ShortCuts-This Week in the Journals

September 2, 2008

392924423_860dafa0a4_m.jpgCommentary by Judith Brenner MD, Associate Editor, Clinical Correlations

In a week dominated by the politics of first, the Democratic National Convention, and later, Senator McCain’s selection of Sarah Palin as the VP nominee, reviewing all of this week’s breaking medical news was a welcome relief. I’ll say no more about politics since this is clearly not the right blog to air my opinions except to relay to you a term I learned this week that put it all together for me: zen koan. You’ll have to read on to find out the why and how of it and then, perhaps, to share your thoughts.

Let’s start with something straight forward: syphilis testing. As reported in the August 15th issue of MMWR, we learn that economics is driving a change in the traditional syphilis testing algorithm. Rather than starting with the nonspecific non-treponemal test as we’re all familiar, several labs are now starting with a temponemal test. MMWR reviews the experience of 4 NYC labs which now approach syphilis testing in this manner. They report that in 116,000 samples tested with an EIA test, 6% were positive. In this reversed algorithm, the 6,500 samples were then tested with the non-specific RPR. Of the 6,500 tested, 44% were positive. This represents a conundrum: what do we do with patients who have a negative RPR and a positive treponemal test? This is a group who, under traditional testing algorithms, would never have been identified. The best guess is that this represents old, treated syphilis. Though the lab approach in this group is not standardized, MMWR reports that the majority in this group went on for second testing with an alternate treponemal specific test and found that, of those tested, 85% were again found to be reactive; therefore RPR neg, tremponemal test positive times two. Thus, treatment is thought to be indicated. Bottom line: Realize that this new algorithm exists, though not uniformly. Exactly how to proceed when discovering one’s patient to be RPR negative and treponemal test positive is unclear at this time.  Side note: Treponemal tests detect antibodies to other treponemal species in addition to T. pallidum. T. pertenue causes yaws and T. carateum causes pinta….so, be on the look out.

Next, another straight forward report, this time on diabetes. (Fellow blog readers, I urge you to enjoy straight forward now because some of the following reports are anything but. ) David Nathan et al from the Joslin Diabetes Center reported in the August 31st issue of Diabetes Care, their study “Translating the A1C Assay into Estimated Average Glucose Values.” Their study essentially defines the mathematical relationship between A1C and average glucose. The obvious and practical need for this is obvious: patients understand “average glucose” much better than a number that tells them the “amount of hemoglobin that is glycosylated.” How many times have you found yourself explaining exactly what HgA1C means only to have a patient respond by telling you that their sugar never goes above 120? With Dr. Nathan’s study, you might realistically be able to tell a patient what their “AG” or “average glucose” is. The other pressing importance of this study is that a new, more stable and specific method for standardization of the A1C assay has been developed. Values are lower and units are different than our current A1C assays. Thus, rather than learn a new system for reporting a non-so-helpful number, why not convert it? In the study, approximately 500 patients had their glucose monitored by two methods: continuous glucose monitoring and sever-point daily self monitoring. Many, many glucoses were obtained on these 500 patients and a linear regression analysis was conducted. The table below is the result:

A1C %  AG* mg/dl
5 97 (76-120)
6 126 (100-152)
7 154 (123-185)
8 183 (147-217)
9 212 (170-249)

*AG=average glucose

The limitations of the study include the relative lack of representation of minorities, including Asian and African Americans. Nonetheless, it’s thought that the mathematical relationship is sound regardless of ethnicity and age. The importance of this is underscored by its inclusion in the most recent consensus statement on hemoglobin A1C measurement.

Surprise, surprise, the next report is about diabetes! I’ll keep this one short, but sweet 😉 As reported in this week’s JAMA, Wiener et al reported a meta-analysis “Benefits and Risks of Tight Glucose Control in Critically Ill Adults.” As a brief background, 2001 marked the start of tight glucose control in critical illness when Van Den Berghe et al published their landmark study demonstrating that tight glucose control reduced hospital mortality by one-third. This is a dramatic reduction and, as such, the next few years saw the incorporation of “tight glucose control” into guidelines for the care of the critically ill. As of 2008, this recommendation is endorsed by 16 professional societies, including, among others, the Surviving Sepsis Campaign guidelines, the Institue for Healthcare Improvement, the Volunteer Hospital Association, and the American Diabetes Association. However, the dark side of “tight glucose control” reared its ugly head, and the downside, namely hypoglycemia, has been reported on consistently as well. Hence, the meta-analysis by Wiener included 29 randomized controlled trials of 8,400 patients. The main conclusion: hospital mortality did not differ between tight glucose control and usual care. The mortality difference was 21.6% in tight control and 23.3% in usual care, with a confidence interval of 0.85-1.03. What to do with this result is a little unclear, but, in reading of the article and accompanying editorial, what is clear is that there is a lack of standardization on measuring and reporting glycemic control. Furthermore, a practice of glucose infusion practiced by some, but not by all, may have influenced the data.

The question becomes: what is a clinician to do? Simple. We await the results of the next trial: NICE SUGAR. Forty one hospitals in Australia, New Zealand, Canada and the Mayo Clinic will participate with early results expected in 2009. Not only is this a huge study (6100 patients), but blood glucose management is standardized in a web-based algorithm. Time will tell whether a mortality benefit is observed. If it is the case, then a web-based algorithm will have real potential to help achieve standardization in achieving tight control.

Finally, the New England Journal published two studies on line in an early release version on the subject of the role of anti-platelet therapy and the ARB, telmisartan, in reducing the risk of recurrent stroke. They are known by their acronym, PRoFESS. The studies are reported separately, but used the same population of 20,332 patients in a double blind randomized trial in a 2-by-2 factorial design. The agents studied were: Aspirin-extended release dipyridamole vs clopidogrel and telmisartan vs. placebo. The primary outcome was recurrent stroke and the secondary outcome was a composite measure of vascular events. In Sacco et al’s study “Aspirin and Extended-Release Dipyridamole (ERDP) versus Clopidogrel for Recurrent Stroke“, the 20,000 patients were randomized to receive aspirin 25 mg plus ERDP 200 mg twice daily vs. clopidogrel 75 mg once daily. The simple background of this article is that there are three anti-platelet contenders in the race for decreasing recurrence of non cardioembolic stroke: aspirin in varying doses, aspirin plus dipyridamole and clopidogrel. Going into this trial, the asa-dipyridamole combination appeared to be superior, but the current investigation was the first head-to-head trial and the results do not support either as being superior. Recurrent stroke occurred in 916 patients in the aspirin-ERDP group and in 898 in the clopidogrel group. This translates to no statistical difference between the two as measured by a hazard ratio of 1.01 with CI of 0/92-1.11. Among secondary outcomes, there was no benefit observed in either group as well. In terms of intracranial hemorrhage, more were noted in the ASA-ERDP group. Also considered in decision making are side effects, namely major hemorrhage.  The background of Yusuf et al’s report of “Telmisartan to Prevent Recurrent Storke and Cardiovascular events” is the knowledge that blood pressure is the strongest risk factor for stroke. There have been indirect observations made suggesting that suggested inhibition of the rennin-angiotensin system might be beneficial to patients with non cardioembolic stroke. However, as with the sister trial, the results do not show benefit. While the BP was lowered in the treatment group, 8.7% of the patients in the telmisartan group had a recurrent stroke (the primary outcome measure) and 9.2% in the placebo arm. The resultant hazard ratio of 0.95 with a CI of 0.86 to 1.04 is not significant. Similar results were observed in the secondary outcomes, a composite of major vascular events.

As a clinician, we’re left wondering what to do and so, I quote our editorialist in this aptly titled: “Stroke Prevention–Insights from Incoherence” as he tries to help the reader make sense of multiple large studies with conflicting results. He says that …”it might be likened to a zen koan–a statement inaccessible to rational understanding, the comtemplation of which may lead to a deeper realization.” For this blogger, this term has particular meaning in light of recent political developments. I won’t say more, but when I googled “Sarah Palin,”one of the first things I learned regarding her qualifications was that in 1984 she was crowned “Miss Wasilla” and went on to place second in the Miss Alaska pageant. I’m going to contemplate now.