College of American Pathologists
Printable Version

  Vitamin D intrigues, but not a done deal


CAP Today




June 2009
Feature Story

Karen Lusky

When Julian Barth, MD, a clinical pathologist in the United Kingdom, attended a recent presentation in the U.S. on studies linking vitamin D to everything from cancer and immunity to heart disease and diabetes, he considered supplementation and wondered how much it might cost. A quick look at a local drugstore gave him an answer: $7 for a 120-day supply of 1,000 international unit (IU) tablets.

At that price, which is fairly typical, vitamin D is one heck of a bargain, if it helps prevent numerous chronic conditions, as a growing body of mostly epidemiological evidence suggests. For now, consumers and physicians interested in the vitamin’s potential health benefits are what’s behind the boost in vitamin D testing.

“In lab diagnostics,” says Ravinder Singh, PhD, co-director of the endocrinology lab at Mayo Clinic, Rochester, Minn., “I have never seen such a rapid growth in a chemistry test. From 2007, the test volume for vitamin D has grown nationally by 80 to 90 percent overall.” Most of the testing is being done for nutritional purposes.

To meet the escalating demand, labs face the usual challenges of selecting a test method based on cost, complexity, and test volume.

The more widespread testing has also drawn attention to the fact that different test methods tend to produce different ranges of test results, which has been confusing to physicians and patients. “The key issue” there, says the U.K.’s Dr. Barth, who presented as part of a panel discussion on vitamin D at the Dark Report’s recent Executive War College in New Orleans (where he investigated supplementation for himself), is the “enormous need for standardization of vitamin D testing”—a point upon which the panel of laboratorians all agreed.

The forecast for vitamin D isn’t all blue skies. Some experts warn that widespread testing and supplementation could be putting the treatment cart before clinical trials proving that long-term supplementation can safely stave off or ease various diseases. There’s also the concern that the testing boom could go bust if payers pull the plug to some extent on the growing stream of vitamin D testing, or if health care consumers lose interest in the vitamin.

To set the record straight, however, vitamin D is actually a hormone that has been labeled as a vitamin because of the need for supplementation, says vitamin D researcher Bruce Hollis, PhD, professor of pediatrics and professor of biochemistry and molecular biology at the Medical University of South Carolina (MUSC), Charleston.

Behind the attention vitamin D is getting is the concern that the amount found in a typical multivitamin and diet fails to meet the nutritional needs of those who don’t get enough sun exposure to make sufficient D3 on their own. (D2 or D3 can be used for nutritional supplementation; pharmacologic treatment is done predominantly with 25-hydroxyvitamin D2.)

The Institute of Medicine in March appointed a committee to examine the evidence for increasing the IOM’s recommended dietary reference intake for vitamin D, which is now 200 IUs a day from birth to age 50, 400 IUs from age 51 to 70, and 600 IUs after age 70. (The IOM committee is also looking at the adequacy of recommended calcium supplementation levels.) The committee is expected to report its findings by May 2010.

The current recommended vitamin D requirement of 200 to 400 units for adults, says Dr. Hollis, who presented on the topic at the War College, was set up for adults 50 years ago with no scientific basis: “They just set it—bing!” he says.

At that time, experts believed that vitamin D deficiency was associated only with skeletal problems. Later, scientists found 2,000 human genes with vitamin D response elements in the region controlling whether the gene is turned on or off, says Dr. Hollis, who developed the first radio­immunoassay for vi­tamin D and now serves as a consultant to DiaSorin of Stillwater, Minn., which sells vitamin D immunoassays.

As for what’s viewed as a normal serum value for vitamin D, Robert Heaney, MD, who is John A. Creighton University Professor and professor of medicine at Creighton University in Omaha, Neb., says most vitamin D scientists in North America today view 32 to 40 ng/mL for 25-hydroxyvitamin D (25OHD) as the lower end of the normal range. 25OHD is the metabolite typically measured to assess vitamin D status. The kidney turns 25OHD into 1,25-dihydroxyvitamin D, the physiologically active form of vitamin D in the body for the hormone’s endocrine function. Many other tissues are able to make the same conversion for local paracrine action.

What’s considered optimal target concentrations of 25OHD is “still open to debate,” however, says Russell Grant, PhD, strategic director of the National Office of Quality and Science at LabCorp, who presented as part of the War College panel on vitamin D testing.

Most everyone agrees that research done to date casts vitamin D in a promising light.

Endocrinologist Carol Zapalowski, MD, PhD, says low vitamin D levels have been associated with osteopenia, osteoporosis, and increased fracture risk. In addition, epidemiologic data suggest low vitamin D status is associated with an increased risk of several common cancers, including colon, prostate, and breast cancers, as well as autoimmune conditions, including ­multiple sclerosis and inflammatory bowel disease, says Dr. Zapalowski, of the Denver Center for Bone Health and a spokesperson for The Endocrine Society.

Also on the list: “Myocardial dysfunction, deaths due to heart failure, and sudden cardiac death ... glucose intolerance and hypertension,” Dr. Zapalowski adds. “Treatment with vitamin D has been shown to decrease the risk of osteoporotic fractures, as reported in a recent meta-analysis” (Bischoff-Ferrari HA, et al. Arch Intern Med. 2009;169:551–561).

Michael Holick, PhD, MD, professor of medicine, physiology, and biophysics at Boston University School of Medicine, says the Women’s Health Initiative study showed that study participants with baseline blood levels of less than 12 ng/mL 25OHD had a 253 percent higher risk of developing colorectal cancer compared with women who started out with 24 ng/mL.

However, says Jean Wactawski-Wende, PhD, one of the study authors and a professor in the Department of Social and Preventive Medicine at the University at Buffalo in New York, the study of 36,282 postmenopausal women did not find that 400 IU of vitamin D combined with calcium over a seven-year period affected the rate at which the women developed colorectal cancer when compared with placebo use.

Vitamin D at a higher dose combined with calcium did appear to lower the rates of cancer of all types and fracture in a population-based, double blind, randomized, placebo-controlled trial in which Creighton’s Dr. Heaney was involved. In the study, 1,179 community-dwelling women 55 years and older in Nebraska received 1,100 IUs of vitamin D3 and supplemental calcium for four years. “In multiple logistic regres­sion models, both treatment and serum 25-hydroxyvitamin D concentrations were significant, independent predictors of cancer risk,” says an abstract of the study published in the American Journal of Clinical Nutrition (Lappe JM, et al. 2008;87:794).

Vitamin D deficiency has also been linked to higher rates of infection, including bacterial vaginosis during pregnancy, according to a study by researchers at the University of Pittsburgh (Bodnar LM, et al. J Nutr. 2009;139:1157–1161).

In research published earlier this year in the Archives of Internal Medicine, Adit Ginde, MD, MPH, of the University of Colorado in Denver, and Harvard researchers found that low 25OHD levels (less than 30 ng/mL) were associated with higher rates of self-reported upper respiratory tract infection. They say the association may be stronger in people with respiratory tract disease (Ginde AA, et al. 2009;169:384–390).

Recently published studies also suggest vitamin D may promote healthy cognitive function. That ­doesn’t surprise Dr. Hollis, given that the brain contains vitamin D receptors. And they aren’t there to make bone—“or hopefully not,” he quips.

Labs have various testing options and related issues to consider in providing the higher volume of vitamin D testing that the recent research appears to have fueled.

Lokinendi Rao, PhD, director of the Core Laboratories at the Univer­sity of Massachusetts Memorial Medical Center, Worcester, says there are basically two methods for testing 25OH vitamin D: immunoassays (CIA, RIA, ELISA) and chromatographic assays (HPLC and LC-MS/MS). As seen in the CAP Surveys, the majority of clinical labs use the DiaSorin Liaison. DiaSorin sells both the Liaison, a chemiluminescence immunoassay, and its original radioimmunoassay. Both tests report a total value that combines 25-hydroxyvitamin D3 and D2.

North Bend Medical Center, a 50-physician, multispecialty practice in Coos Bay, Ore., chose to go with the Liaison, which its reference lab had been using, when the practice brought vitamin D testing in house last November. The test volume had doubled earlier in 2008 when news reports about vitamin D began to hit the mainstream media, says Ed Troyer, MT(ASCP), manager of the practice’s laboratory. The practice is doing about 700 tests a month.

Cleveland Clinic, which uses the DiaSorin RIA, is moving to the Liaison to handle volume that jumped from 1,500 tests a month in 2006 to today’s volume of almost 12,000 a month, says Manjula Gupta, PhD, medical director of the endocrinology lab. “We are getting two Liaisons, although one could handle our volume now,” she says. “The second one will serve as a backup and to accommodate the future increase in our testing volume.”

Immunodiagnostic Systems (IDS) of Fountain Hills, Ariz., offers an RIA and EIA that reports a total 25OHD value combining 25OHD2 and 25OHD3. IDS’ EIA offers a lower-cost alternative to the Liaison, says Kenneth Gibbs, CEO and managing director of IDS. The test “can be implemented aboard open micro­plate analyzers that are available from a number of companies,” he says.

At the AACC annual meeting next month, IDS plans to unveil its iSYS Automated Analyzer, which employs chemiluminescent immunoassay meth­odology designed to compete with DiaSorin’s Liaison, Gibbs reports.

ESA Biosciences of Chelmsford, Mass., sells an FDA-cleared high-pressure liquid chromatography (HPLC) test that separates and measures 25OHD3 and 25OHD2, but reports the sum of the values for clinical decisionmaking, says Tom Bleiziffer, VP of sales. Like some immunoassays, the HPLC method is high complexity. Since HPLC is a physical separation method, he says, it tends to correlate more closely with tandem mass spectrometry methods as well.

ESA Biosciences initially decided to look at selling the test to small and mid-size hospital labs, believing it was a good match for labs doing 200 to 1,000 tests a month. “But then we started getting more and more interest from reference labs and physician office labs,” Bleiziffer says.

Some labs are offering lab-developed liquid chromatography, tandem mass spectrometry (LC-MS/MS) for vitamin D testing. The method, which is high complexity, can provide a total vitamin D value as well as separate results for 25OHD3 and 25OHD2. Mayo’s Dr. Singh says up to 20 percent of patient samples contain vitamin D2.

The LC-MS/MS method, however, tends to produce higher results than the widely used Dia­Sorin immunoassay method with which many physicians are most familiar. “Historically,” says Caroll E. Streetman Jr., president of DiaSorin, “mass spec results have reported out approximately 10 to 20 percent higher than the DiaSorin Liaison vitamin D values.”

Labs offering LC-MS/MS are coming up with different ways to handle the difference in test results that the two methods produced.

As one example, LabCorp took great pains to harmonize its LC-MS/MS for vitamin D testing to the DiaSorin Liaison, which it still uses for the majority of its testing, says Andre Valcour, PhD, director of esoteric immunoassays at LabCorp’s Center for Esoteric Testing and a member of the War College panel of vitamin D experts.

LabCorp sees the harmonization effort as important, in part because the majority of clinical studies on vitamin D were done using DiaSorin assays—either the RIA or the Liaison, Dr. Valcour explains. And physicians can more easily use the harmonized LC-MS/MS results to make clinical decisions based on the cut points described in the literature, he says.

To harmonize the two tests, Dr. Grant says, LabCorp worked directly with DiaSorin to access the same sources of 25-hydroxyvitamin D materials (25-hydroxyvitamin D2 and D3). LabCorp also obtained “the same proprietary calibration matrix directly from DiaSorin and the same protocol for assignment of stock concentrations in manufacturing of our calibrators for LC-MS/MS.”

At UMass Memorial, the Liaison immunoassay had been used for a long time. “We are a large institution with a large outreach clientele,” Dr. Rao says, “and some of our big clients wanted us to report 25OHD by LCMS method only.” For those clients, the testing had been sent to Mayo Clinic. Ultimately, Dr. Rao says, “we developed LCMS assay in our own lab, which was validated against Mayo Clinic method. Now we are offering both LCMS and the immunoassay.”

The laboratory at UMass Memorial conducted a study involving about 600 patients who had testing done simultaneously using the lab-developed LC-MS/MS and the DiaSorin Liaison immunoassay. And as Dr. Rao explained in a pre­sentation at the War College, the lab used the study results to develop a conversion factor so it could provide both the LC-MS/MS value and the converted, estimated immunoassay value. Physicians ordering vitamin D testing had the option of looking at the real LC-MS/MS value and the converted immunoassay value. However, he says, this created confusion for the physicians interpreting the test results, so the converted immunoassay value was removed.

Quest Diagnostics hasn’t harmonized its mass spec with DiaSorin, which the company does offer in one of its business units. Wael Salameh, MD, Quest’s medical director of endocrinology, is quick to note, however, that Quest has put its widely reported problems with LC-MS/MS vitamin D testing inaccuracy behind it and has in place “layers of controls,” including blinded studies, a single companywide standard for doing the testing, stringent operating procedures followed by staff dedicated to doing that particular assay, and weekly reviews to monitor improvements and test consistency.

As for the rationale for eschewing the harmonization route with its LC-MS/MS, Dr. Salameh notes that the public domain contains “multiple publications and DEQAS survey data that the LC-MS/MS all-method mean is higher than the DiaSorin method. This is also true for HPLC.” (DEQAS is the Vitamin D External Quality Assurance Scheme, an international proficiency testing program.)

Dr. Salameh further notes that a cutoff of 30 ng/mL for normal, which Quest and many labs use, is based on “bone and mineral data” derived using a variety of test methods. “That’s different than the epidemiological data linking vitamin D to colorectal cancer, CVD, diabetes, colds in the winter, etc.,” most of which were obtained using DiaSorin assays.

In fact, Dr. Salameh says that since vitamin D tends to be very stable, archival samples from the epidemiological studies using the DiaSorin method should be reanalyzed with LC-MS/MS methods calibrated against the soon-to-be released National Institute of Standards and Technology standards. “Preliminary assessment of a beta release of these standards by Quest showed good agreement with our LC-MS/MS method,” he says. He points out that reanalyzing those samples with technologies such as LC-MS/MS, which can differentiate between D2 and D3, has the benefit of demonstrating potential differential biological potency between the two forms at different tissue targets. The immunoassay method not only reports a “total” value but also tends to have a bias toward underreporting of vitamin D3 and especially of vitamin D2, which explains why LC-MS/MS and HPLC methods report slightly higher overall vitamin D levels.

Apart from the differences in value ranges produced by different vitamin D testing methods, labs using the same method are known to produce variable test results. IDS’ Gibbs says, “Based upon DEQAS vitamin D assay survey data, LC-MS/MS appears to be more widely variable from laboratory to laboratory than FDA 510(k)-cleared commercial kits.”

All of this discourse on vitamin D testing, of course, tends to confound physicians who just want accurate results to guide their prescribing. UMass Memorial’s Dr. Rao says he gets “calls from doctors who say they received different lab results from different labs—and what should they believe?”

As one way to get around that, Dr. Rao encourages physicians to use one lab, which helps ensure they get similar results for the same patient.

James Nichols, PhD, medical director of clinical chemistry at Baystate Health, Springfield, Mass., says if physicians know there is a range of difference between vitamin D tests, they could educate patients who are a few nanograms over the cutoff for normal to consider supplementation—“if this is a significant patient concern.”

Standardizing the different testing methods is the ultimate ticket to achieving more uniform, accurate test results, as the War College vitamin D panel members stressed. And at CAP TODAY press time, the National Institute of Standards and Technology was hoping it would, within two to three weeks, make available standardized reference materials for quality control and calibration of vitamin D testing, says Karen Phinney, PhD, a research chemist at NIST.

“NIST has developed two different standard reference materials for 25OH metabolites [D2 and D3],” Dr. Phinney says. “The first one is SRM 972, a serum-based material that has four different pools or levels of the various metabolites” for quality control. One of the pools is diluted with horse serum to produce a lower level of 25OHD3.

The other SRM that NIST plans to provide is 2972, a suite of solvent-based solutions—one for 25OHD3 and the other for 25OHD2—that can be used for calibration. “The calibration solutions could be diluted and used to prepare a calibration curve to do quantitation,” says Dr. Phinney.

“The standard reference material for calibration is for people who don’t have calibrators readily available or have some concerns” about the calibrators they are using, Dr. Phinney explains. The SRM would provide them ready access to a calibrator with a known concentration for D2 and D3.

As for the QC materials (SRM 972), Dr. Phinney says: “Individual companies have their own QC materials for their tests [as do labs using lab-validated tests]. The NIST standard reference materials aren’t intended to replace those but rather to serve as an adjunct.”

Standardizing assays could help improve testing accuracy industrywide, but a number of experts say the evidence isn’t there yet to make broad-stroke recommendations for using the vitamin in high doses to treat a variety of diseases.

Quest’s Dr. Salameh describes the data for testing and supplementing as “quite solid” for osteoporosis and anything that would lead to vitamin D deficiency and secondary hyperparathyroidism, such as malabsorptive syndrome, bariatric surgery, celiac disease, burns, obesity, drugs that interfere with vitamin D metabolism, and vulnerability in populations such as the elderly. But “our position is that if someone needs a value above 30 ng/mL, the primary reason would be to achieve adequate bone health, prevent muscle weakness, and decrease the number of falls and fractures.”

Dr. Salameh views as promising the data linking vitamin D by association to cardiovascular disease, the immune system, cancer, and diabetes. But “association studies are not necessarily causal,” he adds, noting that folic acid didn’t pan out, as expected, in helping to prevent heart disease. “And there’s also the estrogen story. That’s why one needs to be careful about epidemiological data.” However, “until definitive data become available from these prospective trials, and in light of the safety profile of vitamin D, many physicians will choose—and appropriately so—to treat to higher levels of vitamin D to prevent or mitigate the likelihood of these diseases,” he says.

The University at Buffalo’s Dr. Wactawski-Wende also urges caution: “We don’t know if observational studies showing vitamin D’s association with various health outcomes is a proxy for a healthy lifestyle, for example. Perhaps people with higher levels get more sun exposure because they exercise more and/or have better nutrition overall.”

A few prospective, randomized intervention trials planned or in the works could shed more light on the clinical utility and safety of vitamin D at higher doses.

Harvard researchers plan to conduct a large-scale randomized trial with 20,000 men and women in the U.S. on the impact of vitamin D on preventing cancer and cardiovascular disease. The trial will also look at the impact of vitamin D and marine omega-3-fatty acids in combination and alone.

Study participants will be assigned randomly to one of four treatment groups for five years. One group will receive vitamin D3 (2,000 IU/d) and fish oil (EPA+DHA, 1 g/d). A second group will get vitamin D3 and placebo fish oil; a third group, a placebo vitamin D3 and fish oil; and the fourth group, double placebos.

The researchers plan to collect and store participants’ blood samples at baseline and followup to look at how supplementation affects 25-hydroxy­vitamin D and omega-3 fatty acid levels, as well as to assess effect modification (whether intervention results vary) by baseline levels. They also plan to test genetic/biochemical hypotheses, according to an abstract summarizing the planned trial. Ancillary studies related to cognitive function, diabetes, hypertension, and other outcomes are also planned. Principal investigators for the study are JoAnn Manson, MD, and Julie Buring, ScD, of Brigham and Wo­men’s Hospital and Harvard Medical School.

Dr. Hollis, of the Medical University of South Carolina, reports he is running NIH-funded randomized, controlled studies on vitamin D requirements in pregnancy and lactation. The study on lactation is to date showing that 6,400 IUs a day for the breast-feeding mothers seems to be safe and sufficient for the mother, who then gives the infant adequate amounts, he says.

The studies are important, in his view, because “obstetricians have been taught that vitamin D is poison to pregnant women [and] prenatal vitamins only have 400 IUs.” But if you give that amount to a pregnant woman, “nothing happens. And vitamin D deficiency in utero is believed to have a profound effect. Once it happens, you can’t go back and undo it.”

Another NIH-funded study at Harvard, with which Dr. Hollis is associated, will be looking at high-vitamin D supplementation during pregnancy to prevent childhood asthma.

Clinical trial data showing vitamin D works could provide a sounder scientific rationale for payers to reimburse vitamin D testing for certain indications.

As it stands now, the handwriting may be on the wall for Medicare, which has a national fee limit of about $40 per vitamin D test, to start balking at the higher rates of use.

The first rumble of trouble in that regard came earlier this year when National Government Services, a Medicare contractor, issued a restrictive draft local coverage decision for vitamin D. The draft decision covered only chronic kidney disease, hypercalcemia, osteomalacia, and rickets. It “looked like it came out of a 40-year-old textbook,” says Charles Root, PhD, president of CodeMap in Barrington, Ill.

The Medicare contractor received a lot of comments on the proposal, and, as a result, the final local coverage decision is “fairly reasonable,” Dr. Root says. The allowed codes now include osteoporosis, and 268.9 (unspecified vitamin D deficiency). “This code can be used to cover any patient found to have a vitamin D deficiency during subsequent treatment or monitoring,” he says.

Dr. Root predicts, however, that labs will see other local coverage decisions “slowly go out.”

What about private payers? Dr. Root suspects the higher volume of vitamin D testing may be slipping by them. “My impression is that if a test is under $100, they don’t really pay attention to it.”

Boston University’s Dr. Holick notes that labs using a platform immunoassay method may be offering the testing for $25 to $40, but “if you send the test out to a reference lab by yourself, $200 is not that unusual of a price,” he says. And “some physicians tell me that the cost of the test is overwhelming for some patients without third-party coverage.”

Beyond the science or payment policies, the true wild card determining the fate of vitamin D usage and related testing may be consumers themselves who could move onto that next promising thing, leaving D in the dust.

Troyer says North Bend Medical Center has wondered if the testing is a novelty that will wear off. But “the doctors in the practice think not,” he says. Troyer has also contemplated what might happen when everyone in the area gets their vitamin D checked once, and, if it’s low, rechecked once after supplementation. That could happen, he concedes, but so far the practice’s test volume has not shown any sign of leveling off.

Dr. Nichols, who says Baystate Health has, like most labs, seen its vitamin D testing increase in the past two years, compares vitamin D to other tests that have had as much media hype, such as homocysteine and high-sensitivity C-reactive protein. As for the fate of those tests, hs-CRP “is still fairly high and we still get quite a lot of homocysteine orders, but neither is continuing to increase exponentially like vitamin D.”

Creighton University’s Dr. Heaney waxes optimistic on the long-term prospects for vitamin D supplementation and testing, noting that the volume of evidence supporting the vitamin’s role in various diseases continues to accumulate day by day. “This doesn’t appear to be going away,” he says.

Karen Lusky is a writer in Brentwood, Tenn.