Background: Through recent progress in chemotherapy and bone marrow
transplantation, patients with acute myeloid leukemia(AML) have shown a long-term,
disease free survival. Myelosuppression from therapy and infections have been the major
obstacles in the treatment of acute leukemia. Recently, in order to shorten the duration
of neutropenia and reduce the rate of infection, rhG/GM-CSF(recombinant human
granulocyte or granulocyte/macrophage colony-stimulating factor) has been used in
clinical trials. It has been apparent from earlier studies that rhG/GM-CSF accelerate the
recovery of neutrophils after chemotherapy, but there were variable results in the rate of
infection, febrile duration, and clinical responses to treatment for infection. Thus we
evaluated the effect of rhG/GM-CSF on infection in patients with acute myeloid
leukemia.
Methods: This study reviewed the 262 chemotherapy courses of 121 cases with AML
from January 1990 to June 1995. G-CSF 5㎍/kg and GM-CSF 250㎍/㎡ was
subcutaneously administrated until neutrophil count > 1,000/μL for three consecutive
days. When febrile episodes were noted, stepwise-empiric antimicrobial therapy was
started with microbial culture study. We compared recovery of neutrophils, infection rate,
and clinical response to the treatment of infection in study patients with the results
obtained in historical controls.
Results :
1) Of 262 chemotherapy courses, rhG/GM-CSF was administrated in 85 chemotherap
courses, and 177 courses were not. In the rhG/GM-CSF group, rInG-CSF was
administrated in 37 chemotherapy courses, rhGM-CSF in 48 courses. And rhG/GM-CSF
was administrated for prophylaxis of infection in 49 courses. In the remaining 36
courses, it was used for enhancing the effect of treatment on documented infection.
2) Recovery of neutrophils over 500/μL, 1,000/μL and 1,500/μL, respectively, was
significantly shorter in the rhG/GM-CSF group than in the control group(P value<0.01).
3) The incidence of documented infection was significantly lower in the rhG/
GM-CSF than in the control group(P value=0.03), and the recovery from neutropenia
induced by administration of rhG/GM-CSF is the most important predictor in rate of
infection (P value=0.03). The second important factor is the type of chemotherapy(P
value=0.05).
4) The duration of febrile days and type of infection were not different between the
two groups. Also, the organisms of microbiologically documented infection were not
different.
5) In 55 courses with documented infection, rhG/GM-CSF was administrated in
combination with antimicrobials. The clinical response to treatment of documented
infection in this group was 83.6% , not significantly different from those in the control
group(82.4%).
6) There was no statistical difference in the incidence of infection and clinical
response according to induction timing of hypoplasia after chemotherapy.
Conclusion: With the above results, we have concluded that rhG/GM-CSF
administration after chemotherapy in AML lowers the incidence of infection by
accelerating the recovery of neutrophil counts. There is a need to search for earlier
administration of rhG/GM-CSF to improve clinical response to treatment of documented infection.

Keywords: Acute myeloid leukemia, Recombinant human granulocyte or granulo-cyte/macrophage colony-stimul