Home

see peer reviewed Medical Hypothesis article @;

https://www.sciencedirect.com/science/article/pii/S0306987720312007

Clinical Evidence that EDTA can Reduce COVID-19 Infections

Recent medical articles indicate that there is  lower reported prevalence of asthma and COPD in patients diagnosed with COVID-19 ;see < https://www.thelancet.com/action/showPdf?pii=S2213-2600%2820%2930167-3> and < https://www.nejm.org/doi/pdf/10.1056/NEJMoa2004500?articleTools=true>. In addition, The New York Times also weighed in with an article underscoring that “Asthma Is Absent Among Top Covid-19 Risk Factors, Early Data Shows” ;< https://www.nytimes.com/2020/04/16/health/coronavirus-asthma-risk.html?campaign_id=154&emc=edit_cb_20200416&instance_id=17710&nl=coronavirus-briefing&regi_id=101192890&segment_id=25349&te=1&user_id=43d012015e0efa5e799fec00b69b5c02

Also see peer reviewed article:

Another research paper, “Dominance of SARS-CoV-2 D614G Variant Explained by the Requirement of COVID-19 for Calcium; Proximate Therapeutic Implication(s) for COVID-19” has been published to further explain the potential benefit of EDTA to treat, mitigate and/or prevent COVID-19 . See < https://www.heraldopenaccess.us/openaccess/dominance-of-sars-cov-2-d614g-variant-explained-by-the-requirement-of-covid-19-for-calcium-proximate-therapeutic-implication-s-for-covid-19>. The digital object identifier (DOI) displayed at (< https://dx.doi.org/10.24966/CIIT-8844/1000048 >). 

As described below, computer analysis of SARS-COV-2 (COVID-19) envelope protein has a previously unrecognized domain of very strong amino acid homology to the calcium-binding  regions required for SARS-CoV-1 infection that was recently published in 2017.  This discovery was soon after the first ever calcium dependent fusion process discovered in 2014 for the human airborne pathogen, Rubella (German Measles). In particular, Rubella infection was completely abrogated by the calcium chelating compound called EDTA By recognizing that the “resistance” to COVID-19 infection in asthma patients is linked medications routinely used by patients, we have uncovered a new therapy for COVID-19. That is,  we can directly utilize EDTA to prevent and treat COVID-19 infections in non-asthmatic patients with nebulizer aerosols or by adding EDTA solutions to ventilators. These techniques are well known in respiratory therapy, the equipment is readily available, EDTA has been safely used for treatment of heavy metal poisoning and has been studied by the NIH< https://www.nccih.nih.gov/health/questions-and-answers-the-nih-trials-of-edta-chelation-therapy-for-coronary-heart-disease>.

While there may be several possibilities, Occam’s razor suggests that “the simplest solution is most likely the right one”. Here the  therapies used by patients with chronic respiratory diseases can reduce the risk of infection or of developing symptoms leading to diagnosis. Thus, examination of  common nebulizer therapies used by asthma and COPD patients shows concentrations of EDTA available in nebulizer solutions vary from 0.1 to 0.5 mg/mL < https://www.ncbi.nlm.nih.gov/pubmed/?term=16678502 >.

Moreover, Metaproterenol InhalationSolution USP, a very common asthma drug is formulated specifically with EDTA (edetate disocium) .  Specifically, Metaproterenol Sulfate Inhalation Solution USP is a unit-dose bronchodilator administered by oral inhalation with the aid of an intermittent positive pressure breathing apparatus (IPPB). It contains 0.4% or 0.6% Metaproterenol sulfate in a sterile, acidic, aqueous solution containing edetate disodium, sodium chloride, hydrochloric acid and/or sodium hydroxide for pH adjustment. see < https://www.drugs.com/pro/metaproterenol.html>. Thus, in light of the calcium-dependent infection process described below, the use EDTA in nebulized solutions is the simplest solution  to explain the reduced  risk of infection is the simplest explanation for the  lower reported prevalence of asthma and COPD in patients diagnosed with SARS-CoV-2 /COVID-19.

Overview of inhalation solutions from:
 https://www.ncbi.nlm.nih.gov/pubmed/?term=16678502

HOW EDTA WORKS TO PREVENT COVID-19 INFECTION:

Strategy:

Stop COVID-19 infection by blocking  SARS-COV-2 entry as a result of disrupting the calcium-dependent fusion process with pharmaceutical FDA approved calcium chelating agent called EDTA.

Technique:

Apply the medicine directly to lungs and respiratory track with a Nebulizer. In general, one milligram (mg) of EDTA can be administered with or without a nebulized bronchodilator such as 2.5 mg of Albuterol solution with a nebulizer or mechanical ventilator to stop virus infection at intervals of 2 hours or more. For example, one of the sterile-filled Albuterol products (manufactured by Dey Laboratories) contains 300 micrograms (ug) EDTA in each 2.5 mg dose of bronchodilator. The 300ug amount of EDTA is about one-tenth the concentration known to induce bronchoconstriction. Thus, one way to reach a dosing level of approximately 1 mg EDTA, that is 900 ug, is to repeat the nebulizer treatments three times with the 300 ug EDTA solution in the medicine cup over at least an hour time span.     

Older Reports on EDTA in FDA Approved Nebulizer Solutions

“The sum total of information to date on EDTA reassures us that the potential of a single dose of EDTA to cause bronchoconstriction is minimal with the products now available. The potential for bronchoconstriction during multiple dosing is unknown; however, given the near 10-fold difference between the concentrations of EDTA in available commercial products and the PC20 FEV1, it is likely to be of minimal risk unless multiple doses are administered over a short interval (e.g. , 1 hour, as in the ED treatment of acute airway obstruction)”; (see:< https://www.ncbi.nlm.nih.gov/pubmed/10452789>). Below is an image of a sterile unit-dose vial (manufactured by Day Laboratories) contains 300ug of EDTA per 2.5 mg dose of albuterol.

In addition to the Dey product, the table below shows, commonly used nebulizer bronchodilator solutions available in the United States from < https://www.ncbi.nlm.nih.gov/pubmed/10452789>.

Binary Virus-Receptor Model Limitations

The Coronaviruses are a large family of single-stranded enveloped RNA viruses which can be divided into four major genera; Both SARS-CoV-1 and  SARS-CoV-2 (COVID-19) belong to the β-genus.  It is thought that infection of virus occurs when its outer envelope spike protein guides the coronavirus entry into host cells by binding to a host receptor and then fusing viral and host membranes through a defined receptor-binding domain (RBD) as shown in the copied from < https://www.ncbi.nlm.nih.gov/pubmed/?term=31996437 > .

However, to date, the current published binary virus-receptor models do not, as yet, formally account for the calcium-dependent fusion loop process  corresponding  to the SARS-CoV-2 amino acid residues at  816-837 for the FL1 domain and residues 835-855 for the FL2 domain listed in (Table 2. pBlast Alignment Amino Acid Locations) shown below. Additionally, the less pathogenic Alphacoronavirus 229E (HCoV-229E), accounting for up to 30% of cases of the common, exacerbates chronic obstructive pulmonary disease (COPD) and bronchial asthma, has a similar FL2 domain to the Rubella virus whereas the Influenza H1N1 hemagglutinin (HA) protein has no significant similarity as shown in Table 2.  The similarity the HCoV-229E’s calcium-binding domain to SARS-CoV-2, albeit much less than SARS-CoV-1, suggests attenuation of HCoV-229E, which is consistent with HC0V-229E having crossed species barriers to infect humans decades or centuries ago; see <https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7103735/pdf/TEMI_9_1736644.pdf >.

Here, using the calcium-dependent fusion loops to organize virus relationships helps us find target new therapies using FDA approved medications containing calcium chelators and can help explain differences in viral pathology. That is, using data from calcium-dependent fusion loops can help us overcome the limitations of current models that focus exclusively on the receptor-binding motif (RBD) of SARS-CoV-2, exploit the calcium requirement mechanism, which may be the Achilles heel of the coronavirus.

FIG 1

Structural analysis of human ACE2 recognition by 2019-nCoV and SARS-CoV. (A) Overall structure of human SARS-CoV RBD (year 2002) complexed with human ACE2. PDB ID is 2AJF. ACE2 is in green, the core of RBD (receptor-binding domain) is in cyan, and RBM (receptor-binding motif) is in magenta. (B) Critical residue changes in the RBMs of SARS-CoV and 2019-nCoV. All these five residues in SARS-CoV underwent natural selections and were shown to be critical for ACE2 recognition, cell entry, and host range of SARS-CoV. The residue numbers are shown as in SARS-CoV RBD, with the corresponding residue numbers in 2019-nCoV shown in parentheses. For viral adaption to ACE2, > means “is more adapted”, >>> means “is much more adapted,” and = means “is similarly adapted.” Information about the two most critical residues, 479 and 487, is in red. (C) Experimentally determined structure of the interface between a designed SARS-CoV RBD (optimized for human ACE2 recognition) and human ACE2. PDB ID is 3SCI. (D) Modeled structure of the interface between 2019-nCoV RBD and human ACE2. Here, mutations were introduced to the RBD region in panel C based on sequence differences between SARS-CoV and 2019-nCoV. GenBank accession numbers are MN908947.1 for 2019-nCoV spike, NC_004718.3 for human SARS-CoV spike (year 2002; strain Tor2), AGZ48818.1 for bat SARS-CoV spike (year 2013; strain Rs3367), AY304486.1 for civet SARS-CoV spike (year 2002; SZ3), and AY525636 for human/civet SARS-CoV spike (year 2003; strain GD03). References for the other sequences are in parentheses as follows: civet SARS-CoV spike (year 2005) (9); human SARS-CoV spike (year 2008) (8).

Key Discovery #1: Rubella, The First Virus Found With Calcium-Dependent Fusion Mechanism in 2014

While the vaccine for Rubella, an airborne virus transmitted exclusively by humans and causes “German measles”, has prevented Rubella transmission in the Americas until lately, new studies on Rubella virus have recently provided a key insight into how a virus enters host cells.  Namely, it was discovered that Ca2+ was critical for productive viral entry of Rubella and the presence of EDT, which sequesters any available external calcium (Ca2+), caused a block in the Rubella E1-target membrane insertion and thus Rubella prevented infection by inhibiting virus-membrane fusion. This research, <https://www.ncbi.nlm.nih.gov/pubmed/25474548  > just recently published in December 2014 provides the first example of a Ca2+-dependent viral fusion protein.

Key Discovery #2: SARS-Cov-1 Found To Have Calcium-Dependent Fusion Loops in 2017

COVID-19, or SARS-CoV-2 is a member of the Coronaviruses (CoVs) group that comprises a collection of enveloped viruses that infect humans and animals. These viruses are extremely pathogenic when they jump from animals to humans (zoonotic transmission) and are thus named severe acute respiratory syndrome (SARS). Previously, there was a 2002-2004 SARS outbreak caused by “SARS-CoV-1” in which about 8000 people were known to be infected in 29 different countries and at least 774 people died < https://en.wikipedia.org/wiki/2002%E2%80%932004_SARS_outbreak >.   The enveloped Coranaviruses, including SARS-CoV-1, SARS-CoV-2 and Rubella virus, require an initial viral fusion step to infect host cells. The key experiments published in 2017 showed that calcium is critically important for SARS-CoV-1 membrane fusion in live cells. < https://www.ncbi.nlm.nih.gov/pubmed/?term=29056462 > These researchers found the presence of the chelating agent ethylenediaminetetraacetic acid (EDTA) abrogated the fusion process and proposed a model of fusion protein (FP1, FP2) ( also known as a fusion loops, FL)  as shown in the diagram .

Model of 2 Peptide Domains in Calcium-Dependent Mechanism;

In other words, we can envision the calcium-dependent process utilizing fusion loops as creating a  “barb” where the larger surface spike protein that contacts with the host receptor for the coronavirus is a hook to catch the victims cells.  Importantly they found that the calcium-dependent fusion process was missing from the Influenza spike protein and can biochemically explain the Influenza dichotomy between the airborne SARS-CoV-2 fatality rate that is approximately 200 to 1000 times the .02% fatality rate for Influenza. In other words, the Influenza spike protein acts as a “barbless” hook to “catch” host cells, whereas the calcium-dependent fusion loops in SARS-CoV-2  functions as “barbs” to  more securely infect (“catch”) the cells of the host.

Key Discovery #3: Computationally Based Amino Acid Sequence Homologies For Fusion Loops of Three Viruses; Rubella, SARS-Cov-1 And SARS-Cov-2 (COVID-19) in 2020.

As of April 17, 2020, the airborne COVID-19 pandemic has resulted in 2,249,662 confirmed cases and 154,254 deaths. See < https://coronavirus.jhu.edu/ > for updates. To date, no specific treatment has been proven to be effective for COVID-19 infection. The WHO has classified the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), also named COVID-19, as a β CoV of group 2B.[i]  Sequence results from patient isolates show the new beta-CoV-2 strain have 99.8–99.9% nucleotide identity.[ii] While the overall genetic sequence of the COVID-19 (SARS-CoV-2) has only 80% identity to SARS-CoV,2 there are two calcium dependent binding domains, that act as fusion loops (FL), in the COVID-19’s spike (S) surface protein that are almost 100% homologous to the SARS-CoV fusion domains and appear critical for virus entry into a host.[iii]  According to NCIB protein Blast analysis there is also a significant homology of Rubella Virus membrane glycoprotein E2 amino acids 49 to 55 with the SARS-CoV-2 fusion loop 2 as shown in Tables 1-2.  In 2014, the first virus discovered to have a calcium-requiring viral fusion protein was Rubella Virus , an airborne pathogen.[iv],[v]  Importantly, these researchers demonstrated that the addition of the calcium chelator, EDTA completely abrogated Rubella fusion (i.e., Rubella infection).

Table 1 and Table 2 exhibit the Protein Blast Alignment Tool data from the calcium binding fusion domains, labeled FL13 and FL23 respectively, that compare the spike proteins of   COVID-19 (SARS-CoV-2) with SARS-COV and Rubella utilizing cited reference data; GenBank: QHD43416.1 (CoV-2), NCBI Reference Sequence: NP_828851.1(CoV-1), GenBank: ACN50046.1.(Rubella), GenBank: CAA71056.1(Human coronavirus229E) and GenBank: AD177360.1 (hemagglutinin [Influenza A virus (A/Boston/136/2009(H1N1))]).[vi]

The high peptide homology of FL1 and FL2 domains in COVID-19 compared to SARS-CoV and Rubella justifies investigating calcium chelation therapy (e.g., (Na2EDTA)) to prevent/treat COVID-19 infection. Chelation therapy with disodium EDTA has been used “Off-Label” for more than 50 years [vii]  and EDTA has been added to nebulized bronchodilator solutions in the United States as both non-sterile and sterile-filled products for years.[viii] Nebulizers can be attached to ventilators and standalone nebulizers that aerosol Na2EDTA at a concentration of 1.2 to 12.8 mg/mL may be used provided Albuterol is available to minimize bronchoconstriction. Adding Na2EDTA to alcohol-base hand sanitizers, soaps, and sprays should further reduce COVID-19 virus infectivity.

In conclusion a protein comparison the two relevant amino acid regions from SARS-CoV-1 fusion loop and found 100% and 95% correspondence to similar regions in SARS-CoV-2 ( COVID-19) using the Protein Blast program from the National Center for Biotechnology Information as shown in Table 1 and Table 2 below. Additionally, the Protein Blast program showed significant homology to Rubella Virus as documented in Tale 1 and Table 2.

[i]  Hui DS, I Azhar E, Madani TA, et al. The continuing 2019nCoV epidemic threat of novel coronaviruses to global health-The latest 2019 novel coronavirus outbreak in Wuhan, China.  Int J Infect Dis. 2020;91:264-266

[ii] Zhou P, Yang XL, Wang XG, et al. A pneumonia outbreak associated with a new coronavirus of probable bat origin. Nature  2020; 579(7798):270-273

[iii] Lai AL, Millet JK, Daniel S, Freed JH, Whittaker GR. The SARS-CoV Fusion Peptide Forms an Extended Bipartite Fusion Platform that Perturbs Membrane Order in a Calcium-Dependent Manner. J Mol Biol. 2017 Dec 8;429(24):3875-3892.

[iv] DuBois RM, Vaney MC, Tortorici MA, et al. Functional and evolutionary insight from the crystal structure of rubella virus protein E1. Nature. 2013 Jan 24;493(7433):552-6

[v] Dubé M, Rey FA, Kielian M.Rubella virus: first calcium-requiring viral fusion protein. PLoS Pathog. 2014 Dec 4;10(12)

[vi] Boratyn GM, Camacho C, Cooper PS, et al. BLAST: a more efficient report with usability improvements. Nucleic Acids Res. 2013;41 (Web Server issue):W29-33

[vii] Lamas GA, Goertz C, Boineau R, et al. Effect of disodium EDTA chelation regimen on cardiovascular events in patients with previous myocardial infarction: the TACT randomized trial. JAMA 2013;309(12), 1241–1250

[viii] Asmus MJSherman JHendeles L. Bronchoconstrictor additives in bronchodilator solutions. J Allergy Clin Immunol. 1999;104(2 Pt 2):S53-60

Off-Label Uses Of Sterile Pharmaceutical EDTA

Off-Label uses of sterile pharmaceutical EDTA have been used for more than 50 years[1] and human studies of nebulized EDTA solutions were safely conducted more than 30 years ago. [2]   For example, there were an estimated 500,000 visits for chelation therapy in the U.S. for 1993,[3] and 800,000 in 1997[4]. A Canadian survey found that 8% of patients who had undergone cardiac catheterization had used chelation therapy.[5] In fact, a NIH RFP describes several protocols for EDTA chelation and states “For chelation therapy, the most widely used formulation is a protocol recommended by the American College for Advancement in Medicine (ACAM) (18) that includes disodium EDTA and magnesium chloride. This formulation has yet to be tested in rigorous randomized controlled trials.  EDTA chelates various divalent metal ions with differing affinities. It is poorly absorbed from the gastrointestinal tract. Following intravenous administration, EDTA is found primarily in plasma.  It is distributed in the extracellular fluid, and it does not appear to penetrate cells.  Only about 5% of the plasma concentration is found in spinal fluid. The half-life is 20-60 minutes. It is excreted mainly by the kidney, with about 50% excretion in one hour and more than 95% within 24 hours.  Almost none of the compound is metabolized.  Treatment with EDTA has a low incidence of side effects”.[6] 

While IV administration of EDTA (Figure 2 above), may, or may not, reach the lungs and upper respiratory track, the above NIH RFA indicates that EDTA stays primarily in the plasma. Inhalation of nebulized solutions can deliver Na2EDTA directly into to the very tissues attacked by COVID-19: lung alveoli, bronchi, larynx mouth nose, pharynx and throat. Aerosolized EDTA can be delivered by any of the three main inhalation systems: pressurized metered-dose inhalers (MDIs), dry powder inhalers (DPIs), and nebulizers. However, the prior human inhalation studies with EDTA solutions, [7],[8] using a nebulizer to aerosolize dissolved Na2EDTA at the concentration of 1·2 to 12·8 mg/mL suggest a safe way to test the hypothesis that EDTA treatment can attenuate COVID-19 infection.  Inhaling a nebulized EDTA solution is reasonably safe provided the patient is also monitored for potential bronchoconstriction and Albuterol treatment is available to alleviate the bronchospasm.

Since COVID-19 patients on mechanical ventilators are the most vulnerable, it is critical to be able to deliver aerosolized Na2EDTA to their lower respiratory tracts with precision and consistency to ensure patient safety. To achieve this goal, either jet or ultrasonic nebulizers can be utilized for delivering aerosols to mechanically ventilated patients. Nebulizers can be attached in the inspiratory limb of the ventilator circuit or at the patient Y-piece. Placing the jet nebulizer at a distance from the endotracheal tube has better efficiency than placing it between the patient Y-piece and the endotracheal tube, because the ventilator circuit acts as a spacer for aerosol to collect between inspirations. That said, utilizing standalone nebulizers to aerosol Na2EDTA may provide an supplementary treatment option as the in-patient beds become overloaded with COVID-19 patients.

Without any proven treatment for COVID-19 available, the potential for EDTA to both reduce COVID-19 transmission and treat infection through relatively safe modalities that include nebulizing EDTA solutions (in conjunction with Albuterol to minimize bronchoconstriction if needed), chelation therapy, and adding EDTA to hygienic products warrant further investigation(s).  “Off-Label” use of Na2EDTA can enable evaluation of the clinical response of nebulized Na2EDT dissolved in normal saline at a range up of to 1·2 to 12·8 mg/mL for the treatment of COVID-19 patients on respirators, provided the patients are monitored for signs of bronchial constriction, and administer Albuterol as needed. Alternatively patients without access to a respirator can be similarly treated with Na2EDTA solution through a nebulizer facemask under direct medical supervision.  Then, if tolerated, the more stable patients may be treated at home or as outpatient with a nebulizer and Na2EDTA/Albuterol solutions.

Finally, adding EDTA to alcohol-base hand sanitizers, lotions, sprays and soaps should further reduce COVID-19 virus infectivity. The Cosmetic Ingredient Review Expert Panel found that EDTA ingredients are safe as used in cosmetic formulations. The typical concentration of use of EDTA in cosmetics is less than 2%, and the lowest dose reported to cause a toxic effect in animals was 750 mg/kg/day.[9]

[1] Lamas GA, Goertz C, Boineau R, Mark DB, Rozema T, Nahin RL, et al. Effect of disodium EDTA chelation regimen on cardiovascular events in patients with previous myocardial infarction: the TACT randomized trial. JAMA 2013 309(12), 1241–1250

[2] Beasley CR, Rafferty P, Holgate ST, Bronchoconstrictor properties of preservatives in ipratropium bromide (Atrovent) nebulizer solution. Br Med J (Clin Res Ed). 1987 May 9;294(6581):1197-8.

[3] Grier MT, Meyers DG. So much writing, so little science: a review of 37 years of literature on edetate sodium chelation therapy.  Ann Pharmacother 1993;27:1504-1509

[4] See NIH RFA  < https://grants.nih.gov/grants/guide/rfa-files/RFA-AT-01-004.html> EDTA CHELATION THERAPY FOR CORONARY ARTERY DISEASE Release Date:  April 30, 2001  RFA:  RFA-AT-01-004

[5] Knudtson ML, Wyse DG, Galbraith PD, Brant R, Hildebrand K, Paterson D,  et al. Chelation therapy for ischemic heart disease: a randomized controlled trial. JAMA. 2002 Jan 23-30;287(4):481-6.

[6] See NIH RFA  < https://grants.nih.gov/grants/guide/rfa-files/RFA-AT-01-004.html>EDTA CHELATION THERAPY FOR CORONARY ARTERY DISEASE. Release Date:  April 30, 2001  RFA:  RFA-AT-01-004 National Center for Complementary and Alternative Medicine  <http://nccam.nih.gov>

[7] Asmus MJSherman JHendeles L. Bronchoconstrictor additives in bronchodilator solutions. J Allergy Clin Immunol. 1999 Aug;104(2 Pt 2):S53-60

[8] Beasley CR, Rafferty P, Holgate ST, Bronchoconstrictor properties of preservatives in ipratropium bromide (Atrovent) nebulizer solution. Br Med J (Clin Res Ed). 1987 May 9;294(6581):1197-8.

[9] Lanigan RSYamarik TA. Final report on the safety assessment of EDTA, calcium disodium EDTA, diammonium EDTA, dipotassium EDTA, disodium EDTA, TEA-EDTA, tetrasodium EDTA, tripotassium EDTA, trisodium EDTA, HEDTA, and trisodium HEDTA. Int J Toxicol. 2002;21 Suppl 2:95-142

Coronavirus Survival Is Much Greater Than Influenza; Why HAND WASHING & MASKS Are So Important

5 World Health Organization. Annex C: Respiratory droplets. In: Atkinson J., Chartier Y., Pessoa-Silva C.L., editors. Natural ventilation for infection control in health-care settings. WHO; Geneva: 2009

The lethality SARS-Cov-19 is clearly distinguishable from Influenza and the nonflexible calcium-dependent fusion loops may help explain why the coronaviruses have such a much longer capacity to survive on dry surfaces compared to Influenza and underscore the needs to wear masks and thorough hand washing. For example, viable MERS-CoV was recovered after 48 h, whereas no viable H1N1 was recovered after 1 hour in any conditions tested.  That said, The risk of indirect contact transmission is inexact compared with the other transmission routes, airborne routes, direct contact transmission and droplet dispersion because Influenza virus and SARS-CoV shed into the environment at levels far in excess of the infective dose, but they can survive for extended periods on surfaces, and sampling has identified contamination of hospital surfaces. Contaminated surfaces, which contribute to fomite transmission, can result in onward contamination of hands or equipment, which then initiate inoculation through contact with the nose, eyes, or mouth.  See: < https://www.ncbi.nlm.nih.gov/pubmed/26597631>