| |||||||||||||||||||||||||||||||
The risk factors underlying the development of transfusion-associated GVHD are incompletely defined. Patients clearly at risk7 include those with congenital immune dysfunction, those with acquired immunodeficiency related to high-dose therapies that require bone marrow transplantation for hematopoietic recovery, and those with Hodgkin's disease, perhaps as a result of intrinsic T-cell defects11. In immunocompetent patients, transfusion-associated GVHD has been observed after the transfusion of cellular components from HLA-homozygous donors to recipients heterozygous for that HLA haplotype, usually a first- or second-degree relative12,13,14,15,16,17. In Japan, because of the homogeneity of the population, transfusion-associated GVHD has also been reported after the transfusion of blood from HLA-homozygous donors to unrelated heterozygous recipients18,19,20,21. The risk of transfusion-associated GVHD remains to be established in premature newborns, organ-transplant recipients, patients with hematologic cancers other than Hodgkin's disease, and patients with solid tumors7. No cases have been reported in patients with the acquired immunodeficiency syndrome, despite their profound defects in cellular immunity22.
We describe a case of transfusion-associated GVHD in the United States that occurred after platelets from a nonconsanguineous HLA-homozygous donor were transfused into an immunocompetent patient who was heterozygous for the donor's haplotype. To date, the likelihood that donors who are homozygous for a given HLA haplotype will provide blood for unrelated recipients who share that haplotype has been judged to be remote within the United States, and the general use of gamma irradiation of cellular blood components has not been recommended. Our report helps establish the existence of an HLA-related predisposition to transfusion-associated GVHD in immunocompetent patients and suggests that guidelines for the irradiation of blood components may require reassessment.
Case Report
A 60-year-old man with a smoking history of 100 pack-years presented with hypertrophic osteoarthropathy. A chest film revealed a mass in the right upper lobe with paratracheal adenopathy, and a biopsy confirmed the diagnosis of squamous-cell carcinoma; no distant metastases were present. There was no personal or family history of malignant conditions or immunodeficiency diseases, and the patient was not taking any medications. He had never received a transfusion. Laboratory data included a normal hemogram, minimally elevated serum alkaline phosphatase level (143 IU per liter; normal, 38 to 126), and seronegative status for the human immunodeficiency virus.
The patient received two 28-day cycles of cisplatin (25 mg per square meter of body-surface area), fluorouracil (800 mg per square meter), leucovorin (500 mg per square meter), ifosfamide (1200 mg per square meter), and mesna (1000 mg per square meter); all agents were administered intravenously for 5 days, followed by a subcutaneous injection of granulocyte colony-stimulating factor (5 µg per kilogram of body weight per day) for 14 days. This treatment led to a marked reduction in the lung mass and normalization of the serum alkaline phosphatase level. The patient's nadir leukocyte count was 3500 per cubic millimeter, with 70 percent neutrophils; the lowest measured lymphocyte count was 768 per cubic millimeter six days after the completion of the second treatment cycle. Eleven days after therapy was completed, when the platelet count was 13,000 per cubic millimeter, he received six units of pooled platelet concentrates from random donors; two units of red cells were transfused four days later. Neither blood product was irradiated before the transfusion.
Chills and fever developed 12 days after the transfusion of pooled platelet concentrates. The patient had a temperature of 39 °C; blanchable, macular, and finely papular erythematous eruption limited to the trunk; and mild, nontender enlargement of the liver. Laboratory studies disclosed the following values: white-cell count, 7400 per cubic millimeter, with 88 percent segmented neutrophils, 5 percent bands, 6 percent monocytes, and 1 percent lymphocytes; hemoglobin, 9.5 g per deciliter; platelet count, 273,000 per cubic millimeter; aspartate aminotransferase, 127 U per liter (normal, 5 to 35); alanine aminotransferase, 141 U per liter (normal, 7 to 56); lactic dehydrogenase, 1108 U per liter (normal, 293 to 639); alkaline phosphatase, 158 U per liter (normal, 38 to 126); total bilirubin, 0.7 mg per deciliter (12 µmol per liter; normal, 0.2 to 1.3 mg per deciliter [3.4 to 22.2 µmol per liter]); albumin, 3.4 g per deciliter; prothrombin time, 12.5 seconds; and erythrocyte sedimentation rate, 34 mm per hour.
Despite antibiotic therapy the patient remained febrile. The eruption became dusky and slightly petechial and rapidly progressed to involve the face and upper thighs, sparing the palms and soles. Profuse watery diarrhea started on the second day of hospitalization. The aspartate aminotransferase level peaked (961 U per liter) on day 4; the alanine aminotransferase and lactic dehydrogenase levels peaked (854 and 5528 U per liter, respectively) on day 5; the gamma-glutamyltransferase level peaked (565 U per liter) on day 7; and the alkaline phosphatase level peaked (266 U per liter) on day 8. By day 8, the total bilirubin level had increased to 6.5 mg per deciliter (111.2 µmol per liter) and the serum albumin level had decreased to 1.7 g per deciliter. Blood and urine cultures showed no growth; throat, sputum, and stool cultures grew only normal flora, and stools tested negative for occult blood, fecal leukocytes, Clostridium difficile toxin, and parasites. An abdominal computed tomographic scan was normal. Serologic testing for acute or chronic infection with hepatitis A, B, and C viruses was negative. Blood cultures for herpesvirus, varicella-zoster virus, and cytomegalovirus were negative, and serologic tests for Epstein-Barr virus, toxoplasmosis, and cytomegalovirus were consistent only with the occurrence of previous infections. Antinuclear antibody and rheumatoid factor were not detected.
Transfusion-associated GVHD was suspected on the third day of hospitalization. Although the patient received methylprednisolone intravenously at a daily dose of 1000 mg for four days, the high fever, skin rash, and profuse diarrhea persisted. The results of skin and intestinal biopsies were consistent with a diagnosis of GVHD, and HLA typing of peripheral-blood lymphocytes revealed homozygous A2, B44, Cw5. Seven days after hospitalization, the patient's leukocyte and platelet counts fell to 400 per cubic millimeter (24 percent segmented neutrophils, 12 percent bands, 20 percent monocytes, and 44 percent lymphocytes) and 51,000 per cubic millimeter, respectively. Blood cultures grew Escherichia coli on the ninth day, and the patient died on day 11.
Results
Pathological Findings
A punch biopsy of the skin performed on day 4 of hospitalization (Figure 1) contained prominent epidermal basal cells with mild maturational abnormalities, multifocal dyskeratosis, exocytosis of lymphoid cells, focal adnexal involvement, and a dermal perivascular infiltrate composed predominantly of mononuclear cells. These findings are consistent with the presence of GVHD.
|
|
HLA typing was performed on peripheral-blood lymphocytes from the patient, his immediate family, and all identified donors of the unirradiated blood products that he had received. Although the patient's HLA type had not been determined before GVHD developed, studies of his family (Table 1) suggest that it was A2, B44, Cw5, A26, B38, Cw1. When GVHD was clinically evident (day 5), his HLA type was homozygous A2, B44, Cw5. HLA-typing studies on seven of eight donors of blood components revealed that a single donor (Donor 2 in Table 1) was also homozygous A2, B44, Cw5. HLA-DR typing of the patient's family and this donor suggested that the latter was homozygous and the patient heterozygous for the HLA-DR4 antigen. HLA-A2, B44, and Cw5 antigens were also noted in Donor 7.
|
Fever, diarrhea, liver dysfunction, and a skin rash characteristic of transfusion-associated GVHD developed in our patient within two weeks after he received transfusions of unirradiated platelets and red cells. Laboratory findings as well as the results of skin and duodenal biopsies were consistent with this diagnosis, and other potential causes of the clinical syndrome were excluded. The syndrome was found in a patient with non-small-cell lung cancer who was treated with conventional doses of chemotherapeutic agents, and appears to have occurred after a cellular blood component donated by an unrelated HLA-homozygous donor was transfused into a recipient who shared that haplotype.
Transfusion-associated GVHD has been reported uncommonly in patients with nonhematologic cancers. It was originally recognized in children receiving intensive therapy for neuroblastoma23,24. In one series, transfusion-associated GVHD occurred in 4 of 34 adults with solid tumors (3 with small-cell lung cancer and 1 with germ-cell cancer) who were treated with high doses of chemotherapeutic agents and autologous marrow infusions and subsequently received transfusions of nonirradiated blood products25,26. Recent case reports documenting transfusion-associated GVHD in patients with cervical, renal, and esophageal carcinomas who did not receive aggressive chemotherapy further suggest that a broader spectrum of patients with solid tumors may be at risk5,6,7,8,9,10,15,27.
Although immune deficiency has been postulated to promote the development of transfusion-associated GVHD, the extent and nature of immune compromise that may be necessary are unknown, and reports of this disease in immunocompetent patients cast doubt on this hypothesis. Our patient's nutritional and performance status was excellent, and there was no obvious clinical or laboratory indication of immunodeficiency. Although the lung cancer may have compromised the patient's immune function, it appeared to be only locally advanced, without detectable distant metastases, and was responsive to treatment. It is also possible that the treatment compromised the immune system: ifosfamide may have immunosuppressive effects similar to cyclophosphamide, which was used in the treatment of several of the reported cases of transfusion-associated GVHD23,24,25. Our results, as well as those of others,12,13,14,15,16,17,18,19,20,21 suggest that the presence of a shared HLA haplotype in blood donors and recipients may be more important in the development of transfusion-associated GVHD than is the recipient's immune status.
The incidence and severity of GVHD increase with increasing disparity in the HLA type between bone marrow donor and recipient. In contrast, transfusion-associated GVHD appears to occur when viable donor lymphocytes are not rejected by the host's immune system and consequently engraft in the recipient. It would therefore be more likely to occur if the donor and recipient shared HLA antigens, as is the case with directed donations from family members. A report from Israel13 describes the involvement of a blood-related HLA-homozygous donor who had the same haplotype as the recipient. Charpentier and colleagues reported 2 cases of transfusion-associated GVHD in patients who received platelets donated by their fathers28 and noted that 7 of 26 previously reported cases occurred in patients undergoing chemotherapy who had received blood from first-degree relatives. In addition, there have been recent reports of transfusion-associated GVHD in patients who had received blood from second-degree relatives,16,17 and these patients appear to be at even greater risk than some recipients of blood from first-degree donors. Irradiation of blood components from such directed donations, which is performed in the United States, or the use of blood from unrelated donors may avoid this complication.
In the present study, the patient's HLA type was not known before the development of transfusion-associated GVHD. However, tissue typing performed when clinical manifestations developed revealed only the phenotype of the donor, whose lymphocytes appear to have engrafted in the host. Each of the patient's three daughters inherited the same haplotype (A2, Bw56, Cw1) from her mother (Table 1), and the other haplotype in daughter 3 (A2, B44, Cw5) is identical to the homozygous HLA type identified in the patient. The best explanation for these results is that the patient's HLA haplotypes were A2, B44, Cw5 and A26, B38, Cw1. One of the seven identified blood donors (Donor 2 in Table 1) was homozygous for A2, B44, Cw5. The sharing of this haplotype therefore appears to have resulted in the engraftment of the donor's lymphocytes in the recipient and the development of transfusion-associated GVHD. HLA-DR typing studies also supported this conclusion. The incidence of the A2, B44 haplotype is 0.056 in the North American white population.
The likelihood that blood from unrelated homozygous donors will be transfused into heterozygous recipients has recently been estimated in various populations and ranges from 1 in 874 in Japan to 1 in 16,835 in France29. In Japan, where the relative homogeneity of the population increases the likelihood of such an event, reports of transfusion-associated GVHD in immunocompetent recipients are more frequent, and the incidence may be as high as 1 per 660 patients who receive transfusions18,19,20,21. A case similar to ours has recently been described in Great Britain30. Although the overall frequency with which blood from unrelated American donors transfused into immunocompetent recipients causes transfusion-associated GVHD is low, it may be as high as 1 in 7174 among whites29. Moreover, it may increase with increasing exposure to blood products from homologous donors, which occurs when multiple transfusions are given or when blood components are pooled from multiple donors.
Since transfusion-associated GVHD appears to be preventable and almost universally fatal, its recognition assumes considerable importance. Currently available techniques for the depletion of leukocytes from cellular blood components can result in 1 million residual leukocytes31. However, our lack of knowledge of the precise number or type of T cells needed to mediate transfusion-associated GVHD in humans, coupled with recent reports of the occurrence of this syndrome after transfusion of filtered blood components,32 suggests that leukocyte depletion by itself may not be adequate prophylaxis. Preliminary studies in animals indicate that ultraviolet irradiation can prevent transfusion-associated GVHD,33 but this is not yet proved in humans. Thus, gamma irradiation of cellular components before transfusion is the only proved method for preventing transfusion-associated GVHD. This case report suggests that the definition of groups at risk for transfusion-associated GVHD and related guidelines for gamma irradiation of cellular blood components may require reassessment.
We are indebted to James Crawford, M.D., for preparing the photomicrographs, to Diane Donovan, R.N., for assistance with donor contacts, and to Ms. Bernadette Miner for assistance in the preparation of the manuscript.
Source Information
From the Divisions of Medical Oncology (R.A.S., F.G.H.), Immunogenetics (E.J.K.), and Tumor Immunology (K.C.A.), Dana-Farber Cancer Institute, Boston; the Division of Dermatology, New England Deaconess Hospital, Boston (J.S.D.); the Department of Medicine, Harvard Medical School, Boston (R.A.S., F.G.H., J.S.D., K.C.A.); and the American Red Cross Blood Services, Greater Upstate New York Region, Syracuse (N.L.D.). Presented in abstract form at the American Society of Hematology Annual Meeting, December 4-8, 1992, Anaheim, California.
Address reprint requests to Dr. Anderson at the Division of Tumor Immunology, Dana-Farber Cancer Institute, 44 Binney St., Boston, MA 02115.
References
| |||||||||||||||||||||||||||||||
Related Letters:
Graft-versus-Host Disease
Petz L. D., Anderson K. C., Shivdasani R. A.
Extract |
Full Text
N Engl J Med 1993;
329:664-665, Aug 26, 1993.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
