Association between hepatitis C antiviral treatment and diabetes mellitus: A case series

Abstract

The hepatitis C virus (HCV) infects 185 million people worldwide, and diabetes mellitus (DM) affects 415 million. There has long been a possible association between DM and liver outcomes for patients with HCV infection. We present two cases of worsening glucose resistance and one case of emergent type 1 DM after completion of HCV. The complex interactions between HCV infection and both type 1 and type 2 DM have not yet been elucidated. In addition, consequences and side effects of treatment options for HCV have not been fully studied in the diabetic population. Our case series illustrates a potential complication of HCV treatment, which may warrant additional consideration prior to initiation of therapy as well as monitoring and surveillance post-cure.

The hepatitis C virus (HCV) infects between 110 and 185 million people worldwide and remains a major cause of liver-related death globally (1,2). Type 1 diabetes (T1DM) and type 2 diabetes (T2DM) collectively affect approximately 415 million people worldwide. The presence of insulin resistance and diabetes is known to negatively impact liver disease by accelerating the progression of liver fibrosis and significantly increasing the risk of hepatocellular carcinoma (3–5).

A possible association between T2DM and chronic HCV infection was first noted in the 1970s, and multiple subsequent studies have demonstrated this relationship (6–10). This association has been described in both cirrhotic and non-cirrhotic patients (11). Suggested mechanisms leading to insulin resistance in HCV-infected patients include decreased incretin hormones, the effect of HCV on the pancreas and insulin signaling pathways in the liver, as well as damage done by inflammatory cytokines and reactive oxygen species (12).

There have also been several studies examining the improvement of T2DM after treatment of HCV (13–15). Interferon and ribavirin combination therapy resulting in sustained virologic response (SVR) was associated with improved insulin sensitivity (13).Also, the achievement of SVR with direct-acting antiviral (DAA) treatments (protease inhibitors, nucleotide and non-nucleotide polymerase inhibitors, and NS5A inhibitors) appears to induce a significant amelioration of glycemic control and reduction in fasting blood glucose (FPG) and hemoglobin A_1c (HbA_1c) (14,15).

Though the exact mechanism for the improvement in insulin resistance is not well understood, it is believed that the interaction between viral proteins and insulin signaling, as well as alterations in glucose homeostasis, may play significant roles (12,16,17).

On the other hand, the association between HCV and type 1 diabetes mellitus (T1DM) is less clear. There have been some rare case reports of HCV-related T1DM (18,19) as well as studies of possible autoimmunity due to molecular mimicry and proinflammatory cytokines (20,21).

There are rare reports of new-onset diabetes following HCV treatment (22). The development of fulminant T1DM following interferon-based HCV treatment has also been described (23). The risk, if any, for these outcomes in the context of HCV DAA treatment is unknown. Here, we report a case of new-onset T2DM in a cirrhotic patient after achieving SVR with DAA treatment, a case of T1DM emerging after combination DAA and interferon therapy, and a case of progression of T2DM after DAA treatment.

Case 1

SL is a 61-year-old man with genotype 3 hepatitis C who initially presented to The Ottawa Hospital Viral Hepatitis Clinic in March 2016. His medical history was significant for stage 3 proteinuric chronic kidney disease due to focal segmental glomerular sclerosis, hypertension, chronic obstructive pulmonary disease (COPD), injection drug use, and alcohol abuse more than 20 years ago. He was found to have compensated cirrhosis on transient elastography with a liver stiffness of 23.1 kPa. His controlled attenuation parameter (CAP) was 258, indicating mild to moderate inflammation and steatosis within the parenchyma of the liver. His viral load was 3.67 × 10⁶ IU/mL and there was evidence of active inflammation with elevated alanine transaminase (ALT) of 400 U/L (reference range 17–63 U/L) and aspartate aminotransferase (AST) of 231 U/L (reference range 15–37 U/L). His weight in November 2016, 6 months prior to treatment initiation, was 107.2 kg. His pre-treatment HbA_1c was 5.8% in March 2016.

The patient was started on treatment for his hepatitis C in May 2017 with sofosbuvir, velpatasvir, and ribavirin. He completed 12 weeks of treatment and achieved SVR.

Approximately 1–2 months after starting treatment, SL developed progressive fatigue, generalized weakness, polyuria, polydipsia, increased oral intake, and weight gain. Follow-up bloodwork in September 2017 revealed random glucose of 16.8 mmol/L and an HbA_1c of 6.4%. His hemoglobin was 153 g/L, and his creatinine was 113 μmol/L. He was diagnosed with T2DM and glucose therapy was initiated. Diabetes has been well-controlled on metformin 500 mg daily, with an HbA_1c of 5.1%. His weight in October 2017 was 109.5 kg.

Case 2

DN is a 72-year-old man with genotype 1b hepatitis C who was first seen in the hepatitis clinic in March 2017. His past medical history includes stage 3 proteinuric chronic kidney disease, hypertension, and benign prostatic hypertrophy. He was diagnosed with diabetes in 2016 and was initially treated with metformin alone. His HbA_1c in March 2017 was 8.0%, and his random glucose was 8.2 mmol/L. He was a previous user of intravenous drugs and consumed excess alcohol more than 10 years ago. On assessment, he had a transient elastography score of 47.2 kPa and a CAP of 342, indicating advanced fibrosis and severe inflammation or steatosis. His viral load was 5.87 × 10⁵ IU/mL. His weight in March 2017 was 117 kg.

HCV treatment with elbasvir and grazoprevir was started in May 2017. His treatment course was uncomplicated except for an episode of cellulitis. He completed 12 weeks of treatment in August 2017 and achieved SVR. No HbA_1c was performed at the end of treatment, but random glucose done in June 2017 was 9.1 mmol/L.

In October 2017, 5 months after initiation of HCV treatment, the patient was referred to endocrinology for hyperglycemia following symptoms of blurry vision. He was found to have a random blood glucose of 32.9 mmol/L and HbA_1c of 11.3%. At the time, his hemoglobin was 140 g/L, and his creatinine was at a baseline of 185 μmol/L. Given his chronic kidney disease, his metformin was discontinued, and he was started on insulin. Despite insulin therapy with dose adjustments, DN was admitted 5 days later with a hyperglycemic emergency (serum glucose of 30.1 mmol/L, anion gap of 12, and beta-hydroxybutyrate level of 0.37). Anti-GAD antibodies were negative. At present, his diabetes is relatively well-controlled with insulin lispro Mix30, 42 units at breakfast and 22 units at supper (HbA_1c of 7.7%). His weight was 114 kg in November 2017.

Case 3

YM is a 57-year-old man with genotype 3 HCV who was referred to our clinic in May 2013. His medical history included chronic obstructive pulmonary disease (COPD), vitiligo, and anxiety. He had a family history of autoimmune thyroid disorders, but no type 1 diabetes. He had initially been diagnosed with HCV in 1995. He stopped injection drug use in 1998 but was still drinking up to 15 beers per week in 2013. Bloodwork demonstrated an ALT of 338 U/L, AST of 311 U/L, and the HCV viral load was found to be 1.02 × 10⁷ IU/mL. Transient elastography identified cirrhosis-range liver stiffness (21.1 kPa). There was no clinical evidence of decompensated liver disease. His weight in April 2017 was 83.9 kg.

He initiated pegylated interferon and ribavirin in March 2014. He received a total of 24 weeks of therapy, which was complicated by a rash and elevated transaminases. His viral load was undetectable post-treatment. However, in January 2015, he was found to have a relapse of his HCV infection, with a viral load of 6.01 × 10⁶ IU/mL. Six months later, his ALT and AST had increased to 179 U/L and 147 U/L, respectively. He was started on sofosbuvir, pegylated interferon, and ribavirin in November 2015.

In January 2016, YM described acute onset fatigue, polydipsia, and polyuria. He also reported a 20 lb weight loss. His random blood glucose was found to be 22 mmol/L. His HbA_1c was 5.5% in December 2016 and increased to 6.7% one month later. His hemoglobin was 127 g/L, and his creatinine was 46 μmol/L. Given the acuity of symptom onset, anti-GAD antibodies were ordered and were found to be positive at >30 U/mL. He was diagnosed with T1DM and treated with insulin glargine. His weight in February 2016 was 76.1 kg.

YM completed his second course of HCV treatment in March 2016 and achieved SVR. He continues treatment with insulin glargine and insulin glulisine.

Discussion

Type 2 diabetes mellitus

Though the literature describing an association between HCV and T2DM is extensive, the suggestion of a causal relationship is controversial. One meta-analysis in 2008 reported that HCV was correlated with T2DM with an adjusted odds ratio (OR) of 1.7 (95% CI 1.2 to 2.2) and an adjusted hazard ratio (HR) of 1.7 (95% CI 1.3 to 2.1) (24). A second meta-analysis in 2012 demonstrated an overall OR of 1.7 (95% CI 1.2 to 2.4) (25). A more recent cohort study of 514,791 individuals in South Korea over 12 years showed an HR of 1.68 for HCV (95% CI 1.52 to 1.86) (5) and a meta-analysis of 31 studies determined an OR of 1.58 (95% CI 1.3 to 1.86) (26). However, other studies have disputed this correlation. A retrospective cohort study of 15,128 adults in the United States showed no significant association between HCV and T2DM (27). A more recent population-based study of 21,929 anti-HCV–positive patients in Scotland did not find evidence to indicate that HCV infection conveyed an increased risk of T2DM (28).

Accepting that HCV may perturb glucose homeostasis, it is plausible that HCV clearance with antiviral therapy may also result in subclinical and clinically relevant metabolic outcomes. We described two cases: one, a case of new-onset T2DM after completion of DAA and ribavirin; the second, a case of worsening T2DM after DAA treatment alone. Though certain studies have demonstrated improvement of glycemic control with the eradication of HCV (14,15,29), others have shown no improvements in insulin resistance (30). Our reports of new or worsening glycemic control may indicate a more complex relationship between HCV, DAAs, and diabetes.

The mechanism for the aggravation of insulin resistance and T2DM in patients treated with DAAs is unclear. Some studies have suggested a possible link with increased BMI associated with achieving SVR (31). This weight gain is thought to be due to decreased leptin levels after achieving SVR. However, our patients developed and exacerbated their diabetes too quickly to be associated with an increase in BMI. In fact, two of our patients lost weight after treatment with DAAs. It is also unclear if this phenomenon represents a rare, unknown side effect of DAAs.

The effect of HCV DAA treatments on insulin production and resistance is still unknown, and the risks with regard to diabetes are still controversial. Our cases highlight the close and complex interaction between HCV and diabetes. They underscore the importance of increased glucose monitoring in patients on HCV antiviral therapy, in particular for patients who already have diabetes or are at risk of developing diabetes. Further study is required to elucidate the mechanisms behind the improvements in glycemic control in certain patients, and deteriorations in others. Careful observation will also be important to ensure that the disruption of glucose metabolism is not an unknown side effect of DAAs. These studies will ensure the appropriate care of patients with HCV in the future.

Type 1 diabetes mellitus

Interferon therapy for HCV has been associated with T1DM in rare case reports (22,32,33). Studies in Japan have demonstrated that the mean onset of T1DM after initiation of interferon therapy was approximately 0.7 years, with the majority of patients experiencing a fulminant, sudden-onset form of the disease (34). A study in the Netherlands showed an incidence rate of 2.6% for T1DM in patients with HCV treated with interferon-alpha and ribavirin in combination (35).

In our case, the patient was initially treated unsuccessfully with interferon and ribavirin and subsequently treated again with a combination of interferon, ribavirin, and DAAs. He then developed symptoms of T1DM two months later. The rapidity of symptom onset and deterioration appear to indicate a fulminant T1DM pattern of insulin underproduction previously described in cases associated with interferon therapy (22). The presence of anti-GAD antibodies further confirmed the diagnosis of T1DM. Of note, our patient had a history of vitiligo, as well as a family history of autoimmune thyroid disease, which may indicate a predisposition to T1DM.

The mechanisms through which T1DM develops following interferon therapy remains unclear. Studies have hinted at the role of interferon-alpha being key to the increased expression of HLA-1 in pancreatic islet cells (36). This, in turn, can increase activation of cytotoxic CD8 T cells and a subsequent cascade that leads to the autoimmune destruction of beta islet cells.

In addition, the role of ribavirin in glucose homeostasis has not been studied. Indeed, cases of new-onset diabetes in the context of ribavirin use has always been in combination with interferon (34,35,37). Furthermore, the combination use of ribavirin with DAAs has not been found to be associated with the development of either T1DM nor T2DM (38–42). One case report described a patient who did not develop diabetes with interferon alone but did develop T1DM later with the addition of ribavirin (43). It also appears that the length of time to development of T1DM was significantly shorter with the combination of interferon and ribavirin than interferon alone (34).

The interplay between HCV, its treatments, and diabetes types 1 and 2 is complex. In the majority of cases, liver function, fibrosis, and stiffness improve after SVR through both interferon and DAA treatments (44–47). Though the majority of patients do not experience changes in glucose homeostasis (48), some patients experience improvement of diabetic control after achieving SVR (15,47,49). However, we describe three cases in which the treatment of HCV appears to have induced the development of insulin resistance or islet cell destruction.

Of note, each of our patients presented with symptoms of decompensated hyperglycemia and significantly elevated random glucose. The diagnosis of diabetes or worsening glycemic control should not be solely based on the HbA_1c due to its unreliable nature in certain medical conditions (50). Hemoglobinopathies, hemolytic anemia, iron deficiency, severe renal or hepatic disease can cause inaccurate HbA_1c, which can be misleading.

At this point, we do not know if the disruption in glucose homeostasis in these patients is due to the adverse effect of the treatment or due to the clearance of HCV. Studies suggest that interferon may have an important role in the development of T1DM, possibly through increasing autoimmunity. In the cases of T2DM, increased insulin resistance could represent an unknown and rare adverse effect of DAAs, which may be potentiated by ribavirin. Further research is needed to assess the incidence of T2DM among HCV patients treated with DAAs, with or without ribavirin. Clinicians should be aware of this association. Surveillance may be considered in some at-risk patients during and after HCV antiviral treatment.

Ethics Approval:

N/A

Informed Consent:

N/A

Registry and the Registration No. of the Study/Trial:

N/A

Animal Studies:

N/A

Funding:

No funding was received for this work.

Peer Review:

This article has been peer reviewed.