I was reading the Reddit group “Am I The Asshole?” (which apparently I’m addicted to). Today there was a post that summarized like this:
Woman was dating man and got pregnant. They split up. Man refused to claim son as his child. Later, the man got married and had a daughter. Now the daughter has leukemia, and the man gets in touch with the woman and wants the son to be a bone marrow donor to save the daughter’s life.
Let’s ignore whatever ethical considerations there are, and focus on the science. If the man is not a match for the bone marrow transplant, is it even possible the son (half-sibling to the girl with leukemia) is? In general, how does it work that neither parent is a match, but a full sibling is? I have to assume that the likelihood a half-sibling is a match is much smaller than a full sibling.
Tissue matching is based on Human Leukocyte Antigen (HLA) genes. This region has very high genetic diversity in the population, so matches between unrelated people are unlikely.
The good news is that the HLA genes are physically located close together in a contiguous block on chromosome 6. So HLA genes are usually inherited as an intact stretch of DNA called a haplotype. Each parent carries two HLA haplotypes (we all have two copies of chromosome 6), and randomly passes one of their two haplotypes to each child, usually following simple Mendelian principles. So two full siblings have a 50% chance of inheriting the same HLA haplotype from each parent, multiplying to a 25% chance of both being the same.
Two half siblings have a 50% chance of matching one haplotype, but their chance of a match for their other haplotype is no better than two unrelated people. So more likely to be a full HLA match than two completely unrelated people, but still a much lower probability than full siblings.
You inherit one of a parent’s two haplotypes, so one is a certain match. But you never inherit the other one from the same parent, so the other one is no higher than two unrelated people.
After all, a A-type parent may have AO so could have an O or A child - or not, depending on the other parent (who could be B etc. etc. etc.). Those two are no match for each other. (Other than O compatibility)
I know you explained that very well, but I’m still a bit confused.
The donor has to match both haplotypes that the recipient has, right? Since each parent has provided a haplotype to the child, it’s a certainty that each parent has at least one matching haplotype? Or am I totally misunderstanding?
Yes. A child will be a definite match to each parent on one haplotype, but both must match. For a child to match a parent on both, an unrelated mother and father would have to be a purely chance match at one haplotype.
No, nothing to do with blood type.
This is part of the immune system called the Major Histocompatibility Complex (MHC), aka Human Leukocyte Antigen (HLA) in humans. Although the work leukocyte is in there, MHC-1 proteins are present on all nucleated cells including organ tissue - i.e. all cells other than the red blood cells where blood-type A and B antigens are expressed.
Blood type also generally has to be compatible for transplants, with the usual rules - O is universal donor, etc. But that’s only because the donor’s blood is temporarily present in transplanted tissue. If there is an HLA-matched donor, there are now techniques that allow organ transplantation when blood types are incompatible - either by treatment of the organ to clear the blood antigens, or by temporarily removing antibodies from the recipient.
But if matches between unrelated people are unlikely," but half siblings have at least a 50% chance of having at least one haplotype, then isn’t that still a significant increase in the chances that they might be able to donate?
You kinda made it sound like only full siblings would make any difference at all in terms of likelihood of matching, but that doesn’t really make sense.
The probability of a match between unrelated people is very low, something of the order of 1/100,000 (reference).
So if you think through the numbers, in practise there will only be two ways your are likely to get a match. Either find a relative like a full sibling (or an extended family member, fully related) who has a chance of having both haplotypes identical by descent, or search a large database of thousands of unrelated potential donors.
The 1/100,000 figure cited above for a full match implies (taking the square root) about 1/300 for each haplotype for unrelated people. So for a half sibling, the probability of a match is 1/2 for one haplotype and 1/300 for the other, multiplying to 1/600. Sure, 1/600 is better than 1/100,000, but it’s still very unlikely, and far more likely that you will find your match from a large database of thousands of completely unrelated donors.
It’s a 1 in 600 chance, does that make it definitely worth asking in the situation OP describes where strong emotions are involved? It’s not like it’s the child’s only chance. Most people do find matches from the large databases of unrelated potential donors. If the half sibling agrees to be tested, it’s still 99.8% likely that he won’t be a match and you have to look for an unrelated donor anyway.
What was the story about the child who was conceived as a source of spare parts for the older, medically troubled sibling? …Made into a movie with an excessively convenient alternate ending.
I wonder how likely that was. IIRC they had a number of eggs fertilized in vitro and implanted the one that best matched the troubled older sister.
It wasn’t about specific siblings so it’s probably not the one you’re thinking of, but that idea was the plot of Never Let Me Go. Spoilered because although it’s not a last minute reveal, it’s not at all obvious what’s going on at the beginning.
Just for some background, although the huge diversity of MHC haplotypes in the population is a problem for tissue transplantation, it’s an incredibly cool feature and not a bug.
Early in life, the immune system is trained to recognize and tolerate “self” antigens (everything that occurs naturally in your body), and it will subsequently attack anything “non-self” as a pathogen. If that were the whole story, it would be trivially easy for pathogens to evolve to resemble “self” and evade the immune system. But what happens is that our cells constantly sample their own contents, chopping up samples of proteins that they find and presenting them on their surface bound to the MHC proteins. So what the immune system sees is not the shape of the raw protein sample, but the combined shape of the protein sample in complex with the MHC protein. It is analogous to encryption, where the MHC is the encryption key. And because the MHC is so diverse, there are thousands of different encryption keys spread across the human population.
So, although all human cells of a given tissue type look pretty much the same across the population, it’s not good enough for a pathogen to evolve to look like (say) some part of a kidney cell. Because what the immune system is trained to recognize as “self” is not kidney cell proteins but encrypted kidney cell proteins - with any one of 1000 different encryption keys. If a pathogen adapts to one person’s encrypted kidney protein, as soon as it jumps to someone with a different MHC haplotype = different encryption key, it will be recognized as non-self and outed as a pathogen.
So our immune system is trained early in life to recognize our own “self” proteins in encrypted form, and to know that these are the self antigens it should not react to. This means that otherwise identical donor tissue, encrypted differently by a different MHC haplotype in a recipient, is interpreted by the immune system as a non-self pathogen. That’s transplant rejection.
Yes, it’s still a much higher likelihood than a random match, and since it can mean the difference between life and death, I’d say it’d be entirely reasonable to ask. I mean, there are campaigns calling for people of less common ethnicities (less common in the country they live in, that is) to register on the bone marrow donor, and that’s not going to increase the odds by anything like as much.
But it’s misleading to compare one half sibling to one unrelated person. The question is whether one half sibling increases the chance of a match significantly compared to a large database or unrelated potential donors. And at 1/600 the answer is no.
If there are no other issues to consider, of course we should encourage every potential donor to consider donation. But in OP’s story, there are other considerations. As the parent of that half sibling abandoned by his biological father, if I believed that the situation would have a significant negative emotional effect on my child, or that the type of donation involve significant physical distress or health risk, I would not hesitate to say no.
I know there’s more than one unrelated person involved for general bone marrow appeals, but some tissue types do bring it down to almost that, don’t they?
The parent might be right to say no - depends how old the child is, really; they might want to do it and might resent never being given the option to even be tested, if they find out about it later in life. It’s not life-threatening for them to donate bone marrow, after all. But from the POV of the absent parent with the kid he actually does care about, and also from the POV of his new partner, it would not be weird, unethical or wrong to ask - it’d be weird not to.
I’m not an HLA geneticist, so I don’t know if this explanation makes biological sense.
It’s possible that the haplotype passed down by Dad to the child with leukemia is very rare, so the chance of finding a population match is much lower than 1/100,000. The chance Dad passed the rare haplotype to the abandoned child is still 1/2, so (assuming Mom’s haplotypes are “common”) the chance of the half-sib being a match is still 1/600.
I’m also not a transplant doctor or pediatric oncologist. It might be that a match doesn’t have to be perfect to be useful. If they can’t find a perfect match they may just go with the closest they can get. They may hope that just by chance the half-sib is a closer match than Dad. I’m not sure if it works that way, though.
Yes, I think both of these are valid points that could significantly change the general approximate probabilities that I stated upthread.
It’s also quite possible, of course, that the father doesn’t understand the probabilities. He may be quite genuinely trying to do the best for the sick kid under the mistaken belief the the half-sib is the best or only chance.