Fig. 24.1
The procedure involves the placement of a silicone-made ring around the upper portion of the stomach in order to isolate a small (20–30 ml) proximal gastric pouch which communicates with the distal stomach through a narrow stoma. The inner volume of the ring can be augmented or reduced by injecting or aspirating saline solution from the subcutaneous port which is connected to the band through an intra-abdominal catheter
Mechanism of Action
LAGB is a pure restrictive procedure: the passage of food from the proximal small gastric pouch to the distal stomach is made difficult, with a consequent increase of the intraluminal pouch pressure. This mechanism promotes an early and prolonged sensation of satiety by stimulating of the gastric wall baroreceptors [38]. Thanks to the presence of the subcutaneous port, the surgeon is able to adjust the inner volume of the ring and then the degree of restriction, depending on both the clinical outcome and the comfort of the patient. This implies a close compliance to the follow-up program and the strict observation of dietary recommendations, which are mandatory for a successful LAGB. Solid foods should be preferred to semisolid or liquid ones, because the latter are often hyper caloric and get more easily through the band, without augmenting the pressure in the pouch.
Complications
Mortality: 0.1–0.2 %
Major Complications
Slippage (0.2–12.5 %): it is the prolapse of the gastric wall proximally through the band. Slippage-related symptoms are dysphagia, vomiting, and gastroesophageal reflux disease (GERD). It can be diagnosed by upper gastrointestinal contrast studies showing the improper placement of the band and the dilation of the gastric pouch. The management can be conservative (band deflation, fasting or liquid diet, stomach decompression by nasogastric tube). When the conservative management is not successful, the band must be removed.
Erosion (0–3.7 %): it is a process of chronic erosion of the band through the layers of the gastric wall until it penetrates into the gastric lumen. Clinical presentation may be at variance, epigastric pain, dysphagia, weight regain, and infection of the subcutaneous port. Removal of the band either laparoscopically or with a combined laparoscopic and endoscopic approach is indicated.
Band complications (up to 11 %): the most common ones are the infection of the subcutaneous port and the rupture or disconnection of the catheter. The former can be treated conservatively or through the substitution of the port; the latter is diagnosed by an abdominal x-ray and its management can require surgery.
Outcome
Weight Loss
It has been shown that LAGB can determine an EWL of 40–50 % with an average BMI loss of ten points. However, long-term outcomes are impaired by significant incidence of failures, because of insufficient weight loss (IWL) or WR or psychological intolerance to a “foreign body.” These conditions may lead to revision operations. The best results occur in young women <45 years, BMI <45, no binge eating disorders or sweet eating, and good compliance to follow-up [39].
Comorbidities
After LAGB, arterial hypertension, obstructive sleep apnea syndrome (OSAS), and hyperlipidemia are resolved in 38.4, 94.6, and 71.1 % of patients, respectively. T2DM resolved in <50 % of cases [40].
Advantages
Safe and feasible: LAGB is a simple procedure, with short operative time (about 30 min in high-volume centers) and hospital stay (1 or 2 days). Mortality and postoperative complication rates are very low.
Reversible: the procedure does not involve any resection or anastomosis, thus restoring the original anatomy of the stomach is technically simple.
Disadvantages
Failure rates: the incidence of IWL or WR after LAGB is up to 30 % in some series. When failure occurs, a revision procedure (band removal and conversion to a RYGBP or SG) could be indicated to achieve a satisfying outcome. Revision procedures have higher rates of postoperative complications when compared with primary surgery.
Patients’ compliance: adherence to the follow-up program and observation of a hypocaloric diet play a crucial role to a successful LAGB. An accurate selection and education of the patients undergoing this procedure is mandatory.
24.5.3 Laparoscopic Roux-en-Y Gastric Bypass
Introduction
Gastric bypass (GB) as a bariatric intervention was first proposed by Mason in the 1960s, on the observation that in patients undergoing Billroth II procedures (partial gastrectomy with gastrojejunal anastomosis) a significant and durable weight loss was obtained [41]. Since then, several aspects of the surgical technique have been modified, entailing capacity of the small gastric pouch, Roux-en-Y reconstruction, and variations in the length of the alimentary limb. In 1994, Wittgrove first showed its feasibility with a laparoscopic approach [42] and nowadays laparoscopic RYGBP (LRYGB) represents the most performed bariatric procedure worldwide, accounting for about 50 % of all bariatric interventions [37] (Fig. 24.2).
Fig. 24.2
The stomach is divided to form a small proximal gastric pouch of about 25–30 ml of volume. The transection begins about 6–8 cm below the esophageal gastric junction on the lesser curvature and proceeds first horizontally toward the great curvature, then vertically toward the angle of His. The jejunum is then divided below the ligament of Treitz and the proximal bowel segment (biliopancreatic limb) is connected to the alimentary limb 75–150 cm distal to the gastrojejunostomy. In patients with a BMI >50 kg/m2, a 300 cm-long alimentary limb has been devised to improve the weight loss, by inducing a certain degree of malabsorption. The gastrojejunal anastomosis can be performed either by means of laparoscopic stapler (linear or circular) or by manual suture
Mechanisms of Action
RYGBP exerts its weight loss action through the reduction of caloric intake, achieved by means of concurrent mechanisms: the restrictive effect of the small gastric pouch (25–30 ml), the reduced sense of hunger, and the early feeling of satiety induced by significant changes in the secretion of neuro hormones. In particular it has been shown that RYGBP entails the decrease of ghrelin and the rise of peptide-YY (PYY) and of glucagon-like peptide-1 (GLP-1) serum levels [43]. Ghrelin is an orexigenic hormone which plays a role in both mealtime hunger and long-term regulation of body weight. Bypass surgery promotes the reduction of ghrelin secretion through the process of “override inhibition.” It means that the exclusion of the greater part of the stomach from food passage elicits a persistent signaling that initially super stimulates, then exhausts the secretion of ghrelin. Release of GLP-1 and PYY from L cells of the small bowel is stimulated by the rapid transit of partially digested food in the ileum, eliciting an early sensation of satiety [44].
The abovementioned neurohormonal changes are very similar to those reported after LSG and are involved in the effects of LRYGB on diabetic patients. In fact, while ghrelin impairs both insulin secretion and sensitivity, GLP-1 improves insulin activity, promotes the growth of pancreatic beta-cells, and reduces hepatic glucose production through the inhibition of glucagon secretion [45].
Complications
Mortality: 0.2–0.4 %
Acute Complications (<30days)
Anastomotic leak (2 %): in most cases occurs at the gastrojejunal anastomosis or at the gastric staple line of the pouch. The pathogenesis can be related to the presence of abdominal collections, to a deficient vascularization, or to abnormal tension on the anastomosis. The management can be conservative (percutaneous drainage, enteral/intravenous nutrition, broad-spectrum antibiotics). Uncontained leaks with systemic sepsis and cardiovascular instability require surgery.
Hemorrhage (1.9 %): almost always from the gastrojejunal anastomosis and can occur either outside or inside the gastrointestinal lumen. It can be managed by interventional radiology (embolization) or by endoscopy. When these maneuvers fail, the surgical approach is recommended.
Late Complications (<30days)
Stenosis of the gastrojejunal anastomosis (5 %): it can be successfully managed by means of one or more sessions of endoscopic dilation. The surgical intervention, entailing the revision of the anastomosis, should be considered only in selected cases when endoscopic approach has failed.
Internal hernia (3.7 %): represents the most common cause of bowel obstruction after RYGBP. It is caused by lack or improper closure of the mesenteric defects during primary bypass surgery, in particular when the alimentary limb is transposed upward through the transverse mesocolon (transmesocolic reconstruction). The laparoscopic reduction of the hernia represents the approach of choice; however, the presence of adhesions and of inflated bowel loops may require conversion to open surgery.
Outcome
Weight Loss
Published data report a 60–70 % EWL at a 15-year follow-up [40]. However, up to 20 % of RYGBP patients may present with long-term weight loss failure or regain (%EWL <50).
Comorbidities
More than 75 % of patients undergoing LRYGB achieve the control of their obesity-related comorbidities. T2DM, hypertension, hyperlipidemia, and OSAS resolved in 83, 75, 93, and 86 % of cases, respectively [40].
Advantages
Effectiveness and safety: due to its excellent clinical outcome and its low mortality and morbidity rates, the American Society for Metabolic and Bariatric Surgery (ASMBS) considers LRYGB the bariatric procedure of choice.
GERD improvement: RYGBP improves esophageal reflux by reduced acid secretion from the small gastric pouch and by the diversion of the biliary esophageal reflux. Therefore, most authors perform LRYGB, rather than LAGB or LSG, when GERD and/or HH are present. Moreover LRYGB is considered the revision procedure of choice in patients developing a symptomatic reflux not controlled with medical therapy after LAGB or LSG.
Disadvantages
Complexity: LRYGB is a technically demanding procedure; the available literature suggests an experience of 50–150 cases as a safe and proficient learning curve.
No standard approach in case of failure: unlike patients with failed LSG, who may undergo the second stage of BPD with duodenal switch (BPD-DS) or bypass surgery, there is still no consensus about the surgical strategy to manage WR or IWL after LRYGB. Different revision procedures have been proposed (e.g., resizing or banding the pouch, modifying the length of the alimentary limb), but none of them is consistently supported by published data.
Presence of a “blind” gastric portion: patients undergoing bypass surgery should be carefully informed that a major portion of their stomach, as well as the biliary tract, will be difficult to explore by endoscopy because of the anatomical changes consequent to the procedure.
24.5.4 Laparoscopic Biliopancreatic Diversion
Introduction
In the early 1970s the pioneer works by Scopinaro represented the most systematic scientific approach to the problem of malabsorption as therapy for obesity and were the physiopathological basis for a novel malabsorption procedure [46]. In 1976 Scopinaro performed the first BPD in a patient with morbid obesity. Today BPD is still the most effective bariatric procedure in terms of weight loss and resolution of obesity-related diseases [40]. In order to reduce the high rate of postoperative complications affecting Scopinaro’s BPD, such as peptic ulceration at the anastomosis and malabsorption-related side effects, in 1988 Hess et al. proposed a modification of the original BPD, the BPD-DS (see surgical technique) [47]. According to the Survey of the International Federation for the Surgery of Obesity and Metabolic Disorders (IFSO), in 2011 BPD-DS has accounted for the 2.2 % of the overall bariatric procedures performed worldwide, with a prevalence of BPD-DS on BPD (5,271 vs. 2,324) [37] (Fig. 24.3).
Fig. 24.3
(a) The Scopinaro’s BPD entails a horizontal gastrectomy, performed by means of a linear stapler, leaving a 300–400 ml residual capacity of the stomach. Then the small bowel is measured from the ileocecal valve toward the ligament of Treitz and sectioned at 250–300 cm. The proximal stump (biliopancreatic limb) is anastomosed to the ileum at a distance of 50 cm from the ileocecal valve (common limb), while the distal stump (alimentary limb) is anastomosed to the gastric remnant. Cholecystectomy may be added. (b) In the BPD-DS, gastrectomy is performed in a vertical manner (sleeve gastrectomy) rather than horizontal, and the duodenum is divided 2–3 cm distal to the pylorus, and a duodenal-jejunal anastomosis is performed leaving in place the pylorus sphincter. The length of the common limb is 100 cm, twice as in the original Scopinaro procedure
Mechanism of Action
Scopinaro’s BPD and BPD-DS exert their weight loss and metabolic actions by reducing dramatically the calories intake of the small bowel from the ingested nutrients. In fact, since bile salts and food blend only in the 50 cm of the common limb, the absorption of fats occurs in this short bowel segment. To the contrary the absorption of protein and starch, which is mediated by the enzymatic activity of the intestinal brush border, takes place in the entire intestinal segment comprised between the gastroenterostomy and the ileocecal valve. These changes in the pattern of nutrients absorption entails that, after biliopancreatic diversion, patients absorb an established amount of calories (about 1,700 Kcal in the man and 1,400 Kcal in the woman), irrespective of size and quality of their meals. The only limitation concerns the ingestion of monosaccharides and disaccharides, contained in fruits, candies, milk, sweet drinks, and alcohol, whose absorption is not impaired by BPD [48]. Despite a lower malabsorption effect, BPD-DS when compared to Scopinaro BPD, presents with similar weight loss and comorbidities resolution rates. The incidence of malabsorption-related side effects is lower.
BPD exerts a strong antidiabetic action, owing not only to the large weight loss and to the reduced nutrients absorption but also to the early and durable postoperative changes in the secretion of intestinal hormones. In fact after BPD a significant increase in the serum concentrations of PYY and GLP-1 has been demonstrated, entailing the improvement of both insulin sensitivity and secretion. These modifications are related either to the exclusion of the foregut (duodenum and jejunum) from the passage of food or to the rapid delivery of partially digested nutrients in the distal ileum, stimulating the secreting activity of L cells. Concerning the plasma levels of ghrelin, they are unchanged or increased after Scopinaro’s BPD, while they are significantly reduced after BPD-DS, due to resection of the gastric fundus (sleeve gastrectomy) [49].
Complications
Mortality: 0–2.7 %
Early Complications (<30days)
Hemorrhage and leaks (2–4 %): they can affect either the staple line of the gastric remnant and of the duodenal stumps or the anastomotic sites (gastroileal or duodenum-ileal). Bleeding can occur inside or outside the gastrointestinal lumen and is usually self-limited. Leaks can be managed conservatively by percutaneous drainage. Surgery is indicated in case of high-flow bleeding not responsive to embolization and in uncontained leaks in hemodynamically unstable patients.
Late Complications (<30days)
Peptic ulcer (3 %): affects the gastro-ileal anastomosis in the Scopinaro’s BPD. The healing process can entail development of a fibrotic stenosis of the anastomosis which can be successfully treated by endoscopic dilation.
Nutritional deficiencies (4–6 %): they are related to the malabsorption effect of the procedure. Preventive lifelong oral integration of iron, calcium, and vitamins is the treatment of choice. These complications are less pronounced in BPD-DS patients.
Protein malnutrition (1 %): it is the most feared late complication of BPD. Poor compliance to postoperative follow-up, low proteins diet, or excessive malabsorption are the most frequent causes. Protein malnutrition can be managed by parenteral nutrition and dietary interventions. In severe cases surgical revision by means of common limb elongation or even suppression of the malabsorption effect is indicated.
Outcome
Weight Loss
It has been shown that BPD (both Scopinaro’s procedure and BPD-DS) is the most effective procedure in terms of weight loss, with an EWL greater than 70 % at a very long follow-up (up to 15 years) [40].
Comorbidities
After BPD, hypertension, OSAS, and hyperlipidemia resolved in 83.4, 91.9, and 99 %, respectively [40].
T2DM
Almost all (97 %) diabetic patients definitively stop glucose-lowering drugs early soon after the procedure. The rate of resolution is lower in patients with a longer disease duration, because of a definitive impairment of pancreatic beta-cells function.
Advantages
Modulation of the malabsorptive effect: since the BPD malabsorption depends on the length of the alimentary and common limbs, the degree of this effect can be tailored by modifying the length of the common limb or of the alimentary limb or both.
Long–term results: clinical outcome from BPD series with a follow-up >15 years strongly supports its long-lasting effects in terms of weight loss and comorbidities’ resolution.
Two–step surgery: in high-risk super-obese patients, BPD-DS is usually performed in two subsequent surgical steps: SG first, then BPD-DS after 6–12 months, with a reduction in mortality and morbidity compared to single step BPD-DS.
Disadvantages
Complexity and safety: laparoscopic BPD is a technically demanding procedure involving a partial gastrectomy (horizontal or vertical) and two anastomosis: gastroileal (or duodenojejunal) and jejunoileal. In addition the BPD-DS entails section of the duodenum 2–3 cm from the pylorus. Even if acceptable, the morbidity and mortality rates are higher when compared to other bariatric procedures.
Side–effects: after BPD, patients could experience the development of malabsorption-related side effects, such as diarrhea with a daily number of evacuation up to ten and foul smelling stools, nutritional deficiencies, and flatulence. When these side effects are severe, they can impair the quality of life of the patient, even requiring a surgical correction.
24.5.5 Sleeve Gastrectomy
Introduction
LSG was initially performed as part of a hybrid malabsorptive procedure, the BPD-DS [47]. In early 2000s Micheal Gagner proposed a two-stage model of intervention, LSG first and then BPD-DS after an average 12 months interval, in order to reduce the high rate of complications and mortality registered after single-stage BPD-DS [50]. The good outcome of the first-stage LSG in terms of weight loss and comorbidities resolution encouraged a growing number of surgeons to perform LSG as a sole bariatric operation, and, in 2009, the ASMBS issued a position statement recommending LSG as an approved bariatric procedure [51]. Nowadays the LSG has the fastest growing rate and is the second most performed bariatric procedure worldwide after RYGBP [37] (Fig. 24.4).
Fig. 24.4
The procedure entails the resection of the body and the fundus of the stomach by means of laparoscopic linear stapler. The capacity of the gastric remnant (60–100 ml) is tailored by the placement of a bougie alongside the lesser curvature of the stomach. The staple line could be reinforced by means of buttressing materials or by overlocking with absorbable suture in order to reduce the incidence of postoperative complications such as bleeding or leakage
Mechanisms of Action
Weight loss induced by LSG was initially ascribed solely to a restrictive action. However, it has been demonstrated that both the resection of the gastric fundus and the rapid emptying of the stomach promote significant neuro hormonal changes: in particular the former causes a significant decrease in the circulating levels of ghrelin, the hormone of appetite, while the latter entails an increased satiety through a food-mediated release of GLP-1 and PYY from the L cells of the small intestine. These changes persist 1 year after surgery, showing that LSG can be considered a “food-limiting” operation rather than a restrictive procedure [52]. These neuro hormones, also called incretins, are involved in the antidiabetic effect of LSG. In fact, while ghrelin acts by suppressing the insulin-sensitizing hormone adiponectin, blocking hepatic insulin signaling, and inhibiting insulin secretion, GLP-1 and PYY either improve insulin secretion and sensitivity or reduce hepatic glucose production through the inhibition of glucagon secretion. Serum levels of PYY and GLP-1 significantly augment just 72 h after LSG, before any weight loss or food passage through the alimentary tract has occurred. These early changes, which are associated with an immediate improvement of glucose homeostasis, suggest that there is an intrinsic neuro hormonal effect of the procedure and lead to the formulation of a “gastric hypothesis”: the diminished hydrochloric acid production induced by the significant reduction of oxyntic cell mass stimulates the vagally innervated antral mucosa, left intact by SG, to secret gastrin-releasing peptide which induces the first early phase of GLP-1 secretion [53].
Complications
Mortality: 0.1–0.5 %
Major Postoperative Complications
Hemorrhage (1–2 %): usually occurs within the first 24–48 h. The treatment is interventional radiology or, in fewer instances, open or laparoscopic surgical exploration.
Gastric leak (0–7 %): it is the most feared complication. It can occur within 7 days postoperatively (acute leaks) or within 1–6 weeks (early leaks); later leaks occur less frequently. In 90 % of cases, leaks occur at the upper portion of the gastric suture line suggesting an ischemic pathogenesis. The treatment is usually nonoperative (percutaneous drainage of the abdominal collection, endoscopic stents, and parenteral nutrition). Surgery is indicated in uncontained, symptomatic leaks with signs of general infection and hemodynamic instability.
Stenosis (0.6 %): usually occurs at the corpus-antrum transition zone (incisura angularis) of the gastric tubule. RYGBP reconstruction after failed conservative stricture treatment is a valid therapy.
GERD (6.5 %): published data are not so consistent to determine whether the preoperative diagnosis of GERD and/or HH should be considered a contraindication to LSG. However, it has been shown that the repair of an HH during LSG plays a beneficial role on GERD outcome: the hiatal area inspection should be performed always during LSG [54].
Outcome
Weight Loss
Data from a recent Consensus Summit accounting for more than 46,000 LSGs reported a mean EWL of 50 % at 6 years of follow-up. Long-term WR remains a controversial and largely insufficiently documented issue [55].
Comorbidities
The remission rate of arterial hypertension, OSAS, and hyperlipidemia after LSG ranges from 15 to 93 %, 39 to 100 %, and 5 to 75 %, respectively [55].
Advantages
Feasibility: LSG is quite simple and does not involve anastomosis, entailing shorter learning curve (50 procedures) and operative time (about 60 min) when compared with LRYGBP and BPD. However, there are some technical aspects, such as the complete fundectomy and the accurate exploration of the hiatal area, whose proper execution is of paramount importance for a successful LSG.
Effectiveness and safety: the clinical outcome of LSG in terms of weight loss and comorbidities resolution is comparable with those of RYGBP, while the incidence of major postoperative complication is lower, although not significantly, than in RYGBP and in BPD.
Lack of prosthesis and of “blind” gastric section: LSG does not involve the placement of prosthesis and, unlike RYGBP, the whole gastric remnant is accessible by endoscopy.
Standardization in case of failure: unlike after failed RYGBP, the management of WR or IWL after LSG is standardized, consisting in the second-stage BPD-DS or in the conversion to RYGBP.
Disadvantages
Staple–line leak: its management could be very complex and long lasting, requiring a close cooperation with several different professional figures such as interventional radiologist, endoscopist, nutritionist, and infectious disease specialist. The timing of the treatment is a crucial prognostic factor; in fact, the earlier the diagnosis, the more conservative and successful is the management, avoiding difficult and often unsuccessful surgical reinterventions.
24.6 Outcome
In the Swedish Obesity Study (SOS) long term, prospective and controlled data from over 2,000 bariatric surgery patients and from over 2,000 matched control obese subjects were confronted. Primary end point was mortality rate in the two study groups, and secondary end points were incidence of cardiovascular diseases, diabetes, health-related quality of life, and biliary diseases [60].
At 10- and at 20-year follow-up, weight changes were −25 % in the operated patients and 0 % in the control group. At 16 years follow-up, the SOS study documented a significant reduction in the cumulative mortality of the operated patients (8 %) when compared to the control subjects (12.5 %). At the same time bariatric surgery induced dramatic positive effects on diabetes, cardiovascular diseases, cancer, and quality of life. In another clinical study on a very large sample (7,925 patients/group), the mortality in operated (RYGBP) patients was 40 % lower in confront to that of matched morbidly obese control subjects at a mean follow-up of 7.1 years [61]. The reduction was particularly pronounced for deaths from diabetes (−92 %) and cancer (60 %).
24.6.1 Patients’ Satisfaction
Patients’ satisfaction is of outmost importance in the correct evaluation of bariatric surgery results. Improved quality of life (QoL) and enhanced psychosocial functioning are important goals for bariatric surgery and should be actively pursued and evaluated. A crucial requirement to achieve patient’s satisfaction is a comprehensive and exhaustive informed consent. The patient must be fully aware of the effects and possible side effects of the procedure and, most important, the postoperative expectations must be realistic in order to avoid disappointment and depression.
Although lifestyle intervention programs improve health-related quality of life (HRQoL) in morbidly obese subjects, bariatric surgery has a greater impact on the well-being of these patients.
Surgery outcomes were significantly better in terms of both weight reduction and psychological adjustment compared to highly motivated participants in a weight loss program which included vigorous physical exercises, behavior modification, and nutritional advice [62].
Several papers analyzed the topic of QoL after AGB, LSG, and RYGBP, using Medical Outcomes Survey Short Form 36 (SF-36) and Bariatric Analysis and Reporting Outcome System (BAROS) [63]. HRQoL changes after bariatric surgery compare very favorably with those in obese control groups, irrespective for the specific type of surgery and in all the five dimensions of HRQoL (physical, mental, emotional, obesity symptoms, and symptoms distress).
In a study on 83 LSG patients, BAROS score was “good” to “excellent” in 90 % of patients. In SF-36, “physical functioning” and “general health perception” only had scores significantly better in patients with an EWL >50 % than in patients with EWL <50, suggesting that weight loss is not the only factor influencing the results [64].
Sexual health is a relevant component of patients’ satisfaction but it is infrequently well explored. In a recent prospective cohort study, 2 years following surgery (85 RYGBP and 21 AGB), women reported significant improvements in overall sexual functioning and specific domains of sexual functioning in addition to significant improvements in most domains of quality of life, as well as body image and depressive symptoms [65].
24.6.2 The Diabetes and the “Metabolic Surgery” Debate
The definition of “metabolic surgery” as “a set of gastrointestinal operations used with the intent to treat diabetes (diabetes surgery) and metabolic dysfunctions (which include obesity)” has been put forward by Rubino [66]. The term metabolic surgery describes surgical procedures that treat metabolic diseases, especially T2DM, with mechanisms that are, at least partly, independent by weight loss and it is comprehensive of conventional (AGB, RYGBP, SG, BPD) and of new procedures specifically aimed to determine metabolic effects: duodenojejunal bypass (DJB), mini-gastric bypass (MiniGBP), ileal interposition + SG etc.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
