Case report Granulocyte transfusion as a treatment for ... - Nature

Bone Marrow Transplantation (2003) 31, 69–72 & 2003 Nature Publishing Group All rights reserved 0268-3369/03 $25.00

www.nature.com/bmt

Case report Granulocyte transfusion as a treatment for enterococcal meningoencephalitis after allogeneic bone marrow transplantation from an unrelated donor Y Tsukada1, H Nagayama1, T Mori1, T Shimizu1, N Sato1, N Takayama1, A Ishida2, M Handa2, Y Ikeda1 and S Okamoto1 1 Division of Hematology, Department of Medicine, Keio University School of Medicine, Tokyo, Japan; and University School of Medicine, Tokyo, Japan

Summary: Bacterial meningoencephalitis occurring in the pre-engraftment period after bone marrow transplantation (BMT) is a rare complication, and the feasibility of granulocyte transfusion (GTX) in such cases remains to be elucidated. A 37-year-old man developed enterococcal meningoencephalitis during a severely granulocytopenic pre-engraftment period after BMT. Despite therapy with appropriate antibiotics, cultures of blood and cerebrospinal fluid (CSF) continued to grow Enterococcus faecalis, and he developed rapid mental deterioration and seizure. Granulocytes were collected from his HLAmismatched, ABO-matched sibling with subcutaneous injection of granulocyte colony-stimulating factor (GCSF) and oral dexamethazone. Transfusion of 4.4 1010 granulocytes resulted in a 12-h post-transfusion granulocyte increment of 2.0 109/l, and maintained peripheral blood granulocyte counts above 0.5 109/l for 3 days. A rapid increase of granulocytes in CSF was also observed, and cultures of blood and CSF became negative after GTX. A transient worsening of seizure was observed as a potential side effect of GTX. The patient subsequently developed septic shock because of Pseudomonas aeruginosa and died. Further studies are warranted to evaluate the clinical efficacy of GTX for the treatment of uncontrolled infections in granulocytopenic stem cell transplant recipients. Bone Marrow Transplantation (2003) 31, 69–72. doi:10.1038/sj.bmt.1703780 Keywords: enterococcus; meningoencephalitis; granulocyte transfusion; G-CSF; bone marrow transplantation Bacterial or fungal infections continue to be a major cause of morbidity and mortality during the post-transplant period until engraftment. Recently, the incidence of vancomycin-resistant Enterococcus (VRE)-related infection has been increasing, and its treatment has become a Correspondence: Dr Nagayama, Division of Hematology, Department of Medicine, Keio University School of Medicine, 35, Shinanomachi, Shinjyuku-ku, Tokyo 160-8582, Japan Received 15 April 2002; accepted 30 July 2002

2

Blood Center, Keio

serious problem during the pre-engraftment period following BMT.1–3 However, Enterococci rarely cause CNS infection in granulocytopenic patients, and there are few anecdotal cases of meningoencephalitis due to Enterococcus faecalis after BMT.4–6 Transfusion of normal granulocytes is a logical treatment for severe bacterial infections in neutropenic patients. However, granulocyte transfusion (GTX) had a limited role in the treatment of such infections before the wide clinical use of granulocytes colony stimulating factor (G-CSF). A major reason for the lack of efficacy was that the dose of granulocytes collected and transfused was inadequate. The availability of G-CSF prompted several investigators to re-evaluate the efficacy of GTX in the treatment of uncontrolled bacterial and fungal infection during granulocytopenia. Bielorai et al reported a case of VRE septicemia in a neutropenic patient, which was successfully treated with GTX.3 The phase I/II trials of GTX mobilized by G-CSF for the treatment of infections complicated with hematological disorders also demonstrated its efficacy.7 These trials also demonstrated the migratory activity of transfused granulocytes to the tissue site; however, whether transfused granulocytes migrate to the central nervous system (CNS) remains unknown. In this report, we describe a case of enterococcal meningoencephalitis before marrow engraftment, and discuss the feasibility and efficacy of GTX against CNS infection.

Case report A 37-year-old man was diagnosed with acute myeloblastic leukemia with tri-lineage dysplasia (tMDS-AML) in August 2000. Complete remission was achieved with combination chemotherapy consisting of idarubicin (IDR) and cytosine arabinoside (CA). One additional course of the same chemotherapy was given and the patient was followed as an outpatient without further chemotherapy. He was admitted to the Keio University Hospital in September 2001 to undergo allogeneic bone marrow transplantation (BMT) from an HLA-matched unrelated donor. Although the patient had remained in complete remission, bone marrow examination 3 weeks before the

GTX for meningoencephalitis after BMT Y Tsukada et al

70

transplant showed 21.6% blastic cells, indicating that he was in first relapse. A lumbar puncture on day 12 did not show leukemic cells in the cerebrospinal fluid (CSF), and a combination of methotrexate (MTX), CA, and hydrocortisone sodium succinate was given intrathecally for the prophylaxis of CNS leukemia. Culture of CSF was not performed. A pretransplantation evaluation that included whole body computed tomography (CT), radioisotope imaging, and ophthalmologic, oto-rhino-pharyngeal, dental, and perianal examination failed to detect any infectious foci. Oral ciprofloxacin hydrochloride (CPFX) at a dose of 600 mg/day and fluconazole (FCZ) at a dose of 200 mg/day were started on day 14 as gut decontamination. The conditioning regimen consisted of total body irradiation (2 Gy twice daily for three consecutive days) followed by CA (3 g/m2 twice daily i.v. for four consecutive days). The WBC count was 2.6 109/L on the day of starting the conditioning (day 8), and became less than 0.5 109/L on day 2. Transplantation of 1.8 108/kg of bone marrow mononuclear cells from an HLA-matched unrelated donor was then performed, and intravenous G-CSF at a dose of 5 mg/kg/day was started from day +1. Cyclosporine (CyA) and short-term MTX were administered for GVHD prophylaxis. Acyclovir (ACV) was administered intravenously from day 3 to day +14 of BMT as prophylaxis against herpes viruses. Surveillance cultures before transplantation revealed the presence of methicillin-sensitive Staphylococcus aureus (MSSA) and a-hemolytic streptococcus in the throat, and Staphylococcus epidermidis in the nasal cavity. Repeat urine and stool cultures failed to grow any bacteria, including Enterococcus faecalis. On day 2, the patient began to complain of lower abdominal cramping pain and watery diarrhea. The diarrhea became more frequent on day +2, and was accompanied by an elevated serum CRP value (2.07 mg/dl) and low-grade fever (37.31C). Oral CPFX was discontinued, and intravenous ceftazidime (CAZ) (4 g/day) and arbekacin sulfate (ABK) (150 mg/day once daily) were started. The fever and CRP resolved gradually, but the diarrhea continued. On day +9, his body temperature rose to 39.61C. Because the throat and stool cultures taken on day +5 grew Enterococcus faecalis, the intravenous antibiotics were changed to meropenem trihydrate (MEPM) (1.5 g/day) and vancomycin hydrochloride (VCM) (1.5 g/day). On day +11, he began to complain of severe headache, then developed a grand mal seizure. Results of a CT scan of the brain were unremarkable. The patient was placed on phenytoin and valproate sodium, and intravenous ampicillin sodium (ABPC) (12 g/day), and cefepime dihydrochloride (CFPM) (4 g/day) were started in addition to VCM and ACV. Mechanical ventilation was required on day +13 to control persistent convulsions. Blood cultures on days +9 and +10 grew Enterococcus faecalis. Lumbar puncture yielded cloudy CSF with protein 333 mg/dl, glucose 94 mg/dl (serum glucose 134 mg/dl), and a cell count of 27/mm3 (with monocytes 25/27, and granulocytes 2/27). The culture of CSF also grew Enterococcus faecalis (Table 1). Enterococcus faecalis isolated from both blood and CSF was sensitive (MIC o1 mg/ml using Kirby–Bauer method) to ABPC and VCM, intermediately susceptible (MIC ¼ 2 mg/

Bone Marrow Transplantation

Table 1 Findings of cerebrospinal fluid before and after granulocyte transfusion Before GTX (Day +13)

After GTX (Day +17)

Total cell count (mm3) Monocytes (mm3) Neutrophils (mm3) Total protein (mg/dl) LDH (IC/l) Chloride (mEq/l) Glucose (mg/dl)

27 25 2 333 88 114 94

966 139 827 72 83 114 110

Plasma glucose (mg/dl) Granulocyte counts (PB)

145 100

284 2000

ml) to CPFX in vitro. Cardiac echography failed to detect any vegetations in the heart. Despite therapy with appropriate antibiotics, the infection remained uncontrolled, and GTX was performed using an HLA-mismatched ABO-compatible elder brother as donor. After providing his informed, written consent, the donor received dexamethasone 8 mg orally and G-CSF 10 mg/kg subcutaneously approximately 12 h before the scheduled granulocyte collection. Granulocytes were collected as per standard leukapheresis, processing 8 l of blood. No serious adverse events, apart from bone pain, insomnia, and flushing, were observed in the donor. Granulocytes 44 109 were collected and irradiated at 25 Gy, then transfused to the patient on day +13 (Figure 1). Peripheral blood granulocyte counts rose to 2.0 109/l at 12 h after GTX and remained above 0.5 109/ l for 3 days. Cultures of peripheral blood and CSF became negative and serum CRP levels gradually decreased. Table 1 shows the CSF findings before and after GTX. The absolute granulocyte count in the CSF increased dramatically after GTX, suggesting migratory activity of transfused granulocytes into the CSF. However, a transient worsening of the seizures was observed. On day +16, CFPM was discontinued and intravenous and intrathecal administration of gentamicin sulfate (GM) was started. Blood cultures grew Pseudomonas aeruginosa on day +20, and the patient went into septic shock. Repeated GTX was not planned because he developed multiorgan failure. He died on day +20 without any evidence of marrow recovery. Permission for an autopsy was declined by his family.

Discussion Two important aspects of this case merit attention: the presence of enterococcal meningitis, a rare infectious complication immediately after BMT, and the use of GTX as part of its treatment.1,2 Enterococci are unusual etiologic agents of bacterial meningitis, and previous reports have shown that enterococcal meningitis tends to occur in patients with chronic medical conditions. We speculated that Enterococci entered the blood stream from a damaged gastrointestinal tract,


BMT

GTX

G-CSF

ABK

CFPM ABPC

GM

VCM Seizure

41 BT (°C)

2.0 1.5 1.0 0.5

37 Bacteria grown from cultures Peripheral blood (-) IVH blood Throat Stool S CSF -2

0

(-)

(-)

E E

E E

E E

E

+10

+12

(-) (-) E (-) (-)

Granulocytes (×109/L)

MEPM

CAZ

CPFX

P

E S

E E

0

+2

+4

+6

+8

(-) +14 +16 +18 +20

Days before / after transplantation

Figure 1

Clinical course. S, Staphylococcus epidermidis; E, Enterococcus faecalis; P, Pseudomonas aeruginosa; (), negative culture; GTX, granulocyte transfusion; CPFX, ciprofloxacin hydrochloride; CAZ, ceftazidime; CFPM, cefepime dihydrochloride; MEPM, meropenem trihydrate; ABK, arbekacin sulfate; GM, gentamicin sulfate; ABPC, ampicillin sodium; VCM, vancomycin hydrochloride; ——BT, body temperature (1C).

and then seeded to the meninges. Pretransplant intrathecal chemotherapy and pretransplant conditioning with total body irradiation are possible causes of destruction of the BBB, and may have contributed to the dissemination of Enterococcus from the blood stream to the CNS. In addition, the enterococcal septicemia in the present case may also have been caused by the use of quinolone and third generation of cephalosporine (Ceph 3) as infectious prophylaxis and empiric treatment for febrile neutropenia. In recent years, the value of quinolone prophylaxis has been questioned. Sensitivity of Gram-positive bacteria to CPFX has fallen dramatically. In this case, the Enterococcus isolated from the blood and CSF showed intermediate sensitivity to CPFX. Thus, the use of CPFX and Ceph 3 may have predisposed to the dominant growth of resistant bacteria such as Enterococci. During the pre-G-CSF era, the indication of GTX in granulocytopenic patients with infections not controllable by antibiotics was controversial because of the low yield of granulocytes, the tremendous workload for the donor, and the lack of definitive clinical and cost-effectiveness. However, the use of G-CSF to mobilize granulocytes resulted in a sufficient number of granulocytes for transfusion, minimal repetitions of apheresis, minimal donor work load, and improved functions of transfused granulocytes. Bielorai et al3 reported a case of meningoenterococcal septicemia successfully treated using GTX, after chemotherapy for AML. Recent phase I/II trial of GTX among community blood banks from HLA-mismatched unrelated donors reported 165 transfusions with adequate granulocyte increments in 19-granulocytopenic stem cell transplant recipients.7 In their protocol, dexamethasone (8 mg orally) and G-CSF (600 mg subcutaneously) were administered to the donor approximately 12 h before the scheduled collection. They transfused an average of

81.9 109 granulocytes, which resulted in a mean 1 h post-transfusion granulocytes increment of 2.6 109/l on 165 occasions. They also reported that life-threatening bacterial or fungal infections resolved in eight out of 19 granulocytopenic patients, and they measured the migratory activities of transfused granulocytes by using the number of migrated granulocytes in saliva. We observed the same yield of granulocytes after GTX as that reported by Price et al.7 Further, we were able to clinically confirm the migratory activity of transfused granulocytes into the CSF. The rapid clearance of bacteria from the blood and CSF also suggests that the transfused granulocytes possessed sufficient bactericidal functions. The fact that the granulocyte counts remained above 0.5 109/L in this case suggests that G-CSF-stimulated GTX may reduce the frequency of apheresis, and decrease donor work load. However, this will differ between recipients. Thus, Price et al reported that polymorphonuclear cells of the peripheral blood 1 h after GTX varied from 0.2 to 6.9 109/L.7 One important concern with regard to the use of G-CSFstimulated GTX is that seizures may worsen after GTX, making bacteriological and hematological improvements irrelevant. It is also possible that worsening of the seizures occured as a result of CNS infection, rather than a result of the infusion of HLA-mismatched, G-CSF-stimulated granulocytes. It is well known that serious pulmonary reaction is one of the reasons for limiting the use of GTX in clinical practice. Wright et al reported that, among their 57 patients who received GTX and amphotericin B, 12 patients suffered from acute respiratory distress syndrome as an adverse event.8 Transfusion of G-CSF-activated granulocytes may cause serious organ damage via granulocyte-derived cytokines, free radicals, and enzymes. However, Price et al observed few adverse reactions after GTX.7 Adkins et al reported that adverse reactions were not associated with human leukocyte antigen (HLA).9 Adkins et al also reported that leukocyte compatibility did affect the peak count of polymorphonuclear cells and delayed neutrophil engraftment after stem cell transplantation.9 In the present study, although the patient received granulocytes from an HLA-mismatched sibling and engraftment was not achieved, it is difficult to attribute the engraftment failure to leukocyte incompatibility alone. In summary, we examined the feasibility of GTX for the treatment of serious enterococcal meningoencephalitis and septicemia in the severely granulocytopenic transplant recipient. GTX seemed to be effective in controlling lifethreatening infection, but further evaluation will be needed to confirm the efficacy and safety of GTX for granulocytopenic transplantation recipients.

71

Acknowledgements We thank Mrs Matsuhashi H, MT and the other staff members of the Blood Center of Keio University Hospital for their excellent technical assistance with the PMNC collection. We also thank the staff doctors and nurses at the BMT unit of Keio University Hospital. Bone Marrow Transplantation


72

References 1 Koc Y, Snydman DR, Schenkein DS et al. Vancomycinresistant enterococcal infections in bone marrow transplant recipients. Bone Marrow Transplant 1998; 22: 207–209. 2 Kapur D, Dorsky D, Feingold JM et al. Incidence and outcome of vancomycin-resistant enterococcal bacteremia following autologous peripheral blood stem cell transplantation. Bone Marrow Transplant 2000; 25: 147–152. 3 Bielorai B, Neumann Y, Avigad I et al. Successful treatment of vancomycin-resistant Enterococcus sepsis in a neutropenic patient with G-CSF-mobilized granulocyte transfusions. Med Pediat Oncol 2000; 34: 221–223. 4 Stevenson KB, Murray EW, Sarubbi FA. Enterococcal meningitis: report of four cases and review. Clin Infect Dis 1994; 18: 233–239. 5 Jang TN, Fung CP, Liu CY et al. Enterococcal meningitis: analysis of twelve cases. J Formos Med Assoc 1995; 94: 391–395.

Bone Marrow Transplantation

6 Zeana C, Kubin CJ, Della-Latta P et al. Vancomycin-resistant Enterococcus faecium meningitis successfully managed with Linezolid: case report and review of the literature. Clin Infect Dis 2001; 33: 477–482. 7 Price TH, Bowden RA, Boeckh M et al. Phase I/II trial of neutrophils transfusions from donors stimulated with G-CSF and dexamethasone for treatment of patients with infections in hematopoietic stem cell transplantation. Blood 2000; 95: 3302–3309. 8 Wright DG, Robichaud KJ, Pizzo BS et al. Lethal pulmonary reactions associated with the combined use of amphotericin-B and leukocyte transfusions. N Engl J Med. 1981; 20: 1185–1189. 9 Adkins DR, Goodnough LT, Shenoy Shalini et al. Effect of leukocyte compatibility on neutrophils increment after transfusion of granulocyte colony-mobilized prophylactic granulocyte transfusions and on clinical outcomes after stem cell transplantation. Blood 2000; 95: 3605–3612.