Multi-omics Analysis Reveals the Correlation of Gut Microbiota and Metabolites With Thalidomide Treatment for Chemotherapy-Induced Nausea and Vomiting in Small Cell Lung Cancer.

Introduction

Lung cancer has been found to be one of the most diagnosed cancer and has contributed to the highest cancer deaths worldwide with about 2 million new cases and 1.76 million deaths on an average basis. The still most deadly subtype of lung cancer, roughly 13%–15% of the lung cancer cases, is small cell lung cancer (SCLC), with the 5‐year survival being less than 7%. One of the characteristic features of SCLC is quick development and early systemic distribution, and close to two‐thirds of the patients already have extrathoracic proliferation at the moment of the primary diagnosis [ 1 ]. Still, systemic chemotherapy with platinum‐based drugs and either etoposide or irinotecan is the primary form of treatment of both limited and extensive stage SCLC [ 2 ]. Even though chemotherapy in SCLC may be highly sensitive initially, the nature of nonspecificity of cytotoxic agents often causes various adverse events which may jeopardize tolerance to treatment. Among these toxic reactions, chemotherapy‐induced nausea and vomiting (CINV) and anorexia are particularly common and distressing. The complication does not only reduce the quality of life of the patients, but also leads to low compliance with treatment and can adversely affect clinical outcomes. In its turn, prevention and treatment of CINV and anorexia play a critical role in making sure that patients are able to adhere to the treatment course when it is prescribed to them [ 3 ].

The current National Comprehensive Cancer Network antiemetic guidelines suggest the following approach to result in prevention and management of CINV: a four‐agent regimen with NK1 receptor inhibitor, 5‐HT3 receptor blocker, dexamethasone, and olanzapine should be used [ 4 ]. Nevertheless, the clinical benefits of this antiemetic regimen are still limited. Thalidomide (THD), a barbiturate‐like compound was originally designed as a cure to pregnancy related nausea. Its usage was subsequently abandoned when it caused a devastating epidemiological occurrence due to unforeseen teratogenic consequences [ 5 ]. In more recent times, THD has also been reemerged into the field of oncology and has been adjudicated to treat multiple myeloma where it has shown a reasonable safety profile [ 6 ]. This discovery has also revived the concept of considering THD in case of CINV control. A randomized, double‐blind, phase III trial including a Chinese cohort has shown evidence that THD is an effective agent to prevent delayed CINV in highly emetogenic cancer patients receiving chemotherapy and is also generally well tolerated [ 7 ]. Also, as patients who receive cancer treatment are not in immediate danger of pregnancy or childbirth, THD would be one of the options that would contribute to CINV management programs. The gut microbiota is an intricate, dynamic microbial community that is crucial for intestinal homeostasis and the proper functioning of host physiology. Growing research indicates that alterations in this microbial community are intimately associated with cancer development, disease progression, and therapeutic responsiveness across multiple tumor types [ 8 ]. Probiotics are beneficial microorganisms that influence host health through modulation of the gut microbiota, including coating the intestinal mucosa, preventing pathogen colonization, and regulating immune responses [ 9 , 10 , 11 ].

Herein, the current paper sought to examine how gut microbiota is relevant in the management of CINV and anorexia in patients with SCLC through the combination of metagenomic and metabolomic studies. By comparing stool samples from patients with SCLC who did or did not receive THD, we sought to determine whether improvement in CINV and anorexia was associated with changes in the gut microbial community and metabolic profiles, and to further explore the mechanisms underlying these associations. Our findings may provide new insight into the potential mechanisms by which THD alleviates CINV and anorexia in patients with SCLC and may inform future strategies for their prevention and management.

Methods

Ethics

The Ethics Committee and Scientific Review Board of the Second Hospital of Dalian Medical University (No. 2022–173) approved the study protocol and informed consent forms. All subjects provided written informed consent before enrollment. The research complied with the revised Declaration of Helsinki and relevant ethical standards.

Study Participants and Sample Collection

A total of 17 eligible patients with SCLC were recruited for this study, including 7 individuals in the THD‐treated group (E group) and 10 individuals in the control group (C group). All patients received the same conventional anticancer treatment, namely etoposide plus platinum (EP) chemotherapy, and none received radiotherapy during the study period. Patients in the E group received THD in addition to EP chemotherapy for the management of CINV and anorexia, whereas patients in the C group received EP chemotherapy plus a matching placebo. Specifically, THD was administered orally at a dose of 100 mg twice daily on days 1–5 of each chemotherapy cycle, starting on day 1 of chemotherapy. In addition, both groups received standard antiemetic prophylaxis, consisting of palonosetron (0.25 mg intravenously on day 1) and dexamethasone (12 mg intravenously on day 1 and 8 mg orally on days 2–4). Thus, both groups received the same background treatment, while patients in the E group received THD and those in the C group received a matching placebo.

The inclusion criteria were as follows: (1) histologically confirmed primary SCLC; (2) age between 18 and 75 years, regardless of sex; (3) scheduled to receive highly emetogenic EP chemotherapy with standard antiemetic prophylaxis; (4) Eastern Cooperative Oncology Group performance status of 0–2 and an expected survival of at least 3 months; and (5) willingness to provide written informed consent and comply with follow‐up and sample collection procedures.

The exclusion criteria were as follows: (1) allergy to THD or its components, or contraindications to THD use, including pregnancy or failure to use contraception; (2) use of antibiotics or probiotics within the previous 3 months; (3) severe gastrointestinal disorders or active gastrointestinal infections; (4) diagnosis of another primary malignancy; and (5) infectious diseases, including viral hepatitis, HIV infection, or syphilis.

Fresh fecal samples were collected from all participants at multiple time points during the study. Participants were instructed in the use of sterile fecal collection containers. Samples were placed in insulated containers with ice packs immediately after collection and transported to the laboratory within 30 min whenever possible. After snap‐freezing in liquid nitrogen, all samples were stored at −80°C until metagenomic sequencing and metabolomic profiling were performed.

Metagenomic Sequencing

Shotgun metagenomic sequencing was performed on the Illumina platform using a paired‐end strategy to generate short‐read libraries. Raw sequencing reads were first processed using fastp to remove adapter contamination and low‐quality reads. To eliminate host‐derived contamination, the filtered reads were aligned to the human reference genome using Bowtie2, and matching reads were discarded. The resulting clean reads were then assembled de novo into contigs of at least 300 bp using MEGAHIT, and assembly quality was evaluated with QUAST.

Coding sequences (CDSs) were predicted from the assembled contigs using MetaGeneMark. A nonredundant gene catalog was subsequently constructed using MMseqs2 with a sequence identity threshold of 90% and a coverage threshold of 80%. For functional annotation, the nonredundant genes were compared against public databases, including KEGG, eggNOG, Pfam, SwissProt, CARD, and CAZy.

For taxonomic annotation, the predicted protein sequences of the nonredundant genes were aligned against the NCBI Non‐Redundant Protein Database (Nr database) using DIAMOND (v0.9.29.130) with an E‐value threshold of 1e‐5. The Nr database, created and maintained by NCBI, is a comprehensive nonredundant protein database containing extensive protein sequence information together with corresponding taxonomic annotation. Each gene was assigned taxonomic information according to its best‐matched sequence in the Nr database. Therefore, the reported results at the species level represent annotation‐based taxonomic assignments inferred from the metagenomic gene catalog, rather than direct species identification from individual short reads alone.

To characterize microbial community structure, taxonomic profiles were summarized across different hierarchical levels, including phylum, class, order, family, genus, and taxa with species‐level annotation. Alpha and beta diversity analyses were performed using principal component analysis (PCA), principal coordinates analysis (PCoA), nonmetric multidimensional scaling (NMDS), and unweighted pair‐group method with arithmetic mean (UPGMA) clustering. Differences in community composition between groups were evaluated using ANOSIM and PERMANOVA. Differential taxonomic features were analyzed using Welch's t‐test, analysis of variance (ANOVA), Wilcoxon rank‐sum test, Kruskal‐Wallis test, and metagenomeSeq, as appropriate. All analyses were performed in R using relevant packages, including vegan.

Metabolomic Profiling

Metabolomic analysis was performed using a high‐resolution mass spectrometry system (Waters ACQUITY I‐Class PLUS UPLC coupled with Xevo G2‐XS QTof). The HSS T3 column combined with formic acid in a water solution and acetonitrile was used as a mobile phase, and the chromatographic separation was performed. Sample analyses were done in positive and negative modes of ionization with injection volume 1 µL.

The MSe mode was run on mass spectrometric data; capillary voltage was set to +2000/−1500 V (positive/negative); cone voltage was set to 30 V; desolvation temperature was set to 500°C; and the desired parameters were set to 800 L/h of desolvation gas flow rate. Progenesis QI was used to identify peaks in raw data, align and annotate them. Metabolites were identified by matching against public databases and an in‐house spectral library, which included a mass deviation threshold value of 100 ppm.

All data were normalized to the total peak area. The measurement of the reproducibility to groups was conducted by applying PCA and Spearman correlation. Differential metabolites were identified with the help of orthogonal partial least squares‐discriminant analysis (OPLS‐DA) and confirmed by 200 permutation tests. The criteria used to select the variables were FC >1, p < 0.05 (Student's t‐test), and variable importance in projection >1. Pathway enrichment analysis has been done with KEGG and HMDB and lipid maps database, and significance measured using hypergeometric distribution test.

Statistical Analysis

Data from metagenomic and metabolomic analyses were processed using the Biomarker Cloud Platform ( www.biomarker.com.cn ). Clinical variables were analyzed using GraphPad Prism version 9.5 and SPSS version 25.0. Baseline demographic and clinical characteristics were summarized using descriptive statistics, including means, standard deviations, and frequencies. Two‐group comparisons used Student's t‐test, and associations were analyzed by Spearman's rank correlation. For analyses involving multiple comparisons in microbial taxonomic abundance, nonparametric tests such as the Wilcoxon rank‐sum test were applied as appropriate, and multiple‐testing correction was performed using the Benjamini–Hochberg (BH) procedure to control the false discovery rate (FDR). An adjusted p value ( q value) < 0.05 was considered statistically significant. For LEfSe analysis, we note that the method does not directly apply BH correction to raw P values; instead, it estimates and controls the false discovery rate of the final reported feature list through its internal permutation‐based procedure. All analyses were two‐sided, and p < 0.05 was considered statistically significant unless otherwise specified.

Results

Clinical Characteristics of Patients

A total of 17 patients diagnosed with SCLC participated in this study. All were residents of Dalian, China, and had similar dietary habits. A total of 7 patients were included in the E group and 10 in the C group (Table 1 ).

Effect of THD on CINV and Anorexia

To evaluate the effects of THD on CINV and anorexia, symptom severity was compared between the E group and the C group during platinum‐based chemotherapy. Symptoms were recorded from the day before chemotherapy to day 6 after treatment. Nausea and anorexia were assessed using a 10‐point scale, and vomiting was evaluated by the daily number of episodes.

Compared with the control group, patients in the E group showed significantly improved symptom control. Specifically, the mean nausea score was significantly lower in the E group than in the C group (0.88 ± 0.88 vs. 2.89 ± 2.20, p = 0.0379). The mean anorexia score was also significantly lower in the E group (0.45 ± 0.39 vs. 2.51 ± 2.16, p = 0.0254). In addition, the mean number of vomiting episodes per day during treatment was lower in the E group than in the C group (0.43 ± 0.79 vs. 4.00 ± 4.19, p = 0.0433) (Table 2 ).

Overall, these findings suggest that THD treatment may be associated with improved control of CINV and anorexia, as well as better appetite and overall treatment tolerance, in patients with SCLC. These clinical observations prompted further investigation into whether symptom improvement after THD treatment was associated with changes in gut microbial community structure and metabolic characteristics.

Characterization of the Gut Microbiota in SCLC Patients Using Metagenomic Analysis

THD Treatment Significantly Improves Microbial Richness and Diversity

To evaluate the impact of THD on gut microbial diversity, a total of 38 fecal samples were subjected to metagenomic sequencing, resulting in the annotation of 177 phyla, 152 classes, 304 orders, 667 families, 2613 genera, and 13,401 taxa with species‐level annotation. Alpha diversity was compared between the E and C groups, showing that THD treatment was associated with significant differences in gut microbial composition. Analyses of microbial richness and diversity demonstrated significant differences in both ACE index ( p = 0.022) and Shannon index ( p = 0.0037, Figure 1A ). Overall, our results indicate that THD use is accompanied by higher microbial richness and diversity in the gut microbiota of SCLC patients.

THD Treatment Drives Changes in Specific Microbial Taxa

To further characterize taxonomic alterations in the gut microbiota, we used Welch's t‐test to compare genus‐level microbial abundances between the two groups. Analysis of the data showed significant compositional differences in THD treated and nontreated patients with many genera showing substantial changes of abundance in the THD group (Figure 1B ). LEfSe (Linear Discriminant Analysis Effect Size) analysis also gave further difference in microbial signatures that separated the two groups (Figure 1C,D ). Particularly, Clostridia , Eubacteriales , Firmicutes , Bacteroidales , Bacteroidia , and Oscillospiraceae were more abundant in the E group, and Enterobacteriaceae , Gammaproteobacteria , Proteobacteria , Enterobacterales , and Klebsiella were more abundant in the C group.

Relative abundances of Bacteroides , Clostridium , Faecalibacterium , and Ruminococcus were significantly increased, whereas Escherichia was considerably reduced at the genus level in patients receiving THD (Figure 1E ). Differences were also observed among taxa with species‐level annotation (Figure 1F,G ). The taxa with species‐level annotation that were enriched in the E group included Prevotella_sp._CAG_520 , Eubacterium_sp._AF22_8LB , Blautia_sp._OF03_15BH , and Firmicutes_bacterium_AF22_6AC .

Untargeted Metabolomic Profiling Uncovers Gut Microbial Metabolic Features in SCLC Patients

THD Changes Gut Metabolite Profile

The metabolomic analysis relied on 38 fecal samples in order to perform both qualitative and quantitative analyses and concluded with 991 peaks representing annotated metabolites. To evaluate how THD affects the metabolic profile of SCLC patients, an OPLS‐DA model was used to compare general metabolic trends of the E and C group. The analysis demonstrated a clear separation of metabolic profiles between the two groups, with a high model fit (R 2 Y = 0.986) and satisfactory predictive accuracy (Q 2 Y = 0.507), suggesting that THD treatment was associated with distinct metabolic alterations (Figure 2A,B ).

THD Causes Noticeable Changes in Fecal Metabolite Levels

To identify group‐specific metabolic alterations, volcano plots and radar charts were generated, and significantly upregulated or downregulated metabolites were identified. Hydrogenated MDI, becocalcidiol, b‐octylglucoside, and azelaic acid levels were significantly high in the E group. The other fatty acid combinations, cortisol, sphinganine, and cellobiose, on the other hand, were much higher in the C group (Figure 2C,D ).

Receiver operating characteristic (ROC) curve analysis was conducted to evaluate the ability of differential metabolites to distinguish between the E and C groups. Cortisol exhibited the highest area under the curve (AUC = 0.889), reflecting strong discriminatory performance (Figure 2E ). Also, hydrogenated MDI, kynurenic acid and azelaic acid were well classified with an AUC of 0.847, 0.793 and 0.781, respectively (Figure 2F–H ). These metabolites can be considered candidate biomarkers of information about the therapeutic effects of THD.

Besides, KEGG pathway enrichment analysis of the differentiated metabolites was performed using the clusterProfilerpackage through hypergeometric testing. The findings showed that the metabolic pathways related to THD included the neomycin, kanamycin, and gentamicin synthesis, fatty acid metabolism (through elongation and degradation), and amino acid metabolism (tryptophan metabolism and arginine biosynthesis) (Figure 2I ). These findings suggest that these pathways may be involved in THD‐associated metabolic reprogramming and may provide insight into the underlying biological mechanisms in patients with SCLC. They may also offer a basis for future targeted metabolic investigations.

Integrated Multi‐omics Analysis Uncovers Correlation and Predictive Ability of Gut Microbiota and Metabolites in THD Treatment of CINV

We integrated metagenomic and metabolomic data to assess the relationship between gut microbiota and metabolites. Comparative analysis and dimensionality reduction of the multi‐omics datasets showed significant differences in gut microbial composition, metabolic profiles, and functional gene pathways between THD‐treated patients and controls. These findings suggest that multiple biological processes may be associated with THD‐related differences in CINV and anorexia (Figure 3A ).

Correlation analysis identified strong links between microbial species and metabolites (Figure 3B ). An example of this is Anaerosporobacter which is a bacterial genus that was associated with a positive correlation with the levels of kynurenic acid ( p < 0.001). Co‐inertia analysis of the form of omicake4 was done in R to analyze the co‐distribution of differential microbial taxa and metabolites and identify their interactions and patterns of interacting with each other (Figure 3C ). It was found that microbial species and metabolite profiles are closely related. Microbial species clustered in a manner consistent with overlapping niches, whereas the wide spread of metabolites suggested that their production depends on multiple microbial communities.

To find possible biomarkers to predict the efficacy of THD in the management of CINV and anorexia, we also applied machine learning methods that integrated both metabolomic and metagenomic data. The ranking of feature importance identified a number of important predictors, such as 5,7α‐Dihydro‐1,4,4,7a‐tetramethyl‐4H‐indene, cortisol, Macellibacteroides , farnesyl acetone, and Ruminiclostridium (Figure 3D ). The random forest classifier was trained and the ROC curves were created to evaluate the performance of the model. The integrated model performed well in the cross‐validation of the test set with an AUC of 0.95 ± 0.09 in the k‐fold cross‐validation. This is impressive, as the AUC was 1.00 in the independent validation set indicating great predictive power (Figure 3E ).

These findings indicate that combining multi‐omics data greatly improve the accuracy of predicting the therapeutic effectiveness of THD in CINV and anorexia. Further studies are needed to improve the process of multi‐omics integration and enhance the clinical application of models by enhancing model resilience.

Discussion

Through metagenomic and metabolomic analyses, our study found that THD treatment was associated with reduced CINV and anorexia in patients with SCLC, accompanied by alterations in gut microbial composition and metabolite profiles. THD treatment was associated with increased microbial α‐diversity and with a remodeled gut microbial profile characterized by higher relative abundances of Bacteroides , Clostridium , Faecalibacterium , and Ruminococcus . In parallel, fecal metabolomic analysis showed higher levels of kynurenic acid, hydrogenated MDI, becocalcidiol, b‐octylglucoside, and azelaic acid, whereas cortisol, sphinganine, and cellobiose were increased in the C group. These findings suggest that the gut microbiota‐metabolite axis may be associated with the potential effects of THD on CINV and anorexia.

Inflammatory processes are increasingly recognized as key contributors to the pathophysiology of CINV, which results in the neural sensitization of the enteric and central nervous system [ 12 , 13 ]. Several THD‐enriched genera, including Clostridium and Faecalibacterium , are known producers of short‐chain fatty acids that promote regulatory T cell (Treg) differentiation and suppress intestinal inflammation [ 14 , 15 ]. Butyrate that Faecalibacterium generates also strengthens the integrity of the epithelial barrier and suppresses the activation of nuclear factor kappa‐light‐chain‐enhancer of activated B cells (NF‐κB) [ 16 , 17 ]. Also, Bacteroides have the ability to control serotonin (5‐HT) synthesis through the tryptophan metabolic pathway, which has a direct effect on vagal afferent signaling contributing to the vomiting reflex [ 18 ]. Taken together, these microbiota‐related effects reported in previous studies may provide possible biological explanations for the improved chemotherapy tolerance observed in THD‐treated patients.

Although the present study identified associations among THD treatment, symptom improvement, and changes in gut microbial and metabolomic profiles, the underlying biological mechanisms were not directly tested and should therefore be interpreted as hypotheses informed by previous studies. More specifically, the metabolomic alterations observed in the present study suggest a possible association between THD treatment and the tryptophan‐kynurenine metabolic axis. Among the differential metabolites, kynurenic acid may be of particular relevance because it has been linked to anti‐inflammatory and immunoregulatory signaling, including pathways related to G protein‐coupled receptor 35 (GPR35) [ 19 , 20 ]. Activation of this pathway has been associated with suppression of pro‐inflammatory cytokine signaling and enhancement of Treg and group 2 innate lymphoid cell responses, thereby contributing to an anti‐inflammatory intestinal environment [ 21 , 22 ]. Kynurenic acid has also been reported to participate in the regulation of oxidative and metabolic stress through β‐adrenergic signaling and mitochondrial biogenesis‐related pathways involving peroxisome proliferator‐activated receptor gamma co‐activator 1‐alpha and regulator of G protein signaling 14 [ 23 , 24 ]. As the precursor of 5‐HT synthesis and the upstream substrate for kynurenic acid production, tryptophan may also play an important role in neuro‐immune communication. Kynurenic acid has been reported to act as a noncompetitive inhibitor of N‐methyl‐D‐aspartate receptors and α7‐nicotinic acetylcholine receptors, and may therefore be relevant to pathways involved in neuronal excitability and emetic signaling [ 25 ]. In addition, based on previous literature, one possible hypothesis is that THD treatment may be associated with altered kynurenic acid‐related signaling through modulation of indoleamine 2,3‐dioxygenase activity, which could potentially influence vagal and brainstem pathways involved in nausea and vomiting [ 26 , 27 ]. Taken together, these observations raise the possibility that kynurenic acid‐related neuro‐immune signaling may be involved in the association between THD treatment and improvement of CINV and anorexia, although this mechanism was not directly examined in the present study.

Additionally, azelaic acid, a metabolite elevated in the E group, has been reported to possess strong anti‐inflammatory and antioxidative activities. It can inhibit NF‐κB and p38 mitogen‐activated protein kinase signaling, decrease reactive oxygen species production, and suppress neutrophil‐mediated oxidative bursts [ 28 ]. Becocalcidiol, another differential metabolite identified in our analysis, is a vitamin D3 analogue rather than a canonical endogenous gut microbiota‐derived metabolite [ 29 ]. Vitamin D and its analogues have been implicated in immunomodulatory and anti‐inflammatory processes [ 30 ]. Therefore, although its relevance in the context of THD treatment, CINV, and anorexia remains uncertain, the differential abundance of becocalcidiol may warrant further attention as a potentially biologically relevant signal. By contrast, β‐octylglucoside is more commonly recognized as a nonionic detergent used in biochemical applications than as a host‐microbial metabolite [ 31 ]. Related studies suggest that octyl glucoside compounds may interact with membrane‐associated environments and modulate epithelial permeability under certain conditions [ 32 , 33 ]. Accordingly, its differential abundance may reflect broader changes in membrane‐related physicochemical states or intestinal barrier‐associated properties rather than a well‐defined mechanistic mediator. Hydrogenated MDI is more commonly recognized as an exposure‐related synthetic compound than as a canonical endogenous metabolite [ 34 ]. Therefore, the differential abundance of these less commonly reported metabolites may reflect broader exogenous, exposure‐related, or low‐specificity metabolic signals. Their potential biological relevance should be interpreted cautiously and requires further confirmation in targeted metabolomic and mechanistic studies.

Finally, THD may alleviate CINV and anorexia through coordinated alterations in gut microbial ecology, tryptophan metabolism, and downstream immune‐neural signaling pathways. Taken together, these findings suggest that the gut microbiota‐metabolite axis may represent one of the pathways associated with the potential effects of THD and may provide candidate biomarkers and potential therapeutic targets for improving the management of CINV and anorexia in patients with SCLC. Several limitations of the present study should be acknowledged. First, the sample size was relatively small, with only 7 patients in the E group and 10 in the C group. Given the substantial interindividual variability commonly observed in gut microbiome and metabolomic data, this limited cohort size may reduce statistical power and affect the robustness of the identified differential taxa and metabolites. It may also influence the interpretation of the observed group differences, as some associations could be sensitive to cohort composition and individual‐specific variation rather than reflecting stable population‐level patterns. Second, although 38 fecal samples were collected, they were obtained from a relatively small cohort of 17 patients across multiple time points, which may further constrain the stability and generalizability of the microbiome and metabolomic findings. In addition, a significant imbalance in family history of cancer was observed between the two groups at baseline. Although there is currently no clear evidence that family history of cancer is a major determinant of short‐term CINV outcomes, its potential influence on host background characteristics, including gut microbial and metabolic profiles, cannot be completely excluded. Given the limited sample size, no further statistical adjustment for this baseline imbalance was performed. Third, although multiple‐testing correction was applied in analyses involving multiple taxonomic abundance comparisons, the overall omics findings should still be interpreted with appropriate caution because results derived from a relatively small dataset may remain vulnerable to instability, false‐positive findings, and overfitting. In particular, the performance of the integrated multi‐omics predictive model should be interpreted cautiously, as it may not fully reflect performance in external populations. Taken together, these considerations indicate that the present study should be considered exploratory in nature, and the findings should therefore be regarded as preliminary and interpreted with appropriate caution. Further studies with larger sample sizes, independent validation cohorts, and mechanistic investigations are needed to assess the reproducibility and potential clinical relevance of these observations.

Conclusion

THD treatment was associated with improved control of CINV and anorexia in patients with SCLC, accompanied by alterations in gut microbial composition and metabolite profiles. These findings suggest that the gut microbiota‐metabolite axis may be associated with the potential effects of THD. The identified microbial taxa and metabolites may serve as candidate biomarkers or potential therapeutic targets, although their clinical relevance requires further validation in larger studies.

Author Contributions

Qiang‐Guo Sun , Dan Zang , Yu Xin , Jia Cui , Xu Han : data curation, analysis, interpretation, investigation, writing – original draft. Jun Chen : conceptualization, funding acquisition, methodology, data curation, and writing – review and editing.

Funding

This work was supported by the National Natural Science Foundation of China (82203056), Joint Program Project of Science and Technology Plan of Liaoning Province (2025JH2/101800202) and United Foundation for Dalian Institute of Chemical Physics, Chinese Academy of Sciences and the Second Hospital of Dalian Medical University (DMU‐2&DICPUN202505).

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

Thanks to Beijing Biomarker Technologies for providing a cloud platform for processing metagenomics and metabolomics raw data.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

References

1. Megyesfalvi Z., Gay C. M., Popper H., et al., “Clinical Insights Into Small Cell Lung Cancer: Tumor Heterogeneity, Diagnosis, Therapy, and Future Directions,” CA: A Cancer Journal for Clinicians 73, no. 6 (2023): 620–652.

2. Pillai R. N. and Owonikoko T. K., “Small Cell Lung Cancer: Therapies and Targets,” Seminars in Oncology 41, no. 1 (2014): 133–142, 10.1053/j.seminoncol.2013.12.015.

3. Adel N., “Overview of Chemotherapy‐induced Nausea and Vomiting and Evidence‐based Therapies,” American Journal of Managed Care 23, no. 14 (2017): S259–S265.

4. Xie J., Zhang C., Li S., et al., “Efficacy and Safety of Thalidomide as a Pre‐Medication of Chemotherapy‐Induced Nausea and Vomiting (CINV) Following Highly Emetogenic Chemotherapy (HEC): A Systematic Review and Meta‐Analysis,” Frontiers in Oncology 11 (2021): 818839, 10.3389/fonc.2021.818839.

5. Franks M. E., Macpherson G. R., and Figg W. D., “Thalidomide,” The Lancet 363, no. 9423 (2004): 1802–1811, 10.1016/S0140-6736(04)16308-3.

6. Wang N., Xu P., Liu Y., et al., “Efficacy and Safety of Thalidomide for Chemotherapy‐induced Nausea and Vomiting,” Journal of Cancer 11, no. 15 (2020): 4560–4570, 10.7150/jca.45678.

7. Zhang L., Qu X., Teng Y., et al., “Efficacy of Thalidomide in Preventing Delayed Nausea and Vomiting Induced by Highly Emetogenic Chemotherapy: A Randomized, Multicenter, Double‐Blind, Placebo‐Controlled Phase III Trial (CLOG1302 study),” Journal of Clinical Oncology 35, no. 31 (2017): 3558–3565, 10.1200/JCO.2017.72.2538.

8. Bishehsari F., Voigt R. M., and Keshavarzian A., “Circadian Rhythms and the Gut Microbiota: From the Metabolic Syndrome to Cancer,” Nature Reviews Endocrinology 16, no. 12 (2020): 731–739, 10.1038/s41574-020-00427-4.

9. Yang N., Zhan Y., Wan J., Li Y., Hu X., and Liu W., “Effects of Lacidophilin Tablets, Yogurt, and Bifid Triple Viable Capsules on the Gut Microbiota of Mice With Antibiotic‐Associated Diarrhea,” Canadian Journal of Infectious Diseases and Medical Microbiology 2022 (2022): 1, 10.1155/2022/6521793.

10. Aragon G., Graham D. B., Borum M., and Doman D. B., “Probiotic Therapy for Irritable Bowel Syndrome,” Gastroenterol Hepatol (N Y) no. 6 (2010): 39–44.

11. Nguyen T. T. B., Chung H. J., Kim H. J., and Hong S. T., “Establishment of an Ideal Gut Microbiota to Boost Healthy Growth of Neonates,” Critical Reviews in Microbiology 45, no. 1 (2019): 118–129, 10.1080/1040841X.2018.1561643.

12. Chen W., Zhao Y., Dai Y., and Nie K., “Gastrointestinal Inflammation Plays a Critical Role in Chemotherapy‐induced Nausea and Vomiting,” European Journal of Pharmacology 936 (2022): 175379, 10.1016/j.ejphar.2022.175379.

13. Irving H., Turek I., Kettle C., and Yaakob N., “Tapping Into 5‐HT3 Receptors to Modify Metabolic and Immune Responses,” International Journal of Molecular Science 22, no. 21 (2021): 11910, 10.3390/ijms222111910.

14. Atarashi K., Tanoue T., Shima T., et al., “Induction of Colonic Regulatory T Cells by Indigenous Clostridium Species,” Science 331, no. 6015 (2011): 337–341, 10.1126/science.1198469.

15. Krause F. F., Mangold K. I., Ruppert A. L., et al., “Clostridium Sporogenes‐Derived Metabolites Protect Mice Against Colonic Inflammation,” Gut Microbes 16, no. 1 (2024): 2412669, 10.1080/19490976.2024.2412669.

16. Wei L., Wen X. S., and Xian C. J., “Chemotherapy‐Induced Intestinal Microbiota Dysbiosis Impairs Mucosal Homeostasis by Modulating Toll‐Like Receptor Signaling Pathways,” International Journal of Molecular Science 22, no. 17 (2021): 9474, 10.3390/ijms22179474.

17. Schlein L. J. and Thamm D. H., “Review: NF‐kB Activation in Canine Cancer,” Veterinary Pathology 59, no. 5 (2022): 724–732, 10.1177/03009858221092017.

18. Prochazkova P., Roubalova R., Dvorak J., et al., “The Intestinal Microbiota and Metabolites in Patients With Anorexia Nervosa,” Gut Microbes 13, no. 1 (2021): 1–25, 10.1080/19490976.2021.1902771.

19. Walczak K., Wnorowski A., Turski W. A., and Plech T., “Kynurenic Acid and Cancer: Facts and Controversies,” Cellular and Molecular Life Sciences 77, no. 8 (2020): 1531–1550, 10.1007/s00018-019-03332-w.

20. Zhen D., Liu J., Zhang X. D., and Song Z., “Kynurenic Acid Acts as a Signaling Molecule Regulating Energy Expenditure and Is Closely Associated With Metabolic Diseases,” Frontiers in Endocrinology 13 (2022): 847611, 10.3389/fendo.2022.847611.

21. Arifuzzaman M., Won T. H., Yano H., et al., “Dietary fiber Is a Critical Determinant of Pathologic ILC2 Responses and Intestinal Inflammation,” Journal of Experimental Medicine 221, no. 5 (2024): 20232148, 10.1084/jem.20232148.

22. Jin L., Hu Q., Hu Y., Chen Z., and Liao W., “Respiratory Syncytial Virus Infection Reduces Kynurenic Acid Production and Reverses Th17/Treg Balance by Modulating Indoleamine 2,3‐Dioxygenase (IDO) Molecules in Plasmacytoid Dendritic Cells,” Medical Science Monitor 26 (2020): 926763, 10.12659/MSM.926763.

23. Agudelo L. Z., Ferreira D. M. S., Cervenka I., et al., “Kynurenic Acid and Gpr35 Regulate Adipose Tissue Energy Homeostasis and Inflammation,” Cell Metabolism 27, no. 2 (2018): 378–392, 10.1016/j.cmet.2018.01.004.

24. Park J. C., Luebbers A., Dao M., et al., “Fine‐tuning GPCR‐Mediated Neuromodulation by Biasing Signaling Through Different G Protein Subunits,” Molecular Cell 83, no. 14 (2023): 2540–2558.e12, 10.1016/j.molcel.2023.06.006.

25. Roth W., Zadeh K., Vekariya R., Ge Y., and Mohamadzadeh M., “Tryptophan Metabolism and Gut‐Brain Homeostasis,” International Journal of Molecular Science 22, no. 6 (2021): 2973, 10.3390/ijms22062973.

26. Wang G., Cao K., Liu K., et al., “Kynurenic Acid, an IDO Metabolite, Controls TSG‐6‐mediated Immunosuppression of human Mesenchymal Stem Cells,” Cell Death & Differentiation 25, no. 7 (2018): 1209–1223, 10.1038/s41418-017-0006-2.

27. Zhao L. P., Wu J., Quan W., et al., “DSS‐induced Colitis Activates the Kynurenine Pathway in Serum and Brain by Affecting IDO‐1 and Gut Microbiota,” Frontiers in Immunology 13 (2022): 1089200, 10.3389/fimmu.2022.1089200.

28. Sieber M. A. and Hegel J. K., “Azelaic Acid: Properties and Mode of Action,” Skin Pharmacology and Physiology 27, no. 1 (2014): 9–17, 10.1159/000354888.

29. Helfrich Y. R., Kang S., Hamilton T. A., and Voorhees J. J., “Topical Becocalcidiol for the Treatment of Psoriasis vulgaris: A Randomized, Placebo‐controlled, Double‐blind, Multicentre Study,” British Journal of Dermatology 157, no. 2 (2007): 369–374, 10.1111/j.1365-2133.2007.08037.x.

30. Lin Z. and Li W., “The Roles of Vitamin D and Its Analogs in Inflammatory Diseases,” Current Topics in Medicinal Chemistry 16, no. 11 (2016): 1242–1261, 10.2174/1568026615666150915111557.

31. Lorber B., Bishop J. B., and DeLucas L. J., “Purification of Octyl β‐d‐glucopyranoside and Re‐estimation of Its Micellar Size,” Biochimica et Biophysica Acta (BBA)—Biomembranes 1023, no. 2 (1990): 254–265, 10.1016/0005-2736(90)90421-J.

32. Brismar G., Brismar J., and Cronqvist S., “Complications of Orbital and Skull Base Phlebography,” Acta Radiologica Diagnosis 17, no. 3 (1976): 274–280, 10.1177/028418517601700302.

33. Tirumalasetty P. P. and Eley J. G., “Permeability Enhancing Effects of the Alkylglycoside, Octylglucoside, on Insulin Permeation Across Epithelial Membrane in Vitro,” Journal of Pharmacy & Pharmaceutical Sciences [Computer File] 9, no. 1 (2006): 32–39.

34. Spence M. W., Plehiers P. M., George A., and Hankett J., “Methylene Diphenyl Diisocyanate Occupational Exposure Data in Industry (1998–2020): A Descriptive Summary From an Industrial Hygiene Perspective,” Toxicology and Industrial Health 39, no. 8 (2023): 407–420, 10.1177/07482337231176604.

Associated Data

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.