Lung Biopsy Findings in Severe Pulmonary Illness Associated With E-Cigarette Use (Vaping): A Report of Eight Cases

Abstract

Objectives

The aim of this report is to describe the lung biopsy findings in vaping-associated pulmonary illness.

Methods

Lung biopsies from eight patients with vaping-associated pulmonary illness were reviewed.

Results

The biopsies were from eight men (aged 19-61 years) with respiratory symptoms following e-cigarette use (vaping). Workup for infection was negative in all cases, and there was no evidence for other etiologies. Imaging showed diffuse bilateral ground-glass opacities in all patients. Most recovered with corticosteroid therapy, while one died. Lung biopsies (seven transbronchial, one surgical) showed acute lung injury, including organizing pneumonia and/or diffuse alveolar damage. Common features were fibroblast plugs, hyaline membranes, fibrinous exudates, type 2 pneumocyte hyperplasia, and interstitial organization. Some cases featured a sparse interstitial chronic inflammatory infiltrate. Although macrophages were present within the airspaces in all cases, this feature was not prominent, and findings typical of exogenous lipoid pneumonia were absent.

Conclusions

The histopathology of acute pulmonary illness related to e-cigarette use (vaping) is characterized by acute lung injury patterns, supporting the contention that vaping can cause severe lung damage.

“Vaping” refers to inhalation of the aerosol produced by electronic cigarettes (e-cigarettes, or electronic vaporizers). This aerosol is generated by heating a liquid that usually contains nicotine, flavorings, and other chemicals.¹ E-cigarettes can also be used to deliver marijuana and other drugs. A variation on this theme is “dabbing,” which involves inhalation of an aerosol formed by superheating substances that contain high concentrations of tetrahydrocannabinol (THC) and other plant compounds such as cannabidiol. In the recent past, lung disease related to vaping has emerged as a public health issue in the United States, generating considerable attention in the national news media.² A recent report in the New England Journal of Medicine described the clinical features of a series of 53 patients from Illinois and Wisconsin with vaping-related pulmonary illness.³ This report was a major addition to the literature documenting a link between vaping and lung disease. At the time of this writing (September 2019), 530 possible cases of acute lung injury potentially associated with vaping have been reported from 38 states.^1,4 The Centers for Disease Control and Prevention (CDC) has released interim guidelines on case definitions (discussed in detail in the Materials and Methods section).⁴

Despite the intense spotlight on vaping, the pathologic findings of vaping-related lung disease have not been adequately described. A few reports have appended labels to the lung injury patterns observed in these patients based on imaging and/or bronchoalveolar lavage (BAL) findings, but biopsy findings remain largely unknown.^1,5-7 The aim of this report is to describe and illustrate the histopathologic features of lung disease attributable to vaping.

Materials and Methods

Eight patients with vaping-associated pulmonary illness who underwent lung biopsies at five separate institutions form the basis of this report (seven transbronchial lung biopsies and one surgical lung biopsy). Cases were included if they fulfilled the CDC surveillance case definition for severe pulmonary disease associated with e-cigarette use.⁴ All but one case in this series fulfilled CDC criteria for a “confirmed case,” which are as follows:

1. Using an e-cigarette or dabbing during the 90 days before symptom onset

2. Pulmonary infiltrate on chest radiograph or ground-glass opacities on chest computed tomography (CT)

3. Absence of pulmonary infection on initial workup. Minimum criteria include negative respiratory viral panel, influenza polymerase chain reaction, or rapid test if local epidemiology supports testing. All other clinically indicated respiratory infectious disease testing (eg, urine antigen for Streptococcus pneumoniae and Legionella, sputum culture if productive cough, BAL culture if done, blood culture, human immunodeficiency virus (HIV)–related opportunistic respiratory infections if appropriate) must be negative.

4. No evidence in medical record of alternative plausible diagnoses (eg, cardiac, rheumatologic, or neoplastic process)

CDC criteria for a “probable case” differ from confirmed cases only in that infection may be identified via culture or polymerase chain reaction, but the clinical team believes this is not the sole cause of the underlying respiratory disease process. One case in this report fulfilled criteria for a probable case (case 4).

Results

Clinical Findings

The clinical findings are summarized in Table 1. All eight patients were men (age range, 19-61 years; mean, 29 years) who developed respiratory symptoms following e-cigarette use (vaping). All patients reported vaping THC; two were also vaping nicotine. One individual had a history of dabbing. The most common presenting symptoms were fever (n = 7/8 patients), cough (n = 6/8), and dyspnea (n = 5/8). Most patients had no significant prior medical illness. One patient had a history of inflammatory bowel disease (case 5). Another had a history of chronic pain (on medical marijuana, tramadol, baclofen, and pregabalin), degenerative disc disease, cervical radiculopathy, peripheral neuropathy, and neurogenic bladder (case 3). Four patients were hypoxic at presentation. Empiric broad-spectrum antibiotics were administered in seven. Bilateral crackles were audible in the lungs on auscultation in two individuals and bilateral coarse rhonchi in one. Notable laboratory findings included leukocytosis (n = 6/8) with neutrophil predominance and elevations in erythrocyte sedimentation rate and C-reactive protein in a few individuals. Workup for infection was negative in seven of eight cases. The tests performed were different for each patient but generally included respiratory viral panels, serologic studies for fungi, testing for Legionella and HIV, and microbiologic cultures of sputum, blood, BAL fluid, and biopsied lung tissue. In case 4, there was a weakly positive Legionella immunoglobulin G, but immunoglobulin M testing and urine antigen were negative. Bronchoscopy with transbronchial lung biopsies were performed in seven patients and surgical lung biopsy in one. There was no clinical or laboratory evidence for other plausible etiologies in any patient; autoimmune serologies were negative in all patients tested. Chest CT showed bilateral ground-glass opacities in all patients Image 1. In addition, bilateral consolidation was noted in some. Seven individuals were treated with corticosteroids and one with antibiotics. Most recovered with corticosteroid therapy and were discharged home within a few days. One patient (61-year-old man, case 3) developed severe acute respiratory distress syndrome. His condition continued to deteriorate, with worsening hypoxia and fever, and his course was complicated by methicillin-resistant Staphylococcus aureus ventilator-associated pneumonia on day 21 of admission. Despite receiving appropriate antibiotic therapy and mechanical ventilation, his condition worsened and he died on day 31 of admission.

Table 1

Clinical Features of Patients With Vaping-Associated Pulmonary Illness

Variable	Patient 1	Patient 2	Patient 3	Patient 4	Patient 5	Patient 6	Patient 7	Patient 8
Age, y/sex	35/M	22/M	61/M	20/M	21/M	28/M	19/M	28/M
Main symptoms	Fever, night sweats, weight loss, dyspnea, cough, nausea, vomiting	Fever	Dyspnea, cough, chest heaviness, nausea, fever, dizziness	Dyspnea, cough, nausea, vomiting, diarrhea, fever	Cough, flare of inflammatory bowel disease	Dyspnea, cough, fever, nausea	Nausea, weight loss, abdominal pain, fever	Cough, dyspnea, fever, weight loss, night sweats
Substance vaped	90% THC	THC (brand: “Dank”) and nicotine; recent refill of cartridge with 90% THC	THC/marijuana (vaping) and marijuana wax (dabbing)	THC (cannabis oil)	Purified THC	THC	THC, nicotine	93% THC
Approximate time between vaping and symptom onset	Not known	Switched to “Dank” brand a few weeks prior to admission	2 years (vaping) 2 months (dabbing)	2 weeks	Simultaneous	3 weeks	14 months	Not known
Approximate time from symptom onset to presentation	2 days	1 week	3 days	2 weeks	2 weeks	1 week	6 months	3-4 days
Initial total leukocyte count, /μL	19,900	9,320	9,000	44,900	12,400	20,000	15,000	13,340
ESR, mm/h	72	102	Not available	123	23	Not available	89	Not available
CRP, mg/dL	29.5	35.4	Not available	284	31	Not available	20	Not available
Imaging (chest CT)	Bilateral GGO and consolidation, peripherally located	Bilateral GGO and consolidative opacities in mid to lower lung fields	Diffuse bilateral GGO	Diffuse bilateral GGO and basilar consolidation	Diffuse bilateral GGO	Bilateral GGO, somewhat patchy	Diffuse bilateral GGO and nodular opacities	Diffuse bilateral GGO with solid nodular densities
Therapy	IV methylprednisolone	IV methylprednisolone	High-dose IV corticosteroids	Corticosteroids	Oral prednisone	Oral prednisone	Antibiotics	Antibiotics and methylprednisolone
Outcome	Alive; discharged home on room air on day 3 of steroids	Alive; discharged home on day 3 of treatment	Died on day 31 of admission	Alive; discharged home on day 7	Alive; discharged home on day 8	Alive; discharged home on day 4	Alive; discharged on room air on day 4	Alive; discharged on room air after 3 days of intravenous methylprednisolone

CRP, C-reactive protein; CT, computed tomography; ESR, erythrocyte sedimentation rate; GGO, ground-glass opacities; IV, intravenous; THC, tetrahydrocannabinol.

Image 1

Chest computed tomography findings in severe pulmonary disease associated with e-cigarette use (vaping). A, Case 1: diffuse bilateral ground-glass opacities with peripheral prominence, mimicking eosinophilic pneumonia. B, Case 2: diffuse bilateral ground-glass opacities with areas of consolidation. C, Case 4: bilateral basilar dependent consolidations with diffuse ground-glass opacities and mild smooth septal thickening. D, Case 6: bilateral ground-glass opacities in a somewhat patchy distribution.

Lung Biopsy Findings

The pathologic findings are summarized in Table 2 and illustrated in Image 2, Image 3, and Image 4. All eight biopsy specimens showed acute lung injury, manifesting as organizing pneumonia, diffuse alveolar damage (DAD), unclassifiable organizing acute lung injury, or a combination of these patterns. Four biopsy specimens featured variably prominent areas of organizing pneumonia, characterized by the presence of serpiginous or polypoid fibroblast plugs (Masson bodies) within airspaces (Image 2). This was accompanied by mild interstitial chronic inflammation in three cases and focal acute inflammation within the airspaces in case 8. Two biopsy specimens showed DAD (Image 2). In case 3, the only surgical lung biopsy specimen in this series, areas of DAD were admixed with areas of organizing pneumonia (mixed acute lung injury pattern). Both biopsy specimens with DAD featured hyaline membranes (acute stage of DAD) as well as alveolar septal expansion by fibroblasts (organizing stage of DAD). Two biopsy specimens showed organizing acute lung injury that could not be further classified.

Table 2

Pathologic Findings in Vaping-Associated Pulmonary Illness

Case No.	Biopsy Type	Biopsy Site	Major Histologic Finding	Macrophages in Biopsy Specimen	BAL	Other Histologic Findings
1	TBBX	Right middle lobe, right lower lobe	Diffuse alveolar damage, acute and organizing	Present within airspaces; no evidence of exogenous lipoid pneumonia	BAL cytology: macrophage predominant	Fibrinous exudates in airspaces, interstitial chronic inflammation, rare eosinophils; AFB and GMS negative
2	TBBX	Lingula and left lower lobe	Organizing pneumonia	Present within airspaces; no evidence of exogenous lipoid pneumonia	No BAL cytology	Fibrinous exudates in airspaces, interstitial chronic inflammation, rare eosinophils; GMS negative
3	Surgical (open) lung biopsy	Right middle lobe, right lower lobe	Diffuse alveolar damage (acute and organizing) and organizing pneumonia	Present within airspaces; no evidence of exogenous lipoid pneumonia	No BAL cytology	Organizing fibrinous exudates, squamous metaplasia
4	TBBX	Right lung	Organizing acute lung injury	Few; no evidence of exogenous lipoid pneumonia	Lipid-laden macrophages, oil red O positive	Fibrinous exudates in airspaces, interstitial chronic inflammation, AFB and GMS negative
5	TBBX	Right lung	Organizing pneumonia	Present (foamy); no evidence of exogenous lipoid pneumonia	Lipid-laden macrophages, oil red O positive	Interstitial chronic inflammation, AFB and GMS negative
6	TBBX	Right lower lobe	Organizing acute lung injury	Present (foamy); no evidence of exogenous lipoid pneumonia	Lipid-laden macrophages, oil red O positive	Fibrinous exudates in airspaces, interstitial chronic inflammation, AFB and GMS negative
7	TBBX	Right lower lobe	Organizing pneumonia	Present (foamy); no evidence of exogenous lipoid pneumonia	Macrophage predominant	Fibrinous airspace exudates, GMS and AFB negative
8	TBBX	Right lower lobe	Organizing pneumonia	Present (foamy); no evidence of exogenous lipoid pneumonia	Macrophage predominant	Interstitial inflammation, focal acute inflammation in airspaces, AFB and GMS negative

AFB, acid-fast bacteria; BAL, bronchoalveolar lavage; GMS, Grocott methenamine silver; TBBX, transbronchial biopsy.

Image 2

Lung biopsy findings in vaping-associated pulmonary illness. A, Organizing pneumonia (case 3). A polypoid fibroblast plug fills an airspace (H&E, ×200). B, Organizing pneumonia, case 2 (H&E, ×200). C, Diffuse alveolar damage, case 3. Note hyaline membranes and interstitial thickening (H&E, ×200). D, Diffuse alveolar damage, case 1. Hyaline membranes are beginning to detach into airspaces (H&E, ×200).

Image 3

Fibrinous exudates and interstitial inflammation in vaping-related lung injury. A, B, Fibrinous exudates within airspaces (A, case 1, H&E, ×200; B, case 3, H&E, ×200). C, Interstitial inflammation, case 1 (H&E, ×200). D, Interstitial inflammatory infiltrate at high magnification, showing lymphocytes and a rare eosinophil (case 2, H&E, ×400).

Image 4

Macrophages in vaping-related lung injury. A, Foamy macrophages within airspaces, case 3 (H&E, ×400). B, Few intra-alveolar macrophages without distinctive features, case 2 (H&E, ×400). C, Macrophages in bronchoalveolar lavage fluid do not show prominent intracytoplasmic vacuoles (Papanicolaou stain, ×400). D, Oil red O–positive macrophages, case 4, courtesy of Margaret Compton, MD (oil red O, ×400).

Fibrinous exudates with varying degrees of organization were noted within airspaces in six of eight biopsy specimens (Image 3). In most cases, they were a focal finding. Overall, interstitial chronic inflammation composed mainly of lymphocytes was noted in six of eight cases. In cases 1 and 8, immunohistochemical stains showed that the lymphocytic infiltrate was composed mainly of CD3-positive T cells and very few CD20-positive B cells.

Variable numbers of macrophages were present within airspaces in all cases (Image 4). However, they were not a prominent finding in any biopsy specimen. A CD68 stain (performed in cases 1 and 8) confirmed the presence of several CD68-positive macrophages, mainly within airspaces. The macrophages contained foamy cytoplasm in three cases, but the coarsely vacuolated cytoplasm characteristic of exogenous lipoid pneumonia was not noted in any of the biopsy specimens. Similarly, interstitial lipid deposits surrounded by multinucleated giant cells—as seen in exogenous lipoid pneumonia—were absent in all cases. BAL fluid was macrophage predominant in six cases, without a foamy appearance or coarse vacuoles. Oil red O staining, performed in cases 4, 5, and 6, was positive in all three cases. The lipid index was 12 in case 4 and 6 in cases 5 and 6.

In all biopsy specimens in which special stains for mycobacteria and fungi (Ziehl-Neelsen and/or Grocott methenamine silver) were performed, they were negative. No granulomas or malignant cells were identified, and there were no histologic findings suggesting a plausible alternative etiology.

Discussion

The main finding of this case series—one of the first to report lung biopsy findings in vaping-associated pulmonary illness—is that acute lung injury patterns are the most consistent histologic findings in this entity. Other findings include variable numbers of airspace macrophages and varying degrees of interstitial chronic inflammation. Fibrinous exudates within the alveoli, termed acute fibrinous and organizing pneumonia by some authors, are also seen focally in most cases. These findings are nonspecific with regard to etiology but help to confirm the presence of significant lung damage. Lung biopsies are not currently required for the diagnosis of vaping-associated pulmonary illness, but they provide another modality to exclude the possibility of infection.

From a histologic standpoint, the two major acute lung injury patterns are DAD and organizing pneumonia.⁸ It is well known that a wide variety of toxic inhalants can cause acute lung injury, including amitrole-containing herbicides,⁹ mixtures of household ammonia and bleach,¹⁰ chlorine gas, crack cocaine, hydrogen sulfide, mercury vapor,¹¹ nitric acid fumes, nitrogen dioxide, high concentrations of inspired oxygen, paint remover, smoke,¹² smoke bombs (ZnCl₂), sulfur dioxide, and war gases (phosgene and nitrogen mustard).⁸ It is plausible, therefore, that inhalation of a potentially toxic mix of chemicals into the lungs (vaping and dabbing) could cause these injury patterns. Such chemicals include flavoring compounds such as diacetyl, 2,3-pentanedione, volatile organic compounds, ultra-fine particles, and heavy metals such as nickel, tin, and lead. Since e-cigarette aerosols can reach high temperatures, it is also conceivable that thermal injury may play a role in some patients. Finally, the fact that all patients in our series—and many in the reported literature—report vaping THC-containing products, rather than nicotine-containing products alone, raises the possibility that THC or a component in THC-containing vape fluid may be implicated in vaping-associated pulmonary illness. The lung biopsy findings described in our cases are not specific for vaping. Both DAD and organizing pneumonia have a long list of potential etiologies, which are only rarely diagnosable on the basis of histologic findings.^13,14 Examples of histologically identifiable etiologies of acute lung injury include viral cytopathic changes, fungal organisms, parasites, and aspirated food particles. In contrast, most other etiologies, including toxic inhalants, toxic drugs, aspirated gastric acid, connective tissue diseases, and most viruses, lack pathognomonic pathologic findings.^13-16 A pathologic diagnosis of acute lung injury is therefore a mixed blessing. On one hand, it provides strong and helpful evidence of moderate to severe acute injury to the lung and affords a histologic correlate for abnormal imaging findings. On the other hand, it seldom implicates a specific etiologic agent. These principles also apply to vaping-related lung injury.

The results of this series strongly suggest that vaping does not cause true exogenous lipoid pneumonia. The label “lipoid pneumonia” has been associated with vaping in several reports based mainly on the presence of oil red O–positive lipid-laden macrophages in BAL fluid.^1,5,7 One of the first cases in which oil red O–positive lipid laden macrophages were found was reported by McCauley et al⁷ in 2012. On this basis, the findings were labeled “exogenous lipoid pneumonia” and attributed to vegetable glycerin used in nicotine solutions, although morphologic features of the macrophages were not illustrated and no lung biopsy was performed. A case study from 2018 reported biopsy findings in a surgical lung biopsy specimen from a patient with suspected vaping-associated pulmonary illness; the biopsy specimen was reported as “suggestive of lipoid pneumonia,” although review of the illustrations does not show classic features of exogenous lipoid pneumonia.⁶ A more recent series of six cases of vaping-related lung disease from Utah reported oil red O–positive lipid-laden macrophages in BAL fluid. The authors felt that since imaging in these individuals did not show the low attenuation typical of exogenous lipoid pneumonia, these macrophages could not be explained by aspiration of exogenous lipoid material. They hypothesized that these cells might be a useful marker of the disease.⁵ Similarly, Davidson et al¹ labeled their cases of vaping-related lung disease as “acute lipoid pneumonia” based on the presence of oil red O–positive macrophages on BAL in three of five patients. We would like to emphasize that foamy (lipid-laden) macrophages are a common finding in BAL specimens and lung biopsy specimens and are not specific for aspirated or inhaled lipid. There are many potential explanations for their presence, the most common being airway narrowing or obstruction. “Exogenous” causes are less common, the most frequent being inadvertent aspiration of oily substances such as mineral oil–containing laxatives into the lower respiratory tract. The macrophages seen in biopsy specimens from vaping-related lung injury are most similar in morphology to those commonly seen in so-called endogenous lipoid pneumonia, which is not a true entity but merely a nonspecific histologic lesion characterized by the accumulation of foamy (lipid-laden) macrophages with fine intracytoplasmic vacuoles within alveolar spaces.¹⁷ The lipid in these cells is derived from cholesterol in the cell membranes of endogenous cells and is presumably phagocytosed by macrophages during clearance of cell debris from the lung. In contrast, exogenous lipoid pneumonia is a well-defined entity characterized by areas of low-attenuation (similar to fat) consolidation on chest CT and macrophages with coarse intracytoplasmic vacuoles on histology, along with frequent involvement of the interstitium by lipid vacuoles surrounded by foreign body–type giant cells.^17-19 The difference between macrophage morphology in endogenous lipoid pneumonia and exogenous lipid pneumonia is illustrated in Image 5. With the caveat that our experience is limited, we believe that oil red O staining lacks specificity and is potentially misleading in this setting. However, since this impression is based on only a small number of cases, we do not believe that we are in a position to issue a strong recommendation regarding discontinuation of this stain at this point. A potential area for future studies is to systematically examine the specificity of this stain.

Image 5

Exogenous vs endogenous lipoid pneumonia. A, Exogenous lipoid pneumonia (for comparison only; not from the current series). Note large coarse vacuoles within macrophage cytoplasm (H&E, ×400). B, Endogenous lipoid pneumonia, case 3. Macrophages are “foamy,” with small fine vacuoles (H&E, ×400).

The limitations of this report include the small number of cases and the predominance of transbronchial biopsy specimens. It is conceivable that as more cases are reported, additional pathologic findings will emerge and the histologic spectrum of vaping-related lung injury will widen. It is also possible that examination of surgical lung biopsy specimens or autopsy samples will yield additional histologic information.

In summary, lung biopsy specimens from this small series of patients with severe pulmonary disease associated with electronic cigarette product use (vaping) show histologic features of acute lung injury, supporting the contention that inhalation of the aerosol generated by electronic cigarettes (vaping) can cause lung damage.