Cross‑Study “Establishment” Claims in Scientific Advertising: Literal Falsity by Necessary Implication Absent Proven Comparability, and No Lanham Act Damages Without Proof of Actual Deception

Cross‑Study “Establishment” Claims in Scientific Advertising: Literal Falsity by Necessary Implication Absent Proven Comparability, and No Lanham Act Damages Without Proof of Actual Deception

Case: CareDx, Inc. v. Natera, Inc., Nos. 23-2427, 23-2428 (3d Cir. Aug. 28, 2025) (not precedential)

Court: United States Court of Appeals for the Third Circuit

Panel: Shwartz, Matey, Fisher, JJ. Opinion by Shwartz, J.

Introduction

This appeal arises from a high-stakes false advertising battle in the precision diagnostics market for noninvasive tests that detect kidney transplant rejection by measuring donor-derived cell-free DNA. CareDx (maker of the AlloSure test) sued Natera (maker of the Prospera test) under the Lanham Act, the Delaware Deceptive Trade Practices Act (DTPA), and Delaware unfair competition law, challenging a suite of superiority claims (Claims A–J) that Natera disseminated via press releases, websites, and physician-facing materials.

Natera’s advertising touted Prospera’s performance by placing results from its own Sigdel study alongside CareDx’s Bloom study, asserting greater sensitivity, specificity, area under the curve (AUC), negative predictive value (NPV), and strengths across patient subgroups (including pediatrics). A jury found nine claims literally false, willful, and awarded substantial actual and punitive damages. The District Court denied Natera’s post-trial bid for judgment as a matter of law (JMOL) on liability but vacated damages for lack of proof of actual deception and reliance, entered an injunction by stipulation, and later—on remand—found literal falsity as to all nine claims at issue.

The Third Circuit affirmed across the board: it upheld the finding of literal falsity and the injunction, but agreed that CareDx failed to prove actual deception or reliance necessary for damages, which also doomed the unfair competition verdict and punitive damages.

Summary of the Judgment

  • Liability affirmed: The court held there was sufficient evidence for a reasonable jury to find that Claims A, B, C, D, E, F, G, H, and J were literally false by necessary implication. Although the numeric results reported from the Sigdel and Bloom studies were accurate in isolation, the side-by-side presentations necessarily implied that the studies were comparable and established Prospera’s superiority. Because the studies were methodologically divergent and not shown to be reliably comparable, the superiority claims were literally false.
  • Injunctive relief sustained: With literal falsity of unambiguous claims, deception is presumed for injunctive relief. The court therefore affirmed the stipulated injunction barring the challenged ads.
  • Damages vacated: To recover damages under the Lanham Act, a plaintiff must prove actual deception and reliance by a portion of the buying public. The court found CareDx’s showing—generalized references to confusion, campaign success, and willfulness—insufficient to establish that any customers were actually deceived or that they relied on the claims in choosing Prospera over AlloSure. The actual and punitive damages awards were properly vacated.
  • State law claims: Because CareDx failed to prove causation and harm tied to customer reliance, its Delaware unfair competition claim failed as a matter of law, and punitive damages fell with it. The DTPA liability stood, tracking the Lanham Act analysis, but without damages.

Analysis

Precedents Cited and Their Role

  • Groupe SEB USA, Inc. v. Euro-Pro Operating LLC, 774 F.3d 192 (3d Cir. 2014): Reinforced the two-step literal falsity analysis: determine if a claim is unambiguous and then whether that unambiguous message is false. Also supported reading statements in context (proximity and structure can imply superiority) and the presumption of deception for injunctive relief upon proof of literal falsity.
  • Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharms. Co., 290 F.3d 578 (3d Cir. 2002): Recognized literal falsity by “necessary implication” when the overall message is as clear as an explicit statement. The panel drew on this to treat side-by-side study presentations as necessarily implying comparability and superiority.
  • Castrol, Inc. v. Quaker State Corp., 977 F.2d 57 (2d Cir. 1992): Established the “establishment claim” framework: where an advertiser claims tests prove superiority, a challenger can prove literal falsity by showing the tests are not sufficiently reliable or do not establish the proposition asserted. The Third Circuit applied this framework.
  • Castrol Inc. v. Pennzoil Co., 987 F.2d 939 (3d Cir. 1993): Although principally cited for contextual approach to claim meaning, the panel acknowledged the establishment/non-establishment dichotomy and treated the challenged statements as establishment claims.
  • Apotex Inc. v. Acorda Therapeutics, Inc., 823 F.3d 51 (2d Cir. 2016): Held that combining text and figures derived from different, non-comparable data metrics can render an advertisement literally false. The Third Circuit analogized the cross-study comparisons here to that mismatch.
  • Parkway Baking Co. v. Freihofer Baking Co., 255 F.2d 641 (3d Cir. 1958): Longstanding Third Circuit rule: damages require proof of actual deception and some customer reliance on the falsehood. This rule drove the vacatur of damages.
  • Pernod Ricard USA, LLC v. Bacardi U.S.A., Inc., 653 F.3d 241 (3d Cir. 2011): Confirmed that once literal falsity of an unambiguous claim is shown, deception is presumed for injunctions (but not for damages).
  • Schering-Plough Healthcare Prods., Inc. v. Neutrogena Corp., 702 F. Supp. 2d 266 (D. Del. 2010): DTPA elements track Lanham Act false advertising; used to align state and federal analyses.
  • Total Care Physicians, P.A. v. O’Hara, 798 A.2d 1043 (Del. Super. Ct. 2001): Elements of Delaware unfair competition—expectancy, wrongful interference, causation, harm—were applied; absence of proof of causation and harm compelled judgment for Natera.
  • Agilent Techs., Inc. v. Kirkland, 2009 WL 119865 (Del. Ch. Jan. 20, 2009): Emphasized that compensable harm requires derailment of a business expectancy caused by wrongful acts; bolstered the causation analysis.
  • Unitherm Food Systems, Inc. v. Swift-Eckrich, Inc., 546 U.S. 394 (2006): Cited regarding the propriety of remanding to allow the trial judge to assess literal falsity of the remaining claims in the first instance.

Legal Reasoning

The court’s reasoning proceeds in two integrated tracks: (1) liability (literal falsity under the Lanham Act/DTPA) and (2) remedies (injunctive relief versus damages), with knock-on effects for Delaware unfair competition.

1) Literal falsity of “establishment” claims by necessary implication

The panel treated Natera’s superiority claims as establishment claims—i.e., claims asserting that tests or studies prove superiority. For such claims, a plaintiff proves literal falsity by showing the cited tests are not sufficiently reliable to support the asserted proposition or do not actually establish it. Here, Natera juxtaposed Sigdel (Prospera) and Bloom (AlloSure) study results in side-by-side graphics and text to proclaim “higher sensitivity,” “higher AUC,” “3x fewer rejections missed” (NPV), “more sensitive and specific,” “unparalleled precision,” and “highly sensitive across” patient subgroups (including pediatrics). This presentation necessarily implied that the studies were comparable and that they proved Prospera’s superiority.

The evidence, however, showed material methodological differences: Bloom was multi-site and prospective, matched to biopsies under the standard of care, while Sigdel was single-site and retrospective using banked samples. Expert and party witnesses (from both sides) repeatedly acknowledged that the studies were not “head-to-head,” that their confidence intervals overlapped for AUC, and that comparative claims like “more sensitive and specific” were unsupportable. Internal Natera documents admitted awareness that Prospera’s specificity was lower than AlloSure’s and that differences in patient selection affected sensitivity.

Against this record, the court concluded a reasonable jury could find:

  • Claim A (“More sensitive and specific”): Literally false because both studies showed Prospera’s specificity was lower, and the studies did not reliably establish superiority on either metric.
  • Claim B (“better performance in correctly classifying patients with active rejection”): Literally false by necessary implication; the side-by-side “89% vs. 59%” sensitivity asserted comparability that the record did not support.
  • Claims C, D, G (AUC and sensitivity superiority): Literally false as the AUC differences were not statistically significant and, per defense witnesses’ admissions, could not be deemed “higher” or “superior” across non-comparable studies.
  • Claims E, F (NPV; “3x fewer rejections missed”): Literally false because NPV depends on prevalence and cohort; Natera’s recalculations did not cure the non-comparability of cohorts and methodologies.
  • Claim H (“Unparalleled Precision. Optimized by Prospera”): Unambiguous when read with the adjoining quadrants (sensitivity and NPV claims), and literally false for the same reasons.
  • Claim J (“Highly sensitive across … patients,” including “Below 18 years of age (n=49)”): Literally false because there were no pediatric rejection events; the study could not establish sensitivity in that subgroup, yet the slide unambiguously claimed such performance.

Importantly, the court reiterated that for literal falsity of unambiguous claims—especially by necessary implication—consumer perception evidence is not required. The analysis turns on the message conveyed and its truth or falsity, not on how consumers actually received it.

2) Remedies: injunctions versus damages

While literal falsity sufficed to sustain injunctive relief (with deception presumed), damages require more. Relying on Parkway Baking, the court emphasized that a plaintiff seeking monetary relief must show that the falsehood actually deceived a segment of the purchasing public and that consumers relied on it in making purchasing decisions—i.e., some evidence of causation and reliance.

CareDx’s proof fell short. Generalized testimonies about “confusion,” evidence of campaign effort and success, and internal acknowledgments of willfulness did not demonstrate that any physician or transplant center chose Prospera because of the challenged claims or would have chosen AlloSure absent those claims. The panel expressly declined to infer actual deception for damages from willfulness alone and noted that other circuits’ rebuttable presumptions of deception for deliberate falsehoods have not been adopted in the Third Circuit for monetary relief.

3) Delaware unfair competition and punitive damages

Delaware unfair competition requires proof of interference with a legitimate business expectancy, causation, and harm. Because CareDx did not prove that Natera’s false claims actually diverted business through customer reliance, the unfair competition verdict could not stand. Punitive damages, awarded only on the unfair competition claim, fell as a consequence.

Impact and Practical Significance

A. Scientific advertising and cross‑study comparisons

  • Establishment claims carry heavy evidentiary risk: If you say “studies show” or juxtapose metrics in a way that implies equivalence, you assume the burden of showing the studies are methodologically comparable and that differences are statistically significant in the relevant metrics.
  • Side‑by‑side visuals are powerful—and perilous: Charts and bullets that place non-head-to-head results next to each other will likely be read as head-to-head proof of superiority. Absent rigorous comparability, such presentations risk literal falsity by necessary implication.
  • Metric‑specific cautions:
    • Specificity: Do not imply superiority if your own cited studies show lower specificity. Internal documents acknowledging this discrepancy may be probative of willfulness.
    • AUC: Without statistical testing demonstrating a significant difference, claims of “higher” or “superior” AUC are vulnerable.
    • NPV: Because NPV depends on disease prevalence and cohort composition, advertisers should avoid cross-study NPV comparisons unless they can validly normalize both prevalence and populations—and even then, cohort differences can defeat comparability.
    • Subgroup claims: Do not claim sensitivity in subpopulations (e.g., pediatrics) with zero events—or insufficient power—to estimate the metric.
  • Disclaimers and recalculations may be inadequate: Recomputing metrics to harmonize prevalence or adding fine print will not rescue an inherently non-comparable cross-study comparison when the overall message necessarily conveys proven superiority.

B. Remedies strategy under the Lanham Act

  • Injunctions are achievable with literal falsity: Plaintiffs can obtain injunctive relief without consumer surveys when the claim is unambiguous and literally false.
  • Damages require proof of actual deception and reliance: Plaintiffs must go beyond “campaign success” or “confusion” and present concrete evidence linking the false claims to purchasing decisions. Effective proofs may include:
    • Physician or purchaser testimony that the challenged claim drove adoption or switching;
    • Contemporaneous CRM/sales notes attributing switching to the specific claim;
    • Surveys tailored to purchasing decision-makers showing reliance on the exact claim;
    • Econometric analyses that isolate the causal impact of the challenged campaign on sales, with appropriate controls.
  • No presumption of deception for damages from willfulness: The Third Circuit has not adopted out-of-circuit presumptions of deception for deliberate falsehoods in the damages context.

C. Delaware unfair competition

  • Proof tracks Lanham Act causation: Without evidence that false ads actually derailed CareDx’s business expectancy by influencing customer choices, unfair competition fails. Punitive damages tied to that claim cannot stand.

Complex Concepts Simplified

  • Literal falsity vs. misleadingness: A statement is literally false when its unambiguous message is untrue. If a claim is literally true but misleading, a plaintiff typically needs evidence (often surveys) that consumers were deceived. Here, the court found literal falsity.
  • Necessary implication: Even if not expressly stated, the overall ad can unavoidably communicate a specific message (e.g., that two studies are comparable and prove superiority). If that necessary message is false, the ad is literally false.
  • Establishment claims: Claims that “tests prove” superiority. To defeat them, a challenger can show the tests are unreliable for the asserted proposition or do not establish it.
  • Head‑to‑head comparison: A study design that directly compares products under the same conditions. Cross-study comparisons are risky unless you can demonstrate methodological equivalence and statistical significance.
  • Sensitivity and specificity: Sensitivity measures how often a test correctly identifies true positives (avoids false negatives). Specificity measures how often a test correctly identifies true negatives (avoids false positives).
  • AUC (Area Under the Curve): A combined index of sensitivity and specificity across thresholds. A higher AUC suggests better discrimination, but differences must be statistically significant to claim superiority.
  • NPV (Negative Predictive Value): The probability that a negative test result is truly negative. NPV depends on disease prevalence and sample composition; cross-study NPV comparisons are fraught without rigorous normalization and comparable cohorts.
  • Prospective vs. retrospective studies: Prospective studies define the cohort and then collect outcomes; retrospective studies analyze existing samples and records. Differences in design affect comparability.
  • Actual deception and reliance (damages): For money damages, a plaintiff must show that customers were actually deceived and relied on the falsehood when deciding to purchase.

Conclusion

Although not precedential, the Third Circuit’s decision offers a clear and practical roadmap for litigants and marketers in scientifically complex markets:

  • On liability: Side-by-side presentations of non-head-to-head studies that imply proven superiority are fertile ground for literal falsity findings. Advertisers must ensure methodological comparability and statistical significance before making “establishment” superiority claims. Assertions about subgroups require sufficient data to estimate performance.
  • On remedies: Literal falsity readily supports injunctive relief, but plaintiffs seeking damages must marshal concrete proof that the falsehood caused customers to buy the competitor’s product. Willfulness, confusion in the air, or general sales growth is not enough in the Third Circuit.
  • On state law: Delaware unfair competition claims will rise or fall with proof of causation and harm linked to customer reliance on false statements; punitive damages will not survive where that causal chain is missing.

The decision thus tightens the leash on scientific superiority advertising that relies on cross-study comparisons, and it underscores a sharp evidentiary divide between the requirements for injunctive relief and those for damages under the Lanham Act. Companies in the life sciences and other data-driven sectors should take particular care: when your message is “studies show,” courts will demand that the studies truly show what you say they show—and in a way that is methodologically comparable, statistically sound, and not merely suggestive.

Case Details

Year: 2025
Court: Court of Appeals for the Third Circuit

Comments