Correspondence

Graft-versus-Host Disease

N Engl J Med 1993; 329:664-665August 26, 1993

Article

To the Editor:

In their discussion of graft-versus-host disease (GVHD) associated with the transfusion of blood from unrelated HLA-homozygous donors, Shivdasani et al. (March 18 issue)1 cite data indicating that the risk of transfusion of blood from HLA homozygotes to HLA heterozygotes with a shared haplotype is 1 in 7174 among whites in the United States2. They misinterpret this to mean that the incidence of transfusion-associated GVHD “may be as high as 1 in 7174 among whites.” If this were true, more than 1500 cases would occur in immunocompetent patients each year in the United States, since more than 15 million units of cellular blood products are transfused annually.

The authors state that “in Japan, because of the homogeneity of the population, transfusion-associated GVHD has . . . been reported after the transfusion of blood from HLA-homozygous donors to unrelated heterozygous recipients.” Hundreds of cases of transfusion-associated GVHD, some of which were originally designated as postoperative erythroderma, have been recognized in immunocompetent patients in Japan3,4. However, Takahashi et al.3 calculated that the frequency of HLA homozygosity of a donor for a haplotype shared by an unrelated patient is only approximately two to four times greater in the Japanese population than in European and North American white populations. Ohto et al.2 reported that the relative risks of such an occurrence in Japanese as compared with Canadian whites, Germans, and U.S. whites are 1.9, 3.5, and 8.2, respectively. Other factors are likely to have a role in the high incidence of transfusion-associated GVHD in Japan, particularly the use of “fresh” blood, which is often transfused immediately, and the common use of blood from family members3.

Lawrence D. Petz, M.D.
UCLA Medical Center, Los Angeles, CA 90024

4 References

1. 1

   Shivdasani RA, Haluska FG, Dock NL, Dover JS, Kineke EJ, Anderson KC. Graft-versus-host disease associated with transfusion of blood from unrelated HLA-homozygous donors. N Engl J Med 1993;328:766-770
   Full Text | Web of Science | Medline

2. 2

   Ohto H, Yasuda H, Noguchi M, Abe R. Risk of transfusion-associated graft-versus-host disease as a result of directed donations from relatives. Transfusion 1992;32:691-693
   CrossRef | Web of Science | Medline

3. 3

   Takahashi K, Juji T, Miyazaki H. Post-transfusion graft-versus-host disease occurring in non-immunosuppressed patients in Japan. Transfus Sci 1991;12:281-289
   CrossRef

4. 4

   Hidano A, Yamashita N, Mizuguchi M, Toyoda H. Clinical, histological, and immunohistological studies of postoperative erythroderma. J Dermatol 1989;16:20-30
   Medline

Author/Editor Response

The authors reply:

To the Editor: We did not wish to imply that the incidence of transfusion-associated GVHD among American whites is 1 in 7174 transfusions from unrelated donors. As we discussed, and as Dr. Petz points out, this estimate by Ohto et al.1 refers to the probability that blood from unrelated HLA-homozygous donors will be transfused into heterozygous recipients in that population. It is merely a reasonable theoretical estimate of the frequency of transfusion-associated GVHD; the actual incidence is difficult to determine from largely retrospective studies. It should be recognized that the disease is almost certainly underdiagnosed and underreported.

Although the number of cases of transfusion-associated GVHD reported in Japan far exceeds that in other countries, we know of no evidence that a definitive diagnosis has been made in “hundreds of cases” in immunocompetent patients in Japan. On the basis of retrospective surveys, the incidence after open-heart surgery has been estimated to be between 1 in 300 and 1 in 6592,3. Only a minority of cases, however, have been confirmed by histologic and HLA-typing criteria.

Other factors clearly influence the risk of transfusion-associated GVHD in immunocompetent recipients, including the number of viable and functional lymphocytes transfused. This number may be higher in freshly collected blood products than in those that have been stored before transfusion. Transfusion-associated GVHD has been reported in patients who received transfusions of stored blood products3,4. Our report adds to the evidence indicating that the transfusion of unirradiated blood products from HLA-homozygous donors to unrelated heterozygous recipients, although not a unique risk factor, is of central importance in the development of transfusion-associated GVHD regardless of the immune status of the host.

Kenneth C. Anderson, M.D.
Ramesh A. Shivdasani, M.D., Ph.D.
Dana-Farber Cancer Institute, Boston, MA 02115

4 References

1. 1

   Ohto H, Yasuda H, Noguchi M, Abe R. Risk of transfusion-associated graft-versus-host disease as a result of directed donations from relatives. Transfusion 1992;32:691-693
   CrossRef | Web of Science | Medline

2. 2

   Sakakibara T, Juji T. Post-transfusion graft-versus-host disease after open heart surgery. Lancet 1986;2:1099-1099
   CrossRef | Web of Science | Medline

3. 3

   Takahashi K, Juji T, Miyazaki H. Post-transfusion graft-versus-host disease occurring in non-immunosuppressed patients in Japan. Transfus Sci 1991;12:281-289
   CrossRef

4. 4

   Suzuki K, Akiyama H, Takamoto S, et al. Transfusion-associated graft-versus-host disease in a presumably immunocompetent patient after transfusion of stored packed red cells. Transfusion 1992;32:358-360
   CrossRef | Web of Science | Medline