Plant-Based Meals Modulate the Gut Microbiome During Stem Cell Transplant, Small Trial Finds

Patients with multiple myeloma who undergo autologous hematopoietic cell transplantation face a double assault: the disease itself and the aggressive conditioning regimen that precedes it, which can devastate the gut microbiome. That microbial disruption is not merely a side effect. Research linking gut microbiome disruption to worse HSCT outcomes has established that reduced microbial diversity during transplant correlates with more complications and poorer survival. A new early-phase trial, published June 10, 2026 in Blood Advances, now offers preliminary evidence that a structured plant-based diet can be delivered safely across the transplant window and may shift the microbiome in a favorable direction, though the findings are far from definitive.

Plant-Based Meals Proved Deliverable Across Five Weeks of Auto-HCT

The trial enrolled 22 multiple myeloma patients at a single center and provided them with fresh, pre-prepared plant-based meals for five weeks, starting seven days before transplant and continuing through day 28 post-transplant. Meals arrived three times daily plus snacks, delivered twice weekly by the commercial service Thistle, and were designed to meet Fred Hutchinson Cancer Center food-safety guidelines for immunocompromised patients. That meant no pre-packaged salads, no pre-cut raw fruits or vegetables, and no raw nuts or seeds, though participants were free to eat meat, eggs, or dairy on the side if they chose.

Feasibility, one of the primary endpoints, was met: all 22 participants were successfully recruited, no one withdrew because of study procedures, and 17 of 22 received the full five weeks of deliveries. The median number of delivered meals actually consumed was 29, with a wide range of 2 to 99, illustrating how much individual experience varied. The biggest obstacle was hospitalization: 13 patients were admitted at some point, and gastrointestinal toxicities during the neutropenic nadir sharply limited meal consumption for many of them.

Dose-Dependent Microbiome Shifts in Higher Meal Consumers

The team split participants at the median number of meals consumed into high and low consumer groups and then compared their gut microbiomes using shotgun metagenomic sequencing and stool short-chain fatty acid (SCFA) measurements. SCFAs, molecules produced when gut bacteria ferment dietary fiber, support epithelial barrier integrity and modulate immune function after HCT; their decline post-transplant has previously been linked to mucositis and other complications.

High consumers showed enrichment of SCFA-producing Lachnospiraceae and Bifidobacterium longum, taxa associated with a fiber-responsive, healthier gut environment. Stool SCFA concentrations rose from baseline to the end of the intervention in this group, and the aminobutanoate degradation pathway, which produces SCFAs, was significantly more active in high consumers. Low consumers, by contrast, showed enrichment of Bacteroidetes and the pathobiont Streptococcus mutans.

Alpha diversity, a measure of microbial species richness within a sample, was preserved throughout the study in both groups, with no significant differences between them. Baseline microbiome composition was similar across high consumers, low consumers, and non-study comparators, which suggests the divergence that emerged was driven by diet rather than pre-existing differences.

Fiber was the principal nutritional variable separating the two groups. Higher consumers ended the study period with significantly greater total fiber intake, and higher calorie intake relative to predicted needs during the post-nadir recovery phase. The authors also ran an unbiased screen of 61 dietary variables against all genera; one correlation survived correction for multiple comparisons: dietary alcohol intake correlated positively with a butyrate-producing Roseburia strain. Notably, the alcohol in this cohort came from condiments like wine vinegars, soy sauce, and vanilla extract, not beverage consumption.

Why Microbial Preservation During Transplant Matters

The stakes behind this work extend beyond microbiome curiosity. Loss of gut microbial diversity after hematopoietic cell transplantation has been linked to worse overall survival, higher infection rates, and increased risk of graft-versus-host disease in allogeneic settings. In multiple myeloma patients receiving autologous transplants, dysbiosis is similarly associated with inferior outcomes, making the gut a plausible target for intervention.

Plant-based dietary patterns more broadly have been linked to anti-inflammatory and anticarcinogenic effects through fiber, phytochemicals, and antioxidants, providing biological rationale for why this dietary approach might be particularly relevant in a cancer context. The current trial is not large enough to test survival outcomes; its value is in showing that the intervention can be executed in a genuinely difficult clinical setting and that the microbiome responds in a dose-dependent way.

Quality-of-life scores offered a mixed picture. Global health status and overall quality-of-life scores declined during the study period, consistent with what transplant patients typically experience, but physical, social, and role functioning scores improved across the process. Overall Healthy Eating Index scores remained in the "very poor" range throughout, reflecting how difficult adequate nutrition is during this period even with active support.

Limitations That Make This Preliminary, Not Practice-Changing

The trial is small and single-arm, which means there is no control group to isolate the dietary effect from the natural trajectory of the microbiome during transplant. Twenty-two participants is not enough to draw conclusions about whether microbiome changes translate to clinical benefit. The correlated nature of dietary nutrients makes it difficult to attribute any observed associations to fiber specifically versus other plant-food components. Some findings at the strain level were of uncertain functional significance.

The neutropenic nadir remains the core practical problem: when patients feel worst and are most likely to be hospitalized, they are least able to consume the intervention. Future trials will need a strategy for maintaining meal support during inpatient admissions and for managing GI toxicities that suppress appetite and tolerance.

Next Steps: Scaling Up and Testing Outcomes

The authors frame these results as proof-of-concept for a larger, controlled trial. The key questions are whether microbiome enrichment of SCFA-producing taxa translates into measurable clinical endpoints, such as reduced infections, faster immune reconstitution, or improved survival, and whether delivery logistics can be adapted for patients who require hospitalization during the intervention window. The trial is registered at ClinicalTrials.gov (NCT06559709) and represents one of the first structured attempts to use a commercially delivered plant-based diet as a microbiome intervention specifically during auto-HCT. If a controlled trial can replicate the microbiome signal with a larger sample, it would open the door to dietary modification as a practical, low-toxicity adjunct to standard transplant care.

Evidence rating

INVESTIGATIONAL — 75/100

Caveats:

• Single-arm trial with no control group (n=22); results show association, not causation.
• Small sample size limits statistical power and generalizability.
• Meal consumption was significantly disrupted during the neutropenic nadir due to GI toxicities and hospitalization, reducing adherence in a key study window.
• Funding source was not disclosed by the authors.
• Results have not yet been replicated in an independent cohort or a randomized controlled trial.

Single-arm feasibility trial (n=22) conducted at one center. Results show associations between plant-based meal consumption and gut microbiome shifts, but the absence of a control group means causation cannot be established. Not yet replicated in an independent population or randomized trial.