A Multicenter, Real-World Experience with Polyhexamethylene Biguanide (PHMB) 0.08% Monotherapy for Treating Acanthamoeba Keratitis: A Retrospective Review of 39 Cases Authors: Karl Anders Knutsson, Antonio Di Zazzo, Emilio Pedrotti, Paolo Rama, Vito Romano, Anna Maria Roszkowska, Giacinto Triolo, Francesco Aiello, Stefano De Angelis, Rita Mencucci, Pasquale Aragona, Federica Birattari, Erika Bonacci, Matilde Buzzi, Marco Coassin, Adriano Fasolo, Mariangela Gari, Pia Leon, Carlo Nucci, Giorgio Paganoni, Matteo Pederzolli, Fabiana Secchi, Sabrina Vaccaro, Vincenzo Papa, Antonella Franch Published: 2026 ABSTRACT AbstractIntroductionTopical polyhexamethylene biguanide (PHMB) 0.08% (AKANTIOR®, SIFI Italy) is the only licensed medicinal product for the treatment of Acanthamoeba keratitis (AK), demonstrating an 84.9% cure rate in a pivotal clinical trial. This study aimed to evaluate its effectiveness and safety in a routine clinical practice.MethodsWe retrospectively reviewed patient charts from ten Italian centers for the treatment of AK with PHMB 0.08% monotherapy. The treatment protocol followed that used in the pivotal trial. The primary outcome was medical cure, defined as a healed epithelium without inflammation and treatment for at least four weeks. Treatment failure was defined as the need for surgery or the use of alternative or additional anti-amoebic treatments (AAT). Secondary outcomes included best-corrected visual acuity (BCVA), time to cure, pain, and corneal scarring.ResultsThirty-nine patients (M/F = 14/25) were evaluated. AK was confirmed by confocal microscopy ± polymerase chain reaction or culture in all cases. Sixteen (41%) had stage 3 AK. Thirty-five (89.7%; 95% confidence intervals 80.2–99.3) achieved medical cure. Four failures occurred: three required corneal surgery and one additional AAT. At the end of treatment, mean (SD) LogMAR BCVA was 0.7 (0.8); the median time to cure was 95 days (Q1–Q3: 63–140 days). Two patients (5.6%) reported residual pain, and 27 (73.0%) had corneal scarring. Four serious adverse events (three hypopyon, one corneal perforation) were reported and considered unlikely to be treatment-related.ConclusionsAKANTIOR cured approximately 90% of patients with AK, indicating that the efficacy observed in the clinical trial can be replicated in clinical practice.Trial RegistrationClinicalTrials.gov identifier, NCT06641882.Supplementary InformationThe online version contains supplementary material available at 10.1007/s40123-026-01391-4.Keywords: Acanthamoeba keratitis, Biguanides, Corneal transplantation, Visual acuity Key Summary Points Why carry out this study?Acanthamoeba keratitis (AK) is a serious ocular infection.Polyhexamethylene biguanide (PHMB) 0.08% is the only treatment currently licensed in the EU for the treatment of AK. However, no robust clinical data outside a clinical trial setting are available to date.This study aimed to evaluate the efficacy and safety of PHMB 0.08% in clinical practice, evaluating a large cohort of patients with AK treated during a compassionate-use program in Italy.What was learned from the study?This retrospective study evaluated 39 AK cases and showed a medical cure rate with no surgery of 89.7%; this outcome was comparable to that observed (84.8%) in the controlled setting of a clinical trial.The results provide evidence supporting the efficacy of PHMB 0.08% as first-line therapy for patients with AK. Introduction Acanthamoeba keratitis (AK) is a potentially sight-threatening corneal infection. It accounts for only 5% of infectious keratitis, but it is responsible for approximately 50% of the cases of vision loss [1]. The global annual incidence of AK is estimated to be 2.3 cases per million individuals [2]; it is approximately tenfold higher amongst contact lens wearers [3–6]. AK is caused by a free-living protozoan commonly found in water, soil, dust, and air, which exists in a motile active form (trophozoite) and a dormant cyst form [7, 8]. Both trophozoites and cysts are found in infected tissue, but only trophozoites are infectious. Cysts are the inactive form of Acanthamoeba but are more resistant to treatment, resulting in the persistence of infections. From a clinical perspective, AK may cause severe keratitis if not promptly identified and treated. Patients report nonspecific signs and symptoms at presentations, such as ocular redness, ocular discomfort, photophobia, and reduced visual acuity, that may be confused with other forms of infectious keratitis, such as herpes or fungal keratitis [9, 10]. This may lead to a late diagnosis [6] when worsening of clinical signs and, in many cases, ocular pain disproportionate to clinical findings are recognized as resulting from AK. Late diagnosis is associated with worse outcomes [4, 5, 11]. Several diagnostic methods are used to confirm the diagnosis of AK, including histological staining of corneal samples, corneal culture, polymerase chain reaction (PCR) and in vivo confocal microscopy (IVCM), the latter probably being the most sensitive although it is not available in all corneal units [12]. Medical therapy is considered the first-line therapy for AK, while surgical treatments are reserved for refractory cases [5, 6]. Topical biguanides (polyhexanide [PHMB], chlorhexidine) and diamidines (propamidine, hexamidine) are the most used and effective off-label medical treatments [1, 6, 10, 13, 14]. Biguanides act on both trophozoites and cysts and can also be used as monotherapy, while diamidines are less effective on cysts and, as a result, are usually used in combination with a biguanide [15]. A dual biguanide therapy (PHMB + chlorhexidine) may be an additional option for cases resistant to standard therapy [16]. A realistic medical cure rate (without surgical intervention) achievable with these unlicensed treatments can be derived from the two largest case series published, totaling more than 400 cases, reporting a medical cure in approximately 60% of patients [17, 18]. Keratoplasty may be required in 20–25% of patients in case of imminent corneal perforation or, later, for visual rehabilitation [19, 20]. PHMB 0.08% (AKANTIOR®, SIFI SpA, Italy) is the only drug licensed in Europe for the treatment of AK, while no product is yet licensed elsewhere, including the US. Evidence leading to its approval included the results of the ODAK trial showing a medical cure rate of 84.9% [21]. Outside this trial, successful treatment with PHMB 0.08% has been described in a few case reports [22–24], but no studies with substantial patient numbers have been conducted to date. The aim of this study was to investigate the effectiveness and safety of PHMB 0.08% in clinical practice by analyzing a cohort of patients treated during a compassionate use program (CUP) in Italy. Methods Study Design This was a retrospective single-cohort study to assess the effectiveness and safety of PHMB 0.08% in a real-world clinical setting. Data already recorded in hospital charts of patients with AK treated with PHMB 0.08% eye drops during a CUP were extracted and analyzed. The study was conducted in accordance with the STROBE guidelines for a cohort study. Clinicaltrial.gov identifier: NCT06641882. Clinical Setting The CUP was a named patient program granted by the Italian drug agency; each individual treatment was approved by the local Ethics Committee. The present study covers the period December 2022 to June 2024, during which patients with AK were treated in more than 30 different clinical centers. To ensure data consistency and reduce variability across sites, only the ten centers that had treated at least three patients were selected for the chart reviews. Ethics The study was approved by the CET 1 Comitato Etico Regione Lombardia 1 of the coordinating center (San Raffaele Scientific Institute, Milan, Italy) on April 24, 2024 (approval number CET 148/2024). The study was performed in accordance with the Helsinki Declaration of 1964 and later amendments. Written informed consent was obtained from all patients prior to the commencement of the study, including consent to the use of their clinical data for scientific purposes. Identifying characteristics within the data were replaced with a pseudonym, which does not allow the data subject to be directly identified. Inclusion Criteria Inclusion criteria were as follows: (1) Age > 14 years; (2) a clinical diagnosis of AK confirmed by IVCM, PCR or microbiological culture; (3) able to sign the informed consent; (4) untreated (i.e. treatment-naïve) or already treated with other anti-amoebic treatments (AAT) with a clinically non-progressive condition (stable disease), defined as absence of clinical deterioration (e.g., no progressive corneal melting, perforation risk, or need for immediate rescue therapy) but with persistent disease suitable for switching to PHMB 0.08% once available; (5) treatment approved by the local Ethics Committee between December 1, 2022 and June 30, 2024; (6) available clinical charts including the required demographic, clinical, and imaging data. Only one eye was included in the study; in bilateral cases, both eyes could be treated, but only the right eye was selected for analysis unless there was a difference in disease severity, in which case the more severely affected eye was analyzed. Exclusion Criteria Exclusion criteria were: (1) Age < 14 years; (2) a clinical diagnosis of AK not confirmed by the above-mentioned diagnostic tests; (3) patients scheduled for corneal surgery; (4) treatment approved by the local Ethics Committee after June 30, 2024; (5) incompletely documented clinical charts. Treatment Regimen The treatment regimen recommended was the one used in the pivotal clinical (ODAK) trial [21]. Any additional medication was prescribed at the discretion of each physician on an individual patient basis. Patient evaluation periods and data collection at each visit used routine procedures at each center. Data Collection Patient chart data were entered into an electronic case report form. Data were collected at the start of treatment with PHMB 0.08% (baseline) and at the end of treatment. As in the ODAK trial [21], a follow-up visit was included at least 1 month after stopping PHMB 0.08% and all other treatments, to confirm the outcome. Baseline data recorded were demography, potential risk factors, time from symptom start to initiation of treatment with PHMB 0.08%, diagnostic techniques used, prior treatments, Snellen best-corrected visual acuity (BCVA), slit-lamp findings, and disease staging [25]. The degree of pain was reported descriptively by patients as mild, moderate, or severe, without the use of a standardized scoring system. Conjunctival inflammation was graded by physicians as absent, mild, moderate, or severe, as previously described [26]. End-of-treatment data recorded were other medical treatments (including adjunctive or alternative AAT), surgery, adverse events (AE), BCVA, slit-lamp findings, and degree of pain. Adherence to the dosing protocol was evaluated as well. Outcomes Primary Outcome The primary outcome was treatment success, independent of the functional outcome and defined as the proportion of patients having medical cures without surgical intervention or any change of AAT. Cure was defined as clinical evidence of the elimination of Acanthamoeba (including an intact corneal epithelium and no clinical signs of ocular inflammation, such as absence of corneal inflammation, no or mild conjunctival inflammation, and no limbitis, scleritis, or anterior chamber inflammation) after discontinuing all treatments, confirmed after a minimum follow-up of 30 days. This threshold was based on findings from the ODAK trial [21], in which no recurrences were observed after 1 month off therapy. Treatment with PHMB 0.08% was considered failed in case of surgical interventions, discontinuation due to an adverse drug reaction, switching to (or the addition of) another topical treatment with established anti-amoebic effects (propamidine, hexamidine, miltefosine, chlorhexidine). Secondary Outcomes Secondary outcomes included BCVA, presence of pain, corneal scarring, use of adjunctive treatments, and time to cure. Safety Safety outcomes collected were AE. An AE was classified as serious in case of death, life-threatening, hospitalization, or persistent or significant disability. Data and Statistical Analysis Data analysis was performed using the R software (R core team, version 4.3.2 or higher). All analyses were descriptive in nature and therefore no hypothesis testing was conducted. Demographic and clinical characteristics were listed and summarized. Continuous data were summarized by mean, median, standard deviation, minimum, maximum, quartiles, 95% confidence intervals (CI). Categorical and binary data were summarized by frequency counts and percentages. Pre- and post-treatment BCVA Snellen fractions were converted to logarithm of the minimum angle of resolution (LogMAR) values for statistical evaluation using the conversion tool of Mussa et al. [27]. For very low vision, as counting fingers (CF), hand motion (HM) and light perception (LP), published LogMAR equivalent (2.1, 2.4, and 2.7, respectively) were applied [28]. BCVA change from baseline to the end of treatment was compared using paired t test. A Kaplan–Meier survival analysis of cure rates statistics (median, quartiles, minimum and maximum cure time) was also carried out to establish time to cure. The plot included both cured and not cured patients, with the latter censored at the time of failure. Univariate logistic regression analyses were also performed to examine, individually, the association between the primary outcome (cure rate) and prognostic factors (AK stage and prior use of corticosteroids); odd ratios (OR) and 95% CI for cure were calculated. Due to the limited number of patients and events, a multivariate logistic regression was not performed for a high risk of being underpowered. Results During the CUP, 58 patients were treated by the ten Italian centers that were eligible to contribute to the study group, having met the criterion of treating a minimum of at least three patients. As shown in Figure S1 (Supplementary Material), 19 individuals did not meet the inclusion criteria, resulting in 39 patients for the outcome analysis. Baseline Demographic, Treatments, and Diagnostic Tests These are summarized in Table 1. Fourteen patients were male (35.9%) and 25 were female (64.1%). Almost two-thirds (25/39, 64.1%) were in the 18–50-year age group. Both eyes were affected in 4/39 patients (10.3%). Contact lens use was the risk factor for all patients except two for whom lens wear data were missing. Treatment with PHMB 0.08% from symptom onset was delayed by more than 30 days in 23/39 (59%) of patients with a median of 39 days [interquartile range (IQR) 58 days]. Prolonged delays resulted from the inclusion of 13 patients being diagnosed and treated for AK with alternative AAT before having access to PHMB 0.08% (see Supplementary Material, Table S1 and Table S2, for more details). In this subset, the time from symptom onset to the start of prior AAT was not recorded. All cases were positive for at least one diagnostic test: culture ± PCR ± IVCM. Treatments, not mutually exclusive, used before starting PHMB 0.08% included antibiotics in 31/39 (79.5%), antivirals in 14/39 (35.9%), antifungals in 3/39 (7.7%), corticosteroids in 21/39 (53.8%), and other AAT in 13/39 (33.3%). Only eight patients (20.5%) had stage 1 AK (epithelial involvement only), while the remainder had more advanced disease; 16 (41.0%) had a corneal ring infiltrate (stage 3 AK). Table 1Baseline* demography, treatments, and diagnostic testsNumber of patients (eyes)39Age range (years)n (%) > 184 (10.3)18–3512 (30.8)36–5013 (33.3) > 5010 (25.6)Sexn (%)Male14 (35.9)Female25 (64.1)Bilateral diseasen (%)Yes4 (10.3)No35 (89.7)Contact lens wearn (%)Yes37 (100)No0 (0.0)Unknown2Time from onset of symptoms (days) to the start of treatment with PHMB 0.08%≤ 30 days, n (%)16 (41%) > 30 days, n (%)23 (59%)Median39IQR (Q1, Q3)58 (18,76)Previous treatments1n (%)None3 (7.7)Antibiotics31 (79.5)Corticosteroids21 (53.8)AATs213 (33.3)Antivirals14 (35.9)Antifungals3 (7.7)Diagnostic testsn (%)IVCM alone22 (56.4)IVCM + Culture5 (12.8)IVCM + PCR8 (20.5)IVCM + Culture + PCR1 (2.6)Culture alone2 (5.1)PCR alone1 (2.6)Stage of disease (n, %)3n (%)Stage 18 (20.5)Stage 215 (38.5)Stage 316 (41.0)AAT anti-amoebic treatments, IVCM in vivo confocal microscopy, IQR interquartile range, PCR polymerase chain reaction, PHMB polyhexamethylene biguanide, Q quartile, SD standard deviation*Baseline is the date of the start of treatment with PHMB 0.08%1Not mutually exclusive2PHMB 0.02% n = 8; chlorhexidine 0.02% n = 2; propamidine 0.1% n = 1; hexamidine 0.1% n = 1; PHMB 0.02% + propamidine 0.1% n = 13Stage 1: corneal epitheliopathy only. Stage 2: one or more corneal epithelial defects, perineural infiltrate, stromal infiltrate. Stage 3: a corneal ring infiltrate with one or more features of stage 2 (Robaei et al. [25]) Baseline* demography, treatments, and diagnostic tests AAT anti-amoebic treatments, IVCM in vivo confocal microscopy, IQR interquartile range, PCR polymerase chain reaction, PHMB polyhexamethylene biguanide, Q quartile, SD standard deviation *Baseline is the date of the start of treatment with PHMB 0.08% 1Not mutually exclusive 2PHMB 0.02% n = 8; chlorhexidine 0.02% n = 2; propamidine 0.1% n = 1; hexamidine 0.1% n = 1; PHMB 0.02% + propamidine 0.1% n = 1 3Stage 1: corneal epitheliopathy only. Stage 2: one or more corneal epithelial defects, perineural infiltrate, stromal infiltrate. Stage 3: a corneal ring infiltrate with one or more features of stage 2 (Robaei et al. [25]) Clinical Findings at Baseline These are summarized in Table 2. For the 34 cases in which BCVA was available, the Mean LogMAR (SD) was 1.2 (0.8); 20.6% of eyes had a LogMAR ≤ 0.3 (corresponding to ≥ 20/40 Snellen) while 47.1% of them had low vision (LogMAR ≥ 1.0 (corresponding to ≤ 20/200 Snellen). Eye pain was the most common presenting symptom; the intensity was described by patients as moderate to severe in 82.3% of cases. Conjunctival inflammation was moderate to severe in 30/39 (76.9%) of cases, while common corneal findings were stromal infiltrates (79.5%), epithelial infiltrates (61.5%), epithelial defects (53.8%), limbitis (48.7%), and corneal ring infiltrates (41%). Table 2Baseline* clinical findingsNumber of patients (eyes)39BCVA (LogMAR)Mean (SD) BCVA1.2 (0.8)BCVA ≤ 0.3 (n, %)7 (20.6)BCVA > 0.3 < 1.0 (n, %)11 (32.4)BCVA ≥ 1.0 (n, %)16 (47.1)Unknown5Eye painn (%)Absent4 (11.8)Mild2 (5.9)Moderate22 (64.7)Severe6 (17.6)Unknown5Conjunctival inflammationn (%)Absent3 (7.7)Mild6 (15.4)Moderate19 (48.7)Severe11 (28.2)Slit-lamp findings1n (%)Stromal infiltrates31 (79.5)Epithelial infiltrates24 (61.5)Epithelial defects21 (53.8)Limbitis19 (48.7)Corneal ring infiltrates16 (41.0)Stromal ulcers12 (30.8)Corneal neovascularization9 (23.1)Corneal scars9 (23.1)Perineural infiltrates5 (12.8)Anterior chamber inflammation3 (7.7)Hypopyon2 (5.1)Pseudodendrites1 (2.6)*Baseline is the date of the start of treatment with PHMB 0.08%BCVA best-corrected visual acuity, LogMAR logarithm of the minimum angle of resolution, SD standard deviation1Not mutually exclusive Baseline* clinical findings *Baseline is the date of the start of treatment with PHMB 0.08% BCVA best-corrected visual acuity, LogMAR logarithm of the minimum angle of resolution, SD standard deviation 1Not mutually exclusive Outcomes Full compliance with the ODAK dosing protocol was achieved in 29/39 (74.4%) patients. In the remaining cases, the treatment was further tapered after clinical resolution; however also in such cases the outcome was evaluated after discontinuation of all treatments for at least of 30 days. Table 3 summarizes the study outcomes. The medical cure was obtained in 35/39 patients (89.7%; 95% CI, 80.2–99.3). Treatment failure occurred in four patients. Two cases (both with stage 3 AK) underwent urgent therapeutic penetrating keratoplasty due to uncontrolled disease progression and corneal thinning; another patient (stage 3 AK) developed a corneal perforation managed with a conjunctival flap to maintain globe integrity; 1 patient (stage 2 AK) was nonresponsive to PHMB 0.08% and, after 5 weeks of monotherapy, a second AAT (hexamidine) was added resulting in a cure after a further 8 months. Representative corneal images from two of the 35 successfully treated patients and two of the four that failed are shown in Supplementary Material (Figs. S2 and S3, respectively). Among 37 cases for which BCVA data were available, the mean LogMAR BCVA (SD) at the end of the study was 0.7 (0.8). Figure 1 compares the change in vision from baseline to the end of the study for the 34 patients having both measurements. There was a statistically significant improvement from baseline BCVA of - 0.5 LogMAR (95% CI − 0.8 to − 0.3, p < 0.001, t test). Residual corneal scars were present in 27 cases (73.0%). Pain had resolved in all but two patients [34/36 (94.4%)]. Topical antibiotics, corticosteroids or a fixed combination of both, were given as adjunctive therapy in 53.8%, 25.6%, and 23.1% of cases, respectively. The mean follow-up period (SD) off therapy was 144 days (186); during this interval, four patients having had a medical cure needed a graft for optical reasons (three penetrating keratoplasty and one deep anterior lamellar keratoplasty) to remove scars and restore vision, once AK was resolved. Table 3Clinical outcomesNumber of patients (eyes)39Medical resolutionn (%)Cure135 (89.7)Failure24 (10.3)BCVA (LogMAR)Mean BCVA (SD)0.7 (0.8)BCVA ≤ 0.3 (n, %)17 (45.9)BCVA > 0.3 < 1.0 (n, %)9 (24.3)BCVA ≥ 1.0 (n, %)11 (29.7)Unknown2Corneal scarsn (%)Present27 (73.0)Absent10 (27.0)Unknown2Eye painn (%)Absent34 (94.4)Mild1 (2.8)Moderate1 (2.8)Severe0 (0.0)Unknown3Additional topical medications3n (%)None10 (25.6)Antibiotics21 (53.8)Corticosteroids10 (25.6)Antibiotic/corticosteroid combination9 (23.1)Antifungals3 (7.7)Other AATs41 (2.6)Antivirals1 (2.6)Other55 (12.8)AAT anti-amoebic treatment, BCVA best-corrected visual acuity, SD standard deviation1During the follow-up (144 ± 186 days, mean ± SD), 4 cured patients were subjected to optical keratoplasty (3 penetrating keratoplasty and 1 deep anterior lamellar keratoplasty)2Therapeutic penetrating keratoplasty n = 2; conjunctival flap n = 1; addition of a second AAT (hexamidine) n = 13Not mutually exclusive4Hexamidine was added to PHMB 0.08% in 1 patient (classified as failure)5Non-steroidal anti-inflammatory drugs n = 1; mydriatics n = 2; opioids n = 1; sympathomimetics n = 1 Clinical outcomes AAT anti-amoebic treatment, BCVA best-corrected visual acuity, SD standard deviation 1During the follow-up (144 ± 186 days, mean ± SD), 4 cured patients were subjected to optical keratoplasty (3 penetrating keratoplasty and 1 deep anterior lamellar keratoplasty) 2Therapeutic penetrating keratoplasty n = 2; conjunctival flap n = 1; addition of a second AAT (hexamidine) n = 1 3Not mutually exclusive 4Hexamidine was added to PHMB 0.08% in 1 patient (classified as failure) 5Non-steroidal anti-inflammatory drugs n = 1; mydriatics n = 2; opioids n = 1; sympathomimetics n = 1 Best-corrected visual acuity (BCVA) at the end of the treatment with PHMB 0.08% as function of baseline visual acuity for subjects having both measurements (n = 34). The graph shows that most eyes are below the line meaning they had an improvement of vision (LogMAR) to various degrees. EOS end of the study, SD standard deviations, LogMAR logarithm of the minimum angle of resolution, PHMB polyhexamethylene biguanide Time to Cure Figure 2 includes a Kaplan–Meier curve for the time to cure for all 39 individuals; patients not cured with medical treatment were censored at the time of failure. The analysis illustrates the cumulative probability of remaining uncured over time; the curve shows the progressive reduction of cases not cured. The median time to cure was 95 days (Interquartile range = 77 days), indicating that half of the eyes achieved a complete healing and resolution of inflammation within approximately 3 months. Kaplan–Meier time to cure estimates for 39 patients with Acanthamoeba keratitis treated with PHMB 0.08%. Four patients failed to achieve medical cure and were censored at the time of failure (indicated by an arrow in the plot). The graph illustrates a gradual decline in the survival probability, meaning the probability of remaining not cured. The table below the graph shows estimates of the time to cure. Q quartile Subgroup Analysis Table S1 (Supplementary Material) shows univariate logistic regression analyses for cure rate for each subgroup analyzed (AK stage, prior use of steroids and prior use of AAT). Confidence intervals around estimates are very wide reflecting substantial statistical uncertainty. Table S2 (Supplementary Material) compares demography, baseline characteristics and outcomes of the subset of 26 patients naïve (treated with PHMB 0.08% as their initial AAT) with that of the subset of 13 patients having had an alternative AAT before switching to PHMB 0.08% monotherapy due to its availability and simplified dosing protocol. No meaningful differences were observed between the two groups of patients. Safety Six adverse events were recorded; four of them were categorized as serious (three hypopyon and one perforation), and two as non-serious (one subepithelial infiltrate and one posterior blepharitis). All events were considered potentially related to the underlying disease, except for the subepithelial infiltrate, which was considered by the investigator a possible sign of drug toxicity. Discussion This retrospective single-cohort study evaluates the clinical experience with PHMB 0.08% across ten corneal specialist referral centers in Italy and represents the largest patient series evaluating the effectiveness and safety of this drug in clinical practice. The study reflected real-world referral patterns, including both treatment-naïve patients and those already treated with other AAT showing persistent stable disease. In the latter cases, treatment was switched to PHMB 0.08% monotherapy following its regulatory approval, which offered a standardized, higher-concentration formulation as an alternative to prior compounded or combination AAT. AK diagnosis was confirmed by IVCM, either as a standalone technique or combined with PCR or culture, while AK resolution was a clinical judgement based on the presence of a healed epithelium and the absence of corneal inflammation, because a negative IVCM or PCR is insufficient to provide certainty of disease resolution [21]. The definition of cure after at least 1 month without treatment was aligned with the ODAK trial [21], which reported no recurrences after 1 month off therapy, and represents a pragmatic choice in the context of limited published data on steroid use and the timing of relapse after treatment discontinuation. While this cut-off may not fully exclude late recurrences, patients in our cohort were followed for a longer period after treatment cessation (mean 144 days), strengthening confidence in the durability of the observed outcomes. The treatment success (89.7%) was high and comparable to that reported in the ODAK clinical trial (84.9%) [21]. A significant cure rate (81.3%) was also observed in cases of advanced AK (stage 3), which is considered a predictor of poor prognosis [1, 4, 9]. These favorable outcomes are likely attributable to two key factors: (1) the adherence to a consistent, evidence-based protocol ensuring uniformity in dosing frequency, duration, tapering schedules and patients monitoring across all participating centers [29]; (2) the use of a PHMB formulation containing a pharmaceutical-grade polyhexamethylene biguanide composed by high–molecular-weight long polymer chains, which may have provided superior therapeutic performance compared with compounded preparations [30]. As expected, the duration of treatment was highly variable and reflected the typically prolonged treatment time needed for the management of AK [31]. The Kaplan–Meier survival analysis shows a gradual increase in clinical resolution rate over time; eyes were treated for a median of 3 months, which correlates with a previous case series [23]. BCVA increased in most eyes with a mean improvement of − 0.5 LogMAR. Treatment failure occurred in four patients, two of whom had a therapeutic keratoplasty and one a conjunctival flap. These cases highlight the need for individualized therapy in refractory or advanced disease. However, this patient subset remains at higher risk of requiring therapeutic surgery with its associated poor outcomes [19]. In our series, four patients medically cured subsequently needed an optical keratoplasty to restore vision compromised by central corneal scarring [19, 32]. Overall safety was acceptable; despite the high concentration of PHMB and its prolonged use, only four serious AEs were recorded, most of which were consistent with the natural progression of AK rather than drug-related, thus supporting the favorable risk–benefit profile of PHMB 0.08%. The study's limitations are those common to retrospective data collection from medical records, which in our case resulted in only 39/58 (67%) of subjects being eligible for analysis, limiting the potential study size. An additional limitation was the expected heterogeneity of patients in terms of previous treatment (treatment-naïve vs. pre-treated). Yet, stratified analysis did not reveal major differences between treatment-naïve and pre-treated patients, supporting the overall consistency of the findings. However, the number of patients evaluated remains substantial for what is a rare disease and given that the success rate is comparable to that of the phase 3 ODAK trial, we think it reasonable to consider that these results are likely to reflect what can be expected in a larger series of patients. Conclusions This real-world, multicenter retrospective study confirms that topical PHMB 0.08% is an effective and safe therapy for Acanthamoeba keratitis. The medical cure rate observed across ten Italian corneal centers is consistent with the efficacy reported in the pivotal ODAK trial [21], supporting the clinical relevance and external validity of AKANTIOR as a first-line treatment for AK. Benefits seen under controlled clinical trial conditions can therefore be reproduced in clinical practice, even in the presence of advanced disease. It is important for physicians and patients to recognize the chronic nature of the disease and that after an intensive initial phase, a protracted treatment course is needed to achieve an effective medical cure. Alternative medical treatments may be required in few cases to resolve the infection in the event of an inadequate response, and eye surgery may be mandatory in case of high risk of corneal perforation. Future prospective studies are needed to further confirm these findings. Supplementary Information Below is the link to the electronic supplementary material. Supplementary file1 (DOCX 824 KB) Acknowledgements The authors thank Prof. John K.G. Dart, Honorary Professor at University College of London and Honorary Consultant at Moorfields Eye Hospital of London, for his review which has strengthened the value of this manuscript. The authors also extend their gratitude to all sub-investigators: Chiara Ancona (Eye Unit, ASST Spedali Civili di Brescia and Eye Unit, Department of Medical and Surgical Specialties, Radiological Sciences, and Public Health, University of Brescia), Federico Bortuzzi (Cornea and Ocular Surface Unit, San Raffaele Scientific Institute, Milan), Salvatore Di Vincenzo (Department of Biomedical and Dental Sciences and Morphofunctional Imaging, University of Messina), Martina Maccari (Department of Neuroscience, Psychology, Pharmacology and Child Health-NEUROFARBA-, University of Florence), Guglielmo Marino (Cornea Service, Department of Ophthalmology, University of Rome Campus Bio-Medico, Campus Bio-Medico University Hospital Foundation Roma), Sabina Robilotta (Ophthalmology Unit, Department of Experimental Medicine, University of Rome “Tor Vergata”, Roma), Rosaria Spinella (Department of Biomedical and Dental Sciences and Morphofunctional Imaging, University of Messina). Contract Research Organization (CRES Srl): Davide Integlia, Francesca Saracino, Claudio Ripellino (statistician). Finally, the authors thank all study participants for their involvement in the study. Author Contributions Concept and design: Karl Anders Knutsson, Antonella Franch, Adriano Fasolo and Vincenzo Papa. Data collection and interpretation: Karl Anders Knutsson, Antonio Di Zazzo, Emilio Pedrotti, Paolo Rama, Vito Romano, Anna Maria Roszkowska, Giacinto Triolo, Francesco Aiello, Stefano De Angelis, Rita Mencucci, Pasquale Aragona, Federica Birattari, Erika Bonacci, Matilde Buzzi, Marco Coassin, Adriano Fasolo, Mariangela Gari, Pia Leon, Carlo Nucci, Giorgio Paganoni, Matteo Pederzolli, Fabiana Secchi, Sabrina Vaccaro, Vincenzo Papa and Antonella Franch. Statistical analysis: Claudio Ripellino. Drafting the manuscript: Vincenzo Papa. All authors commented the first draft, read and approved the final version of the manuscript. Funding Sponsorship for this study and the rapid service fee were funded by SIFI SpA (Italy). Data Availability The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. Declarations Conflict of Interest Vincenzo Papa is an employee of SIFI SpA (Italy). Karl Anders Knutsson, Antonio Di Zazzo, Emilio Pedrotti, Paolo Rama, Vito Romano, Anna Maria Roszkowska, Giacinto Triolo, Francesco Aiello, Stefano De Angelis, Rita Mencucci, Pasquale Aragona, Federica Birattari, Erika Bonacci, Matilde Buzzi, Marco Coassin, Adriano Fasolo, Mariangela Gari, Pia Leon, Carlo Nucci, Giorgio Paganoni, Matteo Pederzolli, Fabiana Secchi, Sabrina Vaccaro and Antonella Franch have nothing to disclose. Ethical Approval The study was approved by the CET 1 Comitato Etico Regione Lombardia 1 of the coordinating center (San Raffaele Scientific Institute, Milan, Italy) on April 24, 2024 (approval number CET 148/2024). The study was performed in accordance with the Helsinki Declaration of 1964 and later amendments. Written informed consent was obtained from all patients prior to the commencement of the study, including consent to the use of their clinical data for scientific purposes. Identifying characteristics within the data were replaced with a pseudonym, which does not allow the data subject to be directly identified. References Associated Data Supplementary Materials Supplementary file1 (DOCX 824 KB) Data Availability Statement The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.