Monday, August 9, 2010

The Kingdom of Cockaigne

In a society accustomed to famine, people dream of food. In the Middle Ages, people dreamed of a fantasy world known to the English as the Kingdom of Cockaigne and to the Dutch as Luilekkerland (Fig. 1); there were similar legends in other European countries [1]. Physical work is unnecessary in Cockaigne, because food simply falls into your mouth. No medieval fabulist could possibly have dreamed that his descendants would one day actually come to live in the Kingdom of Cockaigne, otherwise known as the consumer society. Nor could our fabulist have imagined that these descendants might attempt to combat an excess of food by cutting out parts of their stomachs and intestines, and consider this a rational solution.

Simple obesity is a cultural and behavioural problem, strongly linked with the early environment [2] and social deprivation [3]. It is associated with social discrimination, emotional taboos and low self-esteem, while its psychological reverberations may affect life in all its aspects [4]. Last but not least, it is a prime example of allostasis, with forward-feeding behavioural and metabolic consequences that are notoriously intractable to therapy in the great majority of cases. How indeed do you treat the compulsion to put too much food into your own mouth, or those of your children?
Therapeutic failure leads to therapeutic mania, nowhere better exemplified than in the attempted treatment of obesity. Overweight people have always been willing to subject themselves to desperate remedies, and there has been no shortage of therapists, scrupulous or otherwise, willing to cater for this demand. Remedies have included an endless array of useless fad diets, enforced starvation in expensive sanatoria, wiring the jaws, surgical removal of fat, and selling the tapeworm scolex as a ‘natural’ remedy to be taken by mouth. Food consumption is one of our primal drives, and attempts to overcome this with pharmacotherapy have been a catalogue of failure, as witnessed by the recent withdrawal of rimonabant and sibutramine from the European market. Worse still, the pharmacology of obesity has been littered with therapeutic disasters, most notoriously Fen-Phen—a combination of fenfluramine and phentermine—which induced valvular lesions of the heart and fatal pulmonary hypertension [5, 6].
It has been said that there are no heroic surgeons, only heroic patients, and the history of bariatric surgery has had more than its share of these. Early attempts to shorten the small intestine were limited by severe malabsorption and associated complications; the 50% morbidity rate and approximate 10% mortality following jejunoileal bypass were ‘considered sufficient reasons to abandon it as an appropriate operation for the morbidly obese’ [7]. The modern era of bariatric surgery was heralded by the Roux-en-Y gastric bypass, first introduced in 1967 [8]. Even so, the stigma of obesity was so pervasive, and the complication rate so high, that surgery for obesity was considered faintly disreputable.
A few brave pioneers did persevere, however, and improved gastric restriction and bypass procedures coupled with minimally invasive surgery transformed the reputation of bariatric surgery. One result has been a free-for-all for surgical procedures with uncertain long-term consequences for health and well-being. Little seems to have changed since MacDonald concluded: ‘Despite accumulated knowledge from experience, no general agreement as to the optimal procedure has been reached, if one, indeed, does exist for any given morbidly obese patient. Analysis of results of surgical alternatives has always been hindered by a lack of comprehensive data collection, poor long-term patient follow-up, and lack of standardisation of the technical aspects of the procedures and of reporting of results.’ [7].

Benefits and motivation
The benefits of surgery are nonetheless compelling, and its recipients are among the most grateful patients you will encounter, so is there really cause for concern? Weight-reducing surgery does more than reduce weight. It offers psychological benefits [9] and reduces blood pressure, lipids and blood glucose [10, 11], and the need for these to be treated. Deaths from cardiovascular disease and cancer are reduced [12], and many of the secondary consequences of obesity, such as fatty liver [13], musculoskeletal disorders [14], intracranial hypertension [15], sleep apnoea [16] and infertility [17] are ameliorated. Not surprisingly, long-term health costs are also likely to fall [18], including those associated with diabetes [19, 20].
Nonetheless, there are concerns. To begin with, a mechanical remedy is hardly ideal for a condition whose root causes are genetic, social and psychological [21]. One author has vivid memories of a small celebratory meal offered to the first set of patients to undergo gastric banding at his own clinic, a meal during which one patient after another slipped away from the table to throw up in the toilet before coming back for more. There can also be a mismatch between the reasons why some people seek bariatric surgery and the reasons for which it is offered. A retrospective analysis found that candidates fell into two relatively distinct groups: those who claimed to seek gastric banding mainly for medical reasons (52%) and those who said they sought it for reasons primarily related to physical appearance or social embarrassment (32%) [22]. There were no men in the latter group. Similar opinions were expressed by patients seeking gastric bypasses or duodenal switches [23], suggesting that medical concerns predominate in stated motivation for bariatric surgery. These analyses have obvious limitations, not least sex and selection bias, their retrospective nature, and cultural taboos that might prompt men to deny concerns about their own appearance. Human motivation is complex, and we lack well-performed and adequately powered studies into why people seek and undergo surgery; studies conducted by investigators who are independent of the surgical teams involved. Some prospective candidates clearly do seek psychological and social benefit—one wanted to avoid the humiliation of having to call for an extension seatbelt whenever she got on a plane. This is a perfectly understandable motive with which we may all sympathise, but it seems profoundly irrational for her suitability for the procedure to be approved on the grounds of diabetes and hypertension, which caused her little concern. Rules governing access to bariatric surgery vary from one healthcare system to another, but bogus transactions such as this between doctor and patient do not form the basis of good medical care.
A further concern is the trend towards more invasive interventions, which may hold an irresistible appeal for overweight people desperate for a solution, and for surgical teams seeking to improve their outcome measures. Bias towards more invasive treatments exists even within well-regulated surgical settings. At the other end of the scale (and all too often ignored in the literature) are those who perform surgery for personal profit within a largely unregulated environment.
The American Diabetes Association and the National Institute for Health and Clinical Excellence have greeted bariatric surgery with evidence-based caution [24, 25]. Each concluded that it may be a useful treatment option for patients with type 2 diabetes and a BMI of ≥35 kg/m2 who remain in poor control despite adequate lifestyle change and pharmacological treatment, and both highlight the need for lifelong monitoring and for well-designed long-term controlled trials comparing bariatric surgery with non-surgical treatment. Neither saw a role for bariatric surgery in the treatment of diabetes in patients with a BMI of <35>30 kg/m2, a category so broad as to include some 50% of the total diabetic population [26]. Gastric bypass was considered the probable best option, despite the absence of controlled trial data or adequate consideration of the long-term metabolic consequences of this procedure in this group of patients. The extension of surgery beyond its evidence base, otherwise known as ‘indication creep’, thus seems to offer a temptation that many are unable to resist.

Surgery and diabetes
Although psychosocial or cosmetic considerations may loom large in the thinking of those who seek surgery, health concerns generally provide the final stimulus for the patient and the formal justification for the surgeon. Let us therefore consider the risks and benefits of surgery in relation to diabetes. Food consumption in excess of energy requirements underlies the current pandemic [27], and enforced starvation in time of war produced a dramatic reduction in the incidence and mortality of diabetes [28, 29]. Calorie deprivation, whether voluntary or enforced, forms the basis of effective diet regimens and almost every other form of successful weight-reducing therapy. No surprise, therefore, that surgically induced starvation by malabsorption or gastric restriction has proved highly effective in lowering blood glucose and reducing the need for other therapies.
A recent meta-analysis from Buchwald et al. suggests that 78% of patients with type 2 diabetes achieve biochemical ‘remission’ following bariatric surgery [30], but the accompanying commentary emphasised the retrospective and uncontrolled nature of virtually all these data and the heterogeneity between the procedures analysed, some of which would now be considered obsolescent. Ascertainment and follow-up was unsatisfactory in most studies, and a mere 1.6% of the material evaluated qualified as grade 1 evidence [31]. Patient selection also favoured a positive outcome, as many of the patients were young (the mean age was just 40 years), fit and female and were thus unrepresentative of the diabetic population as a whole.
It is remarkable that only one properly designed prospective randomised controlled study of bariatric surgery has ever been conducted with a specific focus on people with diabetes [32]. This reported short-term outcomes with laparoscopic gastric banding vs standard medical treatment in a group of patients with recent-onset type 2 diabetes and an average BMI of 38 kg/m2. After 2 years, 73% of those in the surgical group and 13% of those in the medically treated group were in treatment-free glycaemic remission. Gastric banding also enhanced short-term quality of life as compared with conventional medical treatment for type 2 diabetes [33]. No other relevant controlled trials have been published.

Which is the better operation?
While it remains a matter of opinion which operation is better, some observations can be made. Gastric and intestinal bypass procedures rapidly enhance insulin secretion, probably through short-term effects on gut peptide hormones [34], whereas metabolic improvements after gastric banding are less immediate and may depend more directly on weight loss. Although various authorities have long debated the relative merits of one procedure or another, long-term outcomes have yet to be compared for those with diabetes. As surgeons tend to compete in terms of short-term weight reduction and other surrogate measures, gastric bypass surgery is often the preferred option, despite evidence that gastric banding—a safer and potentially reversible procedure—produces similar weight loss beyond 2 years [35] and may be almost as good at controlling diabetes [32, 36]. Weight loss may indeed be an unreliable surrogate for improved glucose control in diabetes [36], as rapid weight loss also results in a proportionately greater loss of lean tissue [37].
In principle it may seem obvious that safer, reversible procedures should be preferred to more complex irreversible procedures with a wider range of complications, except where the latter have very clear advantages. As many different types of surgical intervention are now on offer, such comparisons are urgently needed. These should be based on longer-term medical outcomes and quality of life, rather than kilograms shed, and the rate and proportion of patients for whom glucose-lowering therapy can be stopped. Short-term withdrawal of hypoglycaemic medication is a naive and inadequate endpoint for surgical studies.
A further concern is that little has been done to establish the most appropriate type of intervention for patients in different age groups and at different stages of diabetes. It is far from clear, for example, that the restoration of euglycaemia by bariatric surgery will benefit older individuals following years of exposure to poor glucose control. Conversely, when bariatric surgery is offered to young people with decades of life in front of them, its longer-term metabolic and nutritional consequences surely demand more active consideration. Uncertainties such as these persist because we lack rigorous, controlled studies. Observational reports of experimental procedures in lean people with diabetes [38] have done little to advance our understanding, other than to suggest that this approach has little to offer.
Bariatric surgery does not ‘cure’ diabetes. A relevant comparison might be with gestational diabetes, in which the escalating insulin demands of pregnancy exceed the capacity of the pancreatic islets to respond, yet normoglycaemia is restored following delivery—a ‘remission’ that, as we know, is often transient. This is consistent with the observation that hyperglycaemia is most effectively controlled by surgery in those with disease of a shorter duration, whereas those with a longer duration of diabetes or higher treatment requirements derive less benefit [32, 39]. In agreement with this, a recent retrospective report highlights the extent to which long-term glycaemic control may eventually deteriorate after gastric bypass [40]. Surgery may be a useful way of buying time, but it should not be regarded as a permanent cure for hyperglycaemia.
Even those in whom diabetes treatment can be reduced or withdrawn still require long-term surveillance of glucose, lipids and blood pressure, not to mention the potential adverse consequences of surgery. The Swedish Obese Subjects (SOS) study included 34 patients who had undergone gastric bypass in the 10 year analysis, and in these systolic blood pressure was reduced by 4.7%, diastolic blood pressure by 10.4% and cholesterol by 12.6% [11]. The 10 year incidences of hypertension and hypercholesterolaemia were unaffected by bariatric surgery [11]. Blood-pressure lowering therapies [41] and statins [42] still provide the largest and best-documented reduction in cardiovascular risk in those with diabetes.

Potential problems of bariatric surgery
Bariatric surgery has come a very long way, and is far safer than in previous years. It is not, however, free from surgical risk (see text box: Potential complications of bariatric surgery). The short-term operative mortality for low-risk patients attending centres with experienced surgical teams is around 1/200 after gastric bypass, as against 1/2,000 for laparoscopic gastric banding [35]. One meta-analysis revealed a 30 day mortality of 0.1% after gastric banding, 0.5% after gastric bypass and 1.1% after biliopancreatic diversion [43]. The authors commented that these risks compare favourably with other forms of surgery, but the question here is whether such risks are justified in the treatment of diabetes. The main short-term causes of death after bariatric surgery are venous thromboembolism and cardiorespiratory disease [44]. Venous thromboembolism affected 0.3% of banding patients and 0.4% of laparoscopic gastric bypass patients in the Longitudinal Assessment of Bariatric Surgery (LABS) [45]. All these risks are increased in older people [46] and in those with diabetes or hypertension [44]. Varela et al. found that hospital mortality rose to 4.7% in a subset of patients aged over 60 with pre-existing cardiac disease [47]. Gastric bypass not only carries a greater risk of death than banding, but is also associated with a greater risk of early postoperative complications and re-admission to hospital [45]. One study reported a 20.6% rate of early re-admission or emergency visit following this form of surgery [48]. Revisional surgery is probably required in at least 5% of patients [49].

The fact that modern bariatric surgery is frequently carried out laparoscopically does nothing to eliminate potentially serious long-term nutritional, medical and psychological consequences (see text box: Long-term nutritional, medical and psychological complications of bariatric surgery). The upper part of the intestine fulfils an essential nutritional function, and gastric and intestinal bypass procedures result in micronutrient depletion [50–52] that requires scrupulous monitoring and lifelong replacement [53]. Here again, systematic data are lacking and the scale of the problem is likely to have been underestimated. The most common concern is anaemia resulting from dietary inadequacy, malabsorption of iron, folic acid, vitamin B12 and vitamin C, and sometimes blood loss from surgical complications. Deficiencies of vitamin D and calcium are also common, with a substantial early decline in bone mineral density [54]; the long-term implications of this are uncertain. Less commonly, peripheral neuropathy and occasional central nervous system damage result from deficiencies of vitamin B12, thiamin or copper or from Guillain–Barré syndrome. A systematic review identified 104 reports of Wernicke’s encephalopathy after bariatric surgery, suggesting that thiamin deficiency is not rare [55]. More unusually, malabsorptive bariatric surgery has been associated with depletion of vitamin A, vitamin K, niacin, zinc and selenium [53], while drug malabsorption is another potential concern [56]. Furthermore, while bariatric surgery generally improves fertility and pregnancy outcomes, intrauterine growth retardation, fetal brain haemorrhage and neural tube defects have been described [57, 58]. Most of these nutritional deficiencies are indeed avoidable, but the limiting factors in their prevention and treatment are human rather than medical, and relate to treatment non-adherence or inadequate specialist follow-up.