Nationwide survey for primary sclerosing cholangitis and IgG4-related sclerosing cholangitis in Japan

Authors


  • This article is a secondary publication based on a study first reported in the Journal of Japan Biliary Association (JBA 2013;27:176–187).

Abstract

Background

We previously conducted nationwide surveys for primary sclerosing cholangitis (PSC) in Japan, and demonstrated several characteristic features of Japanese PSC patients, yet patients with IgG4-related sclerosing cholangitis (IgG4-SC) might be misdiagnosed as PSC. Since the clinical diagnostic criteria of IgG4-SC were established in 2012, we again conducted a nationwide survey to investigate the characteristics of PSC and IgG4-SC lacking pancreatic involvement.

Methods

The design was a questionnaire-based, multi-center retrospective study. The enrolled subjects were patients with PSC and IgG4-SC without pancreatic involvement diagnosed after 2005.

Results

We enrolled 197 PSC and 43 IgG4-SC patients without pancreatic lesions. The male dominance was significantly evident in IgG4-SC (P = 0.006). In patients with PSC, two peaks in age distribution were clearly observed. IgG4-SC was not detected in any patient younger than 45 years of age. At presentation, serum albumin and IgM were significantly higher in PSC, while serum IgG and IgG4 were significantly elevated in IgG4-SC. Inflammatory bowel disease (IBD) was detected in only 68/197 PSC patients (34%). The prognosis of IgG4-SC was considerably better than that of PSC.

Conclusion

We confirmed several interesting clinical details of PSC in Japanese patients: two peaks in the age distribution and lower prevalence of IBD.

Introduction

Primary sclerosing cholangitis (PSC) is a chronic cholestatic liver disease of unknown origin, characterized by the progressive destruction of bile ducts caused by diffuse inflammation and fibrosis that eventually leads to liver cirrhosis [1]. The etiopathogenesis of PSC has not been fully understood, and therefore epidemiological studies of PSC are of vital importance to reduce a significant burden to the health care system posed by PSC [2, 3]. We conducted questionnaire-based nationwide retrospective surveys to elucidate the characteristics of PSC in Japan in 1997 and 2003 [4, 5] and demonstrated several characteristic features of Japanese PSC patients, which had been scarcely reported before by epidemiological studies from Europe and North America. In particular, it is notable that there were two peaks in the age distribution at diagnosis, that is, in the 3rd and 7th decades of life, in Japanese patients with PSC [4, 5]. Most previous epidemiological studies from North America and Europe indicated a single age category as the highest risk for developing PSC [6-9]. Interestingly, the age distribution of PSC complicated with inflammatory bowel diseases (IBD) demonstrated a single peak in the 20s, while patients with AIP exhibited a small peak in the 60s.

Meanwhile, accumulating evidence suggested the presence of a different clinical entity of sclerosing cholangitis with elevated serum IgG4 levels, complicated with/without autoimmune pancreatitis (AIP) [10-12]; this disorder is now defined as IgG4-related sclerosing cholangitis (IgG4-SC) [13]. At the time of 2003 when we conducted the nationwide survey for PSC, the concept of IgG4-SC was not well known among physicians and therefore it was likely that patients with IgG4-SC might be misdiagnosed as PSC in the survey. Indeed, not a few PSC patients consisting of the peak at 60s were reported to be complicated with pancreatitis, indicating that these patients should be considered to be IgG4-SC complicating with AIP, not PSC. This was confirmed by the fact that these patients were reported to have good responses to corticosteroid treatment. Moreover, it is of note that the peak in the 60s persists even if patients with pancreatitis are eliminated. Thus it remains possible that IgG4-SC patients without pancreatic involvement may be also misdiagnosed as PSC and thus coordinately comprise the second peak in the elderly in the 2003 survey.

In 2012, the clinical diagnostic criteria of IgG4-SC were established in Japan [14], based mainly on cholangiographic findings. It was thus possible to discriminate PSC from IgG4-SC using these criteria. At this moment, we conducted the third nationwide survey to investigate the characteristics of PSC as well as IgG4-SC, especially IgG4-SC lacking pancreatic involvement and we reported the results from elsewhere in Japanese [15]. In this article, we aimed to describe the unique features of Japanese patients with PSC identified in the current nationwide survey along with a focus on patients with IgG4-SC.

Patients and methods

Study design and inclusion criteria

This nationwide survey was conducted as a questionnaire-based, multi-center retrospective study as previous surveys in 1997 and 2003. A questionnaire was sent to the 144 facilities in Japan, in which active members of the Japanese Biliary Association, the intractable hepato-biliary disease study group in Japan, and the research committee to establish diagnostic criteria and development of treatment for systemic IgG4-related sclerosing disease in Japan belonged to, and we asked that the questionnaires be sent back when subjects of the study were present in their facilities. The enrolled subjects in this study were patients with PSC and IgG4-SC without pancreatic involvement diagnosed after 2005, when IgG4-SC was supposed to be well known among clinicians in Japan. Since the differential diagnosis of IgG4-SC from PSC is not problematic when AIP is coexistent with IgG4-SC, we enrolled only cases with IgG4-SC apparently lacking pancreatic lesions.

Diagnosis of PSC/IgG4-SC

The diagnosis of PSC and IgG4-SC was independently performed by physicians of each facility. The diagnosis of PSC basically was made according to the diagnostic criteria proposed by Lindor et al. [16], mainly depending on cholangiographic studies, biochemical findings, and exclusion of other possible causes. The diagnosis of IgG4-SC was made using the clinical diagnostic criteria of IgG4-SC established by the Japanese Biliary Association in 2012 [14]. Schematic classification of cholangiographic findings of IgG4-SC was made according to the work by Nakazawa et al. [17]. The presence of pancreatic involvement in cases with IgG4-SC was determined by imaging studies in each facility.

Statistical analyses

As for statistical analyses, continuous variables are presented as means ± standard deviations (SD) if they were normally distributed, or medians if not. Comparison between PSC and IgG4-SC was performed using Student's t-test for normal distributed variables, or Mann–Whitney U-test for non-normal distributed variables. Dichotomous variables were compared using χ2 test. All tests were two-tailed and conducted at a 1% level of significance, considering multiple comparisons. All statistical analyses in the current study were performed using IBM SPSS Statistics version 19 (IBM Japan, Tokyo, Japan). This study protocol was approved by the ethical board committee of Teikyo University School of Medicine (#11-121).

Results

Demographic characteristics

In Table 1, we demonstrate the demographic characteristics of PSC and IgG4-SC without apparent pancreatic involvement. Overall, we enrolled 197 PSC and 43 IgG4-SC patients without pancreatic lesions. The male/female ratio was 106:91 (54%/46%) in PSC and 33:10 (77%/23%) in IgG4-SC, indicating that male dominance was significantly evident in IgG4-SC (P = 0.006). The age distribution (median [min–max]) was 48.1 [4.0–86.3] in PSC and 69.3 [47.6–87.4] in IgG4-SC. In patients with PSC, two peaks in age distribution, one at 35–40 years and the other at 65–70 years, were clearly observed in the current study as well as shown in the previous surveys (Fig. 1A). Thus it was confirmed again that elderly patients (>65 years) are at high risk for developing PSC in Japan. By contrast, IgG4-SC was not detected in any patient younger than 45 years of age; therefore, the elderly population is also considered to be at high risk for IgG4-SC (Fig. 1B). Symptoms at presentation were comparable in both groups, and patients presented as asymptomatic (PSC 55%, IgG4-SC 54%), with jaundice (25%, 22%), with cholangitis (20%, 17%) and with itching (17, 20%). The impaired lesions of bile ducts are shown in Figure 2. Extrahepatic bile ducts were mainly affected in PSC, while intrahepatic bile ducts were dominant lesions in IgG4-SC without pancreatic involvement. Cholangiographic classification of IgG4-SC according to the clinical diagnostic criteria of IgG4-SC in 2012 [14] are demonstrated in Figure 3. Type IV was the most common, probably because cases with IgG4-SC without apparent pancreatic involvement were enrolled in this study.

Figure 1.

The distribution of age at diagnosis of enrolled patients at diagnosis. (A) patients with primary sclerosing cholangitis (PSC). (B) patients with IgG4-SC lacking apparent pancreatic involvement

Figure 2.

The bile duct lesions of primary sclerosing cholangitis (PSC) and IgG4-SC lacking apparent pancreatic involvement

Figure 3.

Cholangiographic classification of IgG4-SC lacking apparent pancreatic involvement according to the clinical diagnostic criteria of IgG4-SC in 2012 [14]

Table 1. Demographic characteristics of primary sclerosing cholangitis (PSC) and IgG4-SC
 PSC (n = 197)IgG4-SC (n = 43)P-value
  1. a Expressed as median [minimum–maximum]
  2. b Expressed as positive/negative cases (positive rate)
  3. ALP alkaline phosphatase, Alb albumin, ALT alanine aminotransferase, ANA anti-nuclear antibodies, AST aspartate aminotransferase, CA19-9 carbohydrate antigen 19-9, cANCA c-anti-neutrophil cytoplasmic antibodies, CEA carcinoembryonic antigen, γGTP gamma-glutamyl transpeptidase, Ig immunoglobulin, pANCA p-anti-neutrophil cytoplasmic antibodies, T.Bil total bilirubin, TP total protein
Sex (male : female)106:9133:100.006
Agea48.1 [4.0–86.3]69.3 [47.6–87.4]<0.001
Symptoms at presentation
None100 (55%)22 (54%)NS
Jaundice46 (25%)9 (22%)NS
Cholangitis37 (20%)7 (17%)NS
Skin itching31 (17%)8 (20%)NS
Laboratory dataa
TP7.5 [4.8–9.8]7.7 [6–11.8]NS
Alb3.9 [1.3–4.9]3.5 [2.2–4.8]<0.001
T.Bil1.0 [0.2–29.4]0.9 [0.4–27.2]NS
AST55 [10–751]44 [17–426]NS
ALT60 [7–927]42 [7–260]NS
ALP (xUNL)2.25 [0.32–17.0]2.05 [0.30-13.74]NS
γGTP236 [11–2975]265.5 [17–1344]NS
IgG1623.5 [508–4456]2303 [680–6615]<0.001
IgG448.9 [3.0–369]519.5 [22.2–2470]<0.001
IgA284 [43.2–1597]272.5 [53–963]NS
IgM119 [24–599]80.5 [20–247]<0.001
IgE177 [4–1816]703 [20–3550]NS
CEA1.99 [0.3–28]2.4 [0.7–14.7]NS
CA19-919.85 [0.6–6957.2]31.2 [2–4862]NS
Detection of autoantibodiesb
ANA58/105 (36%)14/17 (45%)NS
pANCA2/88 (2%)1/11 (8%)NS
cANCA3/48 (6%)0/4 (0%)NS

As for laboratory data at diagnosis (Table 1), serum albumin and IgM were significantly higher in patients with PSC. By contrast, serum IgG, in particular IgG4 (Fig. 4A,B), were significantly elevated in IgG4-SC. Serum IgG4 levels at diagnosis (normal range; <135 mg/dL) were elevated in 15/120 PSC (12.5%) and in 34/38 IgG4-SC (89.5%) patients (Table 2). Thus, the sensitivity and specificity of serum IgG4 levels for diagnosing IgG4-SC without AIP were 89.5% and 87.6%, respectively.

Figure 4.

The distribution of serum IgG4 level at diagnosis of enrolled patients at diagnosis. (A) Patients with primary sclerosing cholangitis (PSC). (B) patients with IgG4-SC lacking apparent pancreatic involvement

Table 2. Comparison of serum IgG4 level between primary sclerosing cholangitis (PSC) and IgG4-S
 PSCIgG4-SC
≤135 mg/dL1054
>135 mg/dL1534

Complications

With respect to complications, IBD was detected in 68/197 PSC patients (34%), indicating a relatively low prevalence of IBD in Japanese PSC patients as demonstrated in the previous surveys (Fig. 5). IBD was diagnosed as ulcerative colitis (UC) and non-specific colitis in 55 and 10 patients, respectively. Crohn's disease was not reported in any patient. Cholangiocarcinoma (CCA) was found in 14/197 PSC (7.3%), which was higher than the prevalence observed in our previous surveys in Japan. The duration between the diagnosis of PSC and development of CCA was shown in Figure 6. It is of note that most cases with CCA (11/14) were found within one year after the diagnosis of PSC. While neither IBD nor CCA was noted in patients with IgG4-SC, other IgG4-related diseases, i.e. dacryoadenitis/sialadenitis and retroperitoneal fibrosis, were detected in 6/43 (14%) and 4/43 (9%) patients, respectively.

Figure 5.

Presence of inflammatory bowel diseases (IBD) in cases with primary sclerosing cholangitis (PSC). CF colonofiberscopy

Figure 6.

Duration between diagnosis of primary sclerosing cholangitis (PSC) and diagnosis of cholangiocarcinoma

Treatment

Medical treatments for PSC are summarized in Table 3. Ursodeoxycholic acid (UDCA) monotherapy was the most frequent choice (n = 89, 45%), followed by combination of UDCA and bezafibrate (n = 28, 14%) and combination of UDCA and prednisolone (PSL; n = 24, 12%). Overall, UDCA, PSL and bezafibrate were used in 150 (76%), 40 (20%) and 39 (20%) out of all PSC cases. Among 40 cases with PSL administration, 18 cases were complicated with ulcerative colitis. As for IgG4-SC, 27 patients out of 36 (75%) in which answers about medical treatment were given were administered with PSL. Initial doses were 30 mg (n = 20), 40 mg (n = 5), 15 mg (n = 1) and mini-pulsed (n = 1). All these cases responded well to PSL.

Table 3. Medical treatment for patients with primary sclerosing cholangitis (PSC)
Medical treatmentn
  1. PSL prednisolone, UDCA Ursodeoxycholic acid
UDCA monotherapy89
UDCA+bezafibrate28
UDCA+PSL24
UDCA+PSL+bezafibrate9
PSL mono7
beza mono2
(not answered)38
Total197
Cases with UDCA150
Cases with PSL40
Cases with bezafibrate39

In Table 4, we summarize endoscopic procedures for patients with PSC and IgG4-SC. Endoscopic dilatation of bile ducts and stenting were performed in 24 (12%) and 46 (23%) PSC patients, respectively. The effect was excellent/fair in 39 and poor/undetermined in 13. In patients with IgG4-SC, dilatation/stenting were performed in four (9%) and 15 (35%) IgG4 patients, respectively. The effect was excellent/fair in 12 and poor/undetermined in two.

Table 4. Endoscopic treatment and efficacy
 PSC (n = 197)IgG4-SC (n = 43)
  1. PSC primary sclerosing cholangitis
Endoscopic dilatation of bile ducts
Yes24 (12%)4 (9%)
No157 (80%)35 (81%)
Unknown/blank164
Endoscopic stenting
Yes46 (23%)15 (35%)
No110 (56%)17 (40%)
Unknown/blank4111
Effects of endoscopic procedures
Excellent or fair3912
Poor or undetermined132
Unknown/blank14529

Prognosis

In Figure 7A–C we demonstrated cumulative survival rate of PSC and IgG4-SC. Since we limited the subjects to those diagnosed after 2005 in the current survey, the median follow-up period was relatively short, being 2.7 ± 2.0 years for PSC patients and 2.3 ± 1.8 years for IgG4-SC patients. The 3-year survival rate was 85.0% for PSC and 90.0% for IgG4-SC, and the 3-year transplantation-free survival rate of PSC was 77.3%. Liver transplantation was not performed in any of the IgG4-SC patients. These figures indicate that the prognosis of IgG4-SC was considerably better than that of PSC, probably due to the excellent response to corticosteroid therapy observed in IgG4-SC patients.

Figure 7.

(A) Cumulative survival rate of primary sclerosing cholangitis (PSC). (B) Cumulative liver transplantation-free survival rate of PSC. (C) Cumulative survival rate of IgG4-SC lacking apparent pancreatic involvement

Discussion

In the current study we demonstrated the results of a nationwide survey for PSC and IgG4-SC without apparent AIP, and again confirmed the second peak in the elderly of patients with PSC (Fig. 1A). In the previous survey in 2003 when the clinical concept of IgG4-SC was not well known among Japanese gastroenterologists, it remained possible that IgG4-SC might be misdiagnosed as PSC and thus compromise the second peak in the elderly. In other words, the second peak in the 60s may not result from special characteristics of Japanese PSC itself; rather, they may be explained by “contamination” of the PSC population with IgG4-SC patients.

Thereafter, based on the great contribution of clinical researchers in Japan, both endoscopists and pathologists, to the establishment of the concept of IgG4-SC, the Japanese Biliary Association established the clinical diagnostic criteria of IgG4-SC in 2012 [14]. The criteria mainly consist of the following four items: (1) characteristic cholangiographic findings; (2) elevated serum IgG4 levels; (3) coexistence of IgG4-related diseases in other parts; and (4) histopathological findings compatible with IgG4-SC; the effectiveness of corticosteroid therapy is an additional option. In the current survey the differential diagnosis of PSC and IgG4-SC was made using these criteria, and thus the misdiagnosis of IgG4-SC as PSC can theoretically be avoided. Therefore the second peak in the distribution truly reflects the presence of patients who developed PSC in the elderly.

Most previous epidemiological studies from North America and Europe indicated a single age category as the highest risk for developing PSC [6-9], except for a recent study in Canada, which suggested two categories, 18–35 and >65, are higher risk groups than other categories, similar to Japanese studies [18]. It remains obscure why the elderly are at high risk for developing PSC in Japan. One of the reasons is supposed to be the difference of genetic background between Japanese and Caucasians. A questionnaire-based epidemiological study conducted in Japan in 2007 suggested the prevalence of PSC in Japan was 0.95 (95% CI; 0.61–1.29) per 100,000 inhabitants [19], indicating low prevalence rate of PSC in Japan compared to Northern Europe and North America [20]. The genetic diversity among races may affect the prevalence rate as well as the age at risk for development of PSC. Alternatively, another subtype of PSC closely resembling “true” PSC and presenting in the elderly may exist in the Japanese population. However, another possibility that cannot be denied is that IgG4-SC might be still misdiagnosed as PSC even in the current study. This study is a questionnaire-based retrospective design and the diagnosis was made by physicians at each facility. Even using the common diagnostic criteria of IgG4-SC the diagnosis of difficult cases may vary among facilities, depending on the experiences and knowledge of each physician.

Besides, in this study we confirmed another characteristic in Japanese PSC, already suggested in the previous surveys: the prevalence of IBD as comorbidity appears to be lower in Japanese PSC patients. In various case series from North America and Northern Europe, IBD was a complication in 47–76% of PSC patients [18, 21-27]. However, in Japanese PSC patients, the presence of IBD was restricted to only 37% (125/388) of patients in the 2003 survey [5] and 34% in the current survey. In general, total colonoscopy is frequently performed in Japanese facilities where the diagnosis of PSC is possible with endoscopic retrograde cholangiography; therefore, a lack of a thorough examination of the colon is not a plausible reason for the lower prevalence of IBD. Indeed, total colonoscopy was performed in 53% (206/388) of the PSC patients in our case series [5], suggesting that almost all PSC patients with any bowel symptoms were examined by total colonoscopy. Even after considering that IBD may be present with little or no clinical bowel symptoms in PSC patients, it is unlikely that the prevalence of IBD in Japanese PSC patients could reach 60–80%. Here again, it is unclear why the presence of IBD as comorbidity in patients with PSC is low in Japan compared to those in North America and Europe. Interestingly, another epidemiological study from Asia also demonstrated a lower prevalence of IBD in PSC patients (2/10; 20%) [28], although the number of patients was relatively small. Thus, genetic diversity between Asia and North America/Europe may contribute to the difference of presence of IBD. By contrast, the “contamination” of IgG4-SC may play a role in the low prevalence of IBD as well. The nationwide survey for PSC in Japan repeatedly demonstrated that the age distribution of PSC complicated with IBD demonstrated a single peak in the 20s, while patients with AIP exhibited a small peak in the 60s [4, 5]. Therefore, the prevalence rate of IBD would be increased if misdiagnosed IgG4-SC patients as PSC in the elderly were eliminated from the results.

In conclusion, the third nationwide survey in 2012 has confirmed several interesting clinical details regarding the apparent differences of PSC in Japanese patients and those from Europe/North America; two peaks in the age distribution and lower prevalence of IBD. To investigate whether Japanese PSC is truly characterized by differing clinical features or not, we need to carefully confirm whether PSC patients from the elderly population are “true” PSC patients by central re-examinations of histopathological studies and cholangiographic findings. In addition, it should be validated whether the clinical diagnostic criteria for IgG4-SC as proposed in Japan are appropriate for patients from Europe and North America. Interestingly, the characteristics of IgG4-SC patients observed in the current survey appear to closely resemble those in IgG4-SC patients in the Mayo clinic [29], suggesting that the Japanese diagnostic criteria may be comparable for both populations. Finally, genetic studies to investigate the susceptible genes that contribute to the development of PSC in Japanese patients are needed in order to clarify whether the genetic background of PSC is similar between Japan and Europe/North America.

Acknowledgments

This study was conducted and supported by the Intractable Hepato-Biliary Disease Study Group in Japan and the Research Committee to establish diagnostic criteria and development of treatment for systemic IgG4-related sclerosing disease supported by the Research Program of Intractable Disease provided by the Ministry of Health, Labor, and Welfare of Japan.

Conflict of interest 

None declared.

Ancillary