Anti-K in patients with a possible Kell null phenotype

[Matthew Kim]

Dear fellow blood bankers,

A 50-year old patient presented with an early gastric cancer to a korean university hospital.

He underwent pre-transfusion testing for pre-op work-up and the antibody identification revealed that anti-K(KEL1) was identified.

The Korean population is known to have only K-k+ phenotype (100% KEL1 negative in several donor cohort studies. 

I thought that the patient might have been immunized to the previous transfusion from foreign donors. 

Suprisingly, his phenotype was K-k-. We repeated the testing, revealing the same result. 

We genotyped Kell groups (targeting biallelic SNPs) and his predicted phenotype was K-k+, Kpa-Kpb+, Jsa-Jsb+ compatible with that of typical Koreans.

Considering anti-ku was not identified in this patient, do you think his phenotype might be a Kell null phenotype?

It is my first time or my country's first time to encounter K-k- phenotype with anti-K.

Could you explain this case  and teach me what I can do for solving the mystery? I need your help. 

[Malcolm Needs ☆]

The Ko phenotype is incredibly rare in all ethnic groups, but, some cases have been published involving a transient loss of Kell antigens, and the concurrent appearance of apparent antibodies directed against one or more of the antigens within the Kell Blood Group System.

"Naturally Occurring" cases of anti-K are not unknown, but, once again are very rare.  A few of these have appeared in the literature, such as:

Morgan P, Bossom EL.  "Naturally Occurring" Anti-Kell (K1):  Two Examples.  Transfusion 1963; 3: 397-398.

Marsh WL, Nichols ME, Oyen R, Thayer RS, Deere WL, Freed PJ, Schmelter SE.  Naturally occurring anti-K stimulated by E. Coli enterocolitis in a 20-day-old child.  Transfusion 1978; 18: 149-154.

Kanel GC, Davis I, Bowman JE.  "Naturally-occurring" anti-K1:  Possible association with mycobacterium infection. Transfusion 1978; 18: 472-473.

Algora M, Barbolla L, Contreras M.  Naturally occurring anti-D, anti-K, anti-Fya and anti_Leab.  Vox Sanguinis 1991; 61: 141.

In each case, you will notice, there is either an accompanying infection, or, in the last case a form of neoplasm.  This may fit with your patient, if the early gastric cancer has allowed the escape of, for example, E coli into his circulation.

I was also interested in the fact that you tested the patient's red cells, and found them to be phenotypically K-k-, but genotypically KEL: -1, 2, -3, 4, -6, 7.  Did you also test the red cells with anti-Kpa, anti-Kpb, anti-Jsa and anti-Jsb, to ensure that these antigens were not detected?  I am not asking this to be facetious, but because there have been examples of an apparent lack of the k antigen due to amino acid residue substitutions either at position 193, usually threonine for the k antigen, or very close to position 193 (see Millard GM, Lopez GH, Turner EM, Lizarazu ME, Roots NM, Liew Y-W, Flower RL, Hyland CA.  Modified expression of the KEL2 (k) blood group antigen attributed to p.Leu196Val amino acid change three residues from the K/k antigen polymorphism site: implications for donor screening.  Transfusion 2019; 59: 1156-1158 and Yazdanbakhsh K, Lee S, Yu Q, Reid ME.  Identification of a defect in the intracellular trafficking of a Kell blood group variant.  Blood 1999; 94 (1): 310-318).  However, the amino acid residue substitutions can be "geographically remote" from position 193 affecting the expression of the k, and other Kell antigens (see Velliquette RW, Hue-Roye K, Lomas-Francis C, Gillen B, Schierts J, Gentzkow K, Peyrard T, von Zabern I, Flegel WA, Rodberg K, Debnath AK, Lee Soohee, Reid ME.  Molecular basis of two novel and related high-prevalence antigens in the Kell blood group system, KUCI and KANT, and their serological and spatial association with K11 and KETI.  Transfusion 2013; 53: 2872-2881)).

To complicate matters further, some anti-k reagents may give weak or negative reactions, while others give apparently normal reactions.  I remember a case I was involved in myself.  We were following a woman with anti-D during her pregnancy.  She was K+k+, and her partner was D+, K-.  Upon delivery, her baby typed as K+k- in our hands (which excluded the father, unless he had a Ko haplotype).  Sadly, he was no longer available to check his red cells again.  I sent a sample of the baby's blood down to the IBGRL, and they made the baby a straightforward K+k+.  Anyway, to cut a long story short, they were using an anti-k from a different clone to the one we were using, so I sent Joyce Poole some of the anti-k we were using, and Lo and Behold, they also got a negative reaction!  I asked if they would perform a KEL gene sequence, and they did find a mutation, miles away from where the KEL2 locus was found, and yet it affected the expression of the k antigen.  Sadly, I can't remember exactly the location of the mutation, and, because we couldn't type Dad again (or sequence his KEL gene, we couldn't prove it was inherited, and so could not write up the case.

So, what to do?

1.  Retest the patient's k antigen using a selection of anti-k reagents with different clones.

2.  If you haven't already done this, test for the expression of the Kp(a), Kp(b), Js(a) and Js(b) antigens, to see if the patient is, at a phenotypic level either a Ko or a Kmod.

3.  It might be worthwhile performing adsorption and elution tests, IF these are negative.

4.  It COULD, POSSIBLY, be worthwhile just checking that the patient has a normal XK gene at position XP21.1, as, of course, it is possible to have the McLeod phenotype without having McLeod syndrome (in other words, these people do not have Chronic Granulomatous Disease [CGD]) - we had a   donor like this where I worked (the only one in the UK).

5.  If transfusion is required in the meantime, give K- IAT cross-match compatible blood.

SORRY FOR THE VERY LONG POST.

[Matthew Kim]

Thank you for Malcom.

Your explanation greatly helps as always.

Unfortunately, we don't have reagents such as anti-k, -Kpb, and -Jsb.

We are thinking about purchasing them from commercial companies.

You recommend using anti-K from different clones, but is it okay to use polyclonal anti-k from Grifols?

Otherwise, we will purchase a monoclonal anti-k from a different clone.

Secondly, is it okay to purchase polyclonal anti-kpa, -kpb, and-Jsb from Grifols?

I am not relevant to Grifols, but I just wonder if I should go with a monoclonal reagent or a polyclonal reagent for

Kell antigen typing.

Once again, I appreciate your insightful explanation. Now I am staying in Liverpool for travel, so we are not far way right

now.

 

[Malcolm Needs ☆]

Hi Matthew,

Thank you so your kind words.

I can see no reason why you should not buy either monoclonal, or polyclonal reagents and, given your genotyping results, I would advise (for this case anyway), that you only buy anti-k, anti-Kpb and anti-Jsb, because you really only need to test for those antigens that you would expect to be expressed.  To be honest, I'm not certain whether monoclonal antibodies are available for all of these specificities.

I hope you enjoy your visit to the UK.

Best wishes,

Malcolm

[Matthew Kim]

Today, We performed Kell antigen phenotyping, revealing k-, Kpb-, Jsb- (genotypically predicted as positive using SNP typing).

Now, I believed his phenotype is Ko. 

I consider further genotyping using Sanger sequencing.

Thank you for your help.

 

Matthew Kim

[Bb_in_the_rain]

On 7/8/2019 at 3:16 AM, Matthew Kim said:

Today, We performed Kell antigen phenotyping, revealing k-, Kpb-, Jsb- (genotypically predicted as positive using SNP 

Now, I believed his phenotype is Ko. 

I consider further genotyping using Sanger sequencing.

Thank you for your help.

 

Matthew Kim

Wow! things must come in a bundle. This week, we just had a case with a broadly specific antibody, non reactive with K0 cells. K-k-Kpa-Kpb- and SNP genotyping predicted a presence of KEL gene. We are on the same page with you on our patient as well, in the process of sequencing.. It will be very interesting to see the sequencing results. Please keep us posted when your sequencing is done. Our patient is a bleeder, so we gave one Ko unit.  yike!! 

Edited July 11, 2019 by Bb_in_the_rain

[gagpinks]

On 7/4/2019 at 11:57 AM, Matthew Kim said:

It is my first time or my country's first time to encounter K-k- phenotype with anti-K.

Hi Malcom

Could it be Matuhasi-Ogata phenomenon?

[Malcolm Needs ☆]

Almost certainly not.  Although Race and Sanger wrote about this phenomenon (very briefly) in their book, "Blood Groups in Man", I remember very distinctly Ruth Sanger telling me that she have grave doubts that such a phenomenon actually existed.