A case of a zebra with spots, rather than stripes?

[Malcolm Needs ☆]

We have been following a pregnancy case from one of our local, but very large London Teaching Hospitals.

The lady in question had a very high titre of anti-D (a quantification of above 36 IU/mL in UK-speak - I don't know the actual titre, as we do not titrate cases of anti-D, but it would have been very high). The lady was also K+k+.

This lady's partner was tested as D+, K-.

When the baby was born, we were surprised to find that, as well as being D+ (as would be expected), the baby was also K+ k- in our hands.

We sent the sample to the International Blood Group Reference Laboratory for confirmation and, phenotypically and genotypically their results were K+k+ and KEL1/KEL2.

I now know the answer to this conumdrum, but would anyone care to have a go and come up with the answer?

:plotting::plotting:

[rwalter]

Did you test the baby's sample with a monoclonal anti-k reagent? Perhaps the baby lacks a common epitope of the k antigen, producing a false negative (admittedly grasping at straws).

[Malcolm Needs ☆]

Not saying yet rwalter.

Give it a couple of days.

[rwalter]

Another thought occurred to me, Malcolm: Perhaps they treated the baby's cells with dithiothreitol (DTT) to remove the maternal anti-D which presumably would have given them a positive DAT. The DTT treatment could have denatured the k antigen, causing a false negative typing.

I'll stop now and wait for you to give us the correct answer. Thanks for making me think this morning.

-Rohn

[Rh-fan]

We have had the same 2 times. But then was the father K+ k- and the child K-k+.

So I think I know the answer, in both our cases it were new versions of this geno/phenotype.

Peter

[Malcolm Needs ☆]

Not saying anything yet!!!!!!!!!!!!!

[David Saikin]

I'm assuming the baby had a +DAT due to anti-D . . . given the large amount of antibody, could the anti-D on the baby's cells be masking the k ag making it inaccessible to anti-k reagent? . . . but I really like the DTT treatment option mentioned above more.

[Malcolm Needs ☆]

The DAT was positive David, but not as "positive" as I expected - about a 2+, whereas I would have expected about a 4+.

[Mabel Adams]

When you say the reference lab results were KEL1/KEL2are we to understand those results are molecular testing or still serological? Is dad truly dad? Maybe Dad was Kpa positive and it suppressed the expression of his K1 antigen-- although I've not heard of Kpa making K1 undetectable, only weaker.

[Malcolm Needs ☆]

Hi Mabel,

The KEL1 and KEL2 results were molecular; the K+k+ results from the IBGRL were serological.

That's all I'll say for now.

[Liz]

• What was the status of the biologic father's cellano (k)?
  
• Was this naturally occuring anti D, or an over dose of RhIg?
  

[Malcolm Needs ☆]

• What was the status of the biologic father's cellano (k)?
  
• Was this naturally occuring anti D, or an over dose of RhIg?
  

Unfortunately, because we grouped the father early in the pregnancy, and we don't routinely go on to type for k, if the K is negative, we were unable to say. When the baby was typed as K+k-, we didn't want to go back and bleed the father, in case he "smelled a rat", and wondered why we were typing him again.

Knowing what we know now, however, we may well ask for a sample from him.

The anti-D was, without doubt, immune. We test anti-D "strength" by quantification, rather than titration. If the anti-D level goes above 1.0 IU/mL, we reckon it to be immune, rather than passive. Up until 4 IU/mL, there is a low risk of clinically significant HDFN. Between 4 and 15 IU/mL, there is a moderate risk. Between 15 and 20 IU/mL, there is a high risk, and above 20 IU/mL there is a risk of hydrops. This lady's anti-D level was around 36-37IU/mL - so there was no chance of it being an overdose of anti-D immunoglobulin.

[Dr. Pepper]

I'm assuming the baby had a +DAT due to anti-D . . . given the large amount of antibody, could the anti-D on the baby's cells be masking the k ag making it inaccessible to anti-k reagent? . . . but I really like the DTT treatment option mentioned above more.

That occured to me as well (although I've read of the masking effect I've never seen it in action), but thought it would have happened with the D antigen as well and the baby would have typed D-. And I would think the International Blood Group Reference Laboratory would be aware of DTT denaturation as well. I'm going on vacation and will have to wait a week for the answer!

[Malcolm Needs ☆]

And I would think the International Blood Group Reference Laboratory would be aware of DTT denaturation as well.

Believe me - my lab is aware of DTT denaturation as well!!!!!!!!!!!!!!!!!!

[Dr. Pepper]

Believe me - my lab is aware of DTT denaturation as well!!!!!!!!!!!!!!!!!!

"the baby was also K+ k- in our hands."

Oops, you were the guys who got the k-! Now I've gone and insulted your lab. I'd better quit while I'm behind.

[Malcolm Needs ☆]

Absolutely no offence taken whatsoever Phil!

[JoyDenver]

My first guess is that the dad is not the father

[Mimi]

You did not mention Dad's k typing. Could baby have inherited a suppressor gene from Dad, or suppressor genes-one from each parent?

Baby would still have the KEL1/KEL2 genes, yet be k- phenotypically.

[Yanxia]

I prefer Mabel's view.

I guess the mother is K Kpb/k kpb and the father is k kpa/?

the baby is K Kpb/ k kpa, when the K is present on the oposite chrimosmone, the Kpa weaken the expression of k, so it is hard to detect it.

Or a more dared guess is the Kpb from mother is nonexpression , the baby is Kpa homozygosity, the k is weaken too.

[Malcolm Needs ☆]

Here is the answer.

When we initially saw the results we were getting serologically with the baby's red cells (K+k-), we first thought that the mum could have been "playing away", and that the male partner was not the father. There was absolutely no doubt, however, that this was not so.

The next thing we did was to get another sample from the baby, just in case there had been a swap of cord samples - but, again, this was doubtful, as the cord was DAT+, with anti-D eluted and, although the host hospital has a Specialist Foetal Medicine Unit, no other lady under their care had given birth that day.

The new sample from the baby gave identical results (K+k-).

As a result, I sent samples to the International Blood Group Reference Laboratory (IBGRL), and their results came back as K+k+ (serologically), and KEL1/KEL2 (molecular). This foxed me a bit.

It just so happened that I was at the Centre where they are based on a training course a couple of days later, and so I went to have a word with my old friend Joyce Poole.

As you may know, Joyce published a paper in 2006 describing a Serine amino acid residue substitution at 193 on the Kell glycopreotein that meant that some anti-K reagents did not detect the "new" K antigen (Poole J, Warke N, Hustinx H, Taleghani BM, Martin P, Finning K, Crew VK, Green C, Bromilow I, Daniels G. A KEL gene encoding serine at position 193 of the Kell glycoprotein results in expression of KEL1 antigen. Transfusion 2006; 46 (11): 187901885.).

I suggested to her that their may be the possibility of a similar mutation on the KEL2 gene that meant that our monoclonal anti-k reagent was not detecting the resultant k antigen. Surprisingly, considering the small amount of evidence that I could produce, she agreed to get Vanja Karamatic Crew to perform a full gene sequence of all 19 exons of the KEL gene.

The sample was found to be heterozygous for 578C>T mutation in exon 6, resulting in a Thr193Met change in the Kell glycoprotein, which is the common (wild-type) characteristic of the K+k+ (KEL1/KEL2 genotype).

In addition, however, there was a novel heterozygous mutation in the KEL gene; 1720G>A in exon 16, encoding an Ala574Thr change in the Kell glycoprotein.

No other mutations were found in the KEL gene.

Now, although position 574 is "geographically remote" from position 193, the Kell glycoprotein is highly folded, and so these positions may not be quite so "remote" from each other as may be thought. We think, therefore, that this substitution of a non-polar amino acid for a polar amino acid may have altered the quarternery structure of the glycoprotein just enough for our monoclonal anti-k not to detect the presence of the k antigen, whilst other monoclonal anti-k reagents do.

There is, of course, a lot more work to be done to absolutely prove all this, but a lot of you got very close to the correct answer.

:clap::clap:

[Rh-fan]

You say that the k antigen canbe neg with some MoAbs but pos with an other MoAb (some kind of epitope loss (sounds like RhD:redface:)), have seen that? I only see k-. And also with a polyclonal anti k it then should be detectable.

I think change in amino acid will give foldin problems an even maybe so much that the expression is very much deminisched. Absortion elution can make the difference between a mod allele or a null allel.

Nice case.

[Malcolm Needs ☆]

Now that we are as certain as we can be that the father really is the father, we are going to try to get another blood sample from him, and work it up from start to finish, including at a molecular level.

Of course, we treated the baby's sample with chloroquine diphosphate, so that we were able to test it with and without the anti-D coating, but we got the same results with both sets of red cells.

I would be great to get to the bottom of the case, but it all depends on whether or not the father would be willing to give another blood sample.

[AMcCord]

Very interesting case. Thanks for sharing Malcolm!

[Dr. Pepper]

You say that the k antigen canbe neg with some MoAbs but pos with an other MoAb (some kind of epitope loss (sounds like RhD:redface:)), have seen that? I only see k-. And also with a polyclonal anti k it then should be detectable.

There is something to be said for the good old days when, aside from a few lectins, we had polyclonal antiserum for everything, and they reacted with a host of variants. Thanks for the case Malcolm.

[Malcolm Needs ☆]

I, sort of, know from where you are coming Phil, but, on the other hand, some of the polyclonal antisera we used to use were, quite frankly, carp (oops, I seem to have mis-spelled that), and it could be a bit of a lottery sometimes as to whether or not antigens were actually there or not. This was particularly true for Lutheran and Kidd typing. Maybe there could be a happy medium - but I have my doubts!!!!!!!!

Thanks for the comment.