Why do some diagnostic tests that seem useful to clinicians in their infancy never fully develop into stars? A number of reasons should be kept in mind when working with a new test in your laboratory or just reading an article or listening to a vendor.

Let me begin with a story about a wonderful test that really never saw the light of day. Our group developed a test that would provide useful information to neonatal physicians about new born patients with increased bilirubin. They agreed the test was what we claimed it to be and the kit was approved by the FDA. Only one physician ever ordered the kit. The rest were accustomed to using a bilirubin of 20 mg/dL as a set point to treat their patients. It was too ‘scary’ to use the new marker.

We often find that as a reason a new, better test is not adopted. Currently I submit cystatin C as one such test. Inertia can be a significant reason some tests do not become stars. Another reason that a diagnostic test may not develop is that the initial testing is done on too few patients or, more likely, too narrow a variety of patients. For example BNP and NT-BNP were early suggested by some to be markers of congestive heart failure. They are, but it was soon found that patients with pulmonary disease and patients with myocardial infarction also had some similar signs and symptoms as well as increased levels of these markers. And the test became less useful. Which is to say the test should be used with knowledge that these are not perfect tests (there being no such test), but used in the context of where and when elevations are seen can be quite helpful.

Another reason is that the initial method, while accurate and precise and reasonably sensitive and specific for certain pathologies, is too expensive, cumbersome or time consuming to perform. One example is myoglobin as first marketed – as a radioimmunoassay that took three hours to perform and a fresh calibration each time. Another more recent example is perhaps myeloperoxidase.

I was recently asked to give a lecture on establishing a reference range. For me, this is an easy task on paper but a difficult task to carry out. By reference range I mean that range from a low value to a high value within which most “normal” people will fall. Usually the reference range includes only those people above the lower 2.5 percent and below the upper 97.5 percent of those people tested—the 95 percent range. Thus, by definition, 5 percent of “normal” people will be excluded from the normal range and be labeled abnormal. Sometimes, diagnostic tests will be run on these persons to verify that they are normal and not abnormal. Any definition, other than including everyone labeled “normal” will result in some people being labeled “abnormal.” The more diagnostics tests run on any person using a cutoff such as the 95 percent range will result in an increasing number of these “abnormal” patients.  If 5 percent are abnormal by Test A and another 5 percent are abnormal by Test B, that yields 10 percent abnormal results.

The next problem is the definition of “normal.” A partial list of those factors that could be considered as part of the definition of normal might include:

• Was the patient fasting? 
• What time of day was the sample drawn?
• What time of month was the sample drawn?
• Was the person on ANY medication (including any vitamin supplement, aspirin, etc.)?
• Does the person drink, and how much?
• Is the patient a smoker?
• What's the patient's age and sex?

If we forget the first points and use only age and sex and ask for, say 100 data points for each sex and each decade beginning at age 20 through age 70, we have 2 times 6 time 100 or 1,200 persons that we much test for any analyte for which we want to establish a reference range. Adding any other variables increases the number of patients we need to test.  

We might define as normal a person with a normal chem.-12 profile.  If that is the only criterion for normal and we use as normal the 95 percent range mentioned earlier and eliminated anyone with an abnormal value, we would lose nearly 50 percent of the patients we tested as being abnormal. The 50 percent is 0.95 raised to the 12^th power.  Think of adding a normal CBC to this criterion and then add a normal urinalysis and imagine how few people for each sex and age and you can see the time, effort and cost involved in establishing a reference or normal range. It is a daunting task. Verifying a reference interval is easier but still involves consideration of the variables and thus the source of the data.

 What's the diagnosis?  

 Answer: diabetic ketoacidosis

 Another one:

Answer: Non-thyroidal illness.  Normal cortisol suggests no issue of the hypothalamus-pituitary-thyroid axis.

Coumadin (Warfarin) has long been used to treat patients with atrial fibrillation. There is no doubt that Coumadin reduces the risk of stroke in the patients. However, the risk of hemorrhage and systemic embolism is increased. In addition to the risk factors associated with the drug, it is difficult to use for some patients. It requires periodic measurement of prothromin time (PT, INR), which can be weekly for some. Effective use of the drug also requires diet restrictions—especially green vegatables.

For some years, there has been a search of a different type of drug that would reduce the risks and be easier to use. In August (www.nejm.org August 30, 2009 ) the RE-LY group reported a large study patients with atrial fibrillation. The group reported that “dabigatran given at a dose of 110 mg was associated with rates of stroke and systemic embolism that were similar to those associated with warfarin, as well as lower rates of major hemorrhage. Dabigatran administered at a dose of 150 mg, as compared with warfarin, was associated with lower rates of stroke and systemic embolism but similar rates of major hemorrhage. The benefit of dabigatran may be explained in part by the twice-daily dosing regimen. Since dabigatran has an elimination half-life of 12 to 17 hours, twice-daily dosing reduces variability in the anticoagulation effect, especially as compared with the anticoagulation effect of warfarin, which is difficult to control. Warfarin broadly inhibits coagulation. By selectively inhibiting only thrombin, dabigatran may have antithrombotic efficacy while preserving some other hemostatic mechanisms in the coagulation system and thus potentially mitigating the risk of bleeding. “

An editorial in the same issue noted that “Because of dabigatran's twice-daily dosing and greater risk of nonhemorrhagic side effects, patients already taking warfarin with excellent INR control have little to gain by switching to dabigatran. In contrast, many other patients who have atrial fibrillation and at least one additional risk factor for stroke could benefit from dabigatran.”

Prostate cancer (PCa) remains the second leading cause of cancer deaths in the U.S. as in most Western societies. Most men with PCa will not die from the disease but with it. Since the mid-1980s, screening with the prostate-specific antigen (PSA) blood test has more than doubled the risk of a prostate cancer diagnosis. A decrease in prostate cancer death rates has been observed since that time, but the relative contribution of PSA testing as opposed to other factors, such as improved treatment, has been uncertain. The recent release of two large randomized trials suggests that if there is a benefit of screening with PSA, it is, at best, small.

Most men who succumb to PCa despite hormone treatments suffer what is known as hormone-refractory prostate cancer. Monitoring patients treated with hormones often proves clinically difficult due to conflicting data from serum levels of PSA, symptoms and bone scans. Recently a new marker has been issued FDA clearance – enumerating circulating tumor cells. Patients with an ‘unfavorable’ pretreatment level of CTCs had a shorter overall survival than those who CTC count was favorable – 11.5 vs. 21.7 months. This test may assist the clinician in predicting a prognosis and monitoring therapy and may aid in testing new drugs.

Another possible marker is PCA3 a gene discovered in 1999. Because of its differential expression between prostate cancer and non-neoplastic tissue, several studies have evaluated its value in PCa diagnosis. These papers are consistent with significant statistical accuracy of measure of the urinary number of PCA3 copies. While the sensitivity of PCA3 is slightly less than that of serum PSA, specificity as well as positive and negative predictive values are quite a bit better.

The concept of multitasking demands our attention. Is it really possible? If so, in what situations? What is known about multitasking?

To begin, quite simply, there are times when multitasking clearly works--most people can walk and chew gum or carry on a conversation while cooking dinner. The reason this is possible is that both the tasks require minimal thinking on our part. Chewing gum borders on breathing--no thinking required--allowing our brain to pay attention to walking in case something arises that needs our immediate attention such as an oncoming car. But can I work on the budget and talk to a colleague at the same time? There are data that suggest that in reality this attempt at multitasking fails. Mistakes occur. Time is wasted trying to switch tasks (which is not multitasking!). To try to enter data correctly into the correct cell while holding a meaningful conversation is likely to cause errors.

In fact, one study found that clinical distractions may account for 43 percent of medication errors. In two hours of observed CPOE use by residents during Medical Intensive Care Unit (MICU) rounds, one distraction or interruption occurred approximately every 5 minutes and preceded two order entry errors. (AMIA Ann Symp Proc. 2006:895).

There are now laws in many states dealing with using phones in automobiles. There are good reasons for them.

In our ‘do more with less” environment, we are faced with a real issue of multitasking. There certainly seem to be times and places where it is possible. We know that with practice it is possible to avoid some (but not all) errors when attempting to multitask.

It is important to study the situations you encounter where you or others are asked to multitasking to determine the feasibility. Know that some situations are just prone to significant errors if multitasking is attempted.

(A worthwhile book on this subject is The Myth of Multitasking by Dave Crenshaw. I have a set of abstracts pertaining to this myth in healthcare that I would be happy to share. E-mail me at Davidplaut@yahoo.com. Also, thank you to Carol Smola for her help with this blog).

Obesity is recognized as a serious chronic disease, but there is pessimism about how successful treatment can be. One pessimistic view is that almost no one succeeds in long-term maintenance of weight loss. For this blog, I will define success as "individuals who have intentionally lost at least 10% of their body weight and kept it off at least one year."

Last year Kruger et al. reported that adults who avoided eating at fast-food restaurants were more successful at weight loss maintenance. Adults who consumed fewer than five fruit and vegetable servings per day and accrued 420 minutes or more per week of physical activity or consumed five or more fruit and vegetable servings and accrued 150 minutes or more per week of activity were more successful at weight loss maintenance.”

Wing and Phelan at Brown University established the national weight loss registry and reported recently that participants in the registry report having lost an average of 33 kg and have maintained the minimum weight loss (13.6 kg)for an average of 5.7 years. Thirteen percent have maintained this minimum weight loss for more than 10 years.

How has keeping unwanted weight off accomplished? Wing and Phelan note that “About one-half reported receiving some type of help with weight loss (commercial program, physician and nutritionist). Eighty-nine percent reported using both diet and physical activity for weight loss. The most common dietary strategies for weight loss were to restrict certain foods, limit quantities and count calories. Approximately 25% counted fat grams.

“Three strategies were reported very consistently: 1) consuming a low-calorie, low-fat diet, 2) doing high levels of physical activity, and 3) weighing themselves frequently. Recently, a fourth behavior was identified: consuming breakfast daily. Registry members maintain their weight loss by continuing to eat a low-calorie, low-fat diet.

“Another characteristic of registry members is high levels of physical activity. These levels of activity would represent approximately 1 hour per day of moderate-intensity activity, such as brisk walking.

Registry members also reported frequent monitoring of their weight. More than 44% report weighing themselves at least once a day.”

The vigilance regarding body weight can be seen as one aspect of the more general construct of degree of conscious control exerted over eating behaviors. There are some Web sites that serve as ‘group therapy’ session to aid in this effort.

Maintaining weight after a loss is difficult, but possible.

Acute pancreatitis usually occurs as a result of alcohol abuse or bile duct obstruction. A careful review of the patient's history and appropriate laboratory studies can help the physician identify the etiology of the condition and guide management. Prompt identification of patients who need intensive care referral or subspecialty consultation is crucial.

The number of serum biomarkers used to detect and assess severity and prognosis of acute pancreatitis (AP) is substantial.

In recent years urinary trypsinogen-2, trypsinogen activation peptide, procalcitonin, interleukin 6, C-reactive protein (CRP), phospholipase A(2), serum amyloid A, have been suggested as replacements for amylase and lipase; even ALT and bilirubin have been used to suggest a gallstone as the cause.

Serum amylase remains the most commonly used biochemical marker for the diagnosis of acute pancreatitis, but its sensitivity can be reduced by late presentation, hypertriglyceridaemia, and chronic alcoholism. The advantages of amylase are its technical simplicity, availability and high sensitivity. Its greatest disadvantage is its low specificity. A normal amylase would usually exclude the diagnosis of AP, with the exception of AP secondary to hyperlipidemia, acute exacerbation of chronic pancreatitis, and when the estimation of amylase is delayed in the course of the disease.

The major advantage of lipase is an increased sensitivity in acute alcoholic pancreatitis and in patients who initially present to the emergency room days after the onset of the disease, as lipase remains elevated longer than amylase.

Simultaneous estimation of amylase and lipase does not improve the accuracy. A recent survey from several laboratories when both amylase and lipase were ordered tends to confirm this – when one is normal the other is most often normal and when one is abnormal so is the other.

Some clinicians feel that serum lipase is a more reliable diagnostic marker of AP than serum amylase; others argue that urinary trypsinogen-2 is convenient, of comparable diagnostic accuracy and provides greater (99%) negative predictive value; others suggest that the serum trypsin level is the most accurate laboratory indicator for AP and still others feel that C-reactive protein (CRP) remains the most useful. To some CRP is still the reference parameter of all predictors indicating severe disease and pancreatic necrosis.

While most parameters peak early and decrease rapidly thereafter, C-reactive protein (CRP), phospholipase A(2), procalcitonin and serum amyloid A are reliable predictors with persistently elevated levels in severe disease. So far, no single parameter has been developed which is suitable for early prediction of infected pancreatic necrosis.

It seems that there is not yet a consensus on which marker(s) to use. My conversations with those in the field suggest that both amylase and lipase in tandem is the most common approach. The data do not suggest that this is the best approach.

Thanks to Andy Mazzara, Beth Field, Sandy Elder and Carol Smola for their help with this entry.

Bacillus Calmette-Guerin (BCG) therapy of bladder cancer is the most successful immunotherapy used against any established human neoplasm. Although how BCG works in the bladder is still being defined, it is clear that BCG orchestrates a vigorous immune response involving sensitized T-lymphocytes, activated macrophages and their cytokines that interact to kill and prevent growth of cancer cells. Instilled into the bladder, BCG eradicates existing tumors, reduces the frequency of tumor recurrences, delays stage progression to invasion and improves survival over other local treatments. How did BCG, a vaccine used to prevent tuberculosis in much of the world’s population, become standard therapy of one of the most prevalent cancers in man? It is a remarkable story, marked by purposeful scientific discovery and serendipitous observation that culminated in sound experimental and clinical investigation.

Discovery and development of the BCG vaccine date back 100 years. In 1908, Albert Calmette, a bacteriologist, and Camille Guerin, a veterinarian, working together at the Pasteur Institute in Lille, France began their life-long quest to develop a TB vaccine against the scourge of tuberculosis. By 1915, they had administered an early form of the strain to several cows, demonstrating protection against TB.

Tuberculosis was noted to have antitumor effects early in the 20^th century. In 1929, Pearl, in an autopsy study at the Johns Hopkins Hospital in Baltimore, reported a lower frequency of cancer among patients with tuberculosis. Pearl concluded that some mutual antagonism existed between TB and cancer, although he was unable to explain its basis. However, the link between tuberculosis and cancer was made.

The possible use of BCG as cancer therapy began with the landmark studies in the late 1950s by Lloyd Old at the Sloan-Kettering Institute in New York and provided the first direct evidence that BCG had antitumor effects.

The clinical use of BCG as cancer therapy began in 1969; the early reports created enormous interest in BCG, and clinical trials flourished using BCG against lung, prostate, colon and kidney cancers. Their promise was not fulfilled, and BCG was soon replaced by more effective therapy. The notable exception was bladder cancer.

Coe and Feldman demonstrated a strong delayed hypersensitivity reaction to BCG in the guinea pig bladder. Zbar established the criteria for successful BCG therapy. Based on these studies, Alvaro Morales, urologist in Canada, had the foresight to try topical BCG in the bladder.  Morales devised the original BCG protocol against bladder tumors in 1972 and in 1976 he published the first use of intravesical BCG against superficial bladder cancer. In 1978, Morales’ encouraging results based on only seven patients spawned the NCI to fund two randomized controlled trials to test the effectiveness of the combined BCG regimen against superficial bladder tumors. In 1990, after local side effects and systemic toxicity were defined in more than 2,500 cases pooled worldwide, the Federal Drug Administration approved the general use of intravesical BCG in patients with superficial bladder tumors.

Thirty years after the first report, BCG therapy remains the recommended standard treatment of high-grade noninvasive bladder cancer. Concern about genetic drift and altered immunogenicity of the bacterium after years of subculture at various facilities has not emerged as a significant problem. Although the dose, frequency and duration of treatments continue to be subjects of debate, BCG therapy has stood the test of time, proved to be superior to other intravesical agents, and favorably altered the natural history of noninvasive bladder tumors. Over the past three decades, many patients have had their bladders spared, quality of life preserved and lives saved.

The February 26, 2009 issue of the New England Journal of Medicine published an article comparing four diets: 1) 20 percent fat, 15 percent protein and 65 percent carbohydrates (low-fat, average-protein); 2) 20 percent fat, 25 percent protein and 55 percent carbohydrates (low-fat, high-protein); 3) 40 percent fat, 15 percent protein and 45 percent carbohydrates (high-fat, average-protein); and 4) 40 percent fat, 25 percent protein and 35 percent carbohydrates (high-fat, high-protein). Goals for all groups were that the diets should include 8 percent or less of saturated fat, at least 20 g of dietary fiber per day and 150 mg or less of cholesterol per 1,000 kcal. Carbohydrate-rich foods with a low glycemic index were recommended in each diet (e.g., fruits, vegetables, whole grains and legumes). Each participant's caloric prescription represented a decrease of 750 kcal per day from base line resting metabolic rate.

Group discussion sessions were held once a week, three of every four weeks during the first six months and two of every four weeks from six months to two years; individual sessions were held every eight weeks for the entire two years. Daily meal plans in two-week blocks were provided. The goal for physical activity was 90 minutes of moderate exercise per week. Participation in exercise was monitored by questionnaire and by the online self-monitoring tool.

The amount of weight loss after two years was similar in participants assigned to a diet with 25 percent protein and those assigned to a diet with 15 percent protein. Weight loss was the same in those assigned to a diet with 40 percent fat and those assigned to a diet with 20 percent fat. There was no effect on weight loss of carbohydrate level through the target range of 35 to 65 percent; the change in waist circumference did not differ significantly among the diet groups.

All the diets reduced risk factors for cardiovascular disease and diabetes at six months and two years. The metabolic syndrome was present in 32 percent of the participants at baseline, and the percentage was lower at two years, ranging from 19 to 22 percent in the four diet groups. Attendance at group sessions strongly predicted weight loss at 2 years (0.2 kg for every session attended) and was similar among the diet groups.

A number of recent studies have demonstrated a major role for inflammation in atherogenesis, the pathophysiology of vulnerable plaque and, consequently, of cardiovascular events. Circulating markers of inflammation (e.g., CRP as measured by an hs-method ) have attracted considerable interest as predictors of cardiovascular risk. High levels of lipoprotein-associated phospholipase A2 [Lp-PLA(2)] are present in inflamed, rupture-prone plaques, and it appears that Lp-PLA(2) is released from these plaques into circulation before the plaque ruptures. 

In other words this marker, like hs-CRP, but unlike the troponins (as a marker of cardiac cell damage and death) is a risk marker. Research has shown that an elevated Lp-PLA(2) level nearly doubles the risk for cardiovascular events. When both CRP and Lp-PLA(2) are increased, they provide an even greater predictive capability to help identify very-high-risk individuals who would benefit from aggressive lipid-lowering therapy. This observation suggests the two markers are largely independent. Studies have shown a positive association with CHF, stroke and Lp-PLA(2). A consensus panel reviewed the literature on the Lp-PLA(2). Consistent with the ATP III guideline recommendations for the use of inflammatory markers, Lp-PLA(2) is recommended as an adjunct to traditional risk assessment in patients at moderate and a high 10-year risk. The panel suggest also that "elevated Lp-PLA(2) levels should prompt consideration of increasing the cardiovascular risk category from moderate to high or high to very high risk, respectively. At this time Lp-PLA(2) should be viewed as an important cardiovascular risk marker whose utility is as an adjunct to the major risk factors. However, currently Lp-PLA(2) cannot be recommended as a target for therapy."

Note: As of this writing, this assay is adaptable to a number of chemistry analyzers with others to follow. An extensive bibliography and/or abstracts are available from the author at davidplaut@yahoo.com.

The most common fatal genetic (inborn) disease in North American among Caucasians is cystic fibrosis (CF). Approximately 30,000 American suffer from CF and another 10 million persons are silent carriers of the defective gene. The gene that should produce a normal protein called CFTR is defective in patients with CF; the CFTR normally channels chloride ions (Cl-) out of cells, especially those in the lungs and pancreas. When the channel is blocked due to improper folding of the CFTR, Cl- cannot be properly moved. A number of disturbances occur – in the lung, for example, the mucus becomes as thick as cooked oatmeal, providing an ideal environment for bacteria such as Pneumococcus to grow.

The defective gene that results in CF was first identified in 1989. Gene therapy approaches, which attempt to introduce normal CFTR genes (using viruses to carry them) to CF patients, emerged in the 1990s as a promising avenue for development. Work is still being done to try to solve the problem of getting these viruses through the epithelial lining in the airways of CF patients.

As the studies of gene therapy continue, a number of drugs are being tested to find ways to treat CF. One interesting approach that may be of value to some patients is the small molecule PTC124. This orally available molecule was designed to read through premature stop codes in RNA translation causing CFTR to be improperly synthesized.

Another approach is to find drugs that can wedge open the gate allowing the chloride ions to pass. However, there are 1,600 known mutations to the CF gene, making the search difficult. There has been progress in opening the gate with at least two new drugs; at present these are being tested with some success in humans. A significant adjunct to this work is that some of the drugs that affect the nonsense codes is that this approach will find application in other genetic diseases such as Duchene Muscular dystrophy.

One of these is PolyHeme, a pyridoxylated, polymerized hemoglobin with a half-life of about 24 hours and a refrigerated shelf life longer than 12 months. Data indicate that fewer allogenic red blood cell transfusions are needed with these products; a study a few years ago showed quite promising data in patients whose hemoglobin was less than 1.0 gm/dL (sic) -- 75 percent of those treated with PolyHeme survived versus 16 percent of historical controls with the same level.

Recently the data from a multi-site Phase III trial of more than 700 patients indicated that although there were more adverse events in the PolyHeme group, the benefit-to-risk ratio of PolyHeme is favorable when blood is needed but not available.

The other first-generation polymerized hemoglobin product, Hemopure, a polymerized form of bovine hemoglobin an intravascular half-life of 8-23 hours and a shelf life of 36 months at room temperature. Hemopure is approved in South Africa for the treatment of adult surgical patients who are acutely anemic. In the United States, phase II trials have been put on hold due to safety issues. Biopure is currently addressing the FDA's questions regarding safety and efficacy of the product.

CA-125 is used mainly to monitor patients with ovarian cancer (OC), in part due to the low level of sensitivity when CA-125 is used alone. In an effort to improve the early detection of OC while keeping the number of false positives low (higher specificity), studies of other biomarkers, either alone or in combination with CA-125, are ongoing. In one such study, apolipoprotein A-1, transthyretin and transferrin were combined with CA-125 to detect early-stage mucinous tumors. This panel detected early-stage cancer with a sensitivity of 96 percent.

Another study of multiple markers included CA-125, macrophage colony-stimulating factor, and OVX1. At least one of the serum markers was elevated in 98 percent of patients with stage I ovarian cancer; CA-125 levels alone were elevated in only 67 percent. By the same criteria, 11 percent of healthy individuals and 51 percent of patients with benign pelvic masses had at least one elevated marker value. The authors concluded that “a panel of these three tumor markers can identify early-stage ovarian cancer with extremely high sensitivity and moderate specificity.”

Studies of annual screening have been reported in the U.S. and U.K. with more than 230,000 followed for up to three years. The British group concluded that “the prevalence screen has established that the screening strategies are feasible. The results of ongoing screening are awaited so that the effect of screening on mortality can be determined.”

While not all groups agree that such screening is efficient, continued research with CA-125 and other markers is needed to validate the encouraging work that has been done.

Shawn Berberich contributed to this entry.