Treatment of Neutropenia-Related Fungal Infections with Granulocyte

Home , Neutropenia

Treatment of neutropenia-related fungal infections with granulocyte colony- stimulating factor-elicited white blood cell transfusions: a pilot study MC Dignani1, EJ Anaissie2, JP Hester3, S O’Brien4, SE Vartivarian5, JH Rex6, H Kantarjian4, DB Jendiroba4, B Lichtiger7, BS Andersson4 and EJ Freireich4

2Section of Infectious Diseases, Department of Medical Specialties, 4Department of Hematology, 7Laboratory Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas, USA; 1Infectious Diseases, Fundaleu, Buenos Aires, Argentina; 3Haemapheresis Consultants, Houston, Texas; 6Department of Internal Medicine, Center for Infectious Diseases, 5The University of Texas Medical School, Houston, Texas, USA

Neutropenia-related fungal infections can be life-threatening and efficacy of rG-CSF-elicited WBC transfusions in patients despite antifungal therapy. We evaluated the role of recombi- with neutropenia-related fungal infections that were refractory nant granulocyte colony-stimulating factor (rG-CSF)-elicited white blood cell (WBC) transfusions in patients with neutro- to therapy with amphotericin B. penia-related fungal infections. Adult patients with hematologic malignancies, absolute neutrophil counts (ANC) Ͻ500/␮l and fungal infections refractory to amphotericin B, received daily Materials and methods transfusions of rG-CSF-elicited and irradiated WBC transfusions from related donors. Donors received 5 ␮g/kg/day of rG-CSF subcutaneously. Donors achieved a mean ANC of Study design ؋ 103 per microliter. The mean yield of neutrophils per 29.4 transfusion was 41 ؋ 109 (range, 10–116). Fifteen patients This pilot study was conducted at The University of Texas MD received a median of eight transfusions (range, 3–16). Fourteen Anderson Cancer Center between May 1993 and May 1994. patients had received rG-CSF for a median of 12 days. The Informed consents were obtained from donors and recipients. median ANC baseline was 20/␮l. Eleven patients had favorable responses and eight of them remained free of infection 3 weeks after therapy. Favorable responses occurred among patients with better Zubrod performance status (median, 3 vs 4) and Patients shorter duration of both profound neutropenia (median, 15 vs 25 days) and active infection (median, 8 vs 17 days). The mean Adult patients were enrolled in the study if they had a hemato- 1- and 24-h post-transfusion ANCs were 594/␮l (range, 98– Ͻ ␮ 1472/␮l) and 396/␮l (range, 50–1475/␮l), respectively. Adverse logic malignancy, neutropenia (ANC 500/ l), and an invas- 1 reactions were observed in nine of 35 donors and in the recipi- ive fungal infection as previously defined. Infections were ents of six of 130 transfusions. rG-CSF-elicited WBC trans- documented by culture, biopsy specimen, or both. Mold fusions may be a safe and promising approach for treating infections were confirmed if the biopsy specimen showed neutropenia-related fungal infections. acute branching septated hyphae but the culture evaluation Keywords: granulocyte transfusion; granulocyte colony-stimulating yielded negative results. factor (G-CSF); fungal infections; leukemia; pneumonia Pretreatment evaluation included history and physical examination, Zubrod scale performance status and hematologic evaluation, liver and kidney function tests, and appropri- Introduction ate roentgenographic, histopathologic, and culture evalu- ations of infected sites. Patients were examined and the Neutropenia-related fungal infections can be life-threatening 1 above-mentioned laboratory parameters were monitored daily in cancer patients despite antifungal therapy. The role of during therapy, at the end of therapy, and 3 weeks after the white blood cell (WBC) transfusions in the treatment of neu- end of therapy. Patients were also evaluated at least 3 months tropenia-related infections has been examined and remains 2–13 after the end of therapy. controversial. Patients were evaluable for response if they received three The low number of donor cells collected probably accounts or more transfusions. Response was evaluated by two investi- for the marginal efficacy of WBC transfusions in some studies. gators (MCD and EJA). Patients were considered to have a fav- A dose–response was established in some experimental14 and 8,11,15 orable response if they had improvement or stabilization of clinical studies. the clinical and laboratory parameters on the last day of WBC Dexamethasone has been the preferred agent for donor transfusion therapy. stimulation. However, administration of recombinant granulo- Response at 3 weeks and 3 months was defined as pre- cyte colony-stimulating factor (rG-CSF) safely increases the viously reported1 and was based on a combination of clinical number of circulating neutrophils in patients with neutro- and laboratory findings. penia16 and results in a significantly higher neutrophil yield per collection and a higher absolute neutrophil count (ANC) (1) Complete response: This indicated the disappearance of in the recipients than the administration of dexamethasone or all clinical and laboratory signs and symptoms of infection no treatment at all.17 including radiologic examination, and negative cultures We conducted this prospective study to evaluate the safety from previously infected sites when possible. The persistence of minimal scarring, as determined by a radiologist not participating in the study, was considered as a complete response if no recrudescence of disease was noted Correspondence: Dr EJ Anaissie, Department of Hematology/ Oncology, The University of Arkansas for Medical Sciences, Little (ie all clinical and laboratory parameters of infection have Rock, AR 72205, USA remained negative for at least 3 months after discontinu- Received 28 October 1996; accepted 26 June 1997 ation of therapy or until patient’s death, whichever came G-CSF-stimulated leukocyte transfusions MC Dignani et al 1622 first, and/or if complete autopsy examination failed to Collection rates varied from 1.5 to 3 ml per min. All WBC show persistent infection). concentrates were irradiated with 3000 rad except for four (2) Complete response with residuae indicated continuous concentrates collected from a bone marrow donor. In this improvement and clinical evolution compatible with the patient, the intent of the WBC transfusion was to both treat natural history of mold infection and no recrudescence of infection and induce myeloid engraftment (patient 13). disease described above. (a) Pulmonary: defervescence of fever, resolution of clinical signs and symptoms of infection, no new lesions White-blood cell transfusions or increase in existing lesions, negative respiratory tract cultures, but with persistent pulmonary cavity. The WBC concentrates were infused within 2 h of collection (b) Sinusitis: defervescence of fever, resolution of clinical over 60–90 min and at least 4 h after the amphotericin B signs and symptoms of infections, negative cultures. infusion. Recipients were premedicated with hydrocortisone The persistence of residual non-specific findings by (100 mg) and diphenhydramine intravenously (25 mg) and sinus X-rays or CT scan did not exclude response if no acetaminophen (650 mg) orally. WBCs were transfused daily recrudescence of disease, as described above, was when possible, and the transfusions were continued until no seen. more donors were available or until patients recovered from (3) Partial response: Partial response indicated clinical myelosuppression (ANC у1000/␮l for 2 successive days with- improvement, and regression of more than 50% in the size out transfusion), died, had no response to therapy, or were of objective parameters of fungal infection. cleared of infection. (4) Stable disease: Indicated stabilization in the clinical signs Complete blood and platelet counts of the concentrates and symptoms of infection, no change in the size of the were obtained with a Coulter automated STKS hematology objective measurements or regression of less than 50% in system (Coulter Corporation, Miami, FL, USA), and a differen- the size of these measurements. tial count was done manually. (5) Failure: Indicated progressive infection (clinical and/or ANCs were obtained for all patients at baseline and at 1 laboratory). and 24 h after the WBC transfusions. Occasionally, WBC (6) Relapse: Documented infection with the same organism transfusions were withheld for 24 h because of suspected at any site occurring at any time after discontinuation of bone marrow recovery or because we could not perform leu- study drug. Same organism implies same antibiogram. kapheresis on Sundays. In those cases, the ANC was obtained A favorable response indicated complete or partial resolution 48 h after transfusion. of the infection. Donor and recipient toxic effects Donors Toxic effects were assessed daily in all recipients and donors. Only related donors were enrolled in the study, except in two Recipients were monitored for adverse reactions during the cases in which spouses were allowed to donate pending the 2 h of WBC infusion and for 6 h after the infusion. The severity arrival of related donors. Human lymphocyte antigen (HLA) of the reactions was classified as: (1) mild: transient, resolved and blood typing were done on all donors, but no attempt without medication; (2) moderate: resolved with medication; was made to select donors on the basis of antigen or blood and (3) severe: potentially life-threatening and requiring dis- type compatibility. All donors were screened and accepted continuation of therapy. according to criteria established by the United States Food and Drug Administration. Donors received 5 ␮g/kg daily of rG-CSF by subcutaneous Statistical analysis injection until the day of the last leukapheresis. We chose rG- CSF instead of recombinant granulocyte–macrophage colony- Univariate logistic regression analyses were performed to stimulating factor because of the lower incidence of adverse determine important predictors of response at the end of ther- events and the higher level of neutrophilia associated with rG- apy. Post-treatment variables thought to have influenced CSF therapy.18 Eighteen to 24 h after their first dose, donors response were analyzed using the χ2 test (to examine differ- underwent leukapheresis. Most donors underwent leukapher- ences in proportions) and the Mann–Whitney test. A P value esis on alternate days, but consecutive donations were of Ͻ0.05 was considered significant. Correlation coefficients allowed for male donors when necessary. among hematologic parameters and between hematologic parameters and day of leukapheresis were computed.

White blood cell collections Results WBC concentrates were collected on a continuous-flow blood cell separator (Spectra; Cobe Laboratories, Lakewood, CO, Patients USA) using 500 ml of 10% low molecular weight starch (Pentaspan; Du Pont Critical Care, Waukegan, IL, USA) per Twenty patients were enrolled in the study; 15 were evaluable procedure, plus 40 ml of 46% citrate concentrate for response (Table 1). Five patients received fewer than three (Haemonetics, Braintree, MA, USA). HES-citrate was infused WBC transfusions because of early death (three patients), bone at 1 ml per liter of donor blood volume per min in a 1:13 marrow transplantation (BMT) after the first transfusion (one anticoagulant/WBC ratio with whole blood. This allowed us patient), and severe pulmonary adverse reaction (one patient). to process approximately 7 liters of blood for each procedure. Seventy-three percent of the patients were classified as level G-CSF-stimulated leukocyte transfusions MC Dignani et al 1623 Table 1 Characteristics of patients at study entry tion and was associated with BMT-induced GVHD (patient 4). None of the eight responders at 3 weeks had relapsing leuke- No. of patients 15 mia or BMT-induced GVHD. Sex: male/female 9/6 Three months after completing therapy, all living patients Median age, years (range) 30 (18–73) Malignancy were free of active disease: patients 2 and 11 were no longer Acute leukemia 9 neutropenic and their malignancies were in remission, and Malignant lymphoma 4 patient 6 remained neutropenic and his leukemia recurred in Chronic leukemia 2 the dermis. All of these patients had initially responsed to Treatment of malignancy WBC transfusion therapy. Chemotherapy 8 Six of the nonsurvivors died of their infections. Three BMT (allogeneic, autologous) 7 (4, 3) GVHD 1 (patient 13) patients were previous responders whose infections recurred Infecting pathogen after they underwent allogeneic BMT (patients 3 and 8) or Mold 11 whose malignancy recurred (patient 9). Patients with unfavor- (Aspergillus spp., Fusarium (7, 3, 1) able outcomes at the end of therapy continued to deteriorate spp., unidentified) 3 weeks and 3 months after therapy. Yeasts 4 Pre-treatment and post-treatment variables other than WBC (Candida spp., Trichosporon (3, 1) spp.) transfusions thought to have influenced response at the end Extent of infection of therapy are shown in Table 5. Patients with an unfavorable Localized 7 response tended to have been profoundly neutropenic (ANC (pneumonia, cellulitis, (4, 2, 1) Ͻ100/␮l) with documented fungal infection for longer periods sinusitis) of time and tended to have a worse performance status. The Disseminated 8 cumulative dose of amphotericin B given at study entry and Appropriate antifungal therapya the additional dose given during WBC transfusion therapy Median duration in days 10 (4–131) Ͼ (range) were comparable in both groups (P 0.5). Median cumulative dose of 900 (290–8600) amphotericin B in mg (range)b Median duration of neutropenia, days Donors ANC Ͻ500/␮l (range) 23 (9–58) ANC Ͻ100/␮l (range)c 14 (8–50) Treatment with rG-CSFd Thirty-five donors (21 men and 14 women) received rG-CSF Median duration, days (range) 12 (1–52) for a mean of 8 days (range, 2–10 days) and donated WBCs Median dose, ␮g/kg (range) 6.5 (5–13) to the 15 patients. The median donor/patient ratio was 2:1 Performance status (range, 1:1–4:1). Each donor provided a median of four WBC No. of patients with Zubrod 11 transfusions (range, 1–5). Characteristics of the donor/patient grade 3 or 4 No. of patients with Zubrod 4 pairs are shown in Table 6. Seventy-one percent of the donors grade 1 or 2 shared two HLA antigens with patients (A and B loci) and donated 72% of the WBC transfusions. All patients had BMT, bone marrow transplantation; GVHD, graft-versus-host received multiple transfusions of red blood cells and platelets, disease; ANC, absolute neutrophil count; rG-CSF, recombinant and alloimmunization could not be ruled out because we did granulocyte colony stimulating factor. not study HLA or leukoagglutinin antibodies at baseline. aIncludes amphotericin B and itraconazole in cases of aspergillosis Donors achieved a median ANC of 29.4 × 103 per microliter and fluconazole in the case of trichosporonosis. (range, 24–35/␮l). bAll three formulations of amphotericin B were considered together. cOne patient had an ANC of 300/␮l and was excluded. dOne patient did not receive rG-CSF and was excluded. White blood cell transfusions 3 or 4 on the Zubrod scale (bedridden for more than 50% of a normal day). Data regarding WBC transfusions and their effect on the 15 Data pertaining to the 15 evaluable patients are shown in evaluable patients are shown in Table 7. The mean and range Table 2, which is arranged by infection diagnosis. All patients, of ANCs taken 1, 24 and 48 h after transfusion include patient except for patient 15, had neutropenia refractory to treatment 10 (the only one who had a baseline ANC Ͼ100/␮l) but when with rG-CSF. Only patient 10 had a baseline ANC Ͼ100/␮l. we excluded this patient from the same analysis, our resulting Detailed information on these 15 patients in regard to WBC values did not differ significantly; they were a mean of 512/␮l transfusion therapy, response of infection, survival, and ulti- (range, 98–1072/␮l) at 1 h, a mean of 320/␮l (range, 50– mate outcome of the infection and the malignancy is shown 932/␮l) at 24 h, and a mean of 196/␮l (range, 0–750/␮l) at in Table 3. Table 4 shows the relationship between outcomes 48 h. Figure 1 shows that there was a strong correlation of infection and recovery from neutropenia observed in the between the donors’ ANCs and the yields of neutrophils per 15 patients. Favorable outcomes were seen in 11 patients at leukapheresis (r =+0.9). the end of therapy. Eight of those patients continued to have Patients with favorable and unfavorable responses at the a favorable outcome 3 weeks later, whereas two patients end of therapy had received a comparable number of total deteriorated (and considered as treatment failures) and one neutrophils per body surface area (mean, 168 × 109 and died of infection. The two patients with treatment failure had 202 × 109/transfusion, respectively; P = 0.6) (data not shown). suffered relapse of leukemia (patient 10) or persistent of neu- The 1-h post-transfusion ANCs of both groups of recipients tropenia and BMT-induced graft-versus-host disease (GVHD) were also comparable in both groups (mean, 636/␮l for (patient 13), whereas death was caused partially by the infec- responders and 480/␮l for nonresponders; P = 0.6). G-CSF-stimulated leukocyte transfusions MC Dignani et al 1624 Table 2 Clinico-pathologic features of the 15 assessable patients at study entry

Patient Age Underlying Infection Laboratory Extent of Days of Days on Z No. (years) disease and diagnosis infection Antifungal treatment neutropenia rG-CSF Sex therapy at SE Days Total mg ANC ANC AMB ABCD AMB ABCD Ͻ500/␮l Ͻ100/␮l ABLC ABLC

1 42 HD (relapsed) Necrotic Skin biopsy 6 × 6cm4—300—1615143 F auto BMT shoulder path and cx D + 14 cellulitis by A. flavus 2 28 AML (relapsed) Sinusitis by Sinus biopsy All sinuses, 108 23 3600 5000 33 10 12 2 F allo BMT A. flavus path and cx nasal cavity, D + 13 nasal cx bone invasion, facial cellulitis 3 18 ALL (relapsed) Pneumonia by BAL cx Lungs: 1 lobe 15 — 900 — 23 22 20 1 M C1 D29 chemo A. terreus sputum cx 4a 32 CML lymphoid Pneumonia Deep ETS cx Lungs: bilateral, 21 — 880 — 35 31 2 3 M BC (refractory) by A. hypoxemia. allo BMT fumigatus (MUD) D + 2 5 73 AML C1 D16 Sinusitis + Deep ETS Lungs: bilateral, 9 — 360 — 13 10 15 4 F chemo pneumonia cx × 3 hypoxemia. by A. Sinuses: fumigatus and bilateral A. terreus ethmoid and sphenoid 6 64 AML C1 D28 Cellulitis by Toenail cx Onychomycoses, 8 — 390 — 21 21 8 2 M chemo A. terreus toe/foot cellulitis 7a 26 HD (relapsed) Disseminated Skin biopsy Lungs: bilateral 9 — 368 — 39 38 8 3 F auto BMT aspergillosis path and cx hypoxemia. D + 40 by A. flavus Skin 8 45 Large cell Disseminated Blood cx × 3 Onychomycoses, 5 — 290 – 9 8 12 1 F lymphoma fusariosis by toe/foot (refractory) F. solani cellulitis, skin, C1 D16 chemo spleen, blood 9 26 AML Disseminated Blood cx × 4 Sinuses: 10 — 575 — 24 14 1 3 M C1 D20 chemo fusariosis by Skin biopsy bilateral max, F. proliferatum path R frontal. Blood, skin 10 23 ALL (4th Disseminated Blood cx × 5 Lungs, skin, 6 4 310 800 58 0 2 3 F relapse) fusariosis by Skin biopsy blood, L eye, C1 D21 chemo F. solani path and cx sinuses: R max Vitrectomy path 11 22 AML (relapsed) Sino- Sinus biopsy Sinuses: L max, — 5 — 1800 16 14 4 3 M C2 D26 chemo pulmonary path L ethmoid, L mold infection sphenoid, bone invasion, facial cellulitis Pneumonia; RUL, RML 12a 42 ALL (relapsed) Disseminated Blood cx × 6 Blood. Lungs: 2 2 154 760 13 13 22 4 M C1 D21 chemo candidiasis Skin bx path bilateral. Skin by C. krusei and cx 13 41 CML Disseminated Blood cx × 6 Blood. Lungs: 1 7 33 1848 25 21 25 4 M (accelerated candidiasis bilateral, phase) by C. hypoxemia allo BMT D + 24 glabrata 14 23 AML (relapsed) Disseminated Blood Blood. Lungs: 1 2 2 930 984 52 50 52 4 M auto BMT candidiasis cx × 10 CSF lobe hypoxemia. D + 51 by C. krusei cx meninges 15 30 Lymphoblastic Disseminated Blood cx × 8 Blood, L 29 13 1155 5265 12 11 0 4 M lymphoma (PR) infection by mastoides, allo BMT + 6 Trichosporon middle and cutaneum external L ear, L facial cellulitis

aPatients 4, 7, and 12 had received itraconazole for 9, 3 and 12 days, respectively. AMB, amphotericin B; ABCD, amphotericin B colloidal dispersion; ABLC, amphotericin B lipid complex; ANC, absolute neutrophil count; SE, study entry; HD, Hodgkin’s disease; AML, acute myelogenous leukemia; ALL, acute lymphocytic leukemia; CML, chronic myelocytic leukemia; BC, blast crisis; PR, partial remission; mo, months; MUD, matched unrelated donor; CLL, chronic lymphocytic leukemia; auto, autologous; BMT, bone marrow transplantation; chemo, chemotherapy; Z, Zubrod (performance status score); LUL, left upper lobe; LLL, left lower lobe; RML, right middle lobe; C, course; D, day; path, pathology; cx, culture; ETS, endotracheal secretions; CNS, central nervous system; rG-CSF, recombinant granulocyte colony-stimulating factor; L, left; R, right; A, aspergillosis; C, candida; F, fusarium; CSF, cerebrospinal ﬂuid. G-CSF-stimulated leukocyte transfusions MC Dignani et al 1625 Table 3 Outcome of infections treated with white blood cell transfusions in 15 assessable patients

Patient No. Mean yield Mean ANC/␮l Reason Infection response Survival Ultimate outcome No. WBCT × of PMN/ after WBCTx to D/C to WBC transfusion in WBCTx ×109 WBCTx therapy days (total dose) after 1 h 24 h At end of 3 weeks study therapy after therapy entry

1 10 31 711 261 DE Improving Death 12 Died neutropenic on day +26 BMT. Cause (286) unrelated to of death: MOF secondary to DIC, VOD, infection and nonbacterial thrombotic endocarditis. No evidence of aspergillosis at autopsy. 2 5 41 968 932 R Stable Complete 390+ Sinusitis recurred while patient was on (517) (underwent response ABCD (total AMB dose 27 g) 80 days surgival with residua after the last WBC transfusion. Relapse debridement) likely secondary to leukopenia following gancyclovir. Responded promptly to rG- CSF. She underwent four subsequent debridements, and one reconstructive surgery b/o oral-nasal ﬁstula. Alive and free of disease 13 months after study entry. 3 5 46 618 455 ND Improving Complete 35 Aspergillosis recurred following allogeneic (182) (while AMB response BMT and neutropenia 24 days after the on hold) last WBC transfusion. He died neutropenic on day +6 of allogeneic BMT. Death likely due to aspergillosis. No autopsy was granted. 4 11 35 622 390 ND Improving Death 26 Developed grade IV GVHD of liver, gut (387) partially and skin. ATG started and MP increased. related to Pneumonia worsened 9 days after last the infection WBC transfusion. Died non-neutropenic of MOF. No autopsy granted. 5 9 63 521 344 R Worsening Death 14 Died of respiratory and renal failure 5 (568) related to days after recovery from neutropenia. No the infection autopsy granted. 6 14 39 378 215 I Improving Complete 67+ Remains in leukemia relapse and (501) response neutropenic (ANC Ͻ100/␮l) on day 96 of course 1 with no evidence of active infection 55 days after the last WBC transfusion. 7 7 38 914 657 DE Worsening Death 9 Died neutropenic on day +49 after (226) related to autologous BMT. Cause of death: the infection disseminated Aspergillus ﬂavus infection. At autopsy fungal infarcts in CNS and lungs. Also BOOP, and renal and splenic infarcts (cx−). 8 8 56 1072 480 ND Improving Complete 71 The patient underwent allogeneic BMT 46 (446) response days after the last WBC transfusion. with residua Disseminated fusariosis recurred after chemotherapy induced neutropenia and after transplantation despite continuous AMB therapy and caused death (autopsy proven) on day +17 after allogeneic BMT. 9 6 27 226 100 T Improving Complete 83 Leukemia recurred 37 days after the last (133) response WBC transfusion. Post course 2, on day with residua 21 of neutropenia and while on AMB he presented with new skin lesions and blood cultures positive for Fusarium spp. He died on course 2 day 31 in aplasia with disseminated fusariosis and C. krusei infection. G-CSF-stimulated leukocyte transfusions MC Dignani et al 1626 Table 3 Continued

10 9 47 1742 1475 PM Improving Failure 38 The patient died with (378) refractory leukemia, worsening of mental status, and without neutropenia 29 days after the last WBC transfusion. CT scan of brain: 1-cm lesion consistent with bleeding or abscess. All fungal cultures remained negative. No autopsy was granted. 11 3 16 150 167 ND Stable Partial 150+ Underwent surgical (49) (cellulitis response debridement. Alive and improved) free of infection 146 days after the last WBC transfusion. 12 9 41 283 79 F Improved Death 14 Died neutropenic of (320) unrelated to MOF 4 days after the the infection last WBC transfusion. Autopsy ﬁnding: extensive lymphoma. Blood and brain cultures positive for multiresistant Pseudomona aeruginosa.No evidence of infection by candidiasis. 13 5 36 229 50 ND Improving Failure 45 Died while on AMB on (142) day +34 after allogeneic BMT with MOF, VOD, neutropenia and persistent fungemia 5 days after the last WBC transfusion. Autopsy not granted. 14 16 41 387 253 F Worsening Death 29 Died on day +81 in BM, (611) partially respiratory, and CNS related to failure 11 days after the the infection last WBC transfusion. Premortem blood culture (−) for Candida krusei (+) for Enterobacter cloacae. No autopsy was granted. 15 7 21 98 94 R Worsening Failure 30 Died on day +37 after (127) allogeneic BMT, no neutropenic, in renal and CNS failure 24 days after the last WBC transfusion. No autopsy.

PMN/WBCTx, polymorphonuclear cells per white blood cell transfusion; ANC, absolute neutrophil count; D/C, discontinuation; BMT, bone marrow transplantation; MOF, multiorgan failure; DIC, disseminated intravascular coagulopathy; VOD, veno-occlusive disease; ABCD, amphotericin B coloidal dispersion; AMB, amphotericin B; GVHD, graft-versus-host disease; BOOP, bronchiolitis obliterans organizing pneumonia; ATG, antithymocyte globulin; BM, bone marrow; CNS, central nervous system; DE, death; R, recovery from neutropenia; ND, no more donors available; I, improvement of the infection; T, toxicity; PM, progression of malignancy; F, failure.

Adverse reactions related to the study treatment two patients with cellulitis (patients 2 and 8). This transient worsening was seen within 24 h of the first WBC transfusion WBC transfusions were well tolerated by the recipients; fewer and was followed by subsequent improvement. Four of the than 5% were associated with adverse reactions. The adverse nine patients with fungal pneumonia had progression of their reactions observed within 6 h of the administration of 130 pulmonary infiltrates during therapy, and two of these patients WBC transfusions to the 20 recipients are shown in Table 8. had a favorable outcome. Four of the five remaining patients Moderate reactions improved when patients were treated with with fungal pneumonia had a favorable outcome. Only one steroids (patient 1), diuretics (patient 6), or both (patient 4). patient tested nonreactive for cytomegalovirus (CMV) anti- Severe reactions that required discontinuation of therapy were body. Because of the potentially fatal and progressive infec- seen after the infusion of 1.5% of the WBC transfusions. tion and the lack of CMV-negative donors, this patient (patient Increase of inflammation at the site of infection was seen in 2) received WBC transfusions from CMV-positive donors. She G-CSF-stimulated leukocyte transfusions MC Dignani et al 1627 Table 4 Outcomes of infection in the 15 assessable patients

Outcomes Total No. of patients (patient No.) No. of patients who remained neutropenic (ANC Ͻ 500/␮l) (patient No.)

At end of therapy Improving 9 (1,3,4,6,8,9,10,12,13) 7 (1,3,4,6,10,12,13)b Stable 2 (2,11) 0 Worsening 4 (5,7,14,15) 2 (7,14) Favorable responsea 11 7 At 3 weeks after therapy Alive 9 Complete response 2 (3,6) 2 (3,6) Complete response with residua 3 (2,8,9) 0 Partial response 1 (11) 0 Failure 3 (10,13,15) 1 (13) Dead 6 Not caused by the infection 2 (1,12) Partially caused by the infection 2 (4,14) Caused by the infection 2 (5,7) Favorable outcome 8 2/6c 3 months follow-up Alive 3 Complete remission 1 (6) 1 (6) Complete remission with residua 2 (2,11) 0 Dead 6 Caused by the infection 6 After relapse 4 (3,8,9,10) After failure 2 (13,15) Favorable outcome 3 1 Survival in days from study entry Ͼ28 days 10 (2,3,6,8,9,10,11,13,14,15) р28 days 5 (1,4,5,7,12) aSee text for definition of favorable response and outcome. bPatients did not recover from neutropenia within 7 days of the last WBC transfusion. cOnly six patients with favorable outcomes remained alive at this time. seroconverted and received prophylactic gancyclovir. She did trophils transfused (three-fold to four-fold higher than con- not develop CMV disease. ventional WBC transfusions). Experimental and clinical Four of 35 donors complained of mild myalgias and arthral- studies have shown that the number of transfused cells is a gias that resolved with the administration of acetaminophen major determinant of the efficacy of WBC transfusion alone. The values of serum lactic dehydrogenase (LDH), alka- therapy.8,11,14,15 line phosphatase (AP), and serum alanine transferase (ALT) Although the mean yield of neutrophils per rG-CSF-elicited increased per day of collection for all donors. After the fifth WBC transfusion was about 50% lower than that obtained in day of donation, the mean values of LDH, AP, and ALT were another study by using donors with chronic myelocytic leuke- 1.8, 2.6, and 3.2 times higher than baseline, respectively (data mia donors,15 the post-transfusion ANCs in our recipients not shown). were not so different. In both studies, transfused neutrophils could be seen in the circulation for as long as 2 days after transfusions. This apparent prolongation of neutrophil survival Discussion might be explained by the following factors: (1) leukocytes harvested from rG-CSF-treated donors include neutrophil pre- This is the first prospective study of rG-CSF-elicited and cursors17 whose proliferation could be enhanced by the rG- irradiated WBC transfusions in the treatment of neutropenia- CSF treatment given to both the donors and the recipients18 related fungal infections. Our findings indicate that the pro- and (2) donor and recipient rG-CSF treatments could have cedure was well tolerated by both donors and patients and prolonged the survival of transfused WBCs by decreasing that these transfusions may be effective in severely ill patients apoptosis.19 We did not systemically assess the degree of with refractory and neutropenia-related fungal infections. The maturity and the functional activity of the collected cells, beneficial effect of these transfusions seemed to be enhanced although, from previous experience at our institution, we did by their administration to patients with good performance find CD34+ cells in some of the WBC concentrates obtained status and early during the period of neutropenia and soon from subjects not included in this study (unpublished data). after the onset of the fungal infection. An earlier study with The status of alloimmunization of the recipients has been conventional WBC transfusions also suggested that the early shown to have an effect on the outcome of infection and on administration of WBC transfusions may be an important pre- the adverse reactions related to this therapy.20 While we did dictor of outcome.3 Transfused neutrophils remained in circu- not assess the alloimmunization status of our patients, all were lation in levels greater than 500/␮l and 200/␮l for 24 and 48 h at risk of being alloimmunized because of prior transfusions. after their infusions, respectively. The fact that most of our donors were related and shared more The beneficial effect of WBC transfusion therapy in this than two HLA antigens could have also contributed to the study could be partially explained by the high number of neu- high response rate obtained in our study. G-CSF-stimulated leukocyte transfusions MC Dignani et al 1628 Table 5 Variables other than WBC transfusion therapy thought to Table 6 Characteristics of the 35 donor/patient pairsa have influenced response at the end of therapy Variable No. of donor/patient No. of WBC Patients with Patients with pairs (%) transfusions (%) favorable response unfavorable given by each (n = 11) response donor category (n = 4) ABO compatibility Variables at study entry Compatible 21 (60) 72 (58) Mean age (range) 35 (18–64) 38 (23–73) Incompatible 14 (40) 52 (42) Sex, M/F 7/4 2/2 HLA antigen sharing Median duration of 15 (8–31) 25 (10–50) 4 antigens 6 (17) 20 (16) profound neutropenia, 3 antigens 3 (8) 6 (5) days (range)a 2 antigens 16 (46) 63 (51) Median duration of the 8 (2–36) 17 (6–30) 1 antigen 1 (3) 4 (3) fungal infection days 0 antigen 2 (6) 3 (2) (range)b Unknown 7 (20) 28 (23) Median cumulative dose 900 (290–8600) 1200 (360–6400) Donor relationship to of AMB, mg (range) patient Median duration of rG- 12 (1–25) 15 (8–52) Parent 8 (23) 29 (24) CSF therapy, days Sibling 13 (37) 42 (34) (range)c Offspring 4 (11) 22 (19) Median Zubrod 3 (1–4) 4 (3–4) Nephew 3 (8) 12 (10) performance status Aunt/uncle 3 (9) 6 (5) grade (range)d Cousin 2 (6) 8 (4) No. of patients with 7 4 Spouse 2 (6) 5 (4) grade 3 or 4 No. of patients with 4 0 aOnly the 15 assessable patients are included. The total number of grade 1 or 2 WBC transfusions given was 124. Post-treatment variables ABO, blood type; HLA, human lymphocyte antigen. Median additional dose 900 (0–4316) 740 (280–4446) of AMB received during WBC transfusion therapy, mg (range) Table 7 WBC transfusions and their effects on 15 assessable No. of patients that 4 (36) 2 (50) recipients recovered from neutropeniae (%) Total No. of WBC transfusions 124 Total No. of donors 35 aAbsolute neutrophil count less than 100/␮l. Median No. of WBC 8 (3–16) bDefined as days since the first sign of fungal infection became transfusions/recipient (range) evident (fever excluded) until study entry. WBC concentrates, mean (range) × 9 cOnly patient 15 did not receive rG-CSF. WBC 55 10 (14–131) × 9 dP value of 0.09. ANC 41 10 (10–116) × 9 eAbsolute neutrophil count Ͼ500/␮l within 7 days of the last WBC Mononuclear cells 14 10 (4–15) × 11 transfusion. Platelets 1.1 10 (0.5–2.7) AMB, amphotericin B, all formulations of AMB were considered Hematocrit 7% (3–14) together; rG-CSF, recombinant granulocyte colony-stimulating Volume 277 ml (150–440) factor. ANC after WBC transfusions, mean (range) 1 h 594 (98–1472) 24 h 396 (50–1475) The four responders in our study who recovered from their 48 ha 231 (0–750) neutropenia would probably not have survived without the support of the WBC transfusion therapy, because they had aThis is the median of 26 values obtained in 14 patients. rapidly progressive and severe infections at study entry. ANC, absolute neutrophil count; WBC, white blood cells. Failures of WBC transfusion therapy were probably caused by the advanced stage of the infection at study entry. The four patients who did not benefit from the WBC transfusions had toms related to the rG-CSF treatment. The incidence was the worst performance status (three of them were on venti- about half that reported in patients treated with rG-CSF after lators at study entry) and had been profoundly neutropenic chemotherapy.16 The rises in alkaline phosphatase and lactate with active fungal infection for much longer than the patients dehydrogenase levels among the donors were expected and who did benefit from WBC transfusions. comparable with the levels observed in patients on day 6 of Six patients had an initially favorable response to the WBC treatment with rG-CSF,21 and may have been related to an transfusion therapy but failed at 21 days and/or 3 months after increased turnover in WBCs. Uric acid serum levels were not the last WBC transfusion. In five of these patients, the pro- measured, but no donor had any clinical manifestation gression or reactivation of their infections occurred after the attributable to hyperuricemia. The rise in alanine amino- discontinuation of the transfusions and was related to greater transferase has not previously been reported as a side-effect impairment of their immune system. It is possible that some related to treatment with either rG-CSF or HES. of these patients could have responded favorably, had they There are several limitations to our study, including the received WBC transfusions during the period of subsequent small sample size and the lack of controls. A selection bias, neutropenia. particularly one based on the availability of related donors, Eleven percent of the donors had musculoskeletal symp- could have also occurred. G-CSF-stimulated leukocyte transfusions MC Dignani et al 1629 persistent neutropenia could benefit from rG-CSF-elicited and irradiated WBC transfusions; however, transfusion therapy is a temporary measure to be used only in patients who are likely to recover from neutropenia and to achieve remission of their underlying malignancies. In such patients, transfusions should be started soon after the onset of the infection. This new promising and safe modality of therapy deserves further investigation in larger controlled clinical trials.

References

1 Anaissie E, Kontoyiannis DP, Vartivarian S et al. Effectiveness of an oral triazole for opportunistic mold infections in cancer patients: experience with SCH 39304. Clin Infect Dis 1993; 17: 1022– 1031. 2 Koza I, Holland J, Cohen E. Histocompatible leukocyte transfusion Figure 1 Mean values of donor’s absolute neutrophil counts (ANC) during granulocytopenia. Neoplasma 1971; 18: 185–191. and leukapheresed cells during recombinant granulocyte-colony sti- 3 Graw RG, Herzig G, Perry S, Henderson E. Normal granulocyte mulating factor treatment. Data from 35 donors. transfusion therapy: treatment of septicemia due to gram-negative bacteria. New Engl J Med 1972; 287: 367–371. 4 McCredie K, Freireich E, Hester J, Vallejos C. Leukocyte transfusion therapy for patients with host-defense failure. Transplant Table 8 Adverse reactions observed within 6 h of the adminis- Proc 1973; V: 1285–1289. tration of 130 rG-CSF-elicited WBC transfusions to 20 recipients with 5 Alavi JB, Root RK, Djerassi I et al. Randomized clinical trial of definite fungal infections granulocyte transfusions for infection in acute leukemia. New Engl J Med 1977; 296: 701–705. Adverse reactions No. of WBC No. of patients (%) 6 Higby DJ, Yates JW, Henderson ES, Holland JF. Filtration leukaph- transfusions (%) eresis for granulocyte transfusion therapy: clinical and laboratory studies. New Engl J Med 1975; 292: 761–766. Mild: transient SOB 1 1 7 Fortuny IE, Bloomfield CD, Hadlock DC, Goldman A, Kennedy Moderate: SOB, 3 3 BJ, McCullough JJ. Granulocyte transfusion: a controlled study in hypoxia, CXR patients with acute nonlymphocytic leukemia. Transfusion 1975; changesa 15: 548–558. Severe: fever, SOB, 2 2 8 Higby DJ, Freeman A, Henderson ES, Sinks L, Cohen E. Granulo- bronchospasm, cyte transfusions in children using filter-collected cells. Cancer hypoxiemia, CXR 1976; 38: 1407–1413. changesb 9 Herzig RH, Herzig GP, Graw RG, Bull MI, Ray KK. Successful granulocyte transfusion therapy for gram-negative septicemia. Total 6 (4.6) 6 (30) New Engl J Med 1977; 296: 701–705. 10 Vogler WR, Winton EF. A controlled study of the efficacy of gra- aIn two patients, the CXR was consistent with pulmonary edema or nulocyte transfusions in patients with neutropenia. Am J Med hemorrhage, and in one with increase of the pre-existent pneu- 1977; 63: 548–555. omonia. 11 Aisner J, Schiffer C, Wiernik P. Granulocyte transfusions: evalu- bIn one patient, the CXR showed a new basal atelectasia, whereas ation of factors influencing results and a comparison of filtration in the other it showed an increase of interstitial markings in the cen- and intermittent centrifugation leukapheresis. Br J Haematol 1978; tral portions of each lung. 38: 121–129. SOB, shortness of breath; CXR, chest X-ray. 12 Winston DJ, Ho WG, Gale RP. Therapeutic granulocyte transfusions for documented infections: a controlled trial in ninety-five infectious granulocytopenic episodes. Ann Intern Med 1982; 97: 509–515. The response rate obtained in our patients was better than 13 Bhatia S, McCullough J, Perry EH, Clay M, Ramsay NKC, Neglia that reported in historical controls who did not receive WBC JP. Granulocyte transfusions: efficacy in treating fungal infections transfusion therapy and than that observed in patients who in neutropenic patients following bone marrow transplantation. received conventional WBC transfusions.13 In the setting of Transfusion 1994; 34: 226–232. profound and persistent neutropenia, the reported mortality 14 Applebaum F, Bowles C, Makuch R, Deisseroth A. Granulocyte transfusion therapy of experimental Pseudomonas septicemia: rate of invasive aspergillosis, fusariosis, and candidiasis is study of cell dose and collection technique. Blood 1978: 52: у90%.22–24 In cases of recovery from neutropenia, these 323–331. values decrease to 40% for aspergillosis, 68% for fusariosis, 15 Freireich EJ, Levin RH, Whang J, Carbone PP, Bronson W, Morse and 62% for candidiasis.22–24 In our study, most patients EE. The function and fate of transfused leukocytes from donors remained neutropenic for a minimum of 7 days after the last with chronic myelocytic leukemia in leukopenic recipients. Ann WBC transfusion and had an overall response rate comparable NY Acad Sci 1965; 113: 1081–1087. 16 Hollingshead LM, Goa KL. Recombinant granulocyte colony-sti- with that reported among patients with similar infections who mulating factor: a review of its pharmacological properties and did recover from neutropenia. prospective role in neutropenic conditions. Drugs 1991; 42: The WBC transfusions were well tolerated by the recipients. 300–330. The incidence of adverse reactions observed in our study 17 Bensinger WI, Price TH, Dale DC et al. The effects of daily recom- compared favorably with that reported with conventional binant human granulocyte-colony stimulating factor adminis- WBC transfusions.2,9,13 We attribute this result to the use of tration on normal granulocyte donors undergoing leukapheresis. Blood 1993; 81: 1883–1888. routine premedication and the careful selection of related 18 Lieschke GJ, Burgess AW. Granulocyte-colony stimulating factor donors. and granulocyte–macrophage colony stimulating factor (first of In conclusion, patients with refractory fungal infections and two parts). New Engl J Med 1992; 327: 28–35. G-CSF-stimulated leukocyte transfusions MC Dignani et al 1630 19 Colotta F, Re F, Polentarutti N, Sozzani S, Mantovani A. Modu- cavitation and massive hemoptysis in invasive pulmonary asper- lation of granulocyte survival and programmed cell death by cyto- gillosis. Am Rev Respir Dis 1985; 131: 115–120. kines and bacterial products. Blood 1992; 80: 2012–2020. 23 Nelson P, Dignani M, Anaissie E. Taxonomy, biology, and clinical 20 Schiffer CA. Transfusion support. Hematol Oncol Ann 1994; 2: aspects of Fusarium species. Clin Microbiol Rev 1994; 7: 479– 277–279. 504. 21 Gabrilove JL, Jakubowski A, Fain K et al. Phase I study of granulo- 24 Maksymiuk AW, Thongprasert S, Hopfer R, Luna M, Fainstein V, cyte-colony stimulating factor in patients with transitional cell car- Bodey GP. Systemic candidiasis in cancer patients. Am J Med cinoma of the urothelium. J Clin Invest 1988; 82: 1454–1461. 1984; 77: 20–27. 22 Albelda S, Talbot G, Gerson S, Miller W, Cassileth P. Pulmonary