Effective Prophylaxis Against Platelet ... - Semantic Scholar

7 downloads 76 Views 507KB Size Report
sion to modify alloimmunization in patients with leukemia. Blood. 622315, 1983. 22. Murphy MF, Metcalfe P, Thomas H, Eve J, Ord J, Lister TA,. Waters AH: Use ...
Effective Prophylaxis Against Platelet Refractoriness in Multitransfused Patients by Use of Leukocyte-Free Blood Components By Ulla M.Saarinen, Riitta Kekomaki, Martti A. Siimes, and Gunnar Myllyla Development of p e r m a n e n t platelet refractoriness is a major problem in multitransfused patients with d i s e a s e s s u c h as leukemia, aplastic anemia, o r pediatric solid tumors. We tried to prevent alloimmunization in t h e s e patients by s y s t e m a t i c u s e of leukocyte-free blood compon e n t s with less t h a n o n e million of contaminating leukocytes per unit of platelets o r r e d cells. Our s t u d y group comprised 26 patients with a minimum of 10 platelet transfusions per patient. T h e s e patients w e r e compared with a historical reference group of 21 patients w h o had received standard blood products. In t h e leukocyte-free g r o u p none developed platelet refractoriness, in c o n t r a s t

to t h e reference g r o u p w h e r e 11 of t h e 21 patients b e c a m e refractory to random platelets. T h e median c o r r e c t e d platelet increment for random pooled platelets was significantly higher in t h e leukocyte-free g r o u p compared with t h e reference group. The increasing number of transfusions did n o t correlate with t h e development of platelet refractoriness; instead we p r o p o s e t h a t t h e lower limit of antigenic e x p o s u r e is important. We conclude t h a t systematic u s e of leukocyte-free blood c o m p o n e n t s effectively prevents development of platelet refractoriness a n d contributes to optimal supportive c a r e of children with cancer. 0 1990 by The American Society of Hematology.

HILE MODERN anticancer therapy is becoming more aggressive, including allogeneic and autologous bone marrow transplant programs and high dose chemotherapy regimens, more vigorous measures are needed to support patients through periods of prolonged pancytopenia. Platelet transfusions are crucial in supportive care. One major obstacle is the development of platelet refractoriness, meaning that no more platelet level increments can be achieved by transfusions.' Permanent platelet refractoriness to random pooled platelets is mostly due to human leukocyte antigen (HLA)alloimmunization, which can be proven serologically.* Even antibodies reactive to platelet-specific antigens can cause destruction of transfused platelets. Of course, a poor response to transfused platelets may also be due to sepsis, fever, disseminated intravascular coagulation, splenomegaly, poor viability of platelets given, or other conditions. Prevention of HLA-alloimmunization and the problems it presents has been tried through utilizing single-donor platelet^,^*^*^ HLA-matched platelets,6 and platelets of family members. These means may be helpful once alloimmunization has occurred but are laborious and expensive. Alloimmunization and refractoriness to platelets are provoked by leukocytes present in routinely prepared platelet concentrate^.^ The risk of alloimmunization is influenced by both the total number of contaminating leukocytes and the types of leukocytes present. The most immunogenic cells in transfused blood are those expressing the major histocompatibility complex class I1 antigens,'.' ie, monocytes, B cells, and dendritic cells. Cells lacking class I1 antigens, such as

platelets and resting T cells, a r e relatively nonimmunogenic.'.'' We have focused on systematic use of leukocyte-free blood products. In 1985 we changed to exclusive use of leukocytefree blood components, including both packed red blood cells and platelet concentrates, in all patients treated a t the pediatric Hematology/Oncology Unit. W e present encouraging data indicating that this policy is effective prophylaxis against platelet refractoriness in multitransfused patients.

From the Children's Hospital, University of Helsinki, and the Finnish Red Cross Blood Transfusion Center. Helsinki, Finland. Submitted March I , 1989; accepted September 11, 1989. Address reprint requests to UIIa M . Saarinen. MD, Children's Hospital, University of Helsinki, Stenbackinkatu 1 1 , 00290 Helsinki, Finland. The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C.section I734 solely to indicate this fact. 0 I990 by The American Society of Hematology. 0006-4971/90/7502-0OaI $3.00/0 512

SUBJECTS AND METHODS

Preparation of leukocyte-free blood components. The preparation and properties of the leukocyte-free platelet and red cell products used have been previously described." Briefly, the leukocytefree red cells were prepared by combining effective buffy coat removal with subsequent filtration. After high-spin centrifugation (6000 g 4 minutes) of a fresh whole blood unit, 40 mL of buffy coat layer was carefully removed. The red cells were then suspended in 100 mL of SAGMAN solution (877 mg sodium chloride, 16.9 mg adenine, 819 mg glucose anhydride, and 525 mg mannito1/100 mL of water; Baxter SA, Le ChLtre, France) and filtered through a cellulose acetate filter (Cellselect R, NPBI, Emmer-Compascuum, The Netherlands). The resulting packed red blood cell product was characterized by harvesting all leukocytes from 39 red blood cell units after ammonium chloride hemolysis, counting the leukocytes in chamber, and typing them using a panel of monoclonal antibodies (MoAbs). The mean number of leukocytes per unit of red blood cells was 0.1 x lo6 cells, and the number per unit was always less than 1 x lo6cells. Of the remaining leukocytes, 40% to 70% were granulocytes, 30% to 50% were T cells, and less than 2% were leukocytes carrying class I1 HLA antigens (monocytes and B cells)." The leukocyte-free platelet products were prepared from 1 to 3 day old standard platelet concentrates by cotton wool filtration. In order to result in a platelet concentrate with consistently less than 1 x lo6leukocytes per unit, the number of leukocytes in thestandard platelet concentrates should be relatively low to start, and the filter should not be overloaded. The mean number of leukocytes was 100 x lo6 cells/unit in our standard platelet concentrates, and no more than four such units were pooled and filtered through one cotton wool filter (Imugard IG-500 Terumo, Tokyo, Japan). The number of platelets decreased by about 20% during the procedure. The resulting pools of leukocyte-free platelet concentrates were characterized by first removing most of the platelets by aggregating them selectively with collagen, and then counting and typing the contaminating leukocytes as described above." The mean number of leukocytes per unit of platelet concentrate was 0.04 x lo6 cells (upper range, Blood, Vol 75, No 2 (January 15). 1990: pp 5 12-5 17

513

PROPHYLAXIS AGAINST PLATELET REFRACTORINESS

0.2 x 106;n = 22). Of the remaining leukocytes, 85%to 98%were T cells, and less than 4% were leukocytes with class I1 HLA antigens (B cells and monocytes).” The methods for enumerating and characterizing white blood cells (WBCs) in the “leukocyte-free’’ products are too cumbersome for routine use. For continuous quality control purposes, microscopic WBC counting is utilized. Two whole hemocytometers (FuchsRosenthal chambers) with a volume of 3.2 pL are counted using 1: 10 dilution for platelet and 1:20 for red blood cell products. During the years of this study, about 350 units of leukocyte-free products have been screened, and the number of leukocytes has remained below the detection limit of the counting system, which corresponds to about 0.8 x IO6 cells/unit. Only two units of red blood cells have exceeded this detection limit. Patients. Our series comprised 47 multitransfused patients, with a minimum of 10 platelet transfusions per patient. The patients were 1 to 17 years of age and were treated at the Unit of HematologyOncology of the Children’s Hospital, University of Helsinki, Finland. None had prior pregnancies. The diagnoses included severe aplastic anemia, leukemia and solid tumors. Of the 47 patients, 36 had received allogeneic or autologous bone marrow transplants and/or high dose chemotherapy (Table 1). The standard platelet group (group I) is a historical reference group of 21 patients, 13 boys and eight girls, who were treated between 1978 and 1984. These patients received standard blood products only (except that during the 1980’s. buffy coat-free red blood cells were used). The leukocyte-free study group (group 11) was given leukocyte-free blood products only and consisted of 26 patients, I 5 boys and 11 girls, treated during or after 1985. Groups I and I1 did not differ regarding age or amount of immunosuppression. In group I1 we had more solid tumors and autologous bone marrow transplants (Table 1). The number of transfusions per patient is shown in Table 2. The number of red blood cell transfusions and the total number of units per patient were not different between groups I and 11. There were more platelet transfusions and less granulocyte transfusions per patient in group 11. The difference in the number of granulocyte transfusions is due to a change in the routine policy to treat sepsis. In group I1 two patients had received granulocyte transfusions on one occasion each (on two consecutive days). These patients were evalulated both before the granulocyte transfusion and as a whole, and because this did not make any difference, we did not exclude them. The individual periods of immunization time, during which all the platelet transfusions were administered, did not differ between

Table 2. Transfusions Given Per Patient Group I. Standard

Platelet Median (Range)

Group 11. Leukocyte-Free Median (Range)

Platelets Transfusions* Units

17 ( 1 0 - 1 14) 79 ( 1 8-644)

Red Blood Cells Transfusions Units*

15 (0-67) 22 (0-88)

20 (4-34) 22 (4-54)

Granulocytes Transfusions Units

3 (0-14) 16 (0-174)

0 (0-2) 0 (0-16)

140 (48-773)

143 (44-474)

Total units

22 (10-63) 114 (28-420)

~~

‘Individual platelet transfusion consisted of generally 1 to 8 units of platelets (occasionally as many a5 12). primarily depending on the size of the patient. **Partial units were used for smaller patients.

groups I and 11. In group I, the median was 6 months with a range from 1 month to 6 years, and in group 11, the median was 8 months with a range from 1 month to 2 years, 4 months. DeJnition of platelet refractoriness. Platelet refractoriness was defined as an increment/unit/m* (corrected increment) of c 2 x 109/L. One-hour posttransfusion increments were used,” unless there clearly was no platelet consumption and thus one-day increments were acceptable. The median corrected increment of, generally, the last five (never less than two) platelet transfusions with random pooled platelets was determined for each patient. If HLAmatched or family member platelets were used, a difference in increment had to be demonstrated compared with random pooled platelets. HLA-antibodies were demonstrated in several refractory patients, but because HLA- and platelet antibodies were not routinely measured in this study, serological data is not included in the definition of platelet refractoriness. We are aware that although alloimmunization to HLA antigens is a major cause of refractoriness, a single one-hour increment may be influenced by multiple clinical factors, eg, disseminated intravascular coagulation (DIC), bone marrow transplantation, prior splenectomy, splenomegaly, and to a lesser extent, by fever.” Statistical methods. The median test and Chi square test were used in the statistical analyses.

Table 1. Patients’ Characteristics

RESULTS

Group I. Standard

Platelets (1978-1984; n = 21)

Median age (range)

lO(1-17)

Diagnoses Aplastic anemia ALL ANLL Solid tumors

6 12 2

Ablative therapy Allogeneic 8MT Autologous BMT High-dose ARA-C

1

Group II. Leukocyte-Free (1985-1988: n = 26)

8 (1-17) 3 7 6 10

15

7

-

8

6

Abbreviations: ALL, acute lymphoblastic leukemia; ANLL, acute nonlymphocytic leukemia; BMT. bone marrow transplantation: ARA-C. cytosine arabinoside.

The one-hour increments to random pooled platelets remained adequate in the leukocyte-free group even after multiple platelet transfusions, while the increments were significantly lower (P < .Ol) in the standard platelet group (Fig l).In the leukocyte-free group, none had an increment of 2 x 109/L/unit/m2 or lower; while in the standard platelet group, the median increment was 2, and 50% of the subjects were below 2, ie, were refractory by our definition (Fig 1). Within the standard platelet group of 21 patients, 13 received HLA-matched platelets, and 14 received platelets from family members by apheresis. Of the 21 patients, 11 were refractory (Table 3). In contrast, in the leukocyte-free group of 26 patients, family member platelets were never needed, and none developed refractoriness (P < .0005). A few patients received occasional HLA-matched platelet trans-

514

SAARINEN ET AL

3c

needed for the development of refractoriness (Fig 2). No correlation was observed when the number of platelet units administered per patient was plotted against the platelet increment/unit/m2 in the standard platelet group and leukocyte-free group separately (Fig 3). These data show that increasing the number of transfusions is not a major factor in platelet refractoriness. Interestingly, the number of granulocyte units administered did not seem to influence the development of platelet refractoriness; within the standard platelet group, those who became refractory to platelets had received no more granulocyte transfusions than those who remained nonrefractory (Fig 4). We did not notice any correlation between the age of the patient or the duration of the immunization time while platelet transfusions were administered and the development of platelet refractoriness. Regarding the role of immunosuppressive therapy, we had in the standard platelet group four patients with severe aplastic anemia who were on conservative and supportive therapy without immunosuppression; nevertheless, one of the four never became refractory.

0

25

2c

0

0

15

0

:

0.

10

me

e

DISCUSSION

0 0

Alloimmunization against HLA-antigens is a major cause of permanent refractoriness to random platelet transfusions in multitransfused patients. Although a study using a histor-

STANDARD LEUKOCYTE- FREE PLATELETS PLATELETS

650



Fig 1. Median one-hour increment/unit/m* for random pooled platelets in the standard platelet group (left1and in the leukocytefree group (right). The symbols represent individual patients, and the horizontal lines indicate median values of the study groups (P < 0.011.

fusions, which, however, gave no better increments than random pooled platelets (Table 3). The number of platelet units administered did not correlate with the development of platelet refractoriness. Within the standard platelet group, there was no difference in the number of platelet units administered between the refractory and nonrefractory individuals (Fig 2). However, a minimum of transfusions seems to be necessary for the development of refractoriness; our cutoff limit of 10 platelet transfusions per patient was based on the fact that none among our patients below that limit was refractory. In terms of units of platelets given, a minimum of about 50 units of standard platelets was Table 3. Refractoriness to Random Pooled Platelets in Multitransfused Patients With 10 or More Platelet Transfusions Group I. Standard Platelets (n = 21)

Received HLA-matched platelets Received platelets from family members Refractory state P < 0.0005

Group 11. LeukocyteFree (n = 26)

13

5

14

0

11.

0'

r

"i W ,

0

:

1001

+ 0 0 0 0

s? t

0

Sa

R EFRA CTORY

0

.

0.0

0 0

0

NONREFRACTORY

Fig 2. The number of units of platelets given per patient in the standard platelet group. The symbols represent individual patients. Left, patients refractory to platelets; right, nonrefractory patients. There is no difference (P = .51).

515

P R O P H W I S AGAINST PLATELET REFRACTORINESS

A

B

STANDARD

25

LEUKOCYTE -FREE

n

lot

.

e* e*

0

/c----l---

100

200

600

l

I

I

I

100

2oo

300

400

NUMBER OF PLATELET UNITS GIVEN

150 -

100 -

50 -

:

0REFRACTORY STATE : YES

*:

NO

Fig 4. The number of granulocyte units given per patient in the standard platelet group. The symbols represent individual patients. Left, patients refractory to platelets; right, nonrefractory patients.

Fig 3. Correlation of the median one-hour platelet increm e n t l u n i t l m ’ for random pooled platelets to the number of platelet units given per patient, in the standard platelet group and in the leukocyte-free group. The symbols represent individual patients. No correlation was observed.

ical reference group might be subject to biases, our data strongly suggests that systematic use of leukocyte-free platelets and red blood cells is effective prophylaxis against platelet refractoriness. None of the 26 patients in the leukocyte-free study group developed refractoriness, although each patient was transfused with 28 to 420 units of platelets (10 to 63 platelet transfusions) and 44 to 474 units of blood products in total. In contrast, half of the patients in the standard platelet group became refractory. In the patient population of immunosuppressed children with cancer, the threshold level for developing refractoriness seemed to be about 10 platelet transfusions, or about 50 units of standard platelet concentrate (Fig 2). Otherwise, the development of refractoriness did not correlate to the increasing alloantigenic load, ie, the increasing number of platelets, red cells, units in total, or even units of granulocytes administered. Unfortunately, there is no way to predict who will become refractory and who will not. About half our patients transfused with standard blood components became refractory to random platelet transfusions. The rate of alloimmunization is rather high but comparable with results published earlier.’4-24On the other hand, the HLA immunization rate might be higher in patients not undergoing ~hemotherapy.’~.’~ It is evident that alloimmunization cannot be prevented with the strongest immunosuppressive or even ablative therapy, as in the bone marrow transplant regimens. We propose that the lower limit of antigenic exposure is important. Depletion of contaminating immunogenic leukocytes has been sought by several investigator^^^^^^ (Table 4). Reduction of the number of leukocytes per unit to 10% or even below 1% of the standard level seems to reduce the incidence of platelet refractorines~.’~*~~-~’ In the present study none of the patients receiving less than 1 x lo6 contaminating leukocytes per unit developed refractoriness (Table 4).

SAARINEN ET

516 Table 4. Refractoriness by Using Standard Platelets and Leukocyte-Depleted Platelets: Review of Literature

Study

No. of Patients

Leukocyteslunit

Refractwy State

A

A

A

B

Contaminating B

B

Eernisse 81 Brand,

19812”

500 x

lo6

5x

lo6 93%

27

68

24%

48

50

31

19 1,060x

lo6

60x

20

20

530 x 64 x

lo6 lo6

200x

lo6

6 x 10‘ 50% 15% 6 x lo6 47% 21% 5 x lo6 9% tl x lo6 6% 0%

Schiffer et al,

1983’’

65 x 10‘ 12 x

lo6

19% 16%

Murphy et al,

1986”

lo6 23%

5%

Sniecinski et al,

1 98823

Andreu et al, 19882435 34 Brand et al, 1 98826 335 Myllyla et al, 1988” 18 21 Present Study

21

26

200 x 10’ < 1 x

lo6 52%

0%

A, standard platelets; B, leukocyte-free/depleted/poor platelets.

Although the minimal immunogenic dose is not known, we seem to be very close, especially when the type of the remaining leukocytes is taken into account. In our blood products, the number of cells carrying major histocompatibility complex class II antigens was very low.” The use of leukocyte-free blood components as prophylaxis for platelet refractoriness is beneficial in several aspects.

AL

Leukocyte-free blood components are available to everybody and can be used independently of the patient’s HLA type or family conditions, as opposed to the use of HLA-matched products or blood donated by family members. It does not require apheresis procedures, as does the use of single-donor platelets. Even the risk of contamination with cytomegalovirus may be reduced in using leukocyte-free blood components.’’ On the other hand, this type of prophylaxis would require systematic use of leukocyte-free blood products in hematology/oncology units in order to include all potential candidates who might become multitransfused, particularly because the tendency to develop refractoriness cannot be individually predicted. The costs of transfusions will inevitably increase with such a policy. These extra costs are, however, balanced by reduced expenses on HLAmatched platelets. We conclude that systematic use of leukocyte-free blood components at the level of less than 1 x lo6of contaminating leukocytes per unit is effective prophylaxis against alloimmunization and platelet refractoriness in multitransfused pediatric patients. Our need for HLA-matched platelets and apheresis of family members has dramatically decreased with this new policy. Decreasing threat of platelet refractoriness should allow better supportive care and more aggressive approaches in anticancer therapy.

REFERENCES

1. National Institutes of Health Consensus Conference: Platelet transfusion therapy. Transfusion Medicine Reviews 1: 195,1987 2. Slichter SJ: Prevention of platelet alloimmunization, in Murawski K, Peetom F (eds): Transfusion Medicine: Recent Technological Advances. New York, NY, Liss, 1986, p 83 3. Schiffer CA, Slichter SJ: Platelet transfusions from single donors. N Engl J Med 307:245, 1982 4. Gmur J, von Felten A, Osterwalder B, Honegger H, Hormann A, Sauter C, Deubelbeiss K, Berchtold W, Metaxas M, Scali G, Frick P: Delayed alloimmunization using random single donor platelet transfusions: A prospective study in thrombocytopenic patients with acute leukemia. Blood 62:473, 1983 5. Gmur J: Random single-donor platelet transfusions: Pros and cons. Plasma Ther Transfus Techno1 7:463, 1986 6. Yankee RA, Grumet FC, Rogentine GN: Platelet transfusion therapy: The selection of compatible platelet donors for refractory patients by lymphocyte HL-A typing. N Engl J Med 281:1208, 1969 7. Claas FHJ, Smeenk RJT, Smidt R, van Steenburgge GJ, Eernisse JG: Alloimmunization against the MHC antigens after platelet transfusion is due to contaminating leukocytes in the platelet suspension. Exp Hematol9:84, 1981 8. Deeg HJ, Torok-Storb B, Storb R, Weider PL, DeRose S, Graham TC, Atkinson K, Thomas ED: Rejection of DLA-identical canine littermate marrow after transfusion-induced sensitization, in Baum, Ledney, Khan (eds): Experimental Hematology Today. Basel, Switzerland, Karger, 1981, p 31 9. Brand A, Class FHJ, Falkenburg JHF, van Rood JJ, Eernisse JG: Blood component therapy in bone marrow transplantation. Semin Hematol21:141, 1984 10. Von Willebrand E, HByry P, Soots A, Parthenais E, Nemlander A: Antigenic, immunogenic and immunosensitive components in rat renal allografts. Transplant Proc 13:1099,1981 11. Vakkila J, Myllyla G: Amount and type of leukocytes in

“leukocyte-free” red cell and platelet concentrates. Vox Sang 53:76, 1987 12. Daly PA, Schiffer CA, Aisner J, Wiernick PH: Platelet transfusion therapy. JAMA 243:435, 1980 13. Bishop JF, McGrath K, Wolf MM, Matthews JP, De Luke T. Holdsworth R, Yuen K, Veale M, Whiteside MG, Cooper IA, Szer J: Clinical factors influencing the efficacy of pooled platelet transfusion. Blood 71:383, 1988 14. Van Eys J, Thomas D, Olivos B: Platelet use in pediatric oncology: A review of 393 transfusions. Transfusion 18:169, 1978 15. Howard JE, Perkins HA: The natural history of alloimmunization to platelets. Transfusion 18:496, 1978 16. Dutcher JP, Schiffer CA, Aisner J, Wiernick PH: Alloimmunization following platelet transfusion: The absence of a doseresponse relationship. Blood 57:395, 1981 17. Dutcher JP, Schiffer CA, Aisner J, Wiernick PH: Long-term follow-up of patients with leukemia receiving platelet transfusions: Identification of a large group of patients who do not become alloimmunized. Blood 58:1007, 1981 18. Holohan TV, Terasaki PI, Deisseroth AB: Suppression of transfusion-related alloimmunization in intensely treated cancer patients. Blood 58:122, 1981 19. Fisher M, Chapman JR, Ting A, Morris PJ: Alloimmunization to HLA antigens following transfusion with leukocyte-poor and purified platelet suspensions. Vox Sang 49:33 I , 1985 20. Eernisse JG, Brand A: Prevention of platelet refractoriness due to HLA antibodies by administration of leukocyte-poor blood components. Exp Hematol9:77, 1981 21. Schiffer CA, Dutcher JP, Aisner J, Hogge D, Wiernick PH, Reilly JP: A randomized trial of leukocyte-depleted platelet transfusion to modify alloimmunization in patients with leukemia. Blood 622315, 1983 22. Murphy MF, Metcalfe P, Thomas H, Eve J, Ord J, Lister TA, Waters AH: Use of leukocyte-poor blood components and HLA-

PROPHYLAXIS AGAINST PLATELET REFRACTORINESS

matched platelet donors to prevent alloimmunization. Brit J Haemato1 62529, 1986 23. Sniecinski I, O’Donnell MR, Nowicki B, Hill L R Prevention of refractoriness and HLA-alloimmunization using filtered blood products. Blood 71:1402, 1988 24. Andreu G, Dewailly J, Leberre C, Quarre MC, Bidet ML, Tardive1 R, Devers L, Lam Y, Soreau E, Boccaccio C, Piard N, Bidet JM, Genetet B, Fauchet R: Prevention of HLA immunization with leukocyte-poor packed red cells and platelet concentrates obtained by filtration. Blood 72:964, 1988

25. Sniecinski I, O’Donnell M: White cell depletion of blood products by cottonwool filtration: An economical method for decreasing alloimmunization in multiply transfused patients. Proceedings of

517

the 19th Congress of International Society of Blood Transfusion, Sydney, Australia, 1986 (abstr) 26. Brand A, Claas FHJ, Voogt PJ, WasserMNJM, Eernisse JG: Alloimmunization after leukocyte-depleted multiple random donor platelet transfusions. Vox Sang 54:160, 1988 27. Myllyla G, Ruutu T, Oksanen V, Rasi V, Kekomaki R: Preparation and properties of leukocyte-free platelet concentrates. 19th Cong Int Soc Blood Transfusion, Sydney, Australia, 1986 (abstr) 28. Verdonck LF, de Graan-Hentzen YCE, Dekker AW, Mudde GC, de Gast, GC: Cytomegalovirus seronegative platelets and leukocyte-poor red blood cells from random donors can prevent primary cytomegalovirus infection after bone marrow transplantation. Bone Marrow Transplantation 2:73, 1987