Cardiovascular mortality following liver transplantation: predictors and temporal trends over 30 years

AIMS

There has been significant evolution in operative and post-transplant therapies following liver transplantation (LT). We sought to study their impact on cardiovascular (CV) mortality, particularly in the longer term.

METHODS AND RESULTS

A retrospective cohort study was conducted of all adult LTs in Australia and New Zealand across three 11-year eras from 1985 to assess prevalence, modes, and predictors of early (≤30 days) and late (>30 days) CV mortality. A total of 4265 patients were followed-up for 37 409 person-years. Overall, 1328 patients died, and CV mortality accounted for 228 (17.2%) deaths. Both early and late CV mortality fell significantly across the eras (P < 0.001). However, CV aetiologies were consistently the leading cause of early mortality and accounted for ∼40% of early deaths in the contemporary era. Cardiovascular deaths occurred significantly later than non-cardiac aetiologies (8.8 vs. 5.2 years, P < 0.001). On multivariable Cox regression, coronary artery disease [hazard ratio (HR) 4.6, 95% confidence interval (CI) 1.2-21.6; P = 0.04] and era of transplantation (HR 0.44; 95% CI 0.28-0.70; P = 0.01) were predictors of early CV mortality, while advancing age (HR 1.05, 95% CI 1.02-1.10; P = 0.005) was an independent predictors of late CV mortality. Most common modes of CV death were cardiac arrest, cerebrovascular events, and myocardial infarction.

CONCLUSION

Despite reductions in CV mortality post-LT over 30 years, they still account for a substantial proportion of early and late deaths. The late occurrence of CV deaths highlights the importance of longitudinal follow-up to study the efficacy of targeted risk-reduction strategies in this unique patient population.

Liver transplantation in Australia and New Zealand

Liver transplantation (LT) in Australia and New Zealand began in 1985. Over this time until December 2014, LT took place in 3700 adults and 800 children. LT is regulated with 1 unit, supported by the government, per state or region. Currently approximately 270 transplants take place per year. Organ donation rates are moderate in Australia (17 per 1 million of population) but very low in New Zealand (11 per 1 million of population). All the units share organ donors for fulminant hepatic failure cases (status 1). Recipient listing criteria and organ allocation criteria are commonly agreed to via National and Trans-Tasman agreements, which are published online. Current survival rates indicate approximately 94% 1-year survival with median survival in adults of approximately 20 years, whereas 75% of children are alive at 20 years. All units collaborate in research projects via the Australia and New Zealand Liver Transplant Registry and have published highly cited articles particularly on the prevention of hepatitis B virus recurrence. Outcomes for indigenous populations have also been analyzed. In conclusion, LT in Australia and New Zealand is well developed with transparent processes related to criteria for listing and organ allocation together with publication of outcomes. Liver Transplantation 22 830-838 2016 AASLD.

The New Zealand Liver Transplant Unit: Auckland District Health Board

New Zealand is a geographically isolated country with 4.55 million inhabitants. It has endemic hepatitis B (HBV) infection that is especially evident in Maori and Pacific Island communities and impacts indications for liver transplantation. The country has a socialised medical system that allows for full coverage of the assessment for, and completion of liver transplants in suitable recipients. Between February 1998 and December 2014, the New Zealand Liver Transplant Unit (NZLTU) had performed 595 liver transplants in 568 patients, indicating a crude re-transplant rate of 4.8%. Overall 1, 5, and 10 year patient survival rates for all adult (96%, 89%, and 81%, respectively) and pediatric (93%, 92%, and 92%, respectively) recipients compare very favourably with international outcomes from Europe and the United States. Eligibility criteria could be modestly expanded if deceased donor rates improved from the current level of around 10 per million of population per year. This somewhat meagre supply of deceased donor organs, along with significant waiting list attrition, has necessitated the use of living donors, which have been used in more than 50 recipients to date. Despite these limitations, the NZLTU has contributed to improvements in the outcome of transplantation for HBV and hepatitis C through the development of effective antiviral prophylaxis regimes. Furthermore, innovative changes have been made to the manner in which pediatric patients are transitioned to the adult service.

Selection pressure on the hepatitis B virus pre-S/S and P open reading frames in Tongan subjects with a chronic hepatitis B virus infection

Identification of the full repertoire of hepatitis B virus (HBV) peptides that are presented to CD8+ T cells by common HLA class I alleles will be useful for designing immunotherapies for chronic hepatitis B. One hundred and seventy five cloned sequences containing the pre-S/S and P open reading frames (ORF) of the HBV were obtained from serum HBV-DNA of HBeAg-positive (n=4) and HBeAg-negative (inactive healthy carriers (IHC), n=16) Tongan subjects with an inactive chronic HBV infection. In addition, 34 and 32 sequences were obtained 5.2±1.4 (mean±SD) years apart from eight subjects. PAML was used to identify codons in the pre-S/S and P ORFs that were under positive selection pressure (ω>1). The number of non-synonymous substitutions in these codons was compared in IHC who were homozygous for either HLA-B∗4001 (n=9) or HLA-B*5602 (n=7), and who were either positive (n=6) or negative (n=10) for HLA-A*02. 34 codons in the pre-S/S and 11 codons in the P ORFs were under positive selection pressure. There was a higher number of non-synonymous substitutions in these codons in HBeAg-negative versus HBeAg-positive subjects in the P (p=0.02) but not the pre-S/S (p=0.64) ORF. There was no association between any HLA class I allele and non-synonymous substitutions in these codons. There was no increase in positive selection pressure on the pre-S/S and P ORFs with time. In conclusion, we could not find HLA class I-restricted selection pressure on any pre-S/S or P ORF amino acid; raising the possibility that peptide-based immunotherapies for chronic hepatitis B may not require peptides from these ORFs.

Evidence for reduced selection pressure on the hepatitis B virus core gene in hepatitis B e antigen-negative chronic hepatitis B

The mechanisms underlying the high levels of hepatitis B virus (HBV) replication that cause hepatitis B e antigen (HBeAg)-negative chronic hepatitis B (e−CHB) are unknown. Impaired anti-HBV immunity, which may be measurable as a relaxation of selection pressure on the virus, is possible. A group of Tongans (n = 345) with a chronic HBV infection, including seven with e−CHB, were genotyped at HLA class I. The repertoire of HBV core-gene codons under positive selection pressure was defined by phylogenetic analysis (by using the paml program) of 708 cloned sequences extracted from the 67 of these 345 subjects with the same repertoire of HLA class I alleles as the seven e−CHB individuals and matched controls (see below). The frequency of non-synonymous mutations at these codons was measured in longitudinal data from 15 subjects. Finally, the number of non-synonymous mutations at these codons was compared in seven groups comprised of one subject with e−CHB and 1–3 HLA class I-matched controls with an inactive, HBeAg-negative chronic HBV infection (e−InD). Nineteen codons in the core gene were under positive selection pressure. There was a high frequency of new non-synonymous mutations at these codons (P<0.0001) in longitudinal data. The mean number of these 19 codons with non-synonymous mutations was lower (P = 0.02) in HBV from subjects with e−CHB (4.4±0.5 codons per subject) versus those with e−InD (6.4±0.4 codons per subject). There is a subtle relaxation in selection pressure on the HBV core gene in e−CHB. This may be due to impaired antiviral immunity, and could contribute to the high levels of viral replication that cause liver inflammation in this disease.

Biliary complications following orthotopic liver transplantation: a 10-year audit

BACKGROUND

Biliary complications following liver transplantation result in major morbidity. We undertook a 10-year audit of the incidence, management and outcomes of post-transplant biliary complications at the New Zealand Liver Transplant Unit.

METHODS

Prospectively collected data on 348 consecutive liver transplants performed between February 1998 and October 2008 were reviewed. The minimum follow-up was 6 months.

RESULTS

A total of 309 adult and 39 paediatric transplants were performed over the study period. Of these, 296 (85%) were whole liver grafts and 52 (15%) were partial liver grafts (24 split-liver, eight reduced-size and 20 live-donor grafts). There were 80 biliary complications, which included 63 (18%) strictures and 17 (5%) bile leaks. Partial graft, a paediatric recipient and a Roux-en-Y biliary anastomosis were independent predictors of biliary strictures. Twenty-five (40%) strictures were successfully managed non-operatively and 38 (60%) required surgery (31 biliary reconstructions, three segmental resections and four retransplants). Seven (41%) bile leaks required surgical revision and 10 (59%) were managed non-operatively. There was no mortality related directly to biliary complications.

CONCLUSIONS

Biliary complications affected one in five transplant recipients. Paediatric status, partial graft and Roux-en-Y anastomosis were independently associated with the occurrence of biliary strictures. Over half of the affected patients required surgical revision, but no mortality resulted from biliary complications.

Low-cost, simultaneous, single-sequence genotyping of the HLA-A, HLA-B and HLA-C loci

New automated DNA sequencing technology has enabled the development of an assay for genotyping the three major HLA class 1 loci from a single sequence of each gene containing exon 3, intron 2 and exon 2. The assay allows 31 subjects (with 3 negative controls) to be genotyped at all three loci simultaneously, using a 96-well plate format. Genotypes were assigned by comparing each sequence to a database of 307 HLA-A, 563 HLA-B and 166 HLA-C alleles. Unequivocal, 4-digit allele assignments were made for 40 of 130 HLA-A genes, 82 of 130 HLA-B genes and 97 of 130 HLA-C genes from 21 European, 20 Tongan and 24 Niuean subjects. Ambiguity in interpretation of the sequence contributed to 66 of the 170 equivocal allele assignments, and 105 equivocal assignments were due to polymorphisms outside exons 2 and 3. All known alternative interpretations of ambiguous genotypes were identified, and seven HLA-B and two HLA-C ambiguities were resolved by reading the out-of-phase exon 2 sequence that followed an indel in intron 2. The genotypes of a subgroup of 27 heterozygous subjects, whose genotypes contained all of the alleles identified in this study, were confirmed with commercial, generic PCR-SSP typing. In European subjects, the repertoire of HLA-B/HLA-C haplotypes was almost identical to previously published data. We identified five new HLA-B/HLA-C haplotypes in the Polynesian subjects, and the remaining haplotypes were of Asian origin. In summary, we are describing a low-cost, sequencing assay for the three major HLA class I loci that provides a level of resolution that is comparable with a commercial PCR-SSP assay.

Genetic Diversity and Linkage Disequilibrium in the Polynesian Population of Niue Island

Isolated populations that recently have been derived from small homogeneous groups of founders should have low genetic diversity and high levels of linkage disequilibrium and should be ideal for mapping ancestral polymorphisms that influence complex genetic disease susceptibility. Populations that fulfill these criteria have been difficult to identify. We have been looking for Polynesian populations with these characteristics, because Polynesians have high rates of complex genetic diseases. In Niue Islanders all ancestral female (mitochondrial HSVI sequence) and 90.4% of ancestral male (Y-chromosome haplogroup) lineages are of Southeast Asian origin. The frequency of European Y-chromosome haplogroups is 7.2%. The diversities of mitochondrial HSV1 sequences (h = 0.18 +/- 0.05) and Y-chromosome haplo-groups (h = 0.18 +/- 0.05) are lower than values published for any other population. Ten autosomal microsatellites spaced over 5.8 cM show low allele numbers in Niue Islanders relative to Europeans (55 vs. 88 total alleles, respectively) and a modest reduction in heterozygous loci (0.71 +/- 0.02 vs. 0.78 +/- 0.02, p = 0.04). The higher linkage disequilibrium (d2) between these loci in Niue Islanders relative to Europeans (p = 0.001) is negatively correlated (r = -0.47, p = 0.01) with genetic distance. In summary, Niue Islanders are genetically isolated and have a homogeneous Southeast Asian ancestry. They have reduced autosomal genetic diversity and high levels of linkage disequilibrium that are consistent with the influence of genetic drift mechanisms, such as a founder effect or bottlenecks. High-powered linkage disequilibrium studies designed to map ancestral polymorphisms that influence complex genetic disease susceptibility may be feasible in this population.

Pre- and postoperative immunonutrition in patients undergoing liver transplantation: a pilot study of safety and efficacy

BACKGROUND & AIMS

Malnutrition is common in patients with end-stage liver disease and is a risk factor for post-transplant morbidity. The goal of this study was to assess the safety of an immune-enhancing diet in patients undergoing liver transplantation and to investigate its effects on nutritional status.

METHODS

Fifteen consecutive patients received oral Impact (0.6l/d) for a median 54 (range 10-168)d pre-transplant and enteral Impact was started early after transplant. Total body protein was measured prior to commencing supplemental Impact, immediately prior to transplant and 10, 15, 30, 90, and 180 days post-transplant. The results were compared with those from 17 patients who received standard nutritional intervention.

RESULTS

All study patients tolerated Impact pre- and postoperatively and there were no safety concerns. Over the preoperative period total body protein increased significantly (P = 0.017). In 7 patients followed for 6 months post-transplant, a significant (P = 0.026) loss of body protein occurred over the first 15 postoperative days which was regained by 6 months. In the patients who did not receive Impact, body protein did not change preoperatively and the loss after surgery was not regained. Infectious complications occurred in 5/15 (33%) Impact patients and 12/17 (71%) non-Impact patients (P = 0.074).

CONCLUSIONS

In patients with end-stage liver disease, our results suggest the possibility that Impact may have a role in improving preoperative nutritional status, hastening recovery after transplant, and reducing postoperative infectious complications. These potential benefits need to be confirmed in a randomised controlled trial.

Intragraft gene and protein expression in rat liver allografts treated with costimulatory blockade alone or in combination with CyA

BACKGROUND

Costimulatory blockade has been shown to prevent acute rejection (AR) and promote long-term graft survival in a number of animal models including nonhuman primates. The effect of concomitant administration of conventional immunosuppressives on long-term liver allograft survival and intragraft expression of immune mediators has not previously been examined.

MATERIALS AND METHODS

A high-responding Dark Agouti to Lewis orthotopic liver transplant (LEW OLT) model was used to compare anti-CD154 alone, or in combination with cyclosporin (CyA) on allograft survival. Donor-specific reactivity was assessed by mixed lymphocyte reaction (MLR) and allogeneic skin grafts. Surviving rats were euthanized on day 150 and intragraft gene (CD80, 86, 152, 154, IFN-gamma, IL-2, IL-6, IL-10, IL-13, TNF-alpha, TGF-beta, IL-7, Fas-ligand, Granzyme B, bax, and bcl(2)) and protein (CD4, CD8, ED1, CD154, CD80, CD86) expression was measured.

RESULTS

Untreated control recipients had a median survival time of 5 days. Recipients treated with anti-CD154 survived to beyond 150 days with no evidence of AR. Concomitant administration of CyA did not alter the long-term survival. There was no difference in the serum aspartate aminotransferase between treatment groups or a change over time. All treated recipients showed a reduction in donor-specific MLR at day 40 and 60 but had persistence of donor reactivity to skin grafts at day 100. Histologically, liver architecture was well preserved despite the presence of a nondestructive mononuclear cell infiltrate. Analysis of intragraft gene expression revealed an inverse relationship between the duration of anti-CD154 therapy and the gene expression of costimulatory molecules and Th1 cytokine transcripts. The pro-apoptotic gene, bax, was increased in recipients treated with anti-CD154, but not CyA, compared with normal liver.

CONCLUSIONS

These data demonstrate that anti-CD154 therapy either alone or in combination with CyA allows for the long-term survival of liver allografts in the rat despite there being a difference in the intragraft gene and protein profile.

Analysis of intragraft gene and protein expression of the costimulatory molecules, CD80, CD86 and CD154, in orthotopic liver transplant recipients

CD40-CD154 and/or CD28-CD80/86 costimulatory blockade induces long-term allograft survival in numerous animal models. Studies examining the expression of costimulatory molecules during acute cellular rejection (ACR) have been limited to renal and cardiac allografts. The aim of this study was to describe the relationship between intragraft costimulatory molecule expression in OLT recipients and ACR. Forty-five liver biopsies were obtained at reperfusion and day 7. Gene and protein expression of CD80, CD86 and CD154 were analyzed by RT-PCR and immunohistochemistry. CD154 protein expression was present in 13 of 18 patients with a RAI score of 4, but in only two of 14 patients with a RAI score of <4. There was a strong association between the RAI score and the presence of CD80 and CD154 immunoreactivity. CD86 protein expression did not correlate with the severity of ACR. In reperfusion biopsies CD154, but not CD80 or CD86, protein expression correlated with the total ischaemic time. There was no association between expression of costimulatory molecule genes and ACR. In conclusion, we have demonstrated an association between CD154 and CD80 protein expression and ACR in orthotopic liver allografts.

Costimulatory blockade prevents early rejection, promotes lymphocyte apoptosis, and inhibits the upregulation of intragraft interleukin-6 in an orthotopic liver transplant model in the rat

Costimulatory pathways have a pivotal role in the T-cell response to alloantigen. The role of costimulatory blockade with anti-CD154 in orthotopic liver transplantation (OLT) has not been examined previously. This study aims to investigate effects of anti-CD154 and CTLA4-immunoglobulin (Ig) in the early post-OLT period using a major histocompatibility complex-disparate fully arterialized OLT model in the rat. Lewis rats underwent OLT with Dark Agouti liver allografts. Recipients were randomized to receive (1) isotype control, (2) anti-CD154, (3) CTLA4-Ig, or (4) cyclosporine A (CyA). Rats were killed day 8, and specimens were obtained for histological examination, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling, immunohistochemistry, and quantitative reverse-transcriptase polymerase chain reaction. An additional five transplant recipients were treated with anti-CD154 for 14 days postoperatively to assess long-term allograft survival. All isotype control animals died on or before day 6 of acute rejection. Apart from four deaths caused by nonimmunologic causes, all treated recipients survived to day 8. The median survival of rats treated for 14 days with anti-CD154 was greater than 150 days. Serum aspartate aminotransferase and bilirubin levels normalized by day 3 in the CyA group and day 5 in transplant recipients treated with costimulatory blockade. Histologically, there was no difference between isotype controls and CTLA4-Ig-treated animals, whereas anti-CD154-treated transplant recipients had a lower Banff score. CD4+ and CD8+ T-cell infiltrates were prominent in transplant recipients treated with costimulatory blockade. Intragraft analysis showed an increase in lymphocyte apoptosis, Fas ligand messenger RNA expression, and reduction in interleukin-6 gene expression in transplant recipients treated with costimulatory blockade. Costimulatory blockade did not alter intragraft gene expression of other mediators of T-cell priming, differentiation, and effector function compared with isotype control animals. In conclusion, costimulatory blockade prevented acute rejection, enabled long-term survival, and increased intragraft lymphocyte apoptosis in a high-responding rat OLT model.

Sequential changes in the metabolic response to orthotopic liver transplantation during the first year after surgery

OBJECTIVE

To quantify the sequential changes in the metabolic response occurring in patients with end-stage liver disease after orthotopic liver transplantation (OLT).

SUMMARY BACKGROUND DATA

Detailed quantification of the changes in energy expenditure, body composition, and physiologic function that occur in patients after OLT has not been performed. Understanding these changes is essential for the optimal management of these patients.

METHODS

Fourteen patients who underwent OLT for end-stage liver disease had measurements of resting energy expenditure, body composition, and physiologic function immediately before surgery and 5, 10, 15, 30, 90, 180, and 360 days later.

RESULTS

Resting energy expenditure was significantly elevated after surgery (24% above predicted), peaking around day 10 after OLT, when it averaged 42% above predicted. A significant degree of hypermetabolism was still present at 6 months, but at 12 months measured resting energy expenditure was close to predicted values. Before surgery, measured total body protein was 82% of estimated preillness total body protein. During the first 10 days after OLT, a further 1.0 kg (10%) of total body protein was lost, mostly from skeletal muscle. Only 54% of this loss was restored by 12 months. Significant overhydration of the fat-free body was seen before OLT, and it was still present 12 months later. Although significant losses of body fat and bone mineral occurred during the early postoperative period, only body fat stores were restored at 12 months. Both subjective fatigue score and voluntary hand grip strength improved rapidly after OLT to exceed preoperative levels at 3 months. At 12 months grip strength was close to values predicted for these patients when well. Respiratory muscle strength improved less markedly and was significantly lower than predicted normal levels at 12 months.

CONCLUSIONS

Before surgery, these patients were significantly protein-depleted, overhydrated, and hypermetabolic. After surgery, the period of hypermetabolism was prolonged, restoration of body protein stores was gradual and incomplete, and respiratory muscle strength failed to reach expected normal values. Our measurements indicate that OLT does not normalize body composition and function and imply that a continuing metabolic stress persists for at least 12 months after surgery.

Mycophenolate mofetil increases cytomegalovirus invasive organ disease in renal transplant patients

The impact of cytomegalovirus (CMV) infection post-transplantation is in part influenced by the degree of immunosuppression. While mycophenolate mofetil (MMF) does not increase the overall incidence of CMV infection, we have questioned whether or not it increases its severity. Using a case control study design in which 29 renal transplant patients developed CMV disease [17 (59%) of which received azathioprine (AZA) and 12 (41%) received MMF], increases in the frequency of organ involvement with CMV (58 vs. 18%; p = 0.03) and in the number of organs involved with CMV were noted in the MMF versus the AZA group (2.0 vs. 1.0; p = 0.015). These results indicate that the increased immunosuppressive activity of MMF impacts the morbidity of CMV infection, thus warranting the use of effective anti-CMV preventive regimens while patients are treated with MMF.

NOD background genes influence T cell responses to GAD 65 in HLA-DQ8 transgenic mice

The major histocompatibility complex (MHC) genes play a significant role in the predisposition to insulin-dependent diabetes mellitus or type 1 diabetes. HLA-DQ8 (DQB1*0302, DQA 1*0301) genes have been shown to have the highest relative risk for human type 1 diabetes. To develop a "humanized" mouse model of diabetes, HLA-DQ8 was transgenically expressed in mice lacking endogenous class II genes. Since non-MHC background genes of the NOD influence the disease process, AP"/DQ8 mice were mated with the NOD strain and backcrossed to generate Abeta degree/DQ8/NOD mice. These mice have DQ8 as the sole MHC class II restriction element with NOD background genes at the N 2 generation. The DQ8 transgenic mice were used to identify T cell epitopes on glutamic acid decarboxylase (GAD 65), an important putative autoantigen in type 1 diabetes. The NOD background genes strongly influenced antigen processing, that is, different T cell epitopes were generated from the processing of GAD 65 in vivo in the Abeta degree/DQ8 and in the Abeta degree/DQ8/NOD mice.

Is cytomegalovirus infection related to mycophenolate mofetil after kidney transplantation? A case-control study

Three multicenter studies have shown that the addition of mycophenolate mofetil (MMF) to an immunosuppressive regime consisting of cyclosporin A (CSA) and prednisone (PRED) decreases the incidence of acute rejection episodes when compared with azathioprine (AZA) or placebo (1-3). In those patients receiving 3 g/d of MMF, the highest dose used in the studies, there was a trend towards an increased incidence of cytomegaloviral sepsis (CMV). We postulated therefore that MMF may represent an independent risk factor for the development of CMV infection in patients receiving renal allografts and MMF at our institution. Having altered the triple drug regime from CSA, AZA (2-2.5 mg/kg/d) and PRED to CSA, MMF (2 g/d) and PRED in July 1995, we elected to study all patients undergoing kidney transplantation for the 33-month period January 1994-September 1996, by undertaking a case control analysis to determine independent risk factors for the development of CMV infection, as defined by CMV viremia or tissue-invasive CMV. Three CMV disease-free control patients were matched to each case, these patients having been randomly selected from the entire pool of patients in the observation period. There were 31 CMV case patients and 102 control patients. Univariate analysis indicated that gender, a concomitant pancreas transplant, acute rejection and CMV seropositivity in the donor were risk factors. However, multivariate analysis indicated that only acute rejection and donor CMV seropositivity were independently linked (p < 0.05) to CMV disease in this sample. Specifically, the odds ratio (OR) for CMV disease between MMF and AZA was 1.0 (95% confidence interval (CI): 0.46-2.18). Therefore, in this case control study we find no evidence that MMF at a dose of 2 g/d is an independent risk factor for primary CMV viremia or tissue invasion in renal allograft recipients.

Diabetes mellitus and difficult laryngoscopy in renal and pancreatic transplant patients

Limited mobility of the cervical spine or temperomandibular joint may contribute to increased difficulty of laryngoscopy in patients who have severe diabetes mellitus. The frequency of difficult laryngoscopy in diabetics undergoing renal and/or pancreatic transplants has been reported to be as high as 32%. We retrospectively reviewed the anesthetic records of all adult patients who underwent renal and/or pancreatic transplant and endotracheal intubation from January 1, 1985 to October 31, 1995. Characteristics specifically reviewed included the presence of diabetes mellitus, type of organ donor, age, gender, body mass index, previous difficult laryngoscopy, known characteristics potentially related to difficult laryngoscopy, and degree of difficulty with laryngoscopy. Laryngoscopy was graded as easy, minimally to moderately difficult, and moderately to extremely difficult to perform. Factors associated with any degree of difficult intubation were univariately assessed by using Fisher's exact test. Of 725 patients, 15 (2.1%) were identified as having difficult laryngoscopies, although all underwent successful endotracheal intubations. Factors associated with difficult laryngoscopy were diabetes mellitus (P = 0.002) and characteristics known to be related to difficult laryngoscopy (P = 0.02). These findings confirm an increase in the frequency of difficult laryngoscopy in diabetic patients undergoing renal and/or pancreatic transplant, although no laryngoscopies were rated as moderately to extremely difficult. We conclude that the frequency of difficult laryngoscopy in these diabetic patients is much lower than previous reports have suggested.

IMPLICATIONS

Previous studies have suggested that airway management of many diabetic patients may be difficult. Our medical record review of patients with severe diabetes undergoing organ transplants showed that extraordinary techniques were not required to successfully manage their airways.

T cell recognition of human pre-proinsulin peptides depends on the polymorphism at HLA DQ locus: a study using HLA DQ8 and DQ6 transgenic mice

HLA DQ8 (DQ A1*0301/DQB1*0302) molecule is implicated in the susceptibility to insulin dependent diabetes mellitus whereas, HLA DQ6 (DQ A1*0103/DQB1*0601) molecule may have a protective effect. In this study we used mice transgenic to HLA DQ8 and HLA-DQ6 to elucidate the T cell determinants on a putative islet cell target antigen, insulin. These mice do not express endogenous mouse class II heterodimers on cell surface. Using overlapping synthetic peptides spanning the complete sequence of huma pre-proinsulin, we identified the sequences recognized by T cells in DQ8 transgenic mice and compared these to those in DQ6 transgenic mice. We observed a differential pattern of recognition of epitopes on human pre-proinsulin (HPI) polypeptide presented by the HLA DQ8 allele as compared to HLA DQ6. The sequences 1-24 and 44-63 were immunodominant in DQ8 transgenic mice while DQ6 transgenic mice primarily recognized sequences 14-33 and 74-93 of HPI. We found that the immune response generated in HLA DQ8 transgenic mice against HPI 1-24 cross-reacted to the mouse pre-proinsulin sequence 1-24. The T cell response were specifically inhibited using anti-CD4 and anti-DQ8 monoclonal antibodies. This cross-recognition of self sequences raises the possibility of modulation of experimental diabetes using this peptide.

Predictors of progression of diabetic nephropathy: implication for timing of kidney transplantation

Multiple risk factors are important predictors in the development of diabetic nephropathy (DN). Once DN has developed, it progresses steadily to renal failure. To determine the rate of renal function decline and the parameters that influence the rate of decline, we retrospectively reviewed the charts of patients with DN who had undergone dialysis or kidney transplantation at the Mayo Clinic from 1983 to 1993. Forty patients were found to have two or more iothalamate clearance (IothmCl) measurements where a slope of renal function decline over time, expressed as mL/ min/month/1.73 m2, can be calculated. The parameters examined included age of onset and duration of diabetes (DM); age at initial presentation, insulin dosage, glycosylated hemoglobin level, proteinuria, blood pressure (BP), number of antihypertensive medications (HTM), use of ACE inhibitors, creatinine, and initial IothmCl. The mean overall decline of clearance was 1.36 +/- 1.1 mL/min/month, corrected. Univariate regression analysis showed that only systolic and mean BP (p < 0.05), use of HTM (p = 0.02), and the number of HTM used (p = 0.0001) correlated with the rate of clearance decline. No other parameter was significant. The decline of IothmCl was 0.72 +/- 0.41, 1.20 +/- 0.9, and 2.34 +/- 1.38 mL/min/month, for patients taking no HTM, < 3 HTM, and > or = 3HTM, respectively. Of the eight patients on HTM who presented with initial IothmCl of < 30 mL/min/1.73 m2, seven (88%) had clearance of < 10 mL/min/1.73 m2 within 1 yr. We conclude that hypertension is an important marker of DN progression, and that the more HTM required for control of BP, the faster the decline of renal function. We recommend that a suitable transplant candidate with DN who presents with hypertension requiring HTM and a clearance of < 30 mL/min should be placed on the transplant waiting list.

Mechanisms of sulfonylurea's stimulation of insulin secretion in vivo: selective amplification of insulin secretory burst mass

Although sulfonylureas enhance insulin secretion, it is unknown whether these hypoglycemic chemicals stimulate insulin secretion through the augmentation of the pulsatile or basal modes of insulin release. Enhanced pulsatile insulin could occur in turn through amplification of the burst mass or an increase in burst frequency. To address the mechanism of sulfonylurea action, we employed a recently validated canine model with a portal vein sampling catheter and flow probe to measure pulsatile insulin secretion in vivo directly in response to tolbutamide infusion or ingestion. After a 16-h fast, seven dogs were studied in the postabsorptive basal state and during a tolbutamide (0.2 mg/min) infusion when their plasma glucose concentrations were clamped at euglycemia. Insulin concentrations in the carotid artery (basal vs. tolbutamide, 85 +/- 12 vs. 325 +/- 66 pmol/l; P < 0.01) and portal vein (basal vs. tolbutamide, 345 +/- 55 vs. 1,288 +/- 230 pmol/l; P < 0.01) increased during tolbutamide infusion, but the portal vein plasma flow did not change. Increased plasma insulin concentrations were achieved by a fourfold increase in the total insulin secretion rate (2.3 +/- 0.2 to 9.4 +/- 1.9 pmol x kg(-1) x min(-1); basal vs. tolbutamide, P < 0.01). The augmented total insulin secretion was achieved mechanistically via a marked and selective increase in the insulin secretory burst mass (basal vs. tolbutamide, 266 +/- 64 vs. 817 +/- 144 pmol/pulse; P < 0.01), with no change in portal-vein insulin pulse frequency (basal vs. tolbutamide, 10.1 +/- 0.6 vs. 11.1 +/- 0.8 pulses/h; P = 0.3). Oral (250 mg) tolbutamide also magnified the endogenous insulin secretion rate by the preferential amplification of the secretory pulse mass (basal vs. tolbutamide, 167 +/- 37 vs. 362 +/- 50 pmol/pulse; P < 0.01). Neither the infusion nor the ingestion of tolbutamide changed the calculated clearance rates of endogenously secreted insulin. We conclude that sulfonylurea (tolbutamide) induced insulin secretion in vivo is achieved by the highly selective amplification of insulin secretory burst mass with no change in basal insulin release or the frequency of the beta-cell-network pacemaker.

A multicenter trial of FK506 (tacrolimus) therapy in refractory acute renal allograft rejection. A report of the Tacrolimus Kidney Transplantation Rescue Study Group

A multicenter trial was conducted to evaluate the efficacy and safety of tacrolimus in the treatment of refractory renal allograft rejection. Renal transplant recipients experiencing biopsy-proven recurrent acute allograft rejection were eligible if the current rejection episode was refractory to corticosteroids. A total of 73 patients were enrolled, of whom 59 (81%) had previously received at least one course of antilymphocyte antibody as rejection therapy. One-year follow-up was available in 93% of patients. Median time to tacrolimus rescue therapy was 75 days after transplantation (range, 18-1448 days). Therapeutic responses to tacrolimus included improvement in 78% of patients, stabilization in 11%, and progressive deterioration in 11%. The risk of experiencing progressive deterioration was related to the pretacrolimus serum creatinine level: serum creatinine < or = mg/dl, 3%; 3.1-5 mg/dl, 16% (P < 0.04); > 5 mg/dl, 23% (P < 0.02). Twelve-month (from the time of initiation of tacrolimus therapy) actuarial patient and graft survival rates were 93% and 75%. Graft loss occurred in 19 patients (25%) at a median time of 108 days. Fourteen episodes of recurrent rejection were diagnosed in 10 patients (14%), at a median time of 101 days. Eleven episodes of recurrent rejection were treated (three patients underwent transplant nephrectomy), with resolution achieved in nine patients. Antilymphocyte antibody therapy was not used to treat recurrent rejection. Serum creatinine values improved during tacrolimus therapy: median serum creatinine level before tacrolimus, 3.2 mg/dl; median at 1 year after tacrolimus, 1.8 mg/dl. Twelve infections were documented in 11 patients (15%), including cytomegalovirus infection in three patients (4%). Posttransplant lymphoproliferative disorder was diagnosed in a single patient. Tacrolimus whole blood levels averaged 15.0 +/- 9.9 ng/ml at day 7 of tacrolimus therapy and 9.4 +/- 5.1 ng/ml at 1 year, and were consistent among individual centers. Treatment outcome did not correlate with tacrolimus blood levels. The most commonly observed adverse events were neurological and gastrointestinal. Seventy-four percent of patients received tacrolimus for at least 1 year. Tacrolimus therapy was discontinued in 18% of patients for rejection (11% for progressive, unrelenting rejection, and 7% for recurrent rejection). Tacrolimus therapy was discontinued in 8% of patients due to adverse events. In conclusion, tacrolimus rescue therapy provides (1) prompt, effective reversal of refractory renal allograft rejection, (2) good long-term renal allograft function, (3) a low incidence of recurrent rejection, and (4) an acceptable safety profile in renal allograft recipients experiencing refractory rejection.

Effects of somatostatin on pulsatile insulin secretion: elective inhibition of insulin burst mass

Although it is well known that somatostatin inhibits net insulin secretion, it is unknown whether this is achieved by regulation of the basal or pulsatile components of insulin secretion and, if the latter, whether this is through modulation of pulse mass or frequency. We addressed these questions with a canine model. Portal vein blood was sampled at 1-min intervals in five dogs for 60 min before (basal) and 90 min after ingestion of 30 g glucose on two different occasions, during a saline (SAL) or a somatostatin (SMS, 175 ng/min) infusion. Plasma glucose concentrations were similar during SAL and SMS. SMS had no effect on pulse frequency before (8.4 +/- 0.7 vs. 9.2 +/- 1.0 pulses/h, SMS vs. SAL, P = 0.54) or after glucose (13.3 +/- 1.1 vs. 11.6 +/- 0.9 pulses/h, SMS vs. SAL, P = 0.22). In contrast, SMS decreased insulin pulse mass in the postabsorptive (84 +/- 28 vs. 214 +/- 73 pmol/pulse, SMS vs. SAL, P < 0.05) and fed states (676 +/- 143 vs. 913 +/- 183 pmol/pulse, SMS vs. SAL, P < 0.05). In the postabsorptive state, SMS decreased insulin clearance by approximately 50% (0.32 +/- 0.04 vs. 0.60 +/- 0.09 l/min, P < 0.05), but after glucose ingestion, insulin clearance was comparable during SMS or SAL (0.72 +/- 0.04 vs. 0.80 +/- 0.08 l/min, P = 0.4). SMS appeared to alter insulin clearance through modulation of insulin pulse amplitude, because in the postabsorptive state clearance was closely correlated to the pulse amplitude (r = + 0.87, P < 0.0001). In conclusion, somatostatin regulates the rate of insulin secretion by selective inhibition of pulsatile insulin secretion. Regulation of secretory burst mass (and amplitude) may secondarily influence transhepatic and thus total body clearance of endogenously secreted insulin and thereby serve as a novel mechanism to dictate the systemic insulin concentration.

The effect of somatostatin 201-995 on the early course of porcine pancreaticoduodenal allotransplantation

This study was undertaken to determine the effects of somatostatin 201-995 (SMS) on the maintenance dose of intravenous cyclosporine and on graft blood flow, exocrine secretion, and rejection after porcine pancreaticoduodenal allotransplantation (PDA). For seven days, 12 pigs (6 control, 6 SMS-treated) were studied to determine the effects of SMS on serum CsA concentrations. Twenty-six pigs (14 control, 12 SMS) with streptozocin-induced diabetes underwent PDA. Blood flow was measured through graft celiac and superior mesenteric arteries 15 and 60 min after reperfusion. SMS (75 micrograms s.c.) was given after the 15-min blood-flow measurement in the SMS group. Sixteen pigs (8 control, 8 SMS) were followed postoperatively with daily measurements of serum glucose and amylase concentrations, and urine amylase and trypsin activities. All pigs were immunosuppressed with azathioprine, prednisone, and i.v. CsA. SMS pigs also received SMS (75 micrograms s.c.) every 8 hr. SMS had no effect on maintenance dose of CsA or on serum amylase, urine amylase, or urine trypsin activities. Mean days to rejection were also not affected. Intraoperative graft blood flow was significantly decreased by SMS, but incidence of graft thrombosis was unchanged. These results suggest that in the porcine PDA model, SMS does not appear to inhibit exocrine secretion and potentially may adversely affect the early course of PDA by decreasing graft blood flow.

Purified canine islet autografts. Functional outcome as influenced by islet number and implantation site

Using a modification of the basic principles of pancreatic intraductal collagenase digestion and density gradient purification to isolate canine islets, in conjunction with simultaneous fluorogenic and dithizone islet staining, we quantified the yield, purity, and viability of the isolated islets. We then determined the combined influences of total and weight-corrected islet counts and implantation site on immediate and long-term functional outcome of purified canine islet autografts. Weight-corrected islet counts were 100% sensitive and specific in differentiating successful and unsuccessful islet autografts implanted to the liver (n = 10) and spleen (n = 10) of pancreatectomized dogs. The threshold number of islets required to achieve normoglycemia in the liver (4400 islets/kg) and spleen (4650 islets/kg) were nearly identical. Islet autografts failed to ameliorate hyperglycemia when implanted to the renal subcapsular space (n = 5) at counts of 4400 to 5500 islets/kg. The mean one- and three-month intravenous glucose tolerance test K-values of dogs with purified islet autografts to the liver (-1.43 +/- 0.27 and -1.69 +/- 0.27, respectively) and spleen (-1.78 +/- 0.36 and -1.64 +/- 0.3, respectively) were also similar. Time needed to achieve normoglycemia , however, was significantly (P less than 0.02) shorter for intrahepatic islets (1.0 +/- 0.0 days posttransplant) than intrasplenic islets (6.8 +/- 2.3 days posttransplant). The long-term durability of islet autograft function was not unlimited. Overall, thirteen canine islet autograft recipients have been followed for greater than or equal to 12 months posttransplant (range 12-18 months), seven canine islet autograft recipients (five intrahepatic and two intrasplenic) have had spontaneous recurrence of hyperglycemia at 2, 6, 11, 13, 14, 8, and 16 months, respectively. The phenomenon depended only on the number of islets implanted. The data underscore the significance of quantitatively defined islet preparations and the importance of islet number and implantation site on immediate and long-term functional outcome of canine islet autografts.

Histologic diagnosis of rejection by using cystoscopically directed needle biopsy specimens from dysfunctional pancreatoduodenal allografts with exocrine drainage into the bladder

To determine the histologic features of rejection and to identify nonrejection causes of human pancreatic allograft dysfunction, we analyzed 31 needle biopsy specimens (17 pancreatic, 14 duodenal) obtained under cystoscopic direction from 15 dysfunctional pancreatoduodenal allografts with exocrine drainage into the bladder. Eight allografts undergoing rejection showed the most common histologic features of rejection to be diffuse mixed inflammatory infiltrates of pancreatic acinar tissue and duodenum wall. Diffuse infiltration of pancreatic acinar tissue by neutrophils was the earliest histologic change in rejection. Seven dysfunctional allografts not undergoing rejection ("nonrejection") showed a normal pancreas or various changes including acinar dilation with inspissation of secretions, fibrosis, cytomegalovirus inclusions, and enzymatic necrosis. The histologic changes in the duodenum paralleled those in the pancreas in both rejection and nonrejection allografts. We conclude that the histologic features of rejection in pancreatoduodenal allografts are distinctive. The changes seen in biopsy specimens accurately reflect the state of the graft and can be used to diagnose rejection and to identify other causes of graft dysfunction. Biopsy samples from the duodenum as well as the pancreas are diagnostically useful. The biopsy findings can be used to guide the clinical management of rejection and in the development of other noninvasive tests for rejection.

Pancreas transplantation at Mayo: II. Operative and perioperative management

Better perioperative and operative management techniques have contributed to an improvement in the success rate of pancreas transplantation. Because of a shortage of donor organs, the criteria for acceptability of the allograft have been liberalized, and the development of techniques such as combined liver and pancreas procurement has increased allograft availability. Major advances have been made in organ preservation. Currently, pancreas allografts can routinely be stored for 18 to 24 hours. The technique of pancreaticoduodenal transplantation with a duodenocystostomy for the exocrine drainage is widely used. Experience with anesthetic and intensive-care unit management of these patients is accumulating. With the evolution of pancreas transplantation and with the help of the excellent transplant centers in our area, we developed a pancreas transplantation protocol and performed transplantation based on this protocol in 16 recipients at the Mayo Clinic from October 1987 through December 1988.

Pancreas transplantation at Mayo: III. Multidisciplinary management

Although pancreas transplantation is a complicated procedure, a good level of success has been achieved because of the introduction of cyclosporine for immunosuppression, improved methods for diagnosing rejection, and a multidisciplinary approach to management. Our immunosuppressive regimen was quadruple therapy with induction by using Minnesota antilymphoblastic globulin. A biopsy technique was instituted in which the pancreas specimens were obtained under cystoscopic direction during episodes of hypoamylasuria. The criteria for rejection episodes were not only biochemical abnormalities but also histologic confirmation and a follow-up to exclude other causes of graft dysfunction. Infectious disease management included use of oral selective bowel decontamination for 3 weeks after transplantation. At the Mayo Clinic between October 1987 and December 1988, 16 patients received pancreaticoduodenal allografts (both kidney and pancreas in 13 and pancreas only in 3 after a prior successful kidney transplantation). In two pancreas and one kidney allograft, function was lost. One patient died of multiorgan failure. The cystoscopically directed biopsy technique was performed 23 times with minimal complications and a 93% success rate. The metabolic results have been excellent; the glycosylated hemoglobulin level was normal 3 to 6 months after transplantation. The quality of life was significantly improved in almost all patients. Nutritional assessment revealed little deterioration after transplantation. With a multidisciplinary approach, the needed answers about the effect of pancreas transplantation on the degenerative complications of diabetes should be forthcoming.