Citations

Abstract Background: Thalidomide and its analogue lenalidomide have high response rates among patients with newly diagnosed as well as previously treated myeloma. Pomalidomide (CC4047) is the newest immunomodulatory (IMiD) agent that has shown single-agent activity in phase I studies. We report on the first Phase 2 trial of pomalidomide combined with low dose dexamethasone (Pom/dex) in patients with relapsed or refractory multiple myeloma. Methods: 37 patients (21 male and 16 female) were enrolled. Pomalidomide was given orally 2 mg daily on days 1–28 of a 28-day cycle. Dexamethasone was given orally at a dose of 40 mg daily on days 1, 8, 15 and 22 of each cycle. Response was assessed by the International Myeloma Working Group Uniform Response criteria. All patients received aspirin 325 mg daily as prophylaxis against DVT. Results: The median age was 66 years (range, 40 – 88). All patients were evaluable for response and toxicity, and all analysis were done on intent to treat basis. All patients had received prior therapy; 38% had 3 prior regimens; 35% had 2 prior regimens and 27% had one prior regimen. 76% had previous autologous stem cell transplant (ASCT) and 24% had 2 prior ASCT. 62% had previous IMiD therapy. Toxicity was mild and consisted primarily of myelosuppression. Grade 3 neutropenia occurred in 31%; grade 3 thrombocytopenia 3%; grade 3 anemia 3 %. Other grade 3/4 toxicities seen in less than 5% pts included: diarrhea, atrial fibrillation, pneumonia, dehydration and renal insufficiency. 16 % had grade 1/2 neuropathy. No grade 3 neuropathy was seen and there have been no thromboembolic events. Thirty (81%) patients are continuing study treatment. Seven patients have discontinued treatment due to: disease progression (5), died on study (1) and the medical doctor’s discretion (1). Twenty three of 37 patients (62%) achieved an objective response to therapy; including 9 (24%) with VGPR; 14 patients (38%) with PR; 6 (16%) with stable disease. Objective responses were seen in 4 of 13 patients (29%) who were refractory to lenalidomide. Conclusions: Pomalidomide plus dexamethasone (Pom/dex) is highly active and well tolerated for treatment of relapsed/refractory multiple myeloma with an objective responserate of 62%, including a 29% response rate among patients who are lenalidomiderefractory.

Background: Primary amyloidosis (AL) is characterized by multi-organ deposition of light chain derived amyloid protein, resulting in organ dysfunction of varying severity. High dose chemotherapy and autologous stem cell transplantation (ASCT) is an effective treatment for AL, but is limited in applicability due to high transplant related mortality in patients with advanced organ dysfunction. Serum LDH, an enzyme that catalyzes the conversion of lactate to pyruvate, is present in large amounts in heart, kidney, liver, and muscle and is elevated with abnormalities in one or more of the organs. It is commonly elevated in a variety of hematologic malignancies. Given the multi-organ involvement in AL, we examined the prognostic value of LDH in AL. Methods: We examined our transplant database to assess the prognostic value of LDH. Laboratory values from pre-ASCT evaluation were used for the analysis. Results: Three hundred and thirteen pts who underwent ASCT between March 1997 and March 2007 were included in this analysis. The median age at ASCT was 57.5 years (range, 33–73); and 183 pts (58.3%) were male. The median duration from diagnosis to transplant was 4.1 months (range, 1–75) and the median follow up post transplant was 28.3 months (range, 1 m to 10 years) for the 229 pts (73%) alive at the time of analysis. The median LDH was 194 IU (range, 92-441) and 138 pts (44%) had an LDH > 200. The median overall survival post-ASCT for patients with an LDH > 200 was 59.4 mos (95% CI; 37.9, 80.8) compared to not reached for the rest (Figure). The prognostic value of LDH for post transplant OS was independent of cardiac troponin (cTnT). The median OS post ASCT for patients with cTnT > 0.035 (n=208) was 37.4 months (95% CI; 17, 57.9) compared to 97.5 months (95% CI; 64.8, 130.2) for the rest. In a Cox proportional hazards model, cTnT > 0.035 was associated with a 2.5 fold risk of death (p = 0.0013) and an LDH > 200 was associated with 1.7 fold risk (p = 0.04). We then examined if LDH added to the cTnT in a prognostic model for survival post ASCT. Pts with both cTnT < 0.035 and LDH < 200 (n=133) had a median OS that was not reached compared to 71 mos (95% CI; 48.3, 94) if one factor was abnormal (n=89) and 37.4 mos (95% CI; 16.1, 58.8) if both cTnT and LDH were abnormal as above (n=45); p < 0.001 (Figure 2). Conclusions: In this study we have demonstrated the prognostic value of serum LDH in AL pts undergoing ASCT. Serum LDH appears to add to the value of troponin in prognostic staging of these patients. These results need to be examined in a larger cohort of patients undergoing different treatments. It is possible that LDH is able to globally assess abnormalities of multiple organs including heart, liver and lung. Prospective studies looking at the different LDH fractions are ongoing. Figure Figure Figure Figure

Abstract Purpose: Flavopiridol is a cyclin-dependent kinase inhibitor with preclinical activity against multiple myeloma cells in vitro and ex vivo. It can induce apoptosis in non-cycling human cell lines, including RPMI-8226, a myeloma cell line. Suggested mechanisms of cytotoxicity include down regulation of anti-apoptotic regulators, direct binding to duplex DNA, disruption of transcription factor/DNA binding, upregulation of p53, inhibition of transcription, and inhibition of angiogenesis. A Phase II multicenter trial was conducted to determine the activity of flavopiridol in patients with relapsed or refractory multiple myeloma. Experimental Design: Eighteen patients were treated with 1 hour flavopiridol infusions (50 mg/m(2)/day) for 3 consecutive days every 21 days. Responses were assessed according to IBMTR criteria each cycle. Serial bone marrow aspirates were collected before and immediately after flavopiridol treatment to measure in vivo and ex vivo effects of flavopiridol on patient plasma cells. Immunoblotting for Mcl-1, Bcl-2, p53, cyclin D, phosphoRNA polymerase II and phosphoSTAT 3 was conducted on total cellular proteins isolated from sorted plasma/myeloma cells. Ex vivo assessment of flavopiridol sensitivity of freshly collected myeloma cells was performed for 5 patients pre-therapy. Results: Median age of patients ranged from 49 to 81 years, with a median of 65 years. Patients had received a median of 3 (range, 1–5) treatment regimens prior to enrollment. Sixty-one percent of patients were refractory to prior therapy and fifty-five percent had received prior high dose therapy with hematopoietic stem cell support. The immunoglobulin isotypes of the patients were as follows: IgG (10), IgA (4), light chain (3), and non-secretory (1). At enrollment, 13 patients had a beta-2 microglobulin greater than 3.5. Seven had a plasma cell labeling index greater than or equal to 1%. Four had an elevated LDH. The trial was stopped at time of interim analysis due to a lack of clinical efficacy. Of eighteen treated patients, 15 progressed (including 1 death on study) and 3 discontinued due to toxicity. All patients have ended the active treatment phase, and the mean number of cycles administered for all enrolled patients was 1.6 cycles (range 1–3). No objective responses were observed. The most frequent adverse events were leukopenia and diarrhea (83% each), followed by thrombocytopenia, nausea, and fatigue (61% each). Ex vivo flavopiridol treatment of pre-therapy patient plasma/myeloma cells led to cytotoxicity, but only after longer exposure times at higher flavopiridol concentrations than were anticipated to be achieved in vivo. Further, immunoblotting demonstrated no indication that known in vitro cellular effects of flavopiridol were recapitulated in vivo. Conclusions: Flavopiridol has little single-agent activity in relapsed or refractory multiple myeloma when administered at this dose and schedule, which is likely due to the inability to achieve biologically relevant concentrations in patients. .

Background: High dose chemotherapy and autologous stem cell transplant (SCT) remains the preferred therapy for eligible patients with multiple myeloma (MM). Several studies have demonstrated an improvement in median survival with use of SCT. Patients with myeloma may undergo SCT immediately following 4–6 cycles of ‘induction’ therapy or at the time of relapse from the initial plateau phase. Available evidence does not suggest a survival difference between the two approaches. Methods: We retrospectively evaluated our experience with autologous single SCT for MM to compare the results of delayed SCT to early SCT. We identified from our transplant database, 202 patients with MM, who underwent SCT between October 1992 and November 2002. 101 patients, who responded to induction chemotherapy, had stem cells collected in plateau phase, received maintenance chemotherapy and had SCT at the time of first relapse (delayed SCT group). The remaining 101 patients, after initially achieving a response underwent upfront SCT (early SCT group). Patients refractory to initial therapy were excluded from this study. Most of the early transplants were done in the recent years and hence this group has a shorter follow up. Results: The study cohort had a median age of 55 years (range 29 – 72) at diagnosis consisting of 126 males (62%), with no significant demographic difference between the two groups. As expected measures of tumor burden (M protein, B2M and marrow plasma cell percentage) were significantly higher in the delayed group. Plasma cell labeling indices and presence of abnormal cytogenetics were higher in the delayed SCT group as well at transplant. TBI containing regimens were used more often in the delayed group reflecting a difference in the time periods of transplant. There was no difference between the groups in terms of overall response to transplant though complete response rate was higher for the early transplants. (Table). There was no difference in the time to overall response between the groups (P=0.13, Kaplan Meier estimate). Though the median progression free survival (PFS) from transplant was shorter for the delyaed SCT group, the overall survival (OS) from diagnosis of MM was comparable for the two groups (Table). The overall survival estimate at 5 year from diagnosis was 60% for both groups. Conclusions: Review of our experience demonstrates comparable overall survival for an early SCT compared to SCT at the time of relapse. Although there is an advantage for early SCT in terms of more chemosensitive disease as shown by a longer PFS from SCT, there is no benefit in OS from time of diagnosis. Patient preference and other co-morbidities should play a role in the decision regarding timing of transplant. This study has the disadvantage of being retrospective in nature and the two groups being spread over different periods in time. Early (n=101) Delayed (n=101) P Overall Response 100 (99%) 94 (94%) 0.07 Complete Response 43 (42%) 28 (27%) 0.04 Median PFS (From Transplant) 26.7 months 14.8 months <0.0001 Median OS (From diagnosis) 67 months 70 months 0.1

Abstract Abstract 4986 The plasma cell disorder POEMS syndrome is defined by the following major features: peripheral neuropathy (P), organomegaly (O), endocrinopathy (E), monoclonal (immunoglobulin) protein expression (M), and skin changes (S). One of the most noticeable characteristics of this disease is that virtually all patients have lambda (γ) rather than the kappa (κ) restricted clonal plasma cells. We hypothesized that this restricted expression and possibly mutated heavy and light chain genes may be linked to the pathogenesis of POEMS. To this end, we analyzed light and heavy chain gene sequences from 21 POEMS patients at Mayo Clinic Rochester. We found that 16 of these patients had dominant light chain gene sequences, and that the expression of light chain was restricted to IGLV 1 subfamily and more specifically, to the IGLV1-40 or the IGLV 1-44 subtype. In contrast, no such stereotypy was observed at the immunoglobulin heavy chain locus (IGHV), though IGHV gene expression indicated a high degree of clonality. We aligned these sequences with their corresponding germline consensus sequences in IMGT database to assess similarity and divergence. IGLV1-40 and IGLV 1-44 sequences from POEMS patients were also aligned with the sequences in the light chain library on Boston University Amyloid Database (http://pulm.bumc.bu.edu/aldb), and POEMS IGHV sequences were compared to IGHV sequences reported in the literature for multiple myeloma, MGUS (monoclonal gammopathy of unknown significance) and CLL (chronic lymphocytic leukaemia). We found that: i) both the restricted light- and the non-restricted heavy chain genes in POEMS patients varied significantly (>2%) from their germline counterparts; ii) the incidence of somatic hypermutation in IGHV was higher in POEMS than that reported for other plasma cell disorders like multiple myeloma, MGUS and CLL, iii) in IGLV, mutations occurred at non-random, clustered hotspots located in specified areas of the gene (such as the structural elements that contribute the antigen binding CDR domain); iv) mutations occurring in the CDR regions tended to be conserved (the amino acid in POEMS was similar to that in germline in its physicochemical properties), and v) the histidine residue at position 40 (H40) in IGLV1-40 and the threonine at position 38 (T38) in IGLV1-44 are POEMS-specific. These observations suggest that an antigen driven somatic hypermutation mechanism may mark a post-germinal precursor plasma cell for selection, but further transformation of this selected cell into a malignant POEMS clone occurs via a different, hitherto unknown, mechanism. They also highlight the significance of H40 (in IGLV 1-40) and T38 (in IGLV 1-44) residues, which seem to be highly POEMS-specific (occurring seldom in other plasma cell disorders), and which may bias the selected plasma cell clone towards POEMS syndrome phenotype. Disclosures Lacy: celgene: Research Funding. Gertz:celgene: Honoraria; millenium: Honoraria, Membership on an entity's Board of Directors or advisory committees. Kumar:CELGENE: Research Funding; MILLENNIUM: Research Funding; BAYER: Research Funding; GENZYME: Research Funding; NOVARTIS: Research Funding.

Abstract Abstract 587 Background: The phosphatidylinositol 3-kinase/mammalian target of rapamycin (mTOR) signal transduction pathway controls cell proliferation and survival. Everolimus is an oral agent targeting raptor mTOR (mTORC1). The trial's goal was to determine the anti-tumor activity and safety of single-agent everolimus in patients with relapsed/refractory Waldenstrom's macroglobulinemia (WM). Patients and Methods: Eligible patients had measurable disease (IgM monoclonal protein >1000 mg/dL with >10% marrow involvement or nodal masses >2 cm), a platelet count ≥75,000 × 106/L, a neutrophil count ≥1,000 × 106/L, and a creatinine and bilirubin ≤2x laboratory upper limit of normal. Patients received everolimus 10 mg PO daily and were evaluated monthly. Tumor response was assessed after cycles 2 and 6 and then every 3 cycles until progression. Results: 50 pts were treated. The median age was 63 years (range, 43-85). The overall response rate (CR+PR+MR) was 70% (95% CI: 55-82%), with a PR of 42% and 28% MR. The median duration of response and median progression-free survival (PFS) has not been reached. The estimated PFS at 6 and 12 months is 75% (95%CI: 64-89%) and 62% (95%CI: 48-80%), respectively. Grade 3 or higher related toxicities were observed in 56% of patients. The most common were hematological toxicities with cytopenias. Pulmonary toxicity occurred in 10% of patients. Dose reductions due to toxicity occurred in 52% of patients. Conclusions: Everolimus has high single-agent activity with an overall response rate of 70% and manageable toxicity in patients with relapsed WM, and offers a potential new therapeutic strategy for this patient group. Disclosures: Ghobrial: Millennium: Honoraria, Research Funding, Speakers Bureau; Celgene: Consultancy, Honoraria, Speakers Bureau; Novartis: Honoraria, Speakers Bureau. Gertz:celgene: Honoraria; millenium: Honoraria, Membership on an entity's Board of Directors or advisory committees. Richardson:Millennium Pharmaceuticals, Inc.: Consultancy, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Celgene: Consultancy, Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Johnson and Johnson: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Novartis: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; Keryx: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau; BMS: Membership on an entity's Board of Directors or advisory committees, Speakers Bureau. Treon:Millennium: Honoraria, Research Funding, Speakers Bureau; Celgene: Honoraria, Research Funding, Speakers Bureau; Genentech: Honoraria, Research Funding, Speakers Bureau. Witzig:Novartis: Research Funding.

Abstract Introduction: Mortality rates associated with stem cell transplantation are considerably higher than those of autologous stem cell transplantation for other hematologic malignancies. The efficacy of stem cell transplantation in amyloidosis largely is determined by treatment-related mortality rates (range, 11%-25%). Appropriate patient selection is essential to avoid this high mortality rate. Patients & Methods: The primary end point of the study was death from any cause between 4 days before transplantation and d+100 after transplantation. Patient characteristics are shown. Patient characteristics (N=271) Characteristic Median (range) Comment Age, y 58 (33–75) Male 159 (59%) Renal involvement 184 (68%) Cardiac involvement 147 (54%) Peripheral neuropathy 30 (11%) Hepatic involvement 28 (14%) Albumin, g/dL 2.8 (0.8–4.4) Creatinine, mg/dL 1.1 (0.6–12.0) 90% of patients had levels ≤1.7 mg/dL Alkaline phosphatase, U/L 93 (31–1,395) 25% of patients had levels >132 U/L Urinary total protein, g/d 3.35 (0.01–35.4) Percentage of plasma cells in the marrow 7 (0–78) 90% of patients had <17% plasma cells Interventricular septal thickness, mm 13 (7–25) Ejection fraction, % 65 (28–84) Time between diagnosis and stem cell transplantation, mo. 4 (0.8–75) Troponins,μg/L 0.01 (0.01–1.0) 75% of patients had levels ≤0.03 g/L 90% of patients had levels ≤0.08 g/L Results: Among 271 patients undergoing stem cell transplantation, troponin T was a powerful predictor of treatment-related mortality. Patients with troponin levels of 0.06 mg/L or higher had a day 100 all-cause mortality rate of 28%. Day-100 Survival Rate (N=271) Patients Troponin level ≥0.06 μg/L, no. of patients (n=40) Troponin level <0.06 μg/L, no. of patients (n=231) Total no. of patients *P<.001 Died before day 100* 11 16 27 Survived beyond d+100 29 215 244 Patients with troponin levels less than 0.06 mg/L had a day-100 all-cause mortality rate of 7% (P<.001). Figure Figure Despite possibly having multiple causes of death, for 7 of 11 patients with elevated troponin levels, cardiac causality could be inferred. For patients with low levels of troponin, clear cardiac causality was observed in 4 of 16. Among the 40 patients with troponin levels of 0.06 mg/L or higher, the conditioning dosage of melphalan was reduced for 31 patients (78%). Among the patients with troponin levels less than 0.06 mg/L, the dosage of melphalan was reduced for 29% of patients. Conclusion: Troponin levels should be measured in all patients before transplantation. Those with troponin levels exceeding 0.06 mg/L should be considered for less toxic therapies until the clinically optimal use of stem cell transplantation is better defined by randomized clinical trials. The chemotherapy dosage was reduced for 77% of patients in the high troponin group, but acceptable treatment-related mortality rates were not achieved. Clearly, a risk-adapted approach that modified the dosage of conditioning chemotherapy did not compensate for poor patient health or advanced cardiac involvement.

Abstract The objective of this study was to analyze molecular changes that occur at the protein level in WM and IgM-MGUS in order to identify proteins dysregulated at disease progression and design new therapies. We employed antibody protein microarrays (BD Clonetech, CA) to measure changes in the patterns of protein expression between WM, IgM-MGUS and normal lymphoplasmacytic cells as control. The antibody array is a novel technique that assays protein differences directly by hybridizing fluorescently labeled protein mixtures from cell extracts onto a glass slide spotted with 512 different monoclonal antibodies specific for human proteins. CD 19+ and CD138+ purified lymphoplasmacytic cells were obtained from cryopreserved bone marrow samples of 5 newly diagnosed patients with WM and 5 patients with IgM-MGUS. Control plasma cells were obtained from 3 pooled CD19+ and CD138+ purified normal donor bone marrow cells. To assess differential expression, the mean of the ratios of Cy5/Cy3 for each sample were analyzed using the Clontech software to calculate an internally normalized ratio. The normalized data were analyzed by the Genespring software and changes of protein expression ≥2 fold in 60% of the samples as compared to normal control were identified in both WM and MGUS groups. There were 20 proteins upregulated in WM and 21 proteins in MGUS as compared to normal control cells These included proteins from the small GTPases such as the Ras, Rho and Ran families (Rab 4, MONA, RCC1, and CDC42GAP), Kinases/signaling proteins (ROK alpha, p62 dok, JAK, 14-3-3e), cyclin dependent kinase (CDK2), fatty acid synthase (HDAC3), NFkB pathway (IKKalpha, IKBe), and histone deacetylase. The WM and IgM MGUS differed in only 3 proteins using the 2-fold threshold. The WM samples showed overexpression of the heat shock protein Hsp90, the Ras family protein CDC25C, and the chemotaxis protein p43/EMAPII as compared to the MGUS samples. This is the first proteomic study of WM and MGUS patients. It identifies novel proteins dysregulated in WM that may serve in future clinical trials as targets for therapy such as Hsp90, NFkB, HDAC and fatty acid synthase. In addition, it identifies 3 proteins that identify disease progression from MGUS to WM, which could be useful for monitoring MGUS patients and designing targeted therapies to prevent progression to WM.

Abstract Abstract 3981 Background: Patients with multiple myeloma (MM) can present with life threatening complications including lethal infections during the disease course. The mortality is high during the early period following diagnosis, mostly as a result of complications of the disease or treatment. It is not clear if we can identify patients at the highest risk of early death based on the presenting clinical and laboratory features. Methods: From among 1545 patients seen at our institution between Jan 1999 and Dec 2008, we identified 265 patients who died within 12 months of diagnosis. For each of these patients we identified two “controls” who had at least 12 months of follow up, were alive at the time of last follow up and were closest to the ‘case' patient in terms of time of diagnosis. We performed logistic regression using the clinical and laboratory features, to identify parameters that predict for 12-month mortality and to determine the best cutoffs. Results: This study included 265 patients as cases and 530 controls. The gender distribution was similar in the two groups. We examined the impact of age, performance status (PS), hemoglobin, platelets, serum creatinine, calcium, LDH, albumin, B2M, free light chain difference, plasma cell labeling index, ISS stage, risk category and exposure to novel agents upfront. Based on the results of univariate and then multivariate analysis – age> 72, ECOG PS>2, Calcium >11.3 and ISS stage 3 were found to be significant. We developed the risk score using 1 point for each adverse factor, Age> 72, PS > 2, ISS 3, and Calcium > 11.3. The odds ratio of patients dying in the first year was 2.7, 9.2 and 37 in the presence of one, two and three or more risk factors compared to none of the risk factors. We then looked at the impact of the initial therapy on outcome. 279 patients (35%) received one of the novel agents (thalidomide, lenalidomide or bortezomib) as initial therapy. The odds ratio for dying in the first year was 0.35 for patients receiving a novel agent compared to the rest. The impact of the novel agent was independent of the risk score, with the risk score predicting outcome in both groups. Conclusions: In newly diagnosed MM advanced age, poor performance status, ISS stage and high calcium were associated with early mortality. If eligible all patients should get the benefit of use of novel agent upfront. Identifying these patients at the time of diagnosis is important and should help develop better risk-adapted strategies. Disclosures: No relevant conflicts of interest to declare.

Abstract Background: POEMS syndrome is a devastating syndrome, characterized by peripheral neuropathy, organomegaly, endocrinopathy, monoclonal plasma cells, skin changes, papilledema, volume overload, sclerotic bone lesions, thrombocytosis, & high VEGF. We noted an unexpectedly high transplant related morbidity, which we have since postulated to be ES. Methods: 30 patients with POEMS were treated with PBSCT at Mayo Clinic Rochester. We retrospectively studied outcomes, with an emphasis on treatment related morbidity. Two definitions of ES were used: Spitzer (BMT 2001) and Maiolino (BMT 2003). Results: Two-thirds were male. Median age was 48, range 20–70. Time from first symptoms and diagnosis was 26 and 4 months, respectively. To mobilize stem cells, CTX/G-CSF was used in 5 & G-CSF in the remainder. Conditioning was Mel200 (n=19), Mel140 (n=10), and BEAM (n=1). Post-transplant 15 had GM-CSF begun day+6. Only 10% remained outpatient, and median time to discharge from hospital was transplant day 17 (range 0–175). Factors predicting for later dismissal included age (p=0.04), abnormal CXR 7 to 17 days post transplant (p<0.0001), & bolus corticosteroids (CS) beyond day 12 post-transplant (p=0.006). Ninety-three percent had fever, although only 9 had bacteremia. Eight satisfied criteria for ES according to Spitzer and 14 according to Maiolino. Another 3 patients received steroid bolus days 8, 12, and 11 for presumed ES but did not meet criteria because of delayed ANC recovery (days 16, 18, and 20, respectively). Of the 5 patients requiring endotracheal intubation, 3 satisfied Maiolino’s criteria for ES. Although these 3 received bolus CS during their course, administration was delayed at 18, 14, and 18 days. The patient whose CS bolus antedated intubation only received prednisone 30 mg/day. In toto, fourteen patients received bolus CS (daily doses between 20 and 1200 mg prednisone equivalents) commencing day 8 to 59. Those 7 patients who received CS before day 13 did better than the 7 who received them day 13 or later (Table). Conclusions: It is essential to recognize that nearly 50% of patients satisfied formal criteria for ES as defined by Maiolino. In these patients ES may run a self-limited course or lead to catastrophe. No Steroid (n=16) Steroid ≤ D12 (n=7) Steroid > D12 (n=7) P *Engraftment syndrome according to definition of Maiolino (M) or of Spitzer (S). ES M / S, n* 5 / 2 4 / 4 5 / 2 NA Wt change, % 0.6 (−.4.2–6.7) 6.7 (3.6–27.2) 11.2 (–2.1–23.2) 0.005 Rash, % 27 71 43 0.13 Diarrhea, % 73 86 86 0.71 Tmax, C 39 (37.8–41) 40.1 (39–41.1) 38.9 (38.7–40.8) 0.08 1st fever, day 10 (6–15) 8 (7–9) 12 (8–146) 0.007 Abn CXR1, % 13 71 71 0.03 Ventilator, % 0 14 71 0.004 First WBC, day 12 (8–21) 14 (12–14) 14 (12–17) 0.03 ANC500, day 15 (12–29) 16 (14–115) 18 (15–45) 0.08 PLT20, day 12 (8–41) 20 (11–115) 24 (9–170) 0.05 PLT50, day 15 (11–192) 32 (16–115) 56 (13–551) 0.03 RBCs, units 3 (2–8) 6 (4–31) 11 (6–64) 0.0008 PLTS, aph. units 2 (1–9) 9 (4–51) 18 (4–60) 0.0004 Hosp dsm, day 15 (13–36) 21 (15–69) 41 (16–175) 0.009