Oct 292019

Jon-Emile S. Kenny MD [@heart_lung] with illustration by Carla M. Canepa MD [@_carlemd_]

Inside the word ‘emergency’ is ‘emerge’; from an emergency, new things come forth."

-Rebecca Solnit


A 32-year-old man is admitted with three days of fevers, chills, malaise, dyspnea and productive cough.  He works in a stressful job in the heart of Manhattan and has recently taken to vaping both THC and CBD to assuage his anxiety.  One week prior to presentation, he received some ‘dank vape’ from a friend with the added flavour of ‘buttered-popcorn’ prior to attending the TriBeCa Film Festival.  He spent the next few days attending many screenings and eating copious amounts of artificially-flavoured popcorn at the theatre.  He has leukocytosis with neutrophil predominance and a CXR demonstrating a bilateral, lower-lobe predominant, interstitial pattern.  His HIV PCR is negative; he is admitted to the observation unit and initiated on therapy for community-acquired pneumonia.  Within 12 hours, however, he rapidly progresses to hypoxemic respiratory failure necessitating intubation and then pronation.  Cautious bronchoalveolar lavage is performed from the right middle lobe with initial results showing few eosinophils, but an abundance of ‘foamy macrophages.’  Oil red O stain is negative.               


Electronic Nicotine Delivery Systems [or ENDS] are, essentially, composed of four parts: 1. a mouthpiece 2. a chamber into which the liquid to be ‘vaped’ is placed 3. a battery to heat the liquid and power 4. the atomizer.  Naturally, this leads me to wonder why patients in prep for bronchoscopy cannot simply ‘vape’ some lidocaine.

The ‘liquid’ that is to be vaped is typically of 3 general categories: 1. nicotine itself 2. synthetic flavours or ‘juices’ and 3. cannabis extracts [e.g. THC, CBD, hash oils extracted using butane – ‘dabs’]; inhalation of cannabis oils requires high temperature.  While flavouring of traditional cigarettes is prohibited, ENDS users have at least 15,000 – 16,000 different kinds of ‘juices’ to inhale; juices may be vaped in isolation or mixed with nicotine or cannabis extracts.

Given the above, it is evident that the combinations and permutations of inhalants is nearly infinite; this is further compounded by the fact that the temperature of aerosolization and the mixing of various juices may result in novel toxic components that are uniquely injurious.  Additionally, the components of the vape pen itself may be toxic, for example different atomizers may contain several different heavy metals.  Importantly, no single compound or ingredient has been identified as the ultimate causative culprit for ENDS or vaping associated lung injury [EVALI].

Epidemiology and presentation

As of October 31, nearly 1900 cases of EVALI have been reported with 37 deaths.  A ‘confirmed case’ of EVALI is defined by 4 criteria:

  1. The use of an ENDS – at most – 90 days prior to symptom onset
  2. Pulmonary Infiltrate(s)
  3. Negative infectious work-up that includes: respiratory viral panel, influenza PCR, pneumococcal and legionella urinary antigen, sputum culture if productive cough, bronchoalveolar lavage culture if done, blood culture and HIV-associated opportunistic infection work up if appropriate.
  4. No other possible medical diagnosis as cause

A ‘probable case’ exists when all of the above criteria are met save for a complete microbiological work-up and/or if part of the microbiological work-up is positive, but the primary care team does not think that an infectious source is the primary cause of lung injury.

A recent, seminal paper in the New England Journal of Medicine described 53 cases of ‘confirmed’ or ‘probable’ EVALI in Wisconsin and Illinois; this was followed by a number of associated commentaries describing the radiology, pathology and bronchoscopic characteristics of small groups of EVALI patients.

Virtually all patients presented with at least one constitutional symptom and at least one chest symptom; gastrointestinal symptoms were also very common.  ENDS-use was typically within 7 days of presentation.  Almost 90% of cases had a peripheral leukocytosis and roughly one-third had an ambient air oxygen saturation of less than 89%.

Of the 53 cases, 41 had extensive exposure interviews at the time of publication; of those interviewed the authors found that products containing THC were the most common ENDS exposure [84%].  In a case series of 6 patients from Pittsburgh, ‘Dank Vape’ [a cannabis product]  was a common exposure.  Critically, however, approximately 1-in-5 in the Wisconsin and Illinois cohort reported using only nicotine-based products, and nearly half used both THC and nicotine-based products.


In a brilliant review of the chest radiology associated with EVALI, Henry and colleagues note the diversity of encountered patterns.  All of the radiographic-patterns below have been documented in at least one case report of EVALI [see illustration].

Lipoid Pneumonia

Aspiration of exogenous lipids usually affects the dependent lung.  Fat attenuation with low Hounsfield Units [less than 30] is classic, but not sensitive.  Indeed, fat attenuation was not reported in the New England Journal case series from Wisconsin and Illinois, though this information was gleaned only by reading radiography reports.  In addition to fat attenuation, lipoid pneumonia is characterized by any or all of: crazy-paving, ground glass opacities and consolidations.

Hypersensitivity Pneumonitis

At least two case reports have noted a pattern consistent with hypersensitivity pneumonitis [HP].  This pattern, classically, is typified by symmetrical mid-to-upper lung ground-glass, mosaicism with ill-defined centrilobular nodules.  HP is an immune response to an environmental antigen, but the antigens associated with EVALI are unknown.

Diffuse Alveolar Hemorrhage

Diffuse alveolar hemorrhage [DAH] may be characterized by centrilobular nodules, ground glass, consolidation, or some combination thereof – typically with sparing of the subpleural lung.  The diagnosis is clinched by BAL demonstrating persistent or increasingly bloody aliquots.

Organizing Pneumonia

Organizing pneumonia is often accompanied by dense or ground glass consolidations with a peripheral, or even peri-lobular distribution; classically, there is subpleural sparing.  Fibroblasts and collagen typify the histopathology of organizing pneumonia which is a regular reaction to lung injury.  Indeed, as described below, the 17 patients with lung biopsy and EVALI were found to have evidence of organizing pneumonia and diffuse alveolar damage.  Importantly, too, these radiographic patterns needn’t be mutually-exclusive.  In one case-report of cannabis-related EVALI, prominent, bilateral ‘tree-in-bloom’ opacities were noted with diffuse alveolar hemorrhage found on BAL [i.e. bloody aliquots] and organizing pneumonia observed on biopsy.

Acute lung injury or diffuse alveolar damage

Radiographic ‘acute lung injury’ during the exudative phase is usually composed of crazy-paving, ground glass and other heterogenous consolidations in a gravity-dependent manner.  The organizing phase is heralded, radiographically, by traction bronchiectasis and reticulations.  Like acute eosinophilic pneumonia, the histology of lung injury is often underpinned by diffuse alveolar damage [DAD], though without eosinophilic infiltration.

Acute Eosinophilic Pneumonia

Symmetrical ground glass with septal thickening and pleural effusions can be seen with AEP, but these are non-specific [e.g. also seen with cardiogenic pulmonary edema].  Importantly, peripheral eosinophilia is usually absent on presentation.  What usually clinches the diagnosis of AEP is a BAL with at least 25% eosinophils and a rapid, dramatic response to steroids.  AEP has the pathological finding of diffuse alveolar damage, but with eosinophilic infiltration.  Both AEP and DAD are rapidly developing acute lung injuries, are associated with inhalational injuries and have intersecting radiographic and histopathological findings.

Illustration: The many radiographic 'flavours' of EVALI

Bronchoscopy and pathology

In the Wisconsin and Illinois cohorts, 24 underwent bronchoalveolar lavage [BAL].  A neutrophil-predominant leukocytosis was common with none having eosinophilia.  There were 14 cytology reports with only 7 noting lipid-laden macrophages by oil red O stain, consistent with lipoid pneumonia.  Importantly, an accompanying letter-to-the-editor [co-authored by @medevidenceblog] reported 6 cases all of which had lipid-laden macrophages.

Yet in another accompanying letter, pathology from 17 confirmed or probable cases of EVALI was reviewed and none showed characteristic findings of lipoid pneumonia.  Rather, histopathological findings revealed multiple patterns of acute lung injury: acute fibrinous pneumonitis, diffuse alveolar damage, or organizing pneumonia – typically around the bronchioles with bronchiolitis.  Lipid-laden macrophages were not identified, rather the macrophages seen were non-specific ‘foamy macrophages.’  The authors concluded that EVALI embodies an airway-centred chemical pneumonitis from one or more unknown inhaled toxic substances rather than lipoid pneumonia.

In part 2 of this EVALI review, some molecular mechanisms of vaping associated lung injury will be described in addition to tentative suggestions on its management.



Dr. Kenny is the cofounder and Chief Medical Officer of Flosonics Medical; he also the creator and author of a free hemodynamic curriculum at heart-lung.org



Get our weekly email update, and explore our library of practice updates and review articles.

PulmCCM is an independent publication not affiliated with or endorsed by any organization, society or journal referenced on the website. (Terms of Use | Privacy Policy)


Vaping-Associated Lung Injury – Part 1