A Taiwanese Nationwide Cohort Study Shows Interferon Based Therapy for Chronic Hepatitis C Reduces the Risk of Chronic Kidney Disease.

Chen YC

, Hwang SJ

, Li CY

, Wu CP

, Lin LC

1

Division of Nephrology, Department of Internal Medicine, Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chiayi, Taiwan;

2

School of Medicine, Tzu Chi University, Hualien, Taiwan;

3

Division of Nephrology, Department of Internal Medicine, Kaohsiung Medical University Hospital, Kaohsiung

4

Department and Graduate Institute of Public Health, College of medicine, National Cheng Hung University, Tainan, Taiwan;

5

Department of Public Health, College of Public Health, China Medical University, Taichung, Taiwan.

6

Public Health Department, New Taipei City Government, Taipei, Taiwan

* Correspondence to:

Yi-Chen Chen, MD

Division of Nephrology, Department of Internal Medicine, Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chiayi

No. 2, Minsheng Rd., Dalin Township, Chiayi County 622, Taiwan.

E-mail: [email protected]. Tel: +886-5-2648000-5665; Fax: +886-5-2648128.

Word count: abstract: 245, text: 3296; 4 tables, 2 figures, 1 supplemental table. [Insert Running title

of <72 characters] 1

4 Running title: Interferon-based therapy for HCV decreases CKD risk.

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

4 ABSTRACT

5 6 7

Hepatitis C virus (HCV) infection is a risk factor for chronic kidney disease (CKD).

8 9

10 However, it remains unclear whether interferon-based therapy (IBT) for HCV was

12 13

14 associated with reduced risk of CKD. From the Taiwan National Health Insurance

15 16

17 Research Database, we identified 919 patients who received 3 months or more of IBT

18 19

20 as our treated cohort. This cohort was propensity score-matched 1:4 with 3676

21 22

23 controls who had never received IBT for HCV infection (untreated cohort).

24 25

26 Cumulative incidences of and hazard ratios (HRs) for CKD were calculated after

27

28 29 adjusting for competing mortality. In the matched HCV cohort, the risk of CKD was

30

31 32 significantly lower in the treated cohort (7-year cumulative incidence, 2.6%; 95% CI,

33

34 35 0.7-6.9%) than in the untreated cohort (4%; 95% CI, 3.5-5.2%) (p<0.001), with an

36

37 38 adjusted HR of 0.42 (95% CI, 0.20-0.92; p=0.03). The results also held in the overall

39

40 41 HCV cohort. The number needed to treat for one fewer CKD at 7 years was 58. The

42 43

44 reduced risk of CKD was greatest (0.35; 0.14-0.87; p=0.024) in HCV-infected

46

47 patients who received six months or more of IBT. Multivariable stratified analysis

49 50

51 verified that greater risk reduction of CKD was present in HCV-infected patients with

52 53

54 hyperlipidemia (0.23; 0.06-0.91; p=0.037), diabetes (0.28; 0.09-0.91; p=0.034), or

55 56

57 hypertension (0.34; 0.12-0.94; p=0.037), and those without coronary heart disease

58 59

60 (0.34; 0.12-0.94; p=0.037). In conclusion, IBT, especially for six or more months, is

11

45

48

4 associated with reduced risk of CKD in HCV-infected patients. Hyperlipidemia,

5 6 7

diabetes, hypertension, and coronary heart disease can modify this association.

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

4 Introduction

6 7

8 Hepatitis C virus (HCV) infection and chronic kidney disease (CKD) are major public

9 10

11 health issues in Taiwan and worldwide.

Apart from major liver complications,

12 13

14 mounting evidence indicates that HCV adversely affects renal function. HCV is

15 16

17 associated with increased risks of CKD

and end-stage renal disease (ESRD),

even in

18 19

20 the absence of cirrhosis.

HCV also accelerates progression of CKD to ESRD in

21

22 23 patients with glomerulonephritis or diabetes.

24

25 26 Studies have reported that HCV infection can induce a state of oxidative stress and

27

28 29 overproduces pro-inflammatory cytokines that play a critical role in insulin

30

31 32 resistance.

Compensatory hyperinsulinemia in an insulin-resistant state further

33 34

35 enhances oxidative stress and promotes endothelial dysfunction, which contributes to

37

38 renal injury.

These observations suggest a biologically plausible mechanism for

40 41

42 increased risk of CKD in certain HCV-infected subjects. Up to 70% of HCV-infected

43 44

45 patients with and without cirrhosis display insulin resistance,

which is

46 47

48 predominately extrahepatic.

Interferon-based therapy (IBT) is considered the

49 50

51 mainstay of HCV treatment.

HCV eradication ameliorates insulin resistance

and

52 53

54 oxidative stress.

Therefore, it seems plausible that HCV-infected patients receiving

55

56 57 IBT also have decreased risk for CKD.

58 59 60 5

36

39

4 To date, there have been no nationwide cohort studies regarding the impact of IBT

5 6 7

on CKD risk in HCV-infected patients, and it is unclear if certain subsets of

8 9

10 HCV-infected patients receiving IBT are more likely to have decreased CKD risk.

12 13

14 Taiwan is a particularly suitable setting for examining the relationship of IBT for

15 16

17 HCV with CKD because it has a high prevalence of both conditions.

Moreover, the

18 19

20 burden of CKD and HCV infection is rising annually.

We hypothesized that IBT for

21 22

23 HCV would reduce CKD risk, given the efficacy of IBT in ameliorating insulin

24 25

26 resistance and oxidative stress.

Hence, we examined this association using

27

28 29 reimbursement claims data from the Taiwan National Health Insurance Research

30

31 32 Database (NHIRD) during a follow-up period of 7 years.

33 34 35 36 37 38

39 Results

41 42

43 Baseline characteristics of the HCV cohort

44 45

46 Baseline characteristics and follow-up status of the overall and matched HCV cohorts

47 48

49 are presented in Table 1. The mean (SD) interval from HCV diagnosis to start of

50 51

52 IBT was 2.4±2.0 years, and the mean duration of IBT was 0.6±0.9 years. In the

53

54 55 matched HCV cohort, there were no significant differences of baseline covariates

56

57 58 between the two cohorts, except for enrollee category (EC); the treated cohort had a

59 60

higher percentage of EC3. In the overall or matched HCV cohort, the percentage of

11

40

4 CKD events and competing mortality was lower in the treated cohort than in the

5 6 7

untreated cohort (all p<0.0001).

8 9 10 11 12 13

14 Cumulative incidences of incident CKD between the treated and

15 16

17 untreated cohorts

18 19

20 In the matched HCV cohort, the 3-, 5-, and 7-year cumulative incidences of CKD in

21 22

23 the presence of competing mortality were 0.71% vs. 1.90%, 1.13% vs. 3.17%, and

24 25

26 2.58% vs. 4.29%, respectively, in the treated cohort compared with the untreated

27

28 29 cohort (all p<0.001) (Table 2). Therefore, the risk of CKD was significantly lower in

30

31 32 the treated cohort (7-year cumulative incidence, 2.6%; 95% CI, 0.7-6.9%) than in the

33

34 35 untreated cohort (4%; 95% CI, 3.5-5.2%) (p=0.008) (Figure 1). The NNT associated

36

37 38 with one fewer CKD after 3, 5, and 7 years were 84, 49, and 58, respectively. The

39

40 41 results were similar in the overall HCV cohort.

42 43 44 45 46 47

48 Risk of CKD in the HCV cohort in the presence of competing mortality

49 50

51 In the matched HCV cohort, CKD was independently associated with male sex (aHR,

52 53

54 1.90; 95% CI, 1.27-2.84, p=0.002) and hypertension (aHR, 2.36; 95% CI, 1.37-4.07,

55 56

57 p=0.002) (Table 3). The treated cohort had a significantly lower risk of CKD than the

58 59 60

[Insert Running title of <72 characters] 7

4 untreated cohort (aHR, 0.42; 95% CI, 0.20-0.92, p=0.03). The results were similar in

5 6 7

the overall HCV cohort.

8 9 10 11 12 13

14 Impact of IBT duration on the risk of CKD

15 16

17 In the matched HCV cohort, the treated cohort who received 6 months or more of IBT

18 19

20 had a significantly lower risk of CKD (aHR, 0.35; 95% CI, 0.14-0.87; p=0.024)

21 22

23 (Table 4), compared with the untreated cohort and the treated cohort who received

24 25

26 less than 6 months of IBT.

27 28 29 30

31 32 Stratified analysis

33

34 35 In the matched HCV cohort, IBT was consistently associated with decreased risk of

36

37 38 CKD across all subgroups (Figure 2). The reduction of risk was significant in subjects

39

40 41 with (aHR, 0.23; 95% CI, 0.06-0.91; p=0.037) versus without (0.63; 0.24-1.61;

42 43

44 p=0.33) hyperlipidemia, with (0.28; 0.09-0.91; p=0.034) versus without (0.63;

46

47 0.22-1.79; p=0.39) diabetes, with (0.34; 0.12-0.94; p=0.037) versus without (0.62;

49 50

51 0.22-1.79; p=0.39) hypertension, without (0.34; 0.12-0.94; p=0.037) versus with (0.62;

52 53

54 0.18-2.07; p=0.43) coronary heart disease, and subjects older (0.40; 0.17-0.92; p=0.03)

55 56

57 versus younger (0.36; 0.05-2.80; p=0.36) than 50 years. However, the magnitude of

58 59

60 risk reduction was more pronounced in the first four comorbidities.

45

48

4 5 6 7

8 Discussion

10 11

12 To our knowledge, this is the first large nationwide cohort study not only to

13 14

15 demonstrate that IBT for HCV infection was significantly associated with a 58%

16 17

18 reduction of incident CKD risk over a 7-year study period after propensity score

19

20 21 matching and adjustment for potential confounders and competing mortality, and that

22

23 24 longer treatment duration (at least 6 months) may be required for IBT to exert its

25

26 27 protecting effect on CKD, but also to characterize HCV-infected patients who are

28

29 30 more likely to benefit from this treatment. The cumulative incidence of CKD was

31 32 33

significantly lower in HCV-infected patients receiving IBT than in those without IBT.

34 35

36 The NNT in association with one patient free of CKD at 3, 5, 7 years after IBT was 84,

38 39

40 49, and 58, respectively. The attenuated risk of CKD was more pronounced in

41 42

43 HCV-infected patients with diabetes, hypertension, hyperlipidemia, and those without

44 45

46 coronary heart disease. These findings suggest that HCV infection may have a role in

47 48

49 the pathogenesis of renal injury, and also implicate that treatment of HCV infection

50 51

52 may improve renal outcome. This information has important clinical implications for

53 54

55 the design of surveillance programs that assess HCV infection and CKD and for the

56

57 58 development of clinical practice guidelines.

59 60

Reported studies

that address the association of IBT for HCV with renal

9

37

4 outcome are few and lacking detail. A hospital-based retrospective cohort study

5 6 7

analyzed 650 HCV-infected cirrhotic Japanese patients who received IBT for periods

8 9

10 of 4-52 weeks and had normal renal function 3 months after IBT termination; the

12 13

14 authors further divided them into sustained virological response (SVR) and non-SVR

15 16

17 groups. The authors found that the development of CKD was associated with

18 19

20 non-SVR rather than with HCV genotype, ribavirin combination, and type of

21 22

23 interferon during a mean follow-up period of 6.5 years. However, IBT intervention

24 25

26 for protecting new development of CKD was not evaluated in that study. A

27 28

29 hospital-based cross-sectional study

analyzed 552 HCV-infected American patients;

30

31 32 the authors found that 2.5% of 159 HCV-infected patients who had ever received IBT

33

34 35 and 12.5% of 393 HCV-infected patients without IBT developed CKD during a

36

37 38 7-year follow-up period, and indicated that history of IBT was associated with

39

40 41 reduced risk of CKD (odds ratio, 0.18; 95% CI, 0.06-0.56). However, the authors did

42 43

44 not report the kind and duration of IBT and the effect of IBT intervention and SVR on

46

47 CKD risk. A nationwide Taiwanese cohort study

indicated that IBT for HCV used

49 50

51 for at least 4 months was associated with reduced risk of ESRD (HR, 0.16; 95% CI,

52 53

54 0.07-0.33) in a diabetic cohort without significant comorbidities during an 8-year

55 56

57 follow-up period. However, this result may not be extrapolated to most HCV-infected

58 59

60 patients because the HCV population is highly comorbid.

Moreover, the authors did

11

45

48

4 not provide NNT associated with one patient free of ESRD. We believe that our

5 6 7

results can be generalized to HCV population because we did not exclude the HCV

8 9

10 cohort with significant comorbidities; the method used to find our HCV cohort was

12

13 similar to that of a prior NHIRD-based nationwide study of HCV cohort.

Moreover,

15 16

17 to evaluate the effect of IBT intervention for HCV on CKD risk, we used a large

18 19

20 nationwide dataset, which afforded considerable statistical power and allowed

21 22

23 long-term tracking of incident CKD events. We evaluated the number of patients

24 25

26 needed to be treated with IBT for 3 months or more for one additional patient to

27 28

29 benefit, which was not evaluated in any of the three above-mentioned studies.

30

31 32 Although the NNT for one fewer CKD at 7 years was 58 in our study, the overall

33

34 35 reduction in CKD burden from the HCV population may be substantial, given that 3

36

37 38 to 4 million people are newly infected each year

and the incidence of CKD was

39

40 41 1.66-fold higher in a HCV cohort than a non-HCV cohort.

42 43

44 The exact mechanism that IBT for HCV is associated with reduced CKD risk is

46

47 unclear. However, the effect of IBT for HCV on the amelioration of insulin resistance

49 50

51 may underlie the association revealed in this study. Most HCV-infected patients with

52 53

54 and without cirrhosis have insulin resistance and compensatory hyperinsulinemia,

55 56

57 which is associated with increased oxidative stress and endothelial dysfunction, and

58 59

60 subsequent renal injury.

The mechanism through which IBT alleviates insulin

11

14

45

48

4 resistance is most likely mediated via viral eradication,

and SVR is an indicator of

5 6 7

successful HCV eradication and clinical cure.

Mounting evidence suggests that

8 9

10 attainment of SVR decreases insulin resistance

that occurs predominantly in

12

13 extrahepatic sites

and oxidative stress markers

in HCV-infected patients. Further

15 16

17 research is warranted to better understand the mechanism because this study could not

18 19

20 directly examine the status of insulin resistance.

21 22

23 Insulin resistance is a hallmark of hyperlipidemia, diabetes, and hypertension, all of

24 25

26 which are components of the metabolic syndrome.

This relationship may account for

27

28 29 our results that HCV-infected patients who had hyperlipidemia, diabetes, and

30

31 32 hypertension obtained more benefits of IBT in CKD risk. The difference in sex had no

33

34 35 influence on achieving SVR,

which may account for our result that there was no

36

37 38 beneficial difference of IBT on CKD risk in male and female HCV-infected patients.

39

40 41 We also found that there was no beneficial difference of IBT on CKD risk in

42 43

44 HCV-infected patients with and without cirrhosis, a result that is inconsistent with

46

47 those of one previous study.

Further research is warranted to better understand this

49 50

51 similarity in outcome between cirrhotic and non-cirrhotic patients. Even though the

52 53

54 NHIRD lacks individual information on HCV genotype and SVR

and we could not

55 56

57 directly show how SVR influences the above-mentioned associations, we believe that

58 59

60 the lower CKD risk resulted from HCV elimination in the treated cohort. Thus we are

11

14

45

48

4 confident of IBT’s efficacy in the treated cohort, because IBT generally achieves SVR

5 6 7

exceeding 70% in Taiwan, where a favorable genetic variation in interleukin-28B is

8 9

10 prevalent.

Moreover, there have been several NHIRD-based nationwide cohort

12 13

14 Taiwanese studies indicating the benefit of IBT on HCV-related liver and extrahepatic

15 16

17 complications.

18 19

20 The major strength of our study is that it was designed to reduce selection bias

21 22

23 (through the use of a large nationwide and highly representative sample with random

24 25

26 sampling and the use of propensity score matching to optimize comparability); reduce

27

28 29 environmental effects (because of the availability of socioeconomic indicators for all

30

31 32 subjects);

avoid detection bias (because of the universal availability of medical

33

34 35 services);

; avoid immortal time bias

(because the time when patients received

36

37 38 IBT was chosen as the entry of observation); and prevent overestimation of nonfatal

39

40 41 outcomes in the untreated cohort by using competing risk analysis.

In addition, the

42 43

44 study population was well defined and follow-up was complete because our design

46

47 relied on the universal coverage of Taiwan’s NHI, which fully reimbursed IBT for

49 50

51 HCV treatment and thus minimized disparity in healthcare accessibility or financial

52 53

54 status as a determinant for receiving IBT. Although unmeasured confounders may still

55 56

57 exist, as with any observational study, we believe the method we used are solid and

58 59 60 11

45

48

4 our finding of decreased risk of CKD following IBT for HCV-infected subjects is

5 6 7

valid.

8 9

10 Our study had some limitations. First, we were unable to document the adverse

12 13

14 reactions related to IBT. Nevertheless, we enrolled HCV-infected patients receiving 3

15 16

17 months or more of IBT

into our analysis to exclude most noncompliant patients.

18 19

20 Second, the actual compliance with medication was unknown. Nonetheless, excessive

21 22

23 prescription is impossible because of the strict regulations for IBT in Taiwan. Third,

24 25

26 misclassification of diseases may occur when an administration database is used.

27

28 29 However, the NHI Administration established an audit and penalty system for quality

30

31 32 monitoring and assurance to ensure accuracy of claims.

Moreover, both CKD and

33

34 35 viral hepatitis are important health problems in Taiwan, so the government has strict

36

37 38 guidelines for diagnosis,

and the diagnoses of CKD and HCV by ICD-9 codes have

39

40 41 been applied in several NHIRD-based nationwide cohort studies.

We also

42 43

44 adopted the standard methodology (one inpatient or two outpatient diagnosis codes) to

46

47 capture CKD patients in claims data.

Fourth, the NHIRD lacks information on

49 50

51 family history of kidney diseases, lifestyle, body weight, and laboratory data (e.g.

52 53

54 SVR, HCV RNA and genotype, creatinine and urinalysis). Thus, we could not include

55 56

57 these variables in the PS analysis and clarify the relationships of SVR, obesity, CKD

58 59

60 severity (stage), viral count and genotype with CKD risk. Finally, although the SVR

11

45

48

4 rates to IBT between the Asian and Western non-HCV genotype-1 (HCV-1) patients

5 6 7

are comparable, the SVR rate to IBT in Asian HCV-1 patients is higher than that in

8 9

10 Western HCV-1 patients,

largely as a result of interleukin-28B genotypic

12

13 polymorphism.

Thus, caution is recommended before applying our results to the

15 16

17 West.

18 19

20 In conclusion, this national cohort study indicates that CKD risk reduction is

21 22

23 greater in HCV-infected patients who receive IBT, as compared with those who do

24 25

26 not receive IBT. This finding implies that HCV infection may have a pathogenic role

27

28 29 in the development of CKD. Noteworthy also, CKD risk is reduced more in

30

31 32 HCV-infected patients receiving IBT for 6 months or more and in those with

33

34 35 hyperlipidemia, diabetes, hypertension, and without coronary heart disease. These

36

37 38 results may provide evidence to justify the long-term use and renal benefit of IBT in

39

40 41 HCV-infected patients.

42 43 44 45 46 47

48 CONCISE METHODS

50 51

52 Database

53

54 55 This cohort study used outpatient and inpatient claims from the NHIRD from 1996 to

56

57 58 2010, which is released by the National Health Research Institutes (NHRI) for

59 60

Taiwan’s National Health Insurance (NHI) Program. The NHI is a government,

11

14

49

4 compulsory-enrolment, single-payer system that had a coverage rate of more than

5 6 7

99% by the end of 2010, and adopts ICD-9 diagnosis codes for provider payment

8 9

10 applications. The NHIRD lacks information on laboratory and lifestyle data and

12 13

14 severity of the disease condition. Our previous research provided details of the

15 16

17 NHIRD.

In brief, the NHIRD has detailed healthcare data of 25.68 million

18 19

20 enrollees (99.9% of the population of Taiwan) based on a random sample of all

21 22

23 enrollees of the NHI program. There were no significant differences in age, sex, or

24 25

26 healthcare costs between the sample group and all enrollees. The NHI Administration

27

28 29 performs a medical quality monitoring and assurance program every month, including

30

31 32 chart reviews, charge audits, and heavy penalties for inappropriate charges or

33

34 35 malpractice. Therefore, it is generally believed that these checks and balances can

36

37 38 ensure accurate coding and further minimize misclassification error.

The study

39

40 41 was approved by our institutional review board. Informed consent was not required

42 43

44 because this is a secondary data analysis.

46 47 48 49 50

51 Study Population

52 53

54 We identified all patients who had a first-time diagnosis of HCV infection

55 56

57 (ICD-9-CM codes 070.41, 070.44, 070.51, 070.54, V02.62)

between 1 January

58 59

60 2004 and 31 December 2007 from the outpatient and inpatient claims. A total of 9639

11

45

4 HCV-infected patients were identified. Patients who were aged less than 18, had

5 6 7

claim-based diagnoses of HCV before 1 January 2004, of HBV (ICD-9-CM codes

8 9

10 070.22, 070.23, 070.32, 070.33, V02.61)

and renal transplantation (ICD-9-CM code

12 13

14 V420) from 1996 to 2010, or of CKD and IBT received before the first HCV

15 16

17 diagnosis were excluded, resulting in a total of 8810 patients with newly diagnosed

18 19

20 HCV.

21 22 23 24 25

26 Study Cohorts

27

28 29 The HCV cohort was divided into two cohorts based on the use of IBT, including

30

31 32 interferon alpha, pegylated interferon alpha-2a and pegylated interferon alpha-2b.

33

34 35 The NHI Administration has been reimbursing IBT (Table S1) for HCV for 6 months’

36

37 38 duration for all genotypes.

Thus, patients who received IBT for 3 months or more

39

40 41 were designated as the treated cohort (n=919). As the combination of ribavirin and

42 43

44 interferon is a common treatment regimen, we also extracted the use of ribavirin; most

46

47 patients (98.9%) were prescribed combination therapy. The index date of the treated

49 50

51 cohort was defined as the date of the start of IBT. Patients who never received IBT

52 53

54 between 2004 and 2010 were designated as the untreated cohort (n=7828). The index

55 56

57 date of the untreated cohort was defined as the first occurrence of a HCV claim during

58 59

60 the entry period. Thus, the overall HCV cohort included 8747 patients.

11

45

48

4 For each treated patient, four untreated patients were selected at or after the day

5 6 7

when IBT was initiated in the treated cohort according to the propensity score that

8 9

10 was calculated to adjust for the baseline differences between patients with and those

12 13

14 without IBT. The propensity score was estimated by the logistic regression built on

15 16

17 the baseline variables including age, sex, comorbidities, geographic region,

18 19

20 urbanization level, enrollee category, number of medical visits, and Deyo-Charlson

21 22

23 Comorbidity Index (CCI) score. The propensity score model was reliable

24 25

26 (Hosmer-Lemeshow test p=0.06) and provided fair discrimination between the

27 28

29 cohorts (c-index, 0.63).

Thus, the matched HCV cohort included 4595 patients.

30 31 32 33

34 35 Definition of CKD

36

37 38 The claims-based diagnosis of CKD was defined by the presence of one inpatient or

39

40 41 two outpatient ICD-9 code 585

in the claims and without catastrophic illness

42 43

44 registration cards for ESRD (indicating the need for renal replacement therapy). The

46

47 ICD-9 code 585 is consistent with the KDOQI/KDIGO (Kidney Disease Outcomes

49 50

51 Quality Initiative/Kidney Disease: Improving Global Outcomes) definition of CKD

52 53

54 stages 1-5, which allows for comparisons of the incidence and prevalence of CKD in

55 56

57 Taiwan and the United States.

However, the CKD stage (severity) cannot be assessed

58 59

60 in the NHIRD.

11

45

48

4 5 6

7 Main Outcome Measurement

9

10 Both cohorts were followed from the index date to the first diagnosis of CKD, death,

12 13

14 or the end of 2010, whichever came first. Because IBT has been shown to decrease

15 16

17 mortality in HCV-infected patients,

censoring resulting from death was regarded as

18 19

20 informative and was adjusted by using competing risk analyses. Death was defined by

21 22

23 withdrawal from the NHI program.

24 25 26 27

28 29 Potential Confounders

30

31 32 We recorded the claims-based diagnoses of comorbidities associated with CKD

33

34 35 between 1 January 1996 and the index date according to ICD-9 codes, including

36

37 38 diabetes (ICD-9 code 250), hypertension (ICD-9 codes 401-405), coronary heart

39

40 41 disease (ICD-9 codes 410-414), hyperlipidemia (ICD-9 codes 272-272.4), and

42 43

44 cirrhosis (ICD-9 codes 571.2, 571.5, 571.6).

CKD was associated with geographic

46

47 region of residence and socioeconomic status.

Thus, we recorded geographic regions

49 50

51 (northern, central, southern, or eastern Taiwan) in order to reduce potential

52 53

54 confounding by differential accessibility of medical care,

and urbanization level

55 56

57 (urban, suburban, and rural) and enrollee category (EC), from EC1 (highest status) to

58 59

60 EC4 (lowest status), as proxy measures of socioeconomic status, to minimize

8

11

45

48

4 environmental effects.

We also considered the number of medical visits

as a

5 6 7

potential confounder and used the CCI score for control of confounding in studies

8 9

10 using administrative databases.

Finally, we considered propensity score in

12 13

14 regression adjustment to control for confounding in healthcare utilization

15 16

17 databases

and to reduce bias in the background covariates between the two

18 19

20 cohorts.

21 22 23 24 25

26 Statistical Analyses

27

28 29 We calculated and compared the cumulative incidences of CKD by use of the

30

31 32 modified Kaplan-Meier method and Gray’s method,

and tested differences in the

33

34 35 full time-to-event distributions between the study cohorts using log-rank test. The

36

37 38 number needed to treat (NNT) represented the number of patients needed to be treated

39

40 41 to yield one fewer CKD; the NNT was calculated with the inverse of the absolute risk

42 43

44 reduction.

After ensuring the assumption of proportional hazards, we used the

46

47 modified Cox proportional hazard model to examine the association of IBT with CKD

49

50 risk,

with adjustment for all covariates (age per year, sex, comorbidities, geographic

52 53

54 region, urbanization level, enrollee category, number of medical visits, CCI score, and

55 56

57 propensity score). We further performed a stratified analysis of the effect of IBT on

58 59

60 CKD risk and compared the effect of IBT duration (<6 vs. ≧6 months) on CKD risk

11

45

48

51

4 in the propensity score-matched HCV cohort. We analyzed all data with SAS (version

5 6 7

9.2; SAS Institute, Inc., Cary, N.C.) and considered a two-sided p-value less than 0.05

8 9

10 as statistically significant.

12 13 14 15 16 17

18 Conflict of Financial Interests

19

20 21 None.

22 23 24 25 26

27 28 Acknowledgements

30 31

32 This study is based partly on secondary data from the National Health Insurance

33 34

35 Research Database provided by the National Health Insurance Administration,

36 37

38 Ministry of Health and Welfare, Taiwan. The interpretations and conclusions herein

39 40

41 do not represent the views of the National Health Insurance Administration, Ministry

42 43

44 of Health and Welfare, or National Health Research Institutes.

45 46 47 48 49

50 51 References

52 53

54 1. Chacko EC, Surrun SK, Mubarack Sani TP, Pappachan JM: Chronic viral

56 57

58 hepatitis and chronic kidney disease. Postgrad Med J 86: 486-492, 2010.

59 60

2. Wen CP, Cheng TY, Tsai MK, Chang YC, Chan HT, Tsai SP, Chiang PH, Hsu

11

29

55

4 CC, Sung PK, Hsu YH, Wen SF: All-cause mortality attributable to chronic kidney

5 6 7

disease: a prospective cohort study based on 462293 adults in Taiwan. Lancet 371:

8 9

10 2173-2182, 2008.

12 13

14 3. Chen YC, Lin HY, Li CY, Lee MS, Su YC: A nationwide cohort study suggests

15 16

17 that hepatitis C virus infection is associated with increased risk of chronic kidney

18 19

20 disease. Kidney Int 85: 1200-1207, 2014.

21 22

23 4. Tsui JI, Vittinghoff E, Shlipak MG, Bertenthal D, Inadomi J, Rodriguez RA,

24 25

26 O'Hare AM: Association of hepatitis C seropositivity with increased risk for

27

28 29 developing end-stage renal disease. Arch Intern Med 167: 1271-1276, 2007.

30

31 32 5. Noureddine LA, Usman SA, Yu Z, Moorthi RN, Moe SM: Hepatitis C increases

33

34 35 the risk of progression of chronic kidney disease in patients with glomerulonephritis.

36

37 38 Am J Nephrol 32: 311-316, 2010.

39

40 41 6. Crook ED, Penumalee S, Gavini B, Filippova K: Hepatitis C is a predictor of

42 43

44 poorer renal survival in diabetic patients. Diabetes Care 28: 2187-2191, 2005.

46

47 7. Harrison SA: Insulin resistance among patients with chronic hepatitis C: etiology

49 50

51 and impact on treatment. Clin Gastroenterol Hepatol 6: 864-876, 2008.

52 53

54 8. Sarafidis PA, Ruilope LM: Insulin resistance, hyperinsulinemia, and renal injury:

55 56

57 mechanisms and implications. Am J Nephrol 26: 232-244, 2006.

58 59 60 11

45

48

4 9. Stättermayer AF, Rutter K, Beinhardt S, Scherzer TM, Stadlmayr A, Hofer H,

5 6 7

Wrba F, Steindl-Munda P, Krebs M, Datz C, Trauner M, Ferenci P: Association of the

8 9

10 IL28B genotype with insulin resistance in patients with chronic hepatitis C. J Hepatol

12 13

14 57: 492-498, 2012.

15 16

17 10. Perico N, Cattaneo D, Bikbov B, Remuzzi G: Hepatitis C infection and chronic

18 19

20 renal diseases. Clin J Am Soc Nephrol 4: 207-220, 2009.

21 22

23 11. Milner KL, van der Poorten D, Trenell M, Jenkins AB, Xu A, Smythe G, Dore

24 25

26 GJ, Zekry A, Weltman M, Fragomeli V, George J, Chisholm DJ: Chronic hepatitis C

27

28 29 is associated with peripheral rather than hepatic insulin resistance. Gastroenterology

30

31 32 138: 932-941, 2010.

33

34 35 12. Rosen HR: Clinical practice. Chronic hepatitis C infection. N Engl J Med 364:

36

37 38 2429-2438, 2011.

39

40 41 13. Conjeevaram HS, Wahed AS, Afdhal N, Howell CD, Everhart JE, Hoofnagle JH:

42 43

44 Changes in insulin sensitivity and body weight during and after peginterferon and

46

47 ribavirin therapy for hepatitis C. Gastroenterology 140: 469-477, 2011.

49 50

51 14. Pearlman BL, Traub N: Sustained virologic response to antiviral therapy for

52 53

54 chronic hepatitis C virus infection: a cure and so much more. Clin Infect Dis 52:

55 56

57 889-900,

2011.

58 59 60 11

45

48

4 15. Serejo F, Emerit I, Filipe PM, Fernandes AC, Costa MA, Freitas JP, de Moura

5 6 7

MC: Oxidative stress in chronic hepatitis C: the effect of interferon therapy and

8 9

10 correlation with pathological features. Can J Gastroenterol 17: 644-650, 2003.

12 13

14 16. Arase Y, Suzuki F, Kawamura Y, Suzuki Y, Kobayashi M, Matsumoto N, Akuta

15 16

17 N, Sezaki H, Hosaka T, Ogawa K, Imai N, Seko Y, Saito S, Ikeda K, Kobayashi M,

18 19

20 Kumada H: Development rate of chronic kidney disease in hepatitis C virus patients

21 22

23 with advanced fibrosis after interferon therapy. Hepatol Res 41: 946-954, 2011.

24 25

26 17. Hsu YC

, Lin JT, Ho HJ, Kao YH, Huang YT, Hsiao NW, Wu MS, Liu YY, Wu

27

28 29 CY: Antiviral treatment for hepatitis C virus infection is associated with improved

30

31 32 renal and cardiovascular outcomes in diabetic patients. Hepatology 59: 1293-1302,

33

34 35 2014.

36

37 38 18. Satapathy SK, Lingisetty CS, Williams S: Higher prevalence of chronic kidney

39

40 41 disease and shorter renal survival in patients with chronic hepatitis C virus infection.

42 43

44 Hepatol Int 6: 369–378, 2012.

46

47 19. Louie KS, St Laurent S, Forssen UM, Mundy LM, Pimenta JM: The high

49 50

51 comorbidity burden of the hepatitis C virus infected population in the United States.

52 53

54 BMC Infect Dis 12: 86, 2012.

55 56

57 20. Hsu CS, Kao JH, Chao YC, Lin HH, Fan YC, Huang CJ, Tsai PS:

58 59

60 Interferon-based therapy reduces risk of stroke in chronic hepatitis C patients: a

11

45

48

4 population-based cohort study in Taiwan. Aliment Pharmacol Ther 38: 415-423,

5 6 7

2013.

8 9

10 21. Liu CH, Kao JH: Nanomedicines in the treatment of hepatitis C virus infection in

12 13

14 Asian patients: optimizing use of peginterferon alfa. Int J Nanomedicine 9: 2051-2067,

15 16

17 2014.

18 19

20 22. Ruderman NB, Carling D, Prentki M, Cacicedo JM: AMPK, insulin resistance,

21 22

23 and the metabolic syndrome. J Clin Invest 123: 2764-2772, 2013.

24 25

26 23. Manns MP, McHutchison JG, Gordon SC, Rustgi VK, Shiffman M, Reindollar R,

27

28 29 Goodman ZD, Koury K, Ling M, Albrecht JK: Peginterferon alfa-2b plus ribavirin

30

31 32 compared with interferon alfa-2b plus ribavirin for initial treatment of chronic

33

34 35 hepatitis C: a randomised trial. Lancet 358: 958-965, 2001.

36

37 38 24. Hsu YC, Ho HJ, Wu MS, Lin JT, Wu CY: Postoperative peg-interferon plus

39

40 41 ribavirin is associated with reduced recurrence of hepatitis C virus-related

42 43

44 hepatocellular carcinoma. Hepatology 58: 150-157, 2013.

46

47 25. Sun Y, Chang YH, Chen HF, Su YH, Su HF, Li CY: Risk of Parkinson disease

49 50

51 onset in patients with diabetes: a 9-year population-based cohort study with age and

52 53

54 sex stratifications. Diabetes Care 35: 1047-1049, 2012.

55 56

57 26. Chen YC, Su YC, Li CY, Wu CP, Lee MS: A nationwide cohort study suggests

58 59

60 chronic hepatitis B virus infection increases the risk of end-stage renal disease among

11

45

48

4 patients in Taiwan. Kidney Int online publication, November 26, 2014;

5 6 7

doi:10.1038/ki.2014.363.

8 9

10 27. Wu CY, Chen YJ, Ho HJ, Hsu YC, Kuo KN, Wu MS, Lin JT: Association

12 13

14 between nucleoside analogues and risk of hepatitis B virus-related hepatocellular

15 16

17 carcinoma recurrence following liver resection. JAMA 308: 1906-1914, 2012.

18 19

20 28. Shariff SZ, Cuerden MS, Jain AK, Garg AX: The secret of immortal time bias in

21 22

23 epidemiologic studies. J Am Soc Nephrol 19: 841-843, 2008.

24 25

26 29. Sheu JJ, Kang JH, Lin HC: Hyperthyroidism and risk of ischemic stroke in

27

28 29 young adults: a 5-year follow-up study. Stroke 41: 961-966, 2010.

30

31 32 30. National Health Insurance Administration, Ministry of Health and Welfare,

33

34 35 Taiwan. Available: http://www.nhi.gov.tw. Accessed 1 August 2014.

36

37 38 31. Collins AJ, Foley RN, Herzog C, Chavers B, Gilbertson D, Ishani A, Kasiske B,

39

40 41 Liu J, Mau LW, McBean M, Murray A, St Peter W, Guo H, Li Q, Li S, Li S, Peng Y,

42 43

44 Qiu Y, Roberts T, Skeans M, Snyder J, Solid C, Wang C, Weinhandl E, Zaun D, Arko

46

47 C, Chen SC, Dalleska F, Daniels F, Dunning S, Ebben J, Frazier E, Hanzlik C,

49 50

51 Johnson R, Sheets D, Wang X, Forrest B, Constantini E, Everson S, Eggers P,

52 53

54 Agodoa L: United States Renal Data System 2008 Annual Data Report. Am J Kidney

55 56

57 Dis 53: S1-374, 2009.

58 59 60 11

45

48

4 32. Chen HF, Chen P, Li CY: Risk of malignant neoplasms of liver and biliary tract

5 6 7

in diabetic patients with different age and sex stratifications. Hepatology 52: 155-163,

8 9

10 2010.

12 13

14 33. Chen YC, Su YC, Lee CC, Huang YS, Hwang SJ: Chronic kidney disease itself

15 16

17 is a causal risk factor for stroke beyond traditional cardiovascular risk factors: a

18 19

20 nationwide cohort study in Taiwan. PLoS One 7: e36332, 2012.

21 22

23 34. Yu ML, Chuang WL: Treatment of chronic hepatitis C in Asia: when East meets

24 25

26 West. J Gastroenterol Hepatol 24: 336-345, 2009.

27

28 29 35. Gershon A, Croxford R, To T, Stanbrook MB, Upshur R, Sanchez-Romeu P,

30

31 32 Stukel T: Comparison of inhaled long-acting beta-agonist and anticholinergic

33

34 35 effectiveness in older patients with chronic obstructive pulmonary disease: a cohort

36

37 38 study. Ann Intern Med 154: 583-592, 2011.

39

40 41 36. Seliger SL: Comorbidity and confounding in end-stage renal disease. Kidney Int

42 43

44 77: 83-85, 2010.

46

47 37. Patorno E, Grotta A, Bellocco R, Schneeweiss S: Propensity score methodology

49 50

51 for confounding control in health care utilization databases. Epidemiol Biostat Public

52 53

54 Health 10: e8940, 2013.

55 56

57 38. D'Agostino RB, Jr.: Propensity scores in cardiovascular research. Circulation

58 59

60 115: 2340-2343, 2007.

11

45

48

4 39. Gray RJ: A class of K-sample tests for comparing the cumulative incidence of a

5 6 7

competing risk. Ann Stat 16: 1141-1154, 1988.

8 9

10 40. Altman DG: Confidence intervals for the number needed to treat. BMJ 317:

12 13

14 1309-1312, 1998.

15 16

17 41. Fine JP, Gray RJ: A proportional hazards model for the subdistribution of a

18 19

20 competing risk. J Am Stat Assoc 94: 496-509, 1999.

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

11

4 Figure Legends

6 7

8 Figure 1. Cumulative incidence of chronic kidney disease (CKD) in the propensity

9 10

11 score-matched HCV cohort with (treated, solid line) and without (untreated, dash line)

12 13

14 interferon-based therapy. Data were compiled after adjustment for competing

15 16

17 mortality.

18

19 20 Figure 2. Stratified analysis for risk of chronic kidney disease (CKD) in association

21

22 23 with interferon-based therapy in the propensity score-matched HCV cohort, with

24

25 26 adjustment for competing mortality.

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[Insert Running title of <72 characters] 29

5

[Insert Running title of <72 characters] 30

42 43 44

45 ScholarOne support: 888-503-1050

1 [First Authors Last Name] Page 30

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

For

Peer

Review

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

For

Peer

Review

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

For

Peer

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

For

Peer

Review

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

For

4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40

For

Peer

1 [First Authors Last Name] Page 36

2

3 4 Table S1. Interferon-based regimens for treatment of HCV infection

5

6 in Taiwan

7

8 Type Interferon-based regimens

9

10 1 Interferon alpha-2a

11

12 2 Interferon alpha-2a + Ribavirin

13 14 3 Interferon alpha-2b

15 16 4 Interferon alpha-2b

17 18 5 Interferon alfacon-1 + Ribavirin

19 20 6 Peginterferon alfa-2a + Ribavirin

21 22 7 Peginterferon alfa-2a

23 24 8 Peginterferon alfa-2a + Ribavirin

25 26 9 Peginterferon alpha-2b

28 10 Peginterferon alpha-2b + Ribavirin

30 Note: These regimens were prescribed serially for 4 to 6 months.

31

32

33

34

35

36

37

27

29

41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[Insert Running title of <72 characters] 36

ScholarOne support: 888-503-1050

e 3 7 o f 3 9

2 3 4 Fig ure 1

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59

Authors Last Name]

Page 37

For

Peer

Review

Run nin g title of

<72 cha ract ers]

37

ScholarOne support: 888-503- 1050

1 2 3

F i g u r e

2

4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2

2 6

29 30 31 32 33 34 35 36 37 38 39 40 41

42 [Insert Running title of

<72 characters] 38

43 44

45 ScholarOne support: 888-503-1050 46

47

Page 38

For

Peer

Review

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

42 [Insert Running title of <72 characters]

45 ScholarOne support: 888-503- 1050

46

47

i

r

s

t

A

u

t

h

o

r

s

L

a

s

t

N

a

m

e

]

P

a

g

e

3

9