotherapy, and those for whom this treatment is unnecessary and could even be harmful given its pronounced toxicity.

MicroRNAs (miRs) regulate gene activity post-transcriptionally by inhibition of protein translation. In cancer, they can function as oncogenes or as tumor suppressors, and miR signatures can serve as promising biomarkers for diagnosis, prognosis, and treatment efficacy monitoring over the disease course [5]. In this context, A miR expression signature containing miR-21, miR-135a, miR-335, miR-206 and miR-let-7a that was associated with the presence of metastatic disease was reported [6]. In this study, we have analyzed the expression of this miR signature in patients with stage III CRC to confirm its correlation with clinical outcome. In the absence of literature data, we explore potential associations with the presence of BRAF and RAS mutations as well as Microsatellite Instability (MSI), in order to correlate miRNA expression signatures with well-known patterns of colorectal cancer subtypes.

2. Materials and methods

2.1 Patients

We analyzed 329 samples out of a total of 472 samples of patients with CRC stage III treated with adjuvant therapy included in a previous study of our group [7]. Tumor samples were taken from the Biobanks of the centers associated to the Cooperative Spanish Group for the Treatment of Digestive Tumors (TTD) and of the Biobank of the Hospital Clinico San Carlos B.0000725 (PT20/00074), integrated in the Spanish National Biobanks Network and they were processed following standard operating procedures with the appropriate approval of the Ethical and Scientific Committees. The study was approved by the Institutional Review Board of the Hospital. All samples collected included the corresponding informed consent following The Code of Ethics of the World Medical Association (Declaration of Helsinki, British Medical Journal 18 July 1964). Unanalyzed samples were excluded from the study due to lack of tumor blocks.

2.2 RNA and DNA extraction

RNA and DNA were extracted from 4–5 cuts (7 $\mu$ m thickness) of FFPE embedded tissue. 250 $\mu$ l of deparaffination solution was used followed by 18 hours of incubation with proteinase K. RNA was extracted using the FFPE RNEasy Kit (Qiagen) using 1250 $\mu$ l of ethanol in order to obtain RNA and miRNA in only one step of purification. DNA was extracted with the QIAamp DNA FFPE Tissue Kit (Qiagen). RNA and DNA quantity were measured by the NanoDrop ND-1000.

2.3 DNA mutational analysis

RAS KRAS (12Val, 12Asp, 13Asp, 12Cys, 12Ser, 12Ala, A146T, K117N, A146V, A183C); NRAS (Q61L, Q61R, Q61K, G12D) and BRAF (600Glu) mutations were detected by End-point Multiplex PCR amplification, followed by hybridization in a low-density microarray, using the CLART ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ CMA KRAS $\cdot$ BRAF system, following the instructions of the company. Thereafter, analysis and interpretation of results were automatically performed by the CAR ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ CLINICAL ARRAY READER (GENOMICA, S.A.U, Madrid Spain).

2.4 Analysis of microsatellite instability

MSI Analysis System, Version 1.2 (Promega) was used to detect microsatellite instability in DNA from FFPE tumor tissue. Following the instructions of the company, PCR products were separated by capillary electrophoresis using an Applied Biosystems ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ 3130 Genetic Analyzer and the output were analyzed with GeneMapper ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ software (Applied Biosystems) to determine MSI status of the samples.

2.5 Analysis of miRNA expression

We analyzed the expression of the miR-21, miR-135a, miR-206, miR-335, miR-let-7a and miR-103 (miR-Control [8]) by RT-qPCR in 150 samples out of 329 initial samples. This selection was based on the quantity ( $>$ 60 ng/ $\mu$ l) and quality (protein and salts ratios close to 2) of the RNA/miRNA obtained. RNA quality was measured using the Bioanalyzer 2000 in combination with the RNA Integrity Number (RIN) software. Samples with a RIN higher than 2.5 were selected (as recommended in [6]). Then, RT was performed using the kit FG, TaqMan(r) microRNA RT Kit following the instructions recommended by the company adding 1 $\mu$ l/sample of Multiscribe Reverse Transcriptase [50 U/ $\mu$ l] at 16 ${}^{\circ}$ C for 30 minutes followed by 30 minutes at 42 ${}^{\circ}$ C and finally 5 minutes at 85 ${}^{\circ}$ C to stop reaction. Then, 15 $\mu$ l of total volume were diluted 6.25-fold to reach 93.75 $\mu$ l.

TaqMan qPCR was carried out using miRs specific primers: miR-21 (ref:000397), miR-135a (ref:000460), miR-206 (ref:000510), miR-335 (ref:000546), miR-Let-7a (ref:000377) and miR-103 (ref:000439) (Thermo Fisher Scientific). The initial volume for each miR was: 1 $\mu$ l for miR-21, miR-103 and miR-let-7a; 3 $\mu$ l for miR-335, 4 $\mu$ l for miR-135a, and 6 $\mu$ l for miR-206 in 20 $\mu$ l total volume. Thus, all the CTs were situated between the 25 and the 40 cycle. For the qPCR, we used the 7500 Fast Real-Time PCR System (Applied Biosystems) associated with the software 7500 vs 2.0.6.

Data was processed using the LinRegPCR software [9] through which the N ${}_{0}$ (starting concentration of each miRNA in the sample) was determined based on their average efficiencies. Finally, N ${}_{0}$ was normalized by the miR-Control (miR-103) [8].

2.6 Cutoff finder

Cutoff points for survival analysis were determined using Cutoff Finder software [10]. Among the five methods available, fit of mixture model procedure was chosen to obtain the cutoff point. This procedure fits two Gaussian distributions to the histogram of the biomarker (in our case, miRNA expression values). Optimal cutoff is established as the point where the probability density functions of both distributions coincide.

2.7 Statistical analysis

Categorical variables were summarized with absolute numbers and the percentage of each group regarding the total population under investigation. Chi-Square-test (or Fisher’s exact test) were used to determine statistical differences in categorical variables distribution between groups. Continuous variables were summarized, and within-group variation was compared by dispersion analysis using pairwise Wilcoxon-Mann-Whitney test. Overall survival and progression free survival was represented using the Kaplan-Meier model. We performed univariate analysis using the Cox proportional hazards regression model to determine the significance of those observed differences. Those variables that presented a significative effect in survival were included in a multiple regression model (along with diagnosis age, due to the well-known impact in prognosis) in order to obtain the effect of each variable in term of hazard ratios (HRs).

All statistical tests were two-tailed, and a $p$ value of $<$ 0.05 was considered statistically significant. Statistical analyses were performed using R v3.6.3 under R-Studio 1.1.383. (R Development Core Team Vienna, Austria; https://www.r-project.org).

3. Results

3.1 Patients

The main characteristics of the study cohorts are shown in Table 1.

Table 1
Patient characteristics ( $N=$ 329)

Characteristics	Parameters	$N$ (%)
Age, years	Median	66
	Range	31–87
Sex	Female	147 (44.68)
	Male	174 (52.89)
	N/A	8 (2.43)
Tumor localization	Colon	292 (88.75)
	Rectum	37 (11.25)
Lymphatic invasion	Yes	101 (30.7)
	No	138 (41.95)
	N/A	90 (27.36)
Vascular invasion	Yes	78 (23.71)
	No	185 (56.23)
	N/A	66 (20.06)
Obstruction/perforation	Yes	63 (19.15)
	No	265 (80.55)
	N/A	1 (0.3)
ECOG PS	0	171 (51.98)
	1	108 (32.83)
	2	15 (4.56)
	N/A	35 (10.64)
Grade of differentiation	Well	108 (32.83)
	Poor	50 (15.2)
	Moderate	131 (39.82)
	N/A	40 (12.16)
T	T1	2 (0.61)
	T2	23 (6.99)
	T3	219 (66.57)
	T4	85 (25.84)
N	N1	221 (67.17)
	N2	108 (32.83)
TNM	IIIA	25 (7.6)
	IIIB	240 (72.95)
	IIIC	64 (19.45)
Prognostic	Good (T1-3/N1)	168 (51.06)
	Poor (T4 or N2)	161 (48.94)
Adjuvant chemotherapy	Yes	314 (95.44)
	No	13 (3.95)
	N/A	2 (0.61)
Treatment	5FU/capecitabine	61 (18.54)
	FOLFOX	136 (41.34)
	XELOX	108 (32.83)
	N/A	11 (3.34)
	No treatment	13 (3.95)
RAS status	Mut	105 (31.91)
	Wt	217 (65.96)
	N/A	7 (2.13)
BRAF status	Mut	18 (5.47)
	Wt	305 (92.71)
	N/A	6 (1.82)
MS status	MSI-H	22 (6.69)
	MSS/MSI-L	307 (93.31)

T, tumor; N, node; TNM, tumor-node-metastasis; ECOG PS, Eastern Cooperative Oncology Group performance status; 5FU, 5-fluorouracil; FOLFOX, 5-fluorouracil plus leucovorin plus oxaliplatin; XELOX, capecitabine plus oxaliplatin; Mut, mutated; Wt, wildtype; N/A, unknown; MS, microsatellite; MSI-H, microsatellite instability High; MSS/MSI-L, microsatellite stability.

3.2 Association between miR expression levels and mutations in RAS, BRAF and MSI

Association between miR-21, miR-135a, miR-206, miR-335 and miR-Let-7a expression levels and mutations in RAS, BRAF and MSI were analyzed. Our results show a significant association of higher expression levels of miR-21 to MSI-H tumors ( $p=$ 0.034). On the other hand, lower expression levels of miR-335 are significantly associated to MSI-H tumors ( $p=$ 0.0061) (Fig. 1). No significant associations were found between miR-135a, miR-206 or miR-Let-7a and MSI-H tumors (data not shown).

Figure 1.

Box plot of the association of miR-21 and miR-335 with MSI and of miR-135a with mutations in RAS. ${}^{*}p<$ 0.05; ${}^{**}p<$ 0.01 white shadow indicates a Kernel density estimation where the wider sections represent a higher probability that the N ${}_{0}$ of the samples will be located there.

Figure 2.

DFS Kaplan-Meier curves according to miR expression. ${}^{*}p<$ 0.05. In all cases, Grey line: N ${}_{0}<$ cutoff point; black line: N ${}_{0}>$ cutoff point. At the bottom of each graph, the number of patients contained in each group that remain throughout the follow up period.

Figure 3.

OS Kaplan-Meier curves according to miR expression. In all cases, Grey line: N ${}_{0}<$ cutoff point; black line: N ${}_{0}>$ cutoff point. At the bottom of each graph, the number of patients contained in each group that remain throughout the follow up period.

Figure 4.

DFS and OS Kaplan-Meier curves according to MS, RAS and BRAF mutations. Grey line: MSS/MSI-L and Wild Type; black line: MSI-H and Mutations in RAS and BRAF. At the bottom of each graph, the number of patients contained in each group that remain throughout the follow up period.

Figure 5.

DFS Kaplan-Meier curves according to miR-21 expression and prognostic statement. A) DFS of the good prognostic group (T1-3/N1): grey line; and the poor prognostic group (T4 or N2): black line. B) DFS of the effect miR-21 levels in the good prognostic group (T1-3/N1). C) DFS of the effect miR-21 levels in the poor prognostic group (T4 or N2). Grey line: N ${}_{0}<$ 6.26; black line N ${}_{0}>$ 6.26; dotted black line the T1-3/N1 or T4 or N2 group as appropriate. ${}^{*}p<$ 0.05; ${}^{**}p<$ 0.01. At the bottom of each graph, the number of patients contained in each group that remain throughout the follow up period.

No significant associations were found between the expression levels of any of the analyzed miRs and the presence of RAS or BRAF mutations with the exception of a trend ( $p=$ 0.0624) found with higher levels of miR-135a and mutations in RAS (Fig. 1). We found a significant association ( $p=$ 0.002) between mutated BRAF and MSI-H (not shown).

3.3 Correlation of miR expression levels with recurrence and survival

To analyze the association of miR expression levels with Disease Free Survival (DFS) and Overall Survival (OS) a cutoff point for each of the analyzed miRs was generated.

The expression levels of the studied miRNAs vary from miR-21 showing the highest expression levels to miR-206 which showed the lowest expression levels. MicroRNA-21 N ${}_{0}$ levels ranged from 0.45 to 12.21, miR-335 goes from 0.38 to 7.59; miR-135a varies from 0.008 to 1.81; miR-Let-7a from 0.13 to 1.29 and miR-206 ranged from 0.00009 to 0.015. A cutoff point was established for each miR and two groups were generated, higher levels than the cutoff point and lower levels than the cutoff point. Thus, we obtained a N ${}_{0}$ cut-off point of 6.26 for miR-21; 2.549 for miR-335; 0.4317 for miR-135a; 0.5718 for miR-Let-7a and 0.006087 for miR-206. Then, association to DFS and OS after a follow up period of three and five years was analyzed. Our results showed that miR-21 high-expression group (N ${}_{0}>$ 6.26) was significantly associated ( $p=$ 0.036) with a better DFS (Fig. 2) and a tendency with a better OS (Fig. 3). No significant associations were found between miR-135a, miR-206, miR-335 or miR-Let-7a expression levels and DFS (Fig. 2) or OS (Fig. 3).

3.4 Association of MSI and mutations in RAS and BRAF with recurrence and survival

Our results showed that there are no significant differences in DFS or OS between wild type and mutated RAS or BRAF and between MSS/MSI-L or MSI-H tumors (Fig. 4).

3.5 Impact of “prognostic grouping” (low risk T1-3/N1) and (high risk T4 or N2) in DFS and association with miR-21 expression

Stage III CRC shows a heterogeneous clinical behavior. Two main groups with different survival, T1-3/N1 and T4 or N2 groups, have been described [2]. Our results shows that there is a significant difference in DFS ( $p<$ 0.01) between the good prognostic group (T1-3/N1) and the poor prognostic group (T4 or N2) (Fig. 5A).

Since we found that miR-21 is associated with DFS we analyzed the effect of prognostic grouping and miR-21 expression (Fig. 5B and C). miR-21 expression levels higher than the cutoff point (N ${}_{0}>$ 6.26) shows significantly ( $p=$ 0.03) better DFS in the poor prognostic group (T4 or N2) (Fig. 5C), but there are no significant differences in the good prognostic group (T1-3/N1) (Fig. 5B).

The impact in DFS of miR-21 expression levels, prognosis group and age of diagnosis was confirmed by a multivariate analysis. The obtained results were HR 2.981: 95% CI 1.196–7.432, $P=$ 0.0191 for miR-21 low expression levels. HR 0.271: 95% CI 0.157–0.467, $P=<$ 0.0001 for good prognostic group (T1-3/N1 group); and HR 1.037: 95% CI 1.014–1.061, $P=$ 0.0015 when considering age of diagnosis as a continuous variable.

4. Discussion

In the present study we have analyzed in patients with stage III CRC, the expression of a panel of miRNAs that was associated to the presence of metastatic disease in CRC [6]. Our results did not show any significant association between miR-135a, miR-206, miR-335 or miR-Let-7a expression levels and DFS or OS. However, we found that high levels of miR-21 correlated with a better DFS. This is conflictive, since a general consensus is that high level of tissue miR-21 expression is associated with poor DFS in patients with CRC [11, 12]. However, in many of these studies individual TNM stages are not considered, stages I to IV are pooled together, stage II is mixed with stage III. Only stage II is analyzed individually in some of these studies. Particularly stage III is the least represented and none of these studies analyzed stage III by itself [11, 12]. This may not be a reliable way to assess OS or DFS, since it is based on a population that is too heterogeneous. In this context, other reports described differences among stages of the effects of high miR-21 expression on DFS or OS. Bovell et al. [13] showed that increased expression of miR-21 correlated with shorter OS of only stage IV patients. Likewise, Kang et al. [14] reported that high miR-21 expression was mainly localized in the stroma and was associated with shorter DFS in stage II but not in stage III CRC. These studies are in accordance with our results. Furthermore, an interesting study in hepatocellular carcinoma shows that deletion of miR-21 in mice promotes the development of hepatocellular carcinoma that is associated with increased expression of oncogenes like Cdc25A [15]. Overexpression of Cdc25A correlates with poor prognosis in cancer and Cdc25A has been validated as a miR-21 target in colon cancer cell lines [16]. These results are in agreement with our data and question the general consensus of the association of high expression of miR-21 and poor clinical outcome.

Another factor to take into account is MSI, the presence of MSI-H is a good prognostic factor [17]. The combination of MSI-H and high levels of miR-21 could contribute to the better DFS we found in our population of stage III CRC patients. Interestingly, we found a significant correlation of MSI-H with high levels of miR-21, to our knowledge this has not been reported previously. In sporadic CRC, MSI-H is predominantly caused by the inhibition of the transcription of the mismatch repair genes, generated by promotor methylation. A different mechanism to decrease mismatch repair genes, could be triggered by miR-21, since it has been reported that miR-21 binds to the 3’-UTRs of MSH2 and MSH6 mRNA to repress protein translation of these mismatch repair genes [18]. This mechanism could explain the association that we found between high levels of miR-21 and MSI-H.

Our results showed a significant association of lower levels of miR-335 and MSI-H tumors, coinciding with Mjelle et al. [19] results, but we did not find an association of miR-335 levels and survival. However, the effect of miR-335 levels in CRC is unclear since some studies indicate miR-335 expression is higher in tumor tissues compared with their matched normal tissue and that miR-335 promotes tumorigenicity of CRC cells by inhibition of RASA1 and activation of the RAS/ERK pathway [20]. Other authors indicate that miR-335 expression levels are higher in tumors that metastasized compared with those that do not [6]. In contrast, other reports suggest that miR-335 inhibit invasion and metastasis in CRC and that miR-335 expression is lower in tumor tissues than in adjacent normal mucosa [21, 22] or report that serum miR-335 levels are lower in stage III and IV compared to stage I and II [23].

Our results did not show any significant association between miR-21, miR-206, miR-335 or miR-Let-7a expression levels and the presence of RAS or BRAF mutations but we found a trend with higher levels of miR-135a and the presence of mutations in RAS.

Post-hoc analysis of the PETACC-8 revealed that MSI and BRAF mutations were not prognostic factors. KRAS mutation was significantly associated with shorter DFS and OS. When the analysis was performed according to MSS/MSI-L or MSI-H, both KRAS and BRAF mutations were independent poor prognostic factors for MSS/MSI-L tumors, while in the MSI-H subgroup, KRAS was not prognostic and BRAF mutation was associated with longer DFS [17, 24]. We did not find significant differences in DFS or OS between wild type and mutated RAS or BRAF and between MSS/MSI-L or MSI-H tumors. A recent report of an ACCENT pooled analysis of 12 adjuvant trials focused on MSI stage III colon cancer showed that MSI-H tumors had better prognosis for N1 stage but not for N2, and the addition of oxaliplatin improved DFS and OS in stage III [25].

Nowadays, the most promising predictive factor for DFS in stage III colon cancer is the analysis of ctDNA after surgery and at the end of adjuvant chemotherapy [26], but it needs to be confirmed in large prospective trials.

Our study has several limitations including the small sample size with few patients in most of the analyzed subgroups and the heterogeneity of this colon cancer population. Only 22 patients had MSI-H tumors and 18 BRAF mutation. Therefore, the conclusions related to those groups need to be taken with caution.

In the future, exploratory studies like this need to be conducted in localized colorectal cancer in order to find prognostic and predictive factors that allow the clinicians to take individualized therapeutic decisions. Patients with stage II MSI-H tumors can be selected to surveillance avoiding adjuvant chemotherapy.

5. Conclusions

Our panel of miRNA expression tested in this study does not allow to select specific subgroups of patients in which changes in clinical practice can be done. Nevertheless, some data such as the association of high expression levels of miR-21 and better prognosis in the poor prognostic group may be of interest and could be explored in future prospective clinical trials.

Footnotes

Acknowledgments

This work was partially funded by: Fondo de Investigación Sanitaria (FIS) Instituto de Salud Carlos III Pi12/00172; IMMUNOTHERCAN Comunidad de Madrid S2017/BMD-3733; Fundacion Mutua Madrileña; Bayer Healthcare; Fundacion 2000 Merck-Serono.

We thank Inmaculada Ruiz from the Cooperative Spanish Group for the Treatment of Digestive Tumors (TTD) and Elena Molina from the Biobank of the Hospital Clinico San Carlos.

Author contributions

Conception: Javier Sastre, Eduardo Diaz-Rubio and Beatriz Perez-Villamil.

Interpretation or analysis of data: Tania Calvo-López, Mateo Paz-Cabezas, Patricia Llovet and Maria Dolores Ibañez.

For the interpretation of clinical data: Vicente Alonso-Orduña, J.M ${}^{\text{a}}$ . Viéitez, Alfonso Yubero, Ruth Vera, Elena Asensio-Martínez, P. Garcia-Alfonso and Enrique Aranda.

Preparation of the manuscript: Tania Calvo-López and Beatriz Pérez-Villamil.

Revision for important intellectual content: Javier Sastre, Eduardo Diaz-Rubio and Beatriz Perez-Villamil.

Supervision: Javier Sastre, Eduardo Diaz-Rubio and Beatriz Perez-Villamil.

References

Sung

Ferlay

Siegel

R.L.

Laversanne

Soerjomataram

Jemal

and Bray

, Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries, CA, Cancer J. Clin. 0 (2021), 1–41. doi: 10.3322/caac.21660.

Sobrero

A.F.

Puccini

Shi

Grothey

Andrè

Shields

A.F.

Souglakos

Yoshino

Iveson

Ceppi

and Bruzzi

, A new prognostic and predictive tool for shared decision making in stage III colon cancer, Eur J Cancer 138 (2020), 182–188. doi: 10.1016/j.ejca.2020.07.031.

Innocenti

Zemla

T.J.

Niedzwiecki

Tam

Mahajan

Goldberg

R.M.

Bertagnolli

M.M.

Blanke

C.D.

Sanoff

Atkins

Polite

Venook

A.P.

Lenz

and Kabbarah

, Mutational analysis of patients with colorectal cancer in CALGB/SWOG 80405 identifies new roles of microsatellite instability and tumor mutational burden for patient outcome, J. Clin. Oncol. 37 (2019), 1217–1227. doi: 10.1200/jco.18.01798.

Sastre

de la Orden

Martínez

Bando

Balbín

Bellosillo

Palanca

Gomez

M.I.P.

Mediero

Llovet

Moral

V.M.

Viéitez

J.M.

García-Alfonso

Calle

S.G.

Ortiz-Morales

M.J.

Salud

Quintero

Lopez

Díaz-Rubio

and Aranda

, Association between baseline circulating tumor cells, molecular tumor profiling, and clinical characteristics in a large cohort of chemo-naïve metastatic colorectal cancer patients prospectively collected, Clin. Colorectal Cancer 19 (2020), e110–e116. doi: 10.1016/j.clcc.2020.02.014.

Rupaimoole

and Slack

F.J.

, MicroRNA therapeutics: Towards a new era for the management of cancer and other diseases, Nat. Rev. Drug Discov. 16 (2017), 203–222. doi: 10.1038/nrd.2016.246.

Vickers

M.M.

Bat

Gorn-Hondermann

Yarom

Daneshmand

Hanson

J.E.L.

Addison

C.L.

Asmis

T.R.

Jonker

D.J.

Maroun

Lorimer

I.A.J.

Goss

G.D.

and Dimitroulakos

, Stage-dependent differential expression of microRNAs in colorectal cancer: Potential role as markers of metastatic disease, Clin. Exp. Metastasis 29 (2012), 123–132. doi: 10.1007/s10585-011-9435-3.

Sotelo

M.J.

Sastre

Maestro

M.L.

Veganzones

Viéitez

J.M.

Alonso

Grávalos

Escudero

Vera

Aranda

García-Alfonso

Gallego-Plazas

Lopez

Pericay

Arrivi

Vicente

Ballesteros

Elez

López-Ladrón

and Díaz-Rubio

, Role of circulating tumor cells as prognostic marker in resected stage III colorectal cancer, Ann. Oncol. 26 (2015), 535–541. doi: 10.1093/annonc/mdu568.

López-Campos

G.H.

Romera-López

Martín-Sánchez

Diaz-Rubio

López-Alomso

and Perez-Villamil

, MicroRNA microarray data analysis in colon cancer: Effects of normalization, Lecture Notes in Computer Science 6692 (2011), 260–267. doi: 10.1007/978-3-642-21498-1_33.

Ramakers

Ruijter

J.M.

Lekanne Deprez

R.H.

and Moorman

A.F.M.

, Assumption-free analysis of quantitative real-time polymerase chain reaction (PCR) data, Neurosci. Lett. 339 (2003), 62–66. doi: 10.1016/s0304-3940(02)01423-4.

10.

Budczies

Klauschen

V Sinn

Györffy

Schmitt

W.D.

and Darb-Esfahani

, Cutoff finder: A comprehensive and straightforward web application enabling rapid biomarker cutoff optimization, PLoS ONE 7 (2012), 1–7. doi: 10.1371/journal.pone.0051862.

11.

Chen

Liu

Jin

Ding

Zheng

and Yu

, Tissue microRNA-21 expression predicted recurrence and poor survival in patients with colorectal cancer – A meta-analysis, Onco. Targets. Ther. 9 (2016), 2615–2624. doi: 10.2147/ott.s103893.

12.

Peng

Zhang

Min

Zou

Shen

and Zhu

, The clinical role of microRNA-21 as a promising biomarker in the diagnosis and prognosis of colorectal cancer: A systematic review and meta-analysis, Oncotarget 8 (2017), 44893–44909. doi: 10.18632/oncotarget.16488.

13.

Bovell

L.C.

Shanmugam

Putcha

B.D.K.

Katkoori

V.R.

Zhang

Bae

Singh

K.P.

Grizzle

W.E.

and Manne

, The prognostic value of MicroRNAs varies with patient race/ethnicity and stage of colorectal cancer, Clin. Cancer Res. 19 (2013), 3955–3965. doi: 10.1158/1078-0432.ccr-12-3302.

14.

Kang

W.K.

Lee

J.K.

S.T.

Lee

S.H.

and Jung

C.K.

, Stromal expression of miR-21 in T3-4a colorectal cancer is an independent predictor of early tumor relapse, BMC Gastroenterol. 15 (2015), 2. doi: 10.1186/s12876-015-0227-0.

15.

de Sousa

M.C.

Calo

Sobolowski

Gjorgjieva

Clément

Maeder

Dolicka

Fournier

Vinet

Montet

Dufour

Humar

Negro

Sempoux

and Foti

, Mir-21 suppression promotes mouse hepatocarcinogenesis, Cancers 13 (2021), 4983. doi: 10.3390/cancers13194983.

16.

Wang

Zou

Zhang

Dulak

Tomko Jr

R.J.

Lazo

J.S.

Wang

Zhang

and Yu

, microRNA-21 negatively regulates Cdc25A and cell cycle progression in colon cancer cells, Cancer Res. 69 (2009), 8157–8165. doi: 10.1158/0008-5472.can-09-1996.

17.

Taieb

Le Malicot

Shi

Penault-Llorca

Bouché

Tabernero

Mini

Goldberg

R.M.

Folprecht

Van Laethem

J.L.

Sargent

D.J.

Alberts

S.R.

Emile

J.F.

Puig

P.L.

and Sinicrope

F.A.

, Prognostic value of BRAF and KRAS mutations in MSI and MSS stage III colon cancer, J. Natl. Cancer Inst. 109 (2017), 1–12. doi: 10.1093/jnci/djw272.

18.

Valeri

Gasparini

Braconi

Paone

Lovat

Fabbri

Sumani

K.M.

Alder

Amadori

Patel

Nuovo

G.J.

Fishel

and Croce

C.M.

, MicroRNA-21 induces resistance to 5-fluorouracil by down-regulating human DNA MutS homolog 2 (hMSH2), Proc. Natl. Acad. Sci. U. S. A. 107 (2010), 21098–21103. doi: 10.1073/pnas.1015541107.

19.

Mjelle

Sjursen

Thommesen

Sætrom

and Hofsli

, Small RNA expression from viruses, bacteria and human miRNAs in colon cancer tissue and its association with microsatellite instability and tumor location, BMC Cancer 19, (2019), 1–12. doi: 10.1186/s12885-019-5330-0.

20.

Yang

Yuan

Liu

Zhang

Hong

Liu

Zhou

Chen

and Li

, Overexpression of miR-335 confers cell proliferation and tumour growth to colorectal carcinoma cells, Mol. Cell. Biochem. 412 (2016), 235–245. doi: 10.1007/s11010-015-2630-9.

21.

Sun

Z.F.

Zhang

Liu

Qiu

Liu

and Dong

, MicroRNA-335 inhibits invasion and metastasis of colorectal cancer by targeting ZEB2, Med. Oncol. 31 (2014), 982. doi: 10.1007/s12032-014-0982-8.

22.

Zhang

and Yang

, MiR-335-5p inhibits cell proliferation, migration and invasion in colorectal cancer through downregulating LDHB, J. BUON. 24 (2019), 1128–1136.

23.

Wang

Zhao

Zhou

and Yan

, Serum miR-1301-3p, miR-335-5p, miR-28-5p, and their target B7-H3 may serve as novel biomarkers for colorectal cancer, 24 (2019), 1120–1127.

24.

Taieb

Zaanan

Le Malicot

Julié

Blons

Mineur

Bennouna

Tabernero

Mini

Folprecht

Van Laethem

J.L.

Lepage

Emile

J.F.

and Laurent-Puig

, Prognostic effect of BRAF and KRAS mutations in patients with stage III colon cancer treated with leucovorin, fluorouracil, and oxaliplatin with or without cetuximab, JAMA Oncol. 2 (2016), 643–653. doi: 10.1001/jamaoncol.2015.5225.

25.

Cohen

Taieb

Fiskum

Yothers

Goldberg

Yoshino

Alberts

Allegra

de Gramont

Seitz

J.F.

O’Connell

Haller

Wolmark

Erlichman

Zaniboni

Lonardi

Kerr

Grothey

Sinicrope

F.A.

André

and Shi

, Microsatellite instability in patients with stage III colon cancer receiving fluoropyrimidine with or without oxaliplatin: An ACCENT pooled analysis of 12 adjuvant trials, J. Clin. Oncol. 39 (2020), 642–651. doi: 10.1200/jco.20.01600.

26.

Tie

Cohen

J.D.

Wang

Christie

Simons

Lee

Wong

Kosmider

Ananda

McKendrick

Lee

Cho

J.H.

Faragher

Jones

I.T.

Ptak

Shaeffer

M.J.

Silliman

Dobbyn

Tomasetti

Papadopoulos

Kinzler

K.W.

Vogelstein

and Gibbs

, Circulating tumor dna analyses as markers of recurrence risk and benefit of adjuvant therapy for stage III colon cancer, JAMA Oncol. 5 (2019), 1710–1717. doi: 10.1001/jamaoncol.2019.3616.

Association of miR-21 and miR-335 to microsatellite instability and prognosis in stage III colorectal cancer

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSIONS:

Keywords

1. Introduction

2. Materials and methods

2.1 Patients

2.2 RNA and DNA extraction

2.3 DNA mutational analysis

2.4 Analysis of microsatellite instability

2.5 Analysis of miRNA expression

2.6 Cutoff finder

2.7 Statistical analysis

3. Results

3.1 Patients

Table 1 Patient characteristics ( N = 329)

3.4 Association of MSI and mutations in RAS and BRAF with recurrence and survival

3.5 Impact of “prognostic grouping” (low risk T1-3/N1) and (high risk T4 or N2) in DFS and association with miR-21 expression

4. Discussion

5. Conclusions

Footnotes

Acknowledgments

Author contributions

References

Table 1
Patient characteristics ( $N=$ 329)