Molecular iodine is not responsible for cytotoxicity in iodophors.

BACKGROUND: Ten percent povidone-iodine (PVP-I) was initially promoted as 'tamed iodine' as the chemical activity of the active biocide, uncomplexed or free molecular iodine (I2), is reduced 30- to 50-fold compared with Lugol's solution. The idea that I2 is responsible for topical iodine staining and irritation remains widely held. However, there are no controlled studies that characterize the cytotoxicity and staining of the hydrophobic I2 species compared with the other hydrophilic iodine species that comprise over 99.9% of the total iodine in topical iodine disinfectants. AIMS: To compare the staining properties of the I2 species with other topical iodine disinfectants; to evaluate if the concentrations of I2 in diluted PVP-I used to reduce severe acute respiratory syndrome coronavirus-2 in the nasal cavity are potentially cytotoxic; and to determine if high concentrations of I2 can be delivered beyond the stratum corneum into the hypodermis, which could provide a mechanistic rationale for I2 out-gassing.
METHODS: Five liquid compositions that contained complexed and uncomplexed (free) I2 in aqueous and non-aqueous carriers were used to evaluate the interaction of I2 with mammalian cells in culture as well as human and pig skin.
FINDINGS: Concentrations of I2 (7800 ppm) that are 1500 times higher than that found in PVP-I can be applied to skin without irritation and staining. I2 is not cytotoxic at concentrations >100 times higher than that found in PVP-I, and does not contribute materially to staining of skin at concentrations found in Lugol's solution (approximately 170 ppm). I2 can partition into hypodermis tissue, remain there for hours and out-gas from skin. PVP-I and Lugol's solution are highly effective topical disinfectants, but do not facilitate diffusion of I2 through the stratum corneum.
CONCLUSION: The maximum concentration of I2 found in diluted PVP, approximately 25 ppm, is not cytotoxic or irritating. The potential clinical utility of I2 has been limited by incorporating this broad-spectrum biocide into acidic aqueous formulations that contain numerous chemical species that contribute toxicity but not biocidal activity. I2 can be delivered topically into hypodermis tissue without irritation.

Introduction

The global pandemic has induced clinicians to identify agents that can be used in the oral and nasal cavities [[1], [2], [3], [4], [5], [6]]. Use of an iodine disinfectant is logical as: (i) iodine has established antiviral and antimicrobial activity; (ii) there is no evidence that bacteria can develop resistance to I2 [7,8] as it reacts with several functional groups, thus providing simultaneous action against multiple molecular targets; (iii) iodine is a naturally occurring biomolecule synthesized in the follicular lumen and other iodine-concentrating tissues; and (d) iodine has a 200-year history of use in and on humans. The nose is the primary nidus of infection for severe acute respiratory syndrome coronavirus-2, and while masks mitigate aerosol spread, they do nothing to reduce viral load in the nasal cavity [9,10]. The 3M Skin and Nasal Antiseptic is 5000 ppm available iodine iodophor that is painted on to the nasal cavity prior to surgery [11], and many clinicians have advocated for use of a diluted iodophor solution with 0.25–1.5% thiosulphate titratable as a prophylactic nasal lavage [[12], [13], [14], [15], [16]] for healthcare personnel. The US Food and Drug Administration has reviewed at least 27 clinical trials that use diluted povidone-iodine (PVP-I) [17] to control coronavirus disease 2019. Oral and nasal application of diluted PVP-I nasal spray and irrigation [18,19] has been adopted, and products with lower total iodine concentrations are also offered for this purpose (ioShield Nasal Barrier Cream, Iotech International, Boca Raton, FL, USA; Halodine Nasal Antiseptic, Miami, FL, USA).

The thiosulphate titratable iodine in PVP-I can vary from the amount stated on the label by 35% as the US Pharmacopeia (USP) definition requires not less than 85% and not more than 120% of the labelled amount of iodine [20]. The term ‘iodine’ in this sentence means thiosulphate titratable iodine. However, triiodide, hypoiodous acid and I2 are all reduced by thiosulphate, and the concentration of active biocide in PVP-I is 0.02–0.06% of the label claim. In addition, there is up to 6.6% iodide ion in PVP-I. Consequently, clinicians do not know the concentration of active biocide in the iodophor they use clinically. Nuckolls has highlighted the risks and limitations inherent in the simple dilution of PVP-I to generate a ‘low concentration’ iodine disinfectant which can have a 10-fold higher concentration of active biocide (i.e. I2) [21].

Gottardi made the surprising discovery that topically applied I2 can diffuse into skin and provide prolonged (>12 h duration) epidermal antibacterial activity due to a continuous flux or ‘back diffusion’ of absorbed I2 from skin [22]. Dilution of an iodophor such as PVP-I results in an increase of I2 concentration [[23], [24], [25]] that also increases absorption of I2 into skin, as the I2 flux is proportional to exposure time and concentration. The interaction of PVP-I with skin and wounds has been studied extensively, but these studies do not provide insight into the interaction of the I2 species with skin. Only a limited number of studies have examined the topical properties of pure I2. Duan et al. published acute dermal toxicity results for a low concentration (35 ppm) I2 composition [26], and Uchiyama et al. examined the efficacy of a 400 ppm I2 composition to reduce meticillin-resistant Staphylococcus aureus in the nasal carriage of mice [27]. The present study compares the interaction of I2 with skin using (i) I2-glycerine compositions and (ii) several commercially available aqueous iodine compositions.

Methods

Staining of human skin by iodine compositions

The forearm of a volunteer was cleaned with a Dacron wipe (75% alcohol) prior to application of the iodine compositions. Four glass vials (VWR Cat#66022-300) were held simultaneously against a human forearm for 3 min. Each vial contained 2 mL of the following compositions: (i) 10% PVP-I, (ii) 300 ppm I2 in water, (iii) 7800 ppm I2-glycerine, (iv) Lugol's solution (3.3% I2–6.6% NaI), and (e) USP tincture of iodine (47% ethanol; 2% I2–2.4% NaI). After removal of the vials, the treated area of skin was cleaned twice with an alcohol Dacron wipe to remove residual material, and a stopwatch was started.

Staining of pig skin with 66,000 ppm I2-glycerine

Thirty microlitres of a 66,000 ppm I2-glycerine solution was applied to the epidermis of a 1.77-cm2 piece of pig skin and spread over the surface for 20 s using a sterile disposable plastic loop. The material was retained on the skin surface for 3 min. The stratum corneum was removed, and a SenSafe Iodine Check test strip (Part No. 480018, Industrial Test Systems, Rock Hill, SC, USA) was contacted with the area of hypodermis tissue that had a purplish hue for 1 min to test for the presence of I2.

Staining of hypodermis pig skin tissue with I2 compositions

The epidermis and dermal layers of pig skin were separated and discarded. Cubes of hypodermis tissue 5 mm in width and length were sectioned, hydrated for 30 min in water, and patted dry. Five cubes of hypodermis tissue were submerged into 1 mL of each of the following three compositions: PVP-I, Lugol's solution, and glycerine with 15,200 ppm I2. Each piece of hypodermis tissue was removed after 15 min and rinsed thoroughly with water until a colour change was not detected. The tissue was contacted with a SenSafe Iodine Check test strip for 5 min to check for I2.

I2 out-gassing from hypodermis tissue

Ten cubes of hypodermis tissue 5 mm in width and length were hydrated for 30 min, patted dry, submerged in 2 mL of 15,200 ppm I2-glycerine for 15 min, and then rinsed thoroughly seven times with water. At 15, 30, 45, 60, 75 and 90 min, and every 30 min thereafter until 210 min, a cube was removed and contacted with a SenSafe Iodine Check test strip for 5 min to test for the presence of I2.

Out-gassing of I2 from pig skin

Circular pieces (1.77 cm2) of pig skin were cut, hydrated for 30 min at room temperature (20 °C), and patted dry with a tissue. The pig skin was treated with 10 μL of a 66,000 ppm I2-glycerine composition, spread over the epidermis for 20 s using a disposable plastic cell spreader. After 10 min, residual material was removed from the epidermis by wiping the top surface of the pig skin with an alcohol Dacron wipe. Controls were treated with 10 μL of glycerine. Pig skin tissues were loaded individually into a vertical diffusion cell (VDC; Copley Transdermal Tester HDT 1000; Cat. No. 7290, Serial #50290) in an open configuration. Tissues were kept at a constant 34 °C in the VDC to mimic the temperature of human skin. Tissues were placed between two support washers, and the donor chamber was screwed into place. The occlusion disc was used intermittently by resting on top of the donor chamber to avoid loss of I2 when not in use. I2 diffusion from skin was measured using 1.0 mM N,N-dimethyl-p-phenylenediamine dihydrochloride (DPD) prepared daily. DPD reagent (1 mL) was loaded on to the pig skin in the donor chamber of the VDC to directly capture I2 out-gassing from the epidermis. The DPD reagent was removed after a defined contact time that ranged from 30 s to 15 min, transferred to a cuvette, and absorbance was determined at 550 nm (Persee Analytics, Inc., Auburn, CA, USA; Model T6V). Pig skin was rinsed with two 1-mL DI-water rinses between DPD sampling to avoid cross-contamination. The I2 flux (in μg/cm2-min) was determined by comparing the absorbance of unknowns with a standard curve prepared daily. The between-day mean (N=25) and standard deviation of the slope and intercept for the standard curve were 0.05217 A/μg ± 0.00236 and 0.01163 A ± 0.00959, respectively.

Cytotoxicity of 24-h exposure to 1500 ppm I2

Three samples of 1500 ppm I2-glycerine were evaluated for cytotoxicity using a direct contact test according to ISO 10993-5:2009 [28]. Sterile filter paper with a flat surface with a total surface area of 1.0 cm2 was saturated with 0.1 mL of the test article, and placed directly on a cell culture monolayer of mouse fibroblasts (ATCC CRL-2648) in the centre of a 10-cm2 well. Triplicate-positive (Encore Latex glove; Ansell Healthcare Products, LLC, Iselin, NJ, USA) and -negative (high-density polyethylene; USP, Lot K0M357) controls were tested in the same manner as the test articles. All wells were incubated for not less than 24 h at 37 ± 1 °C in a humidified incubator with 5 ± 1% CO2. After incubation, the test articles and controls were removed from the wells, and the cell cultures were examined under an inverted microscope with 100X magnification for cytotoxic response.

Out-gassing of I2 from human skin

The forearm of a volunteer was cleaned with a Dacron wipe (75% alcohol) prior to application of Lugol's solution or PVP-I. A borosilicate glass vial with an inner diameter of 1.1 cm was contacted with an area on the forearm for 3 min, and the treated surface was cleaned with an alcohol Dacron wipe. An identical vial that contained 1 mL of DPD reagent was placed on the treated area for 3 min to capture I2 diffusing from the skin. I2 concentrations were determined as described for out-gassing of I2 from pig skin.

Results

A glycerine composition with 7800 ppm of I2 did not stain the skin on the forearm of a 70-year-old white male (Figure 1 ), in contrast to four other topical iodine formulations that contained much lower I2 concentrations. PVP-I, which contains 2–8 ppm I2, discoloured skin slightly more than a freshly prepared aqueous 300 ppm I2 solution. Lugol's solution, which contains 170 ppm free I2, stained skin more deeply than any of the other materials, although it contains less (≈56%) I2 than the aqueous I2 solution and 98% less I2 than the glycerine composition. Lugol's solution and iodine tincture imparted the deepest stains; this implicates triiodide as the staining agent, as both of these products contain high concentrations of uncomplexed triiodide. Staining with I2-glycerine required 15,000 ppm I2 (data not shown), and the discolouration from this higher level of I2 dissipated substantially 10 min post application.

Figure 1

One millilitre of (a) 300 ppm I2 in water; (b) 10% povidone-iodine; (c) 7800 ppm I2-glycerine, (d) Lugol’s solution and (e) iodine tincture were simultaneously contacted to the forearm of a volunteer for 3 min and residual was removed with a Dacron alcohol wipe. Images were taken immediately after application (top) and after 5 min (bottom).

Sodium thiosulphate has been used to remediate skin staining by topical iodine, which suggests that either triiodide, I2 or both iodine species stain skin. I2 presents as a yellowish-brown solution in water, in contrast to the violet colour observed in aprotic (non-polar) solvents or in the vapour phase. This study confirmed that pure I2 can stain skin by applying 30 μL of a 66,000 ppm I2-glycerine composition on pig skin. A deep reddish-brown stain (Figure 2 a) formed on the stratum corneum, and mottled areas were observed in hypodermis tissue that exhibited a purplish colour (Figure 2b). The purplish sections of hypodermis tissue are consistent with the presence of I2 in a non-polar environment. Contacting a SenSafe Iodine Check test strip with the purplish areas of hypodermis indicated the presence of I2. The ability of I2 to partition in hypodermis tissue was further tested by submerging cubes of hypodermis tissue from pig skin in PVP-I, Lugol's solution and 15,200 ppm I2-glycerine for 15 min. PVP-I did not stain hypodermis tissue, and a 5-min contact with a SenSafe Iodine Check test strip did not detect I2 (Figure 3 a). In contrast, staining was observed in hypodermis tissue treated with Lugol's solution and 15,200 ppm I2-glycerine (Figure 3b,c); a much darker stain was observed with the higher concentration of I2 in the glycerine composition. Contacting thoroughly washed hypodermis tissue previously exposed to Lugol's solution or 15,200 ppm I2-glycerine with a SenSafe Iodine Check test strip for 5 min yielded a concentration of 1 and 5 ppm I2, respectively; this is consistent with the idea that I2 diffusion into hypodermis tissue is concentration dependent. After 15 min of exposure to 15,200 ppm I2-glycerine, I2 out-gassing was detected from cubes of hypodermis tissue (Table I ) for up to 90 min.

Figure 2

(a) Iodine staining of pig skin epidermis after application of 30 µL of 66,000 ppm I2-glycerine to a 1.77-cm2 circular piece of pig skin. (b) I2 detected (aqua colour) in purplish hypodermis tissue using a SenSafe Iodine Check test after removal of the epidermal and dermal skin layers.

Figure 3

Cubes of hypodermis tissue (5 x 5 mm) were sectioned, hydrated for 30 min and then patted dry. Cubes of hypodermis tissue were submerged in 1 ml of the following three compositions: (a) 10% povidone-iodine (b) Lugol’s solution, and (c) glycerine with 15,200 ppm I2. Each piece of hypodermis tissue was removed after 15 min and rinsed thoroughly with water until a colour change was not detected. The degree of staining was proportional to I2 exposure.

Table I

Iodine (I2) flux from hypodermis tissue treated with 15,200 ppm I2 vs time

Time post exposure (min)	I2 ppm scorea	Time to final colour (s)b
15	5	25
30	5	20
45	5	18
60	2	120
75	1	180
90	0.5	240
120–210c	0	300

SenSafe Iodine Check test strip incorporates a qualitative scale from 0 to 5 ppm I2.

All observations monitored for 300 s.

The 120-, 150-, 180- and 210-min time points yielded identical results.

I2 out-gassing from pig skin treated with 66,000 ppm I2-glycerine and maintained at 34 °C in a vertical diffusion cell was detected for up to 2 h (Figure 4 ). The I2 flux decreased exponentially with time, as reported previously by Gottardi [2]. The I2 flux from human skin measured after a 3-min 10% PVP-I application was 0.49 ± 0.047 μg/cm2-min (N=5) at 0.5 min post application. Identical measurements made on human skin with Lugol's solution at 0.5, 30 and 80 min post application were 5.39 ± 1.48, 1.32 ± 0.44 and 0.41 ± 0.24 μg/cm2-min, respectively. These results are consistent with data reported previously by Gottardi [22], who demonstrated residual antibacterial activity against bacteria in water.

Figure 4

I2 flux (µg/min-cm2) at 34oC from a hydrated 1.77-cm² piece of pig skin treated with 10 µL of a 66,000 ppm I2-glycerine composition.

A 1500 ppm I2-glycerine composition did not exhibit cytotoxicity in a 24-h direct contact test with mouse fibroblasts, which contrasts with cell-based cytotoxicity reports using commercially available topical iodine products [29].

Discussion

In a series of studies funded by the National Aeronautics and Space Administration, Thrall et al. demonstrated dramatic differences in the pharmacological and toxicological properties of orally administered iodide vs I2 in the rat [[30], [31], [32]]; the results support the principal that the functional properties of an iodine composition depend upon the relative concentration of the different iodine species contained therein. Aceves and others have demonstrated dramatic differences in the behaviour of I2 compared with iodide in different cell lines and tissues [30,31,33,34]. There are more than seven different iodine species in aqueous topical iodine disinfectants [35], in addition to other ingredients (e.g. surfactants, water, pH modifiers, etc.). All of the chemicals in these iodine products can contribute to their functional behaviour [36,37]. The present study observed the interaction of skin with pure I2, and noted differences between the behaviour of pure I2 and complex mixtures of iodine species (e.g. PVP-I).

Thiosulphate has been used successfully to treat iodine burns, which suggests that either triiodide, I2 or both species are implicated as the causal skin staining agent as thiosulphate reduces both iodine species. The differences in human skin staining exhibited by the iodine compositions in this study are inconsistent with the long-held assumption that topical staining is due to I2 [38]. The complete lack of staining from a 7800-ppm composition of I2-glycerine contrasts dramatically with the deep persistent stain imparted by both Lugol's solution and iodine tincture. This lack of staining at 7800 ppm I2 combined with an absence of cytotoxicity with 1500 ppm I2 in a 24-h direct contact test with fibroblasts suggests that high concentrations of I2 can be used safely on skin if properly formulated. Skin staining from Lugol's solution and iodine tincture may be due to the binding of triiodide to glycogen in the epidermis [39].

PVP-I is visible on skin but it does not stain skin, as diffusion of triiodide or I2 into the epidermis is limited due to binding of I3 by PVP. I2 stains skin at higher concentrations but the resulting colour differs from that observed with triiodide staining; a difference in colour is also observed in in-vitro experiments that characterize glycogen-iodine complexes [40]. Systemic iodine exposure from topical iodine disinfectants such as PVP-I is well known [[41], [42], [43], [44]]. The total iodine applied topically in a glycerine-I2 composition would be lower than with a PVP-I composition. For instance, a 1000 ppm I2-glycerine solution would expose 100 μg of I2 to a human hand based on 0.1 mL application on to a 400-cm2 hand, which compares with approximately 10,000 μg of total iodine from PVP-I for an equivalent volume. However, a much higher percentage of the iodine in I2-glycerine may be absorbed compared with PVP-I. Safran and Braverman evaluated thyroid status during daily vaginal douching with PVP-I (0.3% thiosulphate titratable iodine) in 12 euthyroid volunteers for 14 days, and concluded that 5% of the applied iodine was absorbed [45]. A topical administration of 0.2 mL of an I2-glycerine composition containing 1000 ppm I2 used for hand disinfection would administer 200 μg of I2 on to skin. If one assumes 20% absorption as opposed to the 5% measured by Safran and Braverman, 40 μg would cross the stratum corneum. A maximum of 50% of this I2 could be available to partition into the thyroid [[30], [31], [32]], which equates to a total thyroid exposure of 20 μg of iodide which is approximately 15% of the recommended daily allowance. Consequently, it would be important to measure the potential for iodine exposure from repeat applications of a high I2-glycerine composition.

Gottardi first observed out-gassing of I2 from skin after application of Lugol's solution [2]. In the present study, measurements of I2 flux from human skin treated with Lugol's solution are consistent with Gottardi's data. The purplish areas of coloured hypodermis in pig skin treated with 66,000 ppm I2-glycerine suggest the presence of I2, which is consistent with test results from the DPD test strip. The apparent stability of I2 absorbed into skin is ostensibly a surprising outcome as I2 is highly reactive with biological tissue. However, partitioning of I2 into regions of unsaturated lipid in the hypodermis could provide stability to I2, which partitions readily into oil and fat. It was observed that I2 partitioned into hypodermis tissue in proportion to the dose, as tissue treated with 15,200 ppm I2 contained more I2 than tissue treated with Lugol's solution (170 ppm). The duration of out-gassing from pig skin treated with 66,000 ppm I2 was detected for approximately 3.3 h, compared with 4 h for hypodermis tissue treated directly with 15,200 ppm I2.

Gottardi mentioned that absorption and subsequent post-application diffusion of a topically applied antimicrobial over time is unique in the field of skin disinfection. Some researchers have reported a persistent antimicrobial activity on skin after use of iodine-based handwashes [[46], [47], [48], [49]]. Converting hypodermis tissue into a material that releases I2 over time confers potential benefits for topical disinfection, and also offers the potential to: (i) treat infectious diseases, (ii) act as an anti-inflammatory as I2 can interrupt ligand signalling via iodination, and (iii) stimulate wound healing via the sustained delivery of free iodine at concentrations that retain antimicrobial activity without cytotoxicity [50].

It is known that the concentration of I2 increases as PVP-I is diluted up to approximately 1/100 [23]. Diluted PVP-I has been shown to be less cytotoxic than neat PVP-I; therefore, the literature provides indirect evidence that I2 is not the cytotoxic species in PVP-I. Additionally, cadexomer iodine, which contains a high concentration of I2 compared with 10% PVP-I, does not affect fibroblast viability and collage synthesis. In fact, cadexomer iodine stimulated secretion of proinflammatory cytokines (tumour necrosis factor-alpha) in human macrophages [23].

The characterization of PVP-I on product labels is not sufficient to ensure consistent results in cell-based cytotoxicity studies due to variations in pH (can vary between 1.5 and 6.5), iodide concentration, ionic strength, buffering capacity and I2 concentration [24,25]. Additionally, there is a 15% variance allowed in the level of titratable iodine under the definition in the USP. There is a legitimate concern that formulation variations among different PVP-I compositions (manufacturer-to-manufacturer and lot-to-lot) have influenced cytotoxicity study outcomes unbeknownst to researchers conducting these studies [29,51,52]. Equally problematic are studies that dilute a topical iodine composition and draw incorrect conclusions based on the assumption that I2 is stable for days after dilution [53]. Some manuscripts have thoughtfully explored the composition of PVP-I [54], but an insufficient understanding of the fundamental chemical properties of PVP-I almost certainly contributes to unfortunate clinical outcomes [[55], [56], [57]].

If one views I2 as just another antimicrobial chemical, the lack of cytotoxicity observed for the 1500-ppm I2-glycerine material may appear paradoxical given the elevated antimicrobial activity associated with such a high level of I2. However, molecular iodine is a naturally occurring biochemical that can function as an antioxidant [[58], [59], [60]], an anti-inflammatory agent [61,62], an antiproliferative agent [35,[63], [64], [65], [66], [67], [68]] and a differentiation agent [[69], [70], [71], [72], [73]]. There are more than 12 different tissues in humans with the ability to concentrate iodine. Based on a recommended daily allowance of 125 μg/day for iodide, the follicular lumen oxidizes 6.25 μg iodide/h into I2 in a volume of approximately 7.5 mL, which equates to generating 900 ppm I2/h which is approximately 2 orders of magnitude higher than the I2 concentration found in PVP-I. Molecular iodine is the only widely used disinfectant molecule that plays an essential role in mammalian biochemistry [74] and, if formulated properly, can be used clinically at highly elevated concentrations.

In conclusion, the highest concentration of I2 achievable from diluted PVP-I is approximately 25 ppm, which suggests that there is at least a 60-fold margin of safety for I2 with respect to topical toxicity on human mucous membranes of the nasal cavity, as a 1500-ppm I2 composition was not found to be cytotoxic. The staining and irritation associated with topical iodine disinfectants is not due to the I2 species. There are potential clinical benefits that may accrue from the use of high concentrations of topically applied I2.