The Effect of Metformin Usage in Patients With Metastatic Colorectal Cancer Receiving First-line Systemic Therapy

Abstract

Background/Aim

To investigate the effect of metformin usage on progression-free (PFS) and overall (OS) survival in patients with metastatic colorectal cancer (mCRC) receiving first-line systemic therapy.

Patients and Methods

This was a single-center retrospective cohort study of patients aged 18 years or older with histologically confirmed mCRC receiving first-line systemic 5-fluorouracil-based chemotherapy combined with either anti-epidermal growth factor receptor (EGFR) therapy or anti-vascular endothelial growth factor (bevacizumab) between January 2010 and December 2022. This study examined the effect of metformin on PFS and OS of patients with mCRC.

Results

A total of 134 patients were included, with a median age of 59.5 years; 66.4% of the patients were male, and 23.9% had diabetes. Metformin usage was associated with a significant improvement in median PFS (14.0 vs. 9.9 months, p=0.04) but not in median OS (20.7 vs. 19.5 months, p=0.76). In univariate Cox regression for PFS, metformin usage (hazard ratio=0.62, p=0.04) and anti-EGFR therapy (hazard ratio=0.54, p<0.01) were associated with significantly lower hazard ratios.

Conclusion

Metformin usage in patients with mCRC receiving systemic therapy may improve progression-free survival but does not significantly affect overall survival. This result outlines the potential of metformin in mCRC treatment, especially in enhancing the progression-free survival. Further studies are needed to confirm the role of metformin in systemic treatment of mCRC.

Introduction

Colorectal cancer (CRC) is the third most common malignancy in the world and its incidence increases with age (1). As the incidence of diabetes mellitus (DM) also increases with age, it is not surprising that many patients diagnosed with CRC also have DM. In some preclinical and epidemiological studies, DM is noted as a significant risk factor for cancer development including CRC, breast cancer, and other gastrointestinal cancer (2). In comparison with non-diabetic patients, diabetic patients with CRC are thought to have increased complication risk and reduced survival (3,4). Patients with newly diagnosed metastatic CRC (mCRC) with diabetes have been shown to have poorer progression-free survival and overall survival than non-diabetic patients with mCRC (5).

Metformin is the most frequently used oral antidiabetic drug, and there are many studies regarding the effect of metformin on CRC incidence and prognosis, indicating its protective effect. In a recent meta-analysis, metformin usage was shown to reduce cancer incidence, including of CRC (6). Moreover, preclinical studies showed that, by activation of liver kinase B1-related AMPK and reducing mTOR activation, metformin had an anticancer effect (7,8). In a recent preclinical study, metformin was shown to have a glucose-independent inhibitory effect on KRAS proto-oncogene, GTPase (KRAS)-mutant tumor cells, resulting in reduced CRC aggressiveness, and intestinal microbiota changes caused by metformin can inhibit tumorigenesis and metastases (9,10). The potential effects of metformin include its synergistic effect with fluoropyrimidine agents such as 5-fluorouracil (9). Metformin has also been investigated as a potential agent in adjuvant treatment of cancer, and in a strategy to overcome drug resistance (11,12).

Knowledge about the effect of metformin usage in patients with mCRC receiving systemic treatment and its effect on patient survival is limited. Since metformin is a widely used oral anti-diabetic drug with different mechanisms of action, many studies have been done to assess its effect on diseases other than DM. Our study aimed to examine the effect of metformin usage on progression-free and overall survival in patients with mCRC, to identify the patient subgroups for which metformin treatment is most beneficial, and to investigate its effectiveness.

Patients and Methods

The study was planned as a retrospective cohort study. Patients with mCRC admitted to a single academic center’s Oncology Outpatient Clinic receiving first-line systemic therapy with either anti-epidermal growth factor receptor (EGFR) therapy in those with RAS wild-type tumors or anti-vascular endothelial growth factor (VEGF) (bevacizumab) in those with RAS-mutant or wild-type tumors, between January 2010 and December 2022 were assessed for eligibility. Patients ≥18 years of age, with known histological diagnosis of CRC, receiving first-line chemotherapy, and either anti-EGFR or anti-VEGF in the metastatic setting were included. Patients without a pathological diagnosis, with malignities other than mCRC, recurrence of less than 6 months after completion of adjuvant treatment, lacking data on comorbidities and drug usage, and patients without any information of metastatic sites were excluded. Collected data included age, sidedness of primary cancer, metastatic sites, comorbid diseases, therapies administered, time to progression, and overall survival. Patients with ≥10% missing data were not included in the final analysis.

The primary outcome of the study was progression-free survival. The secondary outcomes were overall survival and the effects of RAS or RAF proto-oncogene serine/ threonine kinase (RAF) mutation, sidedness, DM, biological agents administered (anti-VEGF vs. anti-EGFR), and metastatic sites on survival. The survival analyses started from beginning of systemic treatment and ended on progression for progression-free survival, death for overall survival, or at the last follow-up for censored status.

Demographic results are given as the median with interquartile range for continuous variables and percentages for categorical variables. Analyses of categorical variables were performed with chi-square tests, normally distributed continuous variables were analyzed with Student’s t-test, and non-normally distributed variables with Wilcoxon tests. Survival analyses were performed with Kaplan−Meier curves and differences were analyzed with the log-rank test. Univariate and multivariate Cox regression were performed to estimate the hazard ratios (HR) in subgroup analyses. The statistical analyses were carried out with R 4.3.1 (The R Foundation, Vienna, Austria).

Results

A total of 134 patients were included in the study. The median age of patients was 59.5 years, and 89 patients were male. The median age of male patients was 60 years and of female patients was 58 years; the median age did not vary between the sexes (Mann-Whitney U-test, p=0.97). Sixty-three patients (47.0%) had hypertension, 32 patients (23.9%) had DM, and 25 patients (18.7%) had coronary arterial disease as comorbidities, the comorbidities did not differ significantly between the sexes (p=0.11-0.49). In terms of primary tumor location, more patients had left colon and rectal tumors, 51 tumors (38.6%) had left colon origin, 44 tumors (33.3%) had rectal origin, and 37 tumors (28.0%) had right colon origin. All the patients enrolled in this study had an Eastern Cooperative Oncology Group performance score of 0 or 1.

Most of the patients had liver metastasis [106 patients (79.1%)], 47 patients (35.3%) had lymph node metastasis, 44 (32.8%) had peritoneal metastasis, and 32 (23.9%) had lung metastasis. Lung metastasis was significantly more frequent in patients with rectal cancer: 15 patients (34.1%) had lung metastasis with rectal cancer, whilst 17 patients (19.3%) had lung metastasis with right or left colon cancer (p=0.05). Peritoneal metastasis was significantly higher in those with right colon cancer: 18 patients (48.7%) had peritoneal metastasis with right colon cancer compared with 25 patients (26.3%) with left colon cancer and rectal cancer (p=0.03).

RAS mutation was seen in 67 patients (50.0%) and in this cohort, no RAF mutation was observed. As chemotherapy backbone, 84 patients (61.3%) received folinic acid/fluorouracil/oxaliplatin (FOLFOX)/capecitabine with oxaliplatin (CAPOX) and 49 patients (36.8%) received folinic acid/fluorouracil/irinotecan (FOLFIRI). Eighty-nine patients (66.4%) received bevacizumab and 45 patients (33.6%) received cetuximab/panitumumab. Regarding a history of surgery, 28 patients (21.1%) were operated on with palliative intent and 26 patients (19.6%) were operated on with curative intent. Overall, 30 patients (22.4%) were using metformin and all these were diabetic. Demographic and disease-related properties are summarized in Table I.

Table I. Demographic and disease-related properties of study patients.

CAPOX: Capecitabine with oxaliplatin; DM: diabetes mellitus; FOLFIRI: folinic acid, fluorouracil plus irinotecan; FOLFOX: folinic acid, fluorouracil plus oxaliplatin; IQR: interquartile range; RAS: RAS proto-oncogene.

Median progression-free survival was 14.0 months in metformin-receiving patients, and 9.9 months in patients not receiving metformin - this difference was statistically significant (p=0.04) (Figure 1A). Median overall survival was 20.7 months in metformin-receiving patients, and 19.5 months in patients not receiving it - no significant difference was seen (p=0.76) (Figure 1B).

Figure 1.

Kaplan-Meier curves for progression-free and overall survival in patients. (A) Progression-free survival of patients receiving and not receiving metformin. (B) Overall survival of patients receiving and not receiving metformin. (C): Overall survival of metformin users according to primary tumor location.

In univariate Cox regression analysis for progression-free survival, metformin usage and anti-EGFR therapy were significantly protective, with lower HRs of 0.62 (p=0.04) and 0.54 (p<0.01), respectively. RAS mutation and peritoneal metastasis were significantly unfavorable factors, with HRs of 1.71 (p=0.01), and 1.79 (p=0.01), respectively. In multivariate Cox regression, peritoneal metastasis remained significant with an HR of 2.28 (p<0.01), as did anti-EGFR therapy with an HR of 0.54 (p=0.04). Details of univariate and multivariate Cox regression for progression-free survival are presented in Table II.

Table II. Univariate and multivariate Cox regression for progression-free survival in all patients.

EGFR: Epidermal growth factor receptor; DM: diabetes mellitus; VEGF: vascular endothelial growth factor. Statistically significant p-values are shown in bold.

In multivariate Cox regression analysis for overall survival, coronary arterial disease, left colon tumors and anti-EGFR therapy were significantly favorable factors. Coronary arterial disease had an HR of 0.53 (p=0.02), left colon tumors had an HR of 0.59 (p=0.04), and anti-EGFR therapy had an HR of 0.52 (p<0.01). RAS mutation was associated with a significantly unfavorable HR of 2.03 (p<0.01) in univariate analysis. In multivariate Cox regression, only tumor of left colon origin was significant (HR=0.54; p=0.02). Details of univariate and multivariate Cox regression for overall survival are presented in Table III.

Table III. Univariate and multivariate Cox regression for overall survival in all patients.

CI: Confidence interval; DM: diabetes mellitus; EGFR: epidermal growth factor receptor; HR: hazard ratio; RAS: RAS proto-oncogene; VEGF: vascular endothelial growth factor. Statistically significant p-values are shown in bold.

In the subgroup analysis of survival, in patients with RAS-mutant disease, median progression-free survival was 11.6 months in metformin users and 8.8 months in patients not receiving metformin; the difference was not statistically significant (p=0.13). In patients with RAS-mutant disease, median overall survival was 18.2 months in metformin users and 16.3 months in patients not receiving metformin; the difference was not statistically significant (p=0.87). In metformin-using patients, median progression-free survival was 20.9 months in patients with left colon cancer, 15.5 in patients with rectal cancer, and 11.5 months in patients with right colon cancer. The difference in progression-free survival by sidedness was not statistically significant (p=0.16). In metformin-using patients, median overall survival was 31.3 months in those with left colon cancer, 18.5 months in patients with right colon cancer, and 18.2 months in patients with rectal cancer. The difference in overall survival according to sidedness was not statistically significant (p=0.54). Kaplan-Meier curves for progression-free survival by tumor sidedness in metformin users are presented in Figure 1C.

Discussion

In our study, we aimed to investigate the effect of metformin use on survival and prognosis in patients with mCRC with the hypothesis that liver kinase B1-related AMPK activation and mTOR inhibition can increase the response to first-line systemic treatment. More patients in our cohort were male, and about a quarter of the patients had diabetes. Despite the known worse prognosis of diabetic patients with mCRC, our results show significant progression-free survival benefit that may be attributable to metformin usage, but that benefit did not extend to the overall survival of the patients. The main reason for this may be that having more comorbidities of diabetes may lead to earlier death. In our study cohort, we were unable to examine the difference in survival of diabetic patients who did not use metformin because almost all the diabetic patients were using metformin.

The subgroup analyses also showed that there was a relatively small but not statistically significant progression-free and overall survival benefit attributable to metformin use in patients with RAS-mutant disease that may be attributable to the role of RAS mutation in changes of mTOR activity, changes which may lead to cancer progression (13). In a preclinical trial by Sugiura et al., metformin use was shown to reduce mTOR pathway activity via AMPK activation, leading to a decrease in metastasis development and hinting at a potential role in treatment (14). Xie et al. suggested that metformin has a potential inhibitory effect on mCRC cells with KRAS mutation via silencing multidrug and toxin extrusion protein 1 (MATE1) in their preclinical study (15). Our results add to the preclinical insight of these studies. The effects of metformin include changes in gut microbiota. Fan et al. investigated the effect of metformin in patients with CRC and healthy volunteers on the microbiota and M2 macrophage polarization and showed that metformin may lead to changes in gut microbiota. The bacterial species in patients with CRC treated with metformin changed, becoming similar to that of the healthy volunteers, and such a change may lead to suppression of M2 macrophage polarization and inhibition of tumorigenesis (9).

In our cohort, consistent with the literature, peritoneal metastasis was shown to lead to an inferior outcome in CRC. Disseminating tumor cells in the peritoneum are thought to undergo mesenchymal transformation and gain the ability to survive outside the epithelial environment (16). This transformation may lead to a significant change in pharmacokinetics in the peritoneal cavity and may contribute to treatment resistance (17). A preclinical study by Sekino et al. showed that metformin helped to suppress peritoneal metastasis in gastric cancer cell lines, but in our cohort, although metformin was shown to increase progression-free survival, patients with peritoneal metastasis still had worse outcomes (18). In addition, more patients with right-sided colon cancer had peritoneal metastases in our cohort and right-sided colon cancer is known to have worse outcomes, even with curative-intent treatment (19).

Several other studies have suggested that metformin may have a contributory role in treatment response and survival in different types of cancer and systemic treatments. Ciccarese et al. suggested that concomitant usage of metformin with nivolumab increased progression-free and overall survival (20). A combined preclinical and clinical study by Jang et al. showed improved tumor response to radiation in xenograft mouse models of hepatocellular carcinoma and improved survival in patients with hepatocellular carcinoma (21). A review by Saengboonmee et al. discussed the potential effect of metformin in tumor progression, noting AMPK pathway activation leading to modulation of the mTOR pathway and increasing tumor regression, resulting in a reduced risk of carcinogenesis in diabetes; yet the clinical trials showed heterogenous responses (22).

The main limitations of our study are the retrospective nature of its design, and the fact that all the metformin users were also diabetic, which may itself reduce survival and lead to poor response. Moreover, the patient count was relatively low, and only patients receiving biological agents, either anti-EGFR or anti-VEGF, along with 5-fluorouracil were included, leading to a decrease in the generality of our results. Yet as the results of our study are consistent with the literature, it may be reasonable to examine the effect of metformin in treatment of mCRC; randomized controlled trials are needed to confirm our results and the potential role of metformin in mCRC treatment.

Conclusion

Metformin use in mCRC can increase the survival of patients and might have a role in mCRC treatment. Although diabetes leads to a poorer prognosis for colorectal cancer, metformin also positively affected the prognosis in these patients. Further studies are needed to identify the potential role of metformin use in mCRC treatment and confirm our results.

Availability of Data and Materials

Due to the restrictions of the Turkish Personal Data Protection Law, the data cannot be made available in the open environment; the data are available from the corresponding author on reasonable request.

Conflicts of Interest

There are no conflicts of interest to declare.

Authors’ Contributions

ECE: Data collection, analysis, methodology, and writing; MY: data collection and editing; YG: data collection; GU: conceptualization, editing and supervision.