J Public Health Pol https://doi.org/10.1057/s41271-018-0124-1 ORIGINAL ARTICLE

Explaining the discontinuation of a non-tobacco nicotine project at Philip Morris: obstacles to innovation Zachary Cahn1 • Lindsay Eckhaus2

 Macmillan Publishers Ltd., part of Springer Nature 2018

Abstract This paper investigates the fate of the Capillary Aerosol Generator (CAG), a nicotine aerosol device resembling modern e-cigarettes, developed by Philip Morris (PM) in 1994. A debate has emerged as to why this product never made it to market. In Donovan et al. v. Philip Morris USA, Inc., the sides presented opposing views in their arguments. To explore the reasons for CAG abandonment, we searched the Truth Tobacco Industry Documents database for key terms including ‘Capillary Aerosol,’ ‘Project LEAP’ (the project under which the CAG was housed), and ‘Tony Howell’ (CAG inventor). The results show that regulatory, legal, and political concerns deterred CAG development. A lack of progress even after these concerns had abated seems to reveal a larger reluctance to innovate nontobacco products—perhaps due to concerns about competing against tobacco cigarettes. We also explored additional explanations such as technological barriers and lack of consumer interest. Keywords Tobacco
Cigarettes
e-Cigarettes
Capillary aerosol
Philip Morris

Introduction The emergence of e-cigarettes has become a seminal turning point in the evolution of the nicotine marketplace. Their importance derives from the combination of their appeal to consumers and their potential for reducing the hazard of tobacco use due & Zachary Cahn [email protected] 1

Department of Economic and Health Policy Research, American Cancer Society, Inc., 250 Williams St, Atlanta, GA 30303, USA

2

Department of Epidemiology, Emory University Rollins School of Public Health, 1518 Clifton Rd NE, Atlanta, GA 30322, USA

Z. Cahn, L. Eckhaus

to two key product features: (1) e-cigarettes do not contain tobacco (though the nicotine within is derived from tobacco) and (2) e-cigarettes do not involve combustion. Instead, battery power is used to turn a propylene glycol-based liquid into a smoke-like aerosol. Hon Lik, a pharmacist based in Hong Kong developed the modern e-cigarette in 2003, although it should be noted that other inventors applied for patents for similar devices dating back to 1944 [1]. Early prototypes had included electrical heating since the early 1963 [2], with the aim of delivering nicotine from 1964 [3]. With the exception of Favor Smokeless Cigarettes in the 1980s [4], none of these recreational non-tobacco nicotine products made it to market. Tobacco companies developed a similar category of product—heat-not-burn devices—in the 1980s and 1990s, most famously RJ Reynolds’ Premier introduced in 1988. These products differ from e-cigarettes in that they did contain tobacco. It is puzzling that the first viable e-cigarettes were invented outside of the tobacco industry, and that it took many years for the industry to get involved. Why were tobacco companies so slow to develop products like e-cigarettes? We address this question by focusing on a precursor to the modern e-cigarette, the Capillary Aerosol Generator (CAG) [5], that was developed at Philip Morris (PM) during the 1990s. PM developed the CAG as a part of ‘‘Project LEAP,’’ an offshoot of the Ideal Smoke Program. The purpose of Project LEAP was to ‘‘develop a smoking article that satisfies the mission of the Ideal Smoke Program and delivers a minimal set of well-defined compounds’’ [6]. The CAG was intended to be a ‘‘portable,’’ ‘‘safe’’ device that produces ‘‘a ‘smoke like’ aerosol containing… nicotine and/or other ingredients that are subjectively pleasurable’’ [7]. Like modern e-cigarettes, the CAG did not involve combustion and no tobacco was necessary. In addition, this product fell into the same business segment, and regulatory gap, as modern e-cigarettes—a recreational non-tobacco nicotine product. Although the technology in the device was not exactly the same as in modern e-cigarettes, the CAG is suitable for in-depth study because it appeared, both in contemporaneous accounts and in retrospect, to have a strong potential for achieving a reduction in hazard. So, why was CAG development ultimately abandoned? A previous investigation of the CAG by Dutra et al. concluded that it was abandoned largely due to ‘‘political and legal concerns’’ [8]. Our results elaborate on this and introduce some other important factors. There are no direct statements from PM or Altria, but this issue became a point of contention during the Donovan et al. v. Philip Morris USA, Inc. lawsuit [9]. During his deposition and witness statement, PM’s expert witness and former employee, Peter Lipowicz, mentions that technological barriers, consumer acceptance, and an US Food and Drug Administration (FDA) approach that disadvantaged non-tobacco nicotine products were chiefly responsible [10, 11]. Other scholars suggest that different factors were at play. David Hammond, a professor of public health with extensive tobacco control expertise, served as an expert witness for the plaintiff in the Donovan case. He says that CAG development was less attractive due to legal challenges about the addictiveness of nicotine and fears about overarching tobacco industry regulation, more so than regulatory barriers specific to non-tobacco cigarettes (Hammond D. Donovan et al. v. Philip Morris USA Inc., Expert Report of David Hammond, 2015, obtained through

Explaining the discontinuation of a non-tobacco nicotine…

personal communication, hereinafter ‘‘Hammond Expert Report’’). Additionally, Hammond and PM’s former employee William Farone, who served as Director of Applied Research for seven years, suggest that there were business-related obstacles to progress, namely a reluctance to develop and introduce products that would compete with tobacco cigarettes [12, 13]. This argument is compelling in large part because many, if not all, of the other cited obstacles had receded long before PM marketed its first e-cigarette in 2014. However, Dutra et al. argued that PM viewed non-tobacco nicotine products as complements to tobacco smoking rather than substitutes [8]. It is important to investigate the reasons for CAG abandonment to understand what drives and deters tobacco industry’s technological advancement. It can help us explain why technological development was slow in the past, plus it may provide insight into what we should expect from the industry going forward.

Methods We searched the Truth Tobacco Industry Documents database hosted at the University of California, San Francisco, for key words and names. Search terms included ‘‘Capillary Aerosol,’’ ‘‘Project LEAP,’’ ‘‘v2n’’ (a non-tobacco aerosol program at PM in the mid-2000s), and ‘‘Tony Howell’’—the CAG inventor. We selected relevant documents if they offered insight into PM’s concerns with the development of a non-tobacco nicotine consumer product. We deemed early documents defining the scope of the project to be relevant. In addition to PM’s internal communications, we reviewed depositions and expert witness testimony. The initial search yielded 1,274 documents, of which 171 were duplicates, leaving 1,103. Of these, 157 were deemed relevant to answering the research question, and 8 had strong relevance.

Obstacles to CAG Regulation Both Hammond and Lipowicz indicate that regulatory concerns inhibited progress on Project LEAP, but there remains some disagreement about which FDA regulatory decision was responsible (10,11, Hammond Expert Report). During CAG development in the mid-1990s, two distinct FDA regulatory decisions were in different ways problematic for non-tobacco cigarette development. First, tobacco companies had long suspected that The FDA might regulate recreational non-tobacco nicotine products as drugs [4, 14], while products that contained tobacco would likely elude FDA oversight. This situation was epitomized by the short history of ‘‘Favor Smokeless Cigarettes’’ (Favor) which were briefly on the US market in the mid-1980s [4]. Favor was a recreational alternative to tobacco cigarettes containing a nicotine-soaked paper ‘‘plug’’ without tobacco. Although the product was unable to win over most consumers, FDA policy toward non-tobacco nicotine products became a major obstacle to Favor as well. FDA determined that

Z. Cahn, L. Eckhaus

Favor was required to meet stringent pre-market approval standards because it was a ‘‘nicotine delivery system’’ and a ‘‘drug’’ [15]. This episode crystalized the notion among tobacco companies that non-tobacco nicotine would be regulated strictly [4, 16]. This strict approach to non-tobacco nicotine products is what Lipowicz emphasizes when he suggests that the threat of regulatory interference permeated the research and development process at PM [11]. He explains that PM saw no regulatory pathway for a non-tobacco product, and cites the failure of Favor, saying that the FDA was ‘‘not going to allow products like Favor.’’ Instead, ‘‘Philip Morris scientists focused on research and development for nonconventional cigarette technology rather than aerosols’’ [11]. FDA regulation specific to non-tobacco nicotine products was not the only way in which ‘regulation’ presented an obstacle to CAG development. The role of the FDA and how it viewed the regulation of tobacco cigarettes shifted markedly in the early 1990s, impacting the development of non-tobacco cigarettes. In response to pressure from public health groups, and catalyzed by a groundbreaking expose´ from ABC news revealing tobacco industry nicotine manipulation, FDA Commissioner David Kessler began a Herculean effort to assert FDA jurisdiction over all tobacco products, not just non-tobacco nicotine products. The FDA has the authority to regulate market access for ‘‘articles (other than food) intended to affect the structure or any function of the body’’ [17]. As such, Commissioner Kessler concentrated his investigative efforts on demonstrating that tobacco companies engaged in ‘‘manipulation and control’’ of the nicotine in tobacco cigarettes [18]. The tobacco industry, for its part, attempted to downplay the role of nicotine in the smoking experience and contested accusations that they were directly manipulating nicotine levels. In this context, development of the CAG or any other non-tobacco nicotine product would undermine PM’s attempts to deemphasize the role of nicotine in their conventional tobacco products ([4], Hammond Expert Report). Although each of these regulatory barriers to CAG development was important in the mid-1990s, they were less problematic after the FDA v. Brown and Williamson Supreme Court decision in 2000 [19]. The Court found that the FDA did not have the authority to regulate cigarettes ‘‘as customarily marketed,’’ which reduced the potential for non-tobacco cigarette development to spill over into the regulatory fight over all cigarettes. Although the question of how non-tobacco nicotine products would be treated was not resolved definitively, the text of the decision seemed to suggest that recreational nicotine products would be exempt from FDA regulation, whether or not they contained tobacco. This interpretation underpinned the logic behind a District Court decision in 2010, (Smoking Everywhere v. FDA) [20]. Overall, while both of these regulatory barriers were important, it appears that concerns about the regulation of all tobacco products were paramount. First, a PM document not attributed to a particular author, written during Kessler’s war on the tobacco industry, suggested that Project LEAP ‘‘complicates our efforts to resist the FDA’s attempts to regulate the tobacco industry’’ [21]. Second, it does not make sense to have begun CAG development in the first place if the non-tobacco

Explaining the discontinuation of a non-tobacco nicotine…

regulatory issue was seen as insurmountable. Nor would it have made sense to continue engaging in this effort through the mid-2000s [8]. In addition, FDA v. Brown and Williamson [19] created a legal standard that would allow for recreational nicotine products like e-cigarettes. Even if court challenges were required, PM did not have a history of reluctance to engage in legal battles to defend its interests. Litigation PM was reluctant to develop the CAG, a product designed to deliver nicotine, due to 1990s litigation around the question of nicotine addiction and manipulation. Prior to the 1990s, litigation surrounded the health hazards of smoking, and the industry successfully convinced jurors that smokers had assumed the risks of smoking and thus bore responsibility. In tobacco litigation in the early and mid-1990s, plaintiffs tried to undercut this argument by (1) focusing on the addictive nature of nicotine and (2) the industry’s role in engineering their products to be addictive. The most prominent of these cases was the Castano case that hinged on both of these arguments [22]. This new legal strategy shifted the scientific debate from the harms of smoking to the addictiveness of smoking—in simple terms, from tar to nicotine. Therefore, as development of the CAG reached a critical milestone in the mid-1990s, PM’s defense strategy in litigation and in seeking to thwart FDA regulation hinged on denying the addictiveness of smoking and the manufacturer’s role in engineering its products to cause and sustain this addiction. To make matters worse, in 1994, CEO of PM, William Campbell, along with the leaders of six other major tobacco companies, testified before the U.S. House of Representatives’ Subcommittee on Health and the Environment that ‘‘nicotine was not addictive’’ (Hammond Expert Report). In this environment, any evidence that PM was pursuing Project LEAP—which had the explicit goal of delivering a ‘‘subjectively pleasurable’’… ‘‘aerosol containing nicotine’’—had obvious potential to jeopardize PM’s public position and legal defense arguments. The sensitivity of the CAG project was evident in the secrecy through which it was conducted. Nicotine testing for Project LEAP was conducted at INBIFO [23, 24], a research facility in Germany purchased by PM in the 1970s, where highly sensitive research would not be discoverable in U.S. litigation or by U.S. health authorities [12]. Beginning in 1999, PM’s public statements around the harms of smoking shifted toward an admission that smoking was addictive [25]. PM’s legal position underwent a far subtler shift [26]. PM attorneys made the semantic argument that PM had not changed their evaluation of the degree to which smoking and nicotine were habit-forming, but rather that they had changed their position on the meaning of addiction to encompass the long-acknowledged habit-forming nature of smoking and nicotine [25–27]. This pattern of acquiescing to the term ‘addictive’ while attempting to downplay its significance by emphasizing its similarity to benign behaviors and while contrasting it with hard drugs continued in other cases. In 2005, USA v. Philip Morris USA, expert witnesses compared nicotine in cigarettes to caffeine in coffee [27].

Z. Cahn, L. Eckhaus

This shift in public statements did much to alleviate pressure on non-tobacco nicotine innovation. The subtler shift in liability defense arguments also diminished—though did not eliminate—complications for non-tobacco nicotine innovation. To be clear, the issues of addiction and the role of nicotine remained very sensitive to PM and other tobacco companies and the overall issue was, in no way, resolved by these shifts in tactics. However, inasmuch as the specific PM positions with regard to nicotine and addiction had been undermined by the simultaneous pursuit of non-tobacco nicotine products, the shift in PM’s approach to defense was very meaningful, even though the need for defense had not abated. To the extent that the new defense amounted to the argument that nicotine was indeed addictive, but not that addictive, this defense could be extended to nontobacco nicotine products as well. The residual complications arose from the fact that non-tobacco nicotine products might still invite scrutiny of PM’s true views of the importance of nicotine, suggesting a larger role than they were admitting. But PM lawyers now had a way of admitting and describing a role for nicotine in smoking that was far more flexible than the outright denials that had characterized their previous positions [26, 27]. Notably, even though PM’s legal arguments centering on the addictive properties of cigarettes no longer precluded the development of products like the CAG to the extent that they previously had, and PM’s public position posed far less of an obstacle, the development of a nicotine aerosol device progressed at an exceedingly slow pace. Work on such devices did not resume in force until 2004 [8], and the first PM e-cigarette was not marketed until 2014, long after litigation and PR defenses had receded as major obstacles to the innovation of non-tobacco products. Defensive approach to innovation CAG development was a defensive strategy to prepare for product release only if necessitated by competition or regulation, rather than by health concerns. In a 1991–1995 strategic plan, PM’s officials write that novel nicotine products might lead to the ‘‘obsolescence of the current product,’’ and therefore developing new devices should be high priority [28]. However, other documents from the same time period reveal uneasiness with these technologies, presumably because new smoking devices would compete with conventional cigarettes, which dominated the market. In a 1990 new product development meeting with PM’s management team, Executive Vice President for Operations, Mark Serrano, said, ‘‘we have to continue development of these articles [heated, non-burning tobacco, and aerosol generators] from a defensive position, if nothing else’’ [29], adding that the company might enter the market only if new products with ‘‘perceived health benefits’’ were sold by competitors [29]. Indeed, PM was closely monitoring the competitive landscape while Project LEAP was underway [30]. William Farone explains in his testimony: All of our research was done for defensive reasons…Philip Morris was preparing for a time when they were forced – by the government or by competitors in the marketplace – to make meaningful changes to their

Explaining the discontinuation of a non-tobacco nicotine…

products…These techniques were put ‘‘on the shelf’’ until they might become needed, unless they could lead to an immediate profit [12]. These ideas are echoed in the Hammond Expert Report in the Donovan case: ‘‘a ‘health cigarette’ was only acceptable if it had the potential to grow the market and did not threaten existing brands.’’ Consistent with product development for defensive reasons, PM did not introduce their own e-cigarette until after e-cigarettes had proven successful on the market for several years. Tobacco farmers Marketing of a non-tobacco CAG device might have jeopardized PM’s relationship with tobacco farmers. A 1998 document about the direction of Project LEAP states the company’s concern that ‘‘such a product may erode the support of tobacco farmers’’ [21]. As tobacco farmers were only mentioned once in all Project LEAP documents, were not mentioned by Lipowicz or Farone, and moreover, globalization of the tobacco market took off in the 1990s, concern about tobacco farmers is unlikely to have been a serious barrier to CAG development.

Additional purported obstacles Consumer acceptance Lipowicz offers one plausible barrier to CAG development, lack of consumer interest, due largely to the size of the device. While the CAG in the 1990s was not a hand-held device, Chrysalis, a PM-owned spinoff, created a portable prototype for the pharmaceutical industry in the 2000s [31]. Lipowicz argues that it was too large to be a consumer-accepted smoking device, because it was ‘‘the size of your hand’’ [10]. But, we could not find any evidence of consumer testing on the CAG to support this claim. When questioned about market research, Lipowicz mentions a different product, the heat-not-burn Accord, suggesting that its failure was related to ‘‘the weight of it’’ and ‘‘the fact that it was so not like a cigarette’’ [10]. Yet, the similarly sized Accord was both consumer tested and sold in stores, and it seems implausible for PM to have begun CAG development if they were deterred by the fate of Accord. The lack of evidence in both legal testimony and company documents suggests that PM never tested the CAG to determine if it could have market success, though the emergence of e-cigarettes in the 2000s demonstrated substantial consumer interest in a CAG-like product. Technological barriers Another potential obstacle to CAG development was the lack of sufficient technology to develop a portable smoking device. In 1994, Tony Howell successfully developed the bench-top prototype of the CAG, which was patented

Z. Cahn, L. Eckhaus

in 1998 [32]. Although PM planned to build a hand-held CAG for consumer use, Lipowicz argues that battery technology was insufficient for a smaller device [10]. He states that, although lithium-ion batteries used in modern e-cigarettes were invented in 1991, they were not readily available for use in portable devices until the early 2000s, spurred on by the portable electronics industry [10]. Certainly, lithiumion battery technology was both a large improvement over prior battery technology and necessary to the miniaturization of e-cigarettes. PM was, however, able to use nickel–cadmium battery technology in the portable Accord device as early as 1998. Lipowicz says that the power requirements for Accord ‘‘probably exceeded what was done in the CAG,’’ suggesting that its nickel–cadmium battery would have been sufficient to power a portable CAG [11]. Additionally, portability improved over time. At Chrysalis in the early 2000s, the CAG ultimately became ‘‘hand-held, but… somewhat larger than a pack of cigarettes’’ [11]. By 2005, Chrysalis had developed a functional prototype for the Aria inhaler, which used CAG technology and was capable of delivering pharmaceutical drugs [31]. The emergence in 2006 of the Ruyan e-cigarette demonstrated that, by the mid-2000s at the latest, battery technology and portability were sufficient to produce a miniaturized e-cigarette.

Limitations There are a few key limitations in attempting to divine the motivations of PM from the tobacco document record. First, individuals and divisions within the company have their own motivations, apart from those of the company overall, and so internal communications and depositions we have relied on will be colored by motivations of the actors doing the writing and talking. For example, developers of alternative products might feel it wise to emphasize the ways in which those products complement the primary profit centers at PM rather than the ways in which the core business might be threatened. Second, to the extent that litigation strategy and other legal concerns are important aspects of the narrative that we present, the background discussion that might help to illuminate the explicit conflicts between legal strategy and innovation is generally hidden from view due to the privileging of relevant documents. Third, beginning with the leaking of sensitive documents in the late 1980s and early 1990s, and culminating with an agreement to publish millions of documents as a term of the Master Settlement Agreement (MSA) in the late 1990s, there should be an increasing expectation on the part of tobacco industry employees that their internal communications will become public at some point. This likely colors the degree to which such communications involve candid discussions of sensitive topics, particularly after 1998.

Explaining the discontinuation of a non-tobacco nicotine…

Discussion The most important obstacles to CAG development appear to be regulatory and business ‘bottom-line’ concerns. Lipowicz, an expert witness for PM, and Hammond, who was an expert witness for the plaintiffs, agreed at times. The evidence seems to support Hammond where their arguments about the reasons for CAG abandonment conflict. Some of Lipowicz’ claims find little support. He, for example, argues that consumers would not have found the CAG to be acceptable, but we were unable to uncover any mention of research on consumer interest in the CAG to determine the potential commercial viability of the product. Perhaps, Lipowicz is implying that the company’s understanding of the low consumer acceptance of Accord was sufficient to extrapolate that the CAG would also fail. Of course, if that is what took place, such strategic thinking ignored the rise of electronic cigarettes shortly thereafter. Several of Lipowicz’ claims are essentially true, but overstated. Technological barriers were substantial in the 1990s, and PM was certainly not sitting on a finished version of an e-cigarette-like product. At the same time, these obstacles were not insurmountable for a company that featured a combination of high-level scientific expertise and unmatched resources. Indeed, PM successfully launched similar products (e.g., Accord) despite similar technological constraints. In terms of regulation, the most pressing issue, as mentioned by Hammond, was the FDA’s initiative in the 1990s to regulate all tobacco products. Concerns about heightened regulation of non-tobacco products were also evident. Yet, they were not so overwhelming as to prevent PM from engaging in this research; the FDA would have been vulnerable to a legal challenge if PM had sought to push forward. Additionally, a legal pathway to the market was created after the FDA declined in April 2011 to appeal the result of Sottera Inc. v. FDA [33] after losing in both District and Appeals Courts in their attempts to assert jurisdiction over e-cigarettes as ‘‘drug-device combinations’’ [20, 33]. Yet, PM still waited several years before introducing an e-cigarette or acquiring an e-cigarette company. Major obstacles were minimal after 2000 and non-existent after 2011, yet PM was remarkably slow to bring a non-tobacco product to market anyway. This suggests a reluctance to create a product that would compete with tobacco cigarettes. The Hammond Expert Report and Farone testimony [12] both elaborate on the notion, put most succinctly by Farone, that PM ‘‘always worried in the ultimate about losing the damn gold mine they have’’ [13]. The apparent importance of these business incentives do not support the view of Dutra et al. that PM viewed the CAG or related products as a complement to tobacco cigarettes as opposed to a substitute [8]. Logically, PM would prefer these products to be complementary, but their defensive approach was more consistent with a belief that the products would be competing [12, 28]. Two additional lessons are apparent from PM’s history with the CAG. First, PM has been willing to fight much harder for novel products that contain tobacco, such as Accord, than they have for non-tobacco products such as the CAG. Although

Z. Cahn, L. Eckhaus

there may be individuals within the organization who feel differently, the company’s actions and resource allocation strongly suggest that tobacco products were the priority at PM. We may see a similar pattern emerge in the company’s prioritization of the heat-not-burn devices that they have developed—most notably IQOS, a heat-not-burn product developed by PM International and already widely available outside the US—over non-tobacco products like MarkTen, the e-cigarette developed by Altria (parent company of PM USA). Lastly, it is evident that FDA policies and actions are critically important in shaping the nicotine marketplace. The FDA is arguably the most powerful regulatory agency in the world when one considers both the breadth of their jurisdiction and their substantial powers—most notably pre-market approval authority—over many of the product categories that they regulate [34, 35]. As such, the FDA has been at the center of numerous regulatory policy debates. The ‘drug lag debate’ has raged for decades about how best to balance certainty of safety with speed of approval [34]. A modified version of this debate has raged in more recent times concerning medical devices [36, 37]. Notably, ‘flashpoints’ emerged in the battles over drugs for cancer and for AIDS [35], exposing the ways in which a predominant focus on safety over swiftness may poorly serve very high-risk populations. Smoking is different from cancer and AIDS in many important and obvious ways, but the risks of continued smoking and the difficulties associated with quitting are so substantial as to warrant careful consideration of how best to balance innovation and safety for new nicotine products. Current FDA leadership has been explicit about the importance of new product innovation, suggesting a shift in thinking that could avoid the obstacles that contributed to stymying previous generations of non-tobacco nicotine products [38, 39]. We cannot say the exact degree to which PM would have acted differently if FDA policies were more amenable to non-tobacco nicotine, but there is clear evidence that regulatory concerns were discussed. It is likely that competitors, including potential new entrants, would have been similarly deterred. The CAG saga suggests the potential benefits of a new FDA approach that considers both the agency’s role in shaping the course of less-hazardous product innovation along with its role in safeguarding the marketplace from unreasonably dangerous goods. Disclaimer Any views expressed here are those of the authors and do not necessarily represent those of the American Cancer Society or the American Cancer Society—Cancer Action Network.

References 1. Folkman ML. Cigarette article. Shaker Heights, OH; 2,445,476. 1948 [cited 2017 Apr 19]. p. 1–4. http://pdfpiw.uspto.gov/.piw?docid=02445476. 2. Gilbert HA. Smokeless non-tobacco cigarette. Beaver Falls, PA; 3,200,819. 1965 [cited 2017 Apr 19]. p. 1–4. http://pdfpiw.uspto.gov/.piw?Docid=03200819. 3. Ellis CD, Seawards, Dean C, Hughes IW, Battelle Memorial Institute. Smoking device. Columbus, OH; 3,258,015. 1966 [cited 2017 Apr 19]. p. 1–12. http://pdfpiw.uspto.gov/.piw?docid=03258015. 4. Cahn Z, Siegel M. Tobacco harm reduction and FDA: obstacles and opportunities for health promotion. In: Tobacco regulation and compliance: an essential resource. 2011.

Explaining the discontinuation of a non-tobacco nicotine… 5. Howell TM, Sweeney WR. Aerosol and a method and apparatus for generating an aerosol. US5743251 A. 1998 [cited 2017 Apr 19]. http://pdfpiw.uspto.gov/.piw?Docid=00574325. 6. Cox K. Internal ideal smoke 1994 Annual Report for the Leap Activity (1A04). 1995. https://www. industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=hlcp0035. 7. Howell T. Aerosol generator prototype development. 1994. https://www.industrydocumentslibrary. ucsf.edu/tobacco/docs/#id=gzmy0002. 8. Dutra LM, Grana R, Glantz SA. Philip Morris research on precursors to the modern e-cigarette since 1990. Tob Control. 2016 Nov 15. http://tobaccocontrol.bmj.com/content/early/2016/11/15/ tobaccocontrol-2016-053406.abstract. 9. Donovan, et al. v. Philip Morris USA, Inc. Vol. 914, NE 2d. p. 891. 10. Lipowicz P, Donovan v. Philip Morris USA Inc. Expert Report of Peter J. Lipowicz. 06-12234-NG. 2015. https://www.industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=qsgx0223. 11. USDC for the District of MA, Priority-One Court Reporting Services, Inc. Donovan v. Philip Morris USA Inc. Deposition of Peter J. Lipowicz. 06-12234-NG. 2015. https://www. industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=msgx0223. 12. Farone W. United States of America, Plaintiff, V. Philip Morris USA Inc., et al. Defendants. United States’ Written Direct Examination of William A. Farone, Ph.D. Submitted Pursuant to Order # 471 99-CV-02496 (GK). 1999. https://www.industrydocumentslibrary.ucsf.edu/tobacco/docs/#id= hscw0181. 13. Levin M. The Mystery of Philip Morris’ Nicotine Inhaler. Los Angeles Times. 2005 Oct 30 [cited 2015 Aug 28]. http://articles.latimes.com/2005/oct/30/business/fi-chrysalis30/2. 14. Ellis CD. Effects of smoking: proposal for further research contracts with Battelle. 1962. https:// industrydocuments.library.ucsf.edu/tobacco/docs/jygb0223. 15. Dunworth J. Vaping. 1970’s Style: an interview with one of the pioneers. 2014. https://www. ecigarettedirect.co.uk/ashtray-blog/2014/06/favor-cigarette-interview-dr-norman-jacobson.html (Archived by WebCite at http://www.webcitation.org/6wtP6tucn). 16. Charles J. Regulatory Lettter (marginalia). 1987. https://industrydocuments.library.ucsf.edu/tobacco/ docs/#id=lfgb0175. 17. 21 U.S.C. 9 - FEDERAL FOOD, DRUG, AND COSMETIC ACT. [cited 2017 Jun 21]. https://www. gpo.gov/fdsys/granule/USCODE-2011-title21/USCODE-2011-title21-chap9. 18. Kessler D. A question of intent: a great American battle with a deadly industry. Public Affairs; 2002. 19. FDA v. Brown & Williamson Tobacco Corp. 529 U.S. 120 (2000). Justia Law. [cited 2015 Sep 22]. https://supreme.justia.com/cases/federal/us/529/120/case.html. 20. Leon RJ. Smoking Everywhere Inc. v. US Food and Drug Administration. Vol. 680, F. Supp. 2d. 2010. p. 62. 21. LEAP (1A04). 1998. https://www.industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=nhfg0175. 22. Cupp RLJ. Morality play’s third act: revisiting addiction, fraud and consumer choice in third wave tobacco litigation. Univ Kans Law Rev. 1997;1998(46):465. 23. Grubbs H. Interoffice correspondence: product research division- monthly summary for November, 1993. 1993. https://www.industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=mpdh0111. 24. Meeting Notes. 1993. https://www.industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=kxjj0120. 25. Friedman LC. Philip Morris’s website and television commercials use new language to mislead the public into believing it has changed its stance on smoking and disease. Tob Control. 2007;16(6):e9. 26. Cummings KM, Brown A, Douglas CE. Consumer acceptable risk: how cigarette companies have responded to accusations that their products are defective. Tob Control. 2006;15(Suppl 4):iv84–9. 27. Henningfield JE, Rose CA, Zeller M. Tobacco industry litigation position on addiction: continued dependence on past views. Tob Control. 2006;15(Suppl 4):iv27–36. 28. PM USA R&D Strategic Plan 1991-1995. 1991. https://www.industrydocumentslibrary.ucsf.edu/ docs/#id=nypf0189. 29. Serrano MA. Product Development Plan. 1990. https://www.industrydocumentslibrary.ucsf.edu/docs/ #id=frpm0130. 30. Ideal Smoke. 1994. https://www.industrydocumentslibrary.ucsf.edu/tobacco/docs/#id=lycc0023. 31. New Product and Technology Developments. 2005. https://www.industrydocumentslibrary.ucsf.edu/ tobacco/docs/#id=fpyw0219. 32. Cox K. Potential of the Capillary Aerosol Generator. 1998. https://www.industrydocumentslibrary. ucsf.edu/tobacco/docs/#id=ghdb0123. 33. Sottera, Inc. v. Food & Drug Admin. Vol. 627, F. 3d. 2010. p. 891. 34. Ceccoli SJ. Pill politics: drugs and the FDA. Boulder: Lynne Rienner Publishers; 2004. p. 228.

Z. Cahn, L. Eckhaus 35. Carpenter D. Reputation and power: organizational image and pharmaceutical regulation at the FDA. Princeton: Princeton University Press; 2014. p. 825. 36. Curfman GD, Redberg RF. Medical devices–balancing regulation and innovation. N Engl J Med. 2011;365(11):975–7. 37. Fargen KM, Frei D, Fiorella D, McDougall CG, Myers PM, Hirsch JA, et al. The FDA approval process for medical devices: an inherently flawed system or a valuable pathway for innovation? J Neurointerv Surg. 2013;5(4):269–75. 38. Sullivan C, Nicolaou A. Big Tobacco told to read FDA smoke signals on e-cigarettes. Financial Times. 2017 [cited 2017 Dec 15]. https://www.ft.com/content/7d9eed28-7506-11e7-90c090a9d1bc9691. 39. Kaplan S. F.D.A. Delays rules that would have limited e-cigarettes on market. The New York Times. 2017 Jul 28 [cited 2017 Dec 15]. https://www.nytimes.com/2017/07/28/health/electronic-cigarettetobacco-nicotine-fda.html.

Zachary Cahn is a Principal Scientist in the Economic and Health Policy Research Program at the American Cancer Society, Atlanta, Georgia, USA where he studies e-cigarettes and other novel nicotine products. Lindsay Eckhaus is an Epidemiology MPH candidate at the Rollins School of Public Health, Emory University in Atlanta, Georgia, USA.