Acanthamoeba species are ubiquitous protozoa that are widely distributed in the environment. These organisms have an active trophozoite stage and a dormant cyst stage that is resistant to adverse environmental conditions.1 Acanthamoeba can cause a vision-threatening keratitis that is often associated with contact lens wear. An outbreak of Acanthamoeba keratitis (AK) in 2007 resulted in the voluntary recall of a multipurpose contact lens solution (MPS) and prompted the US Food and Drug Administration (FDA) to reassess contact lens safety and the efficacy of MPS against Acanthamoeba.2 Despite the recall in 2007, the number of AK cases continues to increase.3
As discussed by Brocious et al. (Brocious J, Hampton D, Tarver M, et al. Eye & Contact Lens 2018:44:In Press), although both the FDA recognized International Organization for Standardization (ISO) standard 14729 and the FDA guidance for contact lens care products provide guidelines for assessing disinfection efficacy of MPS,4,5 neither document requires testing for efficacy against Acanthamoeba and there is currently no universally accepted protocol for disinfection efficacy testing of MPS against the organism. Among the reasons cited in ISO 14729 for not testing for Acanthamoeba are the rare incidence of infection, preventable source of contamination, and lack of a standard method for propagating and testing trophozoites and cysts in MPS. A voluntary recall of an MPS suspected of being associated with an Acanthamoeba outbreak in 2007 led to recommendations from experts on the Ophthalmic Advisory Panel in 2008, and recommendations regarding parameters to consider at a microbiology workshop cosponsored by FDA, American Academy of Ophthalmology, American Academy of Optometry, American Optometric Association, and Contact Lens Association of Ophthalmologists.2,6,7 With the increase in the number of reported AK infections and using the recommendations from the workshop, FDA published a proposed testing method for Acanthamoeba to evaluate disinfection efficacy.8 The proposed protocol was published as a starting point to test disinfection efficacy against Acanthamoeba, and it was recognized at the time that further refinement of the protocol may be needed. The experts of the advisory panel in the 2009 Microbiological Testing for Contact Lens Care Products workshop noted that the variation to the process of handling and growing the organisms has been an obstacle to yielding a single-test procedure.7 Recommendations from a subsequent Ophthalmic Advisory Panel meeting and cosponsored microbiology workshop in 2014 reiterated the need for testing along with recommended protocol parameters, regarding handling and growing organisms.9,10 Therefore, FDA sought to further optimize the protocol previously published in 2012.8 We incorporated suggestions in optimization of the protocol, incorporating testing of both trophozoites and cysts of Acanthamoeba castellanii (ATCC 50370) and Acanthamoeba polyphaga (ATCC 30461); axenic growth for trophozoites; the use of a starting inoculum of 103 to 104 Acanthamoeba; and the starvation method to generate the cyst form. In an effort to evaluate our method, we incorporated a positive control and a neutralization control into the protocol. The current criteria in ISO 14729 allow for products to be used as a standalone disinfecting product or as part of a cleaning/disinfection regimen to provide basic efficacy against bacteria and fungi. However, because Acanthamoeba have been reported in outbreaks of keratitis associated with contact lens use, some manufacturers would like to assess the anti-Acanthamoeba activity of their MPS. Our intent was to develop a protocol as a first step to determine efficacy of solutions. The goal of this study was to evaluate the interlaboratory and intralaboratory reproducibility of this revised protocol for MPS disinfection efficacy against Acanthamoeba.
American Type Culture Collection (ATCC) human eye isolates used in the experiments were A. castellanii ATCC 50370 and A. polyphaga ATCC 30461 of the T4 genotype (Manassas, VA).
Trophozoites were maintained in ATCC growth medium 712 PYG with additives (formula available online from ATCC) at 25°C. Trophozoites were produced for experiments by subculturing amoeba and harvesting the amoeba after 2 days. Trophozoites were washed 2 times with 10 mL of ATCC Medium 1323 Page's amoeba saline (PAS) (formula available online from ATCC), counted in a hemocytometer, and adjusted to yield 2.0 × 105 to 2.0 × 106 trophozoites/mL in PAS. Cysts were produced by allowing trophozoites to encyst on non-nutrient amoeba saline agar (NNAS) plates (15-gm bacteriological agar in 1,000 mL PAS) for 2 weeks at 30°C. Cysts were harvested by adding 5 mL of PAS to the plates and removing the cysts by scrapping the plate with a sterile swab. The cysts were washed two times with 10 mL of PAS using centrifugation, counted using a hemocytometer, and adjusted to yield 2.0 × 105 to 2.0 × 106 cysts/mL in PAS.
Suspensions of Enterobacter aerogenes ATCC 15038 or 13048 in PAS were used to detect amoeba that survived exposure to the MPS and control solutions. Bacteria were grown overnight in trypticase soy broth at 30 to 35°C with shaking and were harvested by centrifuging for 10 min at 4,000g, and washing 2 times with PAS. The bacteria were suspended in PAS and a densitometer was used to achieve a suspension of bacteria with an OD600 of 0.7 to 0.9. This is the suspension used to inoculate NNAS agar plates and 6- and 12-well tissue culture plates. Non-nutrient amoeba saline agar plates were inoculated with three drops (or approximately 150 μL) of suspension using a sterile plastic Pasteur pipette and spread using a sterile plate spreader. Six-well plates received 1 mL of suspension/well, and 12-well plates received 0.6 mL/well.
Control and Test Solutions
Positive and negative controls were included in each experiment. The positive control consisted of 12.5 μg/mL chlorhexidine (Sigma-Aldrich) in PAS and the negative control was PAS. The concentration of 12.5 μg/mL chlorhexidine used for the positive control was selected from results of preliminary experiments because it killed both cysts and trophozoites, was neutralized by Dey Engley (D/E) broth, and did not induce total kill at time 0 (data not shown). The four different MPS (Solutions A, B, C, and D) that were tested are presented in Table 1. We selected four different MPS each with a unique biocidal agent that are commercially available and contain commonly used biocidal agents. MPSs were used before their expiration dates.
Aliquots (100 μL) of trophozoites or cysts were added to 9.9 mL of the test solution or control solutions in sterile polypropylene tubes to produce initial concentrations of 2.0 × 103 to 2.0 × 104 trophozoites or cysts per milliliter, vortex mixed, and kept at room temperature in the dark. For the MPS-containing hydrogen peroxide, the included neutralization disk was placed into the MPS just before amoeba were added. The case designed for the hydrogen peroxide product was not large enough to hold the volume of solution required for the assay. Therefore, a 50-mL conical polypropylene tube was used. At 0 hr, the manufacturer's recommended soak time (MRST) (Table 1), 8- and 24-hr postinoculation, 1-mL aliquots were removed from each test and control solution and placed in D/E broth for a minimum of 10 min for neutralization (1 in 10 dilution of trophozoites or cysts). D/E broth containing 200 to 500 U catalase was used to neutralize the MPS containing hydrogen peroxide. After neutralization, 1 mL of the inoculated D/E broth was inoculated onto five NNAS plates containing bacterial suspensions (as described above) for each solution and control solution. From the 1 in 10 dilution in D/E broth, a 1 in 100 dilution was prepared in PAS, and from this dilution, 1-mL aliquots were inoculated into 5 wells of 6-well plates containing bacterial suspensions (as described above). Five wells (10 total wells) of 12-well plates containing bacterial suspensions (as described above) were used for the 0.1- and 0.01-mL aliquots, respectively. Plates were then placed into zip-lock plastic bags or sealed with parafilm and incubated at 30°C for 14 days and then checked for growth. For enumeration, a most probable number method was used as described by Beattie et al.11
A control was established to verify that the D/E broths effectively neutralized the test solutions and positive control. 1 mL of test solution was placed into 9 mL of the appropriate neutralizing broth in a polypropylene tube, vortex mixed, and incubated for 10 min. Then, 100 μL of 1 × 103 to 1 × 104 trophozoites or cysts in PAS was added (final concentration of 10–100 amoeba/mL) followed by vortex mixing. For each solution and control, 1 mL was immediately inoculated onto 5 agar 100 × 15-mm Petri plates seeded with bacterial suspensions (as described above). The remainder of the solutions inoculated with amoeba were incubated for the MRST in the dark at room temperature, and another set of five agar Petri plates seeded with bacterial suspensions (as described above) were inoculated, sealed, and incubated as detailed above in efficacy testing. Any plates yielding no growth indicated that the neutralizing broth did not effectively neutralize the test solution.
Testing was performed in parallel in two different FDA laboratories and all experiments were performed by the same individuals at each site. Single lot numbers of each test solution and a common positive control solution were used for all experiments. To establish intralaboratory reproducibility, an experiment was performed in each laboratory in which A. castellanii trophozoites were tested against all test solutions and controls in triplicate. Within each laboratory, the same starting concentration of trophozoites and bacterial suspensions were used for this experiment. All other experiments were performed twice with different initial concentrations of fresh trophozoites or cysts and bacterial suspensions in both laboratories (duplicate testing).
An assay to determine any effect of the PAS (negative control) and a positive control (12.5 μg/mL chlorhexidine) was performed for each experiment. These controls displayed consistent results and trends both within each laboratory and between each laboratory for trophozoites and cysts of both A. castellanii and A. polyphaga. The neutralization control consistently demonstrated the ability of the neutralizing agents (D/E broth with and without catalase) to neutralize the MPS and the control negative demonstrated no inhibition of Acanthamoeba. Results not shown.
Efficacy testing of A. castellanii trophozoites in triplicate demonstrated reproducibility of the protocol both within each laboratory and between the laboratories. The positive control demonstrated total kill at 4 hr of incubation and longer (6, 8, and 24 hr). Total kill was also observed for solutions A, B, and D at the MRST and longer. Solution C did not show total kill at any incubation point. The effectiveness (presented as log kill) for 6 hr (MRST for this solution), 8 hr (likely overnight soak time), and 24 hr is shown in Table 2. Because the protocol allows a range of initial starting concentrations of amoeba of one log (2.0 × 105–2.0 × 106), a difference of greater than one log was considered to be significant. The results in Table 2 demonstrate that, for a solution that does not provide a greater than three-log kill of the amoeba; no difference was observed in the log kill at the MRST and 8 hr exposure to the solution. Results for both laboratories followed similar trends.
Duplicate testing performed with A. castellanii and A. polyphaga trophozoites and cysts demonstrated that the solutions are generally more effective against trophozoites than they are against cysts. For trophozoites of both species, the positive control demonstrated total kill at 4 hr of incubation and beyond (i.e., 6, 8, and 24 hr). Total kill was also observed for solutions A, B, and D at the MRST and longer. Only solution D was able to yield total kill of cysts of both species at the MRST and longer. Solutions A, B, and C did not yield a total kill of cysts at the MRST or at 8 hr (Table 3). Both laboratories generated results with similar trends for controls and solutions with trophozoites. However, in agreement with other investigators, results from testing cysts were more variable.11
Among the solutions we tested, trophozoites and cysts of A. castellanii and A. polyphaga showed similar responses. The only exception was solution C. In both laboratories, this solution had some effect (1.05–1.39 log kill) on A. polyphaga cysts at both 6 and 8 hr. This solution had no effect (<1 log kill) on A. castellanii cysts and A. castellanii trophozoites, at 6 (MRST) and 8 hr and on A. polyphaga trophozoites at 6 hr (Table 4).
The current criteria in ISO 14729 allow for products to be used as a standalone disinfecting product or as part of a cleaning/disinfection regimen to provide basic efficacy and is not intended for solutions to be labeled as effective against any specific bacterium or fungus. However, because Acanthamoeba has been reported in outbreaks of keratitis associated with contact lens use, it would be useful to determine whether MPS is effective against Acanthamoeba. Our intent was to develop a protocol as a first step to determine the efficacy of MPS against Acanthamoeba.
The results of this study demonstrate the validity of our optimized protocol for testing MPS efficacy against Acanthamoeba. We demonstrated that the revised protocol is reproducible within the same laboratory and between two different laboratories. Our revised protocol is presented in detail in the Supplemental Digital Content, http://links.lww.com/ICL/A72. We added a positive control (12.5 μg/mL chlorhexidine) and a neutralization control to the protocol. In addition to the negative control, these controls help ensure the integrity of the results generated with the protocol. Our revised protocol recommends testing the efficacy of MPS against trophozoites and cysts of both A. castellanii ATCC 50370 and A. polyphaga ATCC 30461. Following the recommendations of the expert panel from the 2014 workshop, the protocol also uses axenic growth for trophozoites and the starvation method to generate the cyst form for testing.9,10
We evaluated this protocol by performing triplicate experiments with A. castellanii trophozoites in two different laboratories which demonstrated the interlaboratory and intralaboratory reproducibility of the protocol, especially with the MRST for solutions that did not provide total kill of the amoeba. All other experiments (A. castellanii cysts and A. polyphaga trophozoites and cysts) were performed in duplicate in both laboratories and again demonstrated the same reproducibility of the protocol. Beattie et al. 11found no significant difference between plastic and glass testing bottles in their studies of solutions containing myristamidopropyl dimethylamine and polyhexamethylene biguanide (PHMB). Our study was designed to evaluate the performance of our protocol. Therefore, because it was not our goal in this study to compare the efficacy of individual solutions, we used polypropylene tubes for all solutions. However, appropriate tube materials should be assessed (e.g., as outlined in ISO 14729) when determining efficacy of an MPS.4
As with other published methods for testing efficacy of MPS against Acanthamoeba, trophozoites of both species were more sensitive to MPS than were cysts.11–13 Also in agreement with other published protocols, only the MPS with hydrogen peroxide as the active ingredient (solution D) was consistently and highly effective against cysts.13–15 Results from testing two different species generated similar results except with MPS C that contained PHMB (0.0001%), which seemed to be more effective against A. polyphaga cysts (Table 4). This was an unexpected finding that was reproducibly demonstrated in both laboratories. This result emphasizes the utility of testing MPS against at least two different species or strains of Acanthamoeba.
Our data indicate that our protocol is a well-controlled and reproducible procedure that can effectively evaluate the efficacy of MPS against Acanthamoeba trophozoites. However, testing cysts can be problematic because most solutions, other than those containing hydrogen peroxide as the biocide, have marginal or no activity against cysts and thus results can be variable as previously demonstrated by Beattie et al.11 We recommend that testing be performed in triplicate to compensate for any variability.
To assess agreement between laboratories, we took the difference between the log kill results for trophozoites and cysts and calculated the mean and SD of the differences. Overall, across both species of trophozoites, the mean/SD of the differences was 0.14/0.21 log units for the MRST exposure, 0.24/0.45 log units for the 8 hr exposure, and 0.18/0.35 for the 24 hr exposure. Across both species of cysts, the mean/SD of the differences was 0.60/0.49 log units for the MRST exposure, 0.60/0.59 log units for the 8 hr exposure, and 0.56/0.47 for the 24 hr exposure. For trophozoites, mean differences between laboratories were small with acceptable variability. Differences between laboratories for the cysts were somewhat more variable. This correlates with the activity of solutions against trophozoites and cysts.
In conclusion, we believe that MPS should be tested for efficacy against Acanthamoeba in addition to the current recommendation for testing against bacteria and fungi under the international standard ISO 14729 and the FDA guidance document, Guidance for Industry Premarket Notification (510[k]) Guidance Documents for Contact Lens Care Products.4,5 In fact, experts at a 2008 meeting of the Ophthalmic Advisory Panel and at the 2009 and 2014 FDA workshops for microbiological testing of MPS recommended premarket testing of MPS against Acanthamoeba. 6,7,9 To the best of our knowledge, there are few published testing protocols that are reproducible for both trophozoites and cysts. We believe that the protocol we have revised and evaluated is a well-controlled and reproducible procedure that can effectively evaluate the efficacy of MPS against Acanthamoeba trophozoites and cysts when testing is performed in triplicate. We hope that this protocol will be used by the research community to jointly establish pass/fail criteria. Such improvement in regulatory science would expedite development of MPS effective against Acanthamoeba.
1. Kahn NA. Acanthamoeba
Biology and Pathogenesis. Great Britain, United Kingdom, Caister Academic Press, 2009.
2. Hampton D, Tarver ME, Eydelman MB. Latest Food and Drug Administration's efforts to improve safe contact lens use. Eye Contact Lens 2015;41:1–2.
3. Tu E. Acanthamoeba
keratitis: A new normal. Am J Ophthalmol 2014;158:417–419.
4. International Organization for Standardization. Ophthalmic Optics—Contact Lens Care Products—Microbiological Requirements and Test Methods for Products and Care Regimens for Hygienic Management of Contact Lenses. ISO 14729. Geneva, Switzerland, International Organization for Standardization, 2001.
8. Shoff M, Eydelman MB. Strategies to optimize conditions for testing multipurpose contact lens solution
efficacy against Acanthamoeba
. Eye Contact Lens 2012;38:363–367.
10. Hampton D, Tarver ME, Jacobs DS, et al. Special commentary: Food and Drug Administration, American Academy of Ophthalmology, American Academy of Optometry, American optimetric association and the contact lens association of ophthalmologists cosponsored workshop: Revamping microbiological test methods for contact lenses, products, and accessories to protect health and ensure safety. Eye Contact Lens 2015;41:329–333.
11. Beattie TK, Seal DV, Tomlinson A, et al. Determination of amoebicidal activities of multipurpose contact lens solutions by using a most probable number enumeration technique. J Clin Microbiol 2003;41:2992–3000.
12. Shoff M, Rogerson A, Schatz S, et al. Variable responses of Acanthamoeba
strains to three multipurpose lens cleaning solutions. Optom Vis Sci 2007;84:202–207.
13. Kilvington S, Lam A. Development of standardized methods for assessing biocidal efficacy of contact lens care solutions against Acanthamoeba
trophozoites and cysts. Invest Ophthalmol Vis Sci 2013;54:4527–4537.
14. Johnston SP, Sriram R, Qvarnstrom Y, et al. Resistance of Acanthamoeba
cysts to disinfection in multiple contact lens solutions. J Clin Microbiol 2009;47:2040–2045.
15. Mowrey-McKee M, George M. Contact lens solution
efficacy against Acanthamoeba
Castellanii. Eye Contact Lens 2007;33:211–215.