4. DISCUSSION
4.1. Evaluation of genotoxicity by in vivo micronucleus test.
Increased cancer risk is a serious adverse effect among patients undergoing immunosuppressive therapy. Micronucleus frequency has been reported to be significantly higher in pediatric patients with immunosuppressive therapy after kidney transplant (30). Although the dose challenged in this paper is topical and much lower, this novel RL formulation is being proposed for a chronic disease that may imply long term therapy, which is why assessment of genetic damage is mandatory.
The results presented in Figure 1 indicate a lack of genotoxicity effect, which is consistent with studies reporting that unlike other immunosuppressants, when used specifically for ocular inflammation, rapamycin has been reported to inhibit immunosuppression-induced neoplasia (31). This could be explained by the different mechanism of action of rapamycin versus other immunosuppressants. Oncogenic transformation, one of the most worrying consequences of DNA damage, is favored by the loss of cell cycle control and the activation of growth promoting pathways, such as the signaling path involving the mammalian Target of Rapamycin (mTOR). Rapamycin and other inhibitors of mTOR decelerate cell proliferation and contribute to avoid oncogenic transformation by suppressing the signals required for cell cycle progression, cell growth and proliferation (32). Thus, RL as well as its liposomal vehicle, are not expected to produce genetic damage when administered in mammals.
4.2. Evaluation of mucous irritability potential by HET CAM analysis
HET CAM test is an in vitro alternative to the in vivoDraize test, which is one of the most criticized methods of ocular irritation testing because of the injuries inflicted on test animals. The occurrence of vascular injury or coagulation in egg-chorioallantoic membranes in response to a compound, has a good correlation with the Draize rabbit test (33). This test provides information about immediate effects after administration of a product, mainly by vascular alterations, but also by protein interactions making it an adequate pre-screen method of eye injury hazard potential. The results indicate that the product is not expected to produce irritation effects on ocular membrane when tested in vivo. The obtained results were expected, taking into consideration previous reports of lack of ocular toxicity of rapamycin in animals. However, testing of the vehicle irritation potential and the combination with rapamycin was yet to be proved. These results justify the innocuity of the product in order to be tested in live animals, as well as confirms biocompatibility of the selected nanocarrier.
4.3. Determination of pyrogenicity
Pyrogens are substances that can produce fever when present as contaminants in a drug. Most pyrogens are biological substances derived from microorganisms that trigger an immune response, they can be life-threatening to patients because the produced systemic reaction can go from fever to neurologic effects, shock and death. These results are of great importance since some of the most recognized drawbacks for the application of liposomal technology in medicine are presence of organic solvent residues, difficult pyrogen control and sterility assurance, poor stability and adequate size distribution. It is worth mentioning that the formulations were also tested for sterility with good compliance with FEUM standards (data not published), along with absence of pyrogenic reaction. This means that proprietary methodology used for the preparation of the samples tested in this work is suitable for commercial production.
Aside of ruling out endotoxin contamination, absence of pyrogenicity gives indirect information on acceptable size distribution of RL, since it has been reported that liposomes larger than 200 nm tend to cause non-endotoxin-dependent rise in temperature, probably due to the metabolism of lipid mediators like prostaglandins (34). Lack of pyrogenic reaction suggests a suitable size distribution of the RL liposomes for intraocular injection.
4.4. Subacute toxicity in vivo in male New Zealand rabbits after subconjunctival injection.
Subacute systemic toxicity is defined as the adverse effects occurring after multiple or continuous exposure between 24 h and 28 days. For this kind of studies, the FDA recommends testing the maximum dose intended for therapeutic use, and 3 and 5 times that dose to identify effects associated with overdosing and increased duration of administration (35). Also, in this early stage of our development process it will help to establish a safe dose range for future optimal dose exploration.
In this study the dose range tested was limited to only 3 times the dose intended for therapeutic use and the administration frequency was doubled. Rationale for these modifications rely on the limitations of the route of administration. The maximum volume generally recognized as safe for subconjunctival injection (SCJ) is 0.5 ml, for this study the highest dose was rounded to 0.45 ml that represents 3 times the therapeutic dose. At first, we tried to inject 0.45 ml divided in 2 injections of 0.225 ml, considering that an injection volume of 0.2 ml is commonly used in practice with no remarkable adverse effects. However, significant discomfort was observed in the rabbits when manipulated for the second injection, apparently due to the stress caused by prolonged manipulation. Ultimately, it was decided to perform a single injection for subsequent administrations. Regarding administration frequency, instead of testing a higher dose, weekly administration was established in the protocol, that frequency is twice as frequent as the dose regimen considered for therapeutic use.
4.4.1. Metabolic changes
No difference in mean body weights in the comparisons between treatment groups was found. Although comparisons between final and initial weight show a slight alteration was observed, it is attributable to the innate growth of the rabbits. Hence, no significant metabolic changes were noted.
4.4.2. Clinical evaluation and histopathologic analysis
Macroscopic evaluation only revealed mild alterations in 3 subjects after the second injection. Necropsy detected discrete pulmonary, renal and hepatic congestion in all test and control animals probably related to the euthanasia method. Other organs did not present relevant disturbances. The microscopic examination of histologic slices after necropsy revealed odd pathology results for each tissue, (Figure 5A). For both right and left eyes, lymphoplasmacytic uveitis was detected mainly in control subjects. Due to these alterations, especially in rabbit #1, all the slices were stained with Gram. One subject (rabbit 1, group 3) was detected with presence of Encephalitozoon cuniculi spores, (Figure 4B). This subject exhibited multifocal lymphoplasmacytic infiltrates in stroma and ciliary body of the right eye, this finding was also detected in right lacrimal internal gland, as well as mild multifocal necrosis. The left eye of this specimen showed similar injury, but the infiltration covered also the iris structure. In the left lens slice multiple spores of Encephalitozoon cuniculiwere found, as well as in liver and kidney samples. Optical nerve samples of this subject presented mild multifocal lymphoplasmacytic meningitis.
Based on macroscopic and microscopic findings, it was determined that 6 subjects (2 rabbits from group 2, 3 rabbits from group 3 and 1 rabbit from group 4) presented injuries compatible with encephalitozoonosis but histopathological confirmation of spores was only possible in rabbit #1 from group 3. Most of the injuries observed were similar between groups, this statement was confirmed by statistical analysis that displayed no significant differences among treatment groups so no relationship between treatment and organic damage could be elucidated. Figure 5 summarizes the injuries observed in the microscopic examination of the subjects. It can be inferred that the highest injury levels were presented in group 3 treated with liposomal vehicle. The decreased quantity and level of injuries detected in animals administered with rapamycin containing formulations compared with empty liposomes may be related to the local immunosuppressant and anti-inflammatory activity of rapamycin (36). Lymph nodes in all groups (Figure 5A) exhibited lymphoid hyperplasia at mild multifocal level, these can be appreciated in Figure 4D where germinal centers of the lymph node present discrete hyperplasia. Also, as mentioned, all the rabbits showed lymphoplasmacytic dacryoadenitis. These findings are consistent with the presence of Encephalitozoon infection (37).
Histologic examination of the liver detected hepatocellular degeneration in all groups, Figure 5A. Group 1 showed only mild hepatocellular degeneration and congestion. Furthermore, subjects of groups 2, 3 and 4 presented moderate periportal lymphoplasmacytic hepatitis, Figure 4E. Possibly, hepatitis could be associated with lymphoid activity and hepatocellular degeneration increase and necrosis could be related toEncephalitozoon cuniculi infection, Figure 4F, since these lesions were found near the parasite spores in rabbit #1 (38). It is difficult to attribute these injuries to the administration of the test product since the most affected rabbits were in the control and vehicle groups. Also, it was not possible to determine a statistically significant difference between treatment groups. Mild to moderate multifocal tubular degeneration was observed in the kidneys of all groups. Additionally, lymphocytic interstitial nephritis in groups 2 and 3 was presented with one subject at level 4 and other at level 7, which is represented in Figure 4G. Both, degeneration and nephritis could be a consequence of the parasitic infection. No trend was observed in kidney injuries between groups due to rapamycin administration [37].
Lymphocytic encephalitis and satellitosis were exhibited in the brain sample of one subject of group 2 and granulomatous meningoencephalitis was displayed in one subject of group 3. All cases were at moderate multifocal injury level. Figure 4H displays the encephalic section of rabbit #1 confirmed with E. cuniculi infestation, where it can be observed an inflammatory infiltrate as well as plasmatic cells surrounding blood vessels. The optical nerve exhibited lymphoplasmacytic-meningitis in one rabbit of groups 2 and 3, also satellitosis and gliosis was observed in a rabbit from group 3, Figure 4l. Those pathologies could be linked again to parasitic infection. No damage was observed in groups 1 and 4 either in brain or optical nerve (38-40).
Our animals were certified as healthy by the provider at the beginning of the study, however we detected infestation in our test population.Encephalitozoon cuniculi is a common opportunistic protozoan in laboratory animals which is very hard to identify until symptoms appear (41). It is an obligate intracellular microsporidian parasite whose target organs are kidney, lens and nervous tissue. Immunocompetent animals usually are subclinical carriers, but immunocompromised hosts often present chronic granulomatous inflammation. (42). Regarding this, we could hypothesize that the presence of this infection could have been associated and promoted by immunosuppressive activity of rapamycin, however no microorganisms could be detected in the subjects from groups 1 and 2, administered with rapamycin loaded formulations. As discussed above, parasitic infestation was confirmed only for rabbit #1, administered with 450 µL of liposomal vehicle. Also, all kinds of lesions encountered were statistically similar in all the treatment groups. It has been reported that cerebral lesions can only be observed about 8 weeks after initiation of antibody response to the infestation which takes place within 3 weeks post-infection (43). The duration of our experiment was of 3 weeks, with 10 additional days of quarantine where no obvious clinical signs of disease were observed. Since the age of the animals was of 10 weeks at the beginning of the quarantine period, is probably that these subjects were subclinical carriers that represented an infection focus of dissemination to the rest of the animals from urine excretion of spores that starts 3 to 5 weeks after antibody response (43). Thus, the rest of the animals were probably in a primary stage of the infestation at the moment of the necropsy due to urinary horizontal dissemination at quarantine period. This is also consistent with the presence of kidney injury in 8 of 12 subjects, considering that kidney damage is part of the primary stage of the disease. (44)
Despite the infestation found in one subject, there was no relationship between injuries and treatment according to ANOVA analysis, Figure 5B, which means no effect can be attributable to rapamycin formulations. All the lesions observed were attributable to encephalitozoonosis and no other significant lesions were assessed in the high or low dose test groups. The fact that the confirmed subject was administered with liposomal vehicle without rapamycin suggests that other variable may have participated in the immunosuppression of the animals. In this manner, it has been studied that in assays with live laboratory animals, alterations due to stress often occur during toxicity studies and may interfere with the interpretation of the results. As discussed before, some of the evaluated parameters suggested that the animals were affected by stress during the study which may explain the development of the opportunistic infection. (45)
Finally, it is worth mentioning that the etiopathological origin ofEncephalitozoon Cuniculi in rabbit lens has been reported in literature (46). Ingestion of contaminated food or water is the most likely origin. Trans-placental and respiratory routes are also possible however less likely due to certification of health from the animal provider. We also performed cultivation of the formulation, yielding negative results to this parasite.
4.4.3. Biochemical assay
Samples were analyzed to observe any renal or hepatic alteration caused by administration of liposomal formulations with respect to water for injection as a control. Differences in the aforementioned parameters were analyzed at 0, 10 and 22 days after subconjunctival injection.
Biochemical measured values were between reference levels, therefore no influence from product administration can be assumed. Differences observed in urea and creatinine levels showing an upward trend in time can be considered normal due to increase in body weight of the rabbits due to normal growth. Cholesterol levels were significantly higher for group 3, treated with placebo liposomes. This difference was found because basal measurements in this group presented the highest values for this analyte therefore differences that cannot be associated with product administration. According to literature, administration of rapamycin has been associated with cholesterol serum elevation at rapamycin doses from 1-7 mg/day (47), however there was no increase observed in cholesterol levels in rapamycin treated groups. This could be explained by the very low dose that was locally administered, in conjunction with the liposomal carrier that encapsulates the drug and may inhibit systemic effects (48).
Concerning liver function tests, ALAT results showed difference between groups 1 and 2 for all sampling times because group 1 presented higher values ​​for this parameter since basal measurements. Moreover, since the difference was observed throughout the whole study, it is probably not related to hepatic injury. This is also supported by a lack of difference between treatments for ASAT and AP results, which reaffirms that there is no evidence of damage on hepatocellular integrity.
In a similar manner, there were differences between group 1 and 3 in serum total protein results. Since these findings were observed since basal measurements it lacks significance considering that mean results of each group are in the normal range. These results presented a statistically significant gradual increment between sampling times, associated with normal animal growth. The same upward trend was observed for albumin and calcium, whereas for phosphorous there was observed a downward trend.
Total bilirubin concentration was significantly decreased from basal measurement with respect to second sampling time, but this was considered irrelevant because mean results of each sampling time were in reference range. Also, alkaline phosphatase results did not show alterations, therefore concluding that biliary flux was not affected. None of the serum biochemical parameters related to liver and kidney function showed results indicative of toxicity due to product administration in low or high dose, as well as the liposomal carrier.
Hematologic parameters showed significant differences in hematocrit and erythrocyte count between groups. This difference was due to higher values ​​of group 4, presented in basal measurements. Bearing in mind that this difference was found in basal measurements from the control group, it has no physiological relevance. This may be explained as an effect secondary to a hemoconcentration state of the animals at the beginning of the study. Erythrocyte count was slightly out of normal range in basal measurement of the control group, along with hematocrit results that were in the upper limit of the normal range.
Significant differences in leucocyte count were also observed. Specifically, there was an increase in neutrophils in groups 1, 2 and 3. In all groups, an increase in monocytes and a decrease in lymphocytes at the first sampling time was observed. Even though statistically significant differences were observed, measurements were still in normal range. It has been reported that during toxicity studies with live laboratory animals, stress can affect diverse parameters including hematological assays. These results also suggested the possibility of an acute sub-clinical infectious process in all groups that will be discussed later. Some of the most common stress-related findings are lymphocyte depletion in thymus and spleen; resulting in altered circulating leukocyte counts, including increased neutrophils with decreased lymphocytes. (45) Animal stress is an inherent feature ofin vivo studies and may be triggered by many circumstances, among them, the handling process for dosing. In this case, the subconjunctival administration is a specialized type of injection that requires ocular topical anesthesia and must be performed by a specialist trained for this kind of administration. This process could have been more stressful for the animals than expected, summed with the expected stress generated by the sampling processes for biochemical and hematological parameters. This is consistent with the fact that lymphocytes and monocytes count alterations were also observed in the control group that was handled in the same manner, table 7.
4.5. Evaluation of acute retinal toxicity in vivo in New Zealand rabbits by intravitreal injection.
Intravitreal injection of rapamycin has shown good tolerability in some animal models so far [9,10,14,15]. However, the toxicity evaluation of every new formulation is mandatory being that the excipients of the formulation are not always compatible with the intraocular route (49). In this study, no statistically significant retinal acute toxicity was observed in electrical function evaluation or histological tissue examination. This is consistent with previous efforts to assess intravitreal toxicity with this drug. Fundus pictures showed no evident macroscopic alteration in retinal structure, however traumatic cataract formation was detected in some subjects 3 days after intravitreal administration. By slit-lamp examination, cataract formation was attributed to accidental lens traumatism during intravitreal injection technique since no trend or correlation could be determined between rapamycin amount and lens damage.
3.6.1. Electroretinography
Retinal damage can lead to vision loss due to lack of transmission of visual signal and has been in the top 5 most important causes of new drug candidate dismissal during drug development process (50). Dark-adapted ERG was used as a non-invasive in vivo evaluation of the electrical response of retinal cells. A significant decrease in amplitude of a or b-waves and/or prolongation of implicit times of these waves are indicative of retinal toxicity. In our study no statistically significant (p<0.05) reduction in amplitude, increased implicit time or alteration in waveform was observed between the basal measurements and post injection response. The same lack of significant difference was observed when comparing ERG results of tests groups versus control group 7 days after IVT administration.
This results are of great importance because we are challenging a new formulation of a drug that has been previously proved as safe for intravitreal injection by many other independent research groups, but also, this same drug has also been found to be toxic to the ocular structures when formulated in certain excipients with specific deterioration of retinal function evidenced by unfavorable electroretinography results (50). Interestingly, the b-wave response in the ERG of the group administered with 40 µg of rapamycin was statistically different from control group and of its original basal measurements due to increased amplitude observed in this group after treatment. These may suggest a potential mechanism to improve impaired visual signaling in the retina. This amplitude augmenting effect has been previously reported at least once to our knowledge by De Paivaet al (2019), with the use of sustained-release rapamycin systems and was considered transient and clinically irrelevant (52). Also, a protective effect of rapamycin on visual impairment during inflammation has been reported as the attenuation in a- and b-wave reduction due to induced inflammation (53). Further studies to explore its effects on human retinal signal transmission would be interesting.
3.6.2. Clinical evaluation and histopathologic analysis
Light microscopy showed normal tissue organization and cellularity of retinas in most subjects. Although, non-statistically significant alterations were observed, some subjects showed mild to severe injuries, specifically tissue degeneration and hyperplasia. Also, as can be appreciated in Fig. 11vi, lens fragmentation and presence of Morgagnian globules due to vacuole formation. This was consistent with slit-lamp examination and is probably related to histologic alterations observed since those subjects with cataract formation presented higher levels of histologic injury. In fact, it is noteworthy that the most severe injuries were observed in subjects treated with the intermediate dose of rapamycin, while only mild lesions were observed at the highest rapamycin dose. Despite these not significant (p<0.05) injuries, functionality of retinas was conserved from 40 to 440 µg of liposomal rapamycin as proved by ERG evaluations.
Limitations of this study include the small number of eyes included in each experiment and unexpected development of a subclinical disease common in laboratory animals. Further research in order to elucidate intraocular pharmacokinetics of the formulation and ocular biodistribution of the drug will be determining to achieve aim of clinical testing for the treatment of eye immune mediate diseases. A 40 µg liposomal rapamycin dose appears to have the best toxicity profile to be used by intraocular route, further research of its clinical effectivity is warranted.