|Home | About | Journals | Submit | Contact Us | Français|
Between January 1987 and October 1991, 1466 patients underwent consecutive Orthotopic Liver Transplantation (OLTx) at the University of Pittsburgh. Forty of these patient’s had concomitant splenectomy with OLTx. These patients were compared to 147 randomly selected OLTx patients without splenectomy within the same time period. One-year patient and graft survival (PS and GS) were lower in splenectomized (Splx) patients compared to nonsplenectomized (non-Splx) patients (59% vs 86% PS, 55% vs 80% GS, respectively). One-month and one-year patient mortality in the Splx group was higher than in the non-splx patients (20% vs 3.4%, P < 0.001 for one month; 40% vs 14.3%, P = 0.003 for one year, respectively). One-month and one-year sepsis-related mortality was also high in Splx patients (17.5% vs 2.7%, P = 0.0022, for one month, and 30% vs 11.5%, P = 0.0043, for one year, respectively). We conclude that concomitant splenectomy with OLTx has a significantly higher patient mortality mainly due to its septic complications and, at present, unless there is a specific indication for a splenectomy, the routine addition of this procedure to liver allograft surgery would not be recommended.
There are contradictory reports regarding the effect of splenectomy on allograft rejection following transplantation (1–14). Some have reported a beneficial effect of splenectomy either pre- or peritransplantation (1–4, 6, 7, 9, 10) and have attributed this benefit to a reduction in antibody production (1, 2) or improved leukocyte count and therefore tolerance to azathioprine (AZA) (3, 4, 6, 10, 15). Others have shown no benefit, discouraging the routine use of this procedure because of its potentially lethal complications due to delayed infection and thromboembolic events (5, 8, 13, 16).
Because of the nearly universal presence of pancytopenia in seriously ill liver candidates and consequent problems with AZA dosing, concomitant splenectomy was performed with liver transplantation in most hepatic recipients until 1980. With the advent of nonmyelotoxic agents, such as cyclosporine, the routine use of this frequently difficult adjuvant procedure was discontinued and is rarely performed today.
The objective of our study was to examine the role of splenectomy in liver transplantation and to identify the potential hazards associated with this procedure in modern immunosuppressive regimens.
The records of 1466 patients who underwent OLTx at the University of Pittsburgh between January 1987 and October 1991 were reviewed. Nineteen patients undergoing upper abdominal exenteration and thirteen patients with splenectomy (Splx) following OLTx were excluded from this study. From the remaining 1434 patients, we identified 40 (2.8%) who had undergone combined OLTx + Splx. The course of these patients was compared to 150 randomly selected non-Splx patients from the same transplant period. The sample size was calculated to ensure that a 25% or more difference in mortality could be detected between the Splx and the non-Splx group using methods of unequal group sizes (17). The records of three patients could not be obtained, therefore, the remaining 147 patients served as the non-Splx group. The two groups were comparable with respect to donor and recipient age, sex, HLA-A, B, and DR mismatch, etiology of liver disease, UNOS status, cold ischemic time, platelet count, and immunosuppressive regimen (Tables 1 and and2).2). Although there were more positive lymphocytotoxic cross-matches in the study (Splx) group, the difference was not statistically significant (Tables 1 and and2).2). The two groups were not comparable with respect to the amount of perioperative blood transfusion.
Indications for splenectomy included positive lymphocytotoxic cross-match, (N = 10), portosystemic shunts [distal splenorenal shunt (N = 10), portocaval shunt (N = 1)], technical complications (N = 10), and distal pancreatic tumor resection (N = 2). In seven patients the reason for Splx could not be determined in the retrospective review (Table 3).
The two groups were analyzed with respect to patient and graft survival, as well as biopsy-proven acute allograft rejection within 30 days of transplantation.
The results for continuous variables are presented as means ± SD and for categorical variables as proportions. The standard two-sample t test was used to test the difference between group means, while differences in proportions were tested using Pearson’s chi-square test or Fisher’s exact test. The Wilcoxon rank-sum test, a non-parametric equivalent to the standard two-sample t test, was used for highly skewed data (eg, amount of blood transfusions and platelet count). The odds ratio was calculated as an estimate of the relative risk of acute allograft rejection for Splx patients. An odds ratio >1 indicates increased risk, while an odds ratio <1 indicates a protective effect of splenectomy. The Mantel-Haenszel procedure (18) was used to calculate an adjusted odds ratio based on lymphocytotoxic cross-match. A multivariate logistic regression model was used to adjust the odds ratio for amount of blood transfusions. The 95% confidence intervals were calculated using either the test-based procedure of Miettinen or Woolf’s method (18).
Patient survival was calculated from the date of OLTx until death, and graft survival from the date of OLTx until retransplantation or patient death. Survival curves were generated using the Kaplan-Meier (product-limit) method and were compared using the generalized Wilcoxon (Breslow) test. All tests were two-tailed. P < 0.05 was considered statistically significant. All analyses were performed using SPSS for Windows software.
With a minimum follow-up of one year (median 35, range 12–17 months), the one-year patient survival was 60% (24/40) in the Splx group vs 86% (126/147) in the non-Splx group (P = 0.001), a disparity that was already significant at 30 days (P = 0.001) (Figure 1).
In the Splx group, eight patients died within 30 days of OLTx (8/40,10%) compared to five in the non-Splx group (5/147, 3%, P < 0.0001). Four of the eight deaths were in patients with a positive cross-match (4/10, 40%) and the other four in patients with negative cross-match (4/28, 14%, P = 0.093) (Table 3). The one-year mortality based on positive cross-match was 60% (6/10) in the Splx group and 22% (5/23) in the non-Splx group (P = 0.0037). The one-year mortality based on negative cross-match was 36% (10/28) in the Splx group and 12% (15/121) in the non-Splx group (P = 0.005).
Seven of the eight deaths in the Splx group during the first month (7/40, 17.5%) as well as 12 of the 16 by the end of the first year (12/40, 30%) were due primarily to sepsis compared to 4/147 (2.7%, P = 0.0022) and 17/147 (11.6%, P = 0.0043), in the non-Splx group, respectively.
Three of the 40 (7.5%) patients in the Splx group, including two cross-match-positive patients, underwent retransplantation within 30 days of OLTx because of either primary nonfunction (PNF, N = 1) or hepatic artery thrombosis (HAT, N = 2). One more patient was retransplanted after 30 days, for a total incidence of 10%. Two (50%) of the four attempts succeeded. Fourteen (9.5%) patients in the non-Splx group also underwent retransplantation, including three with a positive cross-match liver. Eight (57%) were successful.
With the equivalent rate of successful retransplantation, the graft survival curves of the Splx and non-Splx groups were both 5% lower than patient survival, and similar at all time points. Graft survival for Splx at one month was 78% vs 93% for non-Splx (P = 0.004), and at one year was 55% vs 80%, respectively (P = 0.001).
Using a multivariate logistic regression analysis, Splx patients had a higher risk of one year mortality (odds ratio = 3.1, P = 0.042) even after adjusting for selected baseline recipient and donor characteristics. However, this multivariate analysis did not reveal a significant association between splenectomy and graft failure at one year.
The incidence of acute allograft rejection one month following OLTx was significantly lower in the Splx group (20/40, 50%) compared to the non-Splx group (108/147, 74%, P = 0.005), an advantage seen with and without a positive lymphocytotoxic cross-match (Table 4).
The Splx patients received significantly more units of blood transfusion within one month of OLTx than non-Splx patients (Table 1); however, in a multivariate logistic regression model, the rejection rate was not affected by the difference in the total amount of blood transfusions (Table 5).
The contribution of the spleen as part of the immune response to the transplanted graft is not clearly known (6, 19–22). The spleen is not only the principle site of antibody synthesis (6, 9, 23), including specific alloantibodies (22, 24–26), but also is a source of specific alloreactive host lymphocytes (23). It has been suggested that transient posttransplant migration of dendritic cells from the transplanted graft to the spleen may be responsible for initiation of graft rejection (20). On the contrary the spleen may play a major role in prolongation of graft survival in different species (19, 21, 23, 27–29), perhaps by generation of specific suppressor cells (30).
Starzl et al (1) were first to suggest the role of splenectomy in prolongation of allograft survival, as four of their five patients treated by thymectomy and splenectomy maintained their renal function for almost six months. Hume et al (3) suggested that splenectomy, if done prior to or at the time of transplantation, could improve leukocyte count and permit administration of large doses of AZA. A report by Gleason and Murray (4), also noted the advantage of splenectomy but only in living related kidney transplant patients. Two reports in the 1980s also noted improved patient and graft survival after splenectomy (7, 8). This is in contrast to several reports in the late 1980s that have cautioned against the routine use of splenectomy in the renal transplant population (9, 10, 14, 16, 31, 32).
This retrospective study was undertaken because of the concern that decisions to perform urgent splenectomy were made only for its historical justification, namely, the mitigation of the spleen-dependent antibody response (25, 26, 33–35), without attention to its hazards.
Our concern about splenectomy is similar to those that have been experienced in renal transplantation. A common theme in the earlier studies was that splenectomy appeared to reduce the incidence of early rejection (4, 7, 8). Similar to the report by Megison (36), Splx patients in our study had a significantly lower rate of rejection compared to non-Splx patients. This difference was still apparent when rejection was stratified by positive and negative lymphocytotoxic cross-match (Table 3). Patients with positive cross-matches may present with a clinical picture of antibody-mediated rejection indistinguishable from PNF (37); however, in our study none of the retransplanted patients for PNF had any histologically documented acute rejection.
In our study splenectomy was associated with more blood transfusion. This extra transfusion requirement could be due to unexpected injury to the spleen or may be the result of preformed lymphocytotoxic antibody, which may lead to excessive elimination of platelets and subsequent hemorrhage (38).
The most emphasized risk of splenectomy in the literature is the increased susceptibility to infection and thrombotic complications. However, in our opinion, the most serious hazard of splenectomy in a liver recipient is aggravation of an already difficult technical operation.
Nevertheless, splenectomy independent of other factors, including blood transfusion, has been associated with an increased risk of mortality (16), with sepsis as the leading cause. Our concern is reinforced by the study reported here and emphasizes that the price of this extra surgical procedure far outweighs any theoretical benefits in OLTx patients. We conclude that, based on multiple life-threatening risk factors associated with splenectomy, concomitant splenectomy with OLTx is not recommended except in the rare occasion of unavoidable splenic injury.
Instead of splenectomy for patients with a positive cytotoxic cross-match, we now recommend intraoperative treatment with a combination of high-dose corticosteroids and prostaglandin E1. This treatment is thought to ameliorate the diffuse inflammatory response that is the basis of hyperacute rejection or its subclinical variations (39, 40). Since instituting this management policy in 1992 (41), the extra hazard of liver transplantation in the case of a positive crossmatch (40, 42) has been all but eliminated (43).
The single center experience with splenectomy in OLTx patients has been reviewed. One thousand four hundred sixty-six patients had OLTx over a period of five years. Forty of these patients had splenectomy at the time of OLTx. These patients were compared to a randomly selected group of OLTx patients who did not have splenectomy within the same time period. The results demonstrate that splx patients are at increased risk for mortality, independent of other factors, mainly a due to increased sepsis-related complications.