In patients with BA who survive with their native liver until adolescence or adulthood, it is difficult to assess the necessity and appropriate timing of LT, due to the slow progression of complications [10]. While some patients have a very good prognosis and do not experience jaundice or any complications or progression of liver fibrosis, other patients develop progressive liver fibrosis and cirrhosis, experience complications, and require the consideration of LT, even though they could survive with their native liver until adulthood [4, 11]. In a previous report of living-donor liver transplant (LDLT) in Japan, the long-term graft survival was worse in recipients aged older than 12 years, and was worst in recipients aged 12–18 years [12]. In addition, there is a shortage of brain-dead donors in Japan [13]; therefore, many adult patients with BA are unable to undergo LT at the time they require it, because of the aging of their related donor candidate, or graft-size mismatch [14]. These situations make it difficult to decide when or whether LT should be performed in patients with BA who survive with their native liver until adolescence.

Our data show that patients with severe liver fibrosis or cirrhosis on follow-up LBx had significantly worse NLS and CFS. Although NLS was not differ between mild and severe fibrosis in patients who had follow-up LBx taken between the ages of 6 and 12 years, CFS in these patients was significantly worse in patients with severe fibrosis, which suggests earlier development of complications of BA. There are two reason why NLS in patients who had LBx between the ages of 6 and 12 years had not significant difference between patients with mild and severe fibrosis. First, the number of these patients was small in the current study. Second, some of these patients who had developed complications of BA were considered for LDLT but could not undergo the procedure because of the lack of suitable related donor candidates. Our data suggests that most patients with BA who have severe liver fibrosis or cirrhosis would require LT or experience complications related to BA, even if they had little or no complications at the time of LBx.

We emphasize that most patients who were graded as F3 or F4 at LBx would suffer from complications of BA during the follow-up period later. In our cohort of school-age LBx, whole patients were graded in Child-Pugh class A and had very low PELD scores, which were generally not considered to be indicated for LT [15]. Especially in Japan, there are few cadaveric donations. Given the current situation of donor shortage and requirement of LDLT for those patients, patients with BA who had severe fibrosis or cirrhosis on LBx should be considered and evaluated for LDLT in the near future, according to the condition of each patient.

The relationship between liver fibrosis and long-term prognosis has been previously reported. However, the liver fibrosis assessed was limited at the time of diagnosis of BA or KPE [6, 16,17,18]. Czubkowski et al. reported lower rates of NLS 10 years post-KPE in patients with severe pathologic fibrosis, but the difference was not significant [16]. Higashio et al. reported a significantly lower NLS of 15 years in patients with severe pathologic fibrosis [17]. The current study is the first report about the relationship between the post-Kasai percutaneous follow-up liver biopsy and the long-term prognosis of patients with BA.

Percutaneous liver biopsy is invasive and major complications such as severe pain, hemorrhage, organ injury or even death has been reported [19,20,21]. There could be higher risks of bleeding in patients with severe liver fibrosis or cirrhosis [21]; therefore, we confirm the diagnosis of cirrhosis in patients who have obtained the result of F4 in 2 separate LBxs, and do not perform LBx after the diagnosis. Since most patients with severe liver fibrosis suffered from complications of BA during the follow-up period, even though their Child-Pugh score and PELD score were low at the time of LBx, we considered that the benefits of detecting liver fibrosis with LBx outweigh the risks of LBx.

Our study has some limitations. First, the study was a retrospective, single-institute study and the number of participant patients was small. Patients with severe fibrosis were younger than those with mild fibrosis because our institute’s LT program started in 1998; therefore, more severe patients were referred to our institute following the program’s initiation. Second, the specimen of the percutaneous LBx did not correctly represent the true fibrotic stage of the whole liver. We treat serum Mac-2 binding protein index (M2BPGi) as an important marker of liver fibrosis in patients with BA [7]; however, we could not include M2BPGi, a relatively new laboratory finding, in the current study. Third, the timing of liver biopsy varies depending on the patient. Patients from school age to adolescents are often busy with school and work. The timing of liver biopsy is subject to change, as we consider the timing of liver biopsy according to the patient’s social situation on the premise that pathologic liver fibrosis does not progress rapidly. Fourth, sampling error may occur in core needle libre biopsy. In our cases, F0 at the latest liver biopsy, two patients suffered from liver failure later. The reason may be that the liver biopsies may only be taken from regenerated nodules and may result in F0 fibrosis.

In conclusion, we found that patients with BA with severe liver fibrosis on follow-up LBx had worse long-term survival and a higher rate of progression of complications of BA, even though they had little or no symptoms at the time of LBx. These results suggest that patients with severe liver fibrosis, namely F3 or F4, should be considered or prepared for LT according to their clinical situation and social situation of donor availability. Follow-up LBxs were useful for determining the patients’ prognosis.