The SEC announced last week that it had filed a complaint against Cassava Sciences, Inc., a “pharmaceutical company with one primary drug candidate, PTI-125, a potential therapeutic for the treatment of Alzheimer’s disease,” for misleading statements about the results of a Phase 2 clinical trial for the potential therapeutic. Also charged in the complaint were the company’s founder and former CEO and its former Senior Vice President of Neuroscience. The complaint highlights and analyzes a number of misleading statements and omissions—an analysis that could be instructive for companies reporting on clinical trial results. In a related Order, the SEC also charged an associate medical professor at the CUNY, who was a consultant and the co-developer of the therapeutic, with manipulating the reported clinical trial results. The company agreed to pay a civil penalty of $40 million. The former CEO and former Senior VP agreed to pay civil penalties of $175,000 and $85,000, respectively, and to officer-and-director bars of three and five years. The consultant agreed to pay a civil penalty of $50,000. They were all charged with violating the antifraud provisions of the federal securities laws; the company was also charged with violating the reporting provisions. It’s been widely reported that, after the announcement of the settlement, the stock price fell by almost 11%. PTI-125 is now reported to be in Phase 3 clinical trials.
According to the complaint, Cassava announced its final Phase 2b results for its primary drug candidate, PTI-125, in September 2020. In the release, the company claimed that “PTI-125 taken for 28 days significantly improved every measured biomarker for Alzheimer’s disease compared with subjects who took a placebo. Cassava also announced that patients who took PTI-125 showed improved cognition compared to patients who took the placebo.” The company then raised $260 million in new funding. The SEC charged that many of the claims in the release were false.
Background. As alleged, the VP and the professor co-discovered PTI-125. In a 13-patient Phase 2a trial, the professor’s results showed “directional improvements in multiple biomarkers, suggesting that the drug may be causing changes in biomarker levels.” The press release and SEC filings announcing those results specifically disclosed that the professor conducted the Phase 2a tests. The Phase 2b trial with 64 patients was to be conducted as a double-blinded clinical trial. Phase 2b participants also took a “battery of cognition tests before treatment and then again after 28 days to assess any changes in cognition.”
Cassava hired an outside European lab to test the samples, but apparently, the lab could not perform all the necessary tests, so the company asked the professor to test for two of the nine biomarkers. None of the European lab tests showed a meaningful effect compared with the placebo or even an effect consistent with the professor’s Phase 2a results. The company filed a Form 8-K disclosing the failure of the Phase 2b study to meet primary endpoints. However, some employees and “outside scientists expressed concern with the European laboratory’s results due to unexplained data variability,” leading the company to announce that it planned to re-analyze the biomarkers. The stock price fell from $7.61 a share to $1.63.
Around the same time, the SEC alleged, the VP sent to the professor data from the biostatistics company showing min, max and change from baseline data for evaluation of the European laboratory’s data; the SEC also charged that the VP knew that the professor had “individual test results identified by patient identification code for the two biomarkers that he had tested for Round 1,” enough information to allow him to “unblind himself to roughly a third of the patients in Phase 2b.” The company then requested that the now-unblinded professor perform a re-analysis of the European lab results and certain additional tests: the tests now “showed statistically significant improvement in all biomarkers in the treatment groups as compared with the placebo group.”
The clinical trial protocol included cognition testing in addition to biomarker analysis. The cognition testing showed no improvement in episodic memory or patient cognition. However, the SEC claimed, the unblinded VP then “first removed patients with missing data and patients who did not take the drug and then engaged in what she described as a ‘sensitivity analysis’ where she removed the highest performing patients and lowest performing patients by baseline score cutoffs across all groups until the results appeared to show separation between the placebo group and the treatment arms.” As alleged, she “ultimately removed 40% of the patient population” from the analysis. According to the SEC, until the company filed a Form 8-K in July 2024, it did not disclose the full results of the cognition tests, but instead reported only the VP’s sensitivity analysis results as the final results, although in some disclosures, the company did say that it had removed from the calculation “the most and least impaired subjects.”
Cognitive testing also included spatial working memory testing as a secondary outcome. But, the SEC claimed, neither of the results for the two key spatial working memory measurements identified by Cassava and the test developer showed a clear benefit. The company did not report those results; instead, the VP “selected, and Cassava reported, another measurement after she received unblinded results—SWM total errors.” The company failed to report the results of the other analyses of trial participants’ spatial working memory, which did not show a clear benefit in the treatment arm.
The company’s September press release announcing the Phase 2b trial results said that treated patients “showed statistically significant (p<0.05) improvements in biomarkers of disease pathology, neurodegeneration and neuroinflammation, versus Alzheimer’s patients who took placebo.” The release also claimed that “Alzheimer’s patients treated with [PTI-125] showed directional improvements in validated tests of episodic memory and spatial working memory, versus patients on placebo.” Notably, the company claimed that “[b]ioanalyses were conducted under blinded conditions to eliminate any possibility of bias” and, without identifying the professor, claimed that an “academic lab generated final results.”
As alleged, the company also released a presentation, claiming that “the biomarker results from Round 1 [were] ‘invalid data,’ in part because some biomarkers in the placebo group ‘moved in opposite directions,’ suggesting simultaneous improving and worsening in the same patients, and that changes in biomarkers in placebo patients were uncorrelated. The presentation claimed that changes in biomarkers from Round 2 were correlated, and therefore valid.” The presentation also “claimed that the Phase 2b was a randomized, double-blind, placebo-controlled, multicenter clinical study.” In the presentation, the company claimed that PTI-125 “appears to stabilize or improve memory,” including improvements in both episodic and spatial working memory as against placebo. Although the presentation did disclose that “‘effect sizes vs. placebo were calculated by Hedge’s g after removing the most and least impaired subjects across all groups by baseline score’ (emphasis added), the presentation did not disclose that the episodic memory results were from a sensitivity analysis and not from the full population.” Nor did the presentation explain that “the episodic memory results were calculated only after removing 40% of the study population,” that the average change in errors for the full episodic memory data set did not show similar directional improvement or that the “key spatial working memory measurements identified by Cassava and the test developer prior to unblinding showed no improvement.” The related conference call took a similar approach, omitting similar information and defending the second bioanalysis as a valid analysis. The company’s stock rose from $3.40 to $8.41.
In November, the company filed its Form 10-Q, which included results from Phase 2b and “continued to claim that Phase 2b was ‘double-blind’” and made claims similar to those in the press release regarding the performance of PTI-125, without disclosing that the episodic memory results excluded data from 40% of the Phase 2b participants. Subsequent SEC filings likewise did not disclose that “any patient data had been removed from episodic memory analysis” or that “the other key spatial working memory measurements showed no improvement compared with placebo.” The company also announced that it planned to proceed to Phase 3. The company offered and sold securities during this period, including under a shelf registration statement that incorporated SEC filings by reference.
According to the complaint, in August 2021, two individuals filed a citizen petition with the FDA, claiming that the professor had “manipulated images of tests known as western blots to various academic journals as well as, collaboratively with Cassava, to the National Institutes of Health to support grant applications.” Both the FDA and Cassava conducted audits of his lab. Cassava’s audit found a number of “critical issues with the laboratory” and with the professor’s practices, including a “‘lack of experiment logbooks/notebooks for all study/research work being performed.’” The company then determined that the professor’s lab was “‘considered unacceptable and temporarily not qualified to provide biomarker analysis and research services for any future Cassava studies,’ (Emphases in original)” and concluded that the lab should not perform further analysis until a close-out audit. Although the CEO and VP were aware of these findings, the company “did not sever its relationship “ with the professor until June 2024, nor did it inform investors of its internal findings regarding the professor.
Charges. The SEC charged that the defendants violated Securities Act Sections 17(a)(2) and 17(a)(3). In its complaint, the SEC alleged that they made the following misleading statements about the results of the Phase 2 clinical trial:
- The claim that “[b]ioanalyses were conducted under blinded conditions to eliminate any possibility of bias” negligently omitted material information that the professor who ran clinical testing for Phase 2b had been negligently given information by the VP sufficient to allow him to partially unblind himself. The information was material, not only because of the claim of lack of bias, but especially because the professor who performed the bioanalyses was the “coinventor of the drug and an individual with a financial stake in its success.”
- The company negligently failed to disclose that the professor who performed the bioanalyses for the Phase 2b trials was the co-inventor of PTI-125, a Cassava consultant and a member of Cassava’s Scientific Advisory Board. The company’s public filings did not name the professor and just characterized his laboratory as an “academic lab,” which “although technically correct, was incomplete and misleading” because it did not disclose that the “scientist performing the analysis had a conflict of interest due to his professional and financial ties to Cassava.”
- The company and the CEO failed to disclose that an audit conducted by the company of the professor’s laboratory in 2022 found that his lab was “unacceptable and temporarily not qualified to provide biomarker analysis and research for services for any future Cassava studies.”
- The company and the VP negligently failed to fully disclose that the VP removed a large portion of patients in reported cognition data, excluding from the reported episodic memory results data from 40% of patients who completed the cognition test. It was also not disclosed that she was “unblinded when she decided which patients to exclude from the reported results.” In addition, the company and the VP failed to disclose that the average change in errors for the full episodic memory data “showed no similar directional improvement” when compared with placebo.
- The company and the VP also negligently failed to disclose that the “spatial working memory measurement reported in the Phase 2b results as showing cognitive improvement of up to 46%” was a measurement selected by the VP after she was unblinded. In addition, they failed to disclose that “other spatial working memory results, including measurements identified as ‘key’ prior to unblinding, did not show directional improvement in patients receiving PTI-125 compared with placebo.”
According to the SEC, these misstatements were material because PTI-125 was the company’s “primary asset and its only realistic potential source of revenue.” In addition, several investment banks had “advised the company that Cassava would be unable to raise sufficient capital for Phase 3 testing until announcing the Phase 2b results. Following the Phase 2b result disclosures, the company’s stock price rose dramatically, enabling the company to raise hundreds of millions of dollars for its Phase 3 testing.”
The SEC also charged that the company violated Section 13(a) of the Exchange Act and Rules 13a-1, 13a-11 and 13a-13 and 12b-20 as a result of misstatements in its disclosure documents, including through disclosures incorporated by reference. Further, the CEO filed quarterly certifications declaring that the disclosures were accurate.
In a separate Order, the SEC brought separate settled charges against the professor. According to the Order, the professor “manipulated the Phase 2b biomarker data, using the knowledge he gained through the unblinding process to show an exaggerated response to the treatment arms as compared to the placebo group.” The SEC charged that the professor violated Securities Act Sections 17(a)(1) and (3) and Exchange Act Section 10(b) and Rules 10b-5(a) and (c), and caused Cassava to violated Sections 17(a)(2) and 17(a)(3). Section 17(a)(1) and Rules 10b-5 require a showing of scienter, which can be satisfied by a showing of reckless conduct.
As noted above, the company agreed to pay a civil penalty of $40 million. The former CEO and former Senior VP agreed to pay civil penalties of $175,000 and $85,000, respectively, and to officer and director bars of three and five years. The consultant agreed to pay a civil penalty of $50,000.