Description

November 24, 2002
Biosite (BSTE) - $30.40
Market Cap - $451 MM
Cash - $68 MM, Debt - $0
2002 P/E – 36X, 2003 P/E – 24X

As a preface to this, I wanted to note that the following is very technical in areas as I collaborated with several cardiologists in my primary research process. Because Biosite has continued to make assertions with respect to physicians’ acceptance of their growth product, the Triage BNP test, the only way in which I could perform quality due diligence was through significant primary research. I encourage any of those reading this to pose questions relating to some of the more technical aspects to the short thesis.

SUMMARY
Biosite should be shorted. The stock is likely to trade down to the mid to low teens for the following reasons:
1) The failure of their growth product, the Triage BNP Test, to gain market acceptance as a marker of Congestive Heart Failure (CHF) due to BNP’s statistical failures to effectively mark patients with CHF as described in the Journal of the American Medical Association.
2) Roche’s recent pro-BNP approval and the 2003 planned entrances of Abbott Labs and Bayer into the BNP test market.
3) Roche’s commitment to competing with Biosite on price and functionality of BNP assay testing.
4) Biosite’s inability to handle large volume BNP testing coupled with Roche’s focus on high volume testing.
5) Approval of Roche’s point of care pro-BNP product – in the event BNP testing receives a CLIA waiver (for testing in physician’s offices).

OVERVIEW OF THE COMPANY
Biosite Inc. develops, manufactures and markets rapid, accurate and cost-effective diagnostic products in drug abuse testing, cardiac testing and microbiology. Biosite has three major product lines which generated approximately 90% of their revenues in Q3 2002:

TRIAGE DRUGS OF ABUSE PANEL (DOA): (34% of Q3 2002 revenues) A rapid, qualitative urine screen which analyzes a single test sample for up to eight different illicit and prescription drugs or drug classes and provides results in approximately 10 minutes. Illicit drugs detected by the Triage Drugs of Abuse Panel include: Amphetamines, Methamphetamines, Cocaine, Opiates, Phencyclidine and Tetrahydrocannabinol (Marijuana).

TRIAGE CARDIAC SYSTEM: (20% of Q3 2002 revenues) A hand-held, point-of-care diagnostic cardiac marker test capable of detecting an acute myocardial infarction (AMI), or heart attack, in about 15 minutes. The Triage Cardiac Panel works by a rapid, quantitative screen for three cardiac markers which are released in the blood when an AMI occurs; creatinine kinase-MB, myoglobin and troponin I.

TRIAGE BNP TEST: (34% of Q3 2002 revenues) A quantitative test for measurement of B-type Natriuretic Peptide, a protein market for Congestive Heart Failure (CHF). B-type Natriuretic Peptide (BNP) is a hormone produced predominantly by the heart’s ventricles. BNP is released from the heart in response to ventricle distention due to blood volume expansion or pressure overload.

WHY THE MARKET HAS LOVED BSTE - FIRST MOVER ADVANTAGE
Biosite’s growth story and high market multiple has been based on its first mover advantage in its early approval of BNP, or B-type natriuretic peptide. BNP and atrial natriuretic peptide, ANP, act as a dual natriuretic system in regulating blood pressure and fluid balance. Studies have demonstrated that the heart is the major source of circulating BNP which is stored in and secreted from membrane granules in the heart’s ventricles. BNP is a cardiac hormone that is released from the heart in response to ventricular volume expansion and pressure overload. BNP is activated by ventricular distension due to increased intracardiac pressure. Biosite suggests and has marketed to hospitals that BNP is an excellent hormonal marker of ventricular systolic and diastolic dysfunction, making it a test marker for congestive heart failure that can replace current tests, including echocardiograms. BNP testing has grown as a novel approach to quickly diagnosing CHF as shown by their quarterly revenues over the last two years:

Q1 01 (Mar): $0.2MM
Q2 01 (Jun): $0.5MM
Q3 01 (Sep): $0.9MM
Q4 01 (Dec): $1.9MM
Q1 02 (Mar): $4.1MM
Q2 02 (Jun): $8.1MM
Q3 02 (Sep): $9.8MM

The last two quarters were particularly impressive to Wall St. as physicians have taken a keen interest in this new product. However, while physicians are often prone to testing new products like Biosite’s BNP test, in order to gain broad acceptance, BNP must show a real usefulness in diagnosing CHF in order to replace other tests such as echocardiography. Biosite’s BNP test does not.

BNP IS NOT USEFUL AS A MARKER OF CHF – THE JAMA STUDY
On September 11, 2002, the Journal of the American Medical Association (JAMA) issued a study entitled Plasma Natriuretic Peptides for Community Screening for Left Ventricular Hypertrophy and Systolic Dysfunction which concluded that plasma brain natriuretic peptide (BNP) was not useful for detection of Congestive Heart Failure (CHF). The study, done by seven physicians (for objective research purposes, not company endorsement) at the Framingham Heart Study in Framingham, MA., is in stark contrast to the assertions from Biosite’s own research.

As a primer, Congestive Heart Failure (CHF) occurs when the heart cannot deliver a sufficient blood supply to the body. The New York Heart Association (NYHA) developed a functional classification system for CHF consisting of four classes based on the severity of the symptoms:

1.Class I patients are characterized as having cardiac disease, but ordinary physical activity does not produce undue fatigue or pain.
2.Class II patients are comfortable at rest, but become symptomatic during ordinary physical activity.
3.Class III patients are comfortable at rest, but become symptomatic upon even minimal physical activity.
4.Class IV patients are in the most severe stage and experience discomfort with any physical activity.

Their study was performed on a sample of 3,177 participants (1707 women, 1470 men) from the Framingham Heart Study who attended a routine examination in 1995-1998. The study examined the usefulness of natriuretic peptides for screening for elevated LV (left ventricular) mass and LVSD (left ventricular systemic dysfunction) in the community. The study was done in order to confirm reports suggesting the usefulness of plasma brain natriuretic peptide (BNP) as a screening test for left ventricular hypertrophy (LVH) and systolic dysfunction (LVSD).

Of most importance, the Framingham Study contended that previous studies (including those commissioned by Biosite) were limited by small sample sizes, selection bias and that none compared the diagnostic performance of these peptides in men and women. The conclusion was as follows:

“In our large community-based sample, the performance of BNP and NT-ANP for detection of elevated LV mass and LVSD was suboptimal, suggesting limited usefulness of natriuretic peptides as mass screening tools.”

To confirm the usefulness of the BNP level, the Framingham doctors looked at a random group of patients as well as a subset of that group with elevated LV mass and LVSD. In testing individual with known elevated LV mass and LVSD, they found BNP levels greater than 45 pg/mL in men and 50 pg/mL in women identified 95% of those with disease. They found significant BNP level overlap between normal individuals (i.e. false positives) and those with elevated LV mass and LVSD, making the test a poor discriminator for CHF determination. Put in simpler terms, the Framingham Study found by using a low value for BNP, they achieved 95% successful case finding for increased elevated LV mass and LVSD, but couldn’t discriminate between those with and without disease.

On Biosite’s website (www.biosite.com) there is a 24 page PDF file describing the Triage BNP Test that reviews the company’s own data from clinical studies and its expected values. Biosite ran two tests, one for individuals without CHF and one with CHF to evaluate the BNP assay’s effectiveness in establishing the presence of CHF. In the test for individuals without CHF, the company used a sample of 1,286 people without CHF (676 women and 610 men) using a decision threshold of 100 pg/mL (more than double that of the Framingham Study’s 46 pg/mL) and found the test to yield a general specificity of the test of greater than 98%, (2% false positives in individuals without CHF) because 98% of normal patients without CHF have BNP levels of less than 100 pg/mL. For individuals with CHF, Biosite ran a test on a sample of 804 patients with varying degrees of CHF again using the cutoff of 100 pg/mL and found that overall 81% of patients with CHF had BNP over 100 pg/mL.

For NYHA Class I patients, only 48% had levels over 100 pg/mL. THUS, MORE THAN HALF OF PATIENTS WITH MILD DISEASE ARE NOT IDENTIFIED BY THE TEST. FOR NYHA CLASS IV PATIENTS (REFER TO PRIOR CLASS IV DESCRIPTION) , 96.3% HAD LEVELS OVER 100 PG/ML, BUT THESE PATIENTS, BECAUSE OF THEIR SYMPTOMS, ARE EASILY IDENTIFIED. Because of the severity of their disease, they inevitably require additional testing such as echocardiograms. Thus, the BNP test does not have great utility in difficult to diagnose cases and does not reduce the need for additional testing in those with severe disease.

At this point, it is crucial to note the difference between the two studies. The Framingham Study chose a lower BNP level (46 pg/mL) in order to find almost all CHF cases resulting in many false positives (their sample included many false positives). The Biosite test used a high BNP level which eliminated false positives, but missed half of the cases with mild disease. If the BNP test is a screening test, the fact that it identifies patients with only severe disease while missing mild cases of CHF (NYHA Class IV Heart Disease) will not increase its use as a diagnostic tool.

IDENTIFYING THOSE WITH SEVERE CHF WITH BNP SCREENING IS NOT NECESSARY. PHYSICIANS CAN LITERALLY LOOK AT A CLASS IV PATIENT AND DETERMINE THE PRESENCE OF CHF. CHF DIAGNOSIS NEEDS A TEST THAT CAN IDENTIFY MILD CASES THAT CANNOT BE EASILY IDENTIFIED BY A PHYSICIAN – BUT BNP CLEARLY IS NOT THAT ANSWER. Echocardiography or other studies are then needed to quantify the severity of the disease rather than a screening test to identify it. The fact that a BNP test can “detect that there is CHF” in Class IV patients does not end the need to know “how much”, which is done through echocardiography.

The Framingham Study has negative consequences for Biosite. The Triage BNP Test is currently the company’s growth driver. It has had short-term hospital interest because of its rapid nature for detecting elevated levels of BNP which are consistent with CHF. Biosite claims the BNP test allows a physician in an triage center to determine quickly whether a patient has CHF¾ much faster than a typical echocardiogram. If the product is really not predictive of CHF as the Framingham study suggests, it is unlikely that it will replace the need for echocardiograms, gain rapid market acceptance or drive revenues for Biosite in the long-term. Ultimately, it will have limited application and long-term hospital acceptance.

HEAVY COMPETITION COMING INTO THEIR ALREADY SUSPECT GROWTH AREA
While Biosite’s growth product not gaining broad market acceptance and remaining a niche test was enough of an issue for the initial short thesis, I am happy to note their situation just recently got worse as they are about to face massive competition. Biosite’s growth in BNP revenues did not go unnoticed by larger, well-capitalized pharmaceutical companies that dominate the market for diagnostic products including Roche, Bayer and Abbott Labs. ROCHE ANNOUNCED APPROVAL OF ITS COMPETITIVE PRODUCT, PRO-BNP, DURING THE NOVEMBER 16TH-18TH, 2002 AMERICAN HEART ASSOCIATION MEETING. ITS EXPECTED RELEASE IS IN THE FIRST QUARTER OF 2003.

VOLUME COMPETITION
Roche’s proBNP product represents a severe problem for Biosite. Biosite has offered its Triage BNP Test in the form of a kit that acts as a point of care (POC) device that uses test strips. Their primary target market is in emergency rooms and clinics. The meter is not automated, requires a person to perform the test, takes 15-20 minutes (with operator present) and thus can only handle low volume.

Roche’s diagnostic division has focused their proBNP test on centralized laboratories that require high throughput automation (high volumes) without human interaction. Roche’s BNP test is run in its Elecsys family of assay test analyzers which already number 7,500 systems worldwide. There is no installation of equipment to use the Roche test. It can be run in existing Roche Elecsys machines (this is important to note because one of Biosite’s assertions is that they have contracts in place with hospitals already for the Triage BNP Test thus allowing them to effectively compete against larger players). The test assay is loaded into the Elecsys. Results are received in 18 minutes without a lab technician being present. This is particularly crucial in high volume hospitals where tests are referred to a central lab and results are then given to the referring physician. A physician/nurse is thus not required to physically run the BNP test, wasting 15-20 minutes handling the test and waiting for results. This makes Biosite’s growth in sales of BNP tests difficult to achieve by volume increases. PUT SIMPLY: NOT ONLY ARE THERE NO BARRIERS TO ENTRY FOR ROCHE, BUT IT IS ALSO, IN FACT, EASIER FOR THEM TO EXPAND THAN BIOSITE NOW THAT THEY HAVE FDA APPROVAL WITH AN ALREADY ENTRENCHED CLIENT BASE.

PRICE COMPETITION
CMS (Centers for Medicare & Medicaid Services, formerly HCFA), the regulatory body which determines reimbursement levels for Medicare and Medicaid, recently announced plans to cross walk BNP to Code#84588, a test for Vasopressin, which increases reimbursement from the current sub-$20 level to $47.43. The new codes go into effect January 1, 2003. With a significant increase in reimbursement poised to take effect in January, CMS’ decision appeared to be a positive for Biosite and would fuel upside to expectations for the company’s 2003 BNP revenue estimates. The average selling price for the Triage BNP Test is around $21. The current reimbursement for the test is $18, making it a losing proposition for hospitals. When reimbursement prices are changed to $46, Biosite could charge up to $5 more per test, which would generate approximately $13m more in revenues in 2003.

While this was a catalyst when no competitors existed, during the AHA meeting in unveiling pro-BNP, ROCHE ANNOUNCED IT INTENDS TO PRICE ITS TEST AT $17 FOR NEW CUSTOMERS AND AS LOW AS $14 FOR THOSE BUYING OTHER ROCHE ASSAYS¾A SIGNIFICANT DISCOUNT TO BIOSITE’S $21 TEST KIT PRICE. THIS MAKES BIOSITE’S GROWTH IN SALES OF BNP TESTS DIFFICULT TO ACHIEVE BY PRICE INCREASES.

BNP VS. PROBNP – DOCTORS SEE NO DIFFERENCE
Biosite and Roche are now currently embroiled in a fierce battle in which each claims that its product (Biosite-BNP, Roche-proBNP) to be superior. The Framingham study would suggest neither works well, but it appears that Roche is winning among physicians. At AHA, physicians polled saw no difference. Roche suggests that clinical data doesn’t indicate that there is a difference. Rather, since each measures a different peptide, a direct value comparison can’t be made. ProBNP has the advantage of having a higher negative predictive value (>97%) compared to a negative predictive value of BNP of 80%. This means proBNP could be used to rule out heart failure, whereas the Biosite BNP test is not specific enough. In terms of stability of the peptide marker itself, BNP may not be stable in serum or heparinized plasma, whereas proBNP is. Thus, proBNP may provide some advantage in sample handling in daily routine, leading to more reliable results. Also, certain new heart failure treatments use administered BNP. If an assay is needed to measure these levels, proBNP would not be influenced by administered BNP levels.

Biosite hosted a generally unimpressive dinner at AHA meeting. It was supposed to serve as a forum to state their case of BNP vs. Roche’s proBNP. Biosite’s main argument was that proBNP is not useful in patients with kidney dysfunction since proBNP is eliminated by the kidney. However, only a limited group of CHF patients (<5%) have renal dysfunction. In a weak argument, Biosite suggested that proBNP is less sensitive than BNP for classifying CHF¾contradicting Roche’s study. Biosite suggested that proBNP’s longer half-life would make it less effective in the immediate diagnosis of CHF, which Roche also denies. Finally, Biosite has made claims that BNP is better than proBNP because it has only one “cut-off” on gender vs. proBNP’s two, one for patients over 75 years old and one for those younger than 75. Biosite claims that this adds more confusion for the proBNP test because physicians must consider a patients’ age before administering a test – a claim that is immaterial. Considering Roche avoided four cut-offs for proBNP (age and gender), it could be viewed as a net positive. Contrary to Biosite’s marketing group, physicians don’t see the negatives of proBNP vs. BNP. (In fact, the positives conceivably outweigh the negatives). If any reader would like to learn additional information on this topic, I suggest calling the Roche Investor Relations department as I did.

In the end, healthcare decisions are made on price. It is easier to add the proBNP assay test to existing Roche Elecsys machines than buy Biosite’s Triage BNP Test. Regardless of the merits of the contenders, the argument is to short Biosite and not to buy Roche as proBNP represents only a small portion of Roche Diagnostic’s revenues.

CLIA WAIVER
Wall St. has noted Roche is a problem for Biosite’s BNP growth, yet 2003 estimates have yet to be lowered. Biosite has suggested that the physician’s office market is their next area of growth via a CLIA (Clinical Laboratory Improvement Amendment) waiver. CMS regulates all laboratory testing (except research) performed on humans in the U.S. through CLIA. CLIA grants waivers on medical tests so that they do not require trained personnel to perform the test. Under its current status of no waiver, in order to do a BNP test in a physician’s office (not a hospital), the physician must have a trained personnel and must comply with stringent government standards, regulations and inspections. Thus, BNP tests are performed in hospitals and not in physician offices. During the AHA meeting, Biosite was cautiously optimistic on the likelihood of a CLIA waiver in 2003 (According to the FDA , less than 10% of tests seeking CLIA waivers actually receive them). REGARDLESS, ROCHE’S DIAGNOSTIC PRODUCTS DIVISION HAS CONFIRMED THAT THEY ALREADY HAVE A POC PRO-BNP TEST UNDER DEVELOPMENT. With each passing day, the CLIA waiver becomes less of a catalyst for Biosite. Roche (and others) will be ready in 2003 for the possible passage of a BNP CLIA waiver.

GROWTH STOCK VALUATION – WITH LIMITED GROWTH
Biosite is currently trading at 36X First Call’s 2002 EPS estimate of $.85/sh and 24X times First Call’s 2003 EPS estimate of $1.29/sh. On an enterprise value to sales basis, Biosite is trading at 3.7X Bank of America’s 2003 sales estimates of $104.3MM and 2.8X their 2003 sales estimates of $137.8MM. These are rather high growth multiples to pay for a stock that has only one product projected to grow revenues significantly year-over-year when that sole area of growth is seriously being threatened. Biosite’s historical revenues and Bank of America’s 2003 revenue estimates are as follows (Note that 93% of the total 2003 expected revenue increase comes from BNP):

PRODUCT REVENUE:
BNP
2001: $3.5MM (no revenue in 2000)
2002: $37.2MM (+963% Y/Y)
2003: $68.3MM (+83% Y/Y)
DOA
2001: $36.7MM (+4% Y/Y)
2002: $36.2MM (-1.4% Y/Y)
2003: $35.5MM (-1.9% Y/Y)
Cardiac
2001: $17.2MM (29% Y/Y)
2002: $18.6MM (8.1% Y/Y)
2003: $23.2MM (25% Y/Y)
Meters/Micro/Contract
2001: $7.7MM (20% Y/Y)
2002: $12.3MM (60% Y/Y)
2003: $10.8MM (-12% Y/Y)

The bottomline is that the bullish analyst community and Biosite are banking nearly all (93% by one analyst’s account) of their hopes on the growth of BNP. My research has shown that in addition to BNP’s growth prospects being suspect absent any competition, the presence of Roche’s newly approved competing product and superior existing infrastructure will negatively affect Biosite’s BNP product sales significantly.

CATALYSTS
1)The impending release of Roche’s proBNP product leading to the slowing of Biosite’s contract signings with hospitals for the Triage BNP Test leading to a probable Q4 2002 or Q1 2003 earnings miss and subsequent Wall St. downgrades.
2)Roche’s Q1 2003 entry into the BNP assay test market with pro-BNP.
3)Roche’s subsequent POC BNP test introduction.
4)Bayer and Abbott Labs 2003 entry into BNP assay market.
5)Continued academic/research studies confirming the failures of BNP as an effective marker of CHF.

Catalyst

1)The impending release of Roche’s proBNP product leading to the slowing of Biosite’s contract signings with hospitals for the Triage BNP Test leading to a probable Q4 2002 or Q1 2003 earnings miss and subsequent Wall St. downgrades.
2)Roche’s Q1 2003 entry into the BNP assay test market with pro-BNP.
3)Roche’s subsequent POC BNP test introduction.
4)Bayer and Abbott Labs 2003 entry into BNP assay market.
5)Continued academic/research studies confirming the failures of BNP as an effective marker of CHF.