Machine learning methods to support personalized neuromusculoskeletal modelling.

Many biomedical, orthopaedic, and industrial applications are emerging that will benefit from personalized neuromusculoskeletal models. Applications include refined diagnostics, prediction of treatment trajectories for neuromusculoskeletal diseases, in silico design, development, and testing of medical implants, and human-machine interfaces to support assistive technologies. This review proposes how physics-based simulation, combined with machine learning approaches from big data, can be used to develop high-fidelity personalized representations of the human neuromusculoskeletal system. The core neuromusculoskeletal model features requiring personalization are identified, and big data/machine learning approaches for implementation are presented together with recommendations for further research.

Individuals with mild-to-moderate hip osteoarthritis walk with lower hip joint contact forces despite higher levels of muscle co-contraction compared to healthy individuals.

OBJECTIVE: To compare hip joint contact forces (HJCF), hip muscle forces, and hip muscle co-contraction levels between individuals with mild-to-moderate hip osteoarthritis (OA) and healthy controls during walking. DESIGN: Eighteen participants with mild-to-moderate hip OA and 23 healthy controls walked at a self-selected speed while motion capture and electromyographic data were synchronously collected. HJCF were computed using a calibrated electromyography-informed neuromusculoskeletal model. Hip joint contact forces, muscle forces, and co-contraction indices for flexor/extensor and adductor/abductor muscle groups were compared between groups using independent sample t-tests (P < 0.05). RESULTS: There was no between-group difference in self-selected walking speed. On average, participants with hip OA walked with 11% lower first peak (mean difference 235 [95% confidence interval (CI) 57-413] N) and 22% lower second peak (mean difference 574 [95%CI 304-844] N) HJCF compared to controls. Hip muscle forces were also significantly lower in the hip OA compared to control group at first (mean difference 224 [95%CI 66-382] N) and second (mean difference 782 [95%CI 399-1164] N) peak HJCF. Participants with hip OA exhibited higher levels of hip muscle co-contraction in both flexor/extensor and adductor/abductor muscle groups. Consistent with existing literature, hip joint angles (extension, adduction) and external moments (flexion, extension, adduction) were lower in hip OA compared to controls. CONCLUSION: Lower HJCF were detected in mild-to-moderate hip OA, primarily due to lower hip muscle force production, and despite higher levels of hip muscle co-contraction. Findings suggest that lower loading of the hip joint during walking is a feature of mild-to-moderate hip OA, which could have implications for the pathogenesis of hip OA and/or disease progression.

Hip joint moments during walking in people with hip osteoarthritis: a systematic review and meta-analysis.

OBJECTIVE: This systematic review and meta-analysis investigated surrogate measures of hip joint loading during walking in people with hip osteoarthritis (OA). METHODS: Five databases were searched up to January 14th, 2018. Studies that measured hip joint moments in the frontal or sagittal plane during walking in people with hip OA and used either a healthy control group or the unaffected leg to compare hip joint moments were included. Standardised mean differences (SMD) in sagittal and frontal plane moments were pooled as appropriate, using a random effect approach. Methodological quality was assessed using the Downs and Black checklist. RESULTS: Thirteen studies with 1,141 participants were eligible and suitable for meta-analyses. Overall, people with hip OA had lower sagittal (SMD -0.55 (95% confidence interval (CI) -1.00 to -0.10) and frontal plane moments (SMD -0.63 (95% CI -0.92, -0.34) compared to controls. However, substantial heterogeneity was observed (I2 ≤ 89%). Results by disease stage suggest that people with end-stage hip OA have lower sagittal (SMD -0.96; -1.30, -0.61; I2 = 69%) and frontal (SMD -1.17; 95% CI -1.71, -0.64; I2 = 85%) plane moments compared to controls. People with less severe hip OA than end-stage disease have comparable sagittal (SMD 0.37; 95% CI -0.17, 0.90; I2 = 69%) and frontal (SMD -0.24; 95% CI -0.76, 0.27; I2 = 51%) plane moments compared to controls. CONCLUSION: Hip joint loading may be dependent on disease stage. People with end-stage hip OA under-loaded compared to controls, while those who were not awaiting hip joint replacement had comparable hip joint loads to controls.

A systematic review and evidence synthesis of qualitative studies to identify primary care clinicians' barriers and enablers to the management of osteoarthritis.

OBJECTIVE: Primary care management of osteoarthritis (OA) is variable and often inconsistent with clinical practice guidelines (CPGs). This study aimed to identify and synthesize available qualitative evidence on primary care clinicians' views on providing recommended management of OA. DESIGN: Eligibility criteria included full reports published in peer-reviewed journals, with data collected directly from primary care clinicians using qualitative methods for collection and analysis. Five electronic databases (MEDLINE, Cochrane Central Register, EMBASE, CINAHL and PsychInfo) were searched to August 2016. Two independent reviewers identified eligible reports, conducted critical appraisal (based on Critical Appraisal Skills Programme (CASP) criteria), and extracted data. Three reviewers independently, then collaboratively, synthesized and interpreted data through an inductive and iterative process to derive new themes. The Confidence in Evidence from Reviews of Qualitative research (CERQual) approach was used to determine a confidence profile for each finding. RESULTS: Eight studies involving approximately 83 general practitioners (GPs), 24 practice nurses, 12 pharmacists and 10 physical therapists, from Australia, France, United Kingdom, Germany and Mexico were included. Four barriers were identified as themes 1) OA is not that serious, 2) Clinicians are, or perceive they are, under-prepared, 3) Personal beliefs at odds with providing recommended practice, and 4) Dissonant patient expectations. No themes were enablers. Confidence ratings were moderate or low. CONCLUSIONS: Synthesising available data revealed barriers that collectively point towards a need to address clinician knowledge gaps, and enhance clinician communication and behaviour change skills to facilitate patient adherence, enable effective conversations and manage dissonant patient expectations. REGISTRATION: PROSPERO (http://www.crd.york.ac.uk/PROSPERO) [4/11/2015, CRD42015027543].

Co-administration of low doses of intranasal PT-141, a melanocortin receptor agonist, and sildenafil to men with erectile dysfunction results in an enhanced erectile response.

OBJECTIVES: To evaluate the safety and pharmacodynamic effect of co-administration of subtherapeutic doses of PT-141, a cyclic heptapeptide melanocortin analogue, and sildenafil to patients with erectile dysfunction. METHODS: Nineteen patients with erectile dysfunction who were responders to either Viagra or Levitra by self-report were given 25 mg sildenafil and 7.5 mg intranasal PT-141, 25 mg sildenafil and an intranasal placebo spray, and a placebo tablet and an intranasal placebo spray in a randomized cross-over design. Erectile activity in response to two 30-minute episodes of visual sexual stimulation was assessed by RigiScan during a 6-hour postdose period. RESULTS: The erectile response induced by co-administration of PT-141 and sildenafil was significantly greater than the response elicited by administration of sildenafil alone. Co-administration of PT-141 and sildenafil was safe and well-tolerated and did not result in new adverse events or adverse events that were increased in frequency or severity compared with monotherapy. CONCLUSIONS: Co-administration of intranasal PT-141 and a phosphodiesterase type 5 inhibitor may constitute a treatment alternative for patients in whom higher doses of a single therapy are not effective or well tolerated.

Evaluation of the safety, pharmacokinetics and pharmacodynamic effects of subcutaneously administered PT-141, a melanocortin receptor agonist, in healthy male subjects and in patients with an inadequate response to Viagra.

PT-141, a cyclic heptapeptide melanocortin analog, was evaluated following subcutaneous administration to healthy male subjects and to patients with erectile dysfunction (ED) who report an inadequate response to Viagra. An inadequate response was defined for this study by patient report indicating that achievement of an erection suitable for vaginal penetration occurred < or =50% of the time while taking 100 mg Viagra. Erectile responses were assessed by RigiScan in healthy subjects in the absence of visual sexual stimulation (VSS) and in ED patients in the presence of VSS. Doses ranging from 0.3 to 10 mg were administered to healthy male subjects, resulting in a statistically significant erectile response at doses greater than 1.0 mg. ED patients were treated with placebo, 4 or 6 mg PT-141 in a crossover design in the presence of VSS. The erectile response induced by PT-141 was statistically significant at both doses. PT-141 was safe and well tolerated in both studies. The erectogenic potential of PT-141, its tolerability profile and its ability to cause significant erections in patients who do not have an adequate response to a PDE5 inhibitor suggest that PT-141 may provide an alternative treatment for ED with a potentially broad patient base.

Double-blind, placebo-controlled evaluation of the safety, pharmacokinetic properties and pharmacodynamic effects of intranasal PT-141, a melanocortin receptor agonist, in healthy males and patients with mild-to-moderate erectile dysfunction.

PT-141, a cyclic heptapeptide melanocortin analog, was evaluated following intranasal administration in healthy male subjects and in Viagra-responsive erectile dysfunction (ED) patients. Erectile response was assessed by RigiScan trade mark in healthy subjects without visual sexual stimulation (VSS) and in Viagra-responsive ED patients with VSS. In healthy subjects, mean C(max) and AUC((0-t)) increased in a dose-dependent manner. Median T(max) was 0.50 h and mean t(1/2) ranged from 1.85 to 2.09 h. In both studies, an erectile response induced by PT-141 administration was statistically significant, compared to placebo, at doses greater than 7 mg, with the onset of the first erection occurring in approximately 30 min. PT-141 was safely administered and well tolerated in both studies. A maximum-tolerated dose was not identified. Flushing and nausea were the most common adverse events reported in both studies and no clinically significant changes in vital signs, laboratory tests, ECGs, or physical exams were observed. Based upon its erectogenic potential and tolerability profile, PT-141 is a promising candidate for further evaluation as a treatment for male ED.

PT-141: a melanocortin agonist for the treatment of sexual dysfunction.

PT-141, a synthetic peptide analogue of alpha-MSH, is an agonist at melanocortin receptors including the MC3R and MC4R, which are expressed primarily in the central nervous system. Administration of PT-141 to rats and nonhuman primates results in penile erections. Systemic administration of PT-141 to rats activates neurons in the hypothalamus as shown by an increase in c-Fos immunoreactivity. Neurons in the same region of the central nervous system take up pseudorabies virus injected into the corpus cavernosum of the rat penis. Administration of PT-141 to normal men and to patients with erectile dysfunction resulted in a rapid dose-dependent increase in erectile activity. The results suggest that PT-141 holds promise as a new treatment for sexual dysfunction.

A human CD46 transgenic pig model system for the study of discordant xenotransplantation.

BACKGROUND: The chronic shortage in the supply of human organs available for allotransplantation has turned attention toward the use of animals as potential donors, with pigs as the most likely species under consideration. Hyperacute rejection, the initial and immediate barrier to a pig-to-primate xenograft, has been addressed by generation of transgenic pigs that express the human membrane-bound complement-regulatory proteins CD59 and/or CD55. Difficulty has been encountered in generation of transgenic animals that express a third membrane-bound complement-regulatory protein, CD46. METHODS: We have generated transgenic animals by using a large genomic construct that encompasses the entire human CD46 gene. RESULTS: We report the first description of transgenic mice and pigs that express high levels of human CD46 in a cell and tissue type-specific manner, resembling patterns of endogenous CD46 expression observed in human tissues. Furthermore, when human CD46 transgenic porcine hearts were transplanted into baboons, the grafts did not succumb to hyperacute rejection, and survival extended for up to 23 days. Under the same conditions, nontransgenic grafts underwent hyperacute rejection within 90 min. CONCLUSIONS: This is the first report to describe generation of transgenic pigs that express human CD46, and the first in vivo demonstration of the ability of human CD46 expressed on pig organs to regulate complement activation and overcome hyperacute rejection upon transplantation of a vascularized organ into nonhuman primates.

Non-depleting anti-CD4, but not anti-CD8, antibody induces long-term survival of xenogeneic and allogeneic hearts in alpha1,3-galactosyltransferase knockout (GT-Ko) mice.

The anti-galactose-alpha1,3-galactose (Gal) antibody (Ab) response following pig-to-human transplantation is vigorous and largely resistant to currently available immunosuppression. The recent generation of GT-Ko mice provides a unique opportunity to study the immunological basis of xenograft-elicited anti-Gal Ab response in vivo, and to test the efficacy of various strategies at controlling this Ab response [1]. In this study, we compared the ability of non-depleting anti-CD4 and anti-CD8 to control rejection and antibody production in GT-Ko mice following xenograft and allograft transplantation. Hearts from baby Lewis rat or C3H mice were transplanted heterotopically into GT-Ko. Non-depleting anti-CD4 (YTS177) and anti-CD8 (YTS105) Abs were used at 1 mg/mouse, and given as four doses daily from day -2 to 1 then q.o.d. till day 21. Xenograft rejection occurred at 3 to 5 days post-transplantation in untreated GT-Ko recipients, and was histologically characterized as vascular rejection. Anti-CD4, but not anti-CD8, Ab treatment prolonged xenograft survival to 68 to 74 days and inhibited anti-Gal Ab as well as xeno-Ab production. In four of the five hearts from anti-CD4 mAbs-treated GT-Ko mice, we observed classic signs of chronic rejection, namely, thickened intima in the lumen of vessels, significant IgM deposition, fibrosis and modest mononuclear cell infiltrate of Mac-1+ macrophages and scattered T cells (CD8>CD4). Xenograft rejection in untreated, as well as anti-CD4- and anti-CD8-treated, recipients was associated with increased intragraft IL-6, IFN-gamma and IL-10 mRNA. C3H allografts were rejected in 7 to 9 days by untreated GT-Ko mice and were histologically characterized as cellular rejection. Treatment with anti-CD4 and anti-CD8 mAb resulted in graft survivals of >94.8 and 11.8 days, respectively. Anti-CD4 mAb treatment resulted in a transient inhibition of alloreactive and anti-Gal Ab production. The presence of circulating alloreactive and anti-Gal Abs at >50 days post-transplant was associated with significant IgM and IgG deposition in the graft. Yet, in the anti-CD4 mAb-treated group, the allografts showed no signs of rejection at the time of sacrifice (>100 days post-transplantation). All rejected allografts had elevated levels of intragraft IL-6, IFN-gamma and IL-10 mRNA, while the long-surviving anti-CD4-treated allografts had reduced mRNA levels of these cytokines. Collectively, our studies suggest that the elicited xeno-antibody production and anti-Gal Ab production in GT-Ko mice are CD4+ T-cell dependent. The majority of xenografts succumbed to chronic rejection, while allografts survived with minimal histological change, despite elevated levels of circulating alloAbs. Thus, immunosuppression with anti-CD4 mAb therapy induces long-term survival of allografts more effectively than to xenografts.

The role of antibodies in dysfunction of pig-to-baboon pulmonary transplants.

OBJECTIVE: Pulmonary transplantation has become the preferred treatment for end-stage lung disease, but application of the procedure is limited because of a paucity of donors. One way to solve donor limitations is to use animal organs as a donor source or xenotransplantation. The current barrier to pulmonary xenotransplantation is the rapid failure of the pulmonary xenograft. Although antibodies are known to play a role in heart and kidney xenograft rejection, their involvement in lung dysfunction is less defined. This project was designed to define the role of antibodies in pulmonary graft rejection in a pig-to-baboon model. METHODS: Orthotopic transgenic swine left lung transplants were performed in baboons depleted of antibodies by one of three techniques before transplantation: (1) ex vivo swine kidney perfusion, (2) total immunoglobulin-depleting column perfusion, and (3) ex vivo swine lung perfusion. Results were compared with those of transgenic swine lung transplants in unmodified baboons. RESULTS: All three techniques of antibody removal resulted in depletion of xenoreactive antibodies. Only pretransplantation lung perfusion improved pulmonary xenograft function compared with lung transplantation in unmodified baboons. CONCLUSIONS: The pathogenesis of pulmonary injury in a swine-to-primate transplant model is different from that in renal and cardiac xenografts. Depletion of antibodies alone does not have a beneficial effect and may actually be detrimental.

CD46 transgene expression in pig peripheral blood mononuclear cells does not alter their susceptibility to measles virus or their capacity to downregulate endogenous and transgenic CD46.

CD46 (or membrane cofactor protein) protects autologous cells from complement-mediated lysis and has been expressed as a transgene in pigs to overcome complement-mediated hyperacute rejection of porcine organs upon transplantation into primates. Since CD46 has been identified as a receptor for measles virus (MV), the susceptibility of CD46-transgenic (tg) pig peripheral blood mononuclear cells (PBMC) to infection with MV strains which do and do not use CD46 as receptor was investigated. Surprisingly, it was found that MV vaccine strains (e.g. Edmonston) bound to tg as well as non-tg pig PBMC. Phytohaemagglutinin-stimulated CD46-tg and non-tg pig PBMC were equally well infected with MV vaccine strains irrespective of CD46 expression. Upon infection, tg CD46 was downregulated from the cell surface. In contrast, the binding capacity for MV wild-type strains to pig and human PBMC was low, irrespective of CD46 expression. These MV strains did not infect tg or non-tg pig cells. Expression of endogenous pig CD46 was detected with polyclonal sera against human CD46. After infection of pig PBMC with MV strain Edmonston, endogenous pig CD46 was also downregulated. This suggests an interaction between MV Edmonston and pig CD46. However, polyclonal CD46 sera did not inhibit infection with MV Edmonston indicating that CD46 may not exclusively act as a receptor for MV on these cells. Interestingly, similar results were observed using human PBMC. Data suggest that CD46 downregulation after interaction with MV may also occur in porcine organs which express endogenous and/or human CD46 as a means of protection against complement-mediated damage.

Liver allotransplantation after extracorporeal hepatic support with transgenic (hCD55/hCD59) porcine livers: clinical results and lack of pig-to-human transmission of the porcine endogenous retrovirus.

BACKGROUND: Whole organ extracorporeal perfusion of a genetically modified humanized (transgenic) pig liver has been proposed as a technology that may sustain patients with severe liver failure while awaiting human liver transplantation. METHODS: We report on two cases of successful extracorporeal perfusion of a transgenic pig liver in patients awaiting transplantation for fulminant hepatic failure. The pig livers used were transgenic for human CD55 (decay-accelerating factor) and human CD59. These transgenic modifications are designed to reduce or eliminate the hyperacute rejection inherent in pig-to-primate xenotransplants. We also report on the results of serial surveillance testing for presence of the porcine endogenous retrovirus (PoERV) in these two patients. RESULTS: Extracorporeal perfusion in two patients was performed for 6.5 and 10 hr, respectively, followed by the successful transplantation of a human liver and resultant healthy patients (18 and 5 months later as of this writing). The porcine livers showed evidence of synthetic and secretory function (decreasing protime and bilirubin, bile production). Serial polymerase chain reaction analysis of these patients' peripheral blood mononuclear cells has failed to show presence of PoERV DNA sequences. CONCLUSIONS: The CD55/CD59 transgenic porcine liver appears capable of safely "bridging" a patient to liver transplantation. Human PoERV infection from these livers has yet to be demonstrated.

The role of anti-Galalpha1-3Gal antibodies in acute vascular rejection and accommodation of xenografts.

BACKGROUND: A major impediment to the transplanting of porcine organs into humans is the susceptibility of porcine organs to acute vascular rejection, which can destroy a vascularized xenograft over a period of hours to days. Acute vascular rejection of porcine-to-primate xenografts is thought to be triggered by binding of xenoreactive antibodies to the graft. We tested whether antibodies, binding to Galalpha1-3Gal epitopes in porcine tissue, initiate this phenomenon. METHODS AND RESULTS: Specific depletion of anti-Galalpha1-3Gal antibodies from the blood of baboons, using extracorporeal perfusion of separated plasma through columns of Sepharose beads covalently linked to the antigenic trisaccharide, Galalpha1-3Galbeta1-4GlcAc, averted the development of acute vascular rejection in porcine organs transgenic for human decay-accelerating factor and CD59. More importantly, after immunodepletion was stopped and Gala1-3Gal antibodies were allowed to return, these same organs continued to function and remained pathologically normal and thus seemed to achieve a state of accommodation. CONCLUSION: These results demonstrate that anti-Galalpha1-3Gal antibodies cause acute vascular rejection and suggest that depletion of these antibodies leads to accommodation of the donor cardiac xenograft and could supply an important model for additional study.

Hearts from transgenic pigs constructed with CD59/DAF genomic clones demonstrate improved survival in primates.

We have previously created transgenic pigs bearing the human complement regulatory proteins CD59 and decay-accelerating factor (DAF) by either the intercellular transfer or the cDNA transgenic method. To achieve more physiologic protein expression, we constructed a new line of transgenic pigs with CD59 and DAF human genomic clones. We transplanted these CD59/DAF transgenic pig hearts into baboons immunosuppressed with cyclosporine, methylprednisone or leflunomide/mofetil mycophenolate. The four wild-type hearts survived for 20-80 min, whereas the four CD59/DAF hearts functioned for 85-130 h. Immunohistochemical staining showed levels of CD59 and DAF protein expression similar to that in human hearts. Wild-type and transgenic hearts demonstrate a similar level of IgM deposition, although transgenic hearts suffered less hyperacute rejection and thus less membrane attack complex deposition. The histology of the transgenic grafts after explant was consistent with acute vascular rejection, with a high level of IgG deposit compared with wild-type control. We conclude that this new line of CD59/DAF transgenic pigs express high levels of the transgene products, which conferred longer survival because of better protection from hyperacute rejection. Similar to previous transgenic pigs, however, these animals suffered from delayed xenograft rejection.

High level transcription of the complement regulatory protein CD59 requires an enhancer located in intron 1.

CD59 is a complement regulatory protein and may also act as a signal-transducing molecule. CD59 transgenic mice have been generated using a CD59 minigene (CD59 minigene-1). Although this minigene contained a 4.6-kilobase pair 5'-flanking region from the human CD59 gene as a promoter, the expression levels of the CD59 mRNA were substantially lower than those observed in humans, suggesting that CD59 gene expression might also require other transcriptional regulatory elements such as an enhancer. To investigate the transcriptional regulation of the CD59 gene, we used three cell lines that express CD59 at different levels. We have identified DNase I-hypersensitive sites in intron 1 in HeLa cells, which express CD59 at high levels, but not in Jurkat (intermediate level) or Raji cells (low level). Furthermore, cell line-specific enhancer activity was detected in a fragment containing these DNase I-hypersensitive sites. The CD59 enhancer was mapped to between -1155 and -888 upstream of the 5'-end of exon 2. To investigate the enhancer activity in vivo, a new CD59 minigene was constructed by the addition of the enhancer fragment into CD59 minigene-1. High expressor CD59 transgenic mice were generated using the new minigene.

Progress in xenotransplantation.

Organ transplantation is considered the most effective treatment for end-stage organ failure; currently it is limited by a severe worldwide shortage of human donor organs. This has led to investigation of the potential use of animals as organ donors. For a number of reasons, the pig represents the most likely organ donor candidate. Transplantation of a vascularised porcine organ into a human or non-human primate results in an immediate and dramatic rejection process, known as hyperacute rejection, which is mediated by the binding of pre-existing antibody to the porcine graft and the subsequent activation of host complement. Strategies aimed at preventing this initial rejection have been largely successful in experimental models. This has allowed attention to turn towards an understanding of the immunological barriers comprising the next phase of xenograft rejection, termed acute vascular rejection. This delayed rejection process appears to be a humoral event, and it is likely that the control of antibody synthesis will play a pivotal role in overcoming the current barrier to successful xenotransplantation.

Role of CD46 in measles virus infection in CD46 transgenic mice.

The susceptibility of CD46 (human membrane cofactor protein) transgenic mice to measles virus (MV) infection was investigated. Cell cultures (lung and kidney) established from transgenic and control mice showed that although both could be infected only those from the CD46+ mice gave fusion. A complete round of replication with the release of infectious virus was detected exclusively in the transgenic cell cultures whose permissiveness to MV was markedly less than that of Vero cells. The ability of MV to replicate in vivo in mice was studied using both vaccine and laboratory-adapted wild-type strains of virus. After intraperitoneal and intranasal inoculations of transgenic mice, virus replication could not be detected. In contrast intracerebral inoculation induced infection in both transgenic and nontransgenic mice. Our results from in vitro infection studies support the hypothesis that CD46 is a major host cell factor involved in the MV-induced fusion process and MV entry. The studies further indicate that MV tropism is not governed solely by the expression of the CD46 gene and that the high efficiency of the replicative cycles characteristic of fully permissive host cells requires additional factors, which are lacking in both transgenic and nontransgenic mice.

The role of antibodies in acute vascular rejection of pig-to-baboon cardiac transplants.

Long-term success in xenotransplantation is currently hampered by acute vascular rejection. The inciting cause of acute vascular rejection is not yet known; however, a variety of observations suggest that the humoral immune response of the recipient against the donor may be involved in the pathogenesis of this process. Using a pig-to-baboon heterotopic cardiac transplant model, we examined the role of antibodies in the development of acute vascular rejection. After transplantation into baboons, hearts from transgenic pigs expressing human decay-accelerating factor and CD59 underwent acute vascular rejection leading to graft failure within 5 d; the histology was characterized by endothelial injury and fibrin thrombi. Hearts from the transgenic pigs transplanted into baboons whose circulating antibodies were depleted using antiimmunoglobulin columns (Therasorb, Unterschleisshein, Germany) did not undergo acute vascular rejection in five of six cases. Biopsies from the xenotransplants in Ig-depleted baboons revealed little or no IgM or IgG, and no histologic evidence of acute vascular rejection in the five cases. Complement activity in the baboons was within the normal range during the period of xenograft survival. In one case, acute vascular rejection of a xenotransplant occurred in a baboon in which the level of antidonor antibody rose after Ig depletion was discontinued. This study provides evidence that antibodies play a significant role in the pathogenesis of acute vascular rejection, and suggests that acute vascular rejection might be prevented or treated by therapies aimed at the humoral immune response to porcine antigens.