Bone Marrow Transplantation (2002) 29, 373–377 2002 Nature Publishing Group All rights reserved 0268–3369/02 $25.00 www.nature.com/bmt Graft-versus-host disease A phase I/II double-blind, placebo-controlled study of recombinant human interleukin-11 for mucositis and acute GVHD prevention in allogeneic stem cell transplantation JH Antin1, SJ Lee1, D Neuberg1, E Alyea1, RJ Soiffer1, S Sonis1 and JLM Ferrara1,2 1Department of Adult Oncology and Biostatistical Science, Dana-Farber Cancer Institute, Departments of Medicine and Surgery, Brigham and Women’s Hospital, Boston, MA, USA Summary: Keywords: interleukin-11; graft-versus-host disease; stem cell transplantation Interleukin-11 (IL-11) decreases cytokine release and increases survival in murine BMT models. In these sys- tems, it reduces gut permeability, partially polarizes T cells to a Th2 phenotype, down-regulates IL-12, pre- GVHD is a major limitation in the successful use of stem vents mucositis, and accelerates recovery of oral and cell transplantation for hematologic diseases. Increasing bowel mucosa. We conducted a randomized double- evidence supports the concept that inflammatory cytokines blind pilot study of rhIL-11 administered with are critical in the initiation and maintenance of acute graft- cyclosporine/MTX prophylaxis after cytoxan/TBI con- versus-host disease (GVHD). One contributor to the induc- ditioning and allogeneic stem cell transplantation for tion of GVHD is mucosal injury to the digestive tract that hematologic malignancies. Patients received rhIL-11, 50 can result in leakage of endotoxin into the body. Naive T ␮g/kg subcutaneously daily or placebo in a 3:1 ratio. cells from the donor are introduced into a milieu that fosters Treatment was administered prior to the start of con- inflammation, where they are activated by cytokines, and ditioning and continued up to 21 days. The study was where they recognize minor histocompatibility antigens in designed to assess safety with stopping rules for cardiac conjunction with HLA on recipient antigen presenting cells. arrhythmias and mortality. Although projected to The resulting expansion of allo-reactive T cells is facilitated accrue 20 patients, only 13 patients (10 IL-11, three by the pro-inflammatory milieu. End organ damage then placebo) were enrolled because the early stopping rule results from the combined effects of primary cytokine- for mortality was triggered. Of 10 evaluable patients induced injury and direct cellular cytotoxicity. Endotoxin who received IL-11, four died by day 40 and one died on (LPS) contributes to this effect by enhancing the production day 85. Deaths were attributable to transplant-related of these inflammatory mediators. The cytokines that have toxicity. One of three placebo recipients died of suicide, been previously implicated include TNF-␣, IL-1, IL-12, the other two are alive. Patients receiving IL-11 had sev- IFN-␥.1–12 Finally, cytokines and their inhibitors affect the ere fluid retention and early mortality, making it polarization of T cells. Th1 cells that produce IL-2 and impossible to determine whether IL-11 given in this IFN-␥ tend to be associated with GVHD, while Th2 cell schedule can reduce the rate of GVHD. Grade B–D predominance tends to diminish the risk of GVHD. Thus, acute GVHD occurred in two of eight evaluable patients manipulations that foster Th2 polarization may reduce the on IL-11 and one of three patients on placebo. The pri- likelihood of acute GVHD.13 mary adverse events of the study were severe fluid Transplant-related morbidity and mortality are substan- retention resistant to diuresis (average weight gain 9 ± tially linked to several inter-related factors that could be 4%) and multiorgan failure in five of 10 evaluable directly ameliorated by IL-11. If IL-11 can result in patients. The use of IL-11 as GVHD prophylaxis in allo- reductions in gut permeability, polarization of T cells to geneic transplantation cannot be recommended as the Th2 phenotype, and down-regulation of IL-12, it may administered in this trial. substantially ameliorate GVHD. Mucositis is a serious Bone Marrow Transplantation (2002) 29, 373–377. DOI: transplant-related morbidity. More than 90% of patients 10.1038/sj/bmt/1703394 develop grade 4 mucositis as manifested by severe erosive mucosal damage and the requirement for parenteral nar- cotics. It is intrinsically valuable to reduce mucositis for Correspondence: Dr JH Antin, Dana-Farber Cancer Institute, 44 Binney patient comfort; however, it may contribute to a reduction St, Boston, MA 02115, USA 2 in GVHD directly by improving barrier function and reduce Current address: Bone Marrow Transplant Program, University of Michi- the risk of intestinal hemorrhage and infection by gan Medical Center, 1500 East Medical Center Dr, Ann Arbor, MI 14–17 48109, USA maintaining intestinal integrity. Received 8 October 2001; accepted 20 November 2001 To determine whether these effects would translate into IL-11 for GVHD prevention JH Antin et al 374 improved outcome after human transplantation, we conduc- phamide infusion or saline placebo given on the same ted a randomized, double-blind pilot study of rhIL-11 schedule. The dose is the standard dose recommended by administered with standard cyclosporine/MTX prophylaxis the manufacturer. It was administered before the start of after cytoxan/TBI conditioning for hematologic malig- conditioning based on data from animal models.14,15 The nancies. IL-11 is a clear, colorless solution with no infusional tox- icities making it indistinguishable from placebo. All patients were treated in HEPA environments in reverse iso- Patients, materials, and methods lation using gut decontamination and acyclovir prophylaxis. Upon engraftment, all patients received P. carinii prophy- The trial was a double blind, randomized, placebo- laxis and CMV prophylaxis if clinically indicated. controlled study of marrow transplantation in patients with hematologic malignancies. Patients were eligible if they Graft-versus-host disease and toxicity evaluation were more than 18 years of age with satisfactory renal and hepatic function, had a histocompatible or one antigen mis- Acute GVHD was graded and scored according to the con- matched family member or unrelated donor. Compatibility sensus grading system19 and IBMTR grading scale, was determined serologically for class I and using standard although the primary endpoint was the IBMTR scale.18 molecular techniques for class II. A single unrelated donor Mucositis was independently evaluated daily by the BMT was mismatched at HLA-C; all other patients were fully attending physician and by the dental service every 3–4 histocompatible. Enrollment took place between February days. Mucositis was scored according to the NCI stan- and September 1999. All patients received bone marrow dardized template: grade 0 = clear, no lesions; grade 1 = stem cells. Patient characteristics are shown in Table 1. The erythema at one site; grade 2 = erythema, one ulcerative study was designed as a pilot, safety trial with stopping site, and pain; grade 3 = two or more ulcerative sites; grade rules for refractory arrhythmias and early mortality by day 4 = three or more ulcerative sites; grade 5 = ulcers covering 40. The secondary objectives were to assess the effects of the entire mouth. Mucositis scores for both ulceration and rhIL-11 on oral mucositis, grade B–D acute GVHD,18 time erythema for the same eight anatomical sites in each patient to neutrophil recovery to Ͼ500 PMN/ml (defined as the were totalled for each day of assessment in order to obtain first of 3 consecutive days of an absolute neutrophil count ‘total daily erythema scores’ and ‘total daily ulceration of у500 cells/␮l), number of platelet transfusions and time scores’ per patient. The total daily ulceration scores and to transfusion independence, and survival at 100 days after total daily erythema scores were then combined to obtain the transplant. Voluntary informed consent was obtained on a ‘total daily mucositis score’ for each patient for each day all patients. The institutional review boards of Dana-Farber of assessment. The ‘average daily mucositis score per Cancer Institute approved the study. patient’ was obtained by totalling each ‘total daily mucos- All patients received cyclophosphamide 1800 mg/m2 on itis score’ for each patient, and dividing by the number of 2 consecutive days followed by total body irradiation, 14 days that the patient was assessed.20 Chronic GVHD was Gy in eight fractions. Cyclosporine was administered graded according to established criteria.21 beginning on day −3 at 2.5 mg/kg intravenously every 12 h. 2 + Methotrexate was given at a dose of 15 mg/m on day 1 Statistical considerations and at a dose of 10 mg/m2 on days +3, +6 and +11. Study drug consisted of either recombinant human IL-11 The trial was a double blind, randomized, placebo-con- (Oprelvekin, provided by Genetics Institute, Boston, MA, trolled study of patients with hematologic malignancies. USA) administered subcutaneously at 50 ␮g/kg per day for The study was initially targeted toward patients with myel- 21 days beginning the evening before the first cyclophos- odysplasia, but generalized to all hematologic malignancies to enhance accrual. The study was designed as a pilot, safety trial with stopping rules for refractory arrhythmias Table 1 Patient characteristics and early mortality by day 40. The historical death rate at day 40 in a similar cohort of patients at this institution was rhIL-11 Placebo 17%. The study was stopped when the lower boundary of the 90% confidence interval exceeded 17%. The secondary Number 10 3 Median age (range)a 38 (19–57) 48, 54, 54 objectives were to assess the effects of rhIL-11 on oral Gender (M/F) 2/8 2/1 mucositis, grade B–D acute GVHD, time to neutrophil Diagnosis recovery to Ͼ500 PMN/ml, number of platelet transfusions CML 5 2 and time to transfusion independence, and survival at 100 Stable phase 4 2 Accelerated phase 1 0 days after the transplant.