Ginseng Compounds Show Promise for Blood Disorders but Face Steep Translational Hurdles

Ginseng has been a cornerstone of East Asian traditional medicine for more than two thousand years, prescribed alone for severe depletion or blended into formulas for patterns described as qi and blood deficiency. Now a systematic synthesis of the laboratory evidence suggests that its active molecules, the ginsenosides, may indeed influence both healthy blood formation and malignant blood cells—but in strikingly different ways depending on the context, and with formidable obstacles standing between the petri dish and the clinic.

A mini review published in Frontiers in Pharmacology by researchers affiliated with Zhejiang Chinese Medical University and Guangxi University of Chinese Medicine set out to integrate a scattered literature that previous summaries had treated piecemeal. Earlier reviews had focused either on individual ginsenosides such as Rg3, Rh2, and Rg1, or on specific conditions like myelosuppression. The new analysis instead compares evidence from non-malignant hematopoietic stress models and malignant hematopoietic models within a single analytical framework, evaluating each finding according to material identity, experimental context, dose, evidence tier, mechanistic validation, and principal limitation. Crucially, the authors excluded crude whole-plant extracts and multi-herbal formulations from the core synthesis, concentrating instead on purified ginsenosides and chemically characterized saponin fractions that permit reproducible dosing and specific attribution of effects.

Ginsenosides are specialized metabolites of Panax species, classified primarily into dammarane-type saponins—including protopanaxadiol (PPD), protopanaxatriol (PPT), and ocotillol types—and oleanane-type saponins. Representative PPD-type molecules include Rb1, Rb2, Rc, Rd, and Rg3, while PPT-type molecules include Rg1, Rg2, and Rh1. Their profiles vary with species, plant part, geography, and processing: steaming converts relatively polar parent ginsenosides into less-polar metabolites such as Rg3, Rg5, Rk1, Rk3, Rh1, and Rh4, fundamentally altering the chemistry of the preparation. This variability is not a footnote—it is a central reason why findings across studies are so difficult to reconcile.

In non-malignant models, the most extensively studied molecule is Rg1. In an immune-mediated murine model of aplastic anemia, purified Rg1 increased the abundance of LSK cells—a population enriched for hematopoietic stem and progenitor activity—boosted erythroid colony formation, and reduced reactive oxygen species accumulation, mitochondrial Bax translocation, and apoptosis. In cyclophosphamide-induced myelosuppression, purified Re and Rk3 partially improved peripheral blood and bone marrow indices and increased ex vivo progenitor colony formation, with several endpoints favoring Rk3. Rg1 pretreatment before 6.5-Gy total-body irradiation also increased SOD2 expression and improved senescence and colony-forming endpoints in Sca-1-positive HSPC-enriched cells. Yet the review is blunt about what these results do and do not show: none of the studies performed competitive transplantation, and the available noncompetitive serial-transplantation data did not establish durable multilineage reconstitution. The evidence, the authors conclude, does not demonstrate restoration of long-term hematopoietic stem cell self-renewal.

The microenvironment tells an equally nuanced story. After birth, the bone marrow is the principal site of hematopoiesis, but hematopoietic stress can induce extramedullary hematopoiesis elsewhere. In cyclophosphamide-treated mice, Rg1 attenuated splenic extramedullary hematopoiesis and increased circulating c-Kit-positive CD45-positive cells and bone marrow LSK cells—but these increases vanished after splenectomy, indicating dependence on the spleen rather than true marrow rejuvenation. In D-galactose-induced senescence models, Rg1 reduced inflammatory signaling in mesenchymal stromal cells and increased the colony-forming activity of cocultured bone marrow cells, with a TLR1/2 agonist partially blocking the effect and implicating TLR2/NF-κB signaling. Direct target engagement and long-term stem cell function, however, were not established.

Immune modulation emerges as another recurring theme. A chemically characterized Panax notoginseng leaf-saponin fraction improved peripheral blood and colony-forming endpoints in an irradiation-assisted lymphocyte-transfer model of aplastic anemia and shifted splenic T-cell profiles from Th1/Th17- toward Th2/Treg-associated responses. A standardized panaxadiol saponin preparation enriched in Rb1, Rb2, and Rc improved blood-count and CD4/CD8 endpoints and reduced NLRP3-associated signaling, with pharmacological support from MCC950 experiments in primary bone marrow cells. Perhaps most intriguingly, oral protopanaxadiol increased phenotypically defined bone marrow HSPCs and selected immune indices in cyclophosphamide-immunosuppressed mice—but these changes were largely absent after gut-microbiota depletion, pointing to an unexpected microbial contribution whose responsible taxa and gut-to-marrow mechanisms remain undefined.

On the malignant side, the picture shifts from protection to destruction. Compound K, at 20 to 40 micromolar, reduced viability and colony formation in acute myeloid leukemia cell lines and induced apoptosis and G1-phase accumulation with decreased BCL2 mRNA. 20(S)-protopanaxadiol suppressed PI3K/AKT/mTOR signaling and activated a sustained integrated stress response in AML cells, with Bax/Bak-dependent apoptosis and enhanced venetoclax activity in vitro. Purified Rd induced morphological and immunophenotypic differentiation in myeloid leukemia cell lines, and 20(S)-Rh2 induced apoptosis and differentiation-associated phenotypes in HL-60 and Kasumi-1 cells that were attenuated by Nur77 knockdown—a rare example of causal mechanistic validation. In ALL cell lines and primary ALL cells, 20(S)-Rh2 induced autophagic flux and mitochondrial apoptosis, and ATG5 knockdown further increased apoptosis, supporting a cytoprotective role for autophagy that might be exploited therapeutically.

But the review repeatedly flags a critical gap: normal CD34-positive hematopoietic stem and progenitor cells, bone marrow stromal cells, and primary patient samples were rarely included, making it impossible to assess malignant-cell selectivity. In the multiple myeloma study, healthy-donor progenitor colonies remained detectable at 60 micromolar, yet quantitative selectivity was not established, and reported oral systemic exposure in rats—approximately 0.61 micromolar—fell below the 20 to 70 micromolar active range. This mismatch between the concentrations that kill cancer cells in a dish and the concentrations actually achievable in plasma after oral dosing is perhaps the single largest translational barrier the review identifies.

Pharmacokinetics compounds the problem. Absolute oral bioavailability remains poorly characterized in humans; available studies indicate low, delayed, and variable plasma exposure. After oral administration of red ginseng products, compound K, Rh2, PPD, and PPT were detected despite being absent from the administered products—evidence of intestinal deglycosylation by gut bacteria—and compound K exposure correlated with fecal compound K-forming activity, underscoring interindividual microbiota-related variability. Bone marrow exposure has never been reported. Safety concerns add further complexity: several purified ginsenosides, including Rs3, Rg5, 20(S)-Rg3, and Rg1, inhibited agonist-induced platelet aggregation in isolated human platelets and attenuated experimental arterial thrombosis, warranting caution in thrombocytopenic or bleeding-prone patients and during concomitant anticoagulant therapy. American ginseng reduced warfarin exposure in one randomized trial, although studies of Asian ginseng products found no significant effect.

Direct clinical evidence remains thin—limited to two small studies in chronic aplastic anemia. In a 24-week multicenter study, 79 patients were randomized to receive a panaxadiol-saponin capsule plus cyclosporine and reduced-dose testosterone undecanoate, or cyclosporine plus full-dose testosterone undecanoate. Protocol-defined response rates were 88.1 percent versus 77.8 percent, a difference that did not reach statistical significance, and female masculinization scores were lower in the saponin group—though the simultaneous androgen dose reduction prevents attribution of that benefit specifically to the ginsenoside preparation. A meta-analysis of 18 randomized trials involving 2,222 patients with solid tumors found adjunctive Rg3 associated with lower reported risks of leukocyte, hemoglobin, platelet, and neutrophil declines, but the underlying trials were predominantly small, methodologically weak, and heterogeneous. The authors’ verdict is measured: the apparent normal-protective, malignant-cytotoxic pattern should be regarded as a context-dependent working hypothesis rather than demonstrated biological selectivity, and future studies must use chemically defined preparations, clinically relevant exposures, normal and malignant comparators, and causal mechanistic designs before clinical relevance can be determined.

Subject of Research: Pharmacological effects of ginsenosides on hematopoietic homeostasis and hematological malignancies

Article Title: Ginsenosides in hematopoietic homeostasis and hematological malignancies: context-dependent pharmacological effects and translational limitations

Article References: Gai, J., Zhang, B., Wan, X., & Yu, X. (2026). Ginsenosides in hematopoietic homeostasis and hematological malignancies: context-dependent pharmacological effects and translational limitations. Frontiers in Pharmacology, 17, Article 1988591. https://doi.org/10.3389/fphar.2026.1988591

Image Credits: AI Generated

DOI: 10.3389/fphar.2026.1988591

Keywords: ginsenosides, ginseng, hematopoiesis, leukemia, aplastic anemia, myelosuppression, apoptosis, autophagy, gut microbiota, pharmacokinetics, drug resistance, translational research