The efficacy of the new drugs has only been tested using surrogate markers such as blood glucose and HbA1c. Cardiac benefits have been claimed [417], but so far none of these drugs has been tested in large-scale, long-term prospective studies on safety and efficacy such as the UKPDS using ultimate clinical endpoints or other patient-oriented outcomes.
For most clinicians insulin monotherapy is a second choice in type 2 diabetes mellitus, usually in the form of conventional insulin therapy. Since type 2 diabetes is associated with insulin resistance, high doses are usually necessary [418,419].
The insulin dose may be reduced by the addition of oral antidiabetic drugs [381,420,421]. Recently, intensified insulin treatment has been recommended [422]. However, the metabolic control achieved using different insulin regimens was comparable [423,424]. The most important adverse effects of insulin in the treatment of type 2 diabetes are hypoglycemia and weight gain. Weight gain may be reduced by combining insulin with metformin [133,425].
Clinical Aspects
Pharmacological therapy should be evidence-based, achieve the goals of therapy, be safe and causal rather than symptomatic, easy and comfortable for the patient, and inexpensive. Only two landmark studies on pharmacological therapy of type2 diabetes mellitus have used mortality and diabetes-related morbidity as ultimate endpoints. The Kumamoto study [131,426] evaluated the effects of intensive insulin therapy on prevention and progression of retinopathy, nephropathy, and neuropathy, and studied the cost-benefit relationship. The UKPDS [130,133,368] compared standard and intensive treatment with human insulin, chlorpropamide, glibenclamide and other sulfonylurea drugs, metformin, and acarbose. Three aggregate endpoints were used: (1) any diabetes-related endpoint, which includes various cardiac diseases, renal failure, amputations, and eye problems; (2) diabetes-related death; and (3) all-cause mortality. The UKPDS also analyzed the cost-benefit relationship and treatment of hypertension.
The main points that can be drawn from these studies are:
1. Intensive blood glucose lowering therapy with glibenclamide, metformin, or insulin (NB not with chlorpropamide [130]) delays the onset and progression of microvascular complications in type 2 diabetes.
2. The HbAlc lowering potency of these drugs was comparable.
3. Intensive treatment of obese type 2 diabetic subjects with metformin can also reduce the risk of macrovascular complications.
In none of the studies was diabetes management ideal. The goals of blood glucose control were rarely met, the rates of chronic complications of diabetes remained high, and adverse events were frequent during sulfonylurea and insulin treatment. The need for combination therapy was realized [125], but the addition of metformin to sulfonylurea had adverse effects. Other combinations were not systematically studied. New drugs could not be investigated. These shortcomings make it difficult to select the best therapy for each individual patient solely on the basis of these important studies. For this reason, the therapeutic options will now also be considered from the practical and patho-physiological points of view.
Intensive treatment with sulfonylurea or insulin has adverse effects. It increases the risk of hypoglycemia. In the past this used to be a minor problem, but it becomes a question of safety and quality of life if the goal of therapy is near-normoglycemia. The other adverse effect is weight gain. Most type 2 diabetic subjects are overweight and are advised to lose weight. The patients experience both hypoglycemia and weight gain as causing distress and impairing their quality of life [130]. Weight gain is a precipitating factor for diabetes and a macrovascular risk factor. Therefore, it is legitimate to ask whether the benefit of lowering HbA1c by insulin or glibenclamide is not jeopardized by the risk of this side effect. Therapy with either drug alone may not be a first choice.
Another concern is the fact that neither human insulin nor sulfonylurea corresponds with the pathogenetic defects of type 2 diabetes. They do not restore early insulin secretion because they are too sluggish; that is, the meal-related early phase of insulin action comes too late and the postprandial insulin action lasts too long.
Both types of drugs may overcome insulin resistance but they do not restore insulin sensitivity. If insulin sensitivity improves, this is not a direct drug effect but a nonspecific effect of lowering blood glucose and decreasing glucose toxicity. In the future, rather than insulin and sulfonylurea, rapid-and short-acting insulin analogues and glinides should be considered to initiate a more physiological insulin substitution, and metformin and glitazones to improve insulin sensitivity. However, the new drugs urgently need to be evaluated in controlled prospective studies using hard endpoints.
Intensive treatment with oral drugs or insulin over three years will lower HbA1c below 7% in only half of the patients. In the majority, combination therapy is indicated. Combinations of oral antidiabetic drugs have already been discussed. The combination of insulin with oral antidiabetic drugs may be even more important. The combination of insulin with sulfonylurea has been extensively studied [381,420,421,427,428]. It is convenient and may save some insulin. If postprandial hyperglycemia after breakfast is the main problem, the morning sulfonylurea dose may be replaced by regular or mixed insulin or short-acting insulin analogues. If fasting hyperglycemia is the main problem, the evening oral antidiabetic drug dose may be replaced by NPH insulin at bedtime. Metabolic control can be improved by various combinations without increasing the risk of hypoglycemia [423,424,427,428]. The combination of insulin with metformin is also possible. This combination offers the advantage of avoiding weight gain [133,425].
Fig. 1.3 Proposed stepwise treatment algorithm based on the pathophysiology of type 2 diabetes. For the rationale of this proposal, see text. The warnings with respect to possible unknown long-term side effects must be taken into account.(Modified after Matthaei et al. [395])