Thursday, September 29, 2011

generic drug seller avoids injunction

Intendis, Inc. v. River's Edge Pharmaceuticals, LLC, 2011 WL 4442849 (D. Md.)

Intendis makes NeoBenz Wash Plus Pack, and defendants make the generic BP 7% Wash External Kit, both of which are prescription drugs. Intendis alleged that defendants make false or misleading statements about the BP wash on its product insert page and package label, and in marketing information distributed to national pharmacies and pharmacy databases. Intendis sought a preliminary injunction requiring defendants to submit supplemental information to the national pharmacy databases to correct or clarify the false or misleading statements made in their prior submissions to these databases and to post statements on their website; the court denied the motion.

The NeoBenz pack contains NeoBenz Micro Wash, a 7% benzoyl peroxide wash, and NeoBenz Micro SD, pre-filled sponge applicators with a 5.5% benzoyl peroxide cream. These are to be used in tandem to treat acne. The product insert states that the active ingredient, benzoyl peroxide, is incorporated into patented porous microspheres to "provide gradual release of active ingredient into the skin and absorb natural skin oils,” but it doesn’t disclose the specific release rate.

The BP Wash Kit contains a 7% benzoyl peroxide wash and a 5.5% benzoyl peroxide cream with sponge applicators. The product insert states that the active ingredient, benzoyl peroxide, is "incorporated into a series of microscopic concentric vesicules of oil and water .... [that] results in the release of the active ingredient into the skin over a sustained period of time." Rather than using the NeoBenz microsponge delivery system, it contains the skin conditioner Incroquat OSC, “a self-emulsifier and polymeric drug delivery system that incorporates the benzoyl peroxide into vesicules of oil and water.”

Intendis discovered the BP Wash Kit via an alert from Wolters Kluwer, a national pharmaceutical database. Wolters Kluwer, like other national pharmacy databases, classifies drugs based on: (1) active ingredient, (2) route of administration, (3) strength of active ingredient, and (4) dosage of active ingredient. The databases assign each drug a generic product identifier code ("GPI" or "GCN"). “Pharmacies rely on a drug's GPI code to make dispensing decisions and will substitute lower priced generic drugs for name-brand drugs with the same GPI unless otherwise prohibited by state law.” Pharmacies thus began to substitute the BP Wash Kit for NeoBenz prescriptions because the price of the former was significantly lower (pharmacies paid $96.47 versus 122.19 as of July 2010).

Intendis alleged a number of falsehoods: (1) the BP wash is a 7% benzoyl peroxide wash; (2) the BP Wash Kit provides sustained release of the active ingredient through the use of microscopic vesicules; (3) the BP Wash Kit contains microscopic vesicules; and (4) the BP Wash Kit has a shelf life of twenty-four months. Statements (1)-(3) are in the product insert and were separately submitted to the national pharmacy databases. The 24-month shelf life claim was in deal sheets submitted to national pharmacy chains and pharmacy database organizations to announce the product release. There were no explicit claims of equivalence or substitutability.

There’s an industry standard for gel and lotion benzoyl peroxide formulations: a gel must contain between 90% and 110% of the claimed label strength of active ingredient, while a lotion must contain between 90% and 125%. There’s no industry standard for washes and creams. Intendis had a testing company test two lots of the BP wash and cream. For the wash, one lot had between 88.0% and 90.0% of the claimed strength, and the other had between 80.1% and 80.5% of claimed label strength. For both lots of the BP cream tested, the results were between 93.8% and 95% of the strength claimed on the label. The margin of error was plus or minus 2%.

Defendants relied on the manufacturer’s tests. These tests on the wash showed results from 98.6% to 101% of label strength. In prior tests, a sample kept at room temperature had remained at 100% of label strength for twelve months, and a sample subjected to accelerated aging had decreased to 98.6% of label strength after three months. In the manufacturer’s tests of the BP cream, one lot started at 100% of label strength and remained at 100% after three months of accelerated aging and six months at room temperature. For other lots not tested by Intendis, the results were all “close, if not equal, to 100% of label strength and well within the industry standard monographs for gels and lotions.”

Dow (the tester Intendis used) calculated the shelf life of the BP wash and cream based on the idea that the shelf life expired when the active ingredient went below 90% of label strength. Using a steady rate of degradation for benzoyl peroxide, it calculated that one lot of the wash had a shelf life of less than two months and the other had an 8-month shelf life. For the cream, Dow calculated a shelf life of nine months and eleven months.

The parties also conducted their own tests of release rates. “The Franz Cell test is used to measure the in vitro permeation rate of active ingredients in solution. The test utilizes a Franz Cell apparatus that contains two distinct chambers separated by a membrane designed to represent human skin. The test solution is placed into the top chamber and measurements are taken from the bottom chamber at regular intervals to measure the amount of active ingredient that has diffused through the membrane.” Defendants’ expert’s tests showed that, on average, the BP wash released at a rate that was 15% slower than the NeoBenz wash.

Dow also conducted a Franz Cell test, but it was deemed inadmissible at the hearing for lack of proper foundation for authentication; this test was supposed to show that the BP wash released at a 50% slower rate.

Intendis had similar troubles introducing photographs that purportedly showed that, at high levels of magnification, microscopic vesicules couldn’t be seen in the BP products. Defendants submitted expert testimony that the photographs did not prove or disprove the presence of concentric vesicules of oil and water because the photos' exposure and lighting conditions rendered them inconclusive. Defendants’ expert opined that “one could infer the existence of concentric vesicules in the BP wash from the fact that it allows for extended release of the active ingredient and the fact that it contains Incroquat OSC.”

Intendis was able to show lost sales, though there’s now a third generic competitor that makes predictions about the impact of an injunction more complicated—if the BP product were enjoined, its sales might go to the new competitor instead of to River’s Edge.

In a footnote, the court explained that, though defendants initially argued that the product inserts weren’t commercial advertising or promotion, the evidence showed that the inserts were submitted to the databases as a way to advertise the BP wash kit.

Literal falsity of 7% benzoyl peroxide claim: Intendis only identified one lot with less than 90% of the claimed label strength, considering the margin of error, and the court wasn’t willing to find likely success on this record.

Sustained release/microscopic vesicules: The BP wash provided sustained release; the questions surrounded its rate and whether the release was accomplished by using microscopic vesicules. Defendants didn’t claim any particular rate or equivalence with NeoBenz’s rate. Thus there was no literal falsity as to rate. The evidence on vesicules was also insufficient; even had the photos been admitted, Intendis wouldn’t have won, because they did appear to contain at least some vesicules and because of defendants’ expert’s testimony.

Shelf life: Dow’s methodology for extrapolating shelf life was unreliable “because it was based on very limited data and incorporated assumptions that were unsubstantiated.” The FDA prefers a trend analysis “where one takes data over a period of time, takes the log of that activity, and then extrapolates over a two-year time period.” The manufacturer’s data was entitled to more weight because it used longer-term data and used an FDA-approved method of calculation. Intendis didn't successfully show falsity.

Intendis also argued that defendants failed to substantiate their establishment claims of product strength and shelf life. Defendants argued that there was no express claim that tests were conducted; to the extent their statements implied that tests were conducted, that’s only meaningful if a single test can prove truth; otherwise there’s no establishment claim. (I take the point that if it’s not clear what the tests were, then a broad range of tests might count as substantiation, but in a litigation context I have no idea what it would mean for a single test to be able to prove truth—it’s always possible to create some FUD about a particular test.)

The court rejected the establishment claim theory. First, the court said, this analysis is typically applied in comparative advertising cases. (What? “Tests prove” is a credible statement whether it’s “tests prove our product works” or “tests prove our product works better”; that’s why advertisers use it instead of “our product works.” Comparison has nothing to do with the underlying justification for the doctrine, which is that “tests prove” gives consumers a material reason to believe the underlying claim.) A defendant isn’t required to provide reliable test data for every factual statement in an ad. (True enough, but some claims necessarily imply scientific substantiation and some don’t; drug claims seem precisely the kind that fit in the former category.) Without an explicit tests prove or comparative superiority claim, there was no establishment claim.

Second, the claim failed on its merits: defendants provided test results confirming the percentage of active ingredient and shelf life. “The mere fact that Intendis conducted similar tests and reached different results is not sufficient to challenge the methodology and results provided by Defendants.”

The court also rejected Intendis’s arguments that defendants deliberately made false statements with the intent and effect of causing pharmacies to treat its BP wash as a substitute for NeoBenz, and made false claims of substitutability. There was no evidence that defendants ever claimed substitutability. (Given the modern pharmaceutical database system, can there have been any other intended result when the defendants submitted a product claimed to be the same across all four relevant attributes?) The data sheets sent to national buyers and retail chains explicitly disclaimed bioequivalence, and pharmacy databases don’t list the products as equivalents.

Comment: Is equivalence the same as substitutability? It seems not, by this description of the system, in which case I’d say there’s a necessarily implied claim of substitutability—but Intendis seems not to have proven that the substitutability claim is false. I think my reading is consistent with the court’s conclusion that “[i]t is not actionable to seek to obtain market share by manufacturing a comparable product that costs less if no false statements are made” (emphasis added).

Without likely success, there wasn’t going to be a preliminary injunction, but the court considered irreparable harm anyway. Intendis argued that, because the parties compete directly, defendants’ false advertising was likely to cause irreparable harm. Past cases applying a presumption of irreparable harm in literal falsity cases, however, predates Winter, in which the Supreme Court emphasized that a plaintiff must always show the likelihood of irreparable harm. Defendants argued that, because pharmaceutical sales are carefully tracked, it would be easy to calculate money damages, and that the entry of a third competitor would prevent Intendis’s market position from changing in any event.

Of course, Intendis couldn’t rely on the presumption of irreparable harm without showing literal falsity. Also, any lost sales could be calculated using information maintained by the national databases. Nor did Intendis submit evidence of less tangible forms of harm.

Naturally, the other factors didn’t favor Intendis either—there’s a public interest in avoiding deception, and in free competition for lower-priced alternatives, but in the end it was the absence of evidence of deception that was key.

No comments: