ACVR2B polymorphism, Adiponectin, and GDF-15 levels as biomarkers for cachexia in gastrointestinal cancer

Abstract

Reversing cancer cachexia remains a challenge. Genetic biomarkers for the early detection of the disease have been explored in order to enable the implementation of preventive measures. We therefore genotyped candidate genes based on cachexia phenotype and quantified adiponectin and GDF-15 levels in cachectic patients with gastrointestinal cancer. Patients with a diagnosis of gastrointestinal cancer were divided into a cachectic and a non-cachectic group after the start of chemotherapy. A control group (no cancer) was also included. We genotyped the following single nucleotide polymorphisms (SNPs) by quantitative PCR: FOXO3 (rs1935949), FOXO3 (rs4946935), ACVR2B (rs2268757), and SELP (rs6136). In addition, we quantified adiponectin and GDF-15 levels by ELISA. The rs2268757 SNP in the ACVR2B gene was associated with the weight loss phenotype in cachectic patients with gastrointestinal cancer (non-cachectic, P = 0.004). Plasma adiponectin levels were higher in cachectic patients compared to controls (P = 0.01) and non-cachectic patients (P = 0.004). GDF-15 was also elevated in cachectic patients compared to controls (P < 0.0001) and non-cachectic patients (P = 0.001). Analysis by sex showed elevated adiponectin levels in men (control, P = 0.01) and cachectic women (control, P = 0.04; non-cachectic, P = 0.01), as well as elevated GDF-15 levels in men (control, P = 0.002) and cachectic women (control, P = 0.002; non-cachectic, P = 0.007). However, there was no significant difference in the levels of these cytokines between cachectic men and women. The results suggest the rs2268757 SNP in the ACVR2B gene, adiponectin, and GDF-15 as potential biomarkers of cachexia in gastrointestinal cancer.

Introduction

The prevalence of weight loss is higher among patients with gastrointestinal cancer (GIC) when compared to other types of tumors^1,2. The location of the tumor can lead to weight loss even before chemotherapy as a result of obstruction of the gastrointestinal tract caused by the tumor mass, malabsorption of nutrients, dysphagia, pain, and gastroparesis^3–6. Furthermore, chemotherapeutic agents exacerbate weight loss and cause weakening of skeletal muscles, mainly due to their indirect side effects on the gastrointestinal tract such as nausea, anorexia, constipation, diarrhea, and vomiting⁷. This weight loss, called cancer-associated cachexia, induces metabolic and endocrine alterations, in addition to activating the immune system, with consequent degradation of muscle and adipose tissues^8,9.

The main diagnostic criterion for cachexia is weight loss¹⁰. However, the complexity of cachexia, which affects multiple organs, and differences in the criteria used for defining weight loss make the diagnosis of this condition difficult. Consequently, the reported prevalence of the disease is low, inconsistency exists between studies and populations, and the development of effective treatments is difficult¹¹. Once diagnosed, nutritional support alone is not capable of reversing cachexia. Despite the multidisciplinary approach and studies trying to identify an effective therapeutic medication, there is still no effective treatment for cachexia¹².

It is not possible to predict which patient will develop weight loss after the start of chemotherapy. As weight loss progresses, the sensitivity of patients to pharmacological and nutritional interventions decreases, resulting in surgical complications, high mortality rates, an increased risk of chemotherapy-associated toxicity, and reduced intensity and efficacy of the anticancer treatment response^13,14. Within this context, genetic biomarkers have been widely studied for the early detection of cachexia and for the implementation of preventive measures that could reduce inflammation and muscle mass loss, which are important characteristics of cachexia¹⁵. In addition to inflammation and muscle loss, cachexia factors are also considered important biomarkers due to their crucial role in signaling between organs, causing disruptions in inter-tissue communication¹⁶. These factors can inhibit and promote tumor progression, such as GDF-15, which suppresses appetite, and adiponectin, which reduces muscle mass and impairs metabolic signaling^17,18.

The aim of the present study was to identify possible genetic biomarkers for GIC-associated cachexia by investigating genetic polymorphisms and the levels of adiponectin and GDF-15.

Results

• Gastrointestinal cancer patient characteristics.

The patients were divided into cachectic (n = 49) and non-cachectic (n = 25) (Table 1). The mean age was 62 ± 12 years in the cachectic group and 61 ± 9 years in the non-cachectic group. Among cachectic patients, 22 were female (44.9%) and 27 were male (55.1%), while among non-cachectic patients, 12 were female (48%) and 13 were male (52%).

Table 1.

Clinical and nutritional characteristics of cachectic and non-cachectic patients.

Characteristics	Cachectic (n = 49)	Non-cachectic (n = 25)
Age (years)	62 ± 12	61 ± 9
Sex
Female	22 (44.9)	12 (48)
Male	27 (55.1)	13 (52)
Comorbidities
Present	24 (48.8)	17 (68)
Absent	25 (51)	8 (32)
Tumor site
Esophagus	3 (6.1)	-
Gastric	8 (16.3)	-
Pancreas	2 (4.0)	-
Colon	19 (38.8)	18 (72)
Rectum	17 (34.7)	7 (28)
Metastasis
Lung	13 (26.5)	3 (12)
Liver	10 (20.4)	4 (16)
Ovarian	4 (8.2)	-
Lymph nodes	4 (8.2)	2 (8)
Other organs	2 (4)	2 (8)
Chemotherapy protocol
mFLOXa	33 (67.3)	16 (64)
Mayo Clinicb	8 (16.3)	6 (24)
FOLFOXc	3 (6.12)	3 (12)
Carboplatin + taxol	2 (4.08)	-
Others	3 (6.12)	-
Height (m)	1.66 ± 0.1	1.63 ± 0.09
Weight (kg)	61 ± 15.2	74 ± 11.1
Body mass index (kg/m²)	21.8 ± 4.9	27.85 ± 4.3
Underweight	25 (51)	2 (8)
Normal weight	12 (24.5)	8 (32)
Overweight	8 (16.3)	11 (44)
Obese	4 (8.2)	4 (16)
Weight loss after chemotherapy (%)	18.9 ± 11.5	-

^a Oxaliplatin + 5-FU + leucovorin.

^b 5-FU + leucovorin.

^c Oxaliplatin + 5-FU + leucovorin, with a rapid injection (bolus) of 5-FU followed by continuous infusion.

- : no data.

Cachectic patients had a mean weight of 61 ± 15.2 kg and a mean BMI of 21.8 ± 4.9 kg/m². This group exhibited a percentage of weight loss of 18.9 ± 11.5% after the start of chemotherapy. Non-cachectic patients had a mean weight of 74 ± 11.1 kg and a mean BMI of 27.85 ± 4.3 kg/m². A control group consisting of individuals without cancer (n = 51) was also included.

• Polymorphisms associated with the cachexia phenotype in patients with gastrointestinal cancer.

We evaluated the frequencies of the risk alleles of SNPs in the FOXO3, SELP, and ACVR2B genes. First, cachectic patients and controls were compared, which did not reveal an association between the SNPs and the groups analyzed (P = 0.06, P = 0.28, P = 0.38 and P = 0.85, respectively) (Table 2). When comparing the risk alleles between cachectic and non-cachectic patients, we observed that the rs2268757 SNP in the ACVR2B gene was associated with the weight loss phenotype, while no significant difference was found for the FOXO3 or SELP gene SNPs (P = 0.004, P = 0.71, P > 0.99 and P = 0.85, respectively) (Table 3).

Table 2.

Genes with variants significantly associated with the weight loss phenotype comparing controls and cachectic patients.

Gene	SNP	Risk allele	OR (95% CI)	P-value
FOXO3	rs1935949	G	1.99 (1.00-3.84)	0.06
FOXO3	rs4946935	A	1.44 (0.79–2.57)	0.28
ACVR2B	rs2268757	C	0.65 (0.27–1.54)	0.38
SELP	rs6136	G	1.15 (0.56–2.42)	0.85

OR: odds ratio; 95% CI: 95% confidence interval. Fisher’s exact test.

Table 3.

Genes with variants significantly associated with the weight loss phenotype comparing cachectic and non-cachectic patients.

Gene	SNP	Risk allele	OR (95% CI)	P-value
FOXO3	rs1935949	G	0.81 (0.40–1.71)	0.71
FOXO3	rs4946935	A	0.95 (0.52–1.72)	> 0.99
ACVR2B	rs2268757	C	2.75 (1.33–5.70)	0.004*
SELP	rs6136	G	0.87 (0.44–1.81)	0.85

OR: odds ratio; 95% CI: 95% confidence interval. *P ≤ 0.05, Fisher’s exact test.

• Adiponectin and GDF-15 levels are higher in cachectic patients.

Plasma adiponectin levels were significantly higher in cachectic patients when compared to the control and non-cachectic groups (P = 0.004 and P = 0.01, respectively) (Fig. 1A). Likewise, GDF-15 levels were higher in cachectic patients compared to the other groups (P < 0.001 and P = 0.001, respectively) (Fig. 1B).

Fig. 1.

Comparison of adiponectin (A) and GDF-15 (B) levels between control, (n = 51), non-cachectic (n = 25), and cachectic patients (n = 45). Kruskal-Wallis test.

We then applied Spearman’s correlation test to evaluate the association of the two cytokines with the percentage of weight loss during the treatment period. The plasma levels of both adiponectin and GDF-15 showed a weak correlation with weight loss after the start of chemotherapy (P = 0.58, r = 0.07 and P = 0.15, r = 0.21, respectively) (Fig. 2A, B).

Fig. 2.

Correlation between plasma levels of adiponectin (A) and GDF-15 (B) and weight loss percentage after chemotherapy in cachectic patients (n = 49). Spearman correlation.

• The higher adiponectin and GDF-15 levels in cachectic patients are independent of sex.

Analysis of the plasma levels of adiponectin and GDF-15 by sex showed elevated levels of the two cytokines in both cachectic men (adiponectin, P = 0.01; GDF-15, P = 0.002) (Fig. 3A, B) and women (adiponectin, P < 0.01; GDF-15, P < 0.001) (Fig. 3A, B) when compared to the control and non-cachectic groups. However, no significant difference in the levels of these cytokines was observed between cachectic men and women.

Fig. 3.

Comparison of plasma levels of adiponectin (A) and GDF-15 (B) between the control, non-cachectic and cachectic groups according to sex. Two-way ANOVA.

Discussion

This study identified potential biomarkers for GIC-associated cachexia, as defined by weight loss ≥ 5%. We found that the risk allele of the SNP in the ACVR2B gene was associated with the weight loss phenotype in cachectic patients, as well as with elevated levels of adiponectin and GDF-15.

Skeletal muscle tissue is the main target of degradation for energy production in cachexia, regardless of the loss of adipose tissue⁹. Activin receptor type 2B (ACVR2B) has been widely studied in the context of weight loss in cachexia since it mediates the signaling of proteins of the transforming growth factor beta (TGF-β) family, such as myostatin and activin A^19,20. The results of this study suggest that cachectic patients with GIC who carry the C allele exhibit increased activation of the ACVR2B receptor by TGF-β family proteins. This activation, in turn, leads to increased phosphorylation and activation of the SMAD2 and SMAD3 transcription factors, an event that result in muscle atrophy through the activation of MuRF-1 and MAFBX. However, more studies are needed to understand the functional effect of an individual genetic SNP. In agreement with our results, a previous study of cachectic patients with different types of tumors demonstrated that the rs2268757 SNP in the ACVR2B gene is associated with weight loss and a reduction in skeletal muscle mass¹⁵. Studies on tumor-bearing mice demonstrated that the blockade of ACVR2B ligands attenuates weight loss and skeletal muscle atrophy, in addition to improving survival due to the maintenance of muscle mass, highlighting the important role of the receptor in cachexia^21,22.

This study did not find an association of SNPs in the FOXO3 or SELP gene with the weight loss phenotype in cachexia. A probable explanation is that most studies on these genes involved European populations whose genetic profiles differ from that of the Brazilian population as a result of migratory factors, genetic drift, and mutations²³. Furthermore, our SNP analysis did not cover the full extent of the FOXO3 and SELP genes, thereby excluding other mutations that could have functional relevance. Another study from our research group involving cachectic patients with breast cancer showed that rs1935949 in the FOXO3 gene was associated with weight loss in these patients, suggesting a possible association of this SNP with the type of cancer (the data is in submission).

Our findings also showed elevated levels of adiponectin in cachectic patients with GIC. Data in the literature on adiponectin levels in cachexia are controversial. One study also reported high levels of adiponectin in cachectic patients with GIC when compared to non-cachectic patients²⁴. In contrast, previous studies did not find a causal relationship between elevated adiponectin levels and weight loss^25–29. This divergence in the literature might be attributed to the lack of uniform methodology, especially in the diagnosis of cachexia, as adiponectin levels have an inverse relationship with adiposity²⁸.

Elevated levels of GDF-15 have been reported in patients with lung, gastrointestinal, and liver cancer^30,31. Elevation in this cytokine has been associated with symptoms of anorexia, increased inflammation, low performance scores, reduced overall survival and increased weight loss, common characteristics of cachexia^31–33. Consistent with the literature, the cachectic patients studied here had elevated levels of GDF-15. However, no difference in adiponectin or GDF-15 levels was observed between male and female patients.

Ponsegromab, a selective GDF-15 inhibitor, is currently considered a promising candidate for the treatment of cachexia. In a phase Ib clinical trial involving cachectic patients with pancreatic, metastatic colorectal, and non-small cell lung cancer undergoing chemotherapy, Ponsegromab was administered subcutaneously for 12 weeks. Treatment resulted in weight gain of 6.6% compared to baseline and improved appetite and physical capacity³⁴. However, the lack of a placebo control and blinding represents a significant limitation of that study.

Our study has some limitations that must considered when interpreting the data, such as the lack of tomography used for diagnosing cachexia, the small sample size, and the wide confidence interval, which suggests the need for additional studies to refine the results. However, taken together, the findings involving the SNP in the ACVR2B gene, GDF-15 levels, and adiponectin significantly contribute to future investigations of potential biomarkers for GIC-associated cachexia.

Methods

All methods were performed in accordance with the relevant guidelines and regulations.

• Study population and clinical variables.

Patients were divided into three groups based on weight loss: cachectic, non-cachectic, and controls. The diagnosis of cachexia followed the criteria of Fearon et al¹⁰.. We included adult and elderly patients of both sexes diagnosed with GIC, with or without metabolic changes, who exhibited weight loss ≥ 5% after the start of chemotherapy (cachectic). Criteria for exclusion were the absence of GIC, weight loss and chemotherapy, and the presence of liver failure, kidney failure, HIV, inflammatory bowel diseases and autoimmune diseases. The same inclusion criteria was applied to the non-cachectic group but these patients showed stable weight or weight gain after the start of chemotherapy. The control group consisted of individuals without cancer.

Data on comorbidities, tumor stage, chemotherapy agents, surgeries, medications, and metastases were obtained from the patients’ medical records. In addition, height and weight were collected at cancer diagnosis and after the start of chemotherapy to calculate the body mass index (BMI).

The study was approved by the Research Ethics Committee of the University of Ribeirão Preto (UNAERP) (protocol number 5.968.150), São Paulo, Brazil. All participants provided written informed consent.

• Genotyping.

DNA samples were extracted using the Illustra Blood Genomic Prep Mini Spin Kit (GE Healthcare, USA) following manufacturer’s instructions. Genotyping was performed by real-time PCR using the TaqMan SNP Genotyping Assays (Sigma-Aldrich, St. Louis, MO, USA) on the Mx3300 qPCR system (Stratagene, San Diego, CA, USA). The candidate SNPs [FOXO3 (rs1935949 > G), FOXO3 (rs4946935 > A), ACVR2B (rs2268757 > C), and SELP (rs6136 > G)] were selected based on the cachexia phenotype^15,35.

• Measurement of adiponectin and GDF-15.

ELISA kits were used for the measurement of plasma concentrations of adiponectin (Sigma Aldrich, USA) and GDF-15/MIC-1 (#DGD150; Sigma Aldrich, USA).

• Statistical analyses.

Data are reported as mean ± standard deviation or number and percentages, as appropriate. Data were tested for normality by the Kolmogorov-Smirnov test. For multiple comparisons, the Kruskal-Wallis test followed by Dunn’s post-hoc test was used. Two categorical variables were compared by two-way ANOVA followed by Bonferroni’s post-hoc test. The correlation between two continuous variables was analyzed using Spearman’s correlation test. Differences in genotype frequencies were assessed using Fisher’s exact test. Statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, San Diego, CA, USA) and a significance level of P < 0.05 was adopted for all hypotheses tested.

Author contributions

L.M.C. contributed to participant selection, conducted experiments, analyzed data, and wrote the manuscript. G.G.S. assisted with participant selection and experiments. L.C.G.F. and J.S.R.M. contributed to the genotyping experiments. R.B reviewed the manuscript. A.F., M.M., and A.L.F. contributed to the experimental design, supervised the study, and reviewed the manuscript. All authors approved the final version of the manuscript.

Funding

FAPESP grant number 2023/15218-3.

Data availability

The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Research Ethics Committee of the University of Ribeirão Preto (UNAERP) (protocol number 5.968.150), São Paulo, Brazil. Written informed consent was obtained from all participants in the study.

Competing interests

The authors declare no competing interests.