With over one-third of US adults considered obese and an even larger proportion of adults and children being classified as overweight, the prevention and management of obesity has never been more crucial. The burden associated with obesity and its related comorbidities (ORC), including type 2 diabetes (T2DM), cardiovascular disease (CVD), hypertension (HTN), and obstructive sleep apnea (OSA), is tremendous, both in terms of direct costs to the healthcare system and indirect costs associated with decreased productivity. Despite this burden, there are relatively few treatments available for obesity. Currently, bariatric surgery (BS) is the most effective treatment modality for obesity and ORC with laparoscopic Roux-en Y gastric bypass (LRYGB) and laparoscopic sleeve gastrectomy (LSG) being the two most commonly performed procedures. Numerous studies support the clinical superiority of BS over intensive medical weight loss, with many patients experiencing upwards of 60% excess body weight loss (EBWL) and complete remission of T2DM and HTN within two years of BS. For these reasons, along with the decreasing morbidity and mortality, the number of BS has exponentially increased during this time to meet the demand for this important, life-saving procedure.
Contiguous with the rise of surgical weight loss, our understanding of the pathophysiology of obesity and adipose tissue physiology has significantly expanded. The Hsueh laboratory has been instrumental in delineating the interconnection between the adipocyte and adipose tissue inflammation in obesity. Pivotal studies in murine models demonstrated that visceral adipose tissue is significantly more inflamed following high fat
diet and is characterized by reductions in immunosuppressive regulatory T cells (Tregs) and increased pro-inflammatory M1-like macrophages, T helper Type 1 cells (Th1s), and neutrophils as compared to chow fed controls. Furthermore, murine adipocytes become pro-inflammatory immune antigen-presenting cells at the expense of their metabolic function during high fat diet feeding. This process is driven primarily by the upregulation of major histocompatibility complex II (MHCII) which stimulates naive helper T cells to differentiate into pro-inflammatory helper T cells (Th1) and thus suppress the production of adipose tissue Tregs. Importantly, mice with an adipocyte-specific MHCII knockout on a pro-atherogenic background demonstrate significant improvements in insulin sensitivity, atherosclerotic plaque development, and fatty liver disease, further emphasizing the adipocyte’s importance as a pro-inflammatory immune cell. We found similar changes in the adipocyte-T cell landscape in obese humans and reported that visceral and subcutaneous adipose tissue depots were markedly more inflamed, with reduced Tregs and increased pro-inflammatory adipocytes and macrophages as compared to lean controls. As with mice, reduced adipose tissue Treg abundance was associated with increased insulin resistance and positively correlated with increasing body mass index. Within our bariatric patients, we further discovered that reduced preoperative visceral Treg abundance was associated with higher disease severity for ORC, including insulin-dependent diabetes, preoperative statin use and need for multiple anti-hypertensive agents at baseline.
While LRYGB and LSG are well-tolerated operations with reportedly excellent outcomes in the short-term, weight regain, comorbidity recidivism, and micronutrient deficiencies are significant post-operative complications that require long-term follow up
and management. Despite the comparative success of BS and our advances in understanding the pathophysiology of obesity, the relationship between adipose tissue inflammation and surgical weight loss and its impact on long-term post-surgical outcomes remains unexplored. We discovered that one year post-operatively over one-third of our LSG patients were within the bottom quartile of EBWL outcomes, corresponding to less than 40% EBWL, which rose to 52% by two years and 69% by 3 years. In fact, this metric is considered “weight loss failure” after BS (<50% EBWL at 24 month post-operatively) and meets criteria for revisional surgery. More importantly, LSG patients with baseline comorbidities experienced a greater than 3-fold increase in the odds of experiencing “weight loss failure” up to three years following BS. As the number of LSG procedures continues to rise in the United States, this study identified an increasing group of patients who may be at risk of failing weight loss surgery.
Given our preliminary work on adipose tissue changes in lean versus obese patients, we hypothesized that post-surgical changes in adipose tissue inflammation would be an important window into examining post-surgical outcomes following BS including insight into poor post-surgical weight loss. We compared subcutaneous adipose tissue obtained at the time of BS and approximately one year after surgery and evaluated changes in adipocytes, macrophages, and adipose resident T cells. In our patient population, post-surgical adipocytes demonstrated a striking upregulation of metabolic gene expression, comparable to that seen in lean patients, with a modest decrease in inflammatory genes and relatively unchanged MHCII-associated gene expression. In addition, adipose tissue Tregs were globally decreased, and adipose tissue macrophages appeared grossly more pro-
inflammatory but exhibited patterns associated with decreased antigen presenting capacity. Within the macrophage population, downregulation of MRC-1, a classical M2-like macrophage marker, and MHCII-related expression, including HLA-DPA1 were positively correlated with increasing weight loss. On the other hand, negative changes in Treg abundance were associated with poorer weight loss outcomes, suggesting a synergism between Treg maintenance and weight loss. In fact, patients with lower visceral adipose tissue Tregs preoperatively demonstrated dramatically reduced weight loss one year after BS compared to those with a higher Treg abundance.
Given that retinoic acid, the active form of vitamin A, plays an important role in stabilizing and stimulating Treg differentiation, we explored the possible ramifications of hypovitaminosis A in terms of adipose tissue Treg abundance further within a cohort of 570 BS patients and discovered that the post-operative prevalence of vitamin A deficiency increased over two-fold by six months, from 15% to 33%, and remained significantly elevated at 12 months following BS. More importantly, post-operative plasma retinol levels were correlated to adipocyte RBP-4 gene expression, suggesting that increased? RBP-4 may reflect decreased adipocyte retinol stores. More importantly, higher adipocyte RBP-4 expression was correlated with reductions in adipocyte CD80 and 86, which are crucial co-stimulators for Treg differentiation, whereas decreasing post-operative plasma retinol levels were associated with a higher reduction in subcutaneous adipose tissue Tregs compared to patients with stable or increased levels.
These findings suggested that not only is there a significant proportion of post-bariatric patients that were vitamin A deficient, but also that these changes were reflected
in adipocyte gene expression which we hypothesize impacts the surrounding immune environment. This is particularly relevant to bariatric care for multiple reasons. While ASMBS recommends a comprehensive micronutrient panel be ordered pre and post-operatively, this is not standardized across bariatric centers. In fact, a number of programs that we reached out to for a multi-institutional study on vitamin A deficiency did not collect vitamin A levels as part of routine care. Furthermore, protocols for replenishing vitamin A are not uniform and are generally targeted at levels less than 1.05 mol, which is based mainly on the development of ocular symptoms. As a result of our findings, further studies evaluating the optimal method of vitamin A delivery and the threshold and rigor for vitamin A supplementation is warranted. To support this, we have shown that high fat diet fed mice demonstrate a complete normalization of insulin sensitivity and a 50% reduction of liver fat stores following six days of subcutaneous retinoic acid injections suggesting that even modest vitamin A supplementation may be beneficial in the context of diet-induced obesity.
The relationship between adipose Treg abundance and post-surgical vitamin A deficiency are exciting and have become the basis for ongoing trials on murine models and vitamin A deficient patients following BS. The objectives of these future studies are not only to elucidate the mechanism by which vitamin A supplementation may be beneficial to Treg maintenance following BS, but also to identify the threshold and time point at which more aggressive supplementation is beneficial. Furthermore, we hope to discover specific clinical and symptom correlates that can help bariatric providers screen for this important deficiency.