| |
Early Menopause is Associated with Lack of Response to Antiviral Therapy in Women with Chronic Hepatitis C: "This suggests that CHC in women should be treated early, disregarding the fact that liver disease is milder in women of reproductive age, as this condition will last only as long as the estrogen-exposed period." - published pdf attached
|
| |
| |
Download the PDF here
Article in Press
Gastroenterology Dec 2010 Ahead of print
"In this prospective study of 1000 consecutive patients with CHC, we found that menopause is independently associated with the severity of liver damage and with a remarkably lower likelihood of achieving SVR (46% vs 67%)......Recent studies suggest that post-menopausal women have accelerated progression of fibrosis,6,7 which is prevented by long-term estrogen exposure from HRT.7 In this large cohort of female HCV patients, we confirmed that liver inflammation and higher levels of GGT (a known surrogate marker of metabolic alterations and of TNF-α up-regulation)21, are independently associated with severe fibrosis. We also identified the length of estrogen deprivation as a strong independent risk factor for fibrosis: the longer the menopausal period, the higher the risk of severe fibrosis. It was 5-fold higher in women who had been menopausal for more than 10 years in comparison with early menopausal women.....
In conclusion, our data from a large cohort of European women with CHC show that menopause is associated with profound changes in TNF-α and IL-6 levels. These changes are greater than those resulting from HCV infection in women of reproductive age and result in a more pronounced inflammatory state, more rapid progression to fibrosis and a hitherto unrecognized resistance to antiviral therapy. This suggests that CHC in women should be treated early, disregarding the fact that liver disease is milder in women of reproductive age, as this condition will last only as long as the estrogen-exposed period. Alternative strategies should be tested for HCV-positive women presenting in early post-menopause. We are currently examining the combination of HRT with PEG IFN/Ribavirin therapy in a controlled randomized trial (EudraCT 2008-001260-36).
"Overall, SVR was achieved by 231 of 442 women (52.2%). SVR occurred in 121 of 263 menopausal women (46.0%) vs. 110 of 163 women of reproductive age (67.5%; OR 2.436, 95% CI 1.620-3.662; P < .0001). The rate of SVR was similar among women of reproductive age who were stratified....By multivariate analysis significant independent baseline predictors of SVR failure in women were the presence of menopause (OR 1.802; 95% CI 1.154-2.813, P = .01), cholesterol level (OR 0.997; 95% CI 0.943-0.991, P = .008), high GGT levels (OR 2.165; 95% CI 1.364- 3.436, P = .001) and genotype 1/4 (OR 3.861, 95% CI 2.433-6.134) P = .006)(Table 4, top).....These data support the hypothesis that menopause determines a switch to a systemic and hepatic pro-inflammatory state25 in which increased IL-6 and TNF-α production contribute to the observed resistance to IFN-based therapies. Women entering menopause would rapidly go from an estrogen-protected environment, where HCV-mediated inflammation is limited, to an estrogen-deprived one in which inflammation becomes less controlled and resistance to antiviral therapy increases remarkably, thus increasing the risk of developing severe fibrosis. Regardless of reproductive status, HCV-positive women had significantly higher serum levels of both TNF-α and IL-6 compared to controls (P < .0001 for all combinations) (Figure 1A). IL-6 levels in women of reproductive age were significantly lower than those in early post-menopausal women [2.6µ1.5 vs 8.9µ10.3 pg/ml; P = .007] and were lower, although to a lesser degree, than those in late post-menopausal women [2.6µ1.5 vs 4.8µ8,2 pg/ml; P = .017]."
Erica Villa1, Aimilia Karampatou1, Calogero Camma2, Alfredo Di Leo3, Monica Luongo1, Anna Ferrari1, Salvatore Petta2, Luisa Losi1, Gloria Taliani4, Paolo Trande5, Barbara Lei1, Amalia Graziosi1, Veronica Bernabucci1, Rosina Critelli1, Paola Pazienza3, Maria Rendina3, Alessandro Antonelli6, Antonio Francavilla3
1. Dept of Gastroenterology, University of Modena and Reggio Emilia, Modena, Italy 2. Dept of Gastroenterology, University of Palermo, Palermo, Italy 3. Dept of Gastroenterology, University of Bari, Bari, Italy 4. Dept of Infectious Diseases, University of Rome "La Sapienza", Italy
5. Dept of Gastroenterology, AUSL, Modena, Italy 6. Department of Internal Medicine, University of Pisa, Italy
Received 20 August 2010; received in revised form 26 October 2010; accepted 2 December 2010. published online 17 December 2010.
Accepted Manuscript
Abstract
Background & Aims
Chronic hepatitis c (CHC) and liver fibrosis progress more rapidly in men and menopausal women than in women of reproductive age. We investigated the associations among menopause, sustained virological response (SVR), and liver damage in patients with CHC.
Methods
We performed a prospective study of 1000 consecutive, treatment-naive patients ≥18 years old with compensated liver disease from CHC. Liver biopsy samples were analyzed (for fibrosis, inflammation, and steatosis) before patients received standard antiviral therapy. From women (n = 442), we collected data on the presence, type, and timing of menopause; associated hormone and metabolic features; serum levels of interleukin-6 (IL-6); and hepatic tumor necrosis factor-α (TNF-α).
Results
Post-menopausal women achieved SVRs less frequently than women of reproductive age (46.0% vs. 67.5%, P<.0001) but as frequently as men (51.1%, P=.178). By multivariate regression analysis, independent significant predictors for women to not achieve an SVR were early menopause (odds ratio [OR]=8.055; 95% confidence interval [CI], 1.834-25.350), levels of γ-glutamyl transpeptidase (OR=2.165, 95% CI, 1.3643-.436), infection with hepatitis C virus (HCV) genotypes 1 or 4 (OR=3.861; 95% CI, 2.433-6.134), and cholesterol levels (OR=0.967; 95% CI 0.943-0.991). Early menopause was the only independent factor that predicted lack of an SVR among women with genotype 1 HCV infection (OR=3.933; 95% CI, 1.274-12.142). Baseline levels of liver inflammation, fibrosis, steatosis, serum IL-6 (P=.04), and hepatic TNF-α (P=.007) were significantly higher among post-menopausal than women of reproductive age.
Conclusions
Among women with CHC, early menopause was associated with a low likelihood of SVR, probably because of inflammatory factors that change at menopause.
"cirrhosis was present in 1.7% of non-menopausal women, 6.0% of early- and 11.0% of late-menopausal women (non menopausal vs late menopausal: P = .0014). Multivariate analysis showed that, in addition to necroinflammatory activity (OR 1.464, 95% CI 1.256-1.707, P < .001), low platelet levels (OR 0.976, 95% CI 0.966-0.986, P < .001) and elevated GGT (OR 1.010, 95% CI 1.003-1.018, P = .008), also a longer duration of estrogen deprivation (5 to 10 years: OR 4.078 (95% CI 1.013-16.409), P = .048; more than 10 years: OR 4.867 (95% CI
1.476-16.042), P = .009) was independently linked with severe fibrosis (Table 2)."
Introduction
Appraisal of the clinical course of chronic hepatitis C (CHC) has revealed several striking differences between men and women. The progression of fibrosis is more than twice as fast in men,1,2 even when potential confounding factors like age, duration of infection, or metabolic features are accounted for by multivariate analysis.1,3 There are conflicting data about SVR rates in women; some studies report that response rates are not significantly different from those of men,4 whereas others have identified female sex as an independent factor associated with SVR.5 This inconsistency might arise from the fact that the female cohorts have always been evaluated as whole, without taking into account differences in responses due to hormonal state.
The reduced rate of fibrosis among women disappears after menopause; in fact, post- menopausal women have accelerated progression of fibrosis,6,7 compared with men, which is slowed by long-term estrogen exposure with hormone replacement therapy (HRT).7 Post- menopausal women are also at higher risk of developing hepatocellular carcinoma (HCC); there is a more balanced ratio of men:women in later life that results from the higher incidence of HCC in older women.8,9
Antiviral therapy greatly improves the natural course of CHC when it results in a sustained virological response (SVR),10 even in patients with established cirrhosis.11 Negative predictive factors for SVR include older age,12-14 which might coincide with menopause. However, no studies have evaluated the impact of menopause itself on response to antiviral therapy.
We therefore evaluated the impact of menopause on SVR and on histological features in a prospective study. We collected data on the type and timing of menopause, parity, and the use and duration of hormone replacement therapy (HRT) in a cohort of 1000 patients with CHC undergoing standard antiviral therapy with pegylated interferon (PEG-IFN) and ribavirin (RBV).
Discussion
In this prospective study of 1000 consecutive patients with CHC, we found that menopause is independently associated with the severity of liver damage and with a remarkably lower likelihood of achieving SVR.
Various lines of evidence support a link between menopause and the severity of fibrosis in CHC patients. Recent studies suggest that post-menopausal women have accelerated progression of fibrosis,6,7 which is prevented by long-term estrogen exposure from HRT.7 In this large cohort of female HCV patients, we confirmed that liver inflammation and higher levels of GGT (a known surrogate marker of metabolic alterations and of TNF-α up-regulation)21, are independently associated with severe fibrosis. We also identified the length of estrogen deprivation as a strong independent risk factor for fibrosis: the longer the menopausal period, the higher the risk of severe fibrosis. It was 5-fold higher in women who had been menopausal for more than 10 years in comparison with early menopausal women.
We found menopause to be significantly correlated with necro-inflammation, steatosis and metabolic alterations (high cholesterol and glucose). Accordingly, a recent clinical study showed a higher prevalence of steatosis in post-menopausal CHC patients who were more than 55 years old.13 Experimental studies are in line with our clinical data on the association between menopause and liver damage, showing that menopause is associated with a pro-inflammatory state that can drive fibrosis progression and lead to HCC.8 In addition, some experimental data also indicate a potential role of estrogen insufficiency in steatogenesis.22,23
Menopause coincides with older age, which has been identified as a negative predictive factor for a SVR.13,14 However, the influence of menopause itself on response to antiviral therapy has not been investigated. We examined age and menopause in multivariate analysis and found that age was not independently correlated with SVR, whereas menopause was. Our findings indicate that menopause is associated with a remarkable and unrecognized resistance to antiviral therapy, especially in carriers of genotype 1 HCV. Moreover, our data indicate that early stages of menopause (estrogen deprivation for less than 5 years) correlate with failure of antiviral therapy.
This is particularly evident in women with genotype 1 HCV, who have less resistance to antiviral therapy with the passage of years after the onset of menopause. A possible explanation for this finding is that hepatic levels of TNF-α and circulating IL-6 are upregulated at the time of menopause. Another, unlikely, explanation is that occasional consumption of NSAIDs or unreported use of alcohol, causes levels of both cytokines to decrease after menopause, and hepatic expression of TNF-α levels become non-significantly different from those of women of reproductive age. (Figure 1C, panel a). A similar phenomenon has been described regarding bone loss and levels of IL-6, which affect bone loss less with time after menopause, becoming insignificant after 10 years.24 It is interesting that in men, who had levels of cytokines that were constant yet greater than those of women at all age groups tested, the OR of not achieving an SVR was not significantly different among those younger than 45 years or older than 55 (i.e. in 2 age groups similar to those of women of reproductive age or menopausal females).
Although our study was not designed to clarify the pathogenesis of this association, the results do suggest several hypotheses. We focused our attention on levels of TNF-α and IL-6; these cytokines undergo large changes during in menopause25 and in HCV infection, during which their levels are greatly up-regulated,26-29 and they are able to interfere with antiviral response. Our data show that the occurrence of menopause is associated with a large additional increase in circulating IL-6, a striking increase in hepatic TNF-α, and an increase in the expression of SOCS3 in the liver. While there are no reports linking serum IL-6 levels with resistance to IFN, TNF-α has been implicated as an independent factor associated with response to IFN,21, 29-30 and hepatic SOCS3 is reportedly the strongest factor influencing the outcome of interferon-based antiviral therapy,31-34 especially in patients with genotype 1.33 SOCS3 is induced by HCV core protein,35 and also by IL- 6 and TNF-α.31,32 These data support the hypothesis that menopause determines a switch to a systemic and hepatic pro-inflammatory state25 in which increased IL-6 and TNF-α production contribute to the observed resistance to IFN-based therapies. Women entering menopause would rapidly go from an estrogen-protected environment, where HCV-mediated inflammation is limited, to an estrogen-deprived one in which inflammation becomes less controlled and resistance to antiviral therapy increases remarkably, thus increasing the risk of developing severe fibrosis.
We have also confirmed that genotype 1 and higher GGT levels are independent negative predictors of SVR, both for the entire cohort and when women are considered as a group. Although hidden alcohol or drug abuse cannot be excluded in principle to explain GGT alteration, the bias, if present, was equally shared by all groups, the methodology of ascertainment being the same for all. More relevant, in this context, is the demonstrated relationship between GGTs levels and hepatic TNF-α mRNA, which in turn strongly related with non response in HCV patients21.
The study has some limitations. In particular we cannot rule-out that lack of data on vitamin D serum levels, a recently recognized factor influencing SVR achievement 36, could affect the interpretation of our results, as well as the absence of data about other adipokines/cytokines.
In conclusion, our data from a large cohort of European women with CHC show that menopause is associated with profound changes in TNF-α and IL-6 levels. These changes are greater than those resulting from HCV infection in women of reproductive age and result in a more pronounced inflammatory state, more rapid progression to fibrosis and a hitherto unrecognized resistance to antiviral therapy. This suggests that CHC in women should be treated early, disregarding the fact that liver disease is milder in women of reproductive age, as this condition will last only as long as the estrogen-exposed period. Alternative strategies should be tested for HCV-positive women presenting in early post-menopause. We are currently examining the combination of HRT with PEG IFN/Ribavirin therapy in a controlled randomized trial (EudraCT 2008-001260-36).
Results
Baseline features of all 1.000 patients according to sex are shown in supplementary table 1. Ninety- nine patients (9.9%) had a diagnosis of cirrhosis: cirrhosis was more frequent in men (12.3%) than women (6.7%) (P = .003), despite the latter being significantly older (52 vs. 48 years, P <.001).
Characteristics of Female Patients
At enrollment, 274 out of 442 women (62.0%) were menopausal. Table 1 reports the baseline characteristics of the female group stratified by reproductive status. In 220 women (80.2%) menopause was spontaneous while in 54 (19.8%) it resulted from surgery. Mean age at the time of menopause was 49.0 µ 2.7 years for spontaneous and 42.5 µ 6.9 years for surgical menopause (P < .0001). Length of estrogen deprivation was significantly longer in women with surgical menopause vs. those with spontaneous menopause (Table 1). Fifty-four of the women (19.7%) had a history of past estrogen or HRT starting soon after the onset of menopause and continuing for a median period of 5 years (range 1-20 years). In 51 of these women (94.5%), HRT was stopped a mean 8.2 µ 3.2 years (median 7 years) before standard antiviral therapy was started.
At baseline, menopausal women had significantly more frequent metabolic alterations (blood glucose, cholesterol) and significantly more histological liver damage than their counterparts who were of reproductive age (Table 1). An increase in the severity of fibrosis was evident between non-menopausal and early menopausal women [staging score: 1.4 (1.0) vs 2.0 (1.0), P = .002], and between early menopausal and late menopausal women [staging score: 2.0 (1.0) to 2.4 (1.2), P = .009]; cirrhosis was present in 1.7% of non-menopausal women, 6.0% of early- and 11.0% of late-menopausal women (non menopausal vs late menopausal: P = .0014). Multivariate analysis showed that, in addition to necroinflammatory activity (OR 1.464, 95% CI 1.256-1.707, P < .001), low platelet levels (OR 0.976, 95% CI 0.966-0.986, P < .001) and elevated GGT (OR 1.010, 95% CI 1.003-1.018, P = .008), also a longer duration of estrogen deprivation (5 to 10 years: OR 4.078 (95% CI 1.013-16.409), P = .048; more than 10 years: OR 4.867 (95% CI 1.476-16.042), P = .009) was independently linked with severe fibrosis (Table 2). Other factors influencing exposure to estrogen, including past pregnancies (OR 0.764, 95% CI .464-1.259, P = .29), more than 2 lifetime pregnancies (OR 0.914, 95%CI 0.607-1.377, P = .66), and HRT use (OR 0.454; 95%CI 0.059-3.508, P = .44) were unrelated to fibrosis severity. Menopausal women also had significantly higher liver necro-inflammatory activity, and a higher rate of steatosis compared with women of reproductive age (Table 1). Multivariate analysis for severe necro-inflammatory activity showed that stage of fibrosis (OR 3.610, 95%CI 1.785-7.301, P=<.001), GGT levels (OR 1.032,95%CI 1.003-1.062, P=.030) and longer duration of estrogen deprivation (5 to 10 years: OR 11.823, 95%CI 2.779-50.302, P= .001; more than 10 years: OR 9.292, 95% CI 2.531-34.111, P = .001) were independently linked with higher inflammatory activity (Table 3). Similarly age (OR 1.133, 95%CI 1.059-1.213, P=<.0001), baseline cholesterol (OR 0.981, 95%CI 1.026-1.333, P P<.0001), BMI (OR 1.170, , 95%CI 1.003-1.062, P=.019) and duration of estrogen deprivation (< than 5 years OR 3,726, 95%CI 1.219-11.385, P=.021; 5 to 10 years: OR 2,648, 95%CI 1.117- 6.276, P=.027; more than 10 years: OR 1.474, 95%CI 0.732-2.969, P=.278) were independently associated with steatosis by multivariate analyses.
Paired liver biopsies taken shortly before and after menopause were available for 19 women (median interval 3 years before and 2 years after); in other 20 women, paired biopsies were obtained while they were still of reproductive ages, at a median interval of 4 years. Analysis of the pairs of menopausal women and those of reproductive age revealed that inflammation had worsened in 8 women (42.1%), increasing by 2 points in 4 women, by 3 points in 3, and by 4 points in 1 of them. It was unchanged in 5 women (26.3%) and improved in 6 (31.6%), decreasing by 2 points in 2 women and 4 points in 4 of them. Fibrosis had progressed in 7 (36.8%), with fibrosis scores increasing by 1 point in 4 women, 2 points in 2, and 3 points in 1 of them. Two women improved by 1 point each and 10 women (52.6%) had no change in fibrosis stage. In the pairs of women who were both of reproductive age, inflammation had worsened in 6 women (30.0%), increasing by 1 point in 3 women and by 2 points in 3. It was unchanged in 5 women (25.0%) and improved in 9 (45.0%), decreasing by 1 point in 4 women and 2 points in 5 of them.
Fibrosis had progressed by 1 point in 7 (35.0%), improved by 1 point in 6 (30.0%) and unchanged in 7 (35.0%).
Results of antiviral treatment
Results are reported as Intention-to-treat analysis. Eight hundred thirty-eight patients completed the antiviral treatment program; forty-nine (4.9%) (34 men/15 women) withdrew because of side effects. SVR was achieved in 511 individuals (51.1%).
Significant independent predictors of SVR failure by multivariate analysis were genotype 1/4, necroinflammatory activity and GGT levels (Supplementary Table 2).
Influence of menopause on response to antiviral therapy
Overall, SVR was achieved by 231 of 442 women (52.2%). SVR occurred in 121 of 263 menopausal women (46.0%) vs. 110 of 163 women of reproductive age (67.5%; OR 2.436, 95% CI 1.620-3.662; P < .0001). The rate of SVR was similar among women of reproductive age who were stratified for parity [nulliparous vs parous 42 of 65 (64.6%) vs 69 of 103 (66.9%; OR 1.200, 95% CI 0.521-2.724; P = .668). Nulliparous menopausal women had worse SVR compared to postmenopausal women with any number of pregnancies [nulliparous vs parous 4 of 32 (12.5%) vs 130 of 242 (53.7%), (OR 0.132, 95% CI 0.029-0.607, P = .009).
Probability of SVR failure between men and unstratified women was similar (OR 0.827, 95% CI 0.615-1.113, P = .210). Instead, women of reproductive age had a significantly lower risk of not achieving SVR (women of reproductive age vs. men: OR 0.452, 0.295-0.693, P < .0001; menopausal women vs. men: OR: 1.212, 95% CI 0.853-1.722, P = .283). Among men, risk of not achieving an SVR was similar after stratification for age groups and comparable to that of women of reproductive age (<45 years) or menopausal (>55 years) women (men aged <45 years vs. men>55 years: OR 1.204, 95% CI 0.964-1.502, P = .101). Percentage of SVR in men younger than 45 years was 59.2% and in those older than 55, 50.0% (p=.114).
By multivariate analysis significant independent baseline predictors of SVR failure in women were the presence of menopause (OR 1.802; 95% CI 1.154-2.813, P = .01), cholesterol level (OR 0.997; 95% CI 0.943-0.991, P = .008), high GGT levels (OR 2.165; 95% CI 1.364- 3.436, P = .001) and genotype 1/4 (OR 3.861, 95% CI 2.433-6.134) P = .006)(Table 4, top). Substituting "duration of estrogen deprivation" for "menopause" in the multivariate model, the OR of SVR failure for women who were post-menopausal for less than 5 years was 8.055 (95% CI 1.834-25.390, P = .006): longer periods of estrogen deprivation were not significantly associated with SVR failure. In addition when replacing length of estrogen deprivation as categorical variable with the linear variable the latest remained significantly associated with lower SVR rate (OR 1.115; C.I. 1.048-1.185).
Restricting analysis to genotype 1-infected women, logistic regression analysis identified only menopause as an independent predictive factor for SVR failure (OR 2.908; 95% CI 1.544- 5.478, P = .001)(Table 4, bottom). Substituting "duration of estrogen deprivation" for "menopause" in the multivariate model revealed that the OR of SVR failure decreased in parallel with increasing time from the menopausal event: less than 5 years: 3.933 (95% CI: 1.274-12.142, P = .017); 5 to 10 years: 2.300 (95% CI 0.982-5.386, P = .055); more than 10 years 1.437 (0.743-2.781), P =.282. Similar to the entire population, in G1, when replacing length of estrogen deprivation as categorical variable with the linear variable the latest remained significantly associated with lower SVR rate (OR 1.088; C.I. 1.006-1.177).
Cytokines levels
Regardless of reproductive status, HCV-positive women had significantly higher serum levels of both TNF-α and IL-6 compared to controls (P < .0001 for all combinations) (Figure 1A). IL-6 levels in women of reproductive age were significantly lower than those in early post-menopausal women [2.6µ1.5 vs 8.9µ10.3 pg/ml; P = .007] and were lower, although to a lesser degree, than those in late post-menopausal women [2.6µ1.5 vs 4.8µ8,2 pg/ml; P = .017]. When only nulliparous women were considered, the difference in IL-6 levels between women of reproductive age and post-menopausal women was greater [3.2µ2.2 vs. 42.3µ18.2 pg/ml; P < .0001] and a small but significant difference was found also in TNF-α levels (27.0µ3.1 vs. 29.2µ2.5; P = .001)(Figure 1A). Among menopausal women, but not women of reproductive age, baseline levels of IL-6 correlated with higher baseline necro-inflammatory scores (menopausal: OR 3.571, 1.494-8.536, P=.004; reproductive age OR .727, .305-1.736, P=.473).
Analysis of the paired serum samples taken shortly before and after menopause revealed significantly different levels for IL-6 (before vs. after: 2.0µ1.2 vs. 9.2µ2.2, P = .03) but not for TNF-α (29.6µ2.4 vs. 28.7µ3.5, P = 0.395)(Figure 1B). The post-menopausal increase in IL-6 levels correlated with fibrosis progression (OR 3.333, 95%CI 1.293-8.591, P = .011) and with worsened necro-inflammatory score (OR 3.667, 95%CI 1.397-9.624, P=.004).
Levels of TNF-α and IL-6 in men were stratified to have age groups comparable to those of females stratified by hormonal status: <45 years (reproductive age); 50-55 years (early menopause); > 60 years (late menopause). Levels found for both cytokines did not differ significantly among the 3 age groups (TNF-α: 28.7µ16.2, 29.2µ21.2, 29.5µ7.5 and IL-6: 27.2µ35.0, 29.0µ42, 34µ51 respectively, NS for all combinations). Levels of TNF-α did not differ significantly between men and women (NS for all combinations), whereas levels of IL-6 were significantly higher among men (p<.0001 for all combinations but vs. nulliparous menopausal females: p=.025, p=.098, p=.393 for <45 years; 50-55 and > 60 year-old men respectively).
Immunohistochemical evaluation of TNF-α and SOCS3
Control liver tissue was negative or had very slight TNF-α staining. TNF-α positivity increased significantly after menopause, the highest values being present in biopsies taken shortly after the occurrence of menopause [liver biopsies scored +/++ for TNF-α: non menopausal vs menopausal: 21 of 93 (22%) vs 36 of 84 (43%), P = .007)] (Figure 1C, panel a).
No significant differences were observed in levels of TNF-α between 20 paired liver biopsies collected from women of reproductive ages (Figure 1C, panel b), whereas in the 19 paired liver specimens taken before and after menopause, staining intensity and the number of positive hepatocytes increased after menopause (Figure 1C, panel b; Figure 2, panel A). Differences in TNF-α scores correlated with circulating levels of IL-6 (2.5µ1.0 vs. 20.6µ41.8 pg/ml in specimens staining negative vs. positive for TNF-α, P = .045).
Staining for SOCS3 was absent from normal liver samples; it was also absent, or very weak, in samples from women of reproductive age, and strong to very strong in samples from post- menopausal women [reproductive age vs menopausal: 17 of 93 (18.2%) vs. 39 of 84 (46.6%), P < .0001]. TNF-α and SOCS3 were co-localized in 60% µ 10% of hepatocytes (Figure 2, Panel B).
|
|
| |
| |
|
|
|
