Redefining Engraftment Syndrome: The New Mayo Clinic Criteria

BACKGROUND:  Engraftment syndrome (ES), characterized by noninfectious fever, rash, diarrhea, and/or respiratory symptoms during engraftment, can be associated with increased morbidity and resource utilization (e.g., IV antibiotics, hospitalization) after autologous stem cell transplant (ASCT).  ES remains a clinical diagnosis, without confirmed biomarkers or standardized treatment recommendations.  Herein, we describe the clinical, laboratory, and outcomes data from a large contemporary cohort of ASCT recipients and suggest new diagnostic criteria.  METHODS:  526 consecutive patients who underwent ASCT at Mayo Clinic from 1/1/2009-12/31/2010 were included.  Diseases included: multiple myeloma, 48%; NHL, 30%; amyloidosis, 12%; HL, 7%; and POEMS, 2%.  Mayo Clinic criteria for ES was defined by a noninfectious fever of 38.3C from 4d prior to WBC 500/ul up until ANC 500/ul and at least one of the following:  rash (erythematous with >25% trunk involvement), diarrhea (>3 loose stools/d), or pulmonary symptoms (O2 requirement, RR > 24).  ES characteristics and consequences were compared for Maiolino and Spitzer hybrid criteria (Maiolino criteria w/in 4d of ANC500 rather than 24 hours of first neutrophils) and Mayo Criteria (MC).  Pretransplant characteristics, ASCT complications, and the dynamics of count recovery were all assessed.  RESULTS:  146 (28%) had ES by MC while 48 (9%) had ES by Maiolino Criteria.  Patients with ES by MC experienced longer hospital stays and received more IV antibiotics (Table 1).  Patients with ES were more likely to have amyloidosis, a trend toward a higher pre-apheresis PB CD34+ count (p=0.053) and pre-apheresis PB WBC count (p=0.052); receive fewer days of G-CSF (P<0.0001), and undergo fewer apheresis sessions to meet collection goals (p=0.012) with more CD34+/kg collected per apheresis (p=0.012). They were also less likely to have undergone chemomobilization.  Too few patients perished within 30 days of ASCT (<1%) for meaningful statistical conclusions, although, 5 patients died within 30d, 3 of whom ES may have contributed.  Patients with ES had shorter times to ANC500 and ANC1000 (p<0.0001) as well as from WBC500 to ANC1000 (p<0.0001).  Patients on corticosteroids for other reasons had fewer episodes of ES (4% vs 29%; RR 0.133; p=0.003).  CONCLUSIONS:  Mayo ES criteria were better able to identify patients with increased resource utilization.  As the rate of ES was significantly less in patients on corticosteroids for reasons other than ES, it is possible that prophylaxis with low doses of corticosteroids in patients at high risk of ES may result in decreased IV antibiotics use and hospitalization.

	Mayo ES	No Mayo ES	p value	Maiolino ES	No Maiolino ES	p value
Days of IV antibiotics	11	8	<0.0001	10	10	0.28
Days hospitalized	6	0	<0.0001	3	2	0.69
Days before dismissal home	20	20	0.65	19	20	0.2