Diagnostic Criteria For Antibody-mediated Rejection Research Articles

Endomyocardial Biopsy Findings in Pediatric Heart Transplants Recipients with Positive HLA Associated Antibodies: Review of a Cohort with Poor Outcomes

Purpose Antibody mediated rejection (AMR) continues to be a major complication after heart transplantation (OHT) and poses a risk for allograft failure, graft vasculopathy and mortality. Diagnostic criteria for AMR have evolved over the years. Incidence of AMR is estimated to be 11-20% after OHT. We sought to evaluate the histological characteristics of AMR in endomyocardial biopsies (EMB) of deceased pediatric patients with positive HLA-associated antibodies. Methods A retrospective chart review was conducted to identify all pediatric patients ( Results During the study period, 86 OHT recipients had HLA associated antibodies, with 16 eventual deaths, 4 of which followed re-transplantation. A total of 152 biopsies from all 16 patients were reviewed. Eight patients (50%) demonstrated no evidence of pAMR. Eight patients (50%) demonstrated pAMR with peak grade of 3 in two patients; grade 2 in one patient, grade 1(i+) in three, and grade 1(h+) in two patients. Mean duration to onset of pAMR was 16.25 months from OHT, range ( Conclusion In our experience, 50% of the patients with HLA associated antibodies developed pAMR. De-novo DSAs were present in all patients that demonstrated pAMR. Unique histologic predictors of poor outcome are unclear from analysis of existing criteria alone. Additional parameters may need to be included to the existing criteria to improve diagnostic sensitivity of pAMR.

Antibody-mediated rejection in pediatric small bowel transplantation: Capillaritis is a major determinant of C4d positivity in intestinal transplant biopsies.

The diagnostic criteria for antibody-mediated rejection (ABMR) after small bowel transplantation (SBT) are not clearly defined, although the presence of donor-specific antibodies (DSAs) has been reported to be deleterious for graft survival. We aimed to determine the incidence and prognostic value of DSAs and C4d in pediatric SBT and to identify the histopathologic features associated with C4d positivity. We studied all intestinal biopsies (IBx) obtained in the first year posttransplantation (N=345) in a prospective cohort of 23 children. DSAs and their capacity to fix C1q were identified by using Luminex technology. Eighteen patients (78%) had DSAs, and 9 had the capacity to fix C1q. Seventy-eight IBx (22.6%) were C4d positive. The independent determinants of C4d positivity were capillaritis grades 2 and 3 (odds ratio [OR] 4.02, P=.047 and OR 5.17, P=.003, respectively), mucosal erosion/ulceration (OR 2.8, P=.019), lamina propria inflammation grades 1 and 2/3 (OR 1.95, P=.043 and OR 3.1, P=.016, respectively), and chorion edema (OR 2.16, P=.028). Complement-fixing DSAs and repeated C4d-positive IBx were associated with poor outcome (P=.021 and P=.001, respectively). Our results support that capillaritis should be considered as a feature of ABMR in SBT and identify C1q-fixing DSAs and repeated C4d positivity as potential markers of poor outcome.

Defining the phenotype of antibody-mediated rejection in kidney transplantation: Advances in diagnosis of antibody injury.

The diagnostic criteria for antibody-mediated rejection (ABMR) are constantly evolving in light of the evidence. Inclusion of C4d-negative ABMR has been one of the major advances in the Banff Classification in recent years. Currently Banff 2015 classification requires evidence of donor specific antibodies (DSA), interaction between DSA and the endothelium, and acute tissue injury (in the form of microvasculature injury (MVI); acute thrombotic microangiopathy; or acute tubular injury in the absence of other apparent cause). In this article we review not only the ABMR phenotypes acknowledged in the most recent Banff classification, but also the phenotypes related to novel pathogenic antibodies (non-HLA DSA, antibody isoforms and subclasses, complement-binding functionality) and molecular diagnostic tools (gene transcripts, metabolites, small proteins, cytokines, and donor-derived cell-free DNA). These novel tools are also being considered for the prognosis and monitoring of treatment response. We propose that improved classification of ABMR based on underlying pathogenic mechanisms and outcomes will be an important step in identifying patient-centered therapies to extend graft survival.

Longitudinal immunological characterization of the first presensitized recipient of a face transplant

Rejection affects greater than 80% of face transplants, yet no diagnostic criteria for antibody-mediated rejection (AMR) following face transplantation have been established. Given that different treatment strategies are required to address AMR and T cell-mediated rejection (TCMR), there is a critical need to delineate the features that can differentiate these two alloimmune responses. Here, we report the longitudinal immunological examination of what we believe to be the first and only highly sensitized recipient of a crossmatch-positive face transplant up to 4 years following transplantation. We conducted gene expression profiling on allograft biopsies collected during suspected AMR and TCMR episodes as well as during 5 nonrejection time points. Our data suggest that there are distinctive molecular features in AMR, characterized by overexpression of endothelial-associated genes, including ICAM1, VCAM1, and SELE. Although our findings are limited to a single patient, these findings highlight the potential importance of developing and implementing molecular markers to differentiate AMR from TCMR to guide clinical management. Furthermore, our case illustrates that molecular assessment of allograft biopsies offers the potential for new insights into the mechanisms underlying rejection. Finally, our medium-term outcomes demonstrate that face transplantation in a highly sensitized patient with a positive preoperative crossmatch is feasible and manageable.

Evaluation of Microvascular Inflammation in ABO-Incompatible Kidney Transplantation.

In ABO-incompatible kidney transplantation, the diagnostic criteria for antibody-mediated rejection remain controversial because C4d deposition is commonly observed. Thus, we investigated microvascular inflammation (MVI score ≥ 2) within 1 year as a predictor of graft outcome. A total of 148 recipients without preformed or de novo donor-specific anti-HLA antibody were stratified based on MVI score less than 2 (n = 117) and MVI score of 2 or greater (n = 31). We found that 5-year graft survival was significantly lower (P = 0.0129) in patients with MVI (89.8%) than in patients without MVI (97.0%). Graft function, as characterized by serum estimated glomerular filtration rate, was also significantly worse for patients with MVI than it was for patients without MVI, between 3 months and 10 years after transplantation (P = 0.048). Multivariate analysis indicated that HLA class II mismatch (P = 0.0085) was an independent marker of MVI. Microvascular inflammation score of 2 or greater is significantly associated with poor graft outcome after ABO-incompatible kidney transplantation. We suggest that MVI score of 2 or greater in ABOi transplantation be used as a basis to diagnose antibody-mediated rejection.

Antibody-Mediated Rejection: A Review.

Chronic antibody injury is a serious threat to allograft outcomes and is therefore the center of active research. In the continuum of allograft rejection, the development of antibodies plays a critical role. In recent years, an increased recognition of molecular and histologic changes has provided a better understanding of antibody-mediated rejection (AMR), as well as potential therapeutic interventions. However, several pathways are still unknown, which accounts for the lack of efficacy of some of the currently available agents that are used to treat rejection. We review the current diagnostic criteria for AMR; AMR paradigms; and desensitization, treatment, and prevention strategies. Chronic antibody-mediated endothelial injury results in transplant glomerulopathy, manifested as glomerular basement membrane duplication, double contouring, or splitting. Clinical manifestations of AMR include proteinuria and a rise in serum creatinine. Current strategies for the treatment of AMR include antibody depletion with plasmapheresis (PLEX), immunoadsorption (IA), immunomodulation with intravenous immunoglobulin (IVIG), and T cell- or B cell-depleting agents. Some treatment benefits have been found in using PLEX and IA, and some small nonrandomized trials have identified some benefits in using rituximab and the proteasome inhibitor-based therapy bortezomib. More recent histologic follow-ups of patients treated with bortezomib have not shown significant benefits in terms of allograft outcomes. Furthermore, no specific treatment approaches have been approved by the US Food and Drug Administration. Other agents used for more difficult rejections include bortezomib and eculizumab (an anti-C5 monoclonal antibody). AMR is a fascinating field with ample opportunities for research and progress in the future. Despite the use of advanced techniques for the detection of human leukocyte antigen (HLA) or non-HLA donor-specific antibodies, alloimmune response remains an important barrier for successful long-term allograft function. Treatment of AMR with currently available therapies has produced a variety of results, some of them suboptimal, precluding the development of standardized protocols. New therapies are promising, but randomized controlled trials are needed to find surrogate markers and improve the efficacy of therapy.

Clinical and molecular significance of microvascular inflammation in transplant kidney biopsies

The diagnostic criteria for antibody-mediated rejection (AMR) are continuously evolving. Here we investigated the clinical and molecular significance of different Banff microvascular inflammation (MVI) scores in transplant kidney biopsies. A total of 356 patients with clinically indicated kidney transplant biopsies were classified into three groups based on MVI scores of 0, 1, 2, or more for Groups 1-3, respectively. Gene expression profiles were assessed using arrays on a representative subset of 93 patients. The incidence of donor-specific anti-HLA antibodies was increased from 25% in Group 1 to 36% in Group 2 and to 54% in Group 3. Acute and chronic AMR were significantly more frequent in Group 3 (15% and 35%) compared with the Group 2 (3% and 15%) and Group 1 (0% and 5%), respectively. Gene expression profiles showed increased interferon-γ and rejection-induced, cytotoxic and regulatory T-cell, natural killer cell-associated and donor-specific antibody (DSA)-selective transcripts in Group 3 compared with Groups 1 and 2. There was no significant difference in gene expression profiles between the Groups 1 and 2. Increased intragraft expression of DSA-selective transcripts was found in the biopsies of C4d- Group 3 patients. Thus, an MVI score of 2 or more was significantly associated with a histological diagnosis of acute and chronic antibody-mediated rejection. Hence, increased intragraft DSA-selective gene transcripts may be used as molecular markers for AMR, especially in C4d- biopsies.

Updates on antibody-mediated rejection in intestinal transplantation.

Antibody-mediated rejection (ABMR) has increasingly emerged as an important cause of allograft loss after intestinal transplantation (ITx). Compelling evidence indicates that donor-specific antibodies can mediate and promote acute and chronic rejection after ITx. However, diagnostic criteria for ABMR after ITx have not been established yet and the mechanisms of antibody-mediated graft injury are not well-known. Effective approaches to prevent and treat ABMR are required to improve long-term outcomes of intestine recipients. Clearly, ABMR after ITx has become an important area for research and clinical investigation.

The utility of circulating donor specific antibody to predict biopsy-proven antibody-mediated rejection and to provide prognostic value after heart transplantation in children

Aim Heart transplantation (HT) is a life-saving option for children with end-stage heart failure. Antibody-mediated rejection (AMR) is a significant cause of morbidity and mortality in HT. The current diagnostic criteria for AMR are based on endomyocardial biopsy (EMB) showing evidence of C4d and C3d deposition, and capillary endothelial changes, including neutrophil and macrophage infiltration, and interstitial edema. Although the presence of donor specific antibody (DSA) is a risk factor for the development of AMR after HT, it is currently not required for diagnosis of AMR. In this study, we aimed to investigate the sensitivity, specificity, predictive values and prognostic implications of circulating DSA using EMB as the gold standard for AMR diagnosis in a cohort of pediatric HT recipients. Methods Retrospective study in pediatric HT patients that had follow-up care between 2009–2013 and had at least one EMB paired with DSA testing within 3 days of the biopsy. Positive DSA was defined at a mean fluorescent intensity (MFI) ⩾ 2000 using single antigen bead testing. Results Of the 66 patients included, 26 (39%) had at least 1 HLA DSA positive test. Of the 26 patients with positive DSA, 18 (69%) had at least one antibody against HLA-DQ. In 7 (27%) of the 26 patients with positive DSA, the only antibody detected was against HLA-DQ. Of 236 DSA EMB pairs analyzed, 132 (56%) pairs were negative for DSA, 23 (10%) pairs had HLA DSA Class I only, 59 (25%) pairs had HLA DSA Class II only, and 22 (9%) pairs had both HLA class I and II DSA. HLA DSA testing had a sensitivity = 92.6%, specificity = 62.2%, PPV = 24.0%, and NPV = 98.5% for biopsy-proven AMR. Kaplan–Meier analysis showed that patients with positive HLA DSA results and biopsy-proven AMR trended toward a higher incidence of cardiovascular mortality, and coronary artery vasculopathy after 3 years compared to patients without circulating DSA and a negative biopsy. Conclusions The results of HLA-DSA testing in this cohort, showing excellent sensitivity and NPV for biopsy-proven AMR suggest for the first time, that DSA testing could be used for non-invasive prediction of AMR absence in heart transplantation, and avoid unnecessary biopsies. The value of DSA testing for biopsy-proven AMR and outcome prognosis in HT is important to study in future multi-center prospective studies.

Antibody testing for cardiac antibody-mediated rejection: Which panel correlates best with cardiovascular death?

Recent efforts are being undertaken to update and refine current diagnostic criteria for antibody-mediated rejection (AMR) in heart transplantation. We believe that the appropriate reactants are those that best predict the adverse consequences of AMR and therefore tested various models using different reactants to find the best predictors of cardiovascular mortality in pathologically defined AMR. The study group included only patients in whom all immunofluorescence antibodies of interest had been tested on biopsy specimens obtained after 2002 when C4d was routinely added. We analyzed our data using 3 Cox proportional hazard models with time-varying covariates using an end point of cardiovascular mortality, as previously defined. In 3,712 biopsy specimens from 422 patients, the 2-antibody model achieved a value of R(2) = 0.930 using C3d and C4d antibodies alone. A model that used 4 antibodies--C3d, C4d, human leukocyte antigen-D related (HLA-DR) and fibrin--was superior (R(2) = 0.988). The model that best predicted cardiovascular mortality included all 6 antibodies: HLA-DR, immunoglobulin (Ig) G, IgM, C3d, C4d, and fibrin (R(2) = 0.989). The models using 4 or 6 antibodies were significantly superior to the model using only C3d and C4d (for each interaction, p < 0.0001). The combination of complement components, HLA-DR and fibrin, is valuable in defining AMR in patients at risk for allograft loss from cardiovascular causes. Fibrin is particularly important for detecting the presence of severe AMR, with a high likelihood of poor long-term patient outcome.

Correlation of Donor-Specific Antibodies, Complement and Its Regulators with Graft Dysfunction in Cardiac Antibody-Mediated Rejection

Antibody-mediated rejection (AMR) is an immunopathologic process in which activation of complement often results in allograft injury. This study correlates C4d and C3d with HLA serology and graft function as diagnostic criteria for AMR. Immunofluorescence staining for C4d and C3d was performed on 1511 biopsies from 330 patients as part of routine diagnostic work-up of rejection. Donor-specific antibodies were detected in 95% of those with C4d+C3d+ biopsies versus 35% in the C4d+C3d- group (p = 0.002). Allograft dysfunction was present in 84% in the C4d+ C3d+ group versus 5% in the C4d+C3d- group (p < 0.0001). Combined C4d and C3d positivity had a sensitivity of 100% and specificity of 99% for the pathologic diagnosis of AMR and a mortality of 37%. Since activation of complement does not always result in allograft dysfunction, we correlated the expression pattern of the complement regulators CD55 and CD59 in patients with and without complement deposition. The proportion of patients with CD55 and/or CD59 staining was highest in C4d+C3d- patients without allograft dysfunction (p = 0.03). We conclude that a panel of C4d and C3d is diagnostically more useful than C4d alone in the evaluation of AMR. CD55 and CD59 may play a protective role in patients with evidence of complement activation.

Antibody-Mediated Rejection of the Pancreas Allograft

Antibody-mediated rejection (AMR) is an entity that is being increasingly recognized in the setting of solid organ transplantation. Clinical, pathologic, and serologic criteria are perhaps better defined and understood in kidney and heart transplants (1), but little is known for other solid organs, including lung, pancreas, liver, and small intestine transplantation. In an effort to understand the impact and significance of alloantibodies in the setting of pancreas transplantation, we systematically screen for donor-specific antibodies (DSA) and immunolabel with C4d all pancreas allograft biopsies at our institution. In a study of 136 recipients of simultaneous pancreas and kidney transplants, we initially reported two patients with histologic evidence of both kidney and pancreas acute AMR, with diffusely C4d-positive peritubular and interacinar capillary staining, respectively. Both patients had documented circulating DSA. One patient lost both kidney and pancreas allografts 1 year posttransplant, and the second patient fully recovered kidney and pancreas allografts with normal function 2 years posttransplant (2). More recently, in our study of 18 pancreas allograft recipients and 27 biopsies, we found a significant correlation between circulating DSA and C4d staining of interacinar capillaries, and no correlation of DSA with biopsy staining of larger arteries, veins, islets, or interstitium (3). In this study, half of the patients returned to insulin therapy and half of the patients retained normal allograft function after a mean follow-up period of 316.3 days. Patients with pure acute cellular rejection fully recovered allograft function after a mean follow-up period of 208.4 days. The decreased allograft survival in AMR+pancreas recipients is in agreement with the decreased survival shown in C4d+AMR patients reported by Munivenkatappa (4). We have also looked at the potential role of C3d, another proposed tissue marker of antibody activation and deposition in kidney transplant (5). In a group of 16 pancreas allografts, C3d proved to be less sensitive than C4d in identifying patients with AMR, and there was no correlation with class I or class II DSA (Poster presentation to the 2009 ATC). We agree with the viewpoints of Dr. Munivenkatappa, Dr. Papadimitriou, and Dr. Drachenberg in their letter to the editor of this edition of Transplantation. In pancreas transplantation (as in other solid organ transplant) there is perhaps a spectrum of clinical scenarios, tissue evidence of antibody activation, and serologic patterns that still need to be defined to better understand AMR as an entity in the pancreas allograft. The diagnostic criteria for AMR in the pancreas (6) will be a subject of important discussions in the next Banff meeting in Alberta, Canada, August 2009, and will undoubtedly undergo refinement in the future. Jose R. Torrealba Department of Pathology University of Wisconsin Madison, WI Jon Odorico Department of Surgery University of Wisconsin Madison, WI

A Clinical Correlation Study of Severity of Antibody-mediated Rejection and Cardiovascular Mortality in Heart Transplantation

The current International Society for Heart and Lung Transplantation (ISHLT) diagnostic criteria for antibody-mediated rejection (AMR) designate AMR as either absent (AMR 0) or present (AMR 1), without grading its severity. Yet, the extent of histologic and immunofluorescence (IF) findings of AMR varies across endomyocardial biopsies (EMBs). In this study, we hypothesized that the severity of AMR, as assessed on EMBs, correlates with cardiovascular mortality in heart transplant recipients. All EMBs from 1985 to 2005 were evaluated. Biopsy specimens were uniformly studied by light microscopy and IF early post-transplant. A comprehensive vascular score (V1: no AMR, to V5: severe AMR) was prospectively assigned to each EMB, based on severity of both histologic and IF findings. Univariate Cox proportional hazards regressions were performed using indicators of vascular scores alone, combined, and cumulatively. Nine hundred six patients were transplanted and included in the study. Mean age was 46.6 +/- 15.5 years and 82% were male. A total of 26,236 EMBs comprised the study data. As expected, histologic and immunopathologic findings of AMR varied in severity. An incremental risk of cardiovascular mortality was found with more severe AMR whether vascular scores were analyzed individually (p = 0.001), in combination (p = 0.01) or cumulatively (p = 0.006). The severity of AMR on EMBs correlates with an incremental cardiovascular mortality risk after heart transplantation, suggesting that AMR should be viewed as a spectrum rather than just as present or absent. Supplementing the ISHLT AMR diagnostic guidelines with a consensus severity scale is warranted.

193: Antibody-Mediated Rejection in Heart Transplantation: More Than Present-or-Absent

Diagnostic criteria for antibody-mediated rejection (AMR) were established as part of the International Society for Heart and Lung Transplantation (ISHLT) endomyocardial biopsy (EMB) grading scheme. This new scheme designates AMR as either absent (AMR 0) or present (AMR 1), without grading severity. Yet, we had previously shown that the prominence of histologic and immunofluorescence (IF) findings of AMR varies across EMBs. In this study, we hypothesized that the severity of AMR on pathology correlates with an increased risk of cardiovascular mortality (CVM) in heart transplant recipients.

Utility of Histologic Parameters in Screening for Antibody-Mediated Rejection of the Cardiac Allograft: A Study of 3,170 Biopsies

Diagnostic criteria for antibody-mediated rejection (AMR) of the cardiac allograft have recently been proposed as part of the International Society for Heart and Lung Transplantation (ISHLT) biopsy grading scheme. Histologic features of vascular adherence of macrophages (VASC) and endothelial activation or swelling in capillaries (ENDO) are proposed as criteria to prompt the immunohistochemical investigation of biopsies for AMR. The aim of this study was to determine whether VASC and ENDO are adequate to act as screening parameters to trigger further AMR investigation. We examined our database of biopsy findings where histologic vascular parameters as well as immunofluorescence (IF) to detect AMR were collected (n = 3,170). Histologic parameters were graded semi-quantitatively on a scale from 1 to 5, where 1 = absence and 5 = obvious and generalized presence of the finding. Seven hundred sixty-eight of 3,170 biopsies had IF findings diagnostic of AMR in the absence of cellular rejection (ISHLT = 0). ENDO had a sensitivity of 63% and a specificity of 80%. VASC had a sensitivity of 30% and specificity of 99%. Combining the interpretation of the 2 tests did not result in a significant improvement of test sensitivity. Neither ENDO, VASC nor the combination of the tests indicated sufficiently high sensitivity to serve as a screening tool before further diagnostic investigation for AMR. Immunohistochemical testing remains necessary in the majority of cases to identify AMR.

C4d in antibody-mediated rejection-how sensitive is it?

Diagnostic criteria for antibody-mediated rejection (AMR) are evolving. We have previously described a method of histologic examination and routine immunofluorescence (IF) with classification of results as cellular, antibody-mediated, or mixed rejection. The aim of this study was to assess performance of C4d in diagnosing AMR.

New Diagnostic Criteria for Antibody-Mediated Rejection in the Renal Allograft: A Significant Advance

New Diagnostic Criteria for Antibody-Mediated Rejection in the Renal Allograft: A Significant Advance

Study 1: characterization of acute rejection episodes after pre-emptive plasmapheresis and cytomegalovirus hyperimmune globulin to abrogate a positive crossmatch and allow live donor renal transplantation

P atients who undergo desensitization protocols to allow renal transplantation despite a positive crossmatch to their living donor remain at risk for acute rejection, specifically, antibody-mediated rejection. Our center uses plasmapheresis and cytomegalovirus hyperimmune globulin (PP-CMVIg) to precondition such patients. In this study, we seek to identify those patients at highest risk for antibodymediated rejection, allowing future modifications to decrease the incidence of acute rejection. All 33 patients treated with PP-CMVIg before live donor renal transplant (Feb 1998–Oct 2002) were analyzed. Episodes of acute rejection were confirmed by biopsy. Diagnostic criteria for antibody-mediated rejection included compatible histologic findings, positive C4d immunohistochemical results, and rising donor-specific antibody titer. Acute rejection occurred in 21 of 33 patients (64%), at a median of 9 days after transplantation, and 12 of 21 were diagnosed with antibody-mediated rejection. Antibody-mediated rejection was treated with steroids and additional PP-CMVIg treatments (mean, 4.3); steroids with or without antilymphocyte therapy were used to treat the 9 other cases of acute cellular rejection. Three patients sustained graft loss, as follows: 1 noncompliance, 1 hemolytic uremic syndrome, and 1 associated with patient death from surgical complications. Patients with an initially positive cytotoxic crossmatch (vs flow, P .007), a positive cytotoxic crossmatch at transplant (vs negative, P .032), and patients who required 6 pretransplant PP-CMVIg treatments (P 0.012) had a higher risk of acute rejection. Occurrence of acute rejection was associated with a longer length of stay (16 vs 5 days, P .0001) and lower 1-week creatinine clearance (43 vs 65 mL/min, P .01). Patients who had an episode of acute rejection had no significant change in their current creatinine clearance (75 vs 73 mL/min at 16 months’ mean follow-up, P .9). Patients with antibody-mediated rejection did not have significantly different outcomes compared with those with cellular rejection alone. Intensity of C4d and donor-specific antibody titer did not predict graft function. Donor-specific antibody remains undetectable in 28 of 30 patients with a functioning allograft. Conclusions: Patients receiving PP-CMVIg to abrogate a positive crossmatch are at risk for acute rejection, especially antibody-mediated rejection, but are responsive to additional PP-CMVIg. Acute rejection does not have an impact on graft function at 1 year. Patients with high levels of donor-specific antibody, as evidenced by an initially positive cytotoxic crossmatch, a positive cytotoxic crossmatch at the time of transplantation, and a requirement of more pretransplant PP-CMVIg treatments, are at highest risk for antibody-mediated rejection and require vigilant observation during the first 2 weeks after transplantation. From the Departments of Surgery, Medicine, and Pathology, The Johns Hopkins University School of Medicine, Baltimore, MD. © 2003 Elsevier Inc. All rights reserved. 0955-470X/03/1704-0000$30.00/0 doi:10.1016/j.trre.2003.10.002