A colleague at an academic medical center in Ohio reports a case of a pregnant woman who during her 19th week of pregnancy is carrying a fetus that appears to experiencing hemolytic disease of the fetus/newborn. According to the Ohio colleague, the mother is direct antiglobulin test (DAT) negative and her serum/plasma contain an antibody with the specificity of anti-K, which is only reactive with K antigen positive red cells using GEL-AHG, PEG-AHG and SALINE-AHG methods. The antibody titer is 512 and the delta OD done on a sample of amniotic fluid is in zone 3 (using a modified Liley graph). The red cells of the purported father of the fetus appear to be positive for both K and k antigens, while the red cells of the mother appear to be negative for K but positive for k antigen. A cordocentesis was done and the fetal hemoglobin was 5 g/dL; as a result of the low hemoglobin value, the fetus received an intrauterine transfusion of an aliquot of K-negative red cells. The aliquot was crossmatch compatible (PEG-AHG) with the mother's plasma. The transfused aliquot was reportedly successful in raising the fetal hemoglobin level to 14 g/dL. The MCV of the ‘fetal’ blood sample used to measure the post transfusion fetal hemoglobin level was higher than the mother’s MCV, and the red cells of that sample did not react with anti-I (the mother’s red cells reacted with anti-I), implying that the hemoglobin value was obtained on fetal (and not maternal) blood. The Ohio colleague expressed surprise that the Kell antigen typing of the fetal blood sample was K-negative, k-positive. Numerous different reagents and methods were used in typing the fetal cells (which were DAT positive), including typing of the cells after performing elution procedures, but in no case did the fetal red cells type as K-positive. However, the fetal cells did show a 2+ positive DAT.

(Editor's note: no mention was made of an antibody specificity being determined in an eluate made from the ‘fetal’ red cells).

Amniocytes were sent for PCR typing and the fetal DNA was determined to be K1 (K) positive. The Ohio colleague mentions an abstract which appeared at the 2002 AABB meeting which in summation indicated that a 28-week old fetus had severe HDN ‘due to anti-K’ (titer 128) yet typed K1 negative by serological techniques, but K1 positive by molecular techniques. The Ohio colleague comments that his understanding of the literature is that anti-K can destroy K1 positive progenitor cells. The Ohio colleague would like to know if others have seen similar cases like their current case and the case described at the AABB in 2002 which appear to involve fetuses that type as K1 negative by serological techniques, yet by molecular testing appear to be K1 positive.

The following response has been received.

ADDENDA Sept. 16, 2003

1. A colleague in Michigan and the Editor both raised questions about the K typing of the fetus under discussion in this case report. What follows are the questions and the answers that were provided by the Ohio colleague:

• Was a blocking phenomenon of the K typing considered? The Michigan colleague is of the opinion that an IgM monclonal anti-K typing serum would give a false negative result if all K sites were blocked by anti-IgG anti-K.

The Ohio colleague responds that fetal K-antigen typing was done using 4 different anti-K reagents, using 56C treated cells and by an indirect reverse typing method, all of which typed the fetal red cells to be K-negative. The Ohio colleague also performed additional testing consisting of an eluate from the fetal red cells and typing of the fetal cells with anti-K typing serum after the fetal cells had been treated by a Gamma Eluate Kit. Testing the Gamma Eluate kit treated fetal cells with anti-K typing serum was 2+ positive (in comparison to the non-reactive result of anti-K typing of untreated fetal red cells), and a saline control was negative. The Ohio colleague then contacted Gamma technical support and they indicated that the K antigen would not be destroyed using the Gamma Eluate Kit procedure. The Ohio colleague says that they did NOT use EDTA-glycine-acid for the elution.

• If an elution procedure is used to remove coating anti-K off the fetal cells in order to type them for K antigen, what elution methods were used to remove IgG anti-K from the fetal cells? If an EDTA-glycine-acid method was used, that reagent is known to destroy KEL-system antigens. Thus, K typing after an EDTA-glycine-acid method elution would render the red cell K-negative.

The Ohio colleague states that they did NOT use EDTA-glycine-acid for the elution, and after doing K-typing of fetal red cells following a Gamma Eluate Kit procedure, the fetal red cells now tested 2+ K-positive.

• What is the specificity of the antibody that is coating the fetal red cells to make those cells 3+ positive by a direct antiglobulin test?

The Ohio colleague reports that anti-K was recovered in the eluate that was made from the fetal red cells.

In addition, the Ohio colleague reports that they incubated the mother's serum with known K-positive panel cells for 1 hour at 37C, then washed the panel cells four times with saline and performed a K-antigen typing of the panel cells. The Ohio colleague reports that the known K-positive panel cells now 'typed' K-negative.

Based on the above results, the Ohio colleague now thinks that the maternal anti-K is able to block K-typing, similar to the way that some examples of anti-D block anti-D typing. Furthermore, the Michigan colleague makes the following comment: "I think the fetal RBC K type was positive!"

Editor's comment: One must wonder if the report at the AABB in 2002 of a fetus whose red cells typed K-negative but whose genetic typing was K-positive may have had a similar explanation. Have others seen a proven example of a 'blocking' anti-K?

ADDENDA Sept. 18, 2003

2. According to M. S. Harvey, Ph.D., Laboratory of the Blood Transfusion Service, Leiden University Medical Centre, Leiden, The Netherlands (attribution used with permission), they have seen 7 cases over the last 10 years in which (verbatim) "fetal blood erythrocytes obtained from fetal cord blood samples from fetuses which were severely affected by maternal Anti-Kell (K1) serotyped Kell Negative." Dr. Harvey continues by reporting "All of these fetuses were retrospectively found to be Kell Positive by PCR genotyping. In 28 other cases of Kell antagonism, the fetal cells could be directly serotyped with monoclonal anti-Kells as being Kell Positive. So, in our practice it looks as though about 20% of the anti-Kells seem to be "blocking".

All the fetal samples were obtained from the cord vein before the first intra-uterine transfusion took place between 18 and 26 weeks gestation time. We were successsful in only a few cases in demonstrating a Kell antigenicity on the "blocked" Kell positive erythrocytes after IgG -elution using the Gamma Eluate kit. In the remaining cases too strong DAT's or too few cells prevented us from carrying out sequential elutions.

The conclusion is, as far as we are concerned:

• The clinical signs are more important than the serology
• Genotype - But a warning: genotyping can also fail as shown by the following case of Family L.

A devoted couple have had 4 (surviving) children, 3 of which were affected in utero by an aggresive maternal anti-Kell (K1). There was evidence of consanquinity in the couple's families. The 3rd child of the series was not affected and serotyped Kell Negative, Cellano Positive. The mother, as expected, serotyped Kell Negative, Cellano Positive and was assigned the probable genotype kk. The father serotyped Kell Positive Cellano Negative, and was assigned (also by a reference laboratory) the genotype KK that is, Kell homozygote. Genotyping by PCR appeared to confirm this. How could the 3rd child be Kell negative? Paternity, we were assured by the treating obstetrician, was not in question. Only after extensive studies, including anti-Kx titrations on paternal erythrocytes (performed by another reference laboratory), could the father's true genotype be assigned as K Ko. The third, and unaffected, child has the genotype Ko,k and thus does not express the Kell antigen."

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator