What Are the Next Steps If a Bone Marrow Transplant Fails?

When a bone marrow transplant fails, the medical team typically shifts to a tiered set of salvage strategies that can range from adjusting medications and infusing additional donor cells to pursuing a second transplant or enrolling in clinical trials. Failure itself comes in two forms: the donated cells never take hold in the first place, or they engraft initially but are later lost or overwhelmed by returning disease. Each scenario calls for a different response, and the options available depend heavily on the patient’s overall health, the type of disease being treated, and how quickly the failure is detected.

How Transplant Failure Is Defined

Transplant failure after an allogeneic stem cell transplant falls into two broad categories. Primary graft failure means the donor cells never successfully engraft, leaving the patient without functioning blood cell production. Secondary graft failure means the donor cells initially took hold but were later lost, sometimes because the patient’s own immune system rejected them, sometimes because the underlying disease came back and crowded them out.¹ The distinction matters because the salvage path for someone whose graft never took is quite different from the path for someone whose leukemia relapsed six months after an initially successful transplant.

Relapse is the more common reason a transplant is considered to have failed. The transplanted immune system was supposed to recognize and destroy residual cancer cells through what’s called the graft-versus-tumor effect. When that immune surveillance falls short, the disease returns. Several factors predict how likely this is and how hard it will be to treat: a shorter interval between transplant and relapse, a higher disease risk at the time of the original transplant, and the presence of graft-versus-host disease before relapse are all associated with worse outcomes afterward.²

Catching Trouble Early With Chimerism Testing

Before failure becomes obvious on a blood count, transplant teams look for early warning signs by measuring chimerism, the proportion of blood cells that come from the donor versus the patient’s own marrow. A drop in donor chimerism can signal either graft rejection or disease relapse weeks before clinical symptoms appear. Research shows that donor chimerism levels measured around day 30 after transplant are significantly associated with the risk of relapse and overall survival, making them a useful early alarm system.³

More granular testing can look at chimerism within specific cell lines. For acute myeloid leukemia, achieving very high donor chimerism in CD33-positive cells (the lineage most relevant to the cancer) within the first 60 days is linked to a lower probability of future relapse. Patients who reached that milestone had an estimated two-year relapse-free rate of about 76%, compared with roughly 42% for those who did not.⁴ The practical value here is that falling chimerism can trigger preemptive action before a full-blown relapse or graft failure takes root.

First-Line Responses When Chimerism Drops or Disease Returns

The least invasive initial step is to taper immune-suppressing drugs. After transplant, patients take immunosuppression to prevent graft-versus-host disease, but those same drugs also suppress the donor immune system’s ability to attack cancer cells. Pulling back on immunosuppression can sometimes reawaken the graft-versus-tumor effect, especially in patients who relapse early, before that immune response has had time to fully develop.⁵ The tradeoff is real: reducing immunosuppression increases the risk of graft-versus-host disease, so the taper has to be carefully managed.

For patients experiencing poor graft function rather than overt relapse, a stem cell boost can help. This involves collecting additional CD34-positive stem cells from the original donor and infusing them without a full conditioning regimen. Early data suggest this approach can improve blood counts without necessarily increasing graft-versus-host disease, since the infused cells are purified to contain mostly stem cells rather than mature immune cells.⁶ More broadly, graft failure may be addressed through intensified conditioning, increased cell doses, or more aggressive immunosuppression to clear whatever is blocking engraftment.⁷

Donor Lymphocyte Infusion

If immunosuppression tapering alone is not enough, the next step up is donor lymphocyte infusion, or DLI. This means collecting immune cells, primarily T cells, from the original donor and infusing them into the patient. The goal is to jumpstart the graft-versus-tumor effect by giving the donor immune system reinforcements. DLI works best in diseases that are inherently more sensitive to immune attack, such as chronic myeloid leukemia, and tends to be less effective in aggressive acute leukemias.

The main risk is graft-versus-host disease, which develops in roughly half to 60% of patients who receive DLI. The likelihood depends heavily on dose: when the infusion contains a high number of T cells, about half or more of recipients develop graft-versus-host disease, while lower doses reduce that risk to under 10%.⁸ This creates a balancing act: enough immune cells to fight the cancer, but not so many that they attack the patient’s healthy tissues. Some centers use escalating doses, starting low and increasing only if there is no response and no graft-versus-host disease.

Targeted Drug Therapies After Relapse

Advances in targeted therapies have added an important layer to post-transplant salvage. For acute myeloid leukemia, the combination of a hypomethylating agent like azacitidine with venetoclax has shown encouraging results. In a study of 67 patients who relapsed after transplant and received this combination, about 48% achieved a complete response or complete response with incomplete count recovery, and among those tested, the vast majority showed no measurable residual disease. Patients who responded had a median overall survival of roughly 20 months, compared with under 3 months for those who did not respond.⁹ Certain genetic mutations in the leukemia, particularly in NPM1 and IDH1 or IDH2, appeared to increase the odds of responding to this combination.

These drug regimens also serve as a bridge, buying time and reducing disease burden before a second transplant or a DLI. A case report described a patient with early post-transplant relapse who achieved a complete remission on venetoclax and azacitidine, then received DLI afterward to reinforce the graft-versus-tumor effect.¹⁰ In a larger multicenter study, patients with relapsed or refractory AML who received azacitidine and venetoclax as salvage and then went on to a transplant had a two-year overall survival of about 61%, with a non-relapse mortality rate around 11%.¹¹

Immunotherapy and CAR-T Cells

For B-cell cancers like acute lymphoblastic leukemia that relapse after transplant, newer immunotherapies offer additional options. Blinatumomab, a bispecific antibody that directs the patient’s T cells toward cancer cells expressing CD19, has been studied specifically in post-transplant relapse. A meta-analysis found a complete response rate of about 48% and a one-year overall survival rate of around 40%, though long-term durability remains a challenge, with two-year survival dropping to roughly 21%.¹²

CAR-T cell therapy, in which immune cells are genetically engineered to recognize and attack cancer, is also being used in the post-transplant setting. Notably, some programs use donor-derived CAR-T cells rather than the patient’s own, which avoids the problem of engineering cells from an immune system that has already failed to control the disease. Early evidence suggests that donor-derived CD19-targeting CAR-T cells can benefit patients with relapsed B-cell malignancies after transplant, with manageable side effects.¹³

A Second Transplant

When other salvage strategies fall short or when the disease comes back aggressively, a second transplant becomes the option with the highest potential for long-term cure and the highest risk. Historically, second transplants carried very high treatment-related mortality, but the use of reduced-intensity conditioning regimens has brought that risk down compared with full myeloablative conditioning.¹⁴ Even so, a recent study of second transplants for relapsed AML reported a non-relapse mortality rate of about 20% at 100 days and 28% at 18 months. Outcomes were substantially worse for patients transplanted with active disease, those who relapsed within six months of the first transplant, and those with significant pre-existing health problems.¹⁵

One of the most consequential decisions around a second transplant is whether to use the same donor or switch to a new one. The answer depends on why the first transplant failed. For true graft failure, where the problem was engraftment rather than disease relapse, switching donors appears to make a large difference. A recent study found that changing donors led to significantly better neutrophil engraftment (about 92% versus 71%), platelet engraftment (about 77% versus 52%), and one-year overall survival (about 62% versus 43%) compared with using the same donor.¹⁶ For disease relapse, the picture is less clear. One study found that selecting a new donor for a second transplant after relapse did not produce a clear survival advantage over using the original donor, though it was not harmful either.¹⁷ Using a mismatched donor for a second transplant may offer a stronger immune mismatch that could translate into a more potent anti-tumor effect, which is one rationale behind haploidentical (half-matched family member) second transplants being explored in clinical trials.¹⁸

Expanded Natural Killer Cells and Ongoing Trials

Beyond established salvage therapies, newer cellular approaches are being tested to prevent or treat post-transplant relapse. One promising avenue involves natural killer (NK) cells, immune cells that can kill cancer without needing prior exposure to a specific target. In a phase I/II trial, patients with myeloid malignancies who received high doses of donor-derived NK cells (expanded in the laboratory) after a haploidentical transplant had a two-year relapse rate of just 4%, compared with 38% in a matched control group.¹⁹ That kind of reduction is striking enough that a larger, multi-center trial (NK-REALM) was launched to test whether expanded NK cells can reliably reduce relapse after transplant for AML and myelodysplastic syndromes.²⁰

Clinical trials are especially relevant for patients whose transplant has failed, because standard options carry significant risk and uncertain benefit. Institutions like St. Jude Children’s Research Hospital run dedicated trials for patients whose cancer has returned or worsened after a prior transplant, testing approaches like haploidentical transplantation with novel conditioning regimens specifically designed for this population. If you or someone you know is facing a failed transplant, asking the transplant team about available trials is one of the most productive conversations to have early on.

The Infection Problem During Salvage

Whatever salvage path is chosen, infection risk runs through all of it. The immune system after a failed transplant is already deeply compromised: neutrophil counts may be dangerously low, the protective lining of the gut and mouth may be damaged, and cell-based immunity against viruses and fungi is often impaired for months. Salvage therapies, whether additional chemotherapy, DLI, or a second transplant, pile further immune damage on top of these existing deficits.²¹ Bacterial infections tend to dominate during the period when neutrophil counts are at their lowest, viral reactivations (particularly cytomegalovirus and Epstein-Barr virus) become a concern as immune reconstitution lags, and fungal infections can emerge at any stage. Aggressive prophylactic antibiotics, antivirals, and antifungals are standard, but infections remain one of the leading causes of death in this setting even with good prophylaxis.

Differences Between Children and Adults

Age plays a real role in how transplant failure unfolds and which salvage strategies are viable. Adults generally experience more transplant-related illness, largely because they have higher rates of acute graft-versus-host disease, though long-term survival rates after an initial transplant have historically been comparable between adults and children.²² When relapse does occur, there is evidence that the underlying immune mechanisms differ between the two groups. In adults who relapse, donor immune cells typically still dominate the blood, suggesting the cancer has found ways to evade a fully engrafted donor immune system. In children under 13, relapse is more often associated with a loss of donor T cells and NK cells, pointing to a distinct biological pathway.²³ This distinction has practical implications: strategies that boost the donor immune system, like DLI, might work through different mechanisms or have different efficacy depending on the patient’s age.

Financial Strain and Practical Barriers

The financial reality of salvage treatment after transplant failure deserves honest discussion. Even before a transplant fails, a significant portion of patients experience financial strain. In one single-center study, over half of patients undergoing transplant evaluation reported at least some financial toxicity, with nearly a quarter reporting moderate to high levels. Those patients cut spending on food and clothing, drained savings, and skipped prescriptions because of costs. Patients with moderate to high financial strain also reported worse quality of life at six months and one year after transplant.²⁴ A second transplant, an extended course of venetoclax and azacitidine, or CAR-T therapy can multiply those costs. Patients with household incomes under $60,000 were at substantially higher odds of financial toxicity. Asking about financial counseling and assistance programs early, before costs compound, is not a luxury but a necessity.

The Emotional Weight on Patients and Caregivers

A failed transplant is not just a medical event. It can be psychologically devastating for both the patient and the people supporting them. Caregivers of transplant patients report anxiety and worry as their most common negative psychological experience. Fear, specifically of graft failure, disease return, and death, runs as an undercurrent through the entire process. As one caregiver described it: “You hear one story after another of whose graft fails and whose departure date came and went and they weren’t discharged. And it really gets to you. So you just live in fear.”²⁵

A particular source of distress that is often underappreciated is the existential burden caregivers face when asked to provide emotional support around fears of dying. In studies, caregivers ranked providing emotional support and discussing concerns about death as the most distressing parts of their role, more difficult than physical caregiving tasks.²⁶ At the same time, caregivers also report personal growth and stronger relationships with the patient, a complexity that simplistic “caregiver burnout” framing can miss.²⁷

When to Talk About Palliative Care

Palliative care is widely misunderstood as giving up. In the transplant world, it is better understood as a parallel track that manages symptoms, supports decision-making, and improves quality of life while active treatment continues. For patients facing post-transplant relapse, especially older adults or those with significant health problems, the evidence points strongly toward incorporating palliative care early rather than reserving it for the very end.²⁸ This means having honest conversations about prognosis, goals of care, and advance directives while still pursuing salvage therapies.

One of the barriers to early palliative care in transplant patients is the treatment culture itself. Transplant medicine is inherently aggressive, oriented toward cure, and the shift to acknowledging that cure may not be possible is difficult for clinicians, patients, and families alike. Proposed solutions include better education for both oncologists and palliative care providers, integrated care models that embed palliative support into the transplant process from the beginning, and institutional commitment to making these conversations routine rather than exceptional.²⁹ If your transplant team has not raised the topic, it is entirely appropriate for you to ask about it. Having a palliative care specialist involved does not mean stopping treatment. It means having more support while you figure out what comes next.