n engl j med 362;16 nejm.org april 22, 2010 1533

e d i t o r i a l s

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

Navigating the Choices for Diabetes PreventionDavid M. Nathan, M.D.

The global epidemic of type 2 diabetes has prompt-ed a large number of clinical trials aimed at re-ducing its incidence.1 Not surprisingly, addressing the underlying lifestyle behaviors — overeating and inactivity — that result in obesity, the pri-mary cause of the epidemic, has had a major and consistent effect in reducing the cumulative inci-dence of diabetes. In addition, lifestyle interven-tions have reduced the cardiovascular risk fac-tors that typically accompany the prediabetic and diabetic states.2,3 Several drugs that are used to treat diabetes have also been studied as a means of reducing the incidence of the development of diabetes. Metformin, the alpha-glucosidase inhibitor acarbose, and the thiazolidinedione rosiglitazone have been shown to reduce the in-cidence of diabetes to variable degrees.1 The Dia-betes Prevention Program (ClinicalTrials.gov number, NCT00004992) and its long-term out-come study (NCT00038727), which are the only studies that can address comparative effective-ness since they included random assignment to lifestyle modification, a medication, and placebo, showed that lifestyle intervention had a substan-tially more powerful effect than did metformin when each was compared with placebo (a reduc-tion in incident diabetes with lifestyle intervention of 58% at 3 years and 34% after 10 years vs. a reduction with metformin of 31% at 3 years and 18% after 10 years).3

Although a reduction in the incidence of dia-betes is important, the major public health im-pact of prevention studies will be determined by whether the prevention of diabetes — or a delay in the development of the disease — will trans-late into a reduction in the diabetes-specific long-term complications affecting the eye, kid-ney, and nervous system and will ameliorate the

less specific cardiovascular disease that is the major cause of death in patients with type 2 dia-betes. Safe, inexpensive, and acceptable inter-ventions that lower glycemia, halt or delay the progression to diabetes, and reduce cardiovas-cular disease are highly desirable.

In this issue of the Journal, the investigators of the Nateglinide and Valsartan in Impaired Glu-cose Tolerance Outcomes Research (NAVIGATOR) study report the results of their international trial that examined the effects of the approved diabetes medication nateglinide, a relatively weak, rapid-acting, sulfonylurea-like drug, and the an-giotensin-receptor blocker valsartan on the de-velopment of diabetes and cardiovascular disease in a high-risk population.4,5 Although the study had an efficient factorial design, the results with respect to nateglinide and valsartan are reported separately in the Journal, as if there were two par-allel trials, since the authors report no interactions between the treatments. This presentation belies one of the stated goals of NAVIGATOR — specifi-cally, to study the combined effects of the two study drugs.6 These effects are reported only brief-ly in Supplementary Appendix 1 for each article.

The rationale for studying nateglinide is that the meglitinide class of hypoglycemic medica-tions, of which nateglinide is a member, has been shown to lower postprandial glycemia, a puta-tive treatment target, since rising postprandial glucose is the most common route to the devel-opment of diabetes.7 Moreover, postprandial hy-perglycemia (or hyperglycemia after an oral glu-cose-tolerance test) has been shown to be more closely associated with the risk of cardiovascular disease than are fasting glucose levels.8 There has been widespread interest in determining whether a causal relationship exists between el-

Downloaded from www.nejm.org on June 15, 2010 . Copyright © 2010 Massachusetts Medical Society. All rights reserved.

T h e n e w e ngl a nd j o u r na l o f m e dic i n e

n engl j med 362;16 nejm.org april 22, 20101534

evated postprandial glycemia and cardiovascular disease, and it has been recommended that tri-als be performed to test this hypothesis.9 Final-ly, owing to its rapid-action profile, nateglinide has a relatively low risk of causing hypoglyce-mia. The rationale behind the choice of valsartan to inhibit the renin–angiotensin axis is less clear, other than the fact that both nateglinide and valsartan are manufactured by the pharma-ceutical sponsor, which also designed the study. Secondary analyses of clinical trials have sug-gested that angiotensin-converting–enzyme (ACE) inhibitors and angiotensin-receptor blockers (ARBs) are associated with a reduced incidence of diabetes; however, the development of diabe-tes was not measured uniformly, and the only clinical trial that directly examined whether ACE inhibition would prevent diabetes failed to show that diabetes was prevented or that there were any beneficial effects on insulin resistance or beta-cell function.10,11

The results of the NAVIGATOR study are large-ly negative. Neither drug (nor the combination, keeping in mind that one quarter of the study co-hort took both drugs) reduced the two coprimary cardiovascular-disease outcomes. The only posi-tive result was a weak, albeit statistically signifi-cant, reduction in the incidence of diabetes with valsartan.5 The relative reduction of 14% and the absolute reduction of 3.7 percentage points in incident diabetes with valsartan, as compared with placebo, over a mean follow-up of 5 years make valsartan the weakest of the drugs studied to date.1

Unfortunately, the NAVIGATOR study has not definitively answered whether lowering postpran-dial glycemia reduces the incidence of cardiovas-cular disease or diabetes, since the mean glucose levels 2 hours after a glucose challenge in the annual oral glucose-tolerance tests were higher in the nateglinide group than in the placebo group. The authors describe this paradoxical finding as a rebound effect, since nateglinide was not ad-ministered on the mornings that the oral glucose-tolerance tests were performed, suggesting that on the 364 other days of the year, when the admin-istration of nateglinide was not delayed, postpran-dial glucose levels were lower in the nateglinide group. However, there are no direct data to sup-port this contention, and no data on glycated he-moglobin are presented, other than in the sub-group that progressed to diabetes, to support a

substantial lowering of overall glycemic levels with nateglinide. The other factor that might have miti-gated any putative salutary effect of nateglinide on the development of diabetes or cardiovascu-lar disease was the provision of a lifestyle inter-vention program for all participants. Consider-ing the powerful effects of lifestyle interventions in other trials,1 this design feature may have re-duced the magnitude of the potential benefit from nateglinide. This explanation is not persua-sive. The lifestyle intervention program was not effectively implemented, as evidenced by the trivial weight loss over the course of the study and the 8% annual incidence of diabetes in the placebo group, which was close to the rates found in control groups from other studies.

The finding that valsartan failed to have an effect on either of the cardiovascular-disease out-comes but had a positive effect on the incidence of diabetes is surprising. Previous studies of ACE inhibitors and ARBs have suggested that these drugs have a beneficial effect on cardiovascular disease in patients with diabetes, and the low-er blood pressure achieved would be expected to result in a reduction in cardiovascular disease. In the NAVIGATOR study, the high rates of loss to follow-up (13%), use of off-study ACE inhibi-tors or ARBs among participants assigned to pla-cebo (24%), and nonadherence to valsartan (34% by study end) could explain the absence of an effect on cardiovascular disease.

The results from the NAVIGATOR study do not support the contention that reducing post-prandial hyperglycemia has a specific role in preventing diabetes or reducing cardiovascular disease. Other than increasing the rate of hypo-glycemia by a factor of two, nateglinide had lit-tle effect. Although the authors suggest that the prevention of diabetes with valsartan might make it a preferred drug as compared with antihyper-tensive drugs that potentially worsen glycemia, valsartan was relatively weak in preventing dia-betes, and it did not lower the rates of cardio-vascular disease. The prevention of diabetes re-mains a critical public health priority, but for now we should steer away from these two drugs and use effective lifestyle interventions and, in select-ed persons, metformin to combat the epidemic.

Disclosure forms provided by the author are available with the full text of this article at NEJM.org.

From the Diabetes Center, Massachusetts General Hospital, Harvard Medical School, Boston.

Downloaded from www.nejm.org on June 15, 2010 . Copyright © 2010 Massachusetts Medical Society. All rights reserved.

editorials

n engl j med 362;16 nejm.org april 22, 2010 1535

This article (10.1056/NEJMe1002322) was published on March 14, 2010, at NEJM.org.

Crandall JP, Knowler WC, Kahn SF, et al. The prevention of 1. type 2 diabetes. Nat Clin Pract Endocrinol Metab 2008;4:382-93.

Goldberg RB, Temprosa M, Haffner S, et al. Effect of pro-2. gression from impaired glucose tolerance to diabetes on cardio-vascular risk factors and its amelioration by lifestyle and met-formin: the Diabetes Prevention Program randomized trial by the Diabetes Prevention Program Research Group. Diabetes Care 2009;32:726-32.

Diabetes Prevention Program Research Group. 10-Year fol-3. low-up of diabetes incidence and weight loss in the Diabetes Prevention Program Outcomes Study. Lancet 2009;374:1677-86. [Erratum, Lancet 2009;374:2054.]

The NAVIGATOR Study Group. Effect of nateglinide on the 4. incidence of diabetes and cardiovascular events. N Engl J Med 2010;362:1463-76.

Idem.5. Effect of valsartan on the incidence of diabetes and cardiovascular events. N Engl J Med 2010;362:1477-90.

Califf RM, Boolell M, Haffner SM, et al. Prevention of diabe-6.

tes and cardiovascular disease in patients with impaired glucose tolerance: rationale and design of the Nateglinide And Valsartan in Impaired Glucose Tolerance Outcomes Research (NAVIGATOR) Trial. Am Heart J 2008;156:623-32.

Meigs JB, Muller DC, Nathan DM, Blake DR, Andres R. The 7. natural history of progression from normal glucose tolerance to type 2 diabetes in the Baltimore Longitudinal Study of Aging. Diabetes 2003;52:1475-84.

Meigs JB, Nathan DM, D’Agostino RB, Wilson PW. Fasting 8. and postchallenge glycemia and cardiovascular disease risk: the Framingham Offspring Study. Diabetes Care 2002;25:1845-50.

Nathan DM, Davidson MB, DeFronzo RA, et al. Impaired 9. fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care 2007;30:753-9.

The DREAM Trial Investigators. Effect of ramipril on the 10. incidence of diabetes. N Engl J Med 2006;355:1551-62.

Hanley AJ, Zinman B, Sheridan P, Yusuf S, Gerstein HC. Ef-11. fect of rosiglitazone and ramipril on β-cell function in people with impaired glucose tolerance or impaired fasting glucose. Diabetes Care 2010;33:608-13.Copyright © 2010 Massachusetts Medical Society.

Improving the Treatment of Necrotizing Pancreatitis — A Step Up

Andrew L. Warshaw, M.D.

Acute pancreatitis in most patients is self-limited and resolves without complications or the need for invasive procedures or surgical intervention. In a minority of patients, perhaps 10%, necrosis of the pancreatic and peripancreatic tissues opens the door to multiple organ failure, infection of the necrotic tissue, or both, with a greatly increased risk of death; this risk has been estimated to be 20 to 30%. Recent reports citing a substantial reduction in the rate of death, ranging as low as 4%,1,2 attribute the improvement to better inten-sive care; avoidance of early intervention to allow resuscitation, stabilization, and demarcation and liquefaction of the damaged areas2,3; and a variety of innovations in drainage and evacuation of fluid and devitalized tissues. The established techniques of laparotomy and open débridement1,2,4 are being challenged by “minimally invasive” endoscopic necrosectomy through transgastric, laparoscopic, and retroperitoneal routes.5 The goals in general are control of infection, maximal evacuation of devitalized tissues that are the culture medium for invasive infection, and promotion of conditions for healing.

The indications for surgical intervention cer-tainly do not include the mere presence of asymp-tomatic necrosis, even if it is substantial.6 Many clinicians argue that infection is the sine qua non,7

but some patients with sterile necrosis nonethe-less have a major systemic inflammatory response and multiple organ failure,2,6 and these patients can benefit from débridement. Other patients have a low-grade inflammatory state or they are “per-sistently unwell” for months,1,2 and as many as 40% of these patients may have unsuspected in-fection of the necrotic tissue.2

There is now a consensus that the best out-comes from intervention are achieved when dé-bridement is delayed until approximately 4 weeks after the onset of pancreatitis, when the damaged area has been walled off and liquefaction has begun.1-3 If necessary, acute sepsis can be prelimi-narily controlled by percutaneous drainage as a bridge to the later evacuation of detritus.5,8 Al-though percutaneous drainage alone has been as-sociated with some success in the treatment of pancreatic “abscesses,” the success rate in defini-tively treating infected necrotic tissue without the need for débridement is only 30 to 35%.8,9

In this issue of the Journal, van Santvoort et al.9 report the results of a randomized trial compar-ing treatment of pancreatic and peripancreatic necrosis by open laparotomy with a hybrid, or “step-up,” approach in which percutaneous drain-age was the first step, with débridement by means of a less invasive video-assisted retroperitoneal

Downloaded from www.nejm.org on June 15, 2010 . Copyright © 2010 Massachusetts Medical Society. All rights reserved.