Effect of type 2 diabetes on risk for malignancies includes hepatocellular carcinoma in chronic hepatitis C

Authors


  • Potential conflict of interest: Dr. Suzuki is on the speakers' bureau of Bristol-Myers Squibb. Dr. Akuta is on the speakers' bureau of MSD and holds intellectual property rights with SRL. Dr. Kumada is on the speakers' bureau of MSD, Mitsubishi Tanabe Pharma, Dainippon Sumitomo Pharma, Bristol-Myers Squibb. He also holds intellectual property rights with SRL. Dr. Arase is on the speakers' bureau of MSD.

Abstract

The aim of this retrospective cohort study was to assess the cumulative development incidence and predictive factors for malignancies after the termination of interferon (IFN) therapy in Japanese patients for hepatitis C virus (HCV). A total of 4,302 HCV-positive patients treated with IFN were enrolled. The mean observation period was 8.1 years. The primary outcome was the first onset of malignancies. Evaluation was performed using the Kaplan-Meier method and Cox proportional hazard analysis. A total of 606 patients developed malignancies: 393 developed hepatocellular carcinoma (HCC) and 213 developed malignancies other than HCC. The cumulative development rate of HCC was 4.3% at 5 years, 10.5% at 10 years, and 19.7% at 15 years. HCC occurred significantly (P < 0.05) when the following characteristics were present: advanced histological staging, sustained virological response not achieved, male sex, advanced age of ≥50 years, total alcohol intake of ≥200 kg, and presence of type 2 diabetes (T2DM). T2DM caused a 1.73-fold enhancement in HCC development. In patients with T2DM, HCC decreased when patients had a mean hemoglobin A1c (HbA1c) level of <7.0% during follow-up (hazard ratio, 0.56; 95% confidence interval, 0.33-0.89; P = 0.015). The cumulative development rate of malignancy other than HCC was 2.4% at 5 years, 5.1% at 10 years, and 9.8% at 15 years. Malignancies other than HCC occurred significantly when patients were of advanced age of ≤50 years, smoking index (package per day × year) was ≥ 20, and T2DM was present. T2DM caused a 1.70-fold enhancement in the development of malignancies other than HCC. Conclusion: T2DM causes an approximately 1.7-fold enhancement in the development of HCC and malignancies other than HCC in HCV-positive patients treated with IFN. In T2DM patients, maintaining a mean HbA1c level of <7.0% reduces the development of HCC. (HEPATOLOGY 2013)

Hepatitis C virus (HCV) is one of the more common causes of chronic liver disease worldwide. Chronic hepatitis C is an insidiously progressive form of liver disease that relentlessly but silently progresses to cirrhosis in 20%-50% of cases over a period of 10-30 years.1,2 In addition, HCV is a major risk factor for hepatocellular carcinoma (HCC).3-7

On the other hand, the prevalence of patients with type 2 diabetes mellitus (T2DM) is increasing in many nations, including Japan.8 Thus, the management of T2DM patients who are chronically infected with HCV is one of the most important issues confronted by physicians. Few studies have reported relationships between T2DM and total malignancies, including HCC in HCV patients. In addition, it is not clear whether the stringent control of T2DM is necessary for protecting the development of malignancies in HCV patients. This issue needs to be confirmed via long-term follow-up of a large cohort of patients at high risk of developing malignancy.

With this background in mind, the present study was initiated to investigate the cumulative incidence and risk factors of malignancies, including HCC after prolonged follow-up in HCV patients treated with interferon (IFN) monotherapy or combination therapy of IFN and ribavirin. The strengths of the current study are the large numbers of patients included and the long-term follow-up of patients.

Abbreviations

CH, chronic hepatitis; CI, confidence interval; HbA1c, hemoglobin A1c; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HR, hazard ratio; IFN, interferon; LC, liver cirrhosis; SVR, sustained virological response; T2DM, type 2 diabetes mellitus; TAI, total alcohol intake.

Patients and Methods

Patients.

The number of patients who were diagnosed with chronic HCV infection and treated for the first time with IFN monotherapy or combination therapy between September 1990 and March 2009 in the Department of Hepatology, Toranomon Hospital, Tokyo, Japan, was 7,205. Of these, 4,302 patients met the following enrollment criteria: (1) no evidence of malignancies by physical examination, biochemical tests, abdominal ultrasonography, gastrofiberscope (or gastrography), or chest X-ray (or computed tomography); (2) features of chronic hepatitis or cirrhosis diagnosed via laparoscopy and/or liver biopsy within 1 year before the initiation of IFN therapy; (3) positivity for serum HCV-RNA before the initiation of IFN therapy; (4) period of ≥1 month to ≤1 year of IFN therapy; (5) negativity for hepatitis B surface antigens, antibody to hepatitis B core, or antimitochondrial antibodies in serum, as determined by radioimmunoassay, enzyme-linked immunosorbent assay, or indirect immunofluorescence assay; (6) age of ≥30 years to ≤80 years; (7) no underlying systemic disease, such as systemic lupus erythmatosus or rheumatic arthritis; and (8) repeated annual examinations during follow-up. Annual examinations included biochemical tests, tumor marker (carcinoembryonic antigen, alpha-fetoprotein, and prostate-specific antigen [only in men]), and abdominal ultrasonography. Patients with were excluded from the study if they had illnesses that could seriously reduce their life expectancy or if they had a history of carcinogenesis.

The primary outcome was the first development of malignancy. The development of malignancies was diagnosed by clinical symptoms, tumor marker, imaging (ultrasonography, computed tomography, or magnetic resonance imaging), and/or histological examination.9-15 All of the studies were performed retrospectively by collecting and analyzing data from the patient records. The physicians in charge explained the purpose, method, and side effects of IFN therapy to each patient and/or the patient's family. In addition, the physicians in charge received permission for the use of serum stores and future use of stored serum. Informed consent for IFN therapy and future use of stored serum was obtained from all patients. The study was approved by the Institutional Review Board of our hospital.

Medical Evaluation.

Body weight was measured in light clothing and without shoes to the nearest 0.1 kg. Height was measured to the nearest 0.1 cm. Height and weight were recorded at baseline, and body mass index was calculated as kg/m2. All patients were interviewed by physicians or nurse staff in the Toranomon Hospital using a questionnaire that gathered information on demographic characteristics, medical history, and heath-related habits, including questions on alcohol intake and smoking history.

The value for hemoglobin A1C (HbA1C) was estimated as a National Glycohemoglobin Standardization Program equivalent value (%). Patients were defined as having T2DM when they had a fasting plasma glucose level of ≥126 mg/dL and/or HbA1C level of ≥6.5%.16

Patients were regarded as hypertensive when systolic blood pressure was ≥140 mm Hg and/or diastolic blood pressure was ≥90 mm Hg for at least three visits. Smoking index (packs per day × year) and total alcohol intake (TAI) were evaluated by the sum of before, during, and after the IFN therapy.

Laboratory Investigation.

Diagnosis of HCV infection was based on detection of serum HCV antibody and positive RNA. Anti-HCV was detected using an enzyme-linked immunosorbent assay (ELISA II; Abbott Laboratories, North Chicago, IL). HCV genotype was examined via polymerase chain reaction assay, using a mixture of primers for the six subtypes known to exist in Japan, as reported.17 HCV-RNA was determined using the COBAS TaqMan HCV test (Roche Diagnostics, Basel, Switzerland). The serum samples stored at −80°C before IFN therapy were used. The linear dynamic range of the assay was 1.2-7.8 log IU/mL, and the undetectable samples were defined as negative. A sustained virological response (SVR) was defined as clearance of HCV-RNA using the COBAS TaqMan HCV test 6 months after the cessation of IFN therapy.

Evaluation of Liver Cirrhosis.

Status of liver was mainly determined on the basis of peritoneoscopy and/or liver biopsy. Liver biopsy specimens were obtained using a modified Vim Silverman needle with an internal diameter of 2 mm (Tohoku University style; Kakinuma Factory, Tokyo, Japan), fixed in 10% formalin, and stained with hematoxylin and eosin, Masson's trichrome, silver impregnation, and periodic acid-Schiff after diastase digestion. The size of specimens for examination was more than six portal areas.18

Follow-up.

The observation starting point was 6 months after the termination of IFN therapy. After that, patients were followed up at least twice a year in our hospital. Physical examination and biochemical tests were conducted at each examination together with a regular checkup. In addition, annual examinations during follow-up were undertaken. When a patient had complaints during follow-up, the physician in charge performed additional examinations based on symptoms. Four hundred eighteen patients were lost to follow-up. The final date of follow-up in 418 patients with loss of follow-up was regarded as the last consulting day. In addition, 881 patients were retreated with IFN. The final date of follow-up in 881 patients re-treated with IFN were regarded as the time of the initiation of IFN retreatment. Thus, 418 patients with loss of follow-up and 881 patients retreated with IFN were counted censored data in statistical analysis.19 The mean follow-up period was 6.8 (SD 4.3) years in 418 patients with loss of follow-up and 7.5 (SD 4.8) years in 881 patients retreated with IFN. Censored patients were counted in the analysis.

Statistical Analysis.

Clinical differences among three groups of patients with HCC with malignancies other than HCC without events were evaluated using the Kruskal-Wallis test. The cumulative development rates of malignancies were calculated using the Kaplan-Meier technique, and differences in the curves were tested using the log-rank test.20,21 Independent risk factors associated with malignancies were studied using the stepwise Cox regression analysis.22 The following variables were analyzed for potential covariates for incidence of primary outcome: (1) age, sex, T2DM, and hypertension at the initiation time of follow-up; (2) HCV genotype, HCV load, and hepatic fibrosis before IFN therapy; (3) average value of body mass index, aspartate aminotransferase, alanine aminotransferase, triglyceride, total cholesterol, and platelet count during follow-up; (4) sum value of smoking and alcohol before, during, and after the IFN therapy; and (5) efficacy of IFN therapy, combination of ribavirin, type of IFN, and total dose of IFN. A P < 0.05 was considered statistically significant. Data analysis was performed using SPSS 11.5 for Windows (SPSS, Chicago, IL).

Results

Patient Characteristics.

Table 1shows the baseline characteristics of the 4,302 enrolled patients at initiation of follow-up. The patients were divided into three groups: with HCC, with malignancies other than HCC, and without events. There were significant differences in several baseline characteristics among the three groups. The SVR rate was 34.4% (985/2,861) in IFN monotherapy and 63.5% (915/1,441) in combination therapy of IFN and ribavirin. Thus, the number of patients with SVR was 1,900. The mean follow-up was 8.1 (SD 5.0) years.

Table 1. Clinical Backgrounds at Initiation of Follow-up in Enrolled Patients
VariableTotalHCC GroupNon-HCC Malignancy GroupWithout Events GroupP
  • Data are presented as no. of patients or mean ± SD.

  • Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; F, female; FPG, fasting plasma glucose; GGT, gamma-glutamyl transferase; HDL, high-density lipoprotein; M, male; NGSP, National Glycohemoglobin Standardization Program.

  • *

    Smoking index is defined as packs per day × year. TAI and smoking index indicate the sum before and after first consultation.

  • Outbreak of IFN monotherapy: recombinant IFN-α2a, n = 220, recombinant IFN-α2b, n = 183, natural IFN-α, n = 1,678, natural IFN-α, n = 691, total dose of IFN = 560 ± 164 megaunit. Outbreak of pegylated IFN monotherapy: pegylated IFN-α2a, n = 89, total dose of pegylated IFN = 7.52 ± 2.24 mg.

  • Outbreak of combination therapy: recombinant IFN-α2b + ribavirin, n = 335, total dose of IFN = 508 ± 184 megaunit, total dose of ribavirin = 160 ± 68 g; natural IFN-β + ribavirin, n = 101, total dose of IFN = 502 ± 176 megaunit, total dose of ribavirin = 156 ± 67 g; pegylated IFN-α2b+ribavirin, n = 1,005 cases, total dose of pegylated IFN = 4.14 ± 1.10 mg, total dose of ribavirin = 206 ± 58 g.

No. of patients4,3023932133,696 
Age, years52.0 ± 11.855.8 ± 7.957.9 ± 9.151.3 ± 12.1<0.001
Sex, male/female2528/1774272/121129/842127/1569<0.001
Height, cm163.0 ± 9.2162.8 ± 8.3163.3 ± 9.1163.0 ± 9.30.772
Weight, kg61.4 ± 13.062.3 ± 10.660.8 ± 10.161.3 ± 13.40.142
BMI23.0 ± 4.023.4 ± 3.022.8 ± 2.823.0 ± 4.10.012
Blood pressure, mm Hg     
 Systolic128 ± 18132 ± 19133 ± 20127 ± 17<0.001
 Diastolic77 ± 1380 ± 1280 ± 1377 ± 13<0.001
TAI, kg*95 ± 92151 ± 101135 ± 8185 ± 89<0.001
Smoking index*6.4 ± 9.410.8 ± 11.112.5 ± 11.85.5 ± 8.7<0.001
AST, IU/L42 ± 4464 ± 5542 ± 3140 ± 42<0.001
ALT, IU/L44 ± 5372 ± 6343 ± 4342 ± 52<0.001
GGT, IU/L54 ± 6163 ± 6556 ± 4553 ± 380.007
Albumin, g/dL4.1 ± 0.34.1 ± 0.34.1 ± 0.24.1 ± 0.20.310
Triglyceride, mg/dL101 ± 53104 ± 54105 ± 50100 ± 520.329
Cholesterol, mg/dL170 ± 32165 ± 31169 ± 33171 ± 320.025
FPG, mg/dL100 ± 22110 ± 26104 ± 2298 ± 21<0.001
HbA1c, %, NSPG5.6 ± 1.25.9 ± 1.45.7 ± 1.45.5 ± 1.1<0.001
T2DM, +/−267/4,03563/33034/179170/3,526<0.001
Platelet count, ×104/mm317.1 ± 5.113.7 ± 4.916.5 ± 5.417.5 ± 5.4<0.001
Staging, LC/non-LC433/3,869113/28527/189293/3,395<0.001
HCV genotype, 1b/2a/2b/other2,721/995/458 /128283/52/20/38121/62/18/122,317/881/420/78<0.001
HCV RNA, log IU/mL6.06 ± 1.056.22 ± 0.526.05 ± 0.866.04 ± 1.050.003
IFN monotherapy/combination therapy2,861/1,441358/35175/382,328/1,368<0.001
Efficacy, SVR/non-SVR1,900/2,40244/34988/1251,768/1,928<0.001

Development and Breakdown of Malignancies.

As shown in Table 1, 606 of 4,302 patients developed malignancies: 393 developed HCC and 213 developed malignancies other than HCC. HCC accounted for 33.3% (44/132) of malignancies in patients with SVR and 73.6% (349/474) in patients without SVR. The breakdown of malignancies other than HCC was as follows: stomach cancer, n = 36; colon cancer, n = 35; lung cancer, n = 20; malignant lymphoma, n = 19; pancreatic cancer, n = 12; prostatic cancer, n = 16; breast cancer, n = 15; other cancers, n = 60.

Predictive Factors for the Development of HCC.

The cumulative development rate of HCC was 4.3% at 5 years, 10.5% at 10 years, 19.7% at 15 years, and 28.0% at 20 years (Fig. 1A). The factors associated with the development of HCC are shown in Table 2. Multivariate Cox proportional hazards analysis showed that HCC occurred when patients had liver cirrhosis (hazard ratio [HR], 5.01; 95% confidence interval [CI], 3.92-6.40; P < 0.001), non-SVR (HR, 4.93; 95% CI, 3.53-6.89; P < 0.001), age increments of 10 years (HR, 1.97; 95% CI, 1.71-2.28; P < 0.001), T2DM (HR, 1.73; 95% CI, 1.30-2.30; P < 0.001), male sex (HR, 1.67; 95% CI, 1.24-2.23; P = 0.001), and TAI of ≥ 200 kg (HR, 1.45; 95% CI, 1.11-1.88; P = 0.007). Fig. 1B-D and Fig. 2A-C show the cumulative development rates of HCC based on difference of IFN efficacy, age, hepatic fibrosis, TAI, sex, and T2DM. The 10-year cumulative rates of HCC after IFN therapy was determined to be 7.1% in 3,869 patients with chronic hepatitis and 37.7% in 433 patients with cirrhosis by using the Kaplan-Meier Method (Fig. 1D). Fig. 2D shows the development rates of HCC in T2DM patients according to difference of mean hemoglobin A1c (HbA1c) level during follow-up. HCC decreased when T2DM patients had a mean HbA1c level of <7.0% during follow-up (HR, 0.56; 95% CI, 0.33-0.89; P = 0.015). The development of HCC was reduced by 44% in T2DM patients with a mean HbA1c level of <7.0% compared with those with a mean HbA1c level of ≥7.0%.

Figure 1.

Cumulative development rate of HCC (A) in total HCV patients treated with IFN therapy and based on the difference of (B) efficacy, (C) age, and (D) hepatic fibrosis.

Figure 2.

Cumulative development rate of HCC based on the difference of (A) TAI, (B) sex, (C) diabetic state, and (D) mean HbA1c level during follow-up in T2DM patients.

Table 2. Predictive Factors for Development of HCC in Enrolled Patients
VariableUnivariate AnalysisCox Regression Analysis
HR (95% CI)PHR (95% CI)P
  • Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; GGT, gamma-glutamyl transferase; HDL, high-density lipoprotein.

  • *

    Smoking index is defined as packs per day × year. TAI and smoking index indicate the sum before and after first consultation.

Age, years (per 10)1.84 (1.64–2.06)<0.0011.97 (1.71–2.28)<0.001
Sex, male/female1.47 (1.18–1.83)<0.0011.67 (1.24–2.23)0.001
BMI, ≥22/<221.37 (1.12–1.66)0.002  
T2DM, +/−2.77 (2.13–3.60)<0.0011.73 (1.30–2.30)<0.001
Hypertension, +/−1.32 (1.02–1.71)0.036  
Smoking index, ≥20/<20*1.43 (1.14–1.79)0.002  
TAI, kg, ≥200/<200*2.13 (1.74–2.61)<0.0011.45 (1.11–1.88)0.007
AST, IU/L, ≥34/<343.00 (2.40–3.89)<0.001  
ALT, IU/L, ≥36/<362.74 (2.16–3.42)<0.001  
GGT, IU/L, ≥109/<1091.79 (1.19–2.46)0.039  
Albumin, g/dL, <3.9/≥3.91.92 (1.37–2.55)0.015  
Triglyceride, mg/dL, ≥100/<1001.14 (0.94–1.37)0.179  
Cholesterol, mg/dL, <150/≥150\1.38 (1.10–1.72)0.004  
Platelet count, ×104/mm3, <15/≥15)3.27 (2.56–4.17)<0.001  
Histological diagnosis, LC/non-LC7.09 (5.59–9.01)<0.0015.01 (3.92–6.40)<0.001
Combination of ribavirin, +/−0.66 (0.45–0.97)0.033  
Type of IFN, α/β1.10 (0.85–1.41)0.474  
Total dose of IFN, MU, ≥500/<5001.12 (0.91–1.38)0.291  
HCV genotype, ½1.67 (1.30–2.14)<0.001  
HCV-RNA, log IU/mL, ≥5/<51.02 (0.98–1.05)0.315  
Efficacy, non-SVR/SVR4.78 (3.47–6.59)<0.0014.93 (3.53–6.89)<0.001

Table 3 shows the development rate of HCC and risk factors in four groups classified by the difference of hepatic fibrosis and efficacy of IFN therapy. The development rate of HCC per 1,000 person years was 1.55 in patients with chronic hepatitis (CH) at baseline and SVR (CH+SVR), 18.23 in patients with liver cirrhosis (LC) at baseline and SVR (LC+SVR), 13.53 in patients with chronic hepatitis at baseline and non-SVR (CH+non-SVR), and 50.43 in patients with LC at baseline and non-SVR (LC+non-SVR). The risk of HCC development in the CH+SVR group was advanced age, male sex, TAI of ≥200 kg, and T2DM. T2DM enhanced the development of HCC with statistical significance in three groups of CH+SVR, CH+non-SVR, and LC+non-SVR.

Table 3. Development Rate of HCC Based on Hepatic Fibrosis and Efficacy of IFN Therapy
VariableCH + SVRLC + SVRCH + Non-SVRLC + Non-SVR
  • Abbreviations: CH + Non-SVR, patients with CH at baseline and non-SVR 6 months after IFN therapy; CH + SVR, patients with CH at baseline and SVR 6 months after IFN therapy; LC + Non-SVR, patients with LC at baseline and non-SVR 6 months after IFN therapy; LC + SVR, patients with LC at baseline and SVR 6 months after IFN therapy.

  • *

    Hazard ratio (95% confidence interval) and P value by Cox proportional hazards analysis.

No. of patients1,7511492,118284
Age, years51.7 ± 12.156.9 ± 9.851.5 ± 11.757.2 ± 9.9
Sex, male/female1,082/66991/581,190/928165/119
HbA1c (%, NSPG)5.5 ± 0.75.8 ± 0.85.7 ± 0.76.1 ± 0.8
TAI, kg86 ± 91104 ± 9997 ± 90129 ± 102
Patients with T2DM741313347
Patients with HCC2222233116
1,000 person years of HCC1.5518.2313.5350.43
Age, years (per 10)*2.60 (1.48–4.58)1.83 (0.95–3.55)2.07 (1.75–2.46)1.09 (0.87–1.37)
 P value0.0010.070<0.0010.477
Sex, male/female*3.42 (1.01–11.63)3.41 (1.00–11.63)1.34 (0.99–1.81)1.93 (1.25–3.00)
 P value0.0490.0500.0580.003
TAI, kg, ≥200/<200*2.68 (1.14–6.34)3.84 (1.83–9.85)2.21 (1.65–2.95)1.54 (1.03–2.31)
 P value0.0240.004<0.0010.038
T2DM, +/−*4.76 (1.60–14.10)2.48 (0.57–10.86)2.53 (1.76–3.65)1.87 (1.16–3.01)
 P value0.0050.228<0.0010.010

Predictive Factors for Development of Malignancies Other than HCC.

The cumulative development rate of malignancies other than HCC was 2.4% at 5 years, 5.1% at 10 years, 9.8% at 15 years, and 18.0% at 20 years (Fig. 3A). The factors associated with the development of malignancies other than HCC are shown in Table 4. Malignancies other than HCC occurred when patients had age increments of 10 years (HR, 2.19; 95% CI, 1.84-2.62; P < 0.001), smoking index of ≥20 (HR, 1.89; 95% CI, 1.41-2.53; P < 0.001), and T2DM (HR, 1.70; 95% CI, 1.14-2.53; P = 0.008). Fig. 3B-D shows the cumulative development rates of malignancies other than HCC based on difference of age, smoking index, and T2DM. Fig. 3E shows the risk of malignancies other than HCC in T2DM patients according to mean HbA1c level during follow-up. The HR of HCC development in patients with mean HbA1c level of <7.0% versus those with mean HbA1c level of ≥7.0% was 0.62 (95% CI, 0.31-1.23; P = 0.170). There was no significant difference in development of malignancies other than HCC based on the difference of mean HbA1c level during follow-up. Table 5 shows the impact based on three factors of age, smoking index, and T2DM for the development of each malignancy other than HCC by using Cox regression analysis. Aging enhanced carcinogenesis of stomach, colon, lung, prostate, breast, and pancreas with statistical significance. Smoking enhanced lung cancer and colorectal cancer with statistical significance. In addition, T2DM enhanced the pancreatic cancer with statistical significance and tended to enhance the gastric cancer.

Figure 3.

Cumulative development rate of malignancies other than HCC (A) in total HCV patients treated with IFN therapy and based on the difference of (B) age, (C) smoking index, (D) diabetic state, and (E) mean HbA1c level during follow-up in T2DM patients.

Table 4. Predictive Factors for Development of Malignancies Other than HCC
VariablesUnivariate AnalysisCox-Regression Analysis
HR (95% CI)PHR (95% CI)P
  • Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BMI, body mass index; GGT, gamma-glutamyl transferase.

  • *

    Smoking index is defined as packs per day × year. TAI and smoking index indicate the sum before and after first consultation.

Age, years (per 10)2.23 (1.88–2.65)< 0.0012.19 (1.84–2.62)<0.001
Sex, male/female1.06 (0.79–1.40)0.759  
BMI, ≥22/<220.97 (0.75–1.24)0.767  
T2DM, +/−2.56 (1.76–3.72)<0.0011.70 (1.14–2.53)0.008
Hypertension, +/−2.33 (1.70–3.18)<0.001  
Smoking index, ≥20/<20*2.74 (2.06–3.65)<0.0011.89 (1.41–2.53)<0.001
TAI, kg, ≥200/<200*1.77 (1.33–2.37)<0.001  
AST, IU/L, ≥34/<340.89 (0.65–1.20)0.412  
ALT, IU/L, ≥36/<360.98 (0.72–1.34)0.891  
GGT, IU/L, ≥109/<1091.26 (0.79–2.01)0.350  
Albumin, g/dL, <3.9/≥3.91.41 (0.90–2.04)0.145  
Triglyceride, mg/dL, ≥100/<1001.28 (1.03–1.60)0.030  
Total cholesterol, mg/dL, <150/≥1501.10 (0.82–1.46)0.548  
Platelet count, × 104/mm3, <15/≥151.39 (1.02–1.91)0.038  
Histological diagnosis, LC/non-LC1.77 (1.13–2.75)0.012  
Combination of ribavirin, +/−0.66 (0.44–0.97)0.034  
Type of IFN, α/β1.05 (0.75–1.47)0.789  
Total dose of IFN, MU, ≥500/<5001.31 (0.96–1.77)0.084  
HCV genotype, ½1.30 (0.80–2.93)0.432  
HCV RNA, log IU/mL, ≥5/<50.89 (0.50–1.23)0.612  
Efficacy, non-SVR/SVR0.85 (0.64–1.12)0.232  
Table 5. Impact Based on Age, Smoking Index, and Diabetes for Development of Malignancies Other than HCC
MalignancyAge, Years (per 10)Smoking Index, ≥20/<20Diabetes, +/−
HR (95% CI)PHR (95% CI)PHR (95% CI)P
Gastric cancer (n = 36)2.48 (1.62–3.78)<0.0011.69 (0.83–3.43)0.1462.29 (0.95–5.52)0.065
Colorectal cancer (n = 35)1.91 (1.28–2.86)0.0022.27 (1.13–4.58)0.0221.78 (0.68–4.66)0.240
Lung cancer (n = 20)2.33 (1.35–4.01)0.0022.90 (1.25–6.74)0.0131.53 (0.45–5.24)0.496
Prostatic cancer (n = 16)2.84 (1.32–6.13)0.0081.89 (0.88–3.15)0.2660.71 (0.09–5.47)0.735
Breast cancer (n = 15)2.86 (1.30–6.29)0.0091.29 (0.17–10.19)0.8081.20 (0.16–9.39)0.859
Malignant lymphoma (n = 19)2.21 (1.26–3.88)0.0061.25 (0.44–3.56)0.6711.39 (0.32–6.12)0.663
Pancreatic cancer (n = 12)3.32 (1.44–7.65)0.0051.41 (0.45–4.82)0.5783.75 (1.02–13.88)0.046

Discussion

This study describes the development incidence of HCC or malignancies other than HCC after the termination of IFN therapy in HCV patients. Patients at Toranomon Hospital comprised mainly government employees, office workers, and business persons. Most patients were regularly recommended to undergo annual multiphasic health screening examinations. In the present study, patients who had undergone annual multiphasic health screening examinations were enrolled. The strengths of the present study are a prolonged follow-up in the large numbers of patients included.

The present study shows several findings with regard to the development incidence and predictive factors for total malignancies after IFN therapy for HCV patients. First, the 10-year cumulative rates of HCC after IFN therapy was determined to be 7.1% in 3,869 patients with chronic hepatitis and 37.7% in 433 patients with cirrhosis using the Kaplan-Meier method. Our previous studies showed via retrospective analysis that the 10-year cumulative rates of HCC were 12.4% for 456 patients with chronic hepatitis and 53.2% for 349 patients with cirrhosis.7,23 Although patient selection bias for IFN treatment versus no treatment had been noted in the previous studies, the results suggest the possibility that IFN therapy reduces the development of HCC in HCV patients. Several historical data in Japan suggest that IFN therapy reduces the development of HCC in HCV patients.24-26

Second, HCC occurred with statistical significance when the following characteristics were present: non-SVR, advanced age, cirrhosis, TAI of ≥200 kg, male sex, and T2DM. T2DM caused a 1.73-fold enhancement in HCC development. Several authors have reported an increased risk of HCC among patients with the following characteristics: non-SVR, cirrhosis, male sex, advanced age, and T2DM.24-28 Our results show that physicians in charge of aged male patients with non-SVR, advanced fibrosis, TAI of ≥200 kg, and T2DM should pay attention to the development of HCC after IFN therapy. In addition, maintaining a mean HbA1c level of <7.0% during follow-up reduced the development of HCC. This result indicates that stringent control of T2DM is important for protecting the development of HCC.

Third, the development rate of HCC per 1,000 person years was about 1.55 in 1,751 patients with chronic hepatitis at baseline and SVR. In these patients, the risk factors associated with HCC were advanced age, male sex, TAI, and T2DM. We compared the HCC development rate in patients with chronic hepatitis at baseline and SVR to the general population. A total of 5,253 individuals without HCV antibody and hepatitis B surface antigen, who underwent annual multiphasic health screening examinations in our hospital were evaluated as controls. Individuals with either of the following criteria were excluded: (1) illness that could seriously reduce their life expectancy or (2) history of carcinogenesis. They were selected by matching 3:1 with patients who had chronic hepatitis at baseline and SVR for age, sex, T2DM, and follow-up periods. In control individuals, the mean age was 51.7 years; the prevalence (number) of male patients was 61.8% (3,246); the prevalence (number) of T2DM patients was 4.2% (222); the mean follow-up period was 8.0 years. The number of development of HCC in control individuals was only five. This result suggests that the development rate of HCC in patients with chronic hepatitis at baseline and SVR is higher than that in the general population.

Fourth, HCC accounted for 33.3% in SVR patients and 73.6% in non-SVR patients. According to Matsuda et al.,29 the outbreak of malignancies in the Japanese male population was observed in the following order in 2005: gastric cancer 20.4% > colon cancer 16.0% > lung cancer 15.4% > prostate cancer 10.9% > HCC 7.4%. On the other hand, the outbreak of malignancies in the Japanese female population was observed in the following order in 2005: breast cancer 18.0% > colon cancer 16.2% > gastric cancer 13.6% > lung cancer 9.3% > uterine cancer 6.8%. Our results show that HCC is the most common cause of malignancy, not only in the non-SVR group but also in the SVR group.

Finally, malignancies other than HCC occurred with statistical significance when patients were of advanced age, were smokers, and had T2DM. Our result indicates that smoking enhances lung cancer and colorectal cancer. Many authors have reported that smoking is a direct cause of cancers of the oral cavity, esophagus, stomach, pancreas, larynx, lung, bladder, kidney, and colon.30,31 In addition, the present study indicates that T2DM enhances pancreatic cancer with statistical significance and tends to enhance gastric cancer. T2DM showed up to about 1.7-fold increase in development of malignancies other than HCC. A recent meta-analysis of cohort studies have revealed that diabetic patients increase risk of pancreatic cancer, HCC, bladder cancer, non-Hodgkin's lymphoma, colorectal cancer, and breast cancer.32-39

Although the role of T2DM in carcinogenesis remains speculative, the following possible mechanisms have been reported: (1) hyperglycemia increases malignancy risk via increasing oxidative stress and/or activating the rennin-angiotensin system40; (2) insulin resistance increases malignancy risk via down-regulation of serine/threonine kinase II to adenosine monophosphate–activated protein kinase pathway41; (3) reduced insulin secretion increases malignancy risk via down-regulation of sterol regulatory element-binding protein-1c with consequent up-regulation of insulin-like growth factor.42

T2DM is increasing dramatically worldwide over the past decades.8 It is estimated that about 7 million people are affected by diabetes mellitus in Japan. Approximately 8%-10% of adults in Japan have T2DM. The risk factors associated with T2DM include family history, age, sex, obesity, smoking, physical activity, and HCV.43-46 In the near future, T2DM will be increasing in HCV-positive patients.

This study is limited in that it was a retrospective cohort trial. Another limitation is that patients were treated with different types of antivirus therapy for different durations. In addition, T2DM patients were treated with different types of drugs during follow-up. Finally, our cohort contains Japanese subjects only. On the other hand, the strengths of the present study are a long-term follow-up in the large numbers of patients included.

In conclusion, T2DM causes an approximately 1.7-fold enhancement in the development of HCC and malignancies other than HCC after IFN therapy. Additionally, in T2DM patients, maintaining a mean HbA1c level of <7.0% during follow-up reduced the development of HCC.

Acknowledgements

We thanks Thomas Hughes for editorial assistance.

Ancillary

Advertisement