AIM: To investigate whether fresh frozen plasma (FFP) transfusion affects outcomes following hepatic resection for hepatocellular carcinoma (HCC) in terms of liver function, postoperative complications and cancer prognosis.

METHODS: We retrospectively compared the incidence of postoperative complications between 204 patients who underwent hepatectomy for HCC with routine FFP transfusion in an early period (1983-1993, Group A) and 293 with necessity for FFP transfusion during a later period (1998-2006, Group B), and also between two subgroups of Group B [22 patients with FFP transfusion (Group B1) and 275 patients without FFP transfusion (Group B2)]. Additionally, only in limited patients in Group B1 and Group B2 with intraoperative blood loss ≥ 2000 mL (Group B1_{≥ 2000 mL} and Group B2_{≥ 2000 mL}), postoperative complications, liver function tests, and cancer prognosis were compared.

RESULTS: No mortality was registered in Group B, compared to 8 patients (3.9%) of Group A. The incidence of morbidity in Group B2 [23.2% (64/275)] was not significantly different from Group B1 [40.9% (9/22)] and Group A [27.0% (55/204)]. The incidence of complications and postoperative liver function tests were comparable between Group B1_{≥ 2000 mL}vs Group B2_{≥ 2000 mL}. Postoperative prognosis did not correlate with administration of FFP, but with tumor-related factors.

CONCLUSION: The outcome of hepatectomy for HCC is not influenced by FFP transfusion. We suggest FFP transfusion be abandoned in patients who undergo hepatectomy for HCC.

Surgical resection is the established optimal treatment for hepatocellular carcinoma (HCC) associated with hepatitis B virus or hepatitis C virus infection. Since HCC usually develops in patients with liver cirrhosis, most such patients present with bleeding tendencies based on chronic liver dysfunction[1,2]. Accordingly, bleeding is a major problem in liver surgery for HCC, and it also affects postoperative mortality and morbidity[3-5].

Fresh frozen plasma (FFP) is human donor plasma, and contains near normal levels of many plasma proteins, including procoagulants and inhibitory components of the coagulation cascades, acute phase proteins, immunoglobulins and albumin. The clinical use of FFP has increased steadily over the last two decades in many countries[6-8]. Furthermore, in the surgical treatment of HCC, FFP has been frequently administered to supply coagulation factors, maintain serum albumin level and circulating blood volume, and prevent postoperative hepatic failure[9-12]. On the other hand, FFP transfusion is reported to induce adverse effects in some patients: transmission of infection, allergic reactions, hemolysis, anaphylaxis, and transfusion-related acute lung injury (TRALI)[13-15]. Moreover, some studies have reported a relationship between perioperative transfusion and postoperative HCC recurrence[16,17]. In addition to these adverse effects, the amount of FFP is limited because of its source from human donation. Therefore, appropriate use of FFP is needed in terms of application and volume, as stated in the guidelines of the Japanese Ministry of Health, Labour and Welfare[18]. Regarding surgery for HCC, recent advances in both surgical and anesthetic techniques that have led to a reduction in intraoperative blood loss, have resulted indirectly in a gradual decrease in the need for FFP perioperatively[19,20]. Considering the reduction in intraoperative blood loss and the aforementioned potential adverse effects of FFP transfusion, we believe there is no need for FFP in surgery for HCC. In order to discuss this need, we first should investigate whether FFP transfusion affects outcomes following hepatic resection for HCC.

In this study, we retrospectively investigate whether FFP transfusion affects outcomes following hepatic resection for HCC in terms of liver function, postoperative complications and cancer prognosis.

Table 1 lists the differences in perioperative characteristics between Group A and Group B. Patients classified as Child-Pugh A were significantly more common among Group A than Group B (P = 0.0176). Intraoperative blood loss in Group A was significantly greater than in Group B (P < 0.0001). While the postoperative mortality was 3.9% (8/204) in Group A, no mortality was recorded in Group B (P = 0.0007). The incidence of postoperative complications was 27.0% (55/204) in Group A and 24.6% (73/297) in Group B, and the incidence did not significantly differ between the two groups.

Various perioperative parameters were compared between Group B1 and Group B2 (Table 1). The preoperative factors were similar in the two groups. The incidence of hepatectomy equal to or more than Hr 2 was significantly higher in Group B1 than in Group B2 (P = 0.0436), and a significantly greater intraoperative blood loss was recorded in Group B1 than in Group B2 (P < 0.0001). There was no operative mortality in either of the two groups. The incidence of postoperative complications was 40.9% (9/22) in Group B1 and 23.2% (64/275) in Group B2, and the incidence did not significantly differ between the two groups. No adverse events related to FFP transfusion were found in Group B1. Postoperative complications and liver function were compared between Group B1 and Group B2 only in patients with intraoperative blood loss of ≥ 2000 mL (Group B1_{≥ 2000 mL}: n = 21, Group B2_{≥ 2000 mL}: n = 29). Comparison of clinical features of patients in these two groups is summarized in Table 2. There were no significant differences in the preoperative factors. Intraoperative blood loss and the frequency of administration of red cell concentrates (RCC) in Group B1_{≥ 2000 mL} were significantly more than those in Group B2_{≥ 2000 mL} (P < 0.0001 and P = 0.0004, respectively). Operative mortality was not encountered in the two groups. The incidence of postoperative complications was 42.9% (9/21) in Group B1_{≥ 2000 mL} and 31.0% (9/29) in Group B2_{≥ 2000 mL}. Table 3 lists the types of complications. Neither postoperative hepatic failure nor postoperative bleeding occurred in the two groups.

Figure 2 demonstrates the perioperative changes in PT and T-Bil in patients with intraoperative blood loss of ≥ 2000 mL. The levels of PT and T-Bil were not significantly different between the two groups, irrespective of the POD.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Perioperative changes in (A) serum prothrombin time, and (B) total bilirubin levels in Group B1_{≥ 2000 mL} and Group B2_{≥ 2000 mL}. NS: Not significant; Preope.: Preoperative; POD: Postoperative day.

Long-term postoperative outcomes, including disease-free survival (DFS) and overall survival (OS) after hepatic resection, were also examined in patients with intraoperative blood loss of ≥ 2000 mL (Table 4). Vascular invasion (absent/present), FFP transfusion (transfused/non-transfused), tumor size (< 5 cm/≥ 5 cm), RCC transfusion (transfused/non-transfused) were significant factors in univariate analysis of DFS among the clinicopathological factors tested (P = 0.0101, 0.0276, 0.0288, and 0.0343, respectively). Multivariate analysis for DFS using the four factors identified vascular invasion as the only significant independent factor (P = 0.0299). The DFS in Group B2_{≥ 2000 mL} was significantly better than in Group B1_{≥ 2000 mL} (P = 0.0276), though the factor was not significant on multivariate analysis (Figure 3A). Next, univariate analysis for OS using various clinicopathological factors demonstrated that vascular invasion (absent/present) and number of nodules (single/multiple) were significant factors (P = 0.0024 and P = 0.0150, respectively). Multivariate analysis for OS using the two factors, identified vascular invasion as the only significant independent factor (P = 0.0185). There was no significant difference in OS between Group B2_{≥ 2000 mL} and Group B1_{≥ 2000 mL} (P = not significant) (Figure 3B).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Disease-free survival (A) and overall survival (B) in Group B1_{≥ 2000 mL} and Group B2_{≥ 2000 mL}. NS: Not significant.

The present study was designed to investigate whether the outcomes of hepatectomy for HCC are influenced by FFP transfusion. Firstly, we compared the incidence of mortality and morbidity between Group A and Group B, indicating no significant difference in the incidence between the two groups. However, the comparison is considered to be difficult because of differences in the background of each period such as surgical and anesthetic techniques. For example, there were significant differences in liver function evaluated by Child-Pugh classification, and in intraoperative blood loss, between the two groups. Therefore, for more justified analysis, we next compared the outcomes between Group B1 and Group B2. The result showed the incidence of mortality and morbidity to be comparable between Group B1 and Group B2. However, since there were significant differences in the surgical procedure and intraoperative blood loss between the two groups, we also compared the postoperative complications between Group B1_{≥ 2000 mL}vs Group B2_{≥ 2000 mL}. The results showed equal rates of postoperative complications in the two groups. In particular, hepatic failure (prevention of which is one of the purposes of FFP administration) was not identified in the two groups. Postoperative residual liver function, represented by PT and T-Bil, was also equal in the two groups. Furthermore, the incidences of postoperative mortality and morbidity in Group B1 and Group B2 were similar to those reported in other studies[5,22-24]. For example, Imamura et al[5] reported the surgical result of 1056 hepatic resections including 532 HCC cases, with incidences of postoperative mortality and morbidity of 0% and 39.0%, respectively. However, they did not report the number of patients who received FFP. FFP was reported to be administered at a rate that exceeded the amount of blood loss by 10% to 20% during surgery; it substituted the amount of protein lost so as to maintain serum total protein level at 6.0 g/dL in that study. Based on our results and those of early studies, it cannot be concluded that FFP administration would contribute to the incidences of postoperative mortality and morbidity, including hepatic failure.

Although some previous studies have reported postoperative complications after hepatectomy for HCC, there are no reports comparing postoperative complications between patients with FFP transfusion and those without FFP transfusion. Martin et al[25] reported the use of FFP after hepatic resection and suggested criteria for FFP transfusion to deal with postoperative complications after treatment of liver metastasis from colorectal cancer, but not for HCC with liver cirrhosis. Accordingly, their criteria should not be necessarily generalized to the use of FFP in hepatectomy for HCC. Therefore, this study is the first report in which the incidence of postoperative complications in HCC patients was compared between patients who received FFP and those who did not receive FFP transfusion.

To date, FFP has been traditionally used at hepatectomy for the purpose of hemostatic effect by correction of deficiency of coagulation factors and maintenance of circulating blood volume by supplementation of albumin, which is mainly responsible for the colloid osmotic pressure of plasma; in addition to the aforementioned purpose of prevention of hepatic failure[6-8]. Firstly, with regard to the hemostatic effect, recent improvements in surgical techniques allow hepatectomy to be performed with minimal bleeding[19,20]. Moreover, coagulopathy requiring FFP transfusion is generally reported to occur at a PT value of more than 2.0 times the control, whereas the mean PT level of patients of Group B1_{≥ 2000 mL} and Group B2_{≥ 2000 mL} in the present study did not drop to the applicable level, even though it was measured after hepatectomy[26-28]. Furthermore, the incidence of postoperative bleeding was low and was not different in the two groups. Taking these results into consideration, routine administration of FFP is not necessary in terms of the hemostatic effect. Secondly, the maintenance of appropriate circulating blood volume is important in order to prevent certain complications such as pulmonary edema and pre-renal type of renal dysfunction. However, albumin products, which can be administered safely compared to FFP, can be substitutes for FFP in terms of maintenance of circulating blood volume. In fact, albumin products were administered perioperatively instead of FFP in this study, especially in the non-transfused group, and the incidence of these complications was not different between Group B1_{≥ 2000 mL} and Group B2_{≥ 2000 mL}. In this context, routine FFP administration is also suggested not to be necessary in terms of maintenance of the circulating blood volume. Thus, we suggest that the routine administration of FFP for the purpose of prevention of hepatic failure, hemostatic effect, and maintenance of circulating blood volume is not necessary.

Many adverse effects related to FFP transfusion have been identified, such as infection, allergic reactions, hemolysis, anaphylaxis, and TRALI[13-15]. In particular, TRALI, which is a rare and serious complication characterized by sudden onset of respiratory distress due to non-cardiogenic pulmonary edema during or following transfusion, can be life-threatening. Fortunately, none of these transfusion-related complications occurred in our patients. However, since some of the reported adverse events can be life-threatening, one should refrain from inappropriate use of FFP.

Since an initial report by Foster et al[29] about survival advantages in patients undergoing colectomy for colon cancer, several other reports have shown that perioperative homologous blood transfusion to be an independent prognostic factor in many kinds of cancers[16,17,30-33]. However, a few suggested that homologous blood transfusion has no significant effect on the prognosis of cancer patients[34,35]. Thus, the association between transfusion and postoperative prognosis is still under debate. In the present study, postoperative prognosis did not correlate with FFP administration, but rather with tumor-related factors. Although the result was not powerful evidence to resolve the controversy, we can at least confirm that FFP administration does not improve prognosis of patients undergoing hepatectomy for HCC.

In fact, the guidelines of the Japanese Ministry of Health, Labour and Welfare state that administration of FFP should be limited only to supplement coagulation factors in those patients with a PT of more than 2.0 times normal or coagulation factor activity of ≤ 30%, and that the use of FFP for supplementation of circulation blood volume is inappropriate[18]. The guidelines do not mention administration of FFP for the prevention of hepatic failure. Thus, our suggestion is to obey the guidelines. Recently, Kaibori et al[36] reported the clinical value of FFP in surgery for HCC. They suggested that FFP transfusion was useful and recommended on the grounds of the results obtained from their analysis that the incidence of postoperative complications in patients with FFP transfusions was lower than that of patients with FFP and RCC transfusions, and was equal to that of non-transfused patients; long-term survival in patients with FFP transfusions was almost equal to that in non-transfused patients. However, their suggestion is perceived as groundless for the following reasons. To begin with, although there were some significant differences in many factors such as liver function and tumor progression among the groups in their study, they simply suggested that the difference in postoperative complications and long-term outcome resulted from the RCC and FFP transfusions. Secondly, since details of postoperative complications were not shown, especially for hepatic failure, postoperative bleeding, pulmonary edema and renal dysfunction, the examination of correlations between complications and FFP transfusions was insufficient. In addition, their suggestion completely ignored the recent guidelines of Japan.

The present analysis did not include HCC patients who underwent liver transplantation for treatment of liver cirrhosis. Therefore, the result of this study is not applicable to liver transplantation surgery. Considering that transfusion is performed for concomitant liver dysfunction at almost all liver transplantation surgery, it seems to be still too early to discuss the necessity of transfusion in such surgery.

In summary, FFP transfusion did not affect outcomes following hepatic resection for HCC in terms of liver function, postoperative complications and cancer prognosis. Considering the previously reported FFP transfusion-related adverse effects in addition to the results of the present study, we suggest that FFP transfusion be abandoned in patients who undergo hepatectomy for HCC.