10 Years of Pediatric Liver Transplantation By Walter Andrews, John Sommerauer, Jay Roden, John Andersen, Colleen Conlin, and Paul Moore

Dallas, Texas

0 Purpose; Pediatric liver transplantation is an accepted therapy for end-stage liver disease, but little long-term data exist. Methods: From October 1984 to October 1994, 202 patients underwent a total of 225 liver transplantations. There were 98 boys and 104 girls, the average age was 5.1 rt 4.9 (range, 0.2 to 19.1) years. Thirty (16%) were under 1 year of age. The diseases that required transplantation included biliary atresia (BA) (45%), metabolic liver disease (MLD) (9.9%) acute hepatic failure (6.9%), and Alagille’s syndrome (AS) (5.4%). Originally the immunosuppression was cyclospo- rine- and steroid-based; the later regimens also included azathioprine and antilymphocyte preparations. All reported survival rates were derived from life-table analysis. Results: The patient survival rates at 1, 5, and 10 years were 76%. 70%, and 61%; the retransplantation rate was 11%. The respective graft survival rates were 71%, 63%. and 59%. There were 60 deaths; 48 (81%) occurred in the first year. These first-year deaths were from sepsis (20; 42%). central nervous system problems (5; II%), intraoperative complica- tions (4; 8%), lymphoproliferative disease (LPD) (2; 4%). rejection (2; 4%), primary nonfunction (2; 4%), and miscella- neous other causes (7; 15%). There were 12 deaths after the first year, from LPD (3; 25%), sepsis (1; 8%). rejection (2; 18%). cancer (1; 9%), secondary hepatic failure (1; 9%). cerebral vascular accident (1; 9%). or pre- or postoperative complications (3; 25%). Compared with the overall survival rate, patients with MLD had a better chance of survival (83%; P c ,012) than did those with AS (45%; P c .OOl). The 5- and IO-year survival rates for patients with BA were 61% and 58%. Over the past 2 years, the survival rate has increased (87% v 72%; P -z .05) as early septic deaths have decreased (from 2.6 to 1.0 per year). Conclusion: Liver transplantation is effective treatment for end-stage liver disease. Decreasing the number of early septic deaths has improved the chance of survival, and better diagnosis and treatment of LPD would improve the late survival rate. Copyright o 1996 by W.B. Saunders Company

INDEX WORDS: Liver transplantation, pediatric, survival, complications, mortality rate.

P EDIATRIC LIVER transplantation has experi- enced a tremendous explosion over the last 10

years. In February 1995,4,139 patients were awaiting liver transplantation, and approximately 20% of them were pediatric patients.’

Information on long-term outcome after pediatric liver transplantation is scant.2 This prompted us to undertake a retrospective review of our lo-year patient and graft survival rates, which includes a detailed analysis of the factors that have a positive or negative effect on survival.

MATERIALS AND METHODS

From October 1984 to October 1994. 202 children received a total of 225 orthotopic liver transplantations. All procedures were

JournalofPed/atricSurgery, Vol31, No 5 (May), 1996: pp 619-624

performed as part of the Joint pediatric liver transplant program between the University of Texas Southwestern Medical Center and the Children’s Medical Center of Dallas. Because of program restructuring, there was a hiatus in transplantation between February and October 1992.

The most common diseases that required transplantation were btliary atresia (45%) metabolic liver disease (9.9%) acute hepatic failure (6.9%) and Alagille’s syndrome (5.4%). The indications for transplantation included (1) the presence of hepatrc encephalopa- thy. (2) complications related to portal hypertension such as uncontrollable variceal hemorrhage and ascites that failed medical management, (3) intractible puritis with excoriations, (4) growth failure despite aggressive nutritional management, (5) metabolic failure (decreasing albumin level and/or increasing prothrombin time), and (6) social invalidism.

Patient pretransplant illness severity was stratified using the UNOS (United Network for Organ Sharing) priority status. A status 1 patient is at home doing well; a status 2 patient is in and out of the hospital, a status 3 patient is confined to the hospital; and a status 4 patient requires intensive care management.

Our maintenance immunosuppression and rejection treatment protocols have evolved over the past 10 years (Table I); our current protocol began in October 1992. Rejection was diagnosed using a combination of clinical and laboratory features and was confirmed by a liver biopsy specimen.

All patients received a variety of prophylactic agents. Since 1992 we have routinely used mechanical bowel preparation, intravenous antibiotics (gram-positive and gram-negative coverage) for 4 days. mycostatin (1 million units orally, four times a day), intravenous acyclovir (15 mg/kg/d) followed by oral acyclovir (60 mg/kg/d) for a total of 6 months, intravenous IgG for 3 months with a tapering frequency, Trimethoprim-sulfa (2 mgikgid on Monday, Wednes- day. and Frrday), and an Hz blocker (ranitidine). Pretransplanta- tion, all donors and recipients are screened for cytomegalovirus (CMV) by IgG viral titers, for Epstein-Barr virus (EBV) by polymerase cham reaction (PCR), and for hepatitis C by enzyme- linked immunosorbent assay (ELISA). Posttransplantation, CMV viral titers are obtained only if CMV disease is suspected. EBV PCR is obtained posttransplantation, every 2 months for 6 months, every 3 months for 6 months, and every 4 months for 1 year.

Whole-organ and reduced-sized orthotopic liver transplanta- tions were performed in a standard fashion, using previously described techniques.3

ABO-incompatible (ABO-I) transplantations were performed according to a standard protocol used since 1988. Pretransplanta- tion, a baseline donor AI30 titer was obtained for the recipient. During the surgery, all packed cells are of the recipient’s blood type

From the Dtvtston qf Pediatric Surgery, Depanments of Surgery and Pediatrics, Umverstty of Texas Southwestern Medical Center and the ChrldrenS Medical Center of Dallas, Dallas, TX.

Presented at the 26th Annual Meeting of the American Pediatric Surgtcal Associahon, Boca Raton, Florida, May 20-23, 1995.

Address reprint requests to Walter S. Andrews, MD, Children‘s Lever Transplant Program, Children’s Medical Center of Dallas, 1935 Motor St, Dallas, TX 75235.

Copyright o 1996 by WB. Saunders Company 0022-346819613105-O002$03.OOlO

619

620 ANDREWS ET AL

Table 1. lmmunosuppression

Pemd Inductron Mamtenance Reiection Theraw

1984-1987 None

1987-1990 None

1990-1992 Antithymocyte preparation (MALG),

15 to 20 mg/kg/d for 3 to 5 days

in patients with pretransplant renal

dysfunction 1992-1994 Antithymocyte preparation (ATGAM),

15 to 30 mglkgld for a minimum of

3 days in all patients

Prednisone: 10 mg/kg in operating room >40 kg: 200 mg-20 mg, tapered over 5 days

<40 kg: 100 mg-20 mg, tapered over 5 days Cyclosporine A: Start (on postoperative day 1)

at 6 mg/kg/d IV. Start oral administration when diet is tolerated, at 20 mg/kg/d.

Target level: 300 to 350 ng/mL by whole blood HPLC

Same as above plus azathioprine, 1 to 2 mglkgld

Triple therapy with prednisone, azathioprine, and cyclosporine A.

Target level: 350 to 400 ng/mL by whole blood TDX (Abbott Laboratories)

Methylprednisolone, 1 mg/kg IV once a day for 3 days, 0.75 mg/kg IV once a day for 2 days, then prednisone, 2 mg/kg/d

tapered to 1 mglkgld in 2 weeks, then to 0.25 mg/kg/d in 1 month, and to 0.1

mg/kg/d by 3 months. Cyclosporine and azathioprine are given as noted above.

1994: Tacrolimus substituted for cyclos- porine A. Dose adjusted to whole blood

IMX (Abbott Laboratories) level of 12 to 15 ng/mL.

and all fresh frozen plasma is of the donor’s blood type. Postopera- tive immunosuppression included antithymocyte gammaglobulin (ATGAM) induction for 7 days, cyclosporine A (recently tacro- limus), steroids, and azathioprine. Donor AI30 titers are obtained daily from the recipient. If the titer is greater than 1:8, the recipient undergoes a double-volume plasma exchange. This is continued until the donor ABO titer and the liver function test results have been stable for 5 days.

Survival curves were generated from Kaplan-Meyer life-table analyses. Significant differences between survival curves were determined by the Mantel-Haenszel x2 technique. Group compari- sons were determined by x2 analysis or the Fisher’s exact test. P values of less than .05 were considered significant.

RESULTS

The mean age of our population was 5.14 + 4.89 years (median, 3.03 years; range, 6 weeks to 19 years).

Mild: 1 g bolus of hydrocortisone or methylprednisolone

Moderate: bolus + steroid taper

Severe: bolus + steroid taper + any

lymphocyte preparation (MALG, ATGAM, OKT3)

Mild/moderate: intravenous

methylprednisolone (10 mg/kg) daily for 3 days

Severe: same as above + antilymphcyte preparation (DKT3)

Sixteen percent were under 1 year of age at the time of transplantation. There were 104 girls and 98 boys. One hundred forty-five (72%) were white, 20 (10%) were black, 33 (16%) were Latin American, and 4 (2%) were of other races.

The overall patient survival rates at 1, 5, and 10 years (Fig 1) were 76%, 70%, and 61% (respectively); the corresponding overall graft survival rates were 71%, 63%, and 59%. Age at the time of transplanta- tion did not have a significant impact on survival. The 5- and 9-year survival rates for children who were under one year of age at the time of the procedure (both 66%) did not differ significantly from those of the older children (70% and 66%, respectively). The 3-year graft survival rates for the various races did not

0 0 2 4 6 8 10 12 Fig 1. Patient and graft survival

Years Post Transplant rates for the entire study group.

LIVER TRANSPLANTATION

differ significantly (67% for whites, 76% for blacks, 64% for Latin Americans, 43% for other races).

The overall survival rate for patients with meta- bolic disease (90% at 5 years) was substantially better than that for patients with acute hepatic failure (43% at 5 years). Patients with biliaxy atresia, hepatitis, and the other diagnoses had 9:year survival rates between 62% and 69%.

Overall patient survival by UNOS status (Fig 2) showed a close grouping of the curves. Graft survival by UNOS status did not differ significantly despite the greater curve separations.

When the patient and graft survival rates for whole-organ transplantation were compared with those of reduced-graft transplantation, the latter were found to be lower (72% v 49% at 1 year, 66% v 49% at 5 years). Retransplantation was not required for any patient who received a reduced-size graft.

ABO-I transplantation has been used at our center to increase the pediatric donor pool. The patient surviva1 rate has not been compromised by the use of ABO-I transplantation (Fig 3). The 7-year graft survival rate was lower (but not significantly) for the ABO-I group (44% v 64% for the ABO-identical group).

The mean operating time decreased from 11.6 2 0.95 hours in 1984 to 7.7 + 2.1 hours in 1994 (P < .OOl), and the mean amount of blood loss decreased from 2.74 + 1.4 volumes in 1984 to 1.85 +- 1.35 volumes (difference not significant) in 1995. The average period of hospitalization decreased from 30.0 2 18.5 days in 1985 to 18.9 2 10 days in 1994 (P < .OOl). These decreases occurred during a pe- riod in which there was a progressive decline in patient pretransplant condition, as reflected by the significant increase in the average UNOS pretrans- plant priority status from 1984 to 1994 (1 to 3.1; P < .OOl).

The incidence of hepatic artery thrombosis (HAT) was 8.3 + 6% per year (0% to 20%). However, the

621

lo-year survival rate was not significantly lower for patients with HAT than for those without it (45% v 61%; P < .6). The incidence of biliary complications was 15.5 f 10.5% per year (0% to 32%) and was associated with a lo-year graft survival rate of 40%.

Retransplantation was required for 23 (11%) pa- tients, because of HAT (6), acute rejection (2), chronic rejection (4) secondary ischemic injury (4), or primary nonfunction (2). Their &year survival rate was lower (but not significantly) than that for patients who required only one graft (50% v 70%).

There were 60 deaths after transplantation. The major cause of death was infection (38%). (There were 17 bacterial, 4 fungal, and 2 viral.) Other causes were secondary hepatic failure (15%) (4 hypotension, 3 rejection, 1 HAT, 1 accelerated rejection), central nervous system complications (lo%, n = 6) posttrans- plant lymphoproliferative disorder (PTLD) (lo%, n = 6), massive intraoperative blood loss (7%, n = 4), and miscellaneous (20%, n = 12). Most deaths (44 of 60, 72%) occurred within the first 6 months after transplantation. Of the 17 that occurred after 6 months, 9 (534) o were related to retransplantation; 4 (24%) patients had been awaiting retransplantation, and 5 (29%) died after retransplantation.

A death analysis was performed by categorizing the deaths into two groups: those that occurred before and those that occurred after the sixth posttransplant month (Table 2). The deaths were then subdivided into four categories according to their potential etiology as described by Shaw et a1.4 Category 1 deaths were related to the patient’s pretransplant condition. Category 2 deaths were related to surgical complications. Category 3 deaths were related to complications of immunosuppression. Category 4 deaths were from various miscellaneous and unpre- dictable causes.

All category 1 deaths that occurred within the first 6 months after transplantation were related to compli- cations of hepatic coma in patients who had acute

5 4oJ. u)

Fig 2. Patient survival rates by 30 -. UNOS status. Status 1 patients are at 20 -- home doing well, status 2 patients 10 -- -status 1 8 2 -status 3 -status 4

require frequent hospitalization, sta- 1

tus 3 patients require continuous hos- 0-t I pitalization, and status 4 patients are 0 2 4 6 0 10 12

in the intensive care unit. Years Post Transplant

622 ANDAEWS ET AL

Fig 3. Patient survival rates by ABO donor-recipient match. Same: donor and recipient blood types match. Com- patible: donor blood type is 0, recipi- ent tv~e is anv. Incompatible: donor

0 2 4 6 6

Years Post Transplant

hepatic failure. The category 2 deaths for this group were intraoperative in five cases and secondary to untreated hepatic artery occlusion in two cases. The category 3 deaths (3) were related to accelerated rejection that started on the fourth postoperative day. The category 4 deaths were attributable to multiple and unpredictable causes.

Most deaths that occurred more than 6 months posttransplantation were category 4. They were pre- dominantly related to the diffuse form of the lympho- proliferative disease.

DISCUSSION

The pretransplantation diagnosis of metabolic dis- ease had a positive effect on the lo-year patient survival rate (90%); the diagnosis of acute hepatic failure had a negative effect (43%). This is not surprising, because patients with metabolic liver dis- ease usually are in good nutritional and physiological shape except for the metabolic defect; however, those with acute hepatic failure require intensive care

10 ~.

12 and recipient blood types do not match.

management, are in stage III or IV hepatic coma, and often have renal compromise. Therefore, the preop- erative condition of the patient, at least at the extremes, does influence outcome. The same conclu- sions were reached by Burdelski et a1,5 Hanid et aI,‘j and Lidofsky, et a1.7

Interestingly, the survival rate for patients with biliary atresia (62% at 10 years) closely approximates that of the total group (61% at 10 years). This is important because before transplantation these pa- tients are nutritionally compromised, have a coagulop- athy, and have had previous surgery. Because post- transplantation survival was not adversely effected, and the Kasai operation yields a 20% to 30% chance of long-term bile drainage, it would be logical to perform a Kasai procedure before liver transplanta- tion. In addition, even though the Kasai procedure can fail, it may function long enough for the patient to grow, thereby increasing the potentially available donor pool. Otte et al2 reported similar results for a series of 198 children with biliary atresia who under-

Table 2. Deaths After Pediatric Liver Transplantation

category 1 Category 2 Category 3 category 4

Deaths < 6 mo 7 Acute hepatic failure with 3 Hepatic artery thrombosis 3 Accelerated rejection 8 Technical complications

posttransplantation stage IV coma 2 Massive blood loss 4 Pulmonary infections (2

(n = 43) 1 Infected cut surface of viral, 2 bacterial)

reduced-size liver 3 Lymphoproliferative

1 Prolonged portal vein disorder

occlusion 2 Pulmonary hemorrhage 1 Cardiac arrest 2 Unknown pulmonary

1 Pulmonary hemorrhage decompensation

2 Gram-negative sepsis 1 ABO-incompatible 1 Myocardial dysfunction 1 lntraabdominal bleeding

(unknown etiology)

Deaths >6 mo posttransplantation (n = 17)

0 1 Hepatic artery thrombosis 7 Chronic rejection (4 died 5 Lymphoprohferative after retransplantation, 3 disorder

died while awaiting 2 Cystic fibrosis

retransplantation) 1 Cancer 1 Pneumonia

LIVER TRANSPLANTATION 623

went liver transplantation; their l- and 5-year survival rates were 82% and 78%.2

The patient survival rate was worse after retrans- plantation than after a primary graft (72% v 50%). This is not surprising, because 76% of the former patients were UNOS status 4 at the time of retrans- plantation. A major reason for retransplantation was HAT. Our current protocol calls for daily Doppler ultrasonography to evaluate hepatic artery flow. If a flow alteration is detected, an arteriogram is per- formed immediately. If thrombosis is found, the artery is declotted with locally delivered thrombolytic agents, followed by hepatic artery revascularization in the operating room. Use of this protocol has success- fully restored long-term arterial flow in three patients and has avoided biliary complications in two.

It has been reported that poor preoperative status has had an adverse effect on patient survival.4 To evaluate this, we compared patient and graft survival rates with respect to pretransplantation UNOS sta- tus. The patient and graft survival rates at 5 and 9 years were 10% to 15% poorer for status 4 patients, but the difference was not statistically significant. These data support our current policy that once accepted for transplantation, a patient would be excluded only because of an acute systemic infection or irreversible brain injury.

infectious complications have occurred at the cut surface of a reduced-size liver. We now pursue an aggressive surveillance and treatment policy for fluid collections in that area. Category 4 deaths are the most difficult to correct because of their unpredictabil- ity. Sepsis and pulmonary complications have been diminished by the introduction of effective therapy for CMV infections and improved critical care man- agement. PTLD remains an unpredictable, perplex- ing, and often devastating problem, but the advent of the use of PCR to detect EBV has allowed prompt intervention with immunosuppressive reduction or cessation. High-dose acyclovir prophylaxis and immu- noglobulin therapy have been used in the last 2 years for CMV and potentially EBV prophylaxis. However, it is too early to assess the effectiveness of this approach on the development of PTLD.

More than 40% of the late deaths were related to failure of immunosuppression, with subsequent chronic rejection and retransplantation. A variety of new immunosuppressive agents currently are under investigation. These agents may lead to the reduction or elimination of chronic rejection in the near fu- ture.ll

The use of ABO-I transplantation is controversial, and the literature yields conflicting results.8-10 The 5-year survival rate for our 23 ABO-I transplant patients is comparable to that of the total group (61% v 69%, respectively). However, graft survival was lower at 7 years (44% v 62%). With the initiation of our current protocol, the 3-year patient and graft survival rates have increased, to 67% and 59%, respectively.

The direct results of our improvements in tech- niques and knowledge are illustrated by the compari- son of our patient and graft survivals between 1984- 1991 and 1993-1994 (patient, 88% v 71%; graft, 82% v 65%).

The best way to improve survival obviously is to prevent the complications that lead to death. In our series, after evaluating the deaths that occurred within 6 months posttransplantation, we altered our protocols. To decrease category 1 deaths, we now use intracranial pressure monitoring in patients with stage IV hepatic encephalopathy. For category 2 patients, we use our new protocol for the detection and treatment of HAT. Our most recent major

Pediatric liver transplantation has undergone a revolution over the last decade. Sacred dicta have become old wives’ tales, and procedures that were believed impossible are now commonplace. Any retro- spective analysis in this type of environment is imme- diately flawed because experience, and therefore, clinical practice are changing constantly. It would be inappropriate to interpret our 61% lo-year survival rate as an indicator of the potential for pediatric liver transplantation. Rather, it should serve as a baseline for future improvements.

ACKNOWLEDGMENT

The authors gratefully acknowledge Anthony Benitez and Del- metria Slone for secretarial support.

REFERENCES

1. Patients Waiting for Transplants. UNOS Update 11:46-47, 1995

2. Otte JB, de Ville de Goyet J, Reding R, et al: Sequential treatment of biliary atresia with Kasai portoenterostomy and liver transplantation: A Review. Hepatology 20:415-485, 1994

3. Andrews W, Wanek W, Fyock B, et al: Pediatric liver transplantation: A 3-year experience. J Pediatr Surg 24:77-82,1989

4. Shaw B Jr, Wood P, Stratta R, et al: Stratifying the causes of death in liver transplant recipients, An approach to improving survival. Arch Surg 124:895-900, 1989

5. Burdelski M, Rodeck B, Latta A, et al: Treatment of inherited metabolic disorders by liver transplantation. J Inherit Metab Dis 14:604-618,199l

6. Hanid M, Davies M, Mellon P, et al: Clinical monitoring of

624 LIVER TRANSPLANTATION

intracranial pressure in fulminant hepatic failure. Gut 21:866-869, 1980

7. Lidofsky S, Bass N, Prager M, et al: Intracramal pressure monitoring and liver transplantation for fulminant hepatic failure. Hepatology 16:1-g, 1992

8. Renard TH, Shimoka S, LeBherz D, et al: ABO incompatible liver transplantation in children: A prospective approach. Trans- plant Proc 25:1953-1956,1993

9. Lo C, Shaked A, Busuttil R: Risk factors for liver transplanta- tion across the ABO barrier. Transplantation 58548-553, 1994

10. Tanaka A, Tanaka K, Kital T, et al: Living related liver transplantation across ABO blood groups. Transplantation 58:548- 553,1994

11. Ferraresso M, Kahan B: New immunosuppressive agents for pediatric transplantation. Pediatr Nephrol7:567-573,1993