Telephone: 34 932607940; FAX: 34 932607603
Impact of immunosuppression without steroids on rejection and hepatitis C virus evolution after liver transplantation: Results of a prospective randomized study
Article first published online: 24 NOV 2008
Copyright © 2008 American Association for the Study of Liver Diseases
Volume 14, Issue 12, pages 1752–1760, December 2008
How to Cite
Lladó, L., Fabregat, J., Castellote, J., Ramos, E., Xiol, X., Torras, J., Serrano, T., Baliellas, C., Figueras, J., Garcia-Gil, A. and Rafecas, A. (2008), Impact of immunosuppression without steroids on rejection and hepatitis C virus evolution after liver transplantation: Results of a prospective randomized study. Liver Transpl, 14: 1752–1760. doi: 10.1002/lt.21629
- Issue published online: 24 NOV 2008
- Article first published online: 24 NOV 2008
- Manuscript Accepted: 19 JUL 2008
- Manuscript Received: 2 APR 2008
- TV3 Marathon Foundation
The purpose of this study was to evaluate the influence of a steroid-free immunosuppression on hepatitis C virus (HCV) recurrence. A total of 198 liver transplantation (LT) patients were randomized to receive immunosuppression with basiliximab and cyclosporine, either with prednisone (steroid [St] group) or without prednisone (no steroids [NoSt] group). The group of 89 HCV-infected patients was followed up with protocol biopsies for 2 years after LT. This group of HCV patients are the patients evaluated in the present study. The rejection rate was 19% (St: 21% versus NoSt: 17%; P = 0.67). Patients in the St group had a slightly higher rate of bacterial infections (59% versus 38%; P = 0.05). Almost all patients had histological HCV-recurrence (St: 39/40 (97%) versus NoSt: 40/41 (97%); P = 1). The percentage of accumulated biopsies with grade 4 portal inflammation at 6 months, 1 year, and 2 years were, 23%, 49%, and 49% in the NoSt group, compared to 33%, 55%, and 69% in the St group, respectively (P = 0.04 at 2 years). The percentage of accumulated biopsies with grade 3 or 4 fibrosis at 6 months, 1 year, and 2 years were 0%, 8%, and 22% in the NoSt group, compared to 8%, 19%, and 31% in the St group, respectively. Immunosuppression without steroids in HCV patients is safe, reduces bacterial infections and metabolic complications, and improves histological short-term evolution of HCV recurrence. Liver Transpl 14:1752–1760, 2008. © 2008 AASLD.
Hepatitis C virus (HCV)-related cirrhosis and its complications are the leading cause of liver transplantation (LT). Disease recurrence has become the main concern in the management of patients having undergone transplantation.1, 2 Recent data have demonstrated a negative impact of HCV infection on both graft and patient survival.3, 4
Several factors have been implicated in the frequency and severity of recurrent HCV liver disease.5, 6 Potent immunosuppression has been blamed for the worse outcome of HCV recurrence,6, 7 although the meaning of “potent” remains open to interpretation. It seems that changes in the degree of immunosuppression rather than its absolute amount have a negative effect in terms of the evolution of HCV recurrence.8 However, no specific agent has been convincingly demonstrated to be more disadvantageous than another. One of the most widely studied factors is steroid use, and it seems clear that the use of steroid boluses is harmful to HCV-infected patients9, 10; nevertheless, conclusions regarding the use of steroids for maintenance immunosuppression are less clear. Recent studies conclude that to improve the outcome of HCV-infected recipients, rapid steroid tapering should be avoided.11, 12
Immunosuppression without steroids is both safe and feasible.13–20 The influence on HCV recurrence has been investigated in few prospective studies,14, 16–19 and only 2 of these included protocol biopsies; follow-up was also limited to 1 year.18, 19 A recent meta-analysis has shown that HCV recurrence is lower with steroid avoidance, although no individual trial reached statistical significance. The small sample size of the trials, the lack of protocol biopsies in most of them, and the heterogeneity of trials precludes drawing definitive conclusions.21
We have previously demonstrated that a steroid-free immunosuppression protocol based on cyclosporine A (CyA) and basiliximab induction is safe and reduces infection and metabolic complications.20 The purpose of this further analysis was to evaluate the specific influence of this protocol on HCV recurrence under the same immunosuppression protocol. To test this we evaluated only HCV-infected patients over a 2-year follow-up with protocol biopsies.
PATIENTS AND METHODS
The study follows the Declaration of Helsinki concerning medical research, and its protocol was approved by the ethics committee at each center. Informed consent to being included in the study was obtained from each patient prior to LT.
The trial was an open-label, prospective, randomized study. Within each center, eligible patients were randomized at the beginning of surgery using sealed envelopes and stratification into 2 cohorts: HCV-negative and HCV-positive.
Patients fulfilling the inclusion criteria were recruited between April 2001 and September 2004. Eligible patients were 18 years of age or older who had undergone LT from a cadaveric donor.
Exclusion criteria were as follows: transplantation for fulminant liver disease, retransplantation, previous other organ transplantation, multiorgan transplantation, autoimmune hepatitis, biliary primary cirrhosis, human immunodeficiency virus (HIV) infection and unlikely compliance with schedule. Patients with autoimmune hepatitis and biliary primary cirrhosis were excluded because we were not confident of the results of not giving steroids after LT to these patients.
Preliminary results about all the patients included in this study have been published.20
Here we evaluate only the evolution of HCV-infected patients over a 2-year follow-up.
All patients received basiliximab (two 20 mg doses). Immunosuppression was maintained with CyA emulsion; CyA was initiated postoperatively at an oral dose of 10 mg/kg/day given twice daily. The dose was adjusted to maintain whole-blood therapeutic C2 levels of 800 to 1200 ng/mL. In the event of renal insufficiency (creatinine higher than 125 μmol/L or creatinine clearance lower than 40), and assuming that the patient had no pancytopenia (hematocrit >26% and platelet count >50,000 cells/mm3), mycophenolate mofetil (MMF) was initiated at a dose of 2000 mg/day, in which case the CyA dose was reduced by at least 50%.
Patients randomized to the steroid (St) group received methylprednisolone (500 mg) as a single intravenous bolus before reperfusion, followed by 0.5 mg/kg/day methylprednisolone until day 5, 0.25 mg/kg/day from day 5 to day 30, and 0.15 mg/kg/day from day 30 to day 90. Steroids were subsequently withdrawn. Patients randomized to the no-steroid (NoSt) group received no steroids after LT.
In the event of rejection (which was always confirmed by core liver biopsy)22 intravenous methylprednisolone therapy (bolus injection of 500 mg/day on 3 consecutive days) was administered. If liver function test results showed no improvement the rejection episode was considered as steroid-resistant and a switch to tacrolimus (0.1 mg/kg/day to achieve 5-10 ng/mL blood levels) was indicated (without any other steroids). In case of no improvement of rejection after steroid treatment, further biopsy was not mandatory as a protocol. However, especially in this HCV group of patients, in case of doubt, a new biopsy was performed before conversion.
Antibiotic, antifungal, and antiviral prophylaxis was administered according to local practice. The most common protocol was the administration of antibiotics (aztreonam and vancomycin) for only 24 hours, oral nistatine for the entire postoperative period, and no universal antiviral prophylaxis. Prophylaxis with valganciclovir 900 mg/24 hours (for 3 months) was administered only to mismatch patients (donor cytomegalovirus [CMV] +/recipient CMV−).
The primary endpoints of the study were biopsy-proven acute rejection incidence and patient and graft survival. The secondary endpoints were the incidence of adverse events, both infections and metabolic decompensations (mainly de novo diabetes mellitus and hypertension), and the incidence and severity of hepatitis C recurrence.
Patients were examined daily from day 0 until the day they left the hospital, with subsequent scheduled examinations at 1, 2, 3, 6, 12, and 24 months.
HCV Recurrence Assessment
HCV recurrence was assessed by protocol biopsies, laboratory data, and quantitative RNA blood levels. All these data, including biopsies, were per protocol performed 6, 12, and 24 months postoperation. Additional biopsies were performed when clinically required. Histological diagnosis of hepatitis C recurrence was graded according to Scheuer's histological grading of chronic hepatitis.24
All biopsies were evaluated by the local histopathologist following the same criteria.
All results from biopsies were analyzed by accumulation of the results.
The objective for the primary endpoints was to demonstrate the equivalence between groups, whereas that for the secondary endpoints was to demonstrate the superiority of the NoSt regime.
A sample size of 200 patients was calculated, and showed no difference between treatment arms for the acute biopsy-proven rejection rates, based on an estimated incidence of 35% for the St group, 5% level of significance, 80% power, and a 10% dropout rate.
Overall analyses have been reported.20 Here, we analyze the cohort of HCV-infected patients, with special attention being paid to protocol biopsies after a minimum follow-up of 2 years.
Analysis was performed on an intention-to-treat basis.
Continuous data were analyzed using the Student t test or nonparametric tests as indicated. Categorical data were analyzed using the chi-square test or Fisher's exact test as required. Analyses of survival were based on Kaplan-Meier estimates and survival curves were compared by means of the log-rank test. A P value <0.05 was considered significant.
Between April 2001 and September 2004, 200 patients were randomized. During this period, 40 patients (16.6%) were not randomized due to some exclusion criteria (4 transplants for fulminant liver disease, 21 retransplants, 1 previous kidney transplant, 6 associated kidney transplants, 3 cases of autoimmune hepatitis, 4 HIV infections, and 1 unlikely compliance with schedule). Two patients were excluded from the study after randomization because of protocol violations. Of these 198 patients, 89 (45%) were HCV-infected patients and formed the cohort of the present analysis. Forty-three were allocated to the NoSt group and 46 to the St group.
Patient demographics and baseline characteristics of the HCV-infected patients were similar between the groups (Table 1). The main operative and initial postoperative evolution was also similar between the groups (Table 2). Half the patients required MMF due to renal insufficiency at some point during follow-up (St: 25 [54%] versus NoSt: 20 [46%]; P = 0.46).
|No Steroids (n = 43)||Steroids (n = 46)||P|
|Mean age, mean ± SD (years)||53.2 ± 7.9||57.1 ± 7.9||0.04|
|Sex, male/female [n (%)]||27 (63)/16 (37)||30 (65)/16 (35)||0.82|
|BMI, mean ± SD (kg/m2)||25.9 ± 4.3||26.9 ± 3.5||0.07|
|Child-Turcotte-Pugh score, A/B/C||14/23/5||14/24/8||0.77|
|Time diagnosis (years)||6.8 ± 5||8.2 ± 7||0.54|
|RNA, mean ± SD||871,359 ± 1,255,688||744,168 ± 1,030,668||0.78|
|CMV+ [n (%)]||41 (95)||39 (85)||0.16|
|DM [n (%)]||9 (21)||12 (26)||0.57|
|Glycated Hb (%)||4.4 ± 0.9||4.6 ± 0.9||0.69|
|Hypertension [n (%)]||4 (9)||9 (19)||0.17|
|Cholesterol (mmol/L)||3.9 ± 0.9||3.8 ± 1.2||0.68|
|No Steroids (n = 43)||Steroids (n = 46)||P|
|Donor age, mean ± SD (years)||47.9 ± 16||51.4 ± 17.5||0.31|
|Ischemic time, mean ± SD (minutes)||419.8 ± 158||461.1 ± 149||0.19|
|RBC transfusion [n (%)]||20 (46)||27 (59)||0.25|
|Ventilation, mean ± SD (hours)||29 ± 68||35 ± 107||0.55|
|Dialysis [n (%)]||1 (2)||2 (4)||1|
|Hospital stay, mean ± SD (days)||15.6 ± 9||14.5 ± 9||0.75|
|ICU stay, mean ± SD (days)||5 ± 6.5||5.14 ± 5.4||0.94|
|Hospital mortality [n (%)]||0 (0)||1 (2)||1|
The overall incidence of biopsy-proven acute rejection was 19%, with no differences between the groups (St: 8/46 [17%] versus NoSt: 9/43 [21%]; P =.67). Eight out of 17 patients with acute rejection (47%) had steroid-resistant rejection (St: 5/8 [60%] versus NoSt 3/9 [33%]; P = 0.32).
The incidence of ductopenic rejection was also similar between the groups (St: 2/46 [4%] versus NoSt: 2/43 [5%]; P = 1.00).
Actuarial patient and graft survival rates were similar between groups (Fig. 1A,B). At 2-year follow-up there were 9 deaths in the NoSt group (21%) compared to 15 deaths in the St group (33%) (P = 0.21). Causes of death were as follows: in the St group: 4 infection, 1 rejection, 2 cardiovascular-neurological, 1 primary nonfunction, 4 HCV recurrence, and 3 other causes; in the NoSt group: 2 infection, 1 vascular, 2 rejection, 1 cardiovascular-neurological, 1 HCV recurrence, and 2 other causes.
The overall incidence of infection was 50.5% (Table 3); bacterial infection was slightly higher in the St group. Viral and fungal infection rates were similar between the groups.
|No Steroids (n = 43)||Steroids (n = 46)||P|
|Infection||23 (53)||31 (67)||0.17|
|Bacterial infection||16 (38)||271 (59)||0.05|
|Viral infection||9 (21)||12 (26)||0.56|
|CMV infection*||1 (2)||4 (9)||0.24|
|Herpes||9 (21)||10 (22)||0.923|
|Fungal||2 (5)||4 (9)||0.44|
|De novo DM|
|Initial||8 (19)||14 (31)||0.19|
|Maintenance||6 (18)||6 (20)||0.85|
|Glycated hemoglobin, month 3||4.5 ± 0.9||5.1 ± 1.1||0.03|
|Glycated hemoglobin, month 6 (%)||4.9 ± 1.2||4.9 ± 0.8||0.57|
|Glycated hemoglobin, month 6 (%)||4.9 ± 1.2||4.9 ± 0.8||0.57|
|Glycated hemoglobin, month 12 (%)||4.6 ± 1.1||4.6 ± 0.7||0.60|
|Glycated hemoglobin, month 24 (%)||4.9 ± 1.0||4.4 ± 0.8||0.29|
|De novo hypertension|
|Initial||12 (28)||23 (50)||0.03|
|Maintenance||19 (59)||15 (51)||0.54|
|Initial||15 (35)||20 (43)||0.40|
|Maintenance||13 (31)||21 (46)||0.15|
|Cholesterol, month 3 (mmol/L)||4.9 ± 1.8||5.6 ± 2.5||0.10|
|Cholesterol, month 6 (mmol/L)||4.3 ± 1.1||4.7 ± 0.9||0.07|
|Cholesterol, month 12 (mmol/L)||4.6 ± 1.2||4.8 ± 1.7||0.73|
|Cholesterol, month 24 (mmol/L)||4.3 ± 1.6||4.8 ± 1.7||0.42|
|Triglycerides, month 3 (mmol/L)||1.7 ± 1.0||2.2 ± 1.1||0.02|
|Triglycerides, month 6 (mmol/L)||1.5 ± 0.8||1.8 ± 0.7||0.06|
|Triglycerides, month 12 (mmol/L)||1.6 ± 0.8||1.8 ± 0.7||0.06|
|Triglycerides, month 24 (mmol/L)||1.2 ± 0.3||1.9 ± 0.4||0.07|
|Neurological||12 (29)||12 (26)||0.74|
|De novo tumor||3 (7)||3 (6.5)||1|
The main adverse events are shown in Table 3. The metabolic profile was worse in the steroid group.
HCV Recurrence Analyses
When all the biopsies performed were considered, both those under protocol and those indicated clinically, nearly all of the biopsies demonstrated evidence of HCV recurrence at the 2-year follow-up, and nearly all of the patients showed HCV recurrence at the 2-year follow-up (St: 39/40 [97%] versus NoSt: 40/41 [97%]; P = 1.00). Six patients in the St group and 2 in the NoSt group were never biopsied because they did not reach the 6-month follow-up point. All the other patients were biopsied at least once.
The accumulated results of biopsies are shown in Table 4. Portal and lobular activity was significantly higher in the St group at 24 months. However, a further analysis focusing on more severe evolution showed that the evolution of both inflammatory activity and fibrosis was worse in the St group, although not statistically significant. The number of patients with grade 4 portal inflammation and those with grade 3 or 4 fibrosis was higher in the St group (Fig. 2). The percentage of patients with grade 4 portal inflammation was 23, 49, and 49 at 6, 12, and 24 months, respectively, in the NoSt group, whereas in the St group the number of patients with grade 4 portal inflammation was 33, 55, and 69 at 6, 12, and 24 months, respectively. This difference was significant at the 24-month follow-up.
|No Steroids (n = 43)||Steroids (n = 46)||P|
|Histological recurrence||39/43 (97)||40/46 (97)||1|
|Month of diagnosis||5.6 ± 2.1||5.4 ± 2.0||0.84|
With respect to fibrosis, 0%, 8%, and 22% of patients in the NoSt group had grade 3 or 4 fibrosis at 6, 12, and 24 months, respectively, whereas in the St group the corresponding figures were 8%, 19%, and 31% of patients at 6, 12, and 24 months, respectively (P = not significant).
There were no statistical differences on viremia (Fig. 3).
Overimmunosuppression is believed to be the principal factor in determining the severity of recurrent HCV liver disease, and thus the poorer outcome of patients after LT. However, there is no consensus about the agent or regime responsible for the severity of evolution of HCV recurrence.8 In this regard, the First International Liver Transplantation Society Consensus Conference published in 200325 concluded that additional studies are needed to define the optimal immunosuppressive strategy in HCV-infected liver transplant recipients.
The role of complete avoidance of steroids after LT in HCV-infected recipients has yet to be clearly demonstrated.
In studying the role of steroid-free immunosuppression after LT, the first step is to demonstrate its feasibility. Randomized studies have demonstrated that steroid-free immunosuppression after LT is safe.13–20 In an earlier prospective randomized multicenter study,20 we compared 2 groups of patients who received immunosuppression based on CyA and basiliximab induction, with or without steroids. The results showed that immunosuppression without steroids is safe and reduces metabolic and infectious complications. A relevant finding of that study was that in HCV-infected patients immunosuppression without steroids did not increase the risk of rejection.
Once the equivalence of the steroid-free protocol in HCV-infected patients had been demonstrated, the next step was to determine whether this type of immunosuppression had an influence on the evolution of HCV recurrence.
In line with our earlier published results, this further analysis, including only HCV patients, confirms that a steroid-free regime based on CyA and basiliximab induction is not associated with an increased rate of allograft rejection. The overall pattern of adverse events was worse in the St group: the incidence of bacterial infections was significantly greater. Although not significant due to the low number of events, the rate of CMV infection was higher in the St group (9% versus 2%).
The relevance of our study rests on 3 main factors: first, it was prospective and randomized; second, the follow-up was 2 years; and third, the evaluation was based on protocol biopsies, a step that is not included in many HCV-recurrence studies.26 The lack of correlation between liver enzymes and histological injury makes protocol biopsies indispensable when studying the evolution of HCV recurrence.
Our study has also some weaknesses. First, the small sample size. Perhaps with more patients some clinically relevant results, such as the differences in fibrosis, may had reached statistical significance. Second, some factors (CMV prophylaxis, biliary complications, and others) with possible influence on HCV recurrence were not evaluated. However, due to the fact that patients were randomized at the time of surgery, it can be assumed that these factors were homogeneously distributed between the groups. In fact, CMV seropositive patients before LT were similar between the groups (Table 1). It has to be noted that CMV infection was slightly more frequent in the St group (although not significant).
To date, few prospective randomized studies have evaluated the influence of steroid-free immunosuppression on HCV recurrence;16–19 only 2 studies18, 19 included protocol biopsies, and their follow-up was limited to 1 year. Filipponi et al.19 found no difference in HCV recurrence, but Kato et al.18 reported a trend toward a lower incidence of advanced fibrosis in 1-year biopsy samples among patients receiving the steroid-free protocol.
All except 2 of our patients who survived for more than 6 months had histological findings of HCV recurrence. The key here is to detect the group of patients with an early and aggressive pattern of recurrence, as it will be they who develop cirrhosis earlier. In this regard, the degree of fibrosis has proven to be the most relevant factor in predicting evolution to cirrhosis3, 27 and it correlates with impaired graft survival.28 Initial necroinflammatory activity may also be of some relevance.27, 28
Our results confirm those of Neumann et al.28 showing that the use of protocol biopsies reveals a rapid and progression of fibrosis mainly during the first year after transplantation. This progression is even more relevant in the St group; however, at the 6-month follow-up there were more patients with grade 3 or 4 fibrosis in the St group and this was confirmed at the 12-month and 24-month biopsies. It should be noted that in the 12-month biopsies (accepted by most authors as an important prognostic factor of poor evolution),27–29 3 out of 39 patients (8%) from the NoSt group had severe fibrosis compared to 7 out of 38 patients (18%) from the St group. Although this difference did not reach statistical significance, we nevertheless regard it as a very relevant clinical finding.
Overall inflammatory activity was worse in the St group. The most significant finding was that grade 4 portal inflammation was higher in the St group throughout the follow-up period. The influence of portal activity as a prognostic factor in the evolution of HCV recurrence has been previously described as a marker of future fibrosis.27, 28
Our 2-year patient and graft survival rates were quite low. It has to be noted that all patients studied were HCV patients, nearly all genotype 1b, and with a median donor age of 51 years old. It has already been described the bad prognosis of similar cohort of patients; all these factors (HCV, genotype 1b, and high donor age) have been related with worse outcome after LT.1, 3, 5 Patient and graft survival rates were similar between groups. Only 1 out of 43 patients from the NoSt group (2.3%) compared to 4 out of 46 (8.6%) from the St group died from HCV-related causes. It has to be noted that in the steroid arm, the 2 leading causes of death were infections and HCV-recurrence (although this was not significant).
Further follow-up is required to determine whether the steroid-free regime will enable us to overcome the previously-described reduced survival rate of HCV-LT patients.3, 4 Our results suggest that both portal activity and fibrosis may be reduced by using a steroid-free immunosuppression regime, and hopefully this may improve the evolution of HCV recurrence. Longer follow-up will allow to draw more definitive conclusions about the influence of steroid avoidance on graft rejection and graft loss in the long term.
RNA levels increased similarly in both groups of patients after LT and, as described by Gane et al.,30 this increase occurred soon after LT. Subsequently, the levels remained stable in the NoSt group but increased further in the St group. The influence of RNA levels on the evolution of HCV recurrence has already been suggested by some authors,31, 32 although it has not been confirmed by others.33
Both the use of steroid bolus and tacrolimus conversion may have some influence on metabolism, infections, and HCV evolution (both RNA levels and histological evolution). Rejection rates, bolus use, and tacrolimus conversion rates were similar between both groups. Thus, we can assume that these factors have not introduced any bias in the analysis, and therefore differences between the study groups can be attributed to either the use of a steroid-free regimen or not.
HCV-recurrence treatment was not protocolized throughout the study. HCV treatment was administered according to local practice. The most common protocol was to treat RNA-positive patients, with at least fibrosis grade 1 (Scheuer's classification), alanine aminotransferase × 2, and no other contraindications. Although it was not fully analyzed, we can say that following these criteria 23 patients from the NoSt group were treated compared to 16 in the St group. Only 4 patients in the NoSt group and 6 in the St group showed virological response. Thus, after this 2-year analysis we think that antiviral treatment did not probably had influence in our results because rate of response was low and similar between groups.
Nowadays it is accepted that improving the outcome of HCV recurrence requires a reduction in overall immunosuppression, although the most important factor is to avoid abrupt variations.8, 11, 12 The alternative would be to give steroids for a long time, a strategy which has been shown to protect against the development of fibrosis.11, 12 However, this approach would probably lead to overimmunosuppression in patients and have a negative influence on important adverse events, for example, metabolic ones. Immunopathology studies suggest that continuous immunosuppression, without abrupt changes, may benefit the overall balance between virus distribution and immune responsiveness.34, 36 In our opinion, the best way to avoid abrupt variations in steroids would be not to give them at the outset.
In conclusion, steroid-free immunosuppression based on CyA and basiliximab induction in HCV-infected patients is safe and reduces both metabolic complications and bacterial infections. This steroid-free regimen allows a slightly lower rate of inflammatory activity and fibrosis 2 years after LT. The influence of these histological results on clinical evolution will be clarified by further clinical and histological follow-up. Longer follow-up and further studies with more number of patients will allow to evaluate the influence of these histological results on patient and graft loss, and define specific clinical guidelines.
The following investigators also participated in this study as part of the THOSIN Study Group: Ignacio Gonzalez-Pinto, Hospital Central de Asturias, Oviedo, Spain; Jose Ignacio Herrero, Hospital Universitario de Navarra, Pamplona, Spain; Dr. Carlos Blesa, Hospital Ramón y Cajal, Madrid, Spain; Dr. Diego Rincón, Hospital Gregorio Marañón, Madrid, Spain; Andrés Valdivieso, Hospital de Cruces, Bilbao, Spain. The acronym THOSIN stands for orthotopic LT without steroids, in Spanish: Trasplante Hepático Ortotópico sin Steroids. We thank Raquel Lastra for her help in data collection and Alan Nance for reviewing the English. We are grateful to Infociencia Clinical Research for monitoring the study, and we especially thank Cati Bonet for the statistical analysis.
- 23Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 1997; 20: 1183–1197.