Electronic cigarettes are handheld devices that heat an inner liquid containing chemicals to be aerosolized and inhaled, and have become a popular alternative to conventional cigarettes. Their use, termed vaping, has been linked to severe injury, with 2711 cases of associated lung injury and 60 deaths reported to the Centers for Disease Control and Prevention at the time of writing. Published case reports and series have emerged detailing clinical and imaging characteristics of vaping-induced lung injury. However, the pathologic characteristics of these induced injuries are still being established, particularly findings occurring over time.
To illustrate the autopsy findings of an older patient who died of vaping-induced injury after prolonged symptomology and to provide a review of the most recent literature regarding the basic science, epidemiology, clinical presentation, imaging characteristics, and pathology of vaping-induced lung injury.
Autopsy case and peer-reviewed literature.
Vaping-induced lung injury has emerged as a public health issue, and this case represents a rare opportunity to evaluate this issue at autopsy. Most commonly, the injury has been attributed to tetrahydrocannabinol product use as opposed to nicotine. This case demonstrates that as today's young and relatively healthy “vapers” grow older and develop the comorbidities that come with advanced age, there is serious risk of chronic lung damage from vaping that could result in death. Further observations and studies, particularly autopsy evidence, are clearly important to understand the possible outcomes.
Electronic cigarettes, or e-cigarettes, are handheld, battery-powered devices that heat an inner liquid containing a number of chemical ingredients to be aerosolized and inhaled. The chemical ingredients principally include nicotine and flavorings, but also contain other additives such as propellants, solvents, and oils. Users are also subject to exposure to chemicals typically used for cleaning.1,2 E-cigarettes can also be used as a method of delivery for tetrahydrocannabinol (THC), cannabidiol, and other drugs. These devices have become an increasingly popular alternative to conventional cigarettes since their entry into the marketplace in the early 2000s and have been generally marketed as a safer alternative to conventional cigarettes as well as a tool in smoking cessation.3,4 However, e-cigarette use, colloquially termed vaping, has been linked to severe injury, including 2711 cases of e-cigarette or vaping product use–associated lung injury with 60 deaths reported to the Centers for Disease Control and Prevention (CDC) as of January 21, 2020.5 The CDC data have also revealed that a large majority (82%) of the cases involved patients who reported use of THC-containing products, with 33% of the total reporting exclusive use of THC-containing products (data as of January 14, 2020).
The patient was a 65-year-old woman with a past medical history of type I diabetes mellitus, coronary artery disease (status post multivessel coronary artery bypass graft procedure), obesity (status post Roux-en-Y gastric bypass), major depressive disorder, chronic pain, and opioid dependence. Six years prior to death, the patient had a low-positive antinuclear antibody titer (1:80). Finally, she had a 20–pack-year history of cigarette smoking, but during the year prior to her death she had been using a cannabinoid vaporizer almost daily. She reported experiencing exertional dyspnea that had been evolving slowly for a year and then acutely presented to an affiliate hospital of our institution reporting nausea, vomiting, shortness of breath, abdominal pain, and chest tightness that was not alleviated by sublingual nitroglycerine. She was tachypneic and hypoxic (91% oxygen saturation while receiving 2 L/min of oxygen via nasal cannula). Computed tomography of the chest revealed bilateral interstitial prominence combined with ground-glass changes (Figure 1), which was concerning for pulmonary alveolar proteinosis or diffuse alveolar hemorrhage. Laboratory studies revealed granulocytosis with a left shift, and she was treated with antibiotics empirically, but blood and respiratory cultures proved to be negative. A cardiac workup did not show evidence of an acute cardiac syndrome. The patient subsequently underwent a right anterior thoracotomy for wedge resection of the right upper lobe and right middle lobe 2 days after admission, and remained intubated following the procedure. Histologic evaluation of the wedge resections revealed acute diffuse alveolar damage (DAD), characterized most prominently by hyaline membrane formation, and organizing DAD, characterized by fibroblastic proliferation and interstitial fibrosis (Figure 2, A). Foamy histiocytes were also observed in scattered alveoli on the biopsy material (Figure 2, B), as were subtle areas of chronic interstitial inflammation with rare eosinophils (Figure 2, C). At this point, there was a strong clinical suspicion that the patient's respiratory illness was due to chronic cannabinoid vaping, and she was treated with a course of high-dose steroids. The steroid therapy resulted in substantial improvement in her respiratory status and regression of the changes that were noted previously on imaging, and the patient was eventually extubated and discharged to a rehabilitation facility 20 days after admission.
Six days after discharge, the patient was readmitted reporting shortness of breath that had again worsened. She was treated with the same approach as in her previous admission, with antibiotics and high-dose steroids. However, despite this, her respiratory status continued to decline, and this ultimately led to a decision to transition to comfort measures. Death was declared 36 days after initial presentation. An autopsy was performed at our institution with an approximate postmortem interval of 42 hours.
Gross examination of the lungs revealed that both were heavy, with the right lung weighing 890 g (reference range, 360–570 g) and the left lung weighing 700 g (reference range,6 325–480 g). On sectioning, the lung parenchyma was diffusely consolidated, with some airspace dilation (Figure 2, D). There were no focal lesions and the bronchi and vasculature appeared grossly unremarkable. Microscopic examination of the lungs showed DAD in all lung lobes. The DAD was present in all temporal phases (exudative and organizing/acute interstitial pneumonitis), characterized by hyaline membranes, organization, areas of extensive squamous metaplasia, bronchiolitis, and bronchiolar proliferation (Figure 2, E and F). Foamy macrophages were observed in a large number of air spaces in sections from both lungs (Figure 2, G). Dense interstitial fibrosis was appreciated in a number of sections from both lungs (Figure 2, H). Sections of tissue sampled from the right middle lobe also showed alveolar hemorrhage and hemosiderin deposition. No foreign material was noted on any of the tissue sections. No significant acute inflammation was noted, there were no viral inclusions, and no bacterial, fungal, or parasitic organisms were noted on routine histology (hematoxylin-eosin stain).
There were a number of additional autopsy findings. The most notable finding on external examination was obesity (body mass index, 33.6 kg/m2). Internally, the most significant additional autopsy finding was severe native coronary artery disease associated with 3 intact and patent bypass grafts, but with severe postgraft atherosclerotic stenosis in the right coronary artery, the left anterior descending artery, and the ramus intermedius of the left coronary artery. Corresponding to this, there was evidence of acute ischemic injury in the myocardium of the interventricular septum and subacute to old myocardial infarctions additionally in the interventricular septum and the left ventricle. The liver also showed steatohepatitis with focal bridging fibrosis. The remainder of the gastrointestinal system showed expected postsurgical changes (status post cholecystectomy and Roux-en-Y gastric bypass). Examination of the thyroid showed mild lymphocytic thyroiditis. Finally, examination of the brain revealed incipient Alzheimer disease neuropathologic changes. Ultimately, the cause of death was determined to be acute on chronic lung injury, likely related to vaping.
REVIEW OF THE LITERATURE
Components in vaporizer products have been demonstrated to impact both animals and human-derived cells and tissues. In mice, a passive, in vivo vapor inhalation system using a top-selling nicotine-based “e-liquid” has been shown to induce platelet hyperactivity,7 demonstrating increased aggregation, increased granule release, enhanced activation of glycoprotein IIb/IIIa, increased phosphatidylserine expression, and Akt and ERK activation. In addition, the platelets demonstrated relative resistance to inhibition by prostacyclin. The vaping-exposed mice therefore had shortened thrombosis occlusion and bleeding times, suggesting that vaping may increase the risk for thrombogenesis. Also, in bronchoalveolar lavage samples from humans, material from chronic “vapers” demonstrated elevations in the activity of neutrophil elastase, matrix metalloprotease 2, and matrix metalloprotease 9 that were essentially equal to the elevations observed in conventional smokers.8 This suggests that chronic vaping may increase proteolysis in the lung and similarly may contribute to the development of chronic lung disease. Further, short-term exposure to flavoring agents used in vaping products have adversely effected endothelial cell phenotype in isolated cells from human participants.9 In terms of specific chemical agents inside vaping fluid, vitamin E acetate has received recent interest and has been identified by the CDC as a potential causative agent in vaping-induced lung injury.5,10 Vitamin E acetate is an oily chemical agent commonly added to THC-containing vaping products, and is thought to be retained by the lungs for prolonged amounts of time, potentially predisposing the lung to injury. States including Washington have initiated bans on the chemical.11,12 In total, these data stand in contrast to the largely marketing-driven viewpoint that e-cigarettes are a safer alternative to conventional cigarettes. Interestingly, though the large majority of known cases of vaping-induced lung injury have involved patients that reported use of THC-containing products,5 epithelial cells, macrophages, and lung fibroblasts exposed to cannabidiol have shown attenuated inflammatory responses, suggesting that cannabidiol may act as an anti-inflammatory agent similar to steroids.13
A recent series published by Layden et al14 highlighted the clinical characteristics of 53 patients with respiratory illness related to e-cigarette use. The average age in that series was 19 years (range, 16–53 years), with about one-third of the patients being younger than 18 years, and 83% of patients were male. The patients presented with varied combinations of respiratory symptoms (98% of cases: shortness of breath, chest pain, cough, and/or hemoptysis), gastrointestinal symptoms (81% of cases: nausea, vomiting, diarrhea, abdominal pain), and constitutional symptoms (100% of cases: subjective fever, chills, weight loss, fatigue/malaise, headaches). The majority (69%) were hypoxic, and 43% were tachypneic. Only 29% had objective fever (temperature ≥38°C). Ninety-four percent of patients were hospitalized, with 87% requiring supplemental oxygen and 32% requiring mechanical ventilation. Ninety percent received antibiotics and 92% received systemic steroids. Of the 53 cases, there was only one death, which is compatible with the low rate of mortality reported by the CDC (2%). There was a variety of e-cigarette product use in the group, with 80% of patients having reported THC product use (37% with THC-only use), 61% with nicotine product use (17% with nicotine-only use), and 44% with a combination of THC and nicotine product use. This corresponds with the most recent CDC data, which implies that the majority of patients who experience vaping-induced lung injury report THC product use.
A comparison between clinical characteristics in the literature and those of the autopsy case is shown in the Table. The patient in the autopsy case largely fits within the frame of presentations in the literature, and she too reported THC product use (THC only). However, of most interest is the duration of her symptomatology. Some of her presenting symptoms were acute (chest tightness and nausea); however, she did report slowly progressive dyspnea that had evolved during the course of a year (before it became severe) that was largely coincident with the time she began vaping. This time frame is substantially longer than the duration of illness reported in the aforementioned series and in several published case reports.15–22 It is also important to note that this patient had a history of severe coronary artery disease with autopsy-proven severe postgraft atherosclerosis and myocardial infarctions, which could have contributed to her symptomatology. Additionally, this raises the important point that symptoms of vaping-induced lung disease may mimic more prevalent cardiorespiratory diseases, particularly in older patients with already established diagnoses. Prior published studies on vaping-induced lung injury have had strict criteria that excluded patients with other comorbidities, therefore including largely younger patients, and have not addressed this question. However, a substantial proportion of vapers, in particular patients who are using the e-cigarette option to attempt to reduce or quit conventional smoking, will be older individuals with other health conditions. Given the lack of data on the clinical diagnosis of vaping-induced lung injury in older patients and/or patients with additional cardiorespiratory comorbidities, at this point it likely should represent a diagnosis of exclusion. It would be of interest to study if there are differentiating characteristics between vaping-induced lung injury and other potential clinical mimickers, and also to investigate whether the clinical, radiographic, and/or pathologic features of vaping-induced injury are more severe in older patients and/or those with other cardiorespiratory comorbidities.
Imaging characteristics of vaping-induced lung injury have also been emerging in the literature. In the series by Layden et al,14 100% of patients (53 of 53) had bilateral infiltrates identified on chest radiograph or computed tomography. Ground-glass opacities, sometimes with subpleural sparing, were characteristic on chest computed tomography. An additional imaging series and review of literature was conducted and recently published by Henry et al,18 totaling 34 cases. In this, 4 imaging patterns were observed in vaping-induced lung injury: acute eosinophilic pneumonia, DAD, organizing pneumonia, and lipoid pneumonia, often also with subpleural sparing. In our case, although the patient did have diffuse, bilateral lung disease as observed in these published cases, there was no subpleural sparing.
Pathology and the Determination of Vaping-Induced Death
The lung pathology observed in this case was that of DAD in both exudative and organizing/acute interstitial pneumonitis phases, with foamy macrophages present within scattered air spaces (comparison with literature findings in the Table). Very little active inflammation was observed and there was no sign of vasculitis or necrosis. In addition, dense interstitial fibrosis was noted in a number of lung sections, indicating some level of chronic disease. Collectively, the lung findings highlight acute lung injury, but also some chronic injury that may provide a histologic correlate to the patient's prolonged symptomatology since beginning vaping. They are also consistent with some of the most commonly observed pathologic findings identified in the literature. These reported findings include multiple combinations of DAD, organizing pneumonia, airway-centric acute fibrinous pneumonitis, interstitial inflammation (chronic, with or without rare eosinophils), fibrinous exudates in air spaces, and peribronchiolar and/or interstitial granulomatous pneumonitis.14,17,19,20,23,24 Bronchiolitis has also been a common accompanying finding. Foci of parenchymal hemorrhage have additionally been observed.17 Previously published imaging results suspicious for acute eosinophilic pneumonia or lipoid pneumonia in published cases of lung injury due to vaping have not been corroborated via pathologic examination of tissue in any current studies.
Unfortunately, the pathologic findings observed in this case and in other published cases are nonspecific, and the potential etiologies for the findings are broad. Therefore, diagnosis of vaping-induced lung injury at the time of biopsy or at autopsy is a challenging endeavor. An accurate clinical history with careful exclusion of other possibilities will likely continue to be the most crucial factor in the establishment of a diagnosis of vaping-associated injury. In our case, though the cannabinoid vaporizer use was the likely culprit for the lung pathology, an occult autoimmune disease was considered given the patient's history of a low-positive antinuclear antibody titer in the distant past. However, the patient in our case only had an antinuclear antibody titer of 1:80, 6 years prior to presentation, with no clinical manifestations of autoimmune disorder in that time and no signs of vascular inflammation or injury at autopsy. It should also be noted that more than a quarter of the healthy general population is antinuclear antibody positive, including greater than 10% with a titer at or above 1:80, and it is believed that a positive antinuclear antibody in isolation is of little diagnostic value for autoimmune disease.25,26
It is important to note that examination of a case at autopsy enables more holistic evaluation of the lungs and full pathologic data on all organ systems. The autopsy is generally regarded as the gold standard for comparison with antemortem clinical observations27,28 and has served as a key tool in the elucidation of numerous diseases, from influenza to lupus to acquired immunodeficiency syndrome.29 From 1912, rates of diagnostic discordance between antemortem diagnoses and postmortem diagnoses at autopsy have been well documented and have remained despite advancements in medicine.30–33 The autopsy should therefore continue as a significant tool for establishing a definitive anatomic diagnosis and/or assessing the ultimate accuracy of any diagnostic test, including radiology and tissue biopsies.34–36
Vaping-induced lung injury has emerged recently as a public health issue, and this case represented a rare opportunity to evaluate the issue at autopsy. Vaping-induced lung injury has thus far typically been reported in younger patients, though this case highlights the potential pathologic effects in older patients and/or those with preexisting cardiorespiratory comorbidities. Most commonly, vaping-induced injury has been attributed to THC product use as opposed to nicotine product use, as it was here, though reports of injury with nicotine-only use have been published. This case additionally demonstrates that as today's young and relatively healthy vapers grow older and develop the comorbidities that come with advanced age, there may be risk of chronic lung damage from vaping that can ultimately contribute to or result in death. Further observations and studies, including autopsy evidence, are necessary to determine if there are more specific diagnostic findings associated with vaping, especially in users over time.
The authors have no relevant financial interest in the products or companies described in this article.