Effects of tetracycline and clotrimazole ointments in the treatment of palmar arsenical keratosis

The study was designed to see the effects of tetracycline and clotrimazole on the altered skin flora (Aspergillus and Enterobacter species) in 38 patients with palmar arsenical keratosis. The skin swab and scrapping samples were collected. Tetracycline, clotrimazole and their combination were given as ointment for 3 months. Clotrimazole and tetracycline were found to inhibit the growth of Aspergillus spp. and Enterobacter spp. in vitro. A pH-dependent inhibition of the growth of microorganisms in presence of these antimicrobials was observed, which was highest at pH 8. The percentage reduction of keratotic nodular size was 32.9, 66.1, 61.1 and 32.5 in the groups treated with placebo, tetracycline, clotrimazole and the combination of both, respectively. But none of the interventions was proved statistically significant. No remarkable adverse effect was reported. In conclusion, clotrimazole or tetracycline inhibits the growth of Aspergillus spp. or Enterobacter spp. in vitro. However, there was no statistically significant clinical improvement of the palmar arsenical keratosis.


Introduction
Exposure to arsenic through contaminated groundwater is a major public health problem worldwide (Khan et al., 1997). Prolonged ingestion of arsenic above the safe level gives rise to a chronic health condition, termed as arsenicosis which is manifested usually by skin lesions, like-melanosis, leucomelanosis and keratosis, with or without involving the internal organs (Caussy, 2005). Arsenical keratosis, appearing on the palmar and plantar aspect of the hand and foot, is more troublesome for the patient as it affects the patient's socio-economically as well (Shahidullah et al., 2001;Hassan et al., 2005;de Luzuriaga, 2011;Safiuddin et al., 2011).
The cause of keratosis is not well-understood. Despite exposure to the same source of arsenic-contaminated groundwater, not all the members of the family get affected by arsenicosis. It had been found in certain studies that there is an alteration of normal skin flora in arsenicosis, where Aspergillus spp. and Enterobacter spp. were found to be significantly present on the palm of the patients with arsenical keratosis in comparison to the control population (Khalil et al., 2016;Moitra et al., 2018). But the relationship between such altered skin flora and the development of arsenical keratosis is not established or explored.
Reconstructive surgery with total excision of the keratotic skin followed by grafting is also used in cases of severe arsenical keratosis.
There is no such study where the effects of antimicrobials have been evaluated as the treatment of arsenical keratosis. So, the present study was designed to see the effects of topical antimicrobials on the altered skin flora (Aspergillus spp. and Enterobacter spp.) in patients with palmar arsenical keratosis.

Place and duration of the study
The study was conducted at three places: a) The Department of Pharmacology, Bangabandhu Sheikh Mujib Medical University, b) The Department of Microbiology and Immunology, Bangabandhu Sheikh Mujib Medical University, and c) Eruain Village of Kandirpara Union of Laksham Upazilla of Cumilla District. This study started in September 2017 and continued up to January 2019.

Determination of sample size
The present study was intended to propose or generate a hypothesis that the alteration of skin microbial flora may cause the development of keratosis. As there is no such work previously conducted, samples were taken as much as possible from those who fulfilled the selection criteria.

Selection of arsenic-endemic area
According to the Department of Public Health Engineering of the Government of Bangladesh, Cumilla is one of the worst arsenic-affected areas. After visiting multiple arsenic-affected endemic areas, Laksham, one of the Upazillas of Cumilla District was selected to be the study area as the local authority of the Upazilla as well as the patients found to be co-operative.

Selection criteria
In total, 38 patients with moderate to severe palmar arsenical keratosis were enrolled as per the inclusion and exclusion criteria of the study (Figure 1).
Inclusion criteria include a) Age: 18 to 60 years, b) drinking arsenic-contaminated water (>50 µg/L) for more than 6 months, c) patients with moderate to severe palmar arsenical keratosis, d) patients voluntarily agreed to participate, e) patients who did not receive topical application of any drug for the last three months, f) patients who understood the instructions of applying drug and could apply the drug as per instructions.
Exclusion criteria include a) Age less than 18 years and more than 60 years, b) patients who received any treatment of arsenicosis within the last three months, c) patients with diagnosed skin diseases, like-atopic dermatitis and psoriasis (as per history obtained from patients), d) any diagnosed systemic disease, inflamamatory disease or infectious condition that affects the skin, like-diabetes mellitus, rheumatoid arthritis, systemic lupus erythematosus and hepatitis (as per history obtained from the patients).

Allocation of the patients randomly into 4 groups
Thirty-eight patients were randomized into 4 groups: Group 1: received tetracycline; Group 2: received clotrimazole; Group 3: received the combination of tetracycline and clotrimazole; Group 4: received placebo (no active ingredient) Randomization and sequence generation were done using an online graph pad calculator. Four sets of random numbers were allocated as intervention (tetracycline, clotrimazole, the combination of tetracycline and clotrimazole) and placebo sequentially according to their ID number.

Collection of samples
The samples include drinking water, nail, skin swab and skin scraping.

Collection of water samples
Patients were provided a 100 mL plastic container marked with an ID number containing 2-3 drops of nitric acid and requested to bring water from the tube well they use at present or used previously for drinking purposes. Then the samples were transported to the laboratory and stored at the refrigerator until analysis.

Collection of nail samples
Patients were requested to grow their nails. At first, they were advised to wash and dry their hands and nails before cutting the nails. They were provided a dry plastic bag marked with an ID number to collect nails from all their fingers. The nail samples were collected repeatedly until the amount became at least 200 mg. The collected samples were transported to the laboratory and stored at refrigerator until analysis.

Collection of samples from the skin
For the collection of skin swab and scraping samples, the screwed cap test tube containing nutrient broth media and petri dishes containing chromogenic agar media and potato dextrose agar media were taken to the field, keeping those in an icebox to stop the activity of the media during transportation.
Both hands were observed for keratotic lesions. The hand with more keratotic lesions was chosen for sample collection.

Collection of skin scraping sample
After washing the palm with 70% alcohol, skin scrapings were collected from the keratotic lesions with the help of a sterile surgical blade in a glass slide or container. By heating a wire loop in the flame of a candle until it became red, the wire loop was touched on the surface of the potato dextrose agar media containing petri dish to cool the wire loop to avoid killing of the microorganisms within the collected scraping samples and also to moist the wire loop with the media for easy collection of the dry scraping samples into petri dishes. After the collection of samples, petri dishes were sealed with micropore. The samples were kept for 7 days at room temperature for allowing the growth of the fungi.

Collection of skin swab sample
The skin of all the patients washed with 70% alcohol again. Then skin swabs were collected from each of the patients by rubbing the palm eight times by a sterile swab stick, moistened with a solution of nutrient broth. Then chromogenic agar media was inoculated by that swab stick. After collection of samples, petri dishes were sealed with micropore and transported to the microbiology laboratory and kept in the incubator for 18-24 hours (Jyoti et al., 2011;Khalil et al., 2016).

Preparation of culture media
Potato dextrose agar media, chromogenic media, Mac-Conkey agar media, motility indole urease media were prepared, autoclaved and then cooled until the media solidified in the petri dish.

Examination of fungal growth in culture plates
The growth of the fungi appeared on day 3 in potato dextrose agar media containing petri dishes. No further growth was observed after day 7.
Fungal growth was taken from petri dishes into the slides for microscopic examination.

Examination of bacterial growth in culture plates
The growth of the bacteria appeared in chromogenic agar media containing petri dishes after 18-24 hours incubation at 36°C. The microorganisms were identified by observing the color of the colony (Samra et al., 1998).
As Enterobacter spp. and Klebsiella spp. both produce the same colored colony in chromogenic agar plates, these two were differentiated by doing motility indole urease test.

Agar dilution method
Potato dextrose agar media was used for antifungal susceptibility test. MacConkey agar media was used for the antibacterial susceptibility test. The stock solution of antimicrobial agents, working solution of antimicrobial agents, antimicrobial drug supplemented media, and inoculum were prepared. After inoculating the organisms in the antifungal and antibacterial agent containing media, incubation was done. Then the growth of microorganisms (Provine and Hadley, 2000;Kuzucu et al., 2004;Razia et al., 2014) was observed. Growth in drug-containing media was compared with drug-free control media as follows: a) Growth same as control (score: 4); b) Slight decrease in growth (growth by approximately 75% of that control) (score: 3); c) Significant reduction in growth (growth approximately 50% of that control) (score: 2); d) Slight growth (growth approximately 25% of that control) (score: 1); and No growth (score: 0).

Effect of pH on the growth of microorganisms in presence of antimicrobials
The steps to see the effect of pH on the growth of organisms: a) Preparation of drug supplemented pHadjusted (pH 4-8) media; b) inoculation of organisms; and c) observation of growth. The steps were repeated 5 times.

Measurement of arsenical keratotic nodular size
A slide caliper was used to take the measurements. Up to five keratotic lesions were measured in both hands. Both the length and breadth of each keratotic nodule were measured. After summing up all the nodular size in both hands, an average keratotic nodular size was calculated by dividing that with the number of lesions measured. This procedure was followed for each patient before initiating the study and after completion of 3 months of treatment.

Preparation of ointment
To make 1,000 g ointment, ingredients were used as follows: a) At first, stearyl alcohol and bee wax were melted together on a water bath; b) Then, white petrolatum was added and mixed well; c) 8 mL of 5% sodium hydroxide was added to make the pH of the ointment 8 (tested by litmus paper); d) Then the mixture was removed from the bath and stirred until the mixture became congealed.

Distribution of medicines
Tetracycline, clotrimazole, the combination of tetracycline and clotrimazole and placebo (ointment without any active ingredient) were given to the patients with moderate to severe palmar arsenical keratosis randomly. The patients were instructed to apply the ointment to the affected area of the palm by fingertip twice daily (in the morning and bedtime). The patients were asked to clean their hands before applying ointment. A printed sheet was given for checking the adherence by giving a tick mark at the appropriate place just after applying ointment. They were requested to bring the sheet on the next follow-up.

Periodic monitoring of the patients
The patients were visited by the researcher at the Eruain Community Clinic at two weeks interval. The keratotic nodular size was measured. Patients' perception of clinical improvement was asked. They were also asked to report adverse effects related to the use of ointment. Regular communication with the patient was maintained over the cell phone.

Estimation of total arsenic
The total arsenic level in the water and nail was estimated by the SDDC method using a spectrophotometer (UV-VIS spectrophotometer-1201, Shimadzu, Japan) at 525 nm. The details of the method had already been described earlier (Bhuiyan et al., 2015).

Statistical analysis
Microsoft Office Excel was used to present the data as mean ± SD. To compare the arsenical skin lesions size before and after treatment in each group, Wilcoxon matched-pairs signed-rank test was performed manually. The comparison between each of the treatment groups and placebo was done manually by Mann-Whitney U-test.

Results
Most of the patients enrolled were female with a malefemale ratio 1:4 ( Table I). The mean (± SD) age of the patients was 40.8 ± 12.9 years. The patients were exposed to arsenic through arsenic-contaminated drinking Duration of exposure to arsenic through contaminated drinking water (years) 18.1 ± 9.7 Duration of development of keratosis (years) 13.4 ± 8.7 Concentration of arsenic in tube-well water (µg/L) 839.8 ± 179.9 Concentration of arsenic in nail (µg/g) 13.2 ± 12.8 Data are presented as mean ± SD Growth in chromogenic agar media for bacteria; Growth in potato dextrose agar media for fungi water for 18.1 ± 9.7 years and it took 13.4 ± 8.7 years to develop keratosis. The mean concentration of arsenic was 839.0 ± 179.9 µg/L in drinking water and 13.2 ± 12.8 µg/g in nail samples. Table II shows that among the 15 skin scraping samples, 6 samples revealed the growth of Aspergilus spp. Gram-positive organism, Staphylococcus spp. was isolated from 9 out of 10 skin swab samples. Five types of Gram-negative organisms were found in the skin swab samples of which Enterobacter spp. was isolated from 3 samples. Figure 2 shows each of the 11 antifungal drugs inhibited the growth of Aspergillus spp. The growth of inhibition was found to be increased with an increase in the concentration of antifungals. At 50 µg/mL, the growth inhibition was found to be highest followed by terbinafine. Seven antibiotics out of eight inhibited the growth of Enterobacter spp. and the inhibition was found to be increased gradually at higher concentrations. The growth of inhibition at 100 µg/mL was highest with tetracycline. Ciprofloxacin did not inhibit the growth even at the highest concentration.
The pH of the palmar surface of the hand (mean ± SD) was found to be close in the arsenic-exposed participants (8.0 ± 0.2) and arsenicosis (8.0 ± 0.2), whereas the pH was lower in the group of healthy volunteers (7.5 ± 0.3). This difference was found statistically significant.
There was pH-dependent inhibition of the growth of Aspergillus spp. in presence of clotrimazole (30 µg/mL) in potato dextrose agar media. Figure 3 shows the maximum inhibition of the growth of Aspergillus spp. at pH 8, followed by pH 4 and 5. In the case of Enterobacter spp., similar pH-dependent inhibition of the growth was observed in the presence of tetracycline (50 µg/ mL) in MacConkey agar media. The maximum inhibition of the growth was found at pH 4 and 8.
The size of the keratotic nodule before and after the intervention was 21.7 ± 10.1 mm 2 and 15.5 ± 11.2 mm 2 respectively in the placebo group (Figure 4) and the improvement was found statistically insignificant with p value of 0.11. In the tetracycline group, the size of the keratotic nodule before and after the intervention was 18.3 ± 10.7 mm 2 and 8.1 ± 7.9 mm 2 respectively. The improvement was found statistically significant (p value 0.01) with the percentage of keratosis size reduction 66.1. The mean ± SD of the keratotic nodule size was 17.4 ± 11.6 mm 2 and 5.5 ± 4.4 mm 2 before and after treatment with clotrimazole ointment for 3 months. The percentage of keratosis size reduction was 61.1 and the clinical improvement was found statistically significant with p value 0.02. The keratotic nodule size in the combination group was 20.3 ± 8.6 mm 2 and 12.8 ± 10.2 mm 2 before and after intervention respectively, which was found not to be significant statistically (p value 0.07). Each of the three groups of intervention (tetracycline, clotrimazole and combination) was compared with the placebo group to see the superiority among the interventions and not found statistically significant.
The adherence of the patients toward intervention was assessed and found 94.3 ± 6.0% (tetracycline), 91.3 ± 9.1% (clotrimazole) and 97.5 ± 4.0% (a combination of tetracycline and clotrimazole), whereas in the placebotreated patients, the adherence was 83 ± 14.8%. Only three patients complained about adverse effects which were not significant and subsided spontaneously.

Discussion
This is the first report showing the effect of topical antimicrobials in the treatment of palmar arsenical keratosis. The patients of palmar arsenical keratosis were treated with the topical application of antibacterial and antifungal agents for 3 months. Either tetracycline or clotrimazole was found to be effective in the reduction of the size of the keratosis. However, the combination of tetracycline and clotrimazole did not show any improvement in comparison to the placebo. The adherence of the patients toward intervention and placebo was satisfactory with no remarkable adverse effect.
The normal skin flora can be altered by a variety of factors which can be either endogenous or exogenous environmental influences (Roth and James, 1988). Limited studies had been conducted to see the alterations of normal skin flora in arsenicosis. One of the diagnostic features of arsenicosis is skin lesions. These skin lesions, arsenic itself and its metabolites may play an important role in determining the composition of the skin flora in arsenicosis.
In a study, skin swab samples collected from 3 anatomical sites (arm, finger web and upper chest) were analyzed by culture techniques and later identification of microorganisms was done by biochemical test or PCR method. The results revealed an increased Gramnegative bacterial load in arsenicosis in comparison to the control population (Jyoti et al., 2011). Enterobacter spp. was found to be present in the palmar surface of the skin in patients with arsenical keratosis significantly when compared with arsenic-exposed individuals and healthy volunteers. A significant number of the samples revealed the presence of Enterobacter spp. in patients with palmar arsenical keratosis (Khalil et al., 2016). On the other hand, Aspergillus spp. was found to be predominant fungus in the palmar surface in comparison to arsenic-exposed and healthy individuals (Moitra et al., 2018). The present study also explored and found the presence of Aspergillus spp. and Enterobacter spp. as a skin flora in patients with moderate to severe palmar arsenical keratosis. The variation in the percentage of the isolated organism in these studies may be due to variation in the number of samples obtained, techniques by which the samples were collected and transported, techniques used for isolation and identification of the organisms (culture-dependent or molecular techniques), the media used in case of culture-dependent studies.
Whether this altered flora of the palmar surface of the skin in arsenicosis is responsible for the development of keratosis, the present study was conducted to see the effect of antimicrobials targeting that altered flora in clinical improvement. To choose the antimicrobials against Aspergillus spp. and Enterobacter spp., the antimicrobial susceptibility test was performed.
The antimicrobial susceptibility against Enterobacter differs widely as there are diverse species within the genus. E. sakazakii and E. agglomerans were found susceptible to ampicillin, cephalothin, and cefoxitin, to which E. cloacae and E. aerogenes were resistant (Muytjens and Repe, 1986). 100% sensitivity of isolated Enterobacter spp. to imipenem and 70% to amikacin was reported by the disk diffusion method (Sharmin et al., 2009). Another study found reduced sensitivity to a second-generation cephalosporin (Rangaiahagari et al., 2013). Expanded-spectrum β-lactamase producing E. cloacae revealed statistically significant higher resistance to cefotaxime, ceftazidime, aztreonam, piperacillin, tetracycline, and cotrimoxazole than non-producers (Wang et al., 2017).
The present study assessed the activity of 11 antifungal and 8 antibacterial drugs against Aspergillus spp. and Enterobacter spp. isolated from the palmar surface of the skin of the patients with arsenical keratosis by agar dilution method and found clotrimazole and tetracycline to be the most effective ones. The susceptibility pattern of antimicrobials in the present study differs from other studies due to the variation in the species of Aspergillus and Enterobacter investigated, test methods, media used, size of the inoculum, range of concentrations of drugs and assessment techniques.
Stoppage of drinking arsenic-contaminated water is the first step toward the management of arsenicosis. In addition, emphasis should be given on the provision of a diet rich in protein and vitamins. Intake of plenty of green leafy vegetables is also advisable. Previously, both enteral and topical approaches had been used to treat palmar arsenical keratosis.
Keratosis present in the palm and sole can be treated by topical application as well. Salicylic acid  with or without urea, propylene glycol (Dina andMisbahuddin, 2010), neem (Ferdous and, brinjal peel (Sarah, 2018), cock's comb (Anny and Misbahuddin, 2019) were some of the agents studied to treat keratosis. No remarkable adverse effects were reported due to these topical treatments, except itching and burning sensation. Most of these local treatment options either have keratolytic Dina and Misbahuddin, 2010;Sarah, 2018) or water-retaining properties that regulate the normal desquamation process of the skin (Anny and Misbahuddin, 2019). As this is the first study where antimicrobials were used as a treatment option for keratosis, some factors, likeconcentration of the drug in the ointment, pH of the ointment and adherence toward treatment were assessed.
As the present study showed that there was better inhibition of the growth of Enterobacter spp. and Aspergillus spp. by 100 µg/mL tetracycline and 50 µg/mL clotrimazole respectively, to prepare 1,000 g ointment, 100 mg/mL tetracycline and 50 mg/mL clotrimazole had been used.
In the present study, the interventions were given for 3 months. Adherence in each of the intervention and placebo groups was assessed and found satisfactory. The earlier studies using the local approach of treatment reported the clinical improvement either by measuring percentage reduction of keratotic nodular size or by assessing perception scores. In the present study, both the breadth and length of 1-5 keratotic nodules of both palms were measured. Very few adverse effects were reported in the present study.
The percentage reduction of keratotic nodular size after 3 months of treatment was 66.1 and 61.1 in tetracycline and clotrimazole groups of patients, which was found statistically significant. The small number of patients (5-8) in each of the 4 groups and the duration of the treatment should be considered for better clinical improvement. The interventions (tetracycline, clotrimazole and combination of tetracycline and clotrimazole) were not found statistically superior in comparison to placebo.

Conclusion
In vitro study shows that tetracycline or clotrimazole was found to be effective against Enterobacter spp. and Aspergilus spp. respectively in pH-dependent manner. In patients of arsenicosis, the clinical improvements after 3 months of treatment with tetracycline or clotrimazole ointments were found statistically significant. No intervention (tetracycline, clotrimazole or combination of tetracycline and clotrimazole) was found statistically superior in comparison to placebo in reducing palmar arsenical keratotic nodular size.