Clonal Hematopoiesis in Allogeneic Transplantation: Should It Influence Donor Selection?

H. Joachim Deeg, MD
Fred Hutchinson Cancer Research Center
Todd E. Druley, MD, PhD
Washington University School of Medicine

Clonal hematopoiesis (CH), the presence of hematopoietic stem/precursor cells with somatic mutations, is a defining feature of hematologic malignancies. However, CH also occurs widely in the absence of clinically recognizable disease.1

The presence of CH with variant allele frequencies (VAFs) of at least 2% in patients who do not meet World Health Organization criteria for a hematologic neoplasm has been dubbed clonal hematopoiesis of indeterminate potential (CHIP). Age-related clonal hematopoiesis (ARCH) is another term used to describe CH that does not require a specific VAF. While rare in people younger than 40, CHIP is estimated to occur in 10% or more of people aged 60 and older.1 The presence of these somatic mutations in the blood has been linked to increased risk of hematologic cancer, incident coronary heart disease, ischemic stroke, and all-cause mortality.1,2

Because CHIP can be transferred from donor to recipient during allogeneic hematopoietic cell transplantation (alloHCT), the use of stem cells from donors with CH has come under scrutiny.3 Currently, though, there are no consensus guidelines on whether donors with CHIP should be considered for alloHCT.

ASH Clinical News asked H. Joachim Deeg, MD, and Todd E. Druley, MD, PhD, to debate the question, “Should donors with clonal hematopoiesis be used for alloHCT?” Dr. Deeg was asked to represent the “yes” opinion and Dr. Druley was asked to represent the “no” opinion.

H. Joachim Deeg, MD: At our center, we do not routinely screen prospective donors for CH, even older donors. There are still many questions ab-out the extent to which factors other than the mutation in hematopoietic stem cells – such as the function of the marrow microenvironment in an individual transplant recipient – will determine how a transplanted clonal hematopoietic cell will behave.

We have known since the 1970s that a small proportion of transplanted patients will develop leukemia or myelodysplastic syndromes derived from donor cells. We did not have mutation markers in those patients, but when the patient and donor were of different gender, sex chromosome markers proved useful for confirming whether a leukemia was donor-derived.4 Looking back, those donor-derived leukemias were probably patients who received cells from a donor with CH; however, in most (if not all) cases, the donor, followed for years, remained healthy, suggesting that recipient factors were more important than the presence of donor CH in contributing to the leukemia.

More recent analyses show that not all new leukemias arising in patients post-transplant (i.e., distinct leukemias rather than relapse of the original disease) are derived from the donor. Some of these leukemias can be of recipient origin, supporting the concept that the microenvironment plays an important role in whether a patient develops a malignancy.5

At this time, then, I would still perform alloHCT with donors with CH, certainly in the absence of an alternative stem cell source.

Todd E. Druley, MD, PhD: I would counter that we should avoid transplanting patients with CH because we do not understand all the implications of CH on the short- and long-term health of transplant recipients.

Data from a 2019 study by Frick and colleagues in Germany showed that receiving donations from related stem cell donors who were 55 or older and had CHIP was associated with a high incidence of chronic graft-versus-host disease (GVHD). This often required intensive therapy.6 We also have evidence that CH by itself correlates with atherosclerosis, which can lead to heart attacks and cardiac dysfunction.7

We have plenty of evidence to suggest that clonal mutations have significant adverse effects on long-term outcomes of patients undergoing alloHCT. If we must transplant hematopoietic stem cells from donors with known CH, we need to be quite cautious.

Dr. Deeg: In reference to the study by Frick and colleagues, while the data demonstrated a higher incidence of chronic GVHD with donors with CH, it also showed that recipients had a lower incidence of relapse or progression, so the overall mortality was similar. So, there was a trade-off.

Dr. Druley: The other issue is that, even if we do screen for CH, there is a debate about how CH is defined. The original paper demonstrating that CH was associated with adverse outcomes looked at more than 17,000 individuals, many of whom had type 2 diabetes.1 The authors performed whole-exome sequencing from these individuals and none had myeloid leukemia, but a substantial percentage had mutations associated with myeloid leukemia in the bloodstream. The sequencing method used by the investigators had a limit of detection of only 2 to 3%, so CH was defined as a VAF of ≥2%. The investigators reported an approximate 1% increase in the risk of acute myeloid leukemia or other hematological malignancies per year for individuals with CH at these levels.

Since that threshold was suggested, I, and others, have argued that 2% is not a biological threshold but more of a technical threshold. Clones may occur at lower frequencies and have significance or at higher frequencies and have negligible impact.

I believe the definition of CH should be divorced from the idea of VAF and instead explain that it is a somatic mutation that has arisen in the hematopoietic system over the course of a person’s lifespan.

Dr. Deeg: I agree. When we talk about the correlation between CH and nonhematologic effects in the transplant setting, such as cardiovascular events, one intriguing question is whether the clone that we transplant is also responsible for the cardiovascular issues in the person from whom the cells are obtained, or whether the clone is a marker for other associated factors in that person, in addition to what we see in the hematopoietic compartment.

Dr. Druley: To answer some of these questions, we need more information from clinical trials. As Dr. Deeg mentioned, the Frick et al. study suggested that CH may augment the graft-versus-leukemia effect, which may be why there is less relapse in recipients who have related CH donors. However, there is a thin line between the graft-versus-leukemia effect and the graft-versus-host effect.

To my knowledge, the only study in alloHCT thus far that is big enough to demonstrate outcomes from CH transfer is the 500-person study by Frick et al. Other studies that have been published range from small cohorts to anecdotal case reports. Our lab published such a study in January 2020, but it included only 25 individuals and was not powered to correlate clinical outcomes.3 If we are talking about outcomes such as cardiac disease or adaptive immune defects, we will need 5, 10, or 20 years of observation in large cohorts to correlate these with clonal transfer.

Dr. Deeg: We do not yet have data to show whether the potential clinical consequences of transplantation from a CH donor outweigh the potential benefits of the transplant, but that is a question with potentially heavy implications. If an intended donor is known to have CH, and if an alternative suitable donor can be identified – a large number of HLA-typed unrelated donors are now available, and various transplant teams would look for an HLA haploidentical relative if a fully HLA matched donor was not available – then I think one may very well go that route, rather than risking the potential adverse effects associated with CH from a related donor.

In fact, there are several publications now showing that transplantation from younger HLA-matched unrelated donors may lead to superior outcomes, compared with transplants from older siblings.8 We don’t know whether those outcomes are related to the absence or lower prevalence of CH in younger donors, so this is an area of ongoing research.

Dr. Druley: Still, I think we have enough evidence now to suggest that if you have a choice between taking hematopoietic stem cells from a donor with CH versus a donor without CH, you would elect to take from the donor without CH. As Dr. Deeg mentioned, we have come to that conclusion empirically over previous decades.

If we have the option of selecting a donor who is 50 years old versus an equivalently matched donor who is 30, we will take the 30-year-old because the recipients seem to do better with stem cells from younger donors.

We may not have always known the reasons for that association, but we know if you use a more stringent definition of CH, say a VAF down to 1 in 5,000, about 30 to 40% of adults under the age of 40 will have similar hematopoietic mutations. In those older than 50, that percentage increases to nearly 100%. So, the transplant community does get concerned about using CH as a donor selection criterion because we are already stretched to find donors. We don’t want to toss out two-thirds of our donor pool.

“Not all mutations are created equal, and it is going to take a lot of data and research to figure out what role each mutation plays.”

–Todd E. Druley, MD, PhD

Dr. Deeg: I know that some U.S. transplant centers are considering screening for CH, especially in potential older donors, or are doing this already on a research basis. It may well be the way to go to further reduce the risk of donor-derived leukemia – as low as it may be. Right now, I think it is impractical to perform this screening across the board. It can be pursued at a few larger centers, and data obtained there may offer a more solid basis on which to make recommendations.

Dr. Druley: Yes – routine screening is technically possible, but real-world evidence will need to show that spending a few hundred to a few thousand dollars for sequencing prior to transplant will avoid the costs of months or years of immunosuppression for chronic GVHD or the management of cardiovascular dysfunction.

Ostensibly, the benefits would outweigh the risks in terms of initial cost, but to do sequencing for every single donor would be a tough sell in 2020. Perhaps we will get there in the future, as the price of sequencing continues to drop. At some point it might be relatively manageable compared with the other expenses related to transplant.

Dr. Deeg: Our approach to screening for CH is rather basic in this respect: Let’s say we have a patient who has an HLA-matched brother or sister who can serve as a donor. If that sibling has a low neutrophil count, for example, or some other peripheral blood abnormalities, we obtain a bone marrow biopsy from that sibling for a more detailed analysis, which, as of late, will include a mutational analysis. If results are “positive” and CH is detected, we would look for a different donor. However, we do not routinely do mutational analysis on potential donors.

Dr. Druley: The method for detecting CH is DNA sequencing, but the question is, “How sensitive do you want DNA sequencing to be?” If you perform standard “off-the-shelf” sequencing with a ~2% limit of detection, you can do that sequencing for a few hundred dollars, but you may miss several potentially meaningful mutations at lower frequencies. More sensitive sequencing adds cost, time, and analysis. The technical capabilities are there, but we do not yet know the appropriate threshold.

At some point in the future, this question will not have a simple “yes” or “no” answer. Instead, we will look at an individual’s mutational profile to determine which are advantageous or disadvantageous. They may be mixed and matched based on a variety of other donor and recipient factors.

For example, current data has shown that cardiac dysfunction is related TET2 mutations but other mutations may pose less of a risk. So, if a donor does not have TET2 mutations, perhaps you do not need to worry as much about cardiac events. There also may be mutations that lead to decreases in recurrence or relapse that are separate from those that increase the risk of chronic GVHD. Not all mutations are created equal, and it is going to take a lot of data and research to figure out what role each mutation plays.


  1. Jaiswal S, Fontanillas P, Flannick J, et al. Age-related clonal hematopoiesis associated with adverse outcomes. N Engl J Med. 2014;371:2488-2498.
  2. Steensma DP. Clinical consequences of clonal hematopoiesis of indeterminate potential. Blood Adv. 2018;2:3404-3410.
  3. Wong WH, Bhatt S, Trinkaus K, et al. Engraftment of rare, pathogenic donor hematopoietic mutations in unrelated hematopoietic stem cell transplantation. Sci Transl Med. 2020;12(526):eaax6249.
  4. Boyd CN, Ramberg RC, Thomas ED. The incidence of recurrence of leukemia in donor cells after allogeneic bone marrow transplantation. Leuk Res. 1982;6:833-837.
  5. Sala-Torra O, Hanna C, Loken MR, et al. Evidence of donor-derived hematologic malignancies after hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2006;12:511-517.
  6. Frick M, Chan M, Arends CM, et al. Role of donor clonal hematopoiesis in allogeneic hematopoietic stem-cell transplantation. J Clin Oncol. 2019;37:375-385.
    Jaiswal S, Matarajan P, Silver AJ, et al. Clonal hematopoiesis and risk of atherosclerotic cardiovascular disease. N Engl J Med. 2017;377:111-121.
  7. Shaw BE, Logan BR, Spellman SR, et al. Development of an unrelated donor selection score predictive of survival after HCT: donor age matters most. Biol Blood Marrow Transplant. 2018;24:1049-1056.