Neurology: May 2010 Archives

In the spirit of diagnostic parsimony, I would not attribute Bret Michaels's recent "warning stroke" or transient ischemic attack (TIA) to a patent foramen ovale (PFO)*—which is a common, and often incidental, finding in the work-up of stroke. It should be noted that my opinion is contrary to that of Michaels's treating MD in Arizona, at least as it is reported by the media. It should also be stressed that Michaels's treating physician knows a helluva lot more about Michaels's condition(s) than I do.

Nevertheless my educated, but armchair, inclination is to ascribe Michaels's TIA to delayed vasospasm due to his subarachnoid hemorrhage (SAH), the cause of which was reportedly never identified. While vasospasm typically occurs 3-5 days after the initial SAH bleed, the risk can continue for several weeks thereafter. The herald symptom of Michaels's SAH, an explosive headache, began on April 22nd. Vasospasm 1 month later is not outside the realm of possibility and more likely, in my mind, than a completely separate process of an embolic TIA due to a PFO.

According to CNN, the presumed embolic TIA of the "Celebrity Apprentice" winner is now being treated with a daily injection of Lovenox (enoxaparin; sanofi-aventis US), an anticoagulant. This is tricky business given that such treatment increases Michaels's risk of bleeding generally, including bleeding into the brain.

Data regarding the follow-up risk of stroke or TIA with PFO and the potential benefit of surgical closure in asymptomatic PFO were discussed here last summer.

* The vestige congenital hole between the left and right atria.

Image from Wikipedia: Horizontal CT cut showing hyperdense subarachnoid blood in the basal cistern.

Monday the Drug and Device Law Blog cited a new lawsuit in the "Judicial Hellhole," Madison County, Illinois (and you know how the DDLB feels about Judicial Hellholes). The suit, covered by the Madison and St. Clair Record* and filed by 19 plaintiffs "from across the country," alleges that Novartis, CIBA, and Sandoz failed to warn African Americans of their genetically increased risk of Stevens Johnson syndrome (SJS), toxic epidermal necrolysis (TEN), and other severe skin disorders during carbamazepine therapy (brand name, Tegretol; abbreviated CBZ). The plaintiffs are represented by Christopher Cueto of neighboring St. Clair county, the brother of controversial circuit court judge Lloyd A. Cueto and the disbarred Amiel Cueto.

In clumsy language, the suit stipulates, according to the Record:

"Defendants knew or should have known a significant portion of deaths and severe side effects, described herein, resulted from carbamazepine products have included African American persons and children who were found to have increased risks of carbamazepine SJS and TEN in the medical literature..."

And,

"Different races, such as African Americans, Hispanics, native Indians, Asians and Caucasians, face an increased risk of developing SJS and TEN because of their genetics. African Americans and Asians remain especially prone to developing the life-threatening skin conditions because of a gene they have a tendency to carry."

The paper also outlines the long, "storied" history of Tegretol's approval, which is evidently solely based on what's written in the complaint.

Problem is: While there are compelling data linking CBZ-induced SJS and TEN in southeast Asians and the presence of the HLA-B*1502 genetic allele, to my knowledge, there are no comparable data in African Americans, persons of African descent, or black Africans. Nor are there comparable data for Hispanics or Native Americans. In fact, published data indicate a negligible or extrememly low rate of the HLA-B*1502 allele in these non-Asian ethnic groups.

The HLA-B*1502 link was first made in 2004 in the journal Nature. Taiwanese investigators showed an astronomical association between the allele and CBZ-related SJS and TEN in Han Chinese (see also, Hung et al). The link was also subsequently found in a Thai population and, to a much lesser extent, in East Indians. The HLA-B*1502 allele, however, does not appear to be a risk factor for SJS/TEN in Caucasians.**  

In other words, the allele is neither necessary nor sufficient for the development of CBZ-related skin disorders: meaning that persons who don't carry the allele can develop SJS and TEN, and persons with the allele don't always develop drug-induced skin disorders. Also an association between HLA-B*1502 and severe skin disorders with other anticonvulsants (eg, oxcarbazepine, phenytoin) was recently reported in Han Chinese.

In December 2007, the FDA issued an alert, informing physicians of the link between HLA-B*1502 and CBZ-induced SJS or TEN. The agency reported, "This allele occurs almost exclusively in patients with ancestry across broad areas of Asia, including South Asian Indians." Genetic testing was advised in "[p]atients with ancestry from areas in which HLA-B*1502 is present," before starting treatment with CBZ. However, "[p]atients who have been taking carbamazepine for more than a few months without developing skin reactions are at low risk of these events ever developing from carbamazepine. This is true for patients of any ethnicity or genotype, including patients positive for HLA-B*1502," the FDA advised. Labeling for CBZ products was accordingly changed with black-box warnings.

Also according to the FDA in 2007, the frequency of the HLA-B*1502 allele in ethnic populations is as follows.

• Chinese, Thai, Malaysians, Indonesians, Filipinos, Taiwanese: 10%-15%
• Other southeast Asians, including Indians: 2%-4%
• Japanese, Koreans: <1%

In 2008, Ferrel and McLeod updated the frequencies of the HLA-B*1502 allele in various Asian and North American populations. Attorney Cueto is evidently hanging his hat on a 0.2% allele rate in African Americans, which hardly constitutes a "tendency." (One at least hopes that the plaintiffs have been tested for the allele.) Curiously enough, Native Americans, who are most likely to have at least a remote Asian ancestry, show a negligible frequency.

Continent	Population/Ethnicity	Allele Frequency (%)	N
Asia	Singapore	11.6	86
	Han Chinese	10.2	572
	Malay	8.4	101
	Thai	6.1	99
	Filipino	5.3	94
	India Khandesh Pawra	6	50
	India North Hindi	2	91
	India Mumbai Marathas	1	72
	Korean	0.5	200
North America	Asian	5.1	396
	African	0.2	251
	European	0	287
	Hispanic	0	240
	Native American	0	235

* Which, by the way, reports that "Pharmaceuticals failed to warn African Americans of drug's dangers" (italics added).

** Nor does the allele increase the risk of maculopapular exanthema in Han Chinese.

Image of bottom-feeding channel catfish from Wikipedia.

While some news sources are claiming that the results from WHO's large INTERPHONE study (yet another study assessing the risk of brain tumors with cell-phone use) are "inconclusive," a review of the published article, in the International Journal of Epidemiology, generally reveals otherwise. (A press release for the article can also be found here.)

Among dozens of case-control analyses in the study, many of which involved small numbers, few demonstrated an increased odds of meningioma or glioma. Moreover, in the vast majority of cases, the odds of these brain tumors were actually lower with cell-phone use.* On the basis of their results, the authors concluded that there is no increased risk of meningioma among cell-phone users.

When considering glioma, however, the odds were increased among users whose cumulative call time exceeded 1639 hours (OR, 1.40; CI: 1.03, 1.89), regardless of how long they had been using cell phones (1-4 years, 5-9 years, or 10+ years).** The highest cumulative call time was also associated with an increased odds of glioma in the temporal lobe (OR, 1.87; CI: 1.09, 3.22) and on the same side as typical cell-phone use (OR, 1.96; CI: 1.22, 3.16). "Still, the evidence for an increased risk of glioma among the highest users was inconclusive," the authors wrote, citing several possible sources of study biases, including selection bias. (Unfortunately several news sources evidently grabbed the word "inconclusive" for their provocative ledes.)

The INTERPHONE study was a multinational study, involving 16 centers in 13 countries (Australia, Canada, Denmark, Finland, France, Germany, Israel, Italy, Japan, New Zealand, Norway, Sweden, and the United Kingdom). Between 2000 and 2004, data were collected from people aged 30-59 years, given that these individuals were likely to have a history of significant cell-phone use. The data were compared with information from matched control populations (who, evidently, didn't use cell phones...ever...! [Addendum "Oops," see below for a clarification of the control pop). The main analyses included 2409 meningioma cases (2662 matched controls) and 2708 glioma cases (2972 matched controls).

The rationale for studying a possible link between cell-phone use and brain-tumor risk is based on the proposal that low-level exposure to radiofrequency (RF) magnetic fields from cell phones increases the risk of brain tumors that absorb this energy. While cell-phone use has increased dramatically since the time of the study, any risks associated with increased cell-phone use are probably mitigated by lower RF emissions, on average, from newer phones and the prevalent use of texting and hands-free devices (which keeps the RF-emitting cell phone away from the head).

* Use of hands-free devices was excluded.

** However, the confidence interval for the 1-4-year odds ratio was very wide. 

Addendum 10/20/37: The reference or control group, defined as "never regular users," included those people with meningioma (32% of cases) or glioma (26%) who never had more than 1 call per week, on average, for 56 months. Never users (11% of meningioma cases and 9% of glioma cases) were, in fact, never users.

Yesterday Orac provided another blog-based review of this study, citing the "inconclusive" moniker as accurate. I disagree, given the overall results of the study (and the extent to which any negative study can ever be called conclusive). The only hedge is provided by the high-end cell-phone users for glioma. But even Orac writes, "There was no compelling evidence of an association between cell phone use and either of these cancers."

Orac also poo-poos the RF rationale for assessing a link between cell-phone use and brain cancer, at least on the basis of DNA breakage. This appears to be an important point to stress (despite the fact that the study was conducted in the first place!).

The INTERPHONE study authors write,

"Much biological research has been done in recent years on possible biological effects of RF fields. This work covers in vitro and in vivo exposure, alone and in combination with other physical or chemical agents, and has found no evidence that RF fields are carcinogenic in laboratory rodents or cause DNA damage in cells in culture. Possible effects of RF fields on other biological endpoints are still being explored."

The authors cite a white paper from the European Commission, "Health Effects of Exposure to EMF," which summarizes,

"It is concluded from three independent lines of evidence (epidemiological, animal and in vitro studies) that exposure to RF fields is unlikely to lead to an increase in cancer in humans. However, as the widespread duration of exposure of humans to RF fields from mobile phones is shorter than the induction time of some cancers, further studies are required to identify whether considerably longer-term (well beyond ten years) human exposure to such phones might pose some cancer risk."

I have a hard time believing that cosmetic doses* of Botox caused the multitude of neurologic and other problems claimed by former OB/GYN Sharla Helton, 48. But yesterday, an Oklahoma jury decided otherwise and granted Helton $15 million, according to the Orange County Register (a paper diligently covering alleged Botox-injury cases). Allergan will reportedly appeal the verdict. (Coverage of the verdict is also provided by NewsOK.com, which shows a photo of a nearly wrinkle-less, smiling Sharla Helton in 2006.)

Helton, who claims to have experienced double vision, respiratory problems, and persistent limb pain from injected Botox, was represented by Ray Chester of the Texas-based firm McGinnis, Lochridge, and Kilgore. Chester also represented the mother of Kristen Spears, who lost her wrongful-death suit against Allergan in March. (For background on this story, start here.) In April, the OC Register reported that Allergan was suing Spears's mother for legal fees.

The Oklahoma jury evidently decided in favor of Allergan with respect to Helton's product liability claim, however. According to a quoted Allergan spokesperson, the jury concluded that "Botox Cosmetic was not a defective product and did not have defective warnings." Consequently "it is not possible for Allergan to have been found negligent in this case," the company reasoned. Punitive damages were apparently not awarded by the jury.

During the 3-week Oklahoma trial, the plaintiff's expert witness, toxicologist Shayne Gad, PhD, was publicly humiliated when it was revealed by Allergan's counsel that Gad had misstated his military record. Gad was convicted of a misdemeanor for falsely claiming that he had been awarded a number of medals, including 3 Purple Hearts. In a plea bargain with the US District Court for eastern North Carolina, Gad avoided prison time and was placed on an 8-month probation, from February to April of last year. A dramatic courtroom apology from Gad to Chester and the plaintiff was printed by the OC Register last month.

Chester's next case (now 3 of 15, if I'm counting correctly) is that of Sondra Bryant, a 70-year-old nurse who died in 2008 after receiving Botox injections for neck pain. The nurse was from Texas, but the trial will begin in Santa Ana, California, in the fall.

Allergan's free-speech case against the government remains pending, but today's search of the court calendar still reveals nothing on the schedule for the remainder of the year.

* Even at somewhat-higher-than-approved doses. According to the OC Register, Helton received a 50-unit treatment for wrinkles in her upper face. The FDA approved total treatment dose is 20 units. It was not reported if Helton was treated by another physician or treated herself.

Bret Michaels's favorable outcome with subarachnoid hemorrhage (SAH) is one thing. But are we better at treating SAH now than 25 years ago?

The happy answer, yes, is suggested by a newly published time-trend study in the journal Neurology.

By using data from 2 vascular-study populations (OSCP and OXVASC), British investigators examined the outcomes of patients who developed aneurysmal SAH in the respective time periods of 1981-1986 and 2002-2008. They found no temporal differences in the incidence or severity of SAH, but a significantly greater proportion of patients surviving to the hospital in the later cohort underwent imaging (67% vs 23%) and aneurysmal treatment (50% vs 19%). The median delay to any interventional treatment was also significantly shorter in the later cohort (2 vs 14 days). Among treated aneurysms, all 5 in the earlier cohort were secured by clips; in the latter cohort, 83% were treated with endovascular embolization.

Age- and sex-adjusted 30-day case fatality tended to be lower in the later cohort (43% vs 67%); but this difference was not statistically significant. Among those patients reaching the hospital alive, the 12-month risk of death or dependency* was reduced by a significant 51%.

The authors also performed a meta-analysis of selected MEDLINE articles and found that the case-fatality rate of SAH has dropped by about 1% per year, from 1980 to 2005.

In an accompanying editorial, neurosurgeon Cargill Alleyne, Jr.,** of the Medical College of Georgia, warns that the numbers of SAH cases in the time-trend study were small (27 and 38 in the respective cohorts); nevertheless, the findings are consistent with the observations of seasoned practitioners.

OSCP = Oxford Community Stroke Project; OXVASC = Oxford Vascular Study.

* Adjusted for age and SAH severity.

** Who performed his fellowship at the Barrow Neurological Institute, where Michaels was treated.

Image from Wikipedia: Horizontal CT cut showing hyperdense subarachnoid blood in the basal cistern.

Three environmental factors appear to increase the risk of multiple sclerosis or exacerbate its course.

1. Early infection with Epstein-Barr virus.
2. Cigarette smoking.
3. Vitamin D deficiency (a factor that may explain the observed relationship between MS risk and residence in sunlight-deficient Northern latitudes).

Whether vitamin D supplementation favorably alters outcomes in MS was recently assessed in a randomized early-phase study of 49 patients.* Results of the open-label trial, which was conducted in Canada, were published last week in an advanced, online issue of Neurology.

Level-raising doses of vitamin D (up to 40,000 IU daily), along with supplemental calcium, appeared to reduce the number of relapses during the 52-week study and reduce the proliferation of proinflammatory T cells—both secondary outcomes. High-dose vitamin D did not cause hypercalcemia, persistent hypercalciuria, or nephrolithiasis and appeared to be well tolerated (the primary outcomes). "No significant" adverse events were observed, despite the fact that the mean level of 25(OH)D peaked at 413 nmol/L, which is well above recognized "toxic" levels (eg, 220 nmol/L).

Given the unblinded design of the trial, the favorable clinical outcomes can only be categorized as class level 4 evidence—in other words, data from an observational study that did not have an adequate control group.** The authors indicate that a multicenter, phase 2 proof-of-concept trial, incorporating clinical and MR outcomes, is currently underway.

The NIH database (clinicaltrials.gov) currently lists 3 trials that are assessing vitamin D supplementation in MS (the Canadian phase 2 study is evidently not registered yet).

A phase 4 Israeli trial of high- and low-dose vitamin D. The Merck Serono-supported study will assess the immunomodulatory effects of vitamin D supplementation on interferon-beta reactions.

A phase 4 Norwegian study of changes in bone-mineral density.

A phase 1 pilot study at Hopkins, which will assess the safety and immunomodulatory effects of low-dose vitamin D.

25(OH)D = 25-hydroxyvitamin D. (Measuring 25(OH)D is the standard for assessing the body's vitamin D status.)

* Forty-five patients had relapsing-remitting disease.

** Patients in the trial's control group were allowed to take vitamin D and calcium at their discretion.

Continued assessment of the medical literature confirms the incomplete publication of clinical trial results. The latest analysis, in last week's open-access Trials journal, reveals that nearly 20% of completed interventional studies in acute ischemic stroke were never published in full. For instance, among the unpublished trials was the PASS-2 study of the GABA derivative piracetam, a widely used drug. The implications of the review, performed by researchers at the University of Edinburgh, is that negative clinical trials are not being published to suppress information that would otherwise undermine the prescription of non-beneficial, or even harmful, drugs.

However, coverage of the analysis at Medscape reveals that many of the unpublished studies were phase 2 trials—the results of which are generally used to rationalize all-important phase 3 investigations. Examples include trials of the potentially neuroprotective agents repinotan and tirilazad. These drugs have been assessed in later-phase trials, and the complete results have, in fact, been published. Bayer announced the discontinuation of its repinotan development program in 2004, after disappointing results of a phase 2b trial. And a randomized, double-blind trial of tirilazad published in 1996 showed no outcome benefit with the drug in acute ischemic stroke. A 2000 Cochrane review indicated that tirilazad actually worsens stroke outcome.

Medscape interviewee Marc Fisher, MD, also suggested that some of the unpublished studies were probably submitted, but rejected, for publication. He cited the low acceptance rate of the journal Stroke, 18%. (Evidently the Edinburgh researchers did not have the submission-rate information for the unpublished studies.)

Other stipulations are also warranted. Among the unpublished stroke trials, 40% enrolled fewer than 50 patients, and 58% enrolled fewer than 100 patients. Although 23 (18%) of the trials were performed in the United States, 16 (13%) were conducted in China. And many of the unpublished studies were dated: About 40% of the trials were more than 10 years old; roughly 30% were reported within the last 5 years.

Horizontal CT image of massive, acute right hemispheric brain infarct with midline shift from Wikipedia.