Strengthening causal inference and analytical rigor in the Wumei Pills-Lactobacillus reuteri-intestinal stem cell axis for chemotherapy-induced mucositis

Abstract

A recent preclinical study reported that Wumei Pills (WMP) and Lactobacillus reuteri (L. reuteri) mitigate 5-fluorouracil-induced intestinal mucositis by promoting intestinal stem cell (ISC)-mediated repair via Wnt/β-catenin signaling. The mechanistic interpretation rests largely on systemic inflammation readouts, correlative microbiota changes, and immunohistochemistry of pathway markers. From a clinical standpoint, chemotherapy-induced mucositis remains a common and burdensome toxicity that leads to dose reductions, treatment delays, and infection risk; current care is largely supportive and does not directly restore ISC-mediated repair. This unmet need motivates rigorous appraisal of the proposed “WMP → L. reuteri → ISC/Wnt” axis. To highlight key methodological considerations that may affect causal inference and analytical rigor in the proposed “WMP → L. reuteri → ISC/Wnt” pathway. This letter critically appraises the study’s design, endpoints, and analyses against current best practices in mucositis biology, microbiome causality testing, Wnt/β-catenin pathway validation, and preclinical statistics, and synthesizes concrete, literature-grounded remedies. Six issues with potential impact on interpretation were identified: (1) Reliance on serum cytokines/lipopolysaccharide to infer local mucosal inflammation, with limited tissue-level indices (e.g., myeloperoxidase, interleukin-1β, immune-cell infiltration); (2) Absence of necessity/sufficiency tests to verify microbiota mediation (e.g., L. reuteri depletion, WMP-donor fecal microbiota transplantation, probiotic add-back); (3) Pathway evidence tiering - Wnt/β-catenin activation not confirmed by β-catenin nuclear translocation or downstream targets (Axin2, c-Myc, cyclin D1), and Lgr5 quantification/specificity insufficient; (4) Statistical design under-specified (power justification, blinded assessment, control of multiple comparisons) and potential cage effects unmodeled; (5) Limited dose-response and safety profiling for WMP/L. reuteri; and (6) Constrained generalizability (single sex/strain/age, lack of ABX-only controls, single time-point). The reported benefits of WMP and L. reuteri in chemotherapy-induced mucositis are promising, but stronger causal and analytical foundations are needed. Incorporating tissue-level inflammation readouts, microbiota loss-/gain-of-function designs, definitive Wnt/β-catenin activation assays, rigorous statistical practices (including mixed-effects models for cage clustering and multiplicity control), dose-response/safety evaluation, and broader experimental scope (sex/age/strain, ABX-only controls, time-course) will yield more robust and translationally relevant conclusions.

Key Words: Chemotherapy-induced mucositis; Wumei Pills; Lactobacillus reuteri; Intestinal stem cells; Wnt/β-catenin

Core Tip: In this methodological critique, we highlight six key areas to bolster the evidence that Wumei Pills (WMP) and Lactobacillus reuteri mitigate chemotherapy-induced mucositis via intestinal stem cell activation. We urge the authors to incorporate tissue-level inflammation indices (myeloperoxidase, interleukin-1β, immune cell infiltration) rather than relying solely on serum cytokines and lipopolysaccharide. We also suggest causality-verification experiments (probiotic depletion or fecal microbiota transplantation) to confirm the WMP-microbiota-intestinal stem cell linkage. The letter recommends more rigorous validation of Wnt/β-catenin signaling (e.g., demonstrating β-catenin nuclear translocation and downstream target gene activation), stronger study design practices (sample size calculation, blinded outcome assessment, accounting for cage effects, and false-discovery rate adjustments), dose-response and safety evaluations for WMP/Lactobacillus reuteri, and expanded experimental conditions to test the robustness of the therapeutic effect across different biological contexts. These improvements will help substantiate the mechanistic claims and translational potential of this promising gut microbiota-mediated therapy.

TO THE EDITOR

Chemotherapy-induced gastrointestinal mucositis remains a significant clinical challenge characterized by epithelial damage and inflammation[1]. We read with interest Lu et al[2], who report that Wumei Pills (WMP, a multi-herb Chinese medicine) and Lactobacillus reuteri (L. reuteri) improve 5-fluorouracil (5-FU)-induced intestinal mucositis in mice via intestinal stem cell (ISC) activation through Wnt/βcatenin signaling, with reductions in serum cytokines/Lipopolysaccharide and higher PCNA+ and Lgr5+ crypt signals. These findings suggest a novel microbiota-mediated strategy to protect gut mucosa from chemotherapy. We commend the authors for integrating traditional medicine with microbiome science and for reporting key components aligned with ARRIVE guidance. To strengthen causal inference and analytical rigor - toward translation where therapies that actively target ISCs are lacking - we outline recommendations below. We organize six recommendations into three themes: (1) Mechanism validation; (2) Experimental design; and (3) Translational potential - and rank them by feasibility: (1) Tissue-level inflammatory markers; (2) Microbiota causality tests [depletion/fecal microbiota transplantation (FMT)/add-back]; (3) Definitive Wnt/β-catenin activation assays; (4) Prespecified statistics (power, blinding, cage random effects, multiplicity); (5) Dose-response; and (6) Safety profiling. Items 1-2 are highest-priority to directly reinforce the proposed causal chain.

LOCAL VS SYSTEMIC INFLAMMATORY MARKERS

The authors primarily measured systemic inflammation [serum tumor necrosis factor (TNF)-α, interleukin (IL)-6] and gut barrier damage (serum lipopolysaccharide) as indicators of mucosal inflammation[2]. While these circulating markers can reflect an inflammatory response and intestinal permeability, they may not fully capture the local mucosal immune status. Mucositis pathology is often compartmentalized, with intense inflammation at the tissue level even when systemic cytokine levels are modest. For instance, 5-FU challenge causes marked neutrophil infiltration and upregulation of pro-inflammatory cytokines within the intestinal mucosa[3]. In the absence of direct mucosal readouts, conclusions about inflammation resolution remain tentative. We suggest incorporating tissue-level inflammatory indices in future studies. Established assays include myeloperoxidase (MPO) activity in gut tissue homogenates as a quantitative measure of neutrophil infiltration, intestinal levels of IL-1β and TNF-α (by enzyme-linked immunosorbent assay or reverse transcription polymerase chain reaction) to gauge local cytokine response, and histological or immunohistochemical detection of inflammatory cells (e.g., Ly6G+ neutrophils or CD45+ leukocytes in the lamina propria)[3]. Indeed, prior mucositis studies have shown 5-FU induces significant increases in intestinal MPO and IL-1β, correlating with mucosal damage[3]. In the WMP-treated mice, assessing these local parameters (e.g., whether WMP/L. reuteri reduced intestinal MPO or IL-1β) would strengthen the claim of ameliorating mucosal inflammation. Notably, a related study using WMP-derived microbiota transfer reported decreased colon tissue IL-1β, MPO, TNF-α, and IL-6 after treatment, underscoring the importance of such local measurements[4]. By moving beyond serum markers to include tissue-centric endpoints, the authors can provide more direct evidence of intestinal inflammatory resolution, thereby solidifying WMP’s therapeutic effect on mucosal immunity.

MICROBIOTA-MEDIATED CAUSAL PATHWAYS

The crux of the study’s hypothesis is that L. reuteri is a key mediator of WMP’s protective effects on ISCs[2]. The data show WMP treatment was associated with L. reuteri colonization and improved ISC-driven regeneration, implying a causal chain: WMP alters the microbiome (notably enriching L. reuteri), which in turn activates ISC proliferation to heal mucosa[2]. While suggestive, these correlations do not fully distinguish whether L. reuteri is necessary and sufficient for the benefits. We encourage the authors to deploy causality-testing experiments that are now standard in microbiome research. One approach is a loss-of-function test: For example, deplete L. reuteri (or broadly the gut microbiota) in WMP-treated mice and observe if the mucosal protection is lost. This could be achieved by administering an antibiotic regimen that specifically targets Lactobacillus or using germ-free or microbiota-depleted mice reconstituted with defined communities. If WMP fails to confer mucosal protection in the absence of L. reuteri, that would strongly support L. reuteri’s causal role. Conversely, a gain-of-function test such as FMT can demonstrate sufficiency: Transferring the fecal microbiome from WMP-treated mice to mucositis-induced recipient mice. Encouragingly, Lu et al[2] have previously employed such a strategy - FMT from WMP-treated donors significantly alleviated 5-FU mucositis in recipients, reducing inflammatory infiltrates and cytokines. This finding implies that WMP’s benefits are indeed microbiota-mediated. Building on that, the authors might perform probiotic add-back experiments: Administer L. reuteri alone to 5-FU-injured mice to see if it recapitulates ISC activation and mucosal healing (as observed in other contexts)[4]. If so, it would confirm L. reuteri’s therapeutic potential and mechanistic link to ISC proliferation. Additionally, including an “antibiotics-only” control group (chemotherapy plus microbiota depletion without WMP) would clarify background effects of microbiota removal on mucositis[5]. Overall, such interventions - removing or transferring the microbiota - are powerful tools to transform a correlative microbiome finding into a causal demonstration. We believe applying these approaches will significantly reinforce the claim that “WMP → microbiota (L. reuteri) → ISC-mediated repair” is a causative pathway rather than an association.

ASSESSMENT OF WNT/Β-CATENIN SIGNALING

The authors attributed the ISC-mediated mucosal repair to activation of the Wnt/β-catenin pathway, evidenced by increased Lgr5+ stem cells and β-catenin expression in intestinal crypts on WMP treatment. While the immunohistochemical (IHC/IF) detection of Lgr5 and β-catenin is a good starting point, a more comprehensive analysis of Wnt signaling status would lend credibility to this mechanism. We note that the current evidence relies on static IHC images of β-catenin and an increase in Lgr5+ cells, without confirming that β-catenin is functionally activated (i.e., translocated to the nucleus and driving transcription). Nuclear translocation of β-catenin is a hallmark of canonical Wnt pathway activation, as cytosolic β-catenin must accumulate and enter the nucleus to trigger Wnt target gene expression[6]. We recommend the authors to demonstrate this by high-resolution confocal microscopy or cellular fractionation: For instance, showing that WMP-treated/L. reuteri-treated crypts have β-catenin concentrated in nuclei (colocalizing with DAPI) compared to controls. Furthermore, measuring downstream Wnt target genes would confirm pathway activation. Classic β-catenin/T-cell factor target genes in the intestinal epithelium include Axin2, c-Myc, cyclin D1, and EphB/Ephrin family members[6]. Quantitative polymerase chain reaction or western blot analysis could be done on intestinal samples to see if these targets are upregulated in WMP-treated mice, which would support the narrative that Wnt signaling is not only present but functionally enhanced. In a similar study, L. reuteri administration in an intestinal injury model significantly increased the mRNA levels of Wnt3a, Axin2, c-Myc, and cyclin D1 in intestinal tissue, alongside higher Lgr5 counts. The authors could emulate this approach: For example, checking whether WMP + L. reuteri raised Axin2 expression in crypts, a sensitive readout of Wnt/β-catenin activity. Additionally, any quantification of the IHC/IF images (e.g., fluorescence intensity of β-catenin or the number of Lgr5+ crypt base cells per crypt) would add rigor, rather than relying on representative images alone. It is also important to ensure specificity - for instance, confirming that increased β-catenin signal is occurring in epithelial cells at crypt bases (where ISCs reside) as opposed to nonspecific staining in other cell types or regions. By demonstrating nuclear β-catenin localization and upregulation of canonical Wnt target genes, the authors can convincingly show that the Wingless pathway is truly activated in their model. This would substantiate the proposed mechanism that WMP and L. reuteri restore gut integrity via boosting ISC proliferation through Wnt/β-catenin signaling.

STATISTICAL AND EXPERIMENTAL RIGOR

Several opportunities exist to enhance experimental design and statistical analysis. First, sample size justification should be provided a priori to ensure adequate power for key endpoints (e.g., histopathology scores, ISC counts). Second, blinding of outcome assessors (histology and imaging analysis) should be explicitly implemented and reported, in line with best practices and ARRIVE guidance[7]. Third, given the nature of microbiota studies, “cage effects” and environmental confounders must be addressed: Co-housed mice share microbiota and physiology, violating independence assumptions. Mixed-effects models with cage as a random effect and balanced distribution of treatment groups across cages are recommended, acknowledging that housing/vendor/bedding can reshape microbiota and inflammatory phenotypes[8]. Finally, because many endpoints and taxa are tested, multiple-comparison control (e.g., FDR) should be applied to reduce false positives, especially for microbiota analyses.

DOSING AND SAFETY CONSIDERATIONS

The therapeutic interventions used - WMP formulation and L. reuteri probiotic - should be further evaluated for dose-response efficacy and safety profile. In the current report, WMP was administered at a certain dose (or two fixed doses) and L. reuteri was presumably given at a set colony count. It is important to determine the optimal dosing as well as any potential toxicity at higher exposures. We suggest performing a dose - response study for both WMP and L. reuteri. For instance, testing a lower dose and a higher dose of WMP decoction could reveal whether the protective effect plateaus or if higher doses yield additional benefit (or conversely, any diminishing returns or toxicity). Similarly, varying the dose or frequency of L. reuteri gavage might identify the minimum effective dose needed to stimulate ISC recovery. Demonstrating a dose-dependent effect would strengthen the evidence that these agents are directly responsible for the observed outcomes and help guide eventual translation (proper human dosing). Moreover, reporting a lack of dose-response (if the effect saturates at a certain level) is equally informative for understanding the mechanism.

In parallel, safety evaluations are prudent, especially for translational outlook. Traditional herbal formulas can have off-target effects, and while L. reuteri is generally regarded as safe, its impact in a chemotherapy-stressed host warrants scrutiny. The authors could monitor basic health indices: Body weight recovery, behavior, and particularly any organ-specific toxicities. Simple serum chemistry panels (liver enzymes alanine aminotransferase/aspartate aminotransferase, kidney markers blood urea nitrogen/creatinine) or histopathological exam of liver, spleen, etc., can reveal if high-dose WMP or probiotic caused any harm. Report any offtarget toxicities transparently; absence of signal at therapeutic doses would support a favorable margin. Confirming that WMP did not adversely affect liver or kidney function in the treated mice (e.g., no elevation in alanine aminotransferase, no pathological changes in hepatic tissue) would alleviate concerns that the treatment’s benefits come at the expense of other organ injury. Likewise, checking for any signs of sepsis or aberrant immune reaction due to L. reuteri translocation is worthwhile given the disrupted gut barrier in mucositis (though L. reuteri is usually benign). Overall, performing these dose-ranging and safety assessments will ensure that the WMP + L. reuteri intervention is not only effective but also within a safe therapeutic window. This risk-benefit analysis is important as we consider moving such interventions toward clinical testing. A clear demonstration that “more is not always better” and identification of an optimal dose will guide clinicians in the future and underscore the thoroughness of the preclinical evaluation.

BROADENING THE EXPERIMENTAL SCOPE

To enhance generalizability and mechanistic clarity, future studies might include sex/age/strain variation, an antibiotics-only control to parse ABX main effects, and time-course sampling to map injury-repair dynamics. Because housing and related factors can alter microbiota composition and immune tone, ensuring these variables are balanced or modeled is essential for robust inference[8].

CONCLUSION

The reported benefits of WMP and L. reuteri in chemotherapy-induced mucositis are promising. Prioritizing tissue-level inflammation indices and microbiota causality tests, followed by definitive Wnt activation assays and prespecified statistical practices, will most directly strengthen causal claims. Subsequent dose-response and safety profiling will enhance translational readiness. Together, these steps will yield more robust and clinically meaningful conclusions.