ObjectiveTo explore safe dosage of single intravitreal injection of ganciclovir (IVG) in healthy rabit eyes, and to explore retinal toxicity of different dosage of ganciclovir after continues intravitreal injection into the vitreous cavity of healthy albino rabbit eyes. MethodsTen healthy New Zealand albino rabbits were divided into 5 groups with 2 rabbits in each group. Each group was injected with 1 mg/0.025 ml, 2 mg/0.025 ml, 5 mg/0.025 ml, 10 mg/0.025 ml ganciclovir or 0.025 ml saline (control group). After 1 week of intervention, rabbits were examined by ultra-wide-angle fundus photography, optical coherence tomography (OCT) and full field electroretinogram (ERG). The maximum mixed response of rod and cone cells (Max-R) was measured under dark adaption conditions, cone response (Cone-R) and 30 Hz flicker response (30 Hz-R) were measured under light adaption conditions. Twenty-four healthy New Zealand albino rabbits were randomly divided into a low-dose experimental group, a low-dose control group, a high-dose experimental group, and a high-dose control group, with 6 rabbits in each group, with the right eye as the experimental eye. The rabbits in the high-dose experimental group were continuously injected with ganciclovir 2 mg/0.025 ml, once a week, for a total of 4 times. The rabbits in the low-dose experimental group were injected with 1 mg/0.025 ml ganciclovir, the induction period was 2 times/week, a total of 4 times; the maintenance period was 1 time/week, a total of 2 times. The rabbits in the high-dose control group and the low-dose control group were injected with 0.025 ml normal saline into the vitreous cavity respectively. Full-field ERG examination was performed 1 day before each injection and 1 week after the last injection. Max-R was measured under dark-adapted conditions, and Cone-R and 30 Hz-R were measured under light-adapted conditions. OCT was recorded before the first injection and one week after the last injection. One week after the last injection, the experimental rabbits in each group were sacrificed for hematoxylin-eosin staining, and the retinal structure was observed under a light microscope. The comparison of a-wave and b-wave amplitude of Max-R, Cone-R and 30 Hz-R amplitude at different time was performed by two independent sample nonparametric test. ResultsThere were no abnormal results of fundus photography, OCT and ERG after single intravitral injection of 1 mg or 2 mg ganciclovir. One week after single 5 mg IVG, fundus photography of rabbits showed vascular occlusion and preretinal hemorrhage and ERG showed slight decrease of amplitude of Max-R, Cone-R and 30 Hz-R. One week after single 10 mg IVG, retinal necrosis and exudative changes were also observed. OCT showed edema and unclear retinal structure in the necrotic area. ERG showed significant decrease of amplitude of Max-R, Cone-R and 30 Hz-R. After continuous IVG in high dose and low-dose experimental group, the amplitude of Max-R a wave (Z=－0.160, 0.000) and b wave (Z=－0.321, 0.000), Cone-R a wave (Z=－0.641,－0.641) and b wave (Z=－0.321, －0.160), and 30 Hz-R (Z=－0.321,－0.160) showed no difference compared to control group. No histologic evidences of retinal microstructure abnormalities were found in both groups. OCT and fundus photography before and after the intervention did not show any difference, either. ConclusionThere was no retinal toxicity of continuous 1 mg or 2 mg IVG recorded in albino rabbits.
ObjectiveTo observe aqueous cytomegalovirus (CMV) DNA load in patients with cytomegalovirus retinitis (CMVR) after allogeneic hematopoietic stem cell transplantation (Allo-HSCT), and to explore influencing factors for transient elevation of CMV-DNA load during the treatment. MethodsA retrospective study. From January 2016 to July 2020, 28 eyes of 19 patients with CMVR after Allo-HSCT diagnosed in the Department of Ophthalmology of Peking University People's Hospital were included in the study. Among them, there were 8 males with 12 eyes, 11 females with 16 eyes; the mean age was 28 years; 10 patients were unilateral and 9 patients were bilateral. During the course of treatment and follow-up, the blood CMV-DNA remained negative. All patients were treated with intravitreal injection of 60 mg/ml ganciclovir 0.05 ml (containing ganciclovir 3 mg), twice a week for two weeks in induction phase and weekly injection in maintenance phase. Aqueous humor sample was collected during injection of ganciclovir (IVG) and CMV-DNA load was determined by real-time quantitative polymerase chain reaction. Intravitreal treatment was terminated if aqueous CMV-DNA load turned negative after the fourth or later intravitreal injection. The patients were followed up every 2 weeks for at least 6 months. Serum CMV-DNA was negative in all patients during treatment and follow-up. All the eyes were divided into continuous decline group and non-continuous decline group depending on whether there was transient elevation of aqueous CMV-DNA load, and data between two groups were compared. Pearson linear regression analysis was used to analyze the correlation between aqueous CMV-DNA load and injection times or treatment duration. ResultsAt the end of treatment, the median number of IVG in the affected eye was 7 (4, 9). The results of correlation analysis showed that the aqueous humor CMV-DNA load of the affected eye was related to the number of treatments [R2=0.385, P＜0.000 1, B=－0.237 log10 copies/(ml · time)], and the duration of treatment [R2=0.394, P <0.000 1, B=－0.301 log10 copies/(ml · week)] were negatively correlated. Among the 28 eyes, 13 eyes (46.4%, 13/28) in the continuous decline group and 15 eyes (53.6%, 15/28) in the non-sustained decline group. Baseline visual acuity (t=－1.223), intraocular pressure (t=1.538), aqueous humor CMV-DNA load (t=－0.109), retinitis lesion area (Z=－0.308) in the continuous decline group and the non-continuous decline group), the number of quadrants involved (Z=－0.024) and whether the macula was involved (Z=－1.826), combined with anterior segment inflammation (Z =－0.499), combined with high intraocular pressure (Z=－1.342), terminal visual acuity (t =－0.845), intraocular pressure (t=－0.068), total IVG times (Z=0.907), age (Z=－0.832), gender composition (Z=－1.074), etc. The difference was not statistically significant (P＞0.05). ConclusionThe CMV-DNA load in aqueous humor decreases by about 50% every week during the treatment of CMVR eyes after Allo-HSCT; the transient increase in the CMV-DNA load in the aqueous humor during treatment does not affect the treatment process and clinical prognosis.