1 Luminex Detected Pre-Formed Donor Specific HLA Antibodies Predict Lung Allograft Failure J.D. Smith, 1 H. Newell, 1 A.J. Danskine, 1 M. Carby, 1 M.L. Rose, 1 1 Transplant Immunology, Harefield Hospital, Harefield, Middx, United Kingdom Purpose: Pre-formed HLA antibodies have historically been detected by complement dependent cytotoxicity (CDC) assays. The introduc- tion of Luminex assays to detect HLA antibodies has resulted in increased detection of sensitisation, however, the clinical impact of Luminex detected HLA antibodies is unclear. The purpose of this study has been to analyse pre-transplant Luminex detected HLA antibodies, their ability to fix human C4d and to determine the effect on lung allograft survival. Methods and Materials: Pre-transplant serum from 394 lung trans- plant patients (144 Heart-Lung, 103 Double Lung and 147 Single Lung) was retrospectively tested for the presence of HLA antibodies using Labscreen (One Lambda) and Lifematch (Tepnel) Luminex assays. All sera had previously been screened by CDC at the time of transplant. C4d deposition on HLA coated Luminex beads was also measured. Results were analysed and correlated with graft survival using Kaplan-Meier and Logrank statistics. Results: Prior to transplantation 6 patients had produced HLA antibodies by CDC, all of which were donor specific (DSA). The Luminex assays detected HLA antibodies in a further 50 patients, 16 of which were donor specific. The one year actuarial graft survival for CDC+ve DSA patients (n=6) was 0% compared to 56.3% for Luminex +ve DSA (n=16), 82.4% for Luminex non-donor specific (n=34) and 71% for Negative patients (n=), p0.0001. Ten % of the luminex positive sera (44 tested to date) fixed C4d. The one year graft survival for C4d+ve/DSA (n=4) was 0%, C4d-ve DSA+ve (n=12) 50%, C4d-ve DSA-ve (n=28) 78.6%, compared with 71% for antibody negative patients(p=0.0004). Conclusions: Detection of Luminex +ve donor specific HLA anti- bodies is a prognostic indicator for poor graft survival. The develop- ment of a technique to detect C4d fixing donor specific Luminex detected antibodies may be a useful tool to identify patients at high risk of accelerated graft failure. 2 Virtual Crossmatch Should Be Utilized in Sensitized Patients To Improve Organ Allocation J. Stehlik, 1 N. Islam, 1 D. Hurst, 1 A. Fuller, 1 J.C. Delgado, 1 L. Ellinger, 1 M.A. Movsesian, 1 A. Kfoury, 1 E.M. Gilbert, 1 F. Bader, 1 P. Fisher, 1 D.G. Renlund, 1 M.E. Hammond, 1 D. Bull, 1 A. Singhal, 1 D.D. Eckels, 11 U.T.A.H. Cardiac Transplant Program, Salt Lake City, UT Purpose: We have previously reported on the use of virtual cross- match (vXM), in which recipient HLA antibodies, identified by LabScreen PRA beads, are compared to the prospective donor HLA-type. This technique resulted in increased non-local organ utili- zation while maintaining favorable outcomes. The aim of this study was to further assess the accuracy of vXM in recipients with preformed Class I HLA antibodies. Methods and Materials: To increase time-efficiency at the time of organ allocation, crossmatch testing is frequently initiated on pre-set trays which contain sera of multiple prospective sensitized recipients awaiting transplantation. While only information for one particular patient is used for clinical purposes, we used the redundant results for our analysis. Results: There were 257 direct T-Cell AHG crossmatch tests per- formed with sera of 12 potential allograft recipients who had preformed HLA-antibodies of known Class I HLA specificities. In 145 instances the vXM was compatible and in 112 instances it was incompatible - Table. The negative predictive value (NPV) of vXM, i.e. the likelihood that a compatible vXM would correctly predict a compatible standard crossmatch, was 92%. The positive predictive value, i.e. the likelihood that an incompatible vXM would correctly identify recipients with incompatible standard crossmatch, was 79%. Table: Crossmatch results in recipients with preformed Class I HLA antibodies. T-cell AHG Crossmatch Incompatible () T-cell AHG Crossmatch Compatible (-) Virtual Crossmatch Incompatible (+) 89 23 PPV=79% Virtual Crossmatch Compatible (-) 12 133 NPV=92% Conclusions: This study confirms a high NPV of vXM and defines its PPV. High PPV makes this a clinically relevant approach as 8 out of 10 compatible organs will be identified as suitable by vXM. These findings further support the use of vXM in situations where standard direct crossmatch can not be performed due to the distant location of the donor organ. vXM should be utilized as it is likely to improve organ allocation in the highly disadvantaged group of sensitized patients. 3 Pre-Implant Patient Triage Using Patient Status with INTERMACS Patient Profiles: Can We Refine Selection Strategy for MCS? F.D. Pagani, 1 L. Warner Stevenson, 2 K.L. Ulisney, 3 V. Jeevanandam, 4 J.J. Teuteberg, 5 M.T. Camacho, 6 T. Martin, 6 K. Aaronson, 1 D.C. Naftel, 7 J.K. Kirklin, 71 University of Michigan, Ann Arbor, MI; 2 Brigham and Womens Hospital, Boston, MA; 3 National Heart Lung and Blood Institute, Bethesda, MD; 4 University of Chicago, Chicago, IL; 5 University of Pittsburgh, Pittsburgh, PA; 6 Newark Beth Israel Medical Center, Newark, NJ; 7 University of Alabama at Birmingham, Birmingham, AL Purpose: Implantable MCS is an established therapy for severe refractory heart failure. Seven patient profiles (PtP) reflecting acuity and severity of pre-implant illness status were created for the INTER- MACS registry to refine pt selection to maximize MCS outcomes. We hypothesized that PtP would simplify and improve assessment of pt risk for MCS above the use of measured hemodynamic (hemo) and laboratory variables. Methods and Materials: A total of 261 pts (74 sites) were entered from 6/23/2006 to 8/31/2007. Pre-implant PtP assignment was based upon the degree of inotropics and hemo stability. Ventricular arrhyth- mias were also assigned a modifier. Results: Most pts were assigned to the inotrope-dependent profile levels 1(44%); 2(38%); 3(8%); with 10% assigned to levels 4-7. Age (table), inotropes, creatinine and bi-VAD implant differed among the PtP. Low LVEF did not differ, possibly due to high inotropes. Fewer pts were NYHA Class IV in levels 5-7. There were no significant differences in pre-implant ventricular arrhythmias between levels although a significantly greater proportion of pts had an ICD at MCS implantation in levels 2-7 (compared to level 1; p=0.001). Actuarial survival for levels 2-7 was improved compared to level 1, but these differences were not significant (p=0.11). However, transplant free survival did differ among the PtP. Conclusions: Assignment of the PtP based upon the pre-implant status of the MCS pt significantly transcends traditional hemo descrip- tors. Early results identified a strong trend toward improved survival for elective implants in more stable pts (levels 2-7). Further develop- S60 ABSTRACTS