The question comes back on breeders' forums regularly: “I have heard that kittens with isoerythrolysis can be given serum from an adult type A cat, but my vet says it is nonsense and cannot possibly work.” I checked what the peer-reviewed research says on the subject. The answer is unambiguous — and at the same time it explains why both sides of this argument have a point.
Before you read on
I hold a doctorate in chemistry and I have been a breeder, I am not a veterinary surgeon. What follows is a description of what particular studies showed — it is not a set of treatment recommendations. Neonatal isoerythrolysis is an emergency in which kittens die within hours, and giving a newborn anything at all, subcutaneously or intraperitoneally, is a veterinary procedure. This text is meant to help you talk to your vet using specifics from the literature, not to replace their decision.
What neonatal isoerythrolysis is
Cats have three blood types: A, B and the rare AB. The A allele is dominant over the b allele, so a phenotypically type A cat may be a homozygote AA or a heterozygote Ab — that is, a carrier of type B. A type B cat is always a bb homozygote.
Whether a cat is type A or type B is decided by the CMAH gene, which encodes an enzyme converting one sugar on the surface of the red cell into another. In type B cats the function of this gene is disrupted (Bighignoli et al. 2007). It is worth adding something the authors themselves note and that gets lost in the abstracts: they were unable to identify the variants corresponding to type AB. They proposed a third allele for the whole system, placed between A and b, but the genetics of that rare type remains unfinished business.
The heart of the problem lies in something that sets cats apart from dogs: cats have natural antibodies against the foreign blood type without any prior contact with it. In this respect they resemble humans rather than dogs — the ABO system works the same way, which is why a first incompatible transfusion in a person can be fatal. In dogs it is the opposite: they have no natural antibodies of clinical significance, so a first transfusion of incompatible blood usually passes without a reaction, and neonatal isoerythrolysis in puppies is practically never described.
Bücheler and Giger (1993) measured just how much these antibodies differ between cats of the two types. Type A cats have weak IgM agglutinins and weak haemolysins. Type B cats have very strong haemagglutinins and haemolysins. It is that asymmetry that kills kittens.
The sequence runs like this:
- The kittens are born healthy — before drinking colostrum neither IgG nor IgA can be detected in their blood (Casal et al. 1996; Levy et al. 2001 confirmed the same in all 43 kittens)
- In the first hours of life they drink their mother's colostrum
- A type B queen's colostrum contains strong anti-A antibodies
- The newborn's gut absorbs them wholesale into the bloodstream
- The antibodies attack the red cells of a type A or AB kitten
- Massive haemolysis follows
Bücheler and Giger detected colostrum-derived antibodies in newborns' blood as early as four hours after birth.
Mortality in isoerythrolysis is high, but it is worth knowing that the picture is not uniform. Silvestre-Ferreira and Pastor (2010) describe a whole spectrum: from sudden death in the first hours, through jaundice, an enlarged spleen and dark brown urine containing haemoglobin, to a form mild enough that the only visible trace left is necrosis of the tail tip — in a newborn that is the one spot not sheltered by the mother's warmth, so that is where cold IgM antibodies act. A kitten with such a tail survived isoerythrolysis, even though nobody recognised it.
In which breeds the risk is real
The frequency of type B differs dramatically. The most frequently cited data come from the American study by Giger, Bücheler and Patterson (1991), which covered 1072 non-pedigree and 1100 pedigree cats:
| Group of cats | Proportion of type B cats |
|---|---|
| Non-pedigree cats (DSH/DLH) | 0.28% |
| Abyssinian, Birman, British Shorthair, Devon Rex, Himalayan, Persian, Scottish Fold, Somali | 15–59% |
| Siamese and related breeds, American Shorthair, Norwegian Forest Cat | not a single one found |
Two caveats. First, the study concerns an American population of more than thirty years ago, and blood type frequencies differ between countries — there is no basis for transferring these figures directly to Poland. Second, a range of 15–59% is so wide that statistics are no substitute for testing. And that is exactly why, in these breeds, blood types are determined before mating rather than left to probability.
Where the immunity problem comes from
The standard preventive approach is simple: keep the kittens away from their mother's colostrum. Silvestre-Ferreira and Pastor (2010) write plainly in their review that the best method of prevention is separating the kitten from its mother for 24 hours.
And here the second problem appears. Kittens separated from their mother for the first day get no colostrum — meaning no antibodies to protect them against infection. This is called failure of passive transfer. You escape the haemolysis, but you leave the litter immunologically defenceless for the first few weeks, exactly when neonatal mortality is at its highest.
And that is precisely what serum from an adult type A cat was meant to answer.
Why 24 hours and not 72
Breeding manuals circulate a recommendation of 24 to 72 hours of separation. Measurements do not support it. Casal and colleagues (1996) checked directly how long a kitten's gut absorbs antibodies: immunoglobulin given up to the 12th hour of life was detectable in the blood, while that given at the 16th hour or later was found in no kitten at all. Since the window closes after some fifteen hours, a type B queen's colostrum stops being dangerous after a day. Keeping kittens away from their mother for another two days deprives them of milk and warmth and gives nothing in return. The authors themselves put it in their summary: kittens at risk of isoerythrolysis should be kept from their mother only for the first day of life.
What the clinical study showed
Contrary to what some breeders are told, the method has a clinical study with control groups behind it. Levy, Crawford, Collante and Papich (2001), published in the Journal of the American Veterinary Medical Association, remains the foundational paper on the question.
Design. Eleven specific-pathogen-free queens and their 43 kittens. The kittens were taken from their mothers at birth, before the first suckling, and randomly assigned to four groups:
- fed their own mother's colostrum
- deprived of colostrum, no treatment
- deprived of colostrum + serum from adult cats intraperitoneally
- deprived of colostrum + serum from adult cats subcutaneously
IgG concentrations were measured by radial immunodiffusion — at birth, after two days, and at 1, 2, 4, 6 and 8 weeks.
Result. No kitten had detectable IgG at birth. Administration of serum — both intraperitoneal and subcutaneous — produced peak IgG concentrations equivalent to those of kittens that suckled normally. Kittens deprived of colostrum and left untreated reached comparable concentrations only at 6 weeks of age.
6 weeks
This is what the method is all about. Six weeks without passive immunity falls exactly across the period of highest neonatal mortality. A kitten does not then die of isoerythrolysis — it dies of an infection it had nothing to defend itself with.
The authors' conclusion, quoted verbatim: “adult cat serum may be used as an immunoglobulin supplement in colostrum-deprived kittens”. They add a caveat of their own, though, which must not be skipped: the minimum IgG concentration that protects a kitten from infection remains unknown. We know the concentrations achieved were comparable to physiological ones; we do not know how much is actually enough.
How much serum
The figure in circulation is 150 ml/kg, given in the review by Silvestre-Ferreira and Pastor (2010) with reference to Levy's paper. I went to Levy's paper itself, and it turns out to be worth knowing where that figure came from — because in this form it is easy to apply wrongly.
Levy did not dose per kilogram at all. Each treated kitten received a fixed 15 ml of serum, in three portions of 5 ml: at birth and 12 and 24 hours later. The kittens weighed 75–120 g, so converted to body weight that came out at an average of 144 ± 24 ml/kg — and that is the figure later rounded to “150 ml/kg”. In total each kitten received around 360 mg of IgG. The authors recorded that no kitten had complications after intraperitoneal or subcutaneous administration.
Why I would rather give the schedule than just “150 ml/kg”
Written as “150 ml/kg”, a hundred-gram kitten comes to 15 ml, and that can be read as a single dose. In the study it was the sum of three portions spread over a day. That also explains a discrepancy circulating in the manuals: some sources give “3–5 ml”, others “15 ml” — and both figures may come from the same paper, one describing a single portion and the other the total. What follows from this for a particular litter is for the attending vet to decide.
Why vets see that it does not work
Here we come to the heart of the misunderstanding — because they are usually right about what they saw.
Giving it by mouth
This is the most important reason for failure. The breeder squirts the serum into the mouth, like milk. But the protocol in the studies is parenteral.
It is worth noting that the sources do not list identical routes. Levy and colleagues tested the intraperitoneal and subcutaneous routes; Münnich, in her review, writes of the intraosseous and subcutaneous ones. Common to both is the subcutaneous route — and that one has both a measurement in Levy and a recommendation in Münnich. Which to choose in a particular case is for the vet to settle; what matters for us is what is not on that list.
The difference has been measured directly. Crawford, Hanel and Levy (2003) divided 77 kittens from thirteen litters into nine groups and compared oral with subcutaneous administration. Kittens given immunoglobulin subcutaneously had significantly higher serum IgG concentrations than those given it orally — and only in the subcutaneous group were the concentrations judged sufficient to protect against infection.
An aside from the same paper
Crawford and colleagues also checked whether cat serum could be replaced with equine immunoglobulin, which is far easier to obtain. Blood concentrations were achieved without difficulty — but the equine antibodies did not support the uptake of bacteria by feline granulocytes, whereas the feline ones clearly did. The authors judged equine immunoglobulin unsuitable for this use. Simply achieving a “nice IgG result” therefore does not yet mean the kitten is protected.
Giving it after gut closure
The second reason is timing. If serum reaches the mouth on the second or third day of life, it cannot possibly work — the gut closed some fifteen hours after birth and the immunoglobulin is simply digested like any other protein. A vet who has seen this drew a logical conclusion. Except that what they observed was an error of execution, not a flaw in the method.
Too small a dose
Forums and older manuals circulate a recommendation of one or two millilitres per kitten. I was unable to trace the paper it is supposed to come from — so I treat it as hearsay, not as data from the literature. Set against 150 ml/kg, meaning around 15 ml for a hundred-gram kitten, it is a difference of an order of magnitude. With a token dose the effect will be token too.
An unscreened donor
Serum taken from a cat whose blood type has not been confirmed in a laboratory, or which has not been tested for FIV and FeLV, is a separate risk — this time one we hand the kittens ourselves.
The donor's genotype does not matter
This question comes back on forums just as often: does the donor have to be genetically AA, or is phenotypically A, which includes Ab, enough?
A confirmed type A phenotype is enough. The reason is simple: serum is the fraction of blood without cells, so the donor's blood group antigens are not present in it at all. All that counts is the antibody profile, and Bücheler and Giger (1993) characterised that by blood type, not by genotype — and found that type A cats have weak anti-B antibodies. Silvestre-Ferreira and Pastor add that in some type A cats their titre cannot be measured at all.
To be honest about it: nobody has directly compared serum from an AA donor with serum from an Ab donor. Biologically there is no reason for them to differ — since serum contains no red cells, the donor's genotype has no way of showing up in it — but that is reasoning from mechanism, not a measurement.
The genotype does become crucial elsewhere, though — in planning matings:
- an AA male × a type B queen — all the kittens will be type A, so all are at risk
- an Ab male × a type B queen — statistically half the kittens type A and half type B, so only part of the litter is at risk
For a breeder of British Shorthairs or Devons that is the basis of planning. For a serum donor it is irrelevant.
What fostering onto another queen cannot replace
The natural instinct is: why serum, when you can foster the kittens onto another nursing type A queen? That works, but only if that queen is in the colostral phase, meaning she has recently given birth herself.
Claus and colleagues (2006) tested this directly. Colostrum-deprived kittens fostered onto queens already in the milk phase had IgG concentrations identical to kittens fed milk replacer — that is, despite suckling they were deficient in passive immunity. Through the first four weeks of life their IgG was significantly lower than in colostrum-fed kittens.
Two papers that disagree — and why it matters
Casal and colleagues (1996) concluded that immunoglobulin concentrations in a queen's colostrum and milk do not differ significantly, and wrote that milk from any period of lactation could replace colostrum. Ten years later Claus and colleagues designed a study precisely to settle this — and showed that queens do after all have a colostral phase: immunoglobulin concentrations on the day of birth were significantly higher than on the seventh day of lactation, and fostering onto a queen in the milk phase did not protect the kittens. The newer paper, aimed squarely at the question, therefore says something different from the older one. I mention it because breeding manuals still circulate that first version.
What science has not settled
Honesty requires pointing out the gaps, and they are sizeable:
- There is no study comparing serum from an AA donor with serum from an Ab donor. Biologically it should make no difference — but no formal comparison has been made.
- It is not known what IgG concentration actually protects a kitten. The authors of the Levy paper say so themselves. We know it matches physiological levels; we do not know how much is enough.
- The rise in IgG is proven, the effect on survival is not. The endpoint in Levy and in Crawford was antibody concentration, not the number of kittens that survived. See the box below: the only study that counted deaths found no difference.
- The description of the Levy study is internally inconsistent about group assignment. The heading says “randomized controlled study”, while the methods say the kittens were assigned “sequentially”. Those are not the same thing, even though both appear in the paper side by side.
- There is no controlled study in kittens that already show signs of isoerythrolysis. The papers concern colostrum-deprived kittens, meaning the prevention of an immunity deficit — not the treatment of haemolysis already under way. That distinction gets lost in online summaries, and it is fundamental.
The latest data: more antibodies does not automatically mean fewer deaths
In June 2026 a study appeared that did not exist before — and it cools the optimism. Chapman and colleagues analysed the fate of 544 orphaned kittens at the San Diego Humane Society shelter. One hundred and ninety-three received 3 ml of plasma subcutaneously on admission; 351 received nothing.
The result: after adjusting for demographics and clinical signs, plasma administration was not associated with a lower risk of death at any point in the first thirty days. It did, however, significantly delay the onset of diarrhoea and weight loss. The strongest prognostic factors turned out to be body weight on admission and weight gain — that is, simply nutrition.
Does that overturn the Levy paper? No, because it is a different question. Levy gave 15 ml in three portions to kittens from the moment of birth; here it was a single 3 ml, five times less, to kittens up to four weeks of age, long after gut closure and after the most critical window. The study was also retrospective and not randomised — who received plasma was decided by the shelter's operational circumstances. The authors themselves write that higher or repeated doses need to be tested.
The honest conclusion for today therefore runs: that serum raises IgG is known. That it saves lives has not been shown.
Veterinary neonatology is in general a chronically underfunded field of research. The method has solid foundations, but it is not documented the way an antibiotic for pneumonia is — and it is better to know that in advance.
In summary
Serum from an adult type A cat does indeed supplement passive immunity in colostrum-deprived kittens. Behind that stand a clinical study (Levy et al. 2001), a review in Veterinary Medicine International (Silvestre-Ferreira and Pastor 2010) and a more recent review of fading kitten syndrome in the Journal of Feline Medicine and Surgery (Münnich 2022), which states plainly that antibody support can be provided by giving serum from an adult type A cat — with the caveat that it produces a lower immunoglobulin concentration than colostrum drunk in the first hours of life.
What has to be added at once is what is not known: the rise in antibodies is proven, a life saved is not. The only study that counted deaths (Chapman et al. 2026) found no lower mortality — although it used a dose five times smaller and gave it considerably later than Levy did.
The conditions on which everything depends:
- Route — parenteral. By mouth it does not work, and that is the most common source of the opinion that the method is worthless.
- Timing — the sooner the better; the gut closes after some fifteen hours.
- Dose — the literature speaks of 150 ml/kg. The common “one or two millilitres per kitten” has no source I was able to trace.
- Donor — a confirmed type A phenotype, healthy and screened.
If a vet says “it does not work”, it is worth asking in what situation they saw that. Usually it turns out to have been given by mouth, after gut closure, or in a token dose — and then it really could not have worked.
And whether it is worth reaching for at all in a particular litter, in what dose and when — that is a question for the attending vet, not for an article. What this text gives you for that conversation is specific papers you can point to.
References
- Levy, J.K., Crawford, P.C., Collante, W.R. & Papich, M.G. (2001). Use of adult cat serum to correct failure of passive transfer in kittens, Journal of the American Veterinary Medical Association, 219(10), 1401–1405doi:10.2460/javma.2001.219.1401
- Casal, M.L., Jezyk, P.F. & Giger, U. (1996). Transfer of colostral antibodies from queens to their kittens, American Journal of Veterinary Research, 57(11), 1653–1658↗
- Silvestre-Ferreira, A.C. & Pastor, J. (2010). Feline neonatal isoerythrolysis and the importance of feline blood types, Veterinary Medicine International, 2010, 753726doi:10.4061/2010/753726
- Claus, M.A., Levy, J.K., MacDonald, K., Tucker, S.J. & Crawford, P.C. (2006). Immunoglobulin concentrations in feline colostrum and milk, and the requirement of colostrum for passive transfer of immunity to neonatal kittens, Journal of Feline Medicine and Surgery, 8(3), 184–191doi:10.1016/j.jfms.2006.01.001
- Crawford, P.C., Hanel, R.M. & Levy, J.K. (2003). Evaluation of treatment of colostrum-deprived kittens with equine IgG, American Journal of Veterinary Research, 64(8), 969–975doi:10.2460/ajvr.2003.64.969
- Bücheler, J. & Giger, U. (1993). Alloantibodies against A and B blood types in cats, Veterinary Immunology and Immunopathology, 38(3–4), 283–295doi:10.1016/0165-2427(93)90088-L
- Giger, U., Bücheler, J. & Patterson, D.F. (1991). Frequency and inheritance of A and B blood types in feline breeds of the United States, Journal of Heredity, 82(1), 15–20doi:10.1093/jhered/82.1.15
- Münnich, A. (2022). Fading kitten syndrome: factors predisposing to 'faders' and treatment options, Journal of Feline Medicine and Surgery, 24(3), 243–256doi:10.1177/1098612X221079710
- Bighignoli, B., Niini, T., Grahn, R.A., Pedersen, N.C., Millon, L.V., Polli, M., Longeri, M. & Lyons, L.A. (2007). Cytidine monophospho-N-acetylneuraminic acid hydroxylase (CMAH) mutations associated with the domestic cat AB blood group, BMC Genetics, 8, 27doi:10.1186/1471-2156-8-27
- Chapman, J., Piraino, A., Min, K., Diniz, P.P.V.P., Noble, J. & Hedge, Z. (2026). Impact of plasma supplementation on survival and clinical illness in orphaned neonatal kittens in a retrospective shelter cohort, Journal of Shelter Medicine and Community Animal Health, 5(1)doi:10.56771/jsmcah.v5.143
- Knottenbelt, C.M. (2002). The feline AB blood group system and its importance in transfusion medicine, Journal of Feline Medicine and Surgery, 4(2), 69–76doi:10.1053/jfms.2001.0162
Frequently asked
Does serum from a type A cat really work against neonatal isoerythrolysis?
It raises antibody levels in colostrum-deprived kittens, and that is confirmed by a clinical study with control groups. Levy and colleagues (2001) gave serum from adult cats to 43 kittens from eleven litters; intraperitoneal and subcutaneous administration produced peak IgG concentrations identical to those of kittens suckling their own mother's colostrum. Three things have to be kept apart, though. First, serum replaces the missing immunity but does not treat the haemolysis itself in a kitten that has already drunk a type B queen's colostrum. Second, what is proven is the rise in antibodies, not a life saved — the only study that counted deaths (Chapman et al. 2026) found no lower mortality, although it used a dose five times smaller and gave it considerably later. Third, whether to use it in a particular litter is a decision for the vet.
Why does my vet say it does not work?
Usually because they have seen it done by mouth or done too late. A kitten's gut stops absorbing antibodies after some fifteen hours of life — Casal and colleagues (1996) showed that immunoglobulin given up to the 12th hour was detectable in the blood, while that given at the 16th hour or later was found in no kitten at all. Serum squirted into the mouth on the second day really will not work. That is not a flaw in the method, though, but in the execution.
How long do the kittens have to be kept away from their mother?
The review by Silvestre-Ferreira and Pastor (2010) speaks of 24 hours — a day, not three days. That is backed by measurement: since the gut stops absorbing antibodies after some fifteen hours, the type B queen's colostrum stops being dangerous after that. Longer separation deprives kittens of their mother's milk and warmth with no immunological gain. The length of separation in a particular litter is a decision for the vet.
Does the serum donor have to be genetically AA, or is Ab enough?
A confirmed type A phenotype is enough. Serum contains no red cells, so the donor's A and B antigens are irrelevant — what counts is the antibody profile. Bücheler and Giger (1993) showed that type A cats have weak agglutinins and weak haemolysins, while type B cats have very strong ones. A phenotypically type A cat has the same safe profile whether it is AA or Ab. The genotype does matter, however, when planning a mating.
Can kittens be fostered onto another queen instead of giving serum?
Only if that queen is in the colostral phase, meaning she has recently given birth herself. Claus and colleagues (2006) tested this directly: colostrum-deprived kittens fostered onto queens in the milk phase had IgG concentrations identical to kittens fed milk replacer — that is, a deficiency of passive immunity despite suckling. A nursing queen with an older litter is no substitute for colostrum.
In which breeds is the risk of isoerythrolysis real?
In those where type B is common. In an American study covering 1072 non-pedigree and 1100 pedigree cats, the proportion of type B cats among domestic cats was just under 0.3%, whereas in breeds such as the British Shorthair, Devon Rex, Persian, Abyssinian, Birman, Scottish Fold and Somali it fell between 15 and 59% (Giger et al. 1991). Frequencies differ between countries, so figures from the American population need not match the Polish one — which is why in these breeds blood types are determined before mating rather than relying on statistics.
