Medical and Pediatric Oncology 6:65-76 (1979)
Granulocyte Transfusions in Infected Neutropenic Children With Malignancies Peter G. Steinherz, MD, and Lilian M. Reich, MD Departmenis of Pediatricsand Medicine, Memorial Sloan-Kettering Cancer Center, New York
Thirty-six febrile neutropenic episodes were treated by granulocyte transfusions in 33 children. Septicemia and mucous membrane ulcerations were most commonly associated with the fever. Infection cleared in 81% of the episodes, eight per cent ended in death from bacterial infections, 11% from nonbacterial infections or hemorrhage. The median number of polymorphonuclear leukocytes given was 1.1 X 10"/m2/ transfusion. Two to twentyeight (median 8.5) transfusions were given over 3-34 days (median 10.5). The source of celk (parental or random) and the method of collection did not seem to affect the outcome. None of the 23 patients whose marrow recovered during the transfusions died of bacterial infections. Infection cleared even without marrow recovery in 62% of the patients, but then only 25% lived for more than two months after clearing of sepsis. In a subgroup of patients with nonlymphoblastic leukemia on the same chemotherapy and antibiotic treatment protocol, 8/11 (73%) survived bacteremia when white cell support was available; only 2/11 (18%) of a historical control group survived when such support was not available. Granulocyte support appears to be a valuable tool in helping neutropenic patients overcome their infections or, at the very least, helping them survive long enough for normal marrow recovery to occur. Key words: childhood malignancy, sepsis, granulocyte transfusions, nonlymphoblastic leukemia, infection INTRODUCTION
With decreasing absolute neutrophil counts there is an increase in the acquisition of severe infections [ l ] . The aggressive chemotherapy regimens that are required for good results in cancer therapy today always cause bone marrow depression, and infection has become the leading cause of death in these patients [2] . A number of investigators have Address reprint requests to Memorial Sloan-KetteringCancer Center, 1275 York Avenue, New York, NY 10021.
0098-1532/79/0601-0065$02.30 0 1979 Alan R. Liss, Inc.
66
Steinherz and Reich
reported improved control of infection in neutropenic patients by means of granulocyte transfusions, [2-lo]. METHODS Patients. Thirty-three neutropenic patients were treated during 36 separate, evaluable febrile episodes. Children with absolute polymorphonuclear leukocyte counts less than SOO/pl and fever above 38°C received broad spectrum antibiotics, usually cephalothin and gentamycin, plus either carbenicillin or another antibiotic, depending on the site of infection or reported microbial sensitivity. The patient’s attending physician had to request the granulocyte support. Children with acute lymphoblastic leukemia during initial remission induction were not considered since these patients rarely die of infection. The patients were given granulocyte transfusions if they were toxic and their fever did not respond to at least 48 hours of antibiotic therapy. Radomization was not performed because we did not feel justified in denying granulocytes to patients. A historical control group will be discussed. AU but one recipient with aplastic anemia were on chemotherapy for various malignancies (Table I). The most frequent underlying disorders were lymphoma with marrow involvement and nonlymphoblastic leukemia. Patients with little or no hope of marrow recovery were not excluded from the study. Eighteen of the courses were in patients with leukemia or lymphoma in relapse. Thirteen of the 18 were after multiple relapses in patients receiving Phase I or Phase I1 agents. The patients were on the pediatric ward in semiprivate rooms with no protective environment or other isolation procedures used.
TABLE 1. Underlying Disordersof Patients Receiving GranulocyteSupport Diagnosis Lymphoma - leukemic phase Nonlymphoblasticleukemia Lymphoblastic leukemia Osteogenic sarcoma Ewing’s sarcoma Neuroblastoma Aplastic anemia Wilms’ tumor
Number entered
Number evaluable 14 (1 1 in relapse) 13 (4 in relapse) 3 (3 in relapse) 2 2 1 1 0
14 13 4 3 2 1 1 1 -
-
39
36
Donors. Parents were used as donors whenever possible. When this was not feasible, random volunteer donors were used (HLA matching was not done). All donors fulfilled the American Association of Blood Banks’ criteria for plasmapheresis [ 11] . Preformed leukoagglutins were not tested. AU transfusions were ABO compatible and no agents to enhance leukocytosis were employed in the donors. Method of collection. In children with surface areas less than 0.7mZ,buffy coats from four units of whole blood from a single donor were combined and transfused (7
Granulocyte Transfusions in Children
67
cases). In other patients, granulocytes were collected by continuous flow centrifugation on an NCI-IBM Model 2990 (International Business Machines, Endicott, NY) cell separator without sedimentation enhancing agents (14 cases) or from Haemonetics Model 30 (Haemonetics Co, Natwick, Massachusetts) cell seprator with hydroxyethyl starch (McGraw Laboratories, Glendale, California) (4 cases). In instances where either donors or the cell separator were not available, 10-20 pooled buffy coats from multiple donors were obtained from the New York Blood Center (1 1 cases). Method of transfusion. The transfusions were given daily whenever possible, with occasional days missed. The granulocytes were transfused over a 2- to 6-hour period and within 24 hours of collection. The recipients were premedicated with diphenhydramine and acetaminophen; meperidine and hydrocortisone were added if the patient had a history of transfusion reactions. The median number of polymorphonuclear leukocytes given was 1.1 X 10'o/m2 /transfusion. The transfusions were terminated when the patients' infections cleared (cultures and visible signs of infection, ie, abscess or infiltrates cleared or the patient defervesced for >48 hours) or the marrow recovered. The median number of transfusions given was 8.5 ranging from 2 to 28. They were given over a 3-34 day period, median 10.5 days. The recovery of the transfused granulocytes did not seem to effect the response rates and these measurements were abandoned early in the study. All granulocyte units were irradiated, 2,000 rads by a cesium source (Gammator M, Isomedix, Parsippany, New Jersey) prior to infusion, as were all blood products infused during the period of granulocyte treatment. This was done to prevent graft versus host reaction without affecting granulocyte function [ 121 . Evaluation of results. When the signs of infection cleared or the patient became afebrile and remained without fever after the discontinuation of the transfusions and the antibiotics he was called a complete responder (Table 11). Patients that died of causes other than bacterial infections (ie, hemorrhage, viral or fungal infections) shown on postmortem examinations were the partial responders. Patients whose bacterial infection did not improve on treatment were the nonresponders.
RESULTS
No significant adverse reactions to donors were noted. One recipient had severe febrile reactions to familial granulocytes and the transfusions were terminated. No other severe reactions occurred due to familial or random donor white cells, although chills and mild febrile reactions were observed. One patient had liver enzyme abnormalities after receiving 263 units of pooled buffy coats. She had received numerous packed cells and platelet transfusions, as well as chemotherapy, antibiotics and abdominal radiation, and remained hepatitis antigen negative. One patient, who died of hemorrhage, showed inclusion bodies in the alveolar cells on postmortem examination, suggesting cytomegalovirus infection. A number of patients required phlebotomy for polycythemia after repeated granulocyte transfusions, due to the number of erythrocytes transfused with the granulocyte concentrates. Thirty-six of 39 episodes were evaluable, two patients died less than 24 hours after entering the study and one patient's treatment was discontinued, as previously mentioned, due to her febrile reactions. Table I1 shows data describing the patient population, sites of infection, source and duration of granulocyte support, and the response to treatment.
68
Steinherz and Reich
TABLE 11. Patients Receiving White Blood Cell Transfusions WBC transfusion No. Duration (times) (days)
Age
Diagnosis
Source
14
AML relapse AML relapse
CFC random CFC random
10
12
Mucous m ulcers
CR
No
12
21
CR
No
AML induction AML induction AML induction
CFC random CFC random CFC random
19
24
CR
Yes
3
3
BC Staph epi CSF Staph epi mucous m ulcers Rectal abscess mucous m ulcers FUO
CR
Yes
6
9
CR
Yes
I
AML induction
B random
6
6
CR
Yes
4
AML relapse
CFC parents
10
11
CR
Yes
3
AML induction
B parents
12
18
NR
No
12
APML relapse
CFC random
28
28
CR
Yes
3
APML induction
B random
11
12
CR
Yes
3
APML induction
CFC random
14
11
NR
No
2
APML induction APML induction ALL relapse ALL relapse ALL relapse Lymphoma relapse
B random B random B pooled B pooled B pooled
8
14
CR
Yes
11
12
CR
Yes
7
I
CR
No
I
I
Rectal abscess BC Pseudo; a mucous m ulcers Mucous m ulcers cellulitis BC prop acnes Pneumonia BC E coli, Diphteroids mucous m ulcers BC staph epi klebsiella peritonitis necrotic bowel mucous m ulcers Rectal abscess BC staph epi, serratia CSF staph epi mucous m ulcers BC strep virid, prop acnes mucous m ulcers BC Clostridia pseudo a, E coli necrotic bowel Abscess mucous m ulcers BC staph aur prop acnes Cellulitis BC prop acnes FUO
CR
Yes
6
I
FUO
CR
No
3
3
CR
Yes
I
I
Mucous m ulcers cellulitis appendicitis BC pseudomonas
CR
Yes
I
12
Mucous m ulcers
CR
No
14
13 13 12
2 18
11 6 15
15 15
Lymphoma relapse Lymphomaa induction
HES parents
B pooled B pooled
Infection
Results ofthe Marrow infection recovery
G r a n u l o c y t e Transfusions in Children
69
Table 11 (continued) WBC transfusion No, Duration (times) (days)
Age
Diagnosis
Source
15
Lymphomaa induction
B pooled
19
29
15
Lymphoma relapse Lymphoma relapse
HES parents B pooled
5
5
13
14
10
Lymphoma relapse
HES parents
12
12
10
Lymphoma relapse Lymphoma relapse Lymphoma relapse Lymphoma relapse Lymphoma relapse Lymphoma relapse
CFC parents HES random CFC random B pooled B parents CFC parents
3
4
8
8
4
5
9
10
2
6
21
34
Lymphoma induction Ewing sarcoma Ewing sarcomaa
HES random B pooled CFC random
9
11
9
9
12
12
9
11
4
I
6
I
I
10
15
8 8
8 5 5
2 21 14
14 14
7 9
B Osteogenic pooled sarcomaa B Osteogenic pooled sarcomaa Neuroblastoma CFC parents Aplastic CFC anemia random
AML, Acute myelocytic leukemia. APML, Acute prolyelocytic leukemia. ALL, Acute lymphoblastic leukemia. B, Buffy coat. BC, Blood culture. CFC, Continuous flow centrifugation CSF, Cerebrospinal fluid.
Infection
Results o f t h e Marrow infection recovery
Cellulitis BC staph epi mucous m ulcers FUO
CR
Yes
CR
Yes
Rectal abscess pneumonia CSF acinetobacter Iwoffi Rectal abscess pneumonia BC candida BC klebsiella
CR
No
CR
No
CR
Yes
Mucous m ulcers pneumonia, viral BC prop acnes
PR**
No
CR
No
BC staph epi
NR
No
BC micrococcus mucous m ulcers Rectal abscess BC E coli mucous m ulcers Abscess pneumonia Mucous m ulcers
CR
Yes
CR
Yes
CR
Yes
CR
Yes
Mucous m ulcers rectal abscess BC strep faec, klebsiella, Bacteroides f enterococcus Mucous m ulcers BC E coli Mucous m ulcers
CR
Yes
PRb
Yes
CR
Yes
Pneumonia, pneumocystis carinii Pneumonia BC staph Bur, candida
PRC
Yes
PRC
No
CR, PR, NR, Complete, partial and n o response. FUO, Fever of unknown origin. HES, Hydroxyethyl starch. aNo marrow involvement by disease. bDied of hemorrhage. 'Died of nonbacterial infections.
70
Steinherz and Reich
The infections cleared completely in 29/36 (81%) of the episodes. Three patients (8%) died from bacterial infections. Two of these patients had necrotic bowel on postmortem examination. Four patients (1 1%) died without evidence of bacterial infection; one had generalized candidiasis after clearing of staphylococcal septicemia, one died of hemorrhage after clearing of E coli septicemia, one succumbed to respiratory failure with no evidence of bacterial infection after clearing of Pneumocystis carinii pneumonia, one child had mucositis and died of pulmonary hemorrhage with viral-like inclusions seen in the lungs at postmortem examination. None of the 23 patients (Table 111) whose marrows recovered during granulocyte transfusions died of bacterial infection. One had granulocyte recovery but remained thrombocytopenic and died of hemorrhage, and one had bone marrow recovery but died of respiratory failure as mentioned above. Twenty of the remaining 21 (95%)with marrow recovery lived more than two months after treatment. One died of disease after another relapse. Three of the 13 (23%) with no marrow recovery died of bacterial infection, one died of candidiasis and one of hemorrhage. The other eight (62%), although recovering from the infectious episodes, remained neutropenic and were soon reinfected. They had a progressively downhill course unless treated again with granulocytes. Three were retreated and one, a patient with lymphoma in her first remission induction, recovered marrow function after 29 days of granulocyte transfusions. Only two (25%)of the eight living with no marrow recovery at the end of the initial treatment survived. The others died 33-61 days after the end of granulocyte support. Severe mucous membrane ulcerations and positive blood cultures were the most commonly associated findings with the febrile episodes (Table IV). The ulcers were accompanied by positive cultures in 13 out of 21 instances. The three deaths that were due to bacterial infections were caused by septicemia. Two had severe mucous membrane ulcerations with necrotic bowel and multiple organisms cultured from the blood. The source of septicemia in the third could not be determined. All patients with more localized bacterial infections, ie, abscesses, cellulities, pneumonia and meningitis, survived. The most common localized infections were rectal abscesses. There were seven cases with rectal abscesses, five of them with positive blood cultures. The organisms cultured from the blood are listed in Table 11.
TABLE 111. Results of Granulocyte Transfusions No marrow recoverv Number of episodes evaluable Infections cleared Death from bacterial infections Death from other causes Survived the episode Survived clearing of infection by two months
Marrow recovery
Total
23 23/23 (100%)
36 33/36 (92%)
3/13 (23%) 2/13 (15%) 8/13 (62%)
0/23 (0%) 2/23 (9%) 21/23 (91%)
3/36 (8%) 4/36 (1 1%) 29/36 (81%)
2/8 (25%)
20/21 (95%)
22/29 (76%)
13 10/13 (77%)
Granulocyte Transfusions in Children
71
There was no difference in response rate whether the granulocytes were obtained from parents (10/11,90%) or from random donors (23/25, 92%). There was one failure in each group given granulocytes obtained by continous flow centrifugation (1/14,7%) single donor buffy coats (1/7, 14%) and pooled multiple donor buffy coats (111 1,9%). There were no failures in the four cases given cells obtained by discontinous flow centrifugation with hydroxyethyl starch. There was no difference in response rate whether or not the circulating granulocyte count increased after transfusion. DISCUSSION
The success of granulocyte transfusions in helping neutropenic patients overcome their infections as reported by many investigators has made it difficult t o justify a randomized trial in which some patients are denied granulocyte transfusions. Without randomized concurrent controls, it is hard to assess the role granulocytes play in the patient’s recovery. The patients entered in this study were frequently in poor condition. They were not considered for treatment unless requested by their private physician and this usually occurred only after prolonged antibiotic treatment, or when the physician felt that the patient was not expected to survive on antibiotics alone. There were only five patients with no marrow involvement by their malignancies. Ninety-two per cent of the bacterial infections cleared. Some would probably have recovered without granulocytes; however, in patients with prolonged, severe infections, the granulocytes had a role in containing the infection. Seventy-seven per cent of the patients without normal marrow recovery cleared their infections, a situation which could not be expected with antibiotics alone (Table 111). Some of the patients acted as their own controls, eg, one patient was treated for ten days and improved without marrow recovery only to develop meningitis and septicemia with Staphylococcus epidermidis two weeks after treatment. He was treated again and the infection cleared. Three weeks later he became septic, still in relapse. The family and physicians decided not to treat him a
TABLE IV. Response to Granulocyte Transfusion by Source of Fever Death due to bacterial infection Marrow recovery aNumber No Yes Mucous membrane ulcerations Septicemia Abscesses or cellulitis Fever of unknown origin Pneumonia, bacterial Meningitis Pneumonia, nonbacterial (candida, pneumocystis, CMV)
21 21
216 (33%)
13
013
4 4 3 3
011 012 012 012
318 (38%)
aMany patients had more than one demonstrable source of fever.
0115 0113 0110 013 012 o/ 1 011
Total 2/21 (10%) 312 1 (14%) 0113 014 014
013 013
72
Steinherz and Reich
third time since he had been in continuous relapse for months and had not responded to experimental agents; he died in ten days. The group was heterogenous in diagnosis, stage of disease, treatment and site of infection. Because concurrent randomized controls were not possible, we used a group of historical controls to evaluate the role of granulocyte transfusions in the improvement of those patients, while clearly recognizing the limitations of historical controls. Due to these limitations, only gross differences between groups were considered and no statistical analysis was performed. Bacteremia in nonlymphoblastic leukemia patients provided a fairly uniform group. From 1975-1977, when granulocytes were available, 14 new patients were entered on study; from 1971-1975, when white cell support was not available, 15 new patients were seen (Table V). Both groups had been treated with the same chemotherapy protocol and similar antibiotic regimens during episodes of fever and neutropenia. We studied the course of these 29 patients through remission-induction, maintenance and eventual relapse. Sepsis during neutropenia occurred more frequently during induction in the group in which white blood cell transfusions were available. During relapse, infections were more frequent in the other group. There were 11 episodes of bacteremia in each group (Table V). The microorganisms cultured from the blood in both groups were similar (Table VI). These included the usual gram-positive and gramnegative pathogens plus some organisms which are not usually pathogenic but may cause disease in neutropenic patients. An occasional positive culture in each group may have been caused by contamination. Six of the seven patients who developed bacteremia during induction were given granulocyte trasnfusions in the study group (Fig. 1) . One patient recovered without white cell support. Five of them survived. Only one of four patients in the control group survived (Table V) and the significance of the patient’s bacter-
TABLE V. Results of Sepsis in Nonlymphobhtic Leukemia 1971-1975
1975-1977
WBC not available
WBC available
Number of consecutive patients
15
14
Positive blood cultures induction relapse
11 4 7
11 7 4
Number given WBC transfusions
0
9
Death with positive blood culture induction relapse
8/11 (73%) 314 (75%) 517 (71%)
2/11 217 (29%) 0/4 (0%)
Time of death after positive blood cultures
4-26 (median 11.5) days
17 & 24 days
Time defervescence
1, 2,9, days
1-32 (median 15) days
Clearing of positive culture
2/10 (20%)
113 (33%)
without M I marrow aBoth deaths due to compromised bowel.
Granulocyte Transfusions in Children
73
emia is questionable; the patient became afebrile one day after Diphteroids were identified in the blood. Five of the seven patients in the control group whose blood cultures were positive in relapse died, while all four during the study period survived (Fig. 2). Three of them received granulocyte support. Figures 1 and 2 show the clinical course of these patients. Of the study patients, nine out of eleven (82%)survived, while only three out of 1 1 (27%)of the control group survived. Both deaths in the study group were due to sepsis secondary to necrotic bowel, a condition that granulocytes could not possibly correct. The 82%survival of sepsis in nonlymphoblastic leukemia is similar to the 86%achieved with bacteremia in the entire study (Table IV). Normal marrow recovery significantly influenced the clearing of positive blood cultures. Only 33% of the patients who received granulocytes survived without normal marrow function, compared to 20%in the untreated nonlymphoblastic leukemia group. All patients with marrow recovery survived. In the 21 patients (Table IV) receiving granulocytes, the immediate survival of a septic episode without marrow recovery was 62%,but this was a transient response. They were soon reinfected and only 25% lived more than two months (Table 111). This seems to indicate that severely infected, neutropenic patients do benefit from granulocyte transfusions when antibiotics alone are ineffective. The effect is clincally significant only if normal hematopoiesis returns. Granulocyte transfusions probably allow the patient to survive longer and thereby permit more time for marrow recovery. The median survival of patients without white cell support was 11.5 days. The median time of clinical improvement in the granulocyte transfused group was 15 days (Table V), and the median time to return of normal marrow function was 18 days. White cell support, in most cases, seemed to prolong life for one week to permit marrow recovery. The results obtained in this study are much better than those reported by Sanders, et a1 [13] and are similar to those of Highby, et al [8] both of whom used filter-collected cells. Those cells, although larger in number, may cause more reactions and have been found by some investigators to be functionally impaired [7, 14-16]. The other studies in the literature are mainly in adults [2-lo].
TABLE VI. Microorganisms Cultured from Blood in Nonlymphoblastic Leukemia
Organisms Escherichia coli Pseudomonas Serratia marcescens Klebsiella Clostridium Streptococcus viridans Diplococcus pneumoniae Staphylococcus aureus Staphylococcus epidermidis Propionibacterium acnes Diphteroids
Number of patients with positive cultures WBC not available [ 11 ] WBC available [ 1 1 ]
2 0 1 1 0
0 0 2
2 2 1 1 1 2 1 1
1
4
1
4
4
1
14
Steinherz and Reich ~~~
M Fever
Blood culture positive WBC transfusion
0
Deoth
BM+ Bone morrow recovery BM- No recovery
WBC ovoiloble Prop. oc.
Slaph.47,:
& BM +
&c
Strw.y:
+p*m
Prop. oc.
o-o BM+ . *WW r. *w -BMt
-
Strep. y. Prcp oc.
Stoph. ou Prop oc.
0
Pseudo.
0
**
* *
*ww+w*
*!3
BM-
0
.
BM+
+b++*3
**
o BM+
339 # 33 M W
Clostridio Pseudo.
e
E co/i I
.*
H3 ##3 *#
Sbph.epi Klebs.
-30
3
I
-20
I
I
-10
4
BM-
I
I
I
I
10
20
30
40
Days before or after 1st positive blood culture
.
WBC not avoiloble Diphtheroids
~4 BM+
LBM-
Stoph. epl:
W...
MBM-
E COli Klebs. I
-30
BMI
1
1
I
I
I
-20 -10 10 20 30 Days before or after 1st positive blood culture
J
40
Fig. 1 . Course of sepsis in patients with nonlymphoblasticleukemia during first remission induction.
White cell transfusions have minimal immediate side effects and the source of the cells (parental or random donors) and the method of collection does not seem to be important as long as enough cells are given. Prolonging the time through which a myelosup pressed patient can be supported is becoming more and more important. Granulocyte support permits the use of more aggressive multiagent chemotherapy protocols which are required for good results in certain pediatric malignancies such as nonlymphoblastic leukemia, lymphoma, osteogenic sarcoma and Ewing’s sarcoma - and it is essential for the proper evaluation of new therapeutic agents in relapsed patients. We feel that, if there is some chance of marrow recovery, all neutropenic, infected children should be
Granulocyte Transfusions in Children
75
MFever
Blood culture positive WBC transfusion 0 Death 0M+ Bone marrow recovery BM- No recovery
C
WBC ovailoble
& BM+
Staph. ou.
I
Y
-51 -/30
I
I
I
I
I
-20
-10 10 20 30 Days before or after 1st positive blood culture
1
40
WBC not ovailoble
-
Dip/.pn.
00 0M-
*
Dibhf.
Ecoli
*
BM-
BM-
oipn,.
BM-
0
-
Prop. oc.
*
0
Dipht
**
Serrotio
sropn. OU. I
-30
I
H
BM-
0M-
0
0 I
I
I
10 20 Days before or after 1st positive cultures
-20
10-
1
0M1
1
30
40
Fig. 2. Course of sepsis in patients with nonlymphoblastic leukemia in relapse.
given granulocyte transfusions when 48 hours of antibiotics have failed. Early institution of white cell support may prevent irreversible changes and death during induction therapy. The white cells should be given daily, 1 X 101O/mZ,from any ABO compatible donor until infection clears or normal granulocytopoiesis returns. Repeated courses of granulocyte transfusions without marrow recovery will not appreciably affect long-term survival. The question of prophylactic granulocyte transfusions prior to the occurrence of overt infections in granulocytopenic patients remains to be answered.
76
Steinherz and Reich
REFERENCES 1. Bodey GP, Buckley M, Sathe YS,Freireich EJ: Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intl Med 64:328-336, 1966. 2. Graw RG, Herzig G, Perry S, Henderson ES: Normal granulocyte transfusion therapy-Treatment of septicemia due to gram negative bacteria, N Engl J Med 287:367-371, 1972. 3. Alavi JB, Root RK, Djerassi I, Evans AE, Gluckman SJ, MacGregor RR, Guerry D, Schreiber AD, Shaw JM, Koch P, Cooper RA: A randomized clinical trial of granulocyte transfusions in infection in acute leukemia N Engl J Med 296:706-711, 1977. 4. Bogs DR: Transfusion of neutrophils as prevention or treatment of infection in patients with neutropenia. N Engl J Med 290: 1055-1062,1974. 5. Herzig RH, Herzig GP, Graw RC Jr, Bull MI, Ray KK: Successful granulocyte transfusion therapy from gram negative septicemia, N Engl J Med 296:701-705, 1977. 6. Hester JP, McCredie KB, Freireich EJ: Granulocyte transfusion: analysis of donor, procedure, recipient variables. Blut 32:253-256, 1976. 7. Highby DJ, Henderson ES: Granulocyte transfusion therapy. Ann Rev Med 26:289-306, 1975. 8. Highby DJ, Freeman A, Henderson ES, Sinks L, Cohen E: Granulocyte transfusions in children using filtercollected cells. Cancer 38:1407-1413, 1976. 9. MacPherson JL, Nusbacher J, Bennett JM: The acquisition of granulocytes by leukapheresis. Transfusion 16:221-228, 1976. 10. Reich LM, VonSickle P, Wright P: Granulocyte transfusion: A preliminary study. In “Leukocytes: Sepration, collection and transfusion. Proceedings of the International Symposium on Leukocyte Separation and Transfusion.” London, September 1974, pp 294-299. 11. American Association of Blood Banks Technical Methods and Procedures: Miller V (ed). Philadelphia: JB Lippincott, 1974. 12. Graw RG, Bucker CD, Whang-Peng J, Leventhal BG, Kruger G, Berard C, Henderson ES: Complications of bone marrow transplantation. Graft-versus-host disease resulting from chronic myelogenous leukaemia leukocyte transfusions. Lancet 2: 338-340, 1970. 13. Sanders J, Wilbur JR, Long T, Etcubanas E, Babin S,Grant J: Granulocyte transfusions in the treatment of infection in children with malignancy. In “Proceedings of the Eleventh Annual Meeting of the American Society of Clinical Oncology.” San Diego, May 1975, p 269. 14. Ts’ao C, Ruder EA: Ultrastructural damage of leukocytes procured by the leukopak: Vulnerability of leukocytes to mechanical injury. Transfusion 16:336-344, 1976. 15. Klok JC, Bainton DF: Degranulation and abnormal bacterial function of granulocytes procured by reversible adhesion to nylon wool. Blood 48:149-161, 1976. 16. Wade PH, Skrabut EM, Vinciguerra L, Valeri CR: In vitro function of granulocytes isolated from blood of normal volunteers using continuous-flow centrifugation in the IBM-Aminco celltrifuge and adhesion-fiitration leukapheresis using nylon fiber. Transfusion 17:136-140, 1977.
Granulocyte Transfusions in Infected Neutropenic Children With Malignancies Peter G. Steinherz, MD, and Lilian M. Reich, MD Departmenis of Pediatricsand Medicine, Memorial Sloan-Kettering Cancer Center, New York
Thirty-six febrile neutropenic episodes were treated by granulocyte transfusions in 33 children. Septicemia and mucous membrane ulcerations were most commonly associated with the fever. Infection cleared in 81% of the episodes, eight per cent ended in death from bacterial infections, 11% from nonbacterial infections or hemorrhage. The median number of polymorphonuclear leukocytes given was 1.1 X 10"/m2/ transfusion. Two to twentyeight (median 8.5) transfusions were given over 3-34 days (median 10.5). The source of celk (parental or random) and the method of collection did not seem to affect the outcome. None of the 23 patients whose marrow recovered during the transfusions died of bacterial infections. Infection cleared even without marrow recovery in 62% of the patients, but then only 25% lived for more than two months after clearing of sepsis. In a subgroup of patients with nonlymphoblastic leukemia on the same chemotherapy and antibiotic treatment protocol, 8/11 (73%) survived bacteremia when white cell support was available; only 2/11 (18%) of a historical control group survived when such support was not available. Granulocyte support appears to be a valuable tool in helping neutropenic patients overcome their infections or, at the very least, helping them survive long enough for normal marrow recovery to occur. Key words: childhood malignancy, sepsis, granulocyte transfusions, nonlymphoblastic leukemia, infection INTRODUCTION
With decreasing absolute neutrophil counts there is an increase in the acquisition of severe infections [ l ] . The aggressive chemotherapy regimens that are required for good results in cancer therapy today always cause bone marrow depression, and infection has become the leading cause of death in these patients [2] . A number of investigators have Address reprint requests to Memorial Sloan-KetteringCancer Center, 1275 York Avenue, New York, NY 10021.
0098-1532/79/0601-0065$02.30 0 1979 Alan R. Liss, Inc.
66
Steinherz and Reich
reported improved control of infection in neutropenic patients by means of granulocyte transfusions, [2-lo]. METHODS Patients. Thirty-three neutropenic patients were treated during 36 separate, evaluable febrile episodes. Children with absolute polymorphonuclear leukocyte counts less than SOO/pl and fever above 38°C received broad spectrum antibiotics, usually cephalothin and gentamycin, plus either carbenicillin or another antibiotic, depending on the site of infection or reported microbial sensitivity. The patient’s attending physician had to request the granulocyte support. Children with acute lymphoblastic leukemia during initial remission induction were not considered since these patients rarely die of infection. The patients were given granulocyte transfusions if they were toxic and their fever did not respond to at least 48 hours of antibiotic therapy. Radomization was not performed because we did not feel justified in denying granulocytes to patients. A historical control group will be discussed. AU but one recipient with aplastic anemia were on chemotherapy for various malignancies (Table I). The most frequent underlying disorders were lymphoma with marrow involvement and nonlymphoblastic leukemia. Patients with little or no hope of marrow recovery were not excluded from the study. Eighteen of the courses were in patients with leukemia or lymphoma in relapse. Thirteen of the 18 were after multiple relapses in patients receiving Phase I or Phase I1 agents. The patients were on the pediatric ward in semiprivate rooms with no protective environment or other isolation procedures used.
TABLE 1. Underlying Disordersof Patients Receiving GranulocyteSupport Diagnosis Lymphoma - leukemic phase Nonlymphoblasticleukemia Lymphoblastic leukemia Osteogenic sarcoma Ewing’s sarcoma Neuroblastoma Aplastic anemia Wilms’ tumor
Number entered
Number evaluable 14 (1 1 in relapse) 13 (4 in relapse) 3 (3 in relapse) 2 2 1 1 0
14 13 4 3 2 1 1 1 -
-
39
36
Donors. Parents were used as donors whenever possible. When this was not feasible, random volunteer donors were used (HLA matching was not done). All donors fulfilled the American Association of Blood Banks’ criteria for plasmapheresis [ 11] . Preformed leukoagglutins were not tested. AU transfusions were ABO compatible and no agents to enhance leukocytosis were employed in the donors. Method of collection. In children with surface areas less than 0.7mZ,buffy coats from four units of whole blood from a single donor were combined and transfused (7
Granulocyte Transfusions in Children
67
cases). In other patients, granulocytes were collected by continuous flow centrifugation on an NCI-IBM Model 2990 (International Business Machines, Endicott, NY) cell separator without sedimentation enhancing agents (14 cases) or from Haemonetics Model 30 (Haemonetics Co, Natwick, Massachusetts) cell seprator with hydroxyethyl starch (McGraw Laboratories, Glendale, California) (4 cases). In instances where either donors or the cell separator were not available, 10-20 pooled buffy coats from multiple donors were obtained from the New York Blood Center (1 1 cases). Method of transfusion. The transfusions were given daily whenever possible, with occasional days missed. The granulocytes were transfused over a 2- to 6-hour period and within 24 hours of collection. The recipients were premedicated with diphenhydramine and acetaminophen; meperidine and hydrocortisone were added if the patient had a history of transfusion reactions. The median number of polymorphonuclear leukocytes given was 1.1 X 10'o/m2 /transfusion. The transfusions were terminated when the patients' infections cleared (cultures and visible signs of infection, ie, abscess or infiltrates cleared or the patient defervesced for >48 hours) or the marrow recovered. The median number of transfusions given was 8.5 ranging from 2 to 28. They were given over a 3-34 day period, median 10.5 days. The recovery of the transfused granulocytes did not seem to effect the response rates and these measurements were abandoned early in the study. All granulocyte units were irradiated, 2,000 rads by a cesium source (Gammator M, Isomedix, Parsippany, New Jersey) prior to infusion, as were all blood products infused during the period of granulocyte treatment. This was done to prevent graft versus host reaction without affecting granulocyte function [ 121 . Evaluation of results. When the signs of infection cleared or the patient became afebrile and remained without fever after the discontinuation of the transfusions and the antibiotics he was called a complete responder (Table 11). Patients that died of causes other than bacterial infections (ie, hemorrhage, viral or fungal infections) shown on postmortem examinations were the partial responders. Patients whose bacterial infection did not improve on treatment were the nonresponders.
RESULTS
No significant adverse reactions to donors were noted. One recipient had severe febrile reactions to familial granulocytes and the transfusions were terminated. No other severe reactions occurred due to familial or random donor white cells, although chills and mild febrile reactions were observed. One patient had liver enzyme abnormalities after receiving 263 units of pooled buffy coats. She had received numerous packed cells and platelet transfusions, as well as chemotherapy, antibiotics and abdominal radiation, and remained hepatitis antigen negative. One patient, who died of hemorrhage, showed inclusion bodies in the alveolar cells on postmortem examination, suggesting cytomegalovirus infection. A number of patients required phlebotomy for polycythemia after repeated granulocyte transfusions, due to the number of erythrocytes transfused with the granulocyte concentrates. Thirty-six of 39 episodes were evaluable, two patients died less than 24 hours after entering the study and one patient's treatment was discontinued, as previously mentioned, due to her febrile reactions. Table I1 shows data describing the patient population, sites of infection, source and duration of granulocyte support, and the response to treatment.
68
Steinherz and Reich
TABLE 11. Patients Receiving White Blood Cell Transfusions WBC transfusion No. Duration (times) (days)
Age
Diagnosis
Source
14
AML relapse AML relapse
CFC random CFC random
10
12
Mucous m ulcers
CR
No
12
21
CR
No
AML induction AML induction AML induction
CFC random CFC random CFC random
19
24
CR
Yes
3
3
BC Staph epi CSF Staph epi mucous m ulcers Rectal abscess mucous m ulcers FUO
CR
Yes
6
9
CR
Yes
I
AML induction
B random
6
6
CR
Yes
4
AML relapse
CFC parents
10
11
CR
Yes
3
AML induction
B parents
12
18
NR
No
12
APML relapse
CFC random
28
28
CR
Yes
3
APML induction
B random
11
12
CR
Yes
3
APML induction
CFC random
14
11
NR
No
2
APML induction APML induction ALL relapse ALL relapse ALL relapse Lymphoma relapse
B random B random B pooled B pooled B pooled
8
14
CR
Yes
11
12
CR
Yes
7
I
CR
No
I
I
Rectal abscess BC Pseudo; a mucous m ulcers Mucous m ulcers cellulitis BC prop acnes Pneumonia BC E coli, Diphteroids mucous m ulcers BC staph epi klebsiella peritonitis necrotic bowel mucous m ulcers Rectal abscess BC staph epi, serratia CSF staph epi mucous m ulcers BC strep virid, prop acnes mucous m ulcers BC Clostridia pseudo a, E coli necrotic bowel Abscess mucous m ulcers BC staph aur prop acnes Cellulitis BC prop acnes FUO
CR
Yes
6
I
FUO
CR
No
3
3
CR
Yes
I
I
Mucous m ulcers cellulitis appendicitis BC pseudomonas
CR
Yes
I
12
Mucous m ulcers
CR
No
14
13 13 12
2 18
11 6 15
15 15
Lymphoma relapse Lymphomaa induction
HES parents
B pooled B pooled
Infection
Results ofthe Marrow infection recovery
G r a n u l o c y t e Transfusions in Children
69
Table 11 (continued) WBC transfusion No, Duration (times) (days)
Age
Diagnosis
Source
15
Lymphomaa induction
B pooled
19
29
15
Lymphoma relapse Lymphoma relapse
HES parents B pooled
5
5
13
14
10
Lymphoma relapse
HES parents
12
12
10
Lymphoma relapse Lymphoma relapse Lymphoma relapse Lymphoma relapse Lymphoma relapse Lymphoma relapse
CFC parents HES random CFC random B pooled B parents CFC parents
3
4
8
8
4
5
9
10
2
6
21
34
Lymphoma induction Ewing sarcoma Ewing sarcomaa
HES random B pooled CFC random
9
11
9
9
12
12
9
11
4
I
6
I
I
10
15
8 8
8 5 5
2 21 14
14 14
7 9
B Osteogenic pooled sarcomaa B Osteogenic pooled sarcomaa Neuroblastoma CFC parents Aplastic CFC anemia random
AML, Acute myelocytic leukemia. APML, Acute prolyelocytic leukemia. ALL, Acute lymphoblastic leukemia. B, Buffy coat. BC, Blood culture. CFC, Continuous flow centrifugation CSF, Cerebrospinal fluid.
Infection
Results o f t h e Marrow infection recovery
Cellulitis BC staph epi mucous m ulcers FUO
CR
Yes
CR
Yes
Rectal abscess pneumonia CSF acinetobacter Iwoffi Rectal abscess pneumonia BC candida BC klebsiella
CR
No
CR
No
CR
Yes
Mucous m ulcers pneumonia, viral BC prop acnes
PR**
No
CR
No
BC staph epi
NR
No
BC micrococcus mucous m ulcers Rectal abscess BC E coli mucous m ulcers Abscess pneumonia Mucous m ulcers
CR
Yes
CR
Yes
CR
Yes
CR
Yes
Mucous m ulcers rectal abscess BC strep faec, klebsiella, Bacteroides f enterococcus Mucous m ulcers BC E coli Mucous m ulcers
CR
Yes
PRb
Yes
CR
Yes
Pneumonia, pneumocystis carinii Pneumonia BC staph Bur, candida
PRC
Yes
PRC
No
CR, PR, NR, Complete, partial and n o response. FUO, Fever of unknown origin. HES, Hydroxyethyl starch. aNo marrow involvement by disease. bDied of hemorrhage. 'Died of nonbacterial infections.
70
Steinherz and Reich
The infections cleared completely in 29/36 (81%) of the episodes. Three patients (8%) died from bacterial infections. Two of these patients had necrotic bowel on postmortem examination. Four patients (1 1%) died without evidence of bacterial infection; one had generalized candidiasis after clearing of staphylococcal septicemia, one died of hemorrhage after clearing of E coli septicemia, one succumbed to respiratory failure with no evidence of bacterial infection after clearing of Pneumocystis carinii pneumonia, one child had mucositis and died of pulmonary hemorrhage with viral-like inclusions seen in the lungs at postmortem examination. None of the 23 patients (Table 111) whose marrows recovered during granulocyte transfusions died of bacterial infection. One had granulocyte recovery but remained thrombocytopenic and died of hemorrhage, and one had bone marrow recovery but died of respiratory failure as mentioned above. Twenty of the remaining 21 (95%)with marrow recovery lived more than two months after treatment. One died of disease after another relapse. Three of the 13 (23%) with no marrow recovery died of bacterial infection, one died of candidiasis and one of hemorrhage. The other eight (62%), although recovering from the infectious episodes, remained neutropenic and were soon reinfected. They had a progressively downhill course unless treated again with granulocytes. Three were retreated and one, a patient with lymphoma in her first remission induction, recovered marrow function after 29 days of granulocyte transfusions. Only two (25%)of the eight living with no marrow recovery at the end of the initial treatment survived. The others died 33-61 days after the end of granulocyte support. Severe mucous membrane ulcerations and positive blood cultures were the most commonly associated findings with the febrile episodes (Table IV). The ulcers were accompanied by positive cultures in 13 out of 21 instances. The three deaths that were due to bacterial infections were caused by septicemia. Two had severe mucous membrane ulcerations with necrotic bowel and multiple organisms cultured from the blood. The source of septicemia in the third could not be determined. All patients with more localized bacterial infections, ie, abscesses, cellulities, pneumonia and meningitis, survived. The most common localized infections were rectal abscesses. There were seven cases with rectal abscesses, five of them with positive blood cultures. The organisms cultured from the blood are listed in Table 11.
TABLE 111. Results of Granulocyte Transfusions No marrow recoverv Number of episodes evaluable Infections cleared Death from bacterial infections Death from other causes Survived the episode Survived clearing of infection by two months
Marrow recovery
Total
23 23/23 (100%)
36 33/36 (92%)
3/13 (23%) 2/13 (15%) 8/13 (62%)
0/23 (0%) 2/23 (9%) 21/23 (91%)
3/36 (8%) 4/36 (1 1%) 29/36 (81%)
2/8 (25%)
20/21 (95%)
22/29 (76%)
13 10/13 (77%)
Granulocyte Transfusions in Children
71
There was no difference in response rate whether the granulocytes were obtained from parents (10/11,90%) or from random donors (23/25, 92%). There was one failure in each group given granulocytes obtained by continous flow centrifugation (1/14,7%) single donor buffy coats (1/7, 14%) and pooled multiple donor buffy coats (111 1,9%). There were no failures in the four cases given cells obtained by discontinous flow centrifugation with hydroxyethyl starch. There was no difference in response rate whether or not the circulating granulocyte count increased after transfusion. DISCUSSION
The success of granulocyte transfusions in helping neutropenic patients overcome their infections as reported by many investigators has made it difficult t o justify a randomized trial in which some patients are denied granulocyte transfusions. Without randomized concurrent controls, it is hard to assess the role granulocytes play in the patient’s recovery. The patients entered in this study were frequently in poor condition. They were not considered for treatment unless requested by their private physician and this usually occurred only after prolonged antibiotic treatment, or when the physician felt that the patient was not expected to survive on antibiotics alone. There were only five patients with no marrow involvement by their malignancies. Ninety-two per cent of the bacterial infections cleared. Some would probably have recovered without granulocytes; however, in patients with prolonged, severe infections, the granulocytes had a role in containing the infection. Seventy-seven per cent of the patients without normal marrow recovery cleared their infections, a situation which could not be expected with antibiotics alone (Table 111). Some of the patients acted as their own controls, eg, one patient was treated for ten days and improved without marrow recovery only to develop meningitis and septicemia with Staphylococcus epidermidis two weeks after treatment. He was treated again and the infection cleared. Three weeks later he became septic, still in relapse. The family and physicians decided not to treat him a
TABLE IV. Response to Granulocyte Transfusion by Source of Fever Death due to bacterial infection Marrow recovery aNumber No Yes Mucous membrane ulcerations Septicemia Abscesses or cellulitis Fever of unknown origin Pneumonia, bacterial Meningitis Pneumonia, nonbacterial (candida, pneumocystis, CMV)
21 21
216 (33%)
13
013
4 4 3 3
011 012 012 012
318 (38%)
aMany patients had more than one demonstrable source of fever.
0115 0113 0110 013 012 o/ 1 011
Total 2/21 (10%) 312 1 (14%) 0113 014 014
013 013
72
Steinherz and Reich
third time since he had been in continuous relapse for months and had not responded to experimental agents; he died in ten days. The group was heterogenous in diagnosis, stage of disease, treatment and site of infection. Because concurrent randomized controls were not possible, we used a group of historical controls to evaluate the role of granulocyte transfusions in the improvement of those patients, while clearly recognizing the limitations of historical controls. Due to these limitations, only gross differences between groups were considered and no statistical analysis was performed. Bacteremia in nonlymphoblastic leukemia patients provided a fairly uniform group. From 1975-1977, when granulocytes were available, 14 new patients were entered on study; from 1971-1975, when white cell support was not available, 15 new patients were seen (Table V). Both groups had been treated with the same chemotherapy protocol and similar antibiotic regimens during episodes of fever and neutropenia. We studied the course of these 29 patients through remission-induction, maintenance and eventual relapse. Sepsis during neutropenia occurred more frequently during induction in the group in which white blood cell transfusions were available. During relapse, infections were more frequent in the other group. There were 11 episodes of bacteremia in each group (Table V). The microorganisms cultured from the blood in both groups were similar (Table VI). These included the usual gram-positive and gramnegative pathogens plus some organisms which are not usually pathogenic but may cause disease in neutropenic patients. An occasional positive culture in each group may have been caused by contamination. Six of the seven patients who developed bacteremia during induction were given granulocyte trasnfusions in the study group (Fig. 1) . One patient recovered without white cell support. Five of them survived. Only one of four patients in the control group survived (Table V) and the significance of the patient’s bacter-
TABLE V. Results of Sepsis in Nonlymphobhtic Leukemia 1971-1975
1975-1977
WBC not available
WBC available
Number of consecutive patients
15
14
Positive blood cultures induction relapse
11 4 7
11 7 4
Number given WBC transfusions
0
9
Death with positive blood culture induction relapse
8/11 (73%) 314 (75%) 517 (71%)
2/11 217 (29%) 0/4 (0%)
Time of death after positive blood cultures
4-26 (median 11.5) days
17 & 24 days
Time defervescence
1, 2,9, days
1-32 (median 15) days
Clearing of positive culture
2/10 (20%)
113 (33%)
without M I marrow aBoth deaths due to compromised bowel.
Granulocyte Transfusions in Children
73
emia is questionable; the patient became afebrile one day after Diphteroids were identified in the blood. Five of the seven patients in the control group whose blood cultures were positive in relapse died, while all four during the study period survived (Fig. 2). Three of them received granulocyte support. Figures 1 and 2 show the clinical course of these patients. Of the study patients, nine out of eleven (82%)survived, while only three out of 1 1 (27%)of the control group survived. Both deaths in the study group were due to sepsis secondary to necrotic bowel, a condition that granulocytes could not possibly correct. The 82%survival of sepsis in nonlymphoblastic leukemia is similar to the 86%achieved with bacteremia in the entire study (Table IV). Normal marrow recovery significantly influenced the clearing of positive blood cultures. Only 33% of the patients who received granulocytes survived without normal marrow function, compared to 20%in the untreated nonlymphoblastic leukemia group. All patients with marrow recovery survived. In the 21 patients (Table IV) receiving granulocytes, the immediate survival of a septic episode without marrow recovery was 62%,but this was a transient response. They were soon reinfected and only 25% lived more than two months (Table 111). This seems to indicate that severely infected, neutropenic patients do benefit from granulocyte transfusions when antibiotics alone are ineffective. The effect is clincally significant only if normal hematopoiesis returns. Granulocyte transfusions probably allow the patient to survive longer and thereby permit more time for marrow recovery. The median survival of patients without white cell support was 11.5 days. The median time of clinical improvement in the granulocyte transfused group was 15 days (Table V), and the median time to return of normal marrow function was 18 days. White cell support, in most cases, seemed to prolong life for one week to permit marrow recovery. The results obtained in this study are much better than those reported by Sanders, et a1 [13] and are similar to those of Highby, et al [8] both of whom used filter-collected cells. Those cells, although larger in number, may cause more reactions and have been found by some investigators to be functionally impaired [7, 14-16]. The other studies in the literature are mainly in adults [2-lo].
TABLE VI. Microorganisms Cultured from Blood in Nonlymphoblastic Leukemia
Organisms Escherichia coli Pseudomonas Serratia marcescens Klebsiella Clostridium Streptococcus viridans Diplococcus pneumoniae Staphylococcus aureus Staphylococcus epidermidis Propionibacterium acnes Diphteroids
Number of patients with positive cultures WBC not available [ 11 ] WBC available [ 1 1 ]
2 0 1 1 0
0 0 2
2 2 1 1 1 2 1 1
1
4
1
4
4
1
14
Steinherz and Reich ~~~
M Fever
Blood culture positive WBC transfusion
0
Deoth
BM+ Bone morrow recovery BM- No recovery
WBC ovoiloble Prop. oc.
Slaph.47,:
& BM +
&c
Strw.y:
+p*m
Prop. oc.
o-o BM+ . *WW r. *w -BMt
-
Strep. y. Prcp oc.
Stoph. ou Prop oc.
0
Pseudo.
0
**
* *
*ww+w*
*!3
BM-
0
.
BM+
+b++*3
**
o BM+
339 # 33 M W
Clostridio Pseudo.
e
E co/i I
.*
H3 ##3 *#
Sbph.epi Klebs.
-30
3
I
-20
I
I
-10
4
BM-
I
I
I
I
10
20
30
40
Days before or after 1st positive blood culture
.
WBC not avoiloble Diphtheroids
~4 BM+
LBM-
Stoph. epl:
W...
MBM-
E COli Klebs. I
-30
BMI
1
1
I
I
I
-20 -10 10 20 30 Days before or after 1st positive blood culture
J
40
Fig. 1 . Course of sepsis in patients with nonlymphoblasticleukemia during first remission induction.
White cell transfusions have minimal immediate side effects and the source of the cells (parental or random donors) and the method of collection does not seem to be important as long as enough cells are given. Prolonging the time through which a myelosup pressed patient can be supported is becoming more and more important. Granulocyte support permits the use of more aggressive multiagent chemotherapy protocols which are required for good results in certain pediatric malignancies such as nonlymphoblastic leukemia, lymphoma, osteogenic sarcoma and Ewing’s sarcoma - and it is essential for the proper evaluation of new therapeutic agents in relapsed patients. We feel that, if there is some chance of marrow recovery, all neutropenic, infected children should be
Granulocyte Transfusions in Children
75
MFever
Blood culture positive WBC transfusion 0 Death 0M+ Bone marrow recovery BM- No recovery
C
WBC ovailoble
& BM+
Staph. ou.
I
Y
-51 -/30
I
I
I
I
I
-20
-10 10 20 30 Days before or after 1st positive blood culture
1
40
WBC not ovailoble
-
Dip/.pn.
00 0M-
*
Dibhf.
Ecoli
*
BM-
BM-
oipn,.
BM-
0
-
Prop. oc.
*
0
Dipht
**
Serrotio
sropn. OU. I
-30
I
H
BM-
0M-
0
0 I
I
I
10 20 Days before or after 1st positive cultures
-20
10-
1
0M1
1
30
40
Fig. 2. Course of sepsis in patients with nonlymphoblastic leukemia in relapse.
given granulocyte transfusions when 48 hours of antibiotics have failed. Early institution of white cell support may prevent irreversible changes and death during induction therapy. The white cells should be given daily, 1 X 101O/mZ,from any ABO compatible donor until infection clears or normal granulocytopoiesis returns. Repeated courses of granulocyte transfusions without marrow recovery will not appreciably affect long-term survival. The question of prophylactic granulocyte transfusions prior to the occurrence of overt infections in granulocytopenic patients remains to be answered.
76
Steinherz and Reich
REFERENCES 1. Bodey GP, Buckley M, Sathe YS,Freireich EJ: Quantitative relationships between circulating leukocytes and infection in patients with acute leukemia. Ann Intl Med 64:328-336, 1966. 2. Graw RG, Herzig G, Perry S, Henderson ES: Normal granulocyte transfusion therapy-Treatment of septicemia due to gram negative bacteria, N Engl J Med 287:367-371, 1972. 3. Alavi JB, Root RK, Djerassi I, Evans AE, Gluckman SJ, MacGregor RR, Guerry D, Schreiber AD, Shaw JM, Koch P, Cooper RA: A randomized clinical trial of granulocyte transfusions in infection in acute leukemia N Engl J Med 296:706-711, 1977. 4. Bogs DR: Transfusion of neutrophils as prevention or treatment of infection in patients with neutropenia. N Engl J Med 290: 1055-1062,1974. 5. Herzig RH, Herzig GP, Graw RC Jr, Bull MI, Ray KK: Successful granulocyte transfusion therapy from gram negative septicemia, N Engl J Med 296:701-705, 1977. 6. Hester JP, McCredie KB, Freireich EJ: Granulocyte transfusion: analysis of donor, procedure, recipient variables. Blut 32:253-256, 1976. 7. Highby DJ, Henderson ES: Granulocyte transfusion therapy. Ann Rev Med 26:289-306, 1975. 8. Highby DJ, Freeman A, Henderson ES, Sinks L, Cohen E: Granulocyte transfusions in children using filtercollected cells. Cancer 38:1407-1413, 1976. 9. MacPherson JL, Nusbacher J, Bennett JM: The acquisition of granulocytes by leukapheresis. Transfusion 16:221-228, 1976. 10. Reich LM, VonSickle P, Wright P: Granulocyte transfusion: A preliminary study. In “Leukocytes: Sepration, collection and transfusion. Proceedings of the International Symposium on Leukocyte Separation and Transfusion.” London, September 1974, pp 294-299. 11. American Association of Blood Banks Technical Methods and Procedures: Miller V (ed). Philadelphia: JB Lippincott, 1974. 12. Graw RG, Bucker CD, Whang-Peng J, Leventhal BG, Kruger G, Berard C, Henderson ES: Complications of bone marrow transplantation. Graft-versus-host disease resulting from chronic myelogenous leukaemia leukocyte transfusions. Lancet 2: 338-340, 1970. 13. Sanders J, Wilbur JR, Long T, Etcubanas E, Babin S,Grant J: Granulocyte transfusions in the treatment of infection in children with malignancy. In “Proceedings of the Eleventh Annual Meeting of the American Society of Clinical Oncology.” San Diego, May 1975, p 269. 14. Ts’ao C, Ruder EA: Ultrastructural damage of leukocytes procured by the leukopak: Vulnerability of leukocytes to mechanical injury. Transfusion 16:336-344, 1976. 15. Klok JC, Bainton DF: Degranulation and abnormal bacterial function of granulocytes procured by reversible adhesion to nylon wool. Blood 48:149-161, 1976. 16. Wade PH, Skrabut EM, Vinciguerra L, Valeri CR: In vitro function of granulocytes isolated from blood of normal volunteers using continuous-flow centrifugation in the IBM-Aminco celltrifuge and adhesion-fiitration leukapheresis using nylon fiber. Transfusion 17:136-140, 1977.
