Donor hematopoietic stem cell engraftment and donor/recipient chimerism after allogeneic bone marrow transplant for leukemia
Before a single donor cell is infused, the recipient's own marrow must be eradicated. Conditioning regimens combine cytotoxic chemotherapy and/or total body irradiation (TBI) to destroy leukemic blasts, empty the bone marrow niche for the incoming graft, and suppress host immunity enough to prevent graft rejection. The intensity of this regimen is one of the most consequential choices in the entire transplant.
Myeloablative conditioning classically pairs an alkylating agent (busulfan or cyclophosphamide) with either additional chemotherapy or TBI. Busulfan cross-links DNA in dividing marrow precursors; cyclophosphamide is metabolized to acrolein and phosphoramide mustard, which alkylate DNA in both leukemic blasts and healthy hematopoietic stem cells. TBI delivers ionizing radiation systemically, killing marrow and residual leukemia cells throughout the skeleton, including sanctuary sites chemotherapy penetrates poorly.
The result is profound, deliberate pancytopenia: the recipient's neutrophils, platelets, and red cell precursors collapse to near zero within days. This is not a side effect to be minimized but the therapeutic goal — without it, residual host stem cells would outcompete the incoming donor graft for marrow niche space, and residual leukemic blasts would survive to cause early relapse.
Immunosuppressive agents (fludarabine, anti-thymocyte globulin, or total lymphoid irradiation) are layered in specifically to deplete recipient T-cells and NK cells, since these are what would otherwise recognize the HLA-mismatched donor graft as foreign and reject it before engraftment can occur.
Conditioning intensity is a dial, not a switch: more intensity kills leukemia and empties the niche faster, but drives materially higher regimen-related toxicity and mortality — the tradeoff explored in reduced-intensity and non-myeloablative protocols.
Not every patient can tolerate full myeloablation. Older patients and those with comorbidities are increasingly offered reduced-intensity or non-myeloablative regimens, which rely more on the graft-versus-leukemia immune effect than on direct cytotoxic tumor killing. The table below summarizes the tradeoffs across the intensity spectrum — the slider in this simulation moves along exactly this axis, reshaping how quickly host marrow clears and how much residual leukemia survives conditioning.
Beyond killing cells, conditioning restructures the marrow microenvironment itself. Ablation of resident hematopoietic and stromal cells transiently disrupts the endosteal and vascular niches, and marrow stromal cells respond by upregulating SDF-1 (CXCL12) and adhesion molecules such as VCAM-1 — the same chemotactic signals that will draw infused donor stem cells into the vacated niches within hours of infusion. In this sense, conditioning is not just demolition but active site preparation for the cellular therapy that follows in Stage 2.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Myeloablative (MAC) | High-dose busulfan/cyclophosphamide or TBI ≥8 Gy | Complete marrow ablation; irreversible without rescue graft | Fastest, deepest leukemia clearance; higher toxicity/mortality |
| Reduced-Intensity (RIC) | Lower-dose alkylator + fludarabine ± low-dose TBI | Partial ablation; relies more on graft-vs-leukemia effect | Lower regimen toxicity; usable in older/frailer patients |
| Non-Myeloablative (NMA) | Fludarabine + very low-dose TBI (~2 Gy) | Minimal cytoreduction; mixed chimerism common initially | Best tolerated; slowest engraftment, higher early relapse risk |
With the recipient's marrow emptied, HLA-matched donor hematopoietic stem cells are infused intravenously — deceptively simple in delivery, remarkably precise in destination. Within hours, circulating stem cells navigate the entire vascular system and selectively lodge in the bone marrow, guided by a well-characterized chemokine signaling axis.
Donor hematopoietic stem cells are collected by one of three routes: direct bone marrow harvest from the posterior iliac crest under anesthesia, peripheral blood stem cell (PBSC) collection via apheresis after mobilizing cells out of the marrow with G-CSF, or banked umbilical cord blood. PBSC is now the most common source, yielding higher CD34+ cell doses and faster neutrophil engraftment than marrow harvest, at the cost of a higher chronic GVHD rate due to the larger T-cell content collected alongside the stem cells.
The graft is infused intravenously through a central line over 30–60 minutes, much like a standard blood transfusion. Despite entering the bloodstream anywhere in the body, stem cells do not engraft randomly — they must actively migrate out of the vasculature and into marrow-specific niches, a process called homing.
Homing begins with rolling adhesion: donor HSCs expressing P-selectin glycoprotein ligand-1 tether loosely to marrow sinusoidal endothelium. Firm adhesion follows via VLA-4 integrin on the stem cell binding VCAM-1 on endothelial cells. The decisive directional cue is chemotactic: bone marrow stromal cells constitutively secrete SDF-1 (CXCL12), which binds the CXCR4 receptor densely expressed on the HSC surface, drawing the cell through the endothelium (transmigration) and into the vacated endosteal niche.
This single chemokine axis is so central to homing that pharmacologic CXCR4 antagonists (used therapeutically to mobilize stem cells for collection) work by transiently disrupting the very same signal that normally retains cells in the marrow.
Once lodged in a vacated niche, each donor HSC is not yet dividing — this stage represents colonization, not expansion. A successful graft distributes stem cells across many niches throughout the axial and long-bone marrow rather than clustering in one location, which improves the odds that enough clones survive early apoptotic loss to sustain durable, polyclonal, long-term hematopoiesis rather than a fragile graft dependent on only a handful of surviving stem cell clones.
Colonized donor stem cells now begin to divide and differentiate down the myeloid and lymphoid lineages, gradually repopulating a marrow left empty by conditioning. Clinically, engraftment is not judged by imaging or biopsy in real time — it is tracked through daily blood counts, and one number matters more than any other in the first weeks after transplant: the absolute neutrophil count.
Each surviving donor HSC undergoes asymmetric division: one daughter cell retains stem-cell identity (self-renewal), while the other commits to a progenitor lineage — myeloid or lymphoid — and begins a cascade of further divisions and maturation steps that ultimately produces mature neutrophils, platelets, and red blood cells. Because neutrophils have the shortest maturation and circulating half-life of the major blood lineages, their reappearance in peripheral blood is the earliest reliable signal that donor hematopoiesis has taken hold.
During the pre-engraftment aplastic window, patients are profoundly neutropenic and are nursed in protective isolation, since even normally harmless organisms can cause fulminant infection when neutrophils are absent.
By clinical convention, neutrophil engraftment is defined as the first of three consecutive days with an absolute neutrophil count above 500 cells/µL, typically occurring day 14–21 after peripheral blood stem cell infusion (marrow grafts and cord blood engraft somewhat later). Platelet engraftment — sustained count above 20,000/µL without transfusion support — usually lags neutrophil recovery by several days to a week.
Recombinant G-CSF (filgrastim) is frequently administered post-infusion to accelerate neutrophil recovery by 3–5 days, shortening the highest-risk infectious window. Reduced-intensity conditioning, by leaving more residual host stromal support intact, can produce a shallower nadir but does not necessarily accelerate the calendar timeline to full donor engraftment.
Rapid ANC recovery is protective against early infection, but the speed of donor hematopoietic takeover is not the only variable that matters. A regimen intense enough to clear host marrow fastest also produces the deepest cytopenic nadir and the most tissue damage, which — as later stages show — primes the immune environment for graft-versus-host disease. Engraftment kinetics, chimerism trajectory, and GVHD risk are mechanistically linked outputs of the same upstream conditioning and HLA-matching choices, not independent dials.
As donor hematopoiesis expands, the recipient's blood and immune system are gradually and almost completely replaced by donor-derived cells — a state called chimerism. This is not merely a passive population replacement: donor T-cells arriving with (or generated from) the graft actively patrol the recipient's tissues, and a subset recognize and destroy residual leukemic cells the conditioning regimen failed to eliminate.
Because donor and recipient cells are otherwise indistinguishable under a microscope, chimerism is tracked molecularly. Short tandem repeat polymerase chain reaction (STR-PCR) exploits the fact that donor and recipient DNA differ at multiple short repetitive sequence loci scattered across the genome. Peripheral blood (often separated into whole blood, T-cell, and myeloid fractions) is amplified at these loci, and the relative peak heights of donor- versus recipient-specific alleles quantify the percentage of donor-origin cells in each compartment.
Full donor chimerism is conventionally defined as more than 95% donor-derived cells; values between roughly 5% and 95% are termed mixed chimerism. Lineage-specific chimerism is particularly informative — declining donor chimerism in the T-cell compartment, even with stable myeloid chimerism, can be an early warning sign of impending relapse or graft rejection, often preceding any change visible on a standard blood count.
Donor T-cells and NK cells recognize residual leukemic blasts primarily through minor histocompatibility antigens — peptide differences arising from normal genetic polymorphism between donor and recipient, presented on HLA molecules that are otherwise matched. Because these minor antigens are ubiquitously expressed in normal tissue as well as leukemia, the same recognition machinery that mediates graft-versus-leukemia killing is mechanistically inseparable from the machinery that can mediate graft-versus-host disease — a central tension explored fully in Stage 5.
NK cells contribute an additional, HLA-independent layer: in haploidentical transplants especially, donor NK cells lacking inhibitory killer-immunoglobulin-like receptors (KIR) for the recipient's remaining HLA class I molecules become potently cytotoxic against leukemic blasts through "missing-self" recognition.
When chimerism monitoring shows a rising recipient fraction, or overt relapse is detected, physicians can administer a donor lymphocyte infusion (DLI) — an infusion of additional donor T-cells collected from the original donor, given without further conditioning, to reinforce the graft-versus-leukemia effect. DLI achieves durable remission in a substantial fraction of relapsed chronic myeloid leukemia cases and a smaller but meaningful fraction of relapsed acute leukemias, though it carries a dose-dependent risk of triggering or worsening GVHD.
The same donor T-cell reactivity that eliminates residual leukemia can turn against the recipient's own healthy tissue — graft-versus-host disease. GVHD is the leading non-relapse cause of morbidity and mortality after allogeneic transplant, and managing it without simply abolishing the graft-versus-leukemia effect that protects against relapse is the defining clinical balancing act of allo-BMT.
Acute GVHD develops through a well-characterized three-phase cytokine cascade. Phase one begins during conditioning itself: chemotherapy and radiation damage host epithelial tissue — particularly gut mucosa — releasing danger-associated molecular patterns and permitting bacterial lipopolysaccharide to translocate into circulation, activating host antigen-presenting cells and priming a highly inflammatory environment even before any donor cell arrives.
Phase two is donor T-cell activation: mature donor T-cells transferred with the graft encounter these activated host antigen-presenting cells, recognize mismatched major or minor histocompatibility antigens, and proliferate vigorously, producing a T-helper-1-skewed cytokine storm (IL-2, IFN-γ, TNF-α).
Phase three is the effector phase: cytotoxic donor T-cells, NK cells, and inflammatory cytokines directly damage target epithelial tissue in the skin (rash, desquamation), gastrointestinal tract (diarrhea, mucosal sloughing), and liver (cholestatic hyperbilirubinemia) — the three classic organs of acute GVHD, graded clinically I through IV by extent and severity.
GVHD and graft-versus-leukemia arise from the same donor T-cell alloreactivity aimed at different targets — minor histocompatibility antigens on healthy tissue versus the same antigens on leukemic blasts. Eliminating GVHD risk entirely (for example, via aggressive T-cell depletion of the graft) reliably increases relapse risk in parallel; effective prophylaxis aims to dampen the former while preserving as much of the latter as possible.
The single strongest predictor of GVHD risk is HLA match quality between donor and recipient across the classical loci (HLA-A, -B, -C, -DRB1, typically assessed as an 8/8 or 10/10 match at high resolution). Each additional mismatched allele increases both the incidence and severity of acute and chronic GVHD, because a larger fraction of donor T-cells will encounter genuinely foreign, rather than self, peptide-HLA complexes on recipient tissue. Matched sibling donors carry the lowest baseline risk; matched unrelated donors carry modestly higher risk; haploidentical (half-matched) and umbilical cord blood transplants carry the highest risk per unit of donor T-cell dose, and correspondingly require more intensive GVHD prophylaxis.
Standard GVHD prophylaxis combines a calcineurin inhibitor (tacrolimus or cyclosporine, blocking IL-2 transcription in activated T-cells) with methotrexate or mycophenolate mofetil, begun before symptoms appear and continued for months post-transplant. Post-transplant cyclophosphamide, given days 3–4 after infusion, has become a particularly effective strategy — it selectively depletes rapidly proliferating alloreactive donor T-cells while sparing resting and regulatory T-cells, and has made haploidentical transplantation far safer.
Ultimately, GVHD prevention is not about eliminating donor T-cell activity but tuning it: sufficient immunosuppression and HLA matching to keep tissue-damaging alloreactivity controlled, while preserving enough graft-versus-leukemia reactivity that residual disease stays in check — the balance this simulation's two sliders are built to explore.