The prevalence of overweight and obesity has steadily increased in the recent decades, regardless of gender, age or level of country development. For this reason, obesity has been considered a major public health problem worldwide. There are more than 2 billion overweight people and 64% of them are in developing countries15.
A successful public health model has not been yet identified to reverse this epidemic, mainly because it is a multifactorial disease, which depends on changes in lifestyle, such as less caloric intake and greater energy expenditure16,17. In the last decades, the relevance of the intestinal microbiota has been suggested as another possible factor involved in the increase of obesity. However, its impact on the treatment of obesity has been negligible18. Special attention should be given to children, due to the considerable reduction in life expectancy and quality related to obesity19. While overweight and obesity showed an increase in prevalence to 27.5% among adults, the infant population’s prevalence reached the high rate of 47.1% between 1980 and 201315.
Brazil is the fifth nation in absolute number of obese individuals, behind the USA, China, India and Russia15. These data have justified the increased investment in public health for the prevention and treatment of obesity and its comorbidities. Obesity is related to several diseases, such as cardiovascular, orthopedic, diabetes and several types of cancer20, being associated with 14–20% of all cancer deaths21. Obesity is the main risk factor for NAFLD, while other factors such as ethnicity (prevalence of 45% in Hispanics; 33% in whites; 24% in blacks), gender (prevalence of 42% in white men; 24% in white women), type 2 diabetes (22.51%), hyperlipidemia (69.16%), hypertension (39.34%), and metabolic syndrome (42.54%) has less impact22,23,24. NAFLD is considered the hepatic phenotype of metabolic syndrome, which is directly related with insulin resistance and disturbances in glucose and lipid metabolism. The high worldwide incidence of NAFLD and its correlation with obesity and its comorbidities emphasize the importance to evaluate the impact of the hepatic features in this affection. To our knowledge, there is no other study that has identified and correlated such detailed aspects of the NAFLD in a large Brazilian cohort. All individuals who underwent bariatric surgery in this study participated in a preoperative weight loss program which lasts 4 to 12 weeks, includes weekly follow-ups, and is carried out by a multidisciplinary team. Individuals underwent surgery once a minimum of 10% preoperative weight loss was achieved, or with the minimal BMI of 35 kg/m2 for subjects with obesity-related morbidities or 40 kg/m2 for those free of comorbidities. This may justify the low mortality rate (0.33%) after bariatric surgery in this study, since the weight loss may assist to attenuate the associated comorbidities.
Bariatric surgery has largely evolved over time and nowadays its perioperative mortality is arguably as low as 0.2%, which is similar to what is observed after routinely performed operations such as gallbladder removal and hysterectomy, and leads to an overall 40% reduction of all-cause mortality25,26. Nonetheless, it is far from being free of risk, although the incidence of major complications reduced over time as well, mostly after the development of minimally invasive approaches. Early morbidity such as anastomotic leaks and thromboembolic events may occur in up to 2–4% of all operated individuals and, as such, are the most dreaded perioperative complications27,28. Late morbidity is more related to nutritional issues, such as protein, iron, and vitamin deficiencies, which are also associated with a poor compliance to follow-up, as well as surgical complications, such as intestinal obstruction caused by adhesions or internal hernias, abdominal wall hernias and gallstones. Significant hepatic impairment is unusual after the mainstream procedures, but there are a number of cases after more malabsorptive operations, mainly the biliopancreatric diversions29,30,31.
Described only in the 1980s, NALFD is currently considered the main liver disease worldwide, with prevalence in the general population of up to 25%, and analyzing only obese patients, the prevalence exceeds 50%24,32,33,34,35. The prevalence of liver diseases secondary to alcohol consumption remained stable, and the better diagnosis and subsequent development of vaccine and treatment for viral hepatitis decreased the liver injury due to this affection worldwide in the recent decades. On the other hand, liver disease associated with obesity and the consequent evolution to steatosis, steatohepatitis, cirrhosis and hepatocellular carcinoma has become one of the main causes of liver failure and transplantation in recent years, with the aggravating possibility of recurrence in the transplanted organ36,37. In our cross-sectional retrospective study, the NAFLD incidence evaluated by histological examination was 46.48% among the morbidly obese patients who underwent bariatric surgery, and 50% in the first timepoint of the paired longitudinal study. This incidence in bariatric patients was verified in other studies with smaller sample sizes38,39,40,41,42,43. While most of the studies evaluated NAFLD after bariatric surgery by ultrasound scan or another image or laboratory methods, our group performed liver biopsy at two distinct timepoints. Some serum levels of liver function markers were decreased in T2 compared to T1, that corroborate with the improvement of NAFLD verified in histological analysis in T2. Thus, there was a decrease in ALT (p = 0.0117), GGT (p = 0.0072), TGC (p = 0.0216) and COL (p = 0.0031) serum levels after bariatric surgery. The decreased serum albumin (p = 0.0014) and protein levels (p = 0.0109) may be explained by the decrease of protein intake and absorption after bariatric surgery.
Since steatosis (NAFL) and steatohepatitis (NASH) are asymptomatic in the vast majority of patients, we have a population of a few billion inhabitants with such diseases without a proper diagnosis44. This condition is aggravated because although the evolution of these changes towards liver cirrhosis and hepatocellular carcinoma is defined24, unfortunately it is not yet possible to predict which patients will remain stable and which will progress to cirrhosis and cancer. Since access to diagnostic tests is scarce in our country for a large part of the population, several patients receive their diagnosis only after the secondary symptoms of liver failure and cirrhosis appear.
Although bariatric surgery provides NAFLD amelioration, the literature is controversial about what surgical technique is more effective or may be equally efficacious in ameliorating NAFLD. All the patients included in this study underwent the same surgical procedure (RYGB bariatric surgery), and the weight loss outcomes were comparable to the literature45. Only 18% of the 895 patients included in the study were considered diabetic at the time of surgery, thus the histopathological changes found and corresponding liver damage may be attributed to the morbid obesity.
Given the importance of assessing liver function and morphology in susceptible patients, several methods have been developed, such as laboratory tests and indexes based on these tests, in addition to imaging tests such as ultrasound scan, nuclear magnetic resonance imaging and elastography46. As a screening method, ultrasound scan is a non-invasive, easily accessible and low-cost method to diagnose patients with hepatic steatosis, associated with the investigation of other causes of liver disease such as viral hepatitis, alcohol consumption and hepatotoxic drugs. However, liver biopsy, despite being invasive and not without complications, has been considered the gold standard for diagnosis of the disease and is the only method able to differentiate non-alcoholic steatohepatitis from steatosis, to evaluate inflammation and to stage fibrosis46.
In addition to the evaluation of NAFLD’s prevalence among morbidly obese patients who were submitted to the RYGB bariatric surgery in a large cohort, the present study identified the histological grading and evolution of the hepatic lesions in a subset of patients. There was a significant reduction of collagen fiber deposition in the liver biopsies after bariatric surgery, besides decreased hepatic steatosis and hepatocyte ballooning, showing NAFLD regression after the procedure. Moreover, immunohistochemistry for α-SMA and EMR1 corroborated these findings. α-SMA is expressed by hepatic stellate cells, reflects their activation to myofibroblast-like cell and has been directly related to experimental liver fibrogenesis and to human fibrosis in chronic liver disease47. In this study, we identified the earliest stages of hepatic fibrosis in morbid obese patients during and after bariatric surgery, showing positive staining to α-SMA in both timepoints, but less evident in T2 when compared to T1. Likewise, the staining for Kupffer cells marker (EMR1) showed positivity in both groups, but less in T2 than in T1. Kupffer cells are central players in the progression of steatosis to non-alcoholic steatohepatitis (NASH) and fibrosis48. Therefore, the immunohistochemistry analysis illustrated the results of the histopathological changes classification of the NAFLD based on H&E and Masson’s Trichome staining during and after the bariatric surgery. These findings better characterized the NAFLD histological regression that occurred in the majority of the cases of our longitudinal paired study.
Most of this study’s limitations are related to its retrospective character. Some variables for the applied histological score were not evaluated by the pathologists and, therefore, were not in the medical records, having been considered not informed, as pointed out in the Supplementary Information Files. Unfortunately, the late follow-up of bariatric surgery is reportedly erratic, since most patients do not appropriately comply with the appointed consultations49. This general limitation of the bariatric practice was also present within the current study, since the individuals which underwent a second biopsy were not regularly accompanied and thus there are no available data in many time points except for both perioperative situations. However, this cohort has a decent size, and a longitudinal dataset as the one presented is of value for the scientific community, as it shows results of liver histology after bariatric surgery. The preoperative weight loss may also have influenced the histological findings and the relatively low prevalence of NAFLD in the entire study population, as well as explain the low BMI at surgery in comparison with other studies.
The blood liver function tests may to estimate, preoperatively, patients who are prone to unfavorable evolution after bariatric surgery. However, the liver biopsy collected during the bariatric surgery with specific histological analysis may better identify the different types of lesions and map the patients with high risk for hepatic complications in the follow-up after this procedure. Moreover, this histologically focused study may bring new insights on the effects of weight loss in the NAFLD regression. There are a few cases of bariatric patients in the literature who develop hepatic injury after the procedure, and this observation needs further investigation based on the molecular and metabolic pathways involved in this process. Based on the currently available evidence, the occurrence of severe liver disease is unusual and even anecdotal after the mainstream bariatric procedures, but not so rare after predominantly malabsportive operations. The mechanisms of this worsening are yet to be determined, but seem to be related to early rapid weight loss, a degree of protein malnutrition, the lack of hepatotrophic factors, and the effect of high levels of mobilized circulating free fatty acids after surgery, as well as changes in gut microbiota and intestinal bacterial overgrowth50,51,52. The benefit of a routine second liver biopsy at a determined timepoint of the follow-up after bariatric surgery should be confirmed in future large longitudinal prospective studies; however, given the extremely low risk of liver biopsy complications in individuals who would otherwise have to undergo another surgery after their primary bariatric procedure (hernioplasty, cholecystectomy etc.), this seems to be justifiable within this context.