AIM: To study pentoxifylline effects in liver and adipose tissue inflammation in obese mice induced by high-fat diet (HFD).

METHODS: Male swiss mice (6-wk old) were fed a high-fat diet (HFD; 60% kcal from fat) or AIN-93 (control diet; 15% kcal from fat) for 12 wk and received pentoxifylline intraperitoneally (100 mg/kg per day) for the last 14 d. Glucose homeostasis was evaluated by measurements of basal glucose blood levels and insulin tolerance test two days before the end of the protocol. Final body weight was assessed. Epididymal adipose tissue was collected and weighted for adiposity evaluation. Liver and adipose tissue biopsies were homogenized in solubilization buffer and cytokines were measured in supernatant by enzyme immunoassay or multiplex kit, respectively. Hepatic histopathologic analyses were performed in sections of paraformaldehyde-fixed, paraffin-embedded liver specimens stained with hematoxylin-eosin by an independent pathologist. Steatosis (macrovesicular and microvesicular), ballooning degeneration and inflammation were histopathologically determined. Triglycerides measurements were performed after lipid extraction in liver tissue.

RESULTS: Pentoxifylline treatment reduced microsteatosis and tumor necrosis factor (TNF)-α in liver (156.3 ± 17.2 and 62.6 ± 7.6 pg/mL of TNF-α for non-treated and treated obese mice, respectively; P < 0.05). Serum aspartate aminotransferase levels were also reduced (23.2 ± 6.9 and 12.1 ± 1.6 U/L for non-treated and treated obese mice, respectively; P < 0.05) but had no effect on glucose homeostasis. In obese adipose tissue, pentoxifylline reduced TNF-α (106.1 ± 17.6 and 51.1 ± 9.6 pg/mL for non-treated and treated obese mice, respectively; P < 0.05) and interleukin-6 (340.8 ± 51.3 and 166.6 ± 22.5 pg/mL for non-treated and treated obese mice, respectively; P < 0.05) levels; however, leptin (8.1 ± 0.7 and 23.1 ± 2.9 ng/mL for non-treated and treated lean mice, respectively; P < 0.05) and plasminogen activator inhibitor-1 (600.2 ± 32.3 and 1508.6 ± 210.4 pg/mL for non-treated and treated lean mice, respectively; P < 0.05) levels increased in lean adipose tissue. TNF-α level in the liver of lean mice also increased (29.6 ± 6.6 and 75.4 ± 12.6 pg/mL for non-treated and treated lean mice, respectively; P < 0.05) while triglycerides presented a tendency to reduction.

CONCLUSION: Pentoxifylline was beneficial in obese mice improving liver and adipose tissue inflammation. Unexpectedly, pentoxifylline increased pro-inflammatory markers in the liver and adipose tissue of lean mice.

Non-alcoholic fatty liver disease (NAFLD) defines a spectrum of hepatic disorders including steatosis or uncomplicated fatty liver and non-alcoholic steatohepatitis (NASH). NAFLD is frequently associated with metabolic syndrome establishment and it occurs more often in males than in females and primarily affects the middle aged and the elderly[1].

Several factors are associated with the development of fatty liver, and NAFLD diagnosis requires the exclusion of secondary etiologies, including alcohol consumption, drug usage, hepatitis B and C[2]. Although simple steatosis is considered benign, NASH can progress to end-stage liver disease, such as fibrosis, cirrhosis and hepatic cancer[3]. The mechanism of steatosis progression to more severe liver injuries is not fully understood, but it is associated with several risk factors, including elevated serum transaminases, inflammation upon liver biopsy, old age, diabetes mellitus, high body mass index (≥ 28 kg/m²), presence of ballooning plus Mallory hyaline or fibrosis upon biopsy and increased visceral adipose tissue[2,4].

The growing epidemic of obesity and an aging population have led to an important demand for a medical therapy for NAFLD, but several decades of pharmacological research have resulted in very few options[2]. As NAFLD is considered a hepatic manifestation of a metabolic syndrome, the first treatment approach is a lifestyle change, including dietary alterations and increased physical activity to reduce adiposity and body weight[5]. Therapeutic drugs are an adjunctive approach to lifestyle changes. Statins are used to control dyslipidemias, metformin and glitazones are used to control diabetes mellitus, and angiotensin receptor blockers are used to control inflammatory cell recruitment and hepatic fibrosis development in addition to their anti-hypertensive effects. In total, these drugs aim to control the symptoms of the metabolic syndrome[1,2,6].

Pentoxifylline is a non-selective phosphodiesterase inhibitor that has been reported to have antioxidant activity and decrease tumor necrosis factor (TNF)-α gene transcription. Pentoxifylline treatment improved the 6-mo survival rate of patients with severe alcoholic hepatitis compared with placebo[7]. Recent studies have shown that pentoxifylline may be a promising drug therapy for NASH treatment[8-10]. Experimental results with pentoxifylline have shown conflicting data depending on the model employed. In NAFLD induced by a choline- and methionine-deficient diet, pentoxifylline treatment was beneficial because it decreased hepatic inflammation and alanine aminotransferase (ALT) levels[11]. However, in a genetic obesity model, pentoxifylline worsened fatty liver in ob/ob mice because it increased intestinal glucose absorption, and thus, hyperglycemia[12].

Considering that obesity is strongly associated with the development of NAFLD and one of the main causes of epidemic obesity is a hyperlipidic and hypercaloric diet, our study evaluated the effects of pentoxifylline in a high-fat diet (HFD)-induced obesity model. Metabolic parameters, hepatic inflammation and adipose tissue alteration were studied after 2 wk of pentoxifylline treatment in mice after 12 wk of a HFD.

NAFLD is currently considered a consequence of obesity, and its prevalence in obese subjects is very high. Sedentary life style and consumption of foods with high-fat and high-caloric content are the main contributing factors to obesity[14]. The reduction of body weight and lifestyle changes are the primary recommendations to control NAFDL, and pharmacological interventions aim to induce weight loss (e.g., orlistat and sibutramine), to improve the antioxidant response (e.g., vitamin E and C, ursodeoxycholic acid) or to ameliorate insulin resistance and glucose and lipid metabolism (e.g., metformin, thiazolidinediones)[15]. Pentoxifylline has been considered an alternative treatment to control NAFDL, as it has been recommended in alcoholic fatty liver disease by acting as an anti-inflammatory drug[16]. Pentoxifylline is a methylxanthine derivative that acts as a nonspecific phosphodiesterase inhibitor to promote an increase in cyclic AMP levels and inhibit TNF-α gene transcription[17,18].

A high-fat diet obesity model is suitable to study metabolic and liver disease associated with adipose tissue expansion and to study potential therapeutics to control obesity. Our results show that Swiss mice fed a HFD for 12 wk present increased body weight, increased adiposity, adipose tissue inflammation, insulin resistance, hyperglycemia, steatosis, inflammation and increased TNF-α and MCP-1 levels in the liver. Pentoxifylline treatment did not change the final body weight but did decrease the liver weight. Visceral (epididymal) adipose tissue increased after pentoxifylline treatment, which may explain why we did not observe a decrease in body weight. Pentoxifylline treatment did not improve glucose homeostasis, but some NAFLD features improved, such as hepatic steatosis, inflammation, TNF-α levels, and serum AST levels. Our results are consistent with a previous study, where Sprague-Dawley rats fed a HFD for 16 wk were treated with pentoxifylline for 4 wk (16 mg/kg per day). The results showed decreased AST levels but not ALT, and improvements in basal glucose but not HOMAIR index[19]. Additionally, in a previous study of Sprague-Dawley fed HFD for 10 wk and treated with pentoxifylline for 6 wk (50 mg/kg per day), hepatic steatosis and plasma levels of TNF-α were reduced[20]. In our work, we evaluated both hepatic and adipose tissue TNF-α levels and found that pentoxifylline treatment reduced both. However, glucose homeostasis did not improve, but TNF-α and IL-6 levels decreased. A balance between leptin and adiponectin have been suggested to have a role in metabolic syndrome and type 2 diabetes[21]. Pentoxifylline treatment could not reverse the alterations in the obesity-induced leptin/adiponectin ratio.

Interestingly, we observed increased TNF-α in the liver and increased PAI-1 and leptin in adipose tissue of lean mice after pentoxifylline treatment. A systematic review of pentoxifylline data in patients with NAFLD revealed that AST and ALT plasma levels and liver histological scores were improved in several studies using pentoxifylline. However, pentoxifylline treatment did not inhibit plasma cytokines levels such as IL-6 in all studies[22]. Although pentoxifylline has been shown to inhibit TNF-α, Zein et al[9] reported that pentoxifylline treatment did not inhibit TNF-α plasma levels in NASH patients. The authors suggested that TNF-α plasma levels may not be related to hepatic levels of this cytokine because they observed histological improvement. Both the adipose tissue and liver of lean mice increased pro-inflammatory cytokine production in response to pentoxifylline treatment, an unexpected result that should be further studied. Interestingly, triglycerides levels presented a tendency to reduction in liver of lean mice. Several findings suggested that triglycerides per se are not toxic, on contrary; they protect liver from lipotoxicity by buffering the accumulation of fatty acids. Triglycerides synthesis inhibition improves steatosis but stimulates oxidizing systems that increase hepatic oxidative stress and liver damage[23]. Although, pentoxifylline is able to decrease oxidative stress and to inhibit lipid peroxidation in patients with NASH[24], we did not rule out the possibility that lean mice had an increase in oxidative response due pentoxifylline treatment. We hypothesize that metabolic status, liver metabolism, adiposity or inflammation degree can interfere with the pentoxifylline response, which could explain the controversial data obtained in different clinical studies of NAFLD patients. In this line of reasoning, pentoxifylline is effective in states of hyperinflammation because relevant anti-inflammatory effects can be achieved only in the presence of sufficient adenosine concentrations[25]. Metabolic stress, hypoxia and inflammation are conditions related to increase adenosine extracellular concentrations[26], and thus, could interfere with the pentoxifylline response.

In conclusion, our results showed that pentoxifylline was beneficial in an obesity-associated NAFLD model by improving liver inflammation and adipose tissue inflammation, but it was not able to improve obesity-induced metabolic disturbances. Unexpectedly, pentoxifylline treatment increased pro-inflammatory markers in the liver and adipose tissue of lean mice.