Who Is More Likely to Respond to Dual Treatment With Pegylated-Interferon and Ribavirin for Chronic Hepatitis C?

A Gender-Oriented Analysis

V. Di Marco; L. Covolo; V. Calvaruso; M. Levrero; M. Puoti; F. Suter; G. B. Gaeta; C. Ferrari; G. Raimondo; G. Fattovich; T. Santantonio; A. Alberti; R. Bruno; C. Mussini; M. Mondelli; F. Donato; A. Craxì

Disclosures

J Viral Hepat. 2013;20(11):790-800. 

In This Article

Results

Baseline Features

Between March 2007 and September 2009, twelve centres enrolled a total of 670 patients (mean patients per centre 46; range, 15–115). Three hundred and sixty-two patients (54%) were infected by genotype 1, 175 patients (26%) by genotype 2, 100 patients (15%) by genotype 3 and 33 patients (5%) by genotype 4. Baseline serum HCV-RNA was measured in 640 patients (95.5%), and 240 of them (37.5%) had levels lower than 400 000 UI/mL. Six hundred and twenty-seven patients (93.7%) were tested for rs8099860 SNP genotypes, and 212 of them (33.8%) carried a C/C genotype. Liver biopsy was performed in 254 (70%) of the 362 genotype 1 patients, in 59 (33.7%) of the 175 genotype 2 patients, in 33 (33%) of the 100 genotype 3 patients and in 18 (54.5%) of the 33 genotype 4 patients.

Adherence to Treatment Schedules

Seventy-five patients (11.2%) stopped treatment because of adverse events. During the first 12 weeks of treatment, 46 patients (6.8%) stopped treatment. Proportion of withdrawal was similar in all HCV genotypes (6.9%, 5.1%, 7% and 12.2% in genotype 1, 2, 3 and 4, respectively). Seventeen genotype 1 patients (4.7%) were withdrawn from treatment between 12 and 24 weeks, and 11 patients (3%) between 24 and 36 weeks. No decompensation of liver disease, or death from other causes, occurred during the treatment.

Effectiveness of Treatment

An intention-to-treat analysis was performed. The rates of SVR were 43.6%, 81.1%, 69% and 39.4% in genotype 1, 2, 3 and 4 patients, respectively. Table 1 shows factors associated with SVR in all the 670 patients at a P-value of 0.10 or less. On multivariate analysis, the demographic factors independently associated with SVR were female gender (OR 1.54) and age (OR 0.937); among the virologic factors, infection with genotype 2 (OR 6.24) or genotype 3 (OR 2.18), serum HCV-RNA levels less than 400 000 IU/mL (OR 1.585); and among the genetic factors, the C/C genotype of the rs12979860 SNP of the IL28B locus (OR 3.063). Analysis of baseline characteristics according to gender showed that the males had a body mass index, ALT and GGT values, levels of serum insulin and HOMA-R score significantly higher than the females. In contrast, the females were older, more frequently had visceral obesity, and had significantly higher serum levels of cholesterol and HDL cholesterol than the males. Regarding the virologic features, a higher proportion of females had serum HCV-RNA level lower than 400 000 IU/mL with respect to males, and there was a significantly higher proportion of males with genotype 3. We found no significant differences in the prevalence of C/C genotype of the rs8099860 SNP between the two gender groups (Table 2).

From the differences found in clinical and virologic factors between males and females, we did separate analyses by gender and virus genotype. We excluded from the analysis genotype 4 patients because of their small number.

Genotype 1 Patients

An SVR was observed in 75 of the 163 females (46%), and in 83 of the 199 males (41.7%), with genotype 1 infection. In the females, age < 50 years (OR 2.12) and C/C genotype of the rs12979860 SNP (OR 2.83) were associated with SVR on multivariate analysis (Table 3). When RVR was included in the model as an independent predictor, similar results were obtained, and an independent association was also found between RVR and SVR (OR 6.21; 95% CI 2.31–16.74; P < 0.001).

In the 199 males infected by genotype 1, absence of visceral obesity (OR 2.491), serum HCV-RNA levels lower than 400 000 IU/mL (OR 2.660) and C/C genotype of rs12979860 SNP (OR 4.969) were independently associated with SVR on multivariate analysis (Table 4). When RVR was included in the model as an independent variable, similar results were obtained and an independent association was found between RVR and SVR (OR 7.028; 95% CI 2.900–17.032; P < 0.001).

Combining favourable baseline variables, the probability of achieving an RVR ranged from 15.5% to 73.6% for females and from 9.5% to 64.3% for males, and the probability of achieving an SVR ranged from 27.6% to 84.2% for females and from 14.36% to 85.7% for males (Table 5) The presence of two favourable factors approximately doubled the likelihood of achieving an SVR in both males and females, and the presence of three factors approximately tripled the likelihood of achieving an SVR in males.

The predictive value of one or more favourable factors for SVR, with respect to having an RVR, is shown in Table 5 and Fig 1, for males and females, respectively.

Figure 1.

Rate of sustained virologic response in genotype 1 Females in relation to favourable factors at baseline and rapid virologic response. Favourable factors at baseline: Age < 50 years and C/C genotype of rs12979860 single-nucleotide polymorphisms (SNP). RVR, rapid virologic response; SVR, sustained virologic response; pts, patients.

In females who experienced an RVR, fewer than 60% of those without favourable factors, but over 85% of those with one or two favourable factors, achieved an SVR (Fig. 1). SVR was observed in fewer than 60% of the male patients, with or without only one favourable factor, who experienced an RVR, and it increased to over 88% in those with 2 or 3 favourable factors who experienced an RVR (Fig. 2). The ability of the rule generated by the model to predict SVR in males was rather high on ROC analysis applied to the same data set (AUC 0.742). The ability of the model to predict SVR in females was slightly lower (AUC 0.70). In the model, the likelihood of SVR gradually decreased from the 'worst' to the 'best' class for each variable. It is noteworthy that only 3 of 77 SVR males (3.89%) were in the class without favourable factors (false negative) and that only 2 of 102 non-SVR males (1.96%) were in the class with three favourable factors (false positive). On the other hand, 16 of 68 female (23.52%) with SVR were in the class without favourable factors (false negative), and 3 of 84 females (3.57%) without SVR were in the class with two favourable factors (false positive).

Figure 2.

Rate of sustained virologic response in genotype 1 Males in relation to favourable factors at baseline and rapid virologic response. Favourable factors at baseline: absence of visceral obesity, C/C genotype of rs12979860 single-nucleotide polymorphisms (SNP) and hepatitis C virus-RNA (HCV-RNA) levels < 400 000 UI/mL. RVR, rapid virologic response; SVR, sustained virologic response; pts, patients.

Genotype 2 Patients

There were 175 genotype 2 patients in the cohort, and an SVR was observed in 79 of the 93 females, and in 63 of the 83 males (84.9% vs 76.8%, P = 0.18). In 92 females, SVR was associated with higher serum ALT values at baseline (P = 0.001), but if RVR was included in the model, it remained the only variable independently associated with SVR (OR 10.987; 95% CI 1.722–70.126; P = 0.011). In 82 males, no baseline variables were associated with SVR, but the SVR was significantly higher in patients who achieved an RVR than in patients without RVR (78% vs 44%, P = 0.045).

Genotype 3 Patients

Among the 100 genotype 3 patients, SVR was achieved in 27 of the 31 females, and in 42 of the 69 males (87.1% vs 60.1%, P = 0.010). In the 31 females, no baseline variable was independently associated with SVR, but the SVR was significantly higher in patients who achieved an RVR than in patients without RVR (100% vs 55%, P = 0.040). In the 69 males, higher values of baseline HDL cholesterol and marginally lower values of baseline GGT were associated with SVR on multivariate analysis. SVR was also significantly higher in patients who achieved an RVR than in patients without RVR (74.5% vs 29.5%, P = 0.0026). When RVR was included in the model as an independent variable, RVR remained the only variable independently associated with SVR (OR 15.360; 95% CI 1.001–257.420; P = 0.047).

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....