Abstract
Purpose
Oral mucositis (OM) is a common and debilitating adverse effect associated with chemoradiotherapy in nasopharyngeal carcinoma. This study aimed to investigate the association between Latexin (LXN) polymorphism and acute toxicity of oral mucositis.
Methods
A total of 238 nasopharyngeal carcinoma (NPC) patients were enrolled. LXN genotypes were analyzed by the Sequenom MassARRAY system. Multivariate logistic regression was performed to assess the association of LXN polymorphisms and chemoradiotherapy-induced toxicities. Multifactor and generalized multifactor dimensionality reduction methods were applied to calculate the SNP-SNP interaction.
Results
Our study showed that the frequency of the AG genotype of rs1492908 was significantly lower in the grade 3–4 oral mucositis group compared to the grade 0–2 group (14.1% vs. 27.1%, respectively). Patients carrying the LXN rs1492908 AG genotype exhibited a decreased risk of severe oral mucositis (OR = 0.452, 95% CI = 0.213–0.959, P = 0.039). Stratification analysis further revealed that the rs1492908 AG genotype conferred protective effects against oral mucositis in specific patient populations, including those aged ≥ 47 years (OR = 0.340, P = 0.041), body mass index (BMI) ≥ 24 (OR = 0.286, P = 0.035), WHO type Ⅲ histology (OR = 0.212, P = 0.007), and receiving a higher radiotherapy dose (planning gross tumor volume of nasopharynx (pGTVnx) ≥ 71 Gy) (OR = 0.158, P = 0.016). Additionally, SNP-SNP interaction analysis identified that the combination of rs1492908, rs9841, rs8455, rs2639655, and rs56321207 was the best multi-locus model for oral mucositis.
Conclusion
This study is the first to establish a link between NPC chemoradiotherapy-induced oral mucositis risk and LXN polymorphisms in the Chinese Han population.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00520-026-10378-7.
Keywords: Nasopharyngeal carcinoma, LXN, Single nucleotide polymorphism, Toxicity reaction, Chinese Han population
Introduction
Nasopharyngeal carcinoma (NPC) is one of the most aggressive head and neck cancers with unique, highly uneven endemic distributions, mainly in East and Southeast Asia [1]. Definitive radiotherapy with concomitant chemotherapy is the predominant treatment for locoregionally advanced NPC. Despite improvements in chemoradiotherapy techniques, their non-specific cytotoxicity causes substantial damage to adjacent tissues and cells, resulting in a high incidence of severe adverse reactions [2]. Oral mucositis (OM) is a common and debilitating adverse effect associated with chemoradiotherapy in head and neck tumors. The overall incidence rate for OM ranges from 50 to 100%, and approximately 30 to 50% of patients experienced severe mucositis [3–7]. Compared with radiotherapy or chemotherapy alone, patients undergoing concurrent chemoradiotherapy have a higher risk of developing OM [8, 9]. OM is mainly characterized by congestion, erythema, and ulceration during the acute phase, which may develop into atrophy and necrosis in the chronic phase. Severe OM causes oral pain, impairing patients’ ability to eat, swallow, and speak. Moreover, the disruption of the oral mucosal barrier increases the risk of systemic infections and may lead to reduced treatment adherence and treatment interruptions. Therefore, identifying patients at high risk of severe OM early is crucial for optimizing management and preventive measures.
Although OM is common, its severity varies markedly among patients receiving similar treatment protocols. As a multifactorial pathological condition, OM is affected by demographics, tumor classification, and chemoradiotherapy regimen [10]. Genomics studies have indicated that genetic variants are associated with adverse effects of chemoradiotherapy [11–13]. Single nucleotide polymorphisms (SNPs), the most common type of genetic variation in DNA sequences, serve as key determinants in diverse biological pathways and account for individual differences. To date, most studies have mainly focused on SNPs in DNA damage repair pathways, whereas the contribution of inflammation-related genes to OM remains insufficiently explored.
OM triggers inflammatory responses, followed by DNA damage repair. The development of OM is closely linked to the increased expression of proinflammatory cytokines, such as TNF-α, IL-1β, and IL-6 [14, 15]. The LXN gene, located on chromosome 3q25.32 and encoding a 222-amino-acid cytosolic protein, has been described as a potential tumor suppressor and is downregulated in several malignancies [16–18]. LXN overexpression has also been shown to sensitize myeloid cells to radiation [19]. Furthermore, recent studies suggest that LXN may participate in inflammatory and cellular stress responses, as it is enriched in mast cells, upregulated by lipopolysaccharides, and capable of modulating IκBα stability and NF-κB activity [20]. However, the potential relevance of LXN to chemoradiotherapy-induced toxicity has not yet been investigated, representing an important knowledge gap.
Tailoring preventive strategies for OM based on genetic susceptibility is an important step toward individualized care. Although several genetic predictors of treatment-related toxicity have been proposed, the overall genetic background of radiation-induced normal tissue injury remains incompletely defined. As inflammation is central to OM pathogenesis and LXN may be involved in inflammatory and stress-response pathways, we sought to examine whether polymorphisms in LXN are associated with the risk of acute OM in NPC patients undergoing chemoradiotherapy.
Materials and methods
Study population
A total of 238 patients were newly diagnosed with NPC and subsequently treated with chemoradiotherapy from Hunan Cancer Hospital and The Affiliated Cancer Hospital of Xiangya School of Medicine, Central South University. The inclusion criteria were as follows: patients pathologically confirmed as NPC; patients were treated with intensity-modulated radiation therapy (IMRT), induction chemotherapy (IC), and platinum-based concurrent chemoradiotherapy (CCRT); the Karnofsky score ≥ 70; samples had complete clinical information and follow-ups; and patients had no presence of distant metastasis, recurrence, and other previous or concomitant malignant malignancy. Exclusion criteria include patients who are pregnant or lactating and those with active infections prior to treatment. The study was approved by the Independent Ethics Committee of the Institute of Clinical Pharmacology, Central South University (CTXY-14000702). All participants in the study signed informed written consent forms for the use of their clinical data in scientific research prior to treatment. All experimental methods were performed following the relevant guidelines and regulations.
Smoking and drinking status were abstracted from structured electronic health record fields and harmonized for analysis. “Smokers” included both current smokers and former smokers with a documented history of cessation. “Non-smokers” were patients explicitly documented as never having smoked. “Drinkers” included current regular drinkers and former drinkers with a recorded history of alcohol use, while “non-drinkers” included never drinkers and those reporting only occasional alcohol intake a few times per year.
Toxic reactions evaluation
All patients were evaluated for the acute CCRT-induced toxic reactions, including oral mucositis, dermatitis, leukopenia, myelosuppression, thrombocytopenia, neutropenia, and anemia during the CCRT treatment period. All toxic reactions were evaluated and classified as 0–4 grade according to the Common Terminology Criteria for Adverse Events (CTCAE 3.0). Grades 0–2 were considered mild toxic reactions, and grades 3–4 were thought to be severe toxic reactions.
Candidate SNP selection
The SNPs in LXN were selected according to the following criteria: (1) with a minor allele frequency (MAF) ≥ 0.05 in a Southern Han Chinese population; (2) with regulatory features identified by HapMap, ENCODE, Ensembl, and other databases. Finally, we chose nine SNPs (rs9841, rs8455, rs7624934, rs1492908, rs6785658, rs2639655, rs56321207, rs4680455, rs17630607) of LXN that have not been reported in NPC for investigation.
DNA extraction and genotyping
Peripheral blood samples (3 ml) from each participant were collected by using an EDTA anti-coagulative tube before treatment. Genomic DNA was extracted using the QIAamp DNA Blood Mini Kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions and then stored at −80 ℃ for genotyping. The concentration and purity of DNA were measured by UV spectrophotometer at 260 nm and 280 nm. Sequenom MassARRAY iPLEX platform (Sequenom, Inc., San Diego, CA, USA) was used for SNP analysis. The detection rate of all SNPs was greater than 95%.
Bioinformatic analysis
We explored the effects of SNPs on the expression of LXN through the expression quantitative trait loci (eQTL) analysis by Genotype-Tissue Expression (GTEx) database (https://www.gtexportal.org). The possible functions of SNPs were predicted using HaploReg (version 4.2) (https://pubs.broadinstitute.org).
Linkage disequilibrium (LD) analysis
Pairwise LD among the nine LXN SNPs was evaluated using Haploview v4.2 based on D′ and r2 statistics, and LD block structures were defined using the confidence interval method.
Statistical analysis
All statistical analyses were performed with the Statistical Package for Social Sciences software package (version 19.0 for Windows; Chicago, IL, USA). Continuous and categorical variables are reported as mean ± standard deviation and frequency (%), respectively. The genotype frequencies of each LXN polymorphism were evaluated for Hardy-Weinberg equilibrium. The association between SNPs and toxic reactions was calculated using multivariate logistic regression, with results expressed as odds ratios (ORs) and 95% confidence intervals (CIs) under various genetic models. Multivariate logistic regression models were adjusted for age, sex, BMI, smoking status, drinking status, histological type, clinical stage, T-staging, N-staging, and irradiation dose, which were selected as potential confounders based on prior evidence of their influence on treatment-related toxicity. The interaction of SNPs on toxic risk was analyzed using multifactor dimensionality reduction (MDR) software (version 3.0.2) and generalized multifactor dimensionality reduction (GMDR) software (version 0.7). P value < 0.05 was considered the cutoff for statistical significance.
Results
Sample characteristics
A total of 238 NPC patients (169 males and 69 females) are included in this study. The clinical and demographic traits are summarized in Table 1. The mean age and BMI of patients were 47.71 ± 8.75 years and 23.29 ± 3.31, respectively. Two hundred sixteen patients (90.8%) were diagnosed at late stages (III and IV), while the remaining 22 patients (9.2%) were at the early stages (I and II). The regimens of IC included docetaxel + cisplatin/nedaplatin (106 cases), 5-fluorouracil + cisplatin/nedaplatin (38 cases), and paclitaxel + cisplatin/nedaplatin (94 cases). NDP was the most used CCRT regimen, and the median total radiation dose was 71.42 ± 3.54 Gy. Toxic analysis revealed that 72 (30.3%) participants developed a grade 3–4 OM. The mean age and BMI were similar between the two groups. Both cohorts’ clinical and demographic traits were evaluated (Table 1).
Table 1.
Patient demographics and clinical characteristics
| Patient characteristics | All patients, n = 238 (%) | Mucositis | |
|---|---|---|---|
| Grade ≤ 2, n = 166 | Grade > 2, n = 72 | ||
| Gender | |||
| Male | 169 (71.0) | 121 (72.9) | 48 (66.7) |
| Female | 69 (29.0) | 45 (27.1) | 24 (33.3) |
| Age, mean ± SD | 47.71 ± 8.75 | 47.73 ± 8.51 | 47.65 ± 9.36 |
| < 47 years | 101 (42.4) | 71 (42.8) | 30 (41.7) |
| ≥ 47 years | 137 (57.6) | 95 (57.2) | 42 (58.3) |
| BMI, mean ± SD | 23.29 ± 3.31 | 23.07 ± 3.04 | 23.78 ± 3.85 |
| < 18.5 | 13 (5.5) | 8 (4.8) | 5 (6.9) |
| 18.5~24 | 126 (52.9) | 92 (55.4) | 34 (47.2) |
| ≥ 24 | 99 (41.6) | 66 (39.8) | 33 (45.8) |
| Smoking status | |||
| Smoker | 118 (49.6) | 83 (50.0) | 37 (51.4) |
| Nonsmoker | 120 (50.4) | 83 (50.0) | 35 (48.6) |
| Drinking status | |||
| Drinker | 43 (18.9) | 135 (81.3) | 60 (83.3) |
| Nondrinker | 195 (81.9) | 31 (18.7) | 12 (16.7) |
| Histological type | |||
| WHO type Ⅱ | 100 (42.0) | 70 (42.2) | 30 (41.7) |
| WHO type Ⅲ | 138 (58.0) | 96 (57.8) | 42 (58.3) |
| Clinical stage | |||
| Ⅰ–Ⅱ | 22 (9.2) | 15 (9.0) | 7 (9.7) |
| Ⅲ–Ⅳ | 216 (90.8) | 151 (91.0) | 65 (90.3) |
| T-staging | |||
| T1–T2 | 107 (45.0) | 77 (46.4) | 30 (41.7) |
| T3–T4 | 131 (55.0) | 89 (53.6) | 42 (58.3) |
| N-staging | |||
| N0–N1 | 41 (17.2) | 29 (17.5) | 12 (16.7) |
| N2–N3 | 197 (82.8) | 137 (82.5) | 60 (83.3) |
| IC regimen | |||
| DP | 106 (44.5) | 73 (44.0) | 33 (45.8) |
| FP | 38 (16.0) | 29 (17.5) | 9 (12.5) |
| TP | 94 (39.5) | 64 (38.6) | 30 (41.7) |
| CCRT regimen | |||
| FP | 36 (15.1) | 27 (16.3) | 9 (12.5) |
| TP | 57 (23.9) | 37 (22.3) | 20 (27.8) |
| DDP | 48 (20.2) | 33 (19.9) | 15 (20.8) |
| NDP | 64 (26.9) | 42 (25.3) | 22 (30.6) |
| DP | 33 (13.9) | 27 (16.3) | 6 (8.3) |
| pGTVnx, mean ± SD | 71.42 ± 3.54 | 71.50 ± 3.39 | 71.25 ± 3.86 |
| < 71.00 Gy | 100 (42.0) | 71 (42.8) | 36 (50.0) |
| ≥ 71.00 Gy | 138 (58.0) | 95 (57.2) | 36 (50.0) |
Abbreviations: SD, standard deviation; BMI, body mass index; CCRT, concurrent chemoradiotherapy; IC, induction chemotherapy; DP, docetaxel + cisplatin/nedaplatin; FP, 5- fluorouacil + cisplatin/nedaplatin; TP, paclitaxel + cisplatin/nedaplatin. GP, gemcitabine + cisplatin/nedaplatin; DDP, cisplatin alone; NDP, nedaplatin alone; pGTVnx, planning gross tumor volume of nasopharynx
Association between LXN polymorphisms and CCRT-related toxicity risk
The basic information and allele frequency distribution of SNPs in LXN are shown in Table 2. All genotype distributions were in accordance with Hardy-Weinberg equilibrium (P > 0.05). Based on HaploReg, these SNPs exhibit regulatory features that may be related to the regulation of expression, motif changes, and other factors. LD analysis identified a single block of approximately 4 kb containing seven SNPs, whereas the remaining two SNPs showed low LD with this block. This pattern indicates that the selected variants capture both the major haplotype structure and additional independent variation within the LXN region (Supplementary Fig. 1). Based on this linkage disequilibrium structure, haplotype analysis of the seven LXN SNPs within the block was performed to evaluate their combined association with chemoradiotherapy-induced toxicities; however, no statistically significant associations were observed (Supplementary Table 1).
Table 2.
The basic information about candidate SNPs of LXN
| SNP | Location | Allelesa | HWE | MAF | Detectable Rate (%) | HaploReg v4.2 |
|---|---|---|---|---|---|---|
| rs9841 | chr3:158672556–158672656 | A/T | 0.55 | 0.148 | 100.0 | Promoter histone marks, DNAse, Proteins Bound, Motifs changed |
| rs8455 | chr3:158670941–158671041 | T/C | 0.78 | 0.352 | 100.0 | Promoter histone marks, Enhancer histone marks |
| rs7624934 | chr3:158671933–158672033 | C/T | 0.55 | 0.148 | 99.6 | Promoter histone marks, Enhancer histone marks, DNAse, Motifs changed |
| rs1492908 | chr3:158668668–158668768 | G/A | 0.59 | 0.148 | 99.6 | Enhancer histone marks, DNAse, Proteins bound, Motifs changed |
| rs6785658 | chr3:158670555–158670655 | T/G | 0.55 | 0.186 | 100.0 | Enhancer histone marks, Motifs changed |
| rs2639655 | chr3:158668177–158668277 | G/C | 0.61 | 0.486 | 99.6 | Enhancer histone marks |
| rs56321207 | chr3:158666894–158666994 | C/A | 0.72 | 0.038 | 100.0 | Enhancer histone marks, Motifs changed |
| rs4680455 | chr3:158667044–158667144 | C/A | 0.13 | 0.205 | 96.2 | Enhancer histone marks, Motifs changed |
| rs17630607 | chr3:158672021–158672121 | C/T | 0.86 | 0.105 | 100.0 | Promoter histone marks, Enhancer histone marks, DNAse, Motifs changed |
aIn the order of wild/mutant
bData source: HaploReg v4.2
Abbreviations: HWE, Hardy–-Weinberg equilibrium; MAF, minor allele frequency (South Han Chinese)
The relationship between LXN polymorphisms and the risk of OM was analyzed by logistic regression analysis adjusted for confounding factors, as presented in Table 3. The results showed that the frequency of the AG genotype of rs1492908 in the OM grade 3–4 group was significantly lower in the grade 0–2 group (14.1% vs. 27.1%, respectively). Individuals carrying the heterozygous AG genotypes of rs1492908 exhibited a decreased risk of severe OM, with OR of 0.452 (95% CI = 0.213–0.959, P = 0.039). This significant association was also observed in the over-dominant (OR = 0.441, 95% CI = 0.208–0.934, P = 0.033) model. Moreover, the rs9841 and rs6785658 were associated with the development of grade 3 mucositis with borderline statistical significance. The AT genotype for rs9841 decreased the risk of OM 2.06-fold in the over-dominant model. Similarly, in the over-dominant model, a protective role of rs6785658 was found in the TG group (OR = 0.485, P = 0.051). To account for multiple comparisons, Bonferroni correction was applied based on the number of SNPs analyzed. After this conservative adjustment, none of the associations remained statistically significant. As shown in Supplementary Table 2 and Supplementary Fig. 2, no significant associations were observed between the selected LXN SNPs and CCRT-induced hematological toxicities or dermatitis.
Table 3.
Association between LXN genotypes and CCRT -induced oral mucositis in NPC patients
| Model | Genotype | Grade≤2, n (%) |
Grade＞2, n (%) |
OR (95% CI) | P |
|---|---|---|---|---|---|
| rs9841 | |||||
| Codominant | AA | 120 (72.3) | 59 (81.9) | 1.00 (reference) | |
| AT | 45 (27.1) | 11 (15.3) | 0.497 (0.240–1.031.240.031) | 0.060 | |
| TT | 1 (0.6) | 2 (2.8) | 4.068 (0.361–45.773.361.773) | 0.260 | |
| Dominant | AA | 120 (72.3) | 59 (81.9) | 1.00 (reference) | |
| AT+TT | 46 (27.7) | 13 (18.1) | 0.543 (0.263–1.122.263.122) | 0.099 | |
| Recessive | AA+AT | 165 (99.4) | 70 (97.2) | 1.00 (reference) | |
| TT | 1 (0.6) | 2 (2.8) | 4.848 (0.368–63.855.368.855) | 0.230 | |
| Over-dominant | AA+TT | 121 (72.9) | 61 (84.7) | 1.00 (reference) | |
| AT | 45 (27.1) | 11 (15.3) | 0.485 (0.234–1.004.234.004) | 0.051 | |
| Allele | A | 285 (85.9) | 129 (89.6) | 1.00 (reference) | |
| T | 47 (14.1) | 15 (10.4) | 0.681 (0.360–1.291.360.291) | 0.239 | |
| rs8455 | |||||
| Codominant | TT | 79 (47.6) | 38 (52.8) | 1.00 (reference) | |
| TC | 75 (45.2) | 26 (36.1) | 0.690 (0.373–1.275.373.275) | 0.236 | |
| CC | 12 (7.2) | 8 (11.1) | 1.468 (0.518–4.158.518.158) | 0.470 | |
| Dominant | TT | 79 (47.6) | 38 (52.8) | 1.00 (reference) | |
| TC+CC | 87 (52.4) | 34 (47.2) | 0.650 (0.360–1.175.360.175) | 0.154 | |
| Recessive | TT+TC | 154 (92.8) | 64 (88.9) | 1.00 (reference) | |
| CC | 12 (7.2) | 8 (11.1) | 1.748 (0.642–4.763.642.763) | 0.275 | |
| Over-dominant | TT+CC | 91 (54.8) | 46 (63.9) | 1.00 (reference) | |
| TC | 75 (45.2) | 26 (36.1) | 0.650 (0.360–1.175.360.175) | 0.154 | |
| Allele | T | 233 (70.2) | 102 (70.9) | 1.00 (reference) | |
| C | 99 (29.8) | 42 (29.1) | 0.963 (0.617–1.502.617.502) | 0.866 | |
| rs7624934 | |||||
| Codominant | CC | 120 (72.3) | 58 (81.7) | 1.00 (reference) | |
| CT | 45 (27.1) | 11 (15.5) | 0.506 (0.244–1.049.244.049) | 0.067 | |
| TT | 1 (0.6) | 2 (2.8) | 4.138 (0.368–46.573.368.573) | 0.250 | |
| Dominant | CC | 120 (72.3) | 58 (81.7) | 1.00 (reference) | |
| CT+TT | 46 (27.7) | 13 (18.3) | 0.555 (0.268–1.149.268.149) | 0.113 | |
| Recessive | CC+CT | 165 (99.4) | 69 (97.2) | 1.00 (reference) | |
| TT | 1 (0.6) | 2 (2.8) | 4.830 (0.364–64.137.364.137) | 0.233 | |
| Over-dominant | CC+TT | 121 (72.9) | 60 (84.5) | 1.00 (reference) | |
| CT | 45 (27.1) | 11 (15.5) | 0.493 (0.238–1.021.238.021) | 0.057 | |
| Allele | C | 285 (85.9) | 127 (89.5) | 1.00 (reference) | |
| T | 47 (14.1) | 15 (10.5) | 0.694 (0.366–1.315.366.315) | 0.262 | |
| rs6785658 | |||||
| Codominant | TT | 120 (72.3) | 59 (81.9) | 1.00 (reference) | |
| TG | 45 (27.1) | 11 (15.3) | 0.497 (0.240–1.031.240.031) | 0.060 | |
| GG | 1 (0.6) | 2 (2.8) | 4.068 (0.361–45.773.361.773) | 0.256 | |
| Dominant | TT | 120 (72.3) | 59 (81.9) | 1.00 (reference) | |
| TG+GG | 46 (27.7) | 13 (18.1) | 0.543 (0.263–1.122.263.122) | 0.099 | |
| Recessive | TT+TG | 165 (99.4) | 70 (97.2) | 1.00 (reference) | |
| GG | 1 (0.6) | 2 (2.8) | 4.848 (0.368–63.855.368.855) | 0.230 | |
| Over-dominant | TT+GG | 121 (72.9) | 61 (84.7) | 1.00 (reference) | |
| TG | 45 (27.1) | 11 (15.3) | 0.485 (0.234–1.004.234.004) | 0.051 | |
| Allele | T | 285 (85.9) | 129 (89.6) | 1.00 (reference) | |
| G | 47 (14.1) | 15 (10.4) | 0.681 (0.360–1.291.360.291) | 0.239 | |
| rs1492908 | |||||
| Codominant | AA | 120 (72.3) | 59 (83.1) | 1.00 (reference) | |
| AG | 45 (27.1) | 10 (14.1) | 0.452 (0.213–0.959.213.959) | 0.039 | |
| GG | 1 (0.6) | 2 (2.8) | 4.068 (0.361–45.773.361.773) | 0.256 | |
| Dominant | AA | 120 (72.3) | 59 (83.1) | 1.00 (reference) | |
| AG+GG | 46 (27.7) | 12 (16.9) | 0.531 (0.261–1.077.261.077) | 0.079 | |
| Recessive | AA+AG | 165 (99.4) | 69 (97.2) | 1.00 (reference) | |
| GG | 1 (0.6) | 2 (2.8) | 4.850 (0.365–64.540.365.540) | 0.232 | |
| Over-dominant | AA+GG | 121 (72.9) | 61 (85.9) | 1.00 (reference) | |
| AG | 45 (27.1) | 10 (14.1) | 0.441 (0.208–0.934.208.934) | 0.033 | |
| Allele | A | 285 (85.9) | 128 (90.2) | 1.00 (reference) | |
| G | 47 (14.1) | 14 (9.8) | 0.645 (0.335–1.241.335.241) | 0.189 | |
| rs2639655 | |||||
| Codominant | GG | 57 (34.3) | 25 (35.2) | 1.00 (reference) | |
| GC | 85 (51.2) | 33 (46.5) | 0.930 (0.483–1.789.483.789) | 0.827 | |
| CC | 24 (14.5) | 13 (18.3) | 1.305 (0.542–3.139.542.139) | 0.552 | |
| Dominant | GG | 57 (34.3) | 25 (35.2) | 1.00 (reference) | |
| GC+CC | 109 (65.7) | 46 (64.8) | 1.010 (0.543–1.876.543.876) | 0.976 | |
| Recessive | GG+GC | 142 (85.5) | 58 (81.7) | 1.00 (reference) | |
| CC | 24 (14.5) | 13 (18.3) | 1.363 (0.620–2.994.620.994) | 0.441 | |
| Over-dominant | GG+CC | 81 (48.8) | 38 (53.5) | 1.00 (reference) | |
| GC | 85 (51.2) | 33 (46.5) | 0.853 (0.474–1.534.474.534) | 0.596 | |
| Allele | G | 199 (59.9) | 83 (58.5) | 1.00 (reference) | |
| C | 133 (40.1) | 59 (41.5) | 1.093 (0.721–1.656.721.656) | 0.676 | |
| rs56321207 | |||||
| Codominant | CC | 157 (94.6) | 70 (97.2) | 1.00 (reference) | |
| CA | 9 (5.4) | 2 (2.8) | 0.523 (0.103–2.648.103.648) | 0.434 | |
| AA | 0 | 0 | / | / | |
| Dominant | CC | 157 (94.6) | 70 (97.2) | 1.00 (reference) | |
| CA+AA | 9 (5.4) | 2 (2.8) | 0.523 (0.103–2.648.103.648) | 0.434 | |
| Recessive | CC+CA | 166 (100) | 72 (100) | 1.00 (reference) | |
| AA | 0 | 0 | / | / | |
| Over-dominant | CC+AA | 157 (94.6) | 70 (97.2) | 1.00 (reference) | |
| CA | 9 (5.4) | 2 (2.8) | 0.523 (0.103–2.648.103.648) | 0.434 | |
| Allele | C | 323 (97.3) | 142 (98.6) | 1.00 (reference) | |
| A | 9 (2.7) | 2 (1.4) | 0.540 (0.111–2.638.111.638) | 0.447 | |
| rs4680455 | |||||
| Codominant | CC | 112 (70.4) | 45 (64.3) | 1.00 (reference) | |
| CA | 45 (28.3) | 24 (34.3) | 1.340 (0.710–2.529.710.529) | 0.366 | |
| AA | 2 (1.3) | 1 (1.4) | 1.614 (0.126–20.689.126.689) | 0.713 | |
| Dominant | CC | 112 (70.4) | 45 (64.3) | 1.00 (reference) | |
| CA+AA | 47 (29.6) | 25 (35.7) | 1.351 (0.723–2.524.723.524) | 0.346 | |
| Recessive | CC+CA | 157 (98.7) | 69 (98.6) | 1.00 (reference) | |
| AA | 2 (1.3) | 1 (1.4) | 1.484 (0.116–18.942.116.942) | 0.761 | |
| Over-dominant | CC+AA | 114 (71.7) | 46 (65.7) | 1.00 (reference) | |
| CA | 45 (28.3) | 24 (34.3) | 1.329 (0.706–2.502.706.502) | 0.378 | |
| Allele | C | 269 (84.6) | 114 (81.5) | 1.00 (reference) | |
| A | 49 (15.4) | 26 (18.5) | 1.279 (0.743–2.201.743.201) | 0.374 | |
| rs17630607 | |||||
| Codominant | CC | 129 (77.7) | 61 (84.7) | 1.00 (reference) | |
| CT | 36 (21.7) | 9 (12.5) | 0.491 (0.215–1.122.215.122) | 0.092 | |
| TT | 1 (0.6) | 2 (2.8) | 4.224 (0.323–55.163.323.163) | 0.272 | |
| Dominant | CC | 129 (77.7) | 61 (84.7) | 1.00 (reference) | |
| CT+TT | 37 (22.3) | 11 (15.3) | 0.584 (0.270–1.266.270.266) | 0.173 | |
| Recessive | CC+CT | 165 (99.4) | 70 (97.2) | 1.00 (reference) | |
| TT | 1 (0.6) | 2 (2.8) | 4.848 (0.368–63.855.368.855) | 0.230 | |
| Over-dominant | CC+TT | 130 (78.3) | 63 (87.5) | 1.00 (reference) | |
| CT | 36 (21.7) | 9 (12.5) | 0.516 (0.234–1.137.234.137) | 0.101 | |
| Allele | C | 294 (88.6) | 131 (91) | 1.00 (reference) | |
| T | 38 (11.4) | 13 (9) | 0.727 (0.366–1.443.366.443) | 0.362 |
OR, odds ratio; 95% CI, 95% confidence interval; P < 0.05 indicates SNP with statistical significance. The highlighted in bold represent significant association. ORs were adjusted for age, sex, body mass index, smoking status, drinking status, clinical stage, T-staging, N-staging, histological type, and irradiation dose
Stratification analyses of the LXN rs1492908 genotypes
To further elucidate the association between rs1492908 genotypes and severe OM risk, we conducted stratified analyses based on potential epidemiological and clinical risk factors, including age, sex, BMI, smoking status, drinking status, histological type, clinical stage, T-staging, N-staging, and irradiation dose (Table 4, Fig. 1). We found that older patients in the severe group had a higher AA+GG genotype frequency compared to those in the mild group (88.1% vs. 71.6%). Similarly, the frequency of the risk AA+GG genotype was significantly higher in severe OM cases and mild cases among males in the WHO type Ⅲ group (90.2% vs. 67.7%) and pGTVnx ≥ 71 Gy groups (94.4% vs. 72.6%) (P < 0.05) (Fig. 1).
Table 4.
Stratified analysis of LXN rs1492908 genotypes according to epidemiological and clinical risk factors
| Genotype | Grade≤2, n (%) |
Grade＞2, n (%) |
OR (95% CI) | P | Grade≤2, n (%) |
Grade＞2, n (%) |
OR (95% CI) | P |
|---|---|---|---|---|---|---|---|---|
| Male | Female | |||||||
| AA+GG | 86 (71.1) | 41 (87.2) | 1.00 (reference) | 35 (77.8) | 20 (83.3) | 1.00 (reference) | ||
| AG | 35 (28.9) | 6 (12.8) | 0.360 (0.140–0.923.140.923) | 0.033 | 10 (22.2) | 4 (16.7) | 0.462 (0.090–2.357.090.357) | 0.353 |
| Age＜47 | Age≥47 | |||||||
| AA+GG | 53 (74.6) | 24 (82.8) | 1.00 (reference) | 68 (71.6) | 37 (88.1) | 1.00 (reference) | ||
| AG | 18 (25.4) | 5 (17.2) | 0.688 (0.198–2.390.198.390) | 0.556 | 27 (28.4) | 5 (11.9) | 0.340 (0.121–0.958.121.958) | 0.041 |
| BMI＜24 | BMI≥24 | |||||||
| AA+GG | 77 (77.0) | 33 (84.6) | 1.00 (reference) | 44 (66.7) | 28 (87.5) | 1.00 (reference) | ||
| AG | 23 (23.0) | 6 (15.4) | 0.580 (0.196–1.716.196.716) | 0.325 | 22 (33.3) | 4 (12.5) | 0.286 (0.089–0.917.089.917) | 0.035 |
| Nonsmoker | Smoker | |||||||
| AA+GG | 58 (69.9) | 32 (86.5) | 1.00 (reference) | 63 (75.9) | 29 (85.3) | 1.00 (reference) | ||
| AG | 25 (30.1) | 5 (13.5) | 0.375 (0.128–1.093.128.093) | 0.072 | 20 (24.1) | 5 (14.7) | 0.448 (0.129–1.552.129.552) | 0.205 |
| Nondrinker | Drinker | |||||||
| AA+GG | 98 (72.6) | 50 (84.7) | 1.00 (reference) | 23 (74.2) | 11 (91.7) | 1.00 (reference) | ||
| AG | 37 (27.4) | 9 (15.3) | 0.477 (0.213–1.065.213.065) | 0.071 | 8 (25.8) | 1 (8.3) | / | 1.000 |
| WHO type Ⅱ | WHO type Ⅲ | |||||||
| AA+GG | 56 (80.0) | 24 (80.0) | 1.00 (reference) | 65 (67.7) | 37 (90.2) | 1.00 (reference) | ||
| AG | 14 (20.0) | 6 (20.0) | 1.056 (0.269–4.149.269.149) | 0.938 | 31 (32.3) | 4 (9.8) | 0.212 (0.068–0.659.068.659) | 0.007 |
| Ⅰ-Ⅱ | Ⅲ-Ⅳ | |||||||
| AA+GG | 9 (60.0) | 6 (85.7) | 1.00 (reference) | 112 (74.2) | 55 (85.9) | 1.00 (reference) | ||
| AG | 6 (40.0) | 1 (14.3) | / | 0.996 | 39 (25.8) | 9 (14.1) | 0.470 (0.213–1.039.213.039) | 0.062 |
| T1-T2 | T3-T4 | |||||||
| AA+GG | 57 (74.0) | 26 (86.7) | 1.00 (reference) | 64 (71.9) | 35 (85.4) | 1.00 (reference) | ||
| AG | 20 (26.0) | 4 (13.3) | 0.350 (0.091–1.343.091.343) | 0.126 | 25 (28.1) | 6 (14.6) | 0.415 (0.147–1.171.147.171) | 0.097 |
| N1-N2 | N3-N4 | |||||||
| AA+GG | 16 (55.2) | 10 (90.9) | 1.00 (reference) | 105 (76.6) | 51 (85.0) | 1.00 (reference) | ||
| AG | 13 (44.8) | 1 (9.1) | / | 0.994 | 32 (23.4) | 9 (15.0) | 0.549 (0.232–1.297.232.297) | 0.172 |
| pGTVnx＜71Gy | pGTVnx≥71Gy | |||||||
| AA+GG | 52 (73.2) | 27 (77.1) | 1.00 (reference) | 69 (72.6) | 34 (94.4) | 1.00 (reference) | ||
| AG | 19 (26.8) | 8 (22.9) | 0.959 (0.326–2.816.326.816) | 0.939 | 26 (27.4) | 2 (5.6) | 0.158 (0.035–0.711.035.711) | 0.016 |
OR, odds ratio; 95% CI, 95% confidence interval; P < 0.05 indicates SNP with statistical significance. ORs were adjusted for age, sex, body- mass index, smoking status, drinking status, clinical stage, T-staging, N-staging, histological type, and irradiation dose
Fig. 1.
Genotype distributions of LXN rs1492908 in mild and severe oral mucositis as stratified by age, sex, BMI, smoking status, drinking status, histological type, clinical stage, T-staging, N-staging, and irradiation dose. Abbreviations: BMI, body mass index; pGTVnx, planning gross tumor volume of the nasopharynx
A potential association was evaluated by logistic regression. Compared with rs1492908 AA+GG, the AG genotype exhibited protective effects on the risk of OM in patients aged ≥ 47 years (OR = 0.340, P = 0.041) and overweight patients (BMI ≥ 24) (OR = 0.286, P = 0.035). Similarly, compared to the AG genotype, the risk of severe OM has increased remarkably to 4.717 times in WHO type III histology (P = 0.007). Additionally, patients with the rs1492908 AG genotype showed an 84.2% lower risk compared to those with the AA+GG genotype (P = 0.0116) in the subgroup receiving higher radiation doses.
SNP-SNP interaction analysis
MDR and GMDR were applied to evaluate potential SNP-SNP interactions among the nine selected LXN variants in relation to oral mucositis risk. The hierarchical clustering of SNP interactions is illustrated in the dendrogram (Fig. 2a), where closer clustering reflects stronger interaction patterns. The Fruchterman-Reingold network (Fig. 2b) visualizes both main effects and pairwise interaction effects among the SNPs.
Fig. 2.
SNP-SNP interaction network of LXN variants. a The dendrogram shows the hierarchical interaction pattern among the nine SNPs, where closer clustering indicates stronger interaction. b The Fruchterman-Reingold plot visualizes both main and pairwise interaction effects. Each node represents a SNP, and connecting lines indicate interactions. Orange edges represent synergistic effects, blue edges represent redundant effects, and thicker edges indicate stronger interaction strength
As shown in Table 5, rs1492908 was identified as the best single-locus model (cross-validation consistency [CVC] = 10/10; testing balanced accuracy = 0.5541), with an information gain of 2.15%. The best multi-locus model consisted of rs1492908, rs9841, rs8455, rs2639655, and rs56321207 (CVC = 10/10; testing balanced accuracy = 0.4861).
Table 5.
LXN SNP-SNP interaction models analyzed by the GMDR method
| Best Model | Training Bal. Acc. | Testing Bal. Acc. | CVC | P |
|---|---|---|---|---|
| rs1492908 | 0.5686 | 0.5541 | 10/10 | 0.104 |
| rs1492908, rs2639655 | 0.5750 | 0.4766 | 4/10 | 0.104 |
| rs8455, rs2639655, rs17630607 | 0.5891 | 0.4735 | 7/10 | 0.111 |
| rs1492908, rs8455, rs2639655, rs56321207 | 0.5948 | 0.4861 | 9/10 | 0.028 |
| rs1492908, rs9841, rs8455, rs2639655, rs56321207 | 0.5948 | 0.4861 | 10/10 | 0.028 |
| rs1492908, rs9841, rs8455, rs7624934, rs2639655, rs56321207 | 0.5948 | 0.4861 | 10/10 | 0.028 |
| rs1492908, rs9841, rs8455, rs7624934, rs6785658, rs2639655, rs56321207 | 0.5948 | 0.4861 | 10/10 | 0.028 |
| rs1492908, rs9841, rs8455, rs7624934, rs6785658, rs2639655, rs56321207, rs4680455 | 0.5948 | 0.4861 | 10/10 | 0.028 |
| rs1492908, rs9841, rs8455, rs7624934, rs6785658, rs2639655, rs56321207, rs4680455, rs17630607 | 0.5948 | 0.4861 | 10/10 | 0.028 |
GMDR, generalized multifactor dimensionality reduction; Bal. Acc., balanced accuracy; CVC, cross-validation consistency; P < 0.05 represent statistical significance
Impact of candidate SNPs on gene expression
We utilized publicly available GTEx bioinformatics datasets to evaluate the expression loci of each SNP. As shown in Fig. 3, we performed eQTL analyses using GTEx whole-blood data to assess whether the selected SNPs are associated with LXN expression. Five variants (rs9841, rs7624934, rs1492908, rs6785658, and rs17630607) showed significant associations with LXN mRNA levels. Among them, carriers of the rs1492908 GG genotype exhibited higher LXN expression compared with those carrying the AA genotype (Fig. 3a). Furthermore, we used Ensembl to analyze the gene expression affected by rs1492908. The most significantly associated genes were RARRES1, MFSD1, GFM1, MLF1, RSRC1, and SHOX2. A string database was utilized to construct a PPI network of these genes (Fig. 3b).
Fig. 3.
The rs1492908 regulates gene expression. a The GTEx whole-blood eQTL analysis identified variants in the LXN locus (rs9841, rs7624934, rs1492908, rs6785658, and rs17630607) as eQTLs associated with LXN expression. b PPI network analysis of genes regulated by rs1492908. Lines between nodes represent interactions between proteins
Discussion
Oral mucositis is a frequent and debilitating toxicity of chemoradiotherapy in NPC, and its increasing severity significantly compromises patients’ quality of life and treatment adherence. Given the involvement of LXN in inflammation and radiation-induced stress responses, we investigated whether LXN polymorphisms are associated with the risk of acute OM during chemoradiotherapy. In our cohort, the rs1492908 AG genotype was associated with a significantly lower likelihood of developing grade 3–4 OM, suggesting a potential genetic contribution to individual differences in mucosal toxicity.
A great number of studies have indicated SNPs could act as biomarker predictors of OM severity in head and neck cancer patients treated with radiotherapy and chemotherapy. For example, XRCC1 was the most frequently analyzed gene. Nanda et al. reported that patients of HNSCC with XRCC1 Arg194Trp polymorphism have higher acute radiation toxicity [21]. Moreover, these patients also have better progression-free survival compared to the wild variant. Similarly, the risk of mucositis was increased in patients with XRCC1−399Gln allele genotypes in the chemoradiotherapy [22]. However, current research has primarily focused on the DNA damage repair pathway, such as XRCC3 (rs1415120657), Rad51 (c.−3392), and MDM2 (rs2279744) [23, 24], while the exploration of inflammatory pathways remains relatively limited. OM initiates from DNA damage induced by chemoradiotherapy, subsequently triggering the release of inflammatory cytokines and signal amplification. Considering that inhibiting inflammatory pathways may successfully prevent severe OM toxicity, it is crucial to investigate the relationship between inflammation-related genes and OM.
The LXN gene was identified as a new potential tumor suppressor in several types of liquid and solid tumors. In LXN-overexpressing cells, cell cycle arrest and cell death induced by chemotherapeutic agents and radiation were augmented [19]. In addition, LXN expression was inversely correlated with chemotherapy resistance in prostate cancer and breast cancer cell lines [25, 26], which was of great clinical significance for target therapy. Based on this biological rationale, we examined LXN variation in NPC patients. In our analysis, carriers of the rs1492908 AG genotype exhibited a lower frequency of severe OM, with similar patterns observed across subgroup analyses.
Currently, various predictive models, including BMI, have been identified with AUCs ranging from 0.67 to 0.782 for acute OM. However, findings regarding the direction of risk remain inconsistent across studies. For instance, Liang et al. reported that low BMI (< 23.9 kg/m2) was associated with increased OM severity [27], whereas Orlandi et al. observed a higher risk in patients with BMI > 30 kg/m2 [28]. In our cohort, using 24 kg/m2 as the population-specific threshold, patients carrying the rs1492908 AG genotype showed a 71.4% reduced risk of severe OM among those classified as overweight (BMI ≥ 24). Beyond BMI, additional subgroup analyses revealed that the protective association of the AG genotype was also observed in patients with WHO type III histology and in those receiving a higher radiation dose (pGTVnx ≥ 71 Gy), where carriers showed notably reduced occurrence of grade 3–4 OM. These findings may imply that the influence of LXN variation could become more discernible under higher biological or treatment-induced stress.
To further explore biological relevance, SNP-SNP interaction analysis suggested potential synergistic effects among LXN variants, indicating that rs1492908 may not act in isolation. Although these interaction results are exploratory, they support a polygenic contribution to mucosal toxicity rather than a single-SNP effect.
Increasing evidence has indicated that SNPs affect the expression level of genes. Consistent with this hypothesis, we observed higher LXN expression in whole blood among rs1492908 GG genotype carriers compared with AA carriers. Therefore, LXN polymorphisms might influence the occurrence of OM by affecting LXN gene expression. It has been demonstrated that LXN mRNA was expressed in different tissues, and this expression correlated with inflammatory response and acute toxicity [29]. Exposure to ionizing radiation, LXN deletion protects against both radiation-induced acute myelosuppression and long-term hematopoietic stem cell damage. However, LXN overexpression attenuates TNF-α-induced inflammatory response through inhibition of p65 activity [20]. Taken together, these findings suggest a potential link between LXN activity, inflammatory regulation, and tissue injury following chemoradiotherapy; however, direct evidence in the context of OM is still lacking.
To further explore potential regulatory mechanisms, in silico analyses indicated that rs1492908 lies within an intronic region with predicted enhancer activity and may intersect with regulatory elements such as enhancer histone marks, DNase accessibility sites, and protein-binding motifs. Based on these annotations, we found that rs1492908 may be located within an intronic regulatory element that could potentially influence nearby or functionally related genes, including RARRES1, MFSD1, GFM1, MLF1, RSRC1, and SHOX2. RARRES1, a podocyte-specific growth arrest gene, has a tumor suppressor role in human cancers, which may be coordinately regulated with LXN [30–33]. In NPC, RARRES1 was downregulated, and 51~90.7% promoter methylation was detected in primary tumors [30]. RARRES1 overexpression improved tumor cell sensitivity to Lenvatinib through the promotion of SPINK2 expression in hepatocellular carcinoma [31]. MLF1 is aberrantly expressed in human malignancies, which was required for the cells to respond to the apoptotic stimulations [34]. MLF1 upregulation would promote cell proliferation, motility, invasiveness, and apoptosis evasion in intrahepatic cholangiocarcinoma [35]. Although these observations provide biological context for our findings, the relevance of these pathways to rs1492908 remains speculative. Therefore, the potential mechanistic role of this variant in chemoradiotherapy-induced OM requires further experimental validation.
Together, these exploratory findings provide a possible mechanistic context for our observations, although they require careful interpretation. While the findings of this study provide preliminary insight into the potential role of LXN variation in NPC treatment tolerance, several aspects warrant further investigation. The sample size, particularly within some stratified subgroups, limits the strength of inference and highlights the need for larger, prospectively designed cohorts to confirm these associations. Additionally, Bonferroni correction attenuated significance, which may reflect limited power rather than lack of biological relevance. Besides, EBV data were not consistently available in this retrospective setting, preventing further evaluation of potential virus-host interactions. Although current evidence does not clearly indicate a direct link between EBV and acute OM susceptibility, incorporating this variable in future analyses may provide additional biological context. Moreover, as the study population consisted exclusively of Han Chinese patients, validation in genetically diverse populations will be essential to assess generalizability. Finally, the biological interpretation of rs1492908 remains hypothesis-generating, and functional experiments will be required to clarify how this variant may contribute to differential mucosal response during chemoradiotherapy.
In conclusion, our study provided a reference and basis for investigating the association of LXN gene variants and chemoradiotherapy-induced OM risk in NPC patients. Carriers of the rs1492908 AG genotype demonstrated a lower observed risk of severe OM, indicating that LXN variation may serve as a potential candidate in future toxicity-prediction models.
Supplementary Information
Below is the link to the electronic supplementary material.
(DOCX 3.11 MB)
Author contributions
Youhong Wang and Siqing Ma contributed equally to this work. Y.H. Wang was responsible for conceptualization, methodology, supervision, funding acquisition, and manuscript revision. S.Q. Ma performed data analysis, investigation, and drafted the original manuscript. J.C. Bao and Z.H. Zhang contributed to clinical data collection and interpretation. J.C. Song, Y. Tai, and Y.T. Gao assisted with data management and literature review. W. Feng and L. An jointly supervised the study, validated the results, and approved the final manuscript. All authors reviewed and approved the final version and take full responsibility for the content.
Funding
This work was jointly supported by the National Natural Science Foundation of China (No. 82204535), the Natural Science Foundation of Xiamen, China (No. 3502Z20227145), the Science and Technology Plan Project of Baotou Municipal Health Commission (No. wsjkwkj030-2021), and the Scientific Research Fund Project of Baotou Medical College (No. BYJJ-ZRQM 202451).
Data availability
The data supporting the findings of this study are available from the corresponding author upon reasonable request.
Declarations
Ethics approval
This study was approved by the Independent Ethical Committee of the Institute of Clinical Pharmacology, Central South University (Approval No. CTXY-140007–2). All procedures involving human participants were performed in accordance with the Declaration of Helsinki.
Consent to participate
Written informed consent was obtained from all participants.
Consent for publication
Not applicable. The manuscript does not contain any identifiable personal data.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Youhong Wang and Siqing Ma contributed equally to this work.
Contributor Information
Wei Feng, Email: bfyyfskf@163.com.
Liang An, Email: 38841031@qq.com.
References
- 1.Chen Y-P, Chan AT, Le Q-T, Blanchard P, Sun Y, Ma J (2019) Nasopharyngeal carcinoma. Lancet 394(10192):64–80 [DOI] [PubMed] [Google Scholar]
- 2.Cramer JD, Burtness B, Le QT, Ferris RL (2019) The changing therapeutic landscape of head and neck cancer. Nat Rev Clin Oncol 16(11):669–683 [DOI] [PubMed] [Google Scholar]
- 3.Chen L, Hu C-S, Chen X-Z, Hu G-Q, Cheng Z-B, Sun Y et al (2012) Concurrent chemoradiotherapy plus adjuvant chemotherapy versus concurrent chemoradiotherapy alone in patients with locoregionally advanced nasopharyngeal carcinoma: a phase 3 multicentre randomised controlled trial. Lancet Oncol 13(2):163–171 [DOI] [PubMed] [Google Scholar]
- 4.Sun Y, Li W-F, Chen N-Y, Zhang N, Hu G-Q, Xie F-Y et al (2016) Induction chemotherapy plus concurrent chemoradiotherapy versus concurrent chemoradiotherapy alone in locoregionally advanced nasopharyngeal carcinoma: a phase 3, multicentre, randomised controlled trial. Lancet Oncol 17(11):1509–1520 [DOI] [PubMed] [Google Scholar]
- 5.Zhang Y, Chen L, Hu G-Q, Zhang N, Zhu X-D, Yang K-Y et al (2019) Gemcitabine and cisplatin induction chemotherapy in nasopharyngeal carcinoma. N Engl J Med 381(12):1124–1135 [DOI] [PubMed] [Google Scholar]
- 6.Li X-Y, Chen Q-Y, Sun X-S, Liu S-L, Yan J-J, Guo S-S et al (2019) Ten-year outcomes of survival and toxicity for a phase III randomised trial of concurrent chemoradiotherapy versus radiotherapy alone in stage II nasopharyngeal carcinoma. Eur J Cancer 110:24–31 [DOI] [PubMed] [Google Scholar]
- 7.Wu Y, Shi W, Li C, Liu X, Jiang Y, Qiu Y et al (2025) Managing strategies of chemotherapy and radiotherapy-induced oral mucositis. Cancer Treat Rev. 10.1016/j.ctrv.2025.102883 [DOI] [PubMed] [Google Scholar]
- 8.Pulito C, Cristaudo A, Porta CL, Zapperi S, Blandino G, Morrone A et al (2020) Oral mucositis: the hidden side of cancer therapy. J Exp Clin Cancer Res 39:1–15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mohammed AI, Fedoruk L, Fisher N, Liu AX, Khanna S, Naylor K et al (2024) Systemic anti-inflammatory agents in the prevention of chemoradiation-induced mucositis: a review of randomised controlled trials. Biomolecules 14(5):560 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Nicol AJ, Ching JC, Tam VC, Liu KC, Leung VW, Cai J et al (2023) Predictive factors for chemoradiation-induced oral mucositis and dysphagia in head and neck cancer: a scoping review. Cancers 15(23):5705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Cavalieri R, de Oliveira HF, de Louvain Souza T, Kanashiro MM (2024) Single nucleotide polymorphisms as biomarker predictors of oral mucositis severity in head and neck cancer patients submitted to combined radiation therapy and chemotherapy: a systematic review. Cancers 16(5):949 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Yang D-W, Wang T-M, Zhang J-B, Li X-Z, He Y-Q, Xiao R et al (2020) Genome-wide association study identifies genetic susceptibility loci and pathways of radiation-induced acute oral mucositis. J Transl Med 18:1–12 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Guo Z, Wang Y-J, He B-S, Zhou J (2022) Linc00312 single nucleotide polymorphism as biomarker for chemoradiotherapy induced hematotoxicity in nasopharyngeal carcinoma patients. Dis Markers. 10.1155/2022/6707821 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Dai M, Dai X, Liang Y, Li X, Huang H, Zhao W (2025) Keratinocyte necroptosis promotes the progression of radiation-induced oral mucositis. BMC Oral Health 25(1):941 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Yoshida K, Kiyomi A, Kurokawa A, Hoshino N, Saito C, Koga M et al (2024) Association between salivary inflammatory mediators and oral mucositis in patients with cancer undergoing chemotherapy. Support Care Cancer 32(9):625 [DOI] [PubMed] [Google Scholar]
- 16.Ni Q-F, Tian Y, Kong L-L, Lu Y-T, Ding W-Z, Kong L-B (2014) Latexin exhibits tumor suppressor potential in hepatocellular carcinoma. Oncol Rep 31(3):1364–1372 [DOI] [PubMed] [Google Scholar]
- 17.Xue Z, Zhou Y, Wang C, Zheng J, Zhang P, Zhou L et al (2016) Latexin exhibits tumor-suppressor potential in pancreatic ductal adenocarcinoma. Oncol Rep 35(1):50–58 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Li Y, Tan Y, Li X, Chen X, Wang L, Zhang L et al (2022) Loss of LXN promotes macrophage M2 polarization and PD-L2 expression contributing cancer immune-escape in mice. Cell Death Discov 8(1):440 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.You Y, Wen R, Pathak R, Li A, Li W, St Clair D et al (2014) Latexin sensitizes leukemogenic cells to gamma-irradiation-induced cell-cycle arrest and cell death through Rps3 pathway. Cell Death Dis 5(10):e1493-e [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Li Y, Huang B, Yang H, Kan S, Yao Y, Liu X et al (2020) Latexin deficiency in mice up-regulates inflammation and aggravates colitis through HECTD1/Rps3/NF-κB pathway. Sci Rep 10(1):9868 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Nanda S, Gandhi A, Rastogi M, Bhatt M, Sahni K, Mishra S et al (2017) Evaluation of XRCC-1 gene polymorphism as a biomarker in head and neck cancer patients undergoing chemoradiation therapy. Int J Radiat Oncol Biol Phys 99(2):S236. 10.1016/j.ijrobp.2017.06.577 [Google Scholar]
- 22.Pratesi N, Mangoni M, Mancini I, Paiar F, Simi L, Livi L et al (2011) Association between single nucleotide polymorphisms in the XRCC1 and RAD51 genes and clinical radiosensitivity in head and neck cancer. Radiother Oncol 99(3):356–61 [DOI] [PubMed] [Google Scholar]
- 23.Werbrouck J, De Ruyck K, Duprez F, Veldeman L, Claes K, Van Eijkeren M et al (2009) Acute normal tissue reactions in head-and-neck cancer patients treated with IMRT: influence of dose and association with genetic polymorphisms in DNA DSB repair genes. Int J Radiat Oncol Biol Phys 73(4):1187–95 [DOI] [PubMed] [Google Scholar]
- 24.Borchiellini D, Etienne-Grimaldi M, Bensadoun R, Benezery K, Dassonville O, Poissonnet G et al (2017) Candidate apoptotic and DNA repair gene approach confirms involvement of ERCC1, ERCC5, TP53 and MDM2 in radiation-induced toxicity in head and neck cancer. Oral Oncol 67:70–6 [DOI] [PubMed] [Google Scholar]
- 25.Zhang M, Osisami M, Dai J, Keller JM, Escara-Wilke J, Mizokami A et al (2017) Bone microenvironment changes in latexin expression promote chemoresistance. Mol Cancer Res 15(4):457–66 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhang H, Ren Y, Pang D, Liu C (2014) Clinical implications of AGBL2 expression and its inhibitor latexin in breast cancer. World J Surg Oncol 12:1–7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Liang Y, Wang X, Shi X, Fei X (2025) Risk factors and prediction models for severe radiation-induced oral mucositis in patients with nasopharyngeal carcinoma undergoing chemoradiotherapy. Discov Oncol 16(1):1–10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Orlandi E, Iacovelli NA, Rancati T, Cicchetti A, Bossi P, Pignoli E et al (2018) Multivariable model for predicting acute oral mucositis during combined IMRT and chemotherapy for locally advanced nasopharyngeal cancer patients. Oral Oncol 86:266–72 [DOI] [PubMed] [Google Scholar]
- 29.Zhang C, Cui X, Liu Y, Wang F, Signer R, Nattamai K et al (2023) Latexin deletion protects against radiation-induced hematopoietic damages via selective activation of Bcl-2 prosurvival pathway. Haematologica 108(12):3464 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yanatatsaneejit P, Chalermchai T, Kerekhanjanarong V, Shotelersuk K, Supiyaphun P, Mutirangura A et al (2008) Promoter hypermethylation of CCNA1, RARRES1, and HRASLS3 in nasopharyngeal carcinoma. Oral Oncol 44(4):400–6 [DOI] [PubMed] [Google Scholar]
- 31.Guo Y, Chai B, Zhang H, Chai X, Chen Y, Xu J et al (2024) RARRES1 inhibits hepatocellular carcinoma progression and increases its sensitivity to lenvatinib through interaction with SPINK2. Biol Direct 19(1):15 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Oldridge E, Walker H, Stower M, Simms M, Mann V, Collins A et al (2013) Retinoic acid represses invasion and stem cell phenotype by induction of the metastasis suppressors RARRES1 and LXN. Oncogenesis 2(4):e45-e [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Maimouni S, Issa N, Cheng S, Ouaari C, Cheema A, Kumar D et al (2018) Tumor suppressor RARRES1-a novel regulator of fatty acid metabolism in epithelial cells. PLoS One 13(12):e0208756 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Sun Y, Fu A, Xu W, Chao J-R, Moshiach S, Morris SW (2015) Myeloid leukemia factor 1 interfered with Bcl-XL to promote apoptosis and its function was regulated by 14-3-3. J Physiol Biochem 71:807–21 [DOI] [PubMed] [Google Scholar]
- 35.Tang Z, Yang Y, Chen W, Liang T (2023) Epigenetic deregulation of MLF1 drives intrahepatic cholangiocarcinoma progression through EGFR/AKT and Wnt/β-catenin signaling. Hepatol Commun 7(8):e0204 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(DOCX 3.11 MB)
Data Availability Statement
The data supporting the findings of this study are available from the corresponding author upon reasonable request.