Tel-Aviv University Sackler School of Medicine Department of Human Microbiology

Transcription

1 Tel-Aviv University Sackler School of Medicine Department of Human Microbiology IN VIVO ASSESSMENT OF AMPHOTERICIN B- LIPID EMULSIONS AS ANTIFUNGAL PREPARATION Thesis submitted for the degree Doctor of Philosophy By Yona Shadkchan Submitted to the Senate of Tel-Aviv University February 2002

2 This work was carried out under the supervision of Professor Esther Segal

3 TO MY LATE MOTHER ELENA To my husband Alex, who helped me immensely and without complaint To my father Valentin To my mother and father Zoya and Michael, who helped, help, and will help forever To my sister Lea And to my sons Ariel and Ron, the light of my life, who, despite seeing me only at dinner for months on end, still remembered to call me Mama

4 I would like to convey special thanks to my supervisor professor Esther Segal for her very much appreciated help and support throughout my work

5 I wish to thank the people who helped me in my work: Edi Sionov Dr Hani Sandovsky-Losica Yael Gov Dr Zeev Zaslavsky Dr Gabi Shavit Professor Yona Keisari Dr Noam Kariv and all employees of Beit Hayot The secretary and all staff of the Department of Human Microbiology for their practical help and advice.

6 ABSTRACT CONTENTS INTRODUCTION 1 I. Systemic fungal infections 1 II. Candidiasis. 2 II- 1. Candida albicans 3 II- 2. Non-albicans species 4 III. Antifungal therapy 5 III-1. Amphotericin B 5 III-2. History of AMB 6 III-3. Structure of AMB 6 III-4. Mechanism of action of AMB 7 III-5. Pharmacokinetics of AMB 8 III-6. Dosage and administration of AMB 9 III-7. Adverse effects of AMB 10 III-8. Immune response of antifungal therapy 11 IV. Formulations of AMB 12 V. Lipid emulsions 14 VI. AMB and lipid emulsions 16 VII. Characterization and in vitro activity of our formulation of AMB and 16 Intralipid AIMS OF THE STUDY 19 MATERIAL AND METHODS 20 Antifungal compounds 20 Fat emulsion 20 Organisms 20 Growth of fungi 21 Preparation of AMB-Intralipid admixtures 21 I. In vitro Activity of AMB-IL. 21 Preparation of inocula 22 Drug dilution 22 Performance of test 22 II. Normal (naïve) animal model 23 Animals 23 Animal model 23 Mortality 24 Morbidity 24 Determination of fungal elements in tissues 24 Candida CFU enumeration 25 Treatment of normal mice with AMB formulations 25 Evaluation of in vivo toxicity of AMB-IL 26 III. Compromised animal model 26 Induction of a compromised state 26 Induction of experimental candidiasis 26 Treatment of candidiasis with AMB formulations in compromised mice. 27 IV. Bioavailability and pharmacokinetics of AMB-IL 27 Drugs and drug preparations 27

7 Animals 28 Determination of AMB concentration 28 Extraction of AMB from blood 28 Extraction of AMB from organs 29 HPLC assay 29 V. Cytokines expression 30 Animals 30 Drugs 30 Total RNA isolation 30 RT-PCR system 31 Separation of the PCR products 33 VI. Statistical Analysis 34 RESULTS 35 General 35 I. In vitro Activity of AMB-IL 36 II. Normal (naïve) animal model 38 Experimental Animal Model 38 Treatment of Systemic Murine Candidiasis with Low Doses of AMB 42 Treatment of Systemic Murine Candidiasis with High Doses of AMB 46 III. Compromised animal model 51 Induction of compromised state 51 Animal model of systemic candidiasis in compromised mice 53 Treatment of C.albicans infection in compromised mice with low doses of AMB 58 Treatment of C.albicans infection in compromised mice with high doses of AMB 60 Treatment of C.glabrata and C.tropicalis infections in compromised mice 62 with AMB preparations IV. Bioavailability and pharmacokinetics of AMB-IL 67 AMB standard curve 67 AMB concentrations in serum 69 Levels of AMB in organs 72 Bioavailability of AMB in animals treated with higher doses of AMB 78 V. Cytokines expression 80 DISCUSSION 86 REFERENCES 94

8 CONTENTS OF FIGURES Figure 1. Mean survival rate (%) at day 42 of normal mice treated with 43 low doses of AMB formulations Figure 2. Course of infection in normal mice treated with low doses of 44 AMB formulations Figure 3. Mean survival time (follow up 42 days) of normal mice 45 treated with low doses of AMB formulations Figure 4. Mean survival rate (%) at day 42 of mice treated with high doses 48 of AMB formulations Figure 5. Course of infection in normal mice treated with high doses of 49 AMB formulations Figure 6. Mean survival time (follow up 42 days) of mice treated with high 50 doses of AMB formulations. Figure 7. Effect of CY pretreatment on mice 52 Figure 8. Treatment of C.albicans infection in CY compromised mice with 59 low doses of AMB formulations Figure 9. Treatment of C.albicans infection in CY compromised mice with 61 high doses of AMB formulations Figure 10. Treatment of C.tropicalis infection in CY compromised mice 63 with low doses of AMB formulations Figure 11. Treatment of C.tropicalis infection in CY compromised mice 64 with high doses of AMB formulations Figure 12. Treatment of C.glabrata infection in CY compromised mice with 65 low doses of AMB formulations Figure 13. Treatment of C.glabrata infection in CY compromised mice with 66 high doses of AMB formulations Figure 14. Calibration curve of AMB 68 Figure 15. AMB concentration in serum 70 Figure 16. AMB concentration (ug/gr) in the kidneys 73 Figure 17. AMB concentration (ug/gr) in the liver 74 Figure 18. AMB concentration (ug/gr) in the spleen 75 Figure 19. AMB concentration (ug/gr) in the lungs 76 Figure 20. AMB concentration (ug/gr) in the heart 77

9 CONTENTS OF TABLES Table 1. MIC values of Candida strains 37 Table 2. Survival of mice (at day 42), following IV administration of 46 Fungizone or AMB-IL Table 3. Inoculum of fungi that caused 100% mortality 53 Table 4. Murine systemic candidiasis in naive and compromised animals 57 Table 5. Pharmacokinetics of AMB preparations administered IV to mice 71 Table 6. AMB concentration in blood (µg/ml) and organs (µg/gr) after 79 treatment of mice with AMB-IL at the dose of 2mg/kg

10 CONTENTS OF PICTURES Picture 1. Infected visceral organs from C.albicans infected normal mouse 39 Picture 2. Calcofluor stained kidney s homogenates from C.albicans 40 infected normal mouse Picture 3. Images of PAS stained paraffin section of kidneys from 41 C.albicans infected normal mouse Picture 4. Kidneys from CY compromised mouse infected with Candida spp 54 Picture 5. Pas stained section of kidney from C.albicans infected CY 55 compromised mouse Picture 6. Mouse spleen RNA 81 Picture 7. Expression of cytokines in untreated mice 82 Picture 8. Expression of cytokines in normal mice treated with AMB 83 preparations Picture 9. Expression of cytokines in CY untreated mice 84 Picture 10. Expression of cytokines in CY compromised mice treated with AMB preparations 85

11 ABBREVIATIONS and SYMBOLS AIDS Acquired immunodeficiency syndrome AMB Amphotericin B. AMB-IL AmBisome C.albicans CFU CY DOC Fungizone HIV HPLC hrs IL ip IV MIC MST N OD PBS PMN RBC RT-PCR spp TNF WBC Amphotericin B-Intralipid admixtures Commercial liposomal preparation of AMB Candida albicans Colony forming units Cyclophosphamide Desoxycholate Standard formulation of AMB with DOC Human immunodeficiency virus High performance liquid chromatography Hours Interleukin Intraperitoneal Intravenous Minimal inhibitory concentration. Mean survival time Number Optical density Phosphate buffer saline Polymorphonuclear leukocytes Red blood cell Reverse transcriptase polymerase chain reaction Species Tumor necrosis factor White blood cell

12 ABSTRACT Background During the last 30 years there has been a dramatic increase in the number of patients with serious fungal infections, particularly in compromised and immunsupressed individuals.the most common of the opportunistic infections are systemic candidiasis, aspergillosis, cryptococcosis, and zygomycosis. Candida species are now the most common cause of serious fungal infections and the clinical manifestations are varied. Even though there are more than 150 species of Candida, C.albicans is the most pathogenic and most commonly encountered species among all. However, the frequencies of candidaemia due to non-albicans Candida spp. such as C.glabrata, C.krusei, C.tropicalis, C.parapsilosis, and C.lusitaniae have increased in recent years. The current gold standard agent used in treatment of invasive fungal infections remains Amphotericin B (AMB), whose activity is based on binding to sterol in the fungal membrane and causing leakage of cell components. AMB has the broadest spectrum of activity of any antifungal drug against the main fungal pathogens. AMB treatment is associated with significant adverse effects. Toxicity can be attributed to the high affinity of the drug to membrane sterol and the consequent formation of pores in the cellular membranes of various organs. Side effects associated with IV administration of AMB are common and include among others fever, chills, weight loss, nausea and vomiting. The most significant adverse effect of the drug is nephrotoxicity, reported in up to 80% of the patients. Since the major limiting factor in AMB therapy is its toxicity, several strategies have been explored for reducing toxic effects. The common strategy is to develop vehicles to deliver the drug other than deoxycholate, which is the vehicle in the standard IV formulation Fungizone. Incorporating AMB into liposomal or lipid complexes alters the pharmacological properties of drug and reduces the toxicity of AMB. Three new lipid-based formulations of AMB have been i

13 developed and are now available: Abelcet, Amphotec and AmBisome. However these new formulations are extremely expensive, which limits their use. An alternative delivery system could be lipid emulsions, such as Intralipid or Lipofundin, which are often used for parenteral nutrition. Attempts to prepare formulations by mixing AMB with emulsions prior to infusion to critically ill patients did not include characterization of these mixtures or assessment of their stability. In view of this information, we developed in our previous study Amphotericin B-Intralipid (AMB-IL) admixtures, which were prepared by vigorous overnight agitation of AMB with lipid emulsions. These formulations had several important advantages: they are composed of components that are approved for clinical use and commercially available at reasonable cost. Furthermore, they are stable, exhibited high in vitro efficacy against Candida species and had significantly reduced acute toxicity. The results obtained in our previous in vitro study led us to the present study in which we focused on more elaborate in vitro studies, and particularly, concentrated on evaluation of the efficacy of the AMB-IL preparations in vivo in an experimental systemic candidiasis model. Aims Thus, the main goal of this study is the evaluation of AMB-IL admixtures in vivo against experimental murine systemic candidiasis. The specific objectives included: Elaborate evaluation of in vitro activity of AMB-IL against various Candida species, including species known for resistance to AMB. The in vivo activity of AMB-IL against experimental systemic candidiasis induced by C.albicans and non-albicans spp in naive animals. The in vivo activity of AMB-IL against experimental systemic candidiasis induced by C.albicans and non-albicans spp in compromised animals. Biodistribution and pharmacokinetics of the AMB administered as AMB-IL. ii

14 Expression of cytokines as a result of AMB-IL therapy. Experimental Design and Experimental Data Systemic murine candidiasis in normal animals The initial experiments were devoted to the adaptation of a suitable animal model. Our model was based on previous research in our laboratory with C.albicans CBS 562, using IV inoculation. Our experiments revealed that using 4-6 weeks old female ICR mice injected with 5x10 4 organisms/mouse resulted in 100% of mortality within 5-10 days. Systemic candidiasis was determined by demonstration of fungal elements in visceral organs. Treatment experiments in normal animals Following the adaptation of a suitable animal model we treated infected mice 48 hrs post Candida inoculation with either AMB-IL or Fungizone at a concentration of 0.4mg AMB/kg in comparison to untreated animals. Treatment was administered during 5 consecutive days by IV injections. Animals were surveyed up to 42 days. Data obtained from 4 experiments with 108 mice showed that both formulations significantly increased the survival of the mice as compared with the sham treated controls. Furthermore over 50% of mice survived when treated with AMB- IL. The mean percentage of surviving mice at day 42 was 0, 24 and 52 for the untreated, Fungizone and AMB-IL treated groups, respectively. The follow-up of the course of infection indicated that the Fungizone and AMB-IL increased the survival time of the treated mice. The mean survival time (MST) was 7, 25 and 30 days for the untreated control group, Fungizone and AMB-IL treated groups, respectively. As AMB-IL treatment at these low doses did not save almost half of the animals from experimental candidiasis, we planned experiments with higher doses of AMB. Towards this aim we first had to establish the maximal doses of AMB to which the mice would be tolerant. We found that the maximum tolerated dose of Fungizone was 1 mg/kg/day x 5. Higher doses (1.2 iii

15 mg/kg) caused immediate death in 100% of the mice. In contrast AMB-IL at the dose of 2mg/kg/day x 5 (total 10mg/kg) did not cause death during an observation period of six weeks. Based on these results we then performed 4 additional treatment experiments with 104 mice with higher doses of AMB-IL. AMB-IL at the higher concentration of 2mg AMB/kg/day x 5 significantly increased the survival of the mice, almost up to 100% vs. Fungizone that led to only 38% survival. In addition, AMB-IL preparations at the higher dose prolonged the survival time. The results of this part of the study revealed that AMB-IL is significantly more effective in treating systemic murine candidiasis than Fungizone. In addition to C.albicans infection, we attempted to induce murine systemic candidiasis by inoculating mice with non-albicans species: C.glabrata, C.tropicalis, C.krusei and C.lusitaniae. However, all our attempts to elicit systemic candidiasis with non- albicans species in naive mice failed and we therefore turned to experiments in compromised animals. Systemic murine candidiasis in a compromised animal model We induced a compromised state in animals by the immunosuppressive agent cyclophosphamide (CY). Four weeks old ICR female mice were treated ip with CY at a dosage of 200mg/kg. A compromised state was observed on day 4 by a decreased number of white blood cell (from 8x10 6 /ml to 9.1x10 5 /ml) and by loss of body weight (from 26.2 gr. to 21.6gr). Hence, on day 4 post CY treatment the animals were inoculated IV with C.albicans, C.glabrata or C.tropicalis. The optimal inoculum for C.albicans was determined as 1x10 4 /mouse, for C.tropicalis 5x10 5 /mouse and 5x10 6 /mouse for C.glabrata. These concentrations caused 100% mortality within 7-16 days (MST was 9.36). Evaluation of systemic candidiasis was performed during a follow-up period of 42 days using criteria similar to those as for the naive animal model Treatment experiments in CY compromised mice Forty-eight hours after fungal inoculation various doses of AMB or AMB-IL admixtures (from 0.4mg/kg to 2mg/kg during 5 days) were administered IV and the survival rate and MST iv

16 were evaluated during an observation period of up to 42 days. Although treatment with low doses of AMB-IL preparations was more effective than conventional AMB, nevertheless, only 43.3% of thus treated mice survived. Treatment of animals with higher doses of AMB showed that all AMB formulations at high doses increased significantly the survival of the mice. Furthermore, AMB-IL at the higher concentration of 2mg/kg x 5 was very effective, 100% of treated mice survived during the experimental period, while among Fungizone treated mice only 48% of the mice survived. Biodistribution of AMB and AMB-IL in mice Towards the aim of assessing availability of AMB-IL we adapted a sensitive HPLC method for assaying AMB in blood and organs. Evaluation was carried out in comparison to animals treated with Fungizone or AmBisome. We found that the levels of AMB in the blood of mice injected with AMB-IL were consistently higher in comparison to that in animals treated with Fungizone, but similar to those treated with AmBisome. Furthermore, at 48 hrs no detectable levels of AMB in animals administered with Fungizone were noted, while those treated with AMB-IL or AmBisome revealed presence of AMB up to 72-hrs post treatment. We found that the highest AMB concentrations in animals treated with AMB-IL were present in the organs of the reticuloendothelial system: liver and spleen. In contrast, concentrations of AMB in the kidneys and lungs were lower in animals receiving lipid formulations of AMB. Cytokine expression At this stage of the study we investigated the expression of pro-inflammatory cytokines and non-inflammatory cytokines in animals treated with conventional AMB, commercially available liposomal AMB-AmBisome or AMB-IL. Total RNAs were purified from spleen and analyzed by one step RT-PCR. RNA samples were amplified using specific mouse cytokine v

17 primers for the following cytokines: TNF-α, IL-1β, IL-2, IL-6, and β-actin as a control. We found that treatment with AMB increased the production of pro-inflammatory cytokines in comparison to non-treated mice. In animals treated with AMB-IL or AmBisome the expression of proinflammatory cytokines was lower than in AMB treated mice. SUMMARY 1. The data presented in this study show that in normal animals treatment with AMB-IL increased significantly the survival rate and /or prolonged the survival time in comparison to treatment with conventional AMB. 2. AMB-IL can be administered at doses at which AMB is not tolerated. 3. In CY compromised animals AMB-IL admixtures were more effective than conventional AMB, especially when high doses of AMB-IL were used. 4. This was true not only for C.albicans infection but also for non-albicans spp, which could produce infection only in compromised animals. 5. HPLC experiments showed that concentration of AMB in serum from mice treated with AMB-IL is higher than in those treated with Fungizone, but similar to that in those treated with AmBisome. 6. Administration of lipid formulations of AMB resulted in higher concentration of AMB in the liver and spleen, but lower concentration in the kidneys and lungs. 7. In animals treated with AMB-IL the expression of pro-inflammatory cytokines was lower than in AMB treated mice. In conclusion, our results show that AMB-IL admixtures are effective for treatment of systemic candidiasis caused by C.albicans and non-albicans species, both in naive and compromised animals. Thus, Amphotericin B-Intralipid admixture, being a simple, standardized and inexpensive formulation, has great promise for treatment of candidiasis. vi

18 INTRODUCTION I. Systemic fungal infections During the last 30 years there has been a dramatic increase in the numbers of patients with serious fungal infections [11, 17, 22, 41, 42, 66, 73, 113]. Invasive fungal infections continue to be a major source of morbidity and mortality in compromised and immunsupressed hosts [42, 46, 98]. Several factors predispose patients to fungal diseases, either immune defects or mechanical defects [11, 16-18, 22, 48, 50, 62]. Prolonged neutropenia is a major immune factor contributing to fungal infections. Neutropenia may be associated with leukemia or may result from chemotherapy for other malignancies [2, 18, 65]. The incidence of fungal infections among patients with acute leukemia is 20-25% [11, 16, 17, 62, 65]. Another type of immune defect involves impaired cell-mediated immunity, which is associated with AIDS, lymphomas, chemotherapy and steroid therapy. Up to 80% of AIDS patients develop at least one fungal infection, and 10-20% of them die as a result [1, 11, 66, 85, 113]. Mechanical defects that cause fungal infections include skin and membrane alterations resulting from chemotherapy, invasive procedures such as placement of catheters, and events such as trauma, surgery and burns [1, 11, 22, 48, 92]. Susceptibility to fungal infections also increased with broad-spectrum antibiotic, immunosuppression or corticosteroids therapy [66]. Organ transplantat patients are also at high risk for fungal infections because the immunosuppression therapy is 1

19 necessary to prevent rejection of organs and incidence of fungal infections among transplant patients is very high 20-40% [58, 84]. The most common of the opportunistic infections are systemic candidiasis, aspergillosis, cryptococcosis, and zygomycosis [51, 98, 122]. In addition to these opportunistic pathogens, there are a growing number of fungal infections such as histoplasmosis, blastomycosis and coccidioidomycosis in compromised hosts [42, 81, 122]. The recent increase in systemic fungal infections is mostly due to the greater number of immunocompromised hosts such as AIDS patients and patients receiving aggressive chemotherapy for malignant diseases [1, 22, 58, 62]. II. Candidiasis. Candida is a family of organisms commonly found in the gastrointestinal tracts and in oral mucosa of normal individuals [15]. However, candida can produce infections and now Candida species are the most common cause of serious fungal diseases in the immunocompromised hosts [1, 22, 46]. The clinical spectrum of candidiasis ranges from superficial cutaneous and mucosal manifestations to deepseated infections of various visceral organs and even a systemic disease. Mucocutaneus candidiasis is manifested as oral thrush and vaginitis. Candida infections arising from the gastrointestinal tract also may manifest as esophagitis or gastrointestinal candidiasis. Candida may cause disseminated infections that can affect many organs, including kidneys, liver, spleen, heart, lungs, and brain in immunosuppressed patients [92-94, 99, 108]. Not only patients with a granulocytopenia or impaired cellular immunity (AIDS and cancer patients) are at risk of Candida infections but also those 2

20 who received broad spectrum antibiotics, chemotherapy or corticosteroids treatment for a long period [22, 48, 58, 83, 84, 92-94, 99, 108]. Central nervous system infections with Candida may result in brain abscess or meningitis. Other forms of invasive infections may produce endocarditis, myositis, arthritis, osteomyelitis, pneumonitis, or endophtalmitis [22, 48, 58, 83, 84]. Candida spp are now the fifth most common isolates from blood and the fourth most commonly recovered pathogens from all sites in intensive care units [84, 93, 94, 108]. Overall they account for 5-8% of all nosocomial pathogens and for 8-15% of all nosocomial bloodstream infections [93, 94]. The mortality in patients with invasive candidaemia is very high; it is reported that the overall mortality among patients with nosocomial candidaemia to be higher than 30% [1, 84]. II- 1. Candida albicans Even though there are more than 150 species of Candida, no more than ten cause disease in humans with any frequency [15]. Candida albicans is the most pathogenic and most commonly encountered species among all [1, 46]. C.albicans is the most common yeast colonizing damaged skin and mucosa and the most common causing fungaemia [17, 22, 46, 48, 66, 73, 84, 122]. Its ability to adhere to host tissues, produce secretory aspartyl proteases and phospholipase enzymes, as well as its ability to transform from yeast to hyphal phase are the major determinants of its pathogenesis [15, 88]. C.albicans is a commensal fungus that can be a component of the microbial floras of the oral cavity, gastrointestinal tract, or vagina [15, 88, 98]. Under conditions 3

21 of host immune suppression or when natural barriers (skin) to infections are degraded, colonization of C. albicans can give rise to opportunistic infections [92, 99]. These infections can be superficial, involving the mucous tissues or can be hematogenously disseminated, resulting in systemic candidiasis, which has a high mortality rate [1, 84]. People with severely compromised immune systems, such as cancer patients or organ transplant recipients, represent a major group at risk for serious C.albicans infections [1, 17, 18, 22, 92]. C.albicans still accounts for about 50-70% of disseminated infections, but there is a rising tide of opportunistic infections caused by non-albicans Candida species [20, 32]. II- 2. Non-albicans species While C. albicans is the most common agent that causes systemic candidiasis, the frequencies of candidaemia due to non-albicans Candida spp. such as C.glabrata, C.krusei, C.tropicalis, C.parapsilosis, and C.lusitaniae have increased significantly in recent years [14, 20, 32, 41, 42, 46, 51, 66, 73, 81, 98, 122]. Non-Candida albicans species have a tendency of decreased susceptibility to antifungal agents and cause infections associated with high morbidity and mortality. The distribution of species varies markedly between the countries [46, 66, 73, 93]. The frequency of candidaemia due to non-albicans spp is variable due to influences of many factors including underlying diseases, cytotoxic chemotherapy and usage of prolonged antifungal agents [21, 46, 66, 73, 93, 94, 98]. Among the nonalbicans species, C. glabrata and C.tropicalis are the most common. The frequency of non-albicans species (in different countries) causing candidiasis were15-30% for 4

22 C.tropicalis, 10-30% for C.parapsilosis, 15-35% for C. glabrata, 1-5% for C. krusei, 1-3% for C.lusitaniae and1-3% for C. kefyr [14, 20, 32, 41, 42, 46, 51, 66, 73, 81, 98, 122]. But overall 50-70% were attributable to C. albicans [17, 22, 46, 48, 66, 73, 84, 122]. III. Antifungal therapy The standard antifungal therapy includes the polyenes and the azoles, both of which target the ergosterol, a sterol component of the cell membrane of fungi [2, 97]. The current gold standard agent used in invasive fungal infections remains Amphotericin B [2, 52, 53, 110]. III-1. Amphotericin B In the management of serious systemic mycoses, the commonest choice of treatment is Amphotericin B (AMB), a polyene antifungal, given by intravenous infusion (IV) [2, 52, 53, 74, 75, 91, 97, 110]. AMB has a fungicidal activity by preferential binding to ergosterol that alters membrane permeability of fungi and causes leakage of cell components such as potassium and proteins [74, 75]. It has the broadest spectrum of activity of any antifungal drug against the main fungal pathogens [91, 110]. It is active against most common fungal infections, such as disseminated candidiasis, aspergillosis, zygomycosis, cryptococcosis, disseminated blastomycosis, coccidioidomycosis, histoplasmosis, sporotrichosis and others [12, 13, 52, 53, 75, 86, 91, 110]. 5

23 Aggressive early empiric use of AMB has also reduced the mortality in febrile neutropenic patients who have not responded to conventional antibiotic therapy [65, 68, 75]. The use of AMB is continuously growing as a result of the increasing incidence of life-threatening fungal infections particularly in immunocompromised hosts [2, 18, 22, 92, 97, 110, 113]. Furthermore, the use of AMB therapy is the standard of practice for the management of immunocompromised patients with persistent fever unresponsive to antibacterial treatment. III-2. History of AMB Amphotericin A and B are both polyene compounds that were originally isolated from Streptomyces nodosus, an Actinomycete, cultured from the soil of the Orinoco Valley in Venezuela [56]. AMB is a polyene antifungal agent, first isolated by Gold et al in The classic amphotericin B deoxycholate (Fungizone) formulation has been available since AMB is now also available in several lipid formulations. Amphotericin B has a very broad range of activity against most pathogenic fungi. III-3. Structure of AMB AMB is an amphoteric compound composed of a hydrophilic polyhydroxyl chain along one side and a lipophilic polyene hydrocarbon chain on the other. AMB is a polyene macrolide that consists of seven conjugated double bonds, an internal ester, a free carboxyl group and a glycoside chain with a primary amino group [9, 26, 36]. 6

24 Chemical structure of AMB It is either fungistatic or fungicidic, depending on the concentration of the drug achieved in the serum or tissues and the susceptibility of the pathogens [5, 26, 36, 39]. Its activity is maximal over the ph range of 6.0 to 7.5 [26]. AMB is highly lipophilic and practically insoluble in water and cannot be applied orally [36]. III-4. Mechanism of action of AMB Amphotericin B binds to sterols, preferentially to the primary fungal cell membrane sterol, ergosterol. This binding disrupts osmotic integrity of the fungal membrane, resulting in leakage of intracellular potassium, magnesium, sugars, and metabolites and then cellular death [2, 9, 10, 12, 26, 28, 47, 52, 74, 75, 82, 86, 107]. Human cell membranes also contain sterols (mainly cholesterol, which has lower affinity to AMB than ergosterol), and the damage to human and fungal cells may share common mechanisms [19, 47, 74]. Another proposed mechanism of the 7

25 chemotherapeutic effects of Fungizone is the oxidation-dependent, AMB-induced stimulation of macrophages [47, 119, 131]. This immunomodulation is augmented by oxidative metabolites such as hydrogen peroxide and may be due to auto-oxidation of the drug with formation of free radicals, which may also increase membrane permeability, especially to monovalent cations. III-5. Pharmacokinetics of AMB Following intravenous administration, AMB is bound to proteins (91% to 95%), primarily to albumin and to plasma lipoproteins and then redistributed from the blood into tissues [5, 9, 12, 36]. The most important depots of AMB in body tissues are liver (14 to 41% of dose), lungs (1.2 to 6%), and kidneys (0.3 to 2%), but interindividual variations are great [5, 36]. AMB follows a biphasic pattern of elimination from the plasma, with an initial half life of 24 to 48 hours, followed by a long elimination half life of up to 15 days, probably due to extremely slow release of the drug from peripheral tissues [64]. Detectable levels of the drug have been found in bile for up to 12 days and in urine for 27 to 35 days after administration [5, 9, 36]. The total body clearance of AMB is about 30 ml/min. (1.8 L/h), while renal clearance is very low, about 1 ml/min. (0.06 L/h) [5, 9]. AMB was detectable in the liver, spleen and kidneys for up to 1 year after termination of the therapy, although urine excretion at that time was too low to be detected. The pharmacokinetic profile of AMB is somewhat different in children than in adults [12, 13]. It is also noteworthy that despite crossing the placenta, AMB has been 8

26 administered during all stages of pregnancy without detectable adverse effects to the fetus [27]. III-6. Dosage and administration of AMB The drug is conventionally formulated as Fungizone. Each vial contains a yellow, fluffy lyophilized powder consisting of 50,000 units (50 mg) AMB and sodium deoxycholate (DOC) (approximately 41 mg), in sodium phosphate buffer. AMB is reconstituted with sterile water for injection without preservatives. The drug may then be diluted with 5% dextrose injection for IV infusion. AMB is usually administered by slow IV infusion over a period of approximately 6 hours [2, 52, 75]. The targeted daily dose varies from mg/kg of body weight depending on the severity of the infection, the nature of the pathogen, renal function and drug tolerance [86, 110]. Within the range of mg/kg the daily dose is maintained at the highest level that is not accompanied by unacceptable toxicity. Higher dosages, up to 1.5 mg/kg, are necessary in seriously ill patients, for infections caused by filamentous fungi, particularly zygomycosis and invasive aspergillosis [18]. Alternative routes of administration of AMB are very limited due to the poor absorbance of the drug: oral administration is suitable only for the prevention of fungal colonization of the gastrointestinal tract [110]. Intraocular local instillation has been recommended for the treatment of patients with fungal endophtalmitis. Nebulized AMB has been used for the treatment of patients with fungal infection in the respiratory tract [68]. AMB has been used in an attempt to eradicate fungal colonization of the respiratory tract and has been used to prevent pulmonary 9

27 Aspergillus infection in neutropenic patients [40]. Administration of AMB by the intraperitoneal route has been used to treat fungal peritonitis in patients receiving dialysis. Other uses of Fungizone include local instillation for the treatment of fungal infections of the ear, eye, lung cavities and joint spaces [11, 75]. III-7. Adverse effects of AMB The activity of AMB is based on its affinity to bind to the fungal membrane sterol ergosterol, thereby changing the permeability of the fungal cell membrane and leakage of cell contents, leading to fungal death [10, 12, 38, 107, 119]. In addition, AMB binds, although less, to membrane-sterol cholesterol of mammalian cells, which led to various toxic effect of this drug [19, 28, 47, 77, 107, 126]. Side effects associated with IV administration are relatively common and include fever, headache, anorexia, weight loss, nausea and vomiting, rigors, muscle and joint pain, dyspepsia, epigastric pain, local venous pain at the injection site with phlebitis and trombophlebitis, normochromic and normocytic anemia, and hypokalaemia [10, 12, 28, 119, 126]. The most significant adverse effect of the drug is nephrotoxicity, reported in up to 80% of the patients [28, 107, 126], including abnormal renal function such as hypokalaemia, hypomagnesaemia, azotaemia, renal tubular acidosis or nephrocalcinosis. It has been shown that the drug causes changes in tubular cell permeability to ions both in vivo and in vitro [19]. Hence, one possible explanation for AMB-induced nephrotoxicity (azotaemia) is tubuloglomerular feedback, a mechanism whereby increased delivery and reabsorbtion of chloride ions in the distal tubule initiates a decrease in the glomerular filtration rates of the nephron [28, 126]. 10

28 The following adverse reactions occur less frequently or rarely: anuria; cardiovascular toxicity including arrhythmia, ventricular fibrillation, cardiac arrest, hypo-or hypertension; coagulation defects; thrombocytopenia; leucopenia; agranulocytosis; convulsions and other neurological symptoms; anaphylactic reactions and acute liver failure [107]. III-8. Immune response to antifungal therapy Use of AMB is limited by severe side effects such as fever, chills, and hypotension. It has been hypothesized that these adverse effects of AMB are mediated through induction of proinflammatory cytokines [3, 123]. It s also reported that AMB increases the production of proinflammatory cytokines by human mononuclear cells, and this induction is mediated through activation of transcription of tumor necrosis factor- α (TNF- α) and interleukin-1 β (IL-1β) [123]. In addition, it is well accepted that polymorphonuclear leukocytes (PMN) play an important role in the first line defense against most opportunistic fungal infections including candidiasis and that PMN are activated by several cytokines, particularly TNF-α and IL-1 [85, 102]. TNF-α is known as one of the inflammatory cytokines principaly produced by activated macrophages and PMN themselves. TNF-α also demonstrated by several investigators to induce various inflammatory responses such as fever, enhanced blood coagulation, and inflammatory tissue destruction, ultimately resulting in toxic effects on the host [3, 123]. Thus the immunopharmacologic effects of AMB, including its antifungal activity and its toxic effects, may be mediated by inflammatory cytokines produced in the host [85, 102]. 11

29 IV. Formulations of AMB Since the major limiting factor in AMB therapy is its toxicity, several strategies have been explored for reducing toxic effects. The common strategy is to develop vehicles other than DOC to deliver the drug [8, 23, 25]. The search for effective, less toxic derivatives of AMB or for different delivery systems which may decrease its toxicity to the host without affecting its antifungal activity led to the development of formulations in which AMB was associated with liposome or lipid complexes [8, 25, 30, 49, 55, 59, 63, 67, 80, 100, 132]. Incorporating AMB into liposomal or lipid complexes revealed to reduction of toxicity without loss of antifungal activity of drug [30, 54, 59, 63, 89, 90, 128, 129]. Moreover, the reduced toxicity enabled to enlarge the dosage of AMB, thereby increasing the therapeutic efficacy of the drug [25, 49, 55, 67]. These lipid based preparations were explored in experimental models and also used in clinical practice and included the discoidal mixed micelles made of AMB and cholesterol sulfate (ABCD), and the ribbon-like mixed aggregates made of AMB with dimyristoylphosphatidyl choline and dimyristoylphosphatidyl glycerol (ABLC) [54, 124]. In addition, liposomal preparations were developed, which are AMB-containing small unilamelar vesicles [57, 72, 80]. 12

30 There are three lipid-based formulations of AMB now available in the USA and Europe: Abelcet TM, Amphotec TM (Amphocil TM ) and AmBisome [35, 59, 63, 67]. These commercial preparations allow patients to receive higher doses of AMB for long periods of time with decreased toxicity than that associated with conventional Fungizone [35, 37, 43, 59, 63, 72, 77, 100, 124, 125, 132]. The major problem of these lipid-associated preparations is their cost. They are extremely expensive, which prohibits their use as a drug of choice and results in a limited use of these preparations [96, 100, 132]. Commercially available formulations of AMB Formulation Manufacturer Carrier Colloidal type Target dosage mg/kg Daily cost $ (patient 70 kg) AMB with DOC Bristol-Myers Squibb, DOC 1 Micelles Fungizone Princeton, NJ, USA AMB lipid Liposome Company, DMPC 2 & Ribbons Princeton, NJ, USA DMPG 3 complex (ABLC) Abelcet TM ) AMB colloidal Sequus Cholesteryl Discs dispersion (ABCD; Amphocil TM or Amphotec TM ) Pharmaceuticals, Menlo Park, CA, USA or Liposome Company, Princeton, NJ, USA sulfate Liposomal AMB AmBisome Vestar Pharmaceuticals, San Dimas, CA, USA DSPG 5 unilamellar liposome Deoxycholate, 2 Dimyristoylphosphatidyl choline; 3 Dimyristoylphosphatidyl glycerol; 4 Distearoylphosphatidil glycerol 13

31 The exact mechanism(s) responsible for the reduced toxicity of AMB, when given as a liposomal or lipid complex formulation are not known. Selective transfer of AMB from these formulations to the target fungal cells, with reduced uptake into human cells, is believed to be an important mechanism of reduced toxicity [37, 49, 63, 71, 72, 129]. Both the phospholipid/amb ratio and the type of phospholipid used are important determinants of toxicity as well as of fungicidal activity. V. Lipid emulsions The extreme cost of the lipid-based AMB preparations led to the continuation of the search for alternative delivery systems. Such systems may possibly be based on lipid emulsions available for parenteral nutrition in clinical practice, such as Intralipid [6,7]. In the early 1960s, Arvid Wretlind and his colleagues developed the first safe parenteral fat emulsion for clinical nutrition, called Intralipid [133]. This product provides concentrated energy and essential fatty acids to patients who cannot eat. Now, the clinical use of parenteral fat emulsions is commonly accepted as a part of nutrition therapy [6, 7, 38, 44, 103]. 14

32 A lipid emulsion consists of a water phase with droplets composed of a triglyceride core (diameter nm) stabilized with a phospholipid monolayer (2 3 nm) [38, 101]. There are several commercially available non-toxic fat emulsions including Intralipid 10% and 20% (Kabi Vitrum) and Lipofundin (Braun). These fat emulsions containing long chain triglycerides (LCT, Intralipid ) are mixtures, made of soybean or egg lecithin and soybean oil in an aqueous solution containing glycerol [6]. Intralipid is available in three different compositions, all containing about 12g/L of phospholipid emulsified either in 100g/L triacylglycerol (10%); 200g/L (20 %) or 300 g/l (30%) [101]. Intralipid emulsions are used for IV nutrition in critically ill patients including infants [6], surgical patients [69], patients with acute renal failure [44] and patients with traumatic brain injury [103]. No adverse clinical effects and no evident toxicity were apparent in patients receiving IV Intralipid and Lipofundin [6, 7, 38, 44, 69, 103]. 15

33 VI. AMB and lipid emulsions. The relatively low toxicity of lipid emulsions together with the extensive use in total parenteral nutrition has made the lipid emulsions an attractive choice as a carrier system for lipophilic drugs [38, 39]. Several investigators mixed Fungizone with the lipid emulsions prior to infusion to the serious ill patients with neutropenia or HIV patients, without appropriately characterizing the formulations in terms of their structure and stability [29, 33, 87]. In several experimental models lipid emulsion mixed with AMB prior to injections improved tolerability and decreased toxicity when compared with Fungizone [70, 76, 120]. However, it was demonstrated that these mixtures are not stable [130]: they spontaneously separated into a yellowish precipitate and a white floating emulsion. VII. Characterization and in vitro activity of our formulation of AMB and Intralipid Based on indications that such formulations could be stabilized, we performed a study [117] where AMB-Intralipid (AMB-IL) admixtures have been characterized physically and chemically, stability under different conditions assessed, in vitro sensitivity of several Candida spp evaluated in comparison to AMB, and in-vitro toxicity to mammalian cells tested. We prepared AMB-IL admixtures by vigorous agitation of Fungizone with three different lipid emulsions and showed that the our formulations remained 16

34 homogeneous for at least 1 month, both at 4 C and at 24 C. Centrifugation of the various emulsions always resulted in floatation of a lipidic cream. Chemical analysis revealed that in the emulsion made by the extended agitation most of the AMB was associated with this lipidic phase, whereas in the preparations made by gentle shaking, most of the AMB precipitated and only a quarter of the AMB was lipid-associated. Hence, prolonged agitation of Fungizone with lipid emulsions increased markedly and irreversibly the association of AMB with emulsion particles. The in vitro antifungal activity of the AMB emulsions admixtures against various Candida spp was compared to that of Fungizone. The minimal inhibitory concentration (MIC) of all the admixtures was similar to that of Fungizone. Two weeks of storage at either 24 C or 4 C decreased the efficacy of the preparations made by gentle mixing whereas the efficacy of the preparations made by prolonged agitation was retained. The AMB-IL preparations were less toxic to human erythrocytes than Fungizone : whereas 4-5µg/ml Fungizone caused complete hemolysis, none of the emulsion-based formulations caused hemolysis at AMB concentrations of up to 80µ g/ml. Furthermore, 1.5µg Fungizone /ml induced considerable K + -efflux from RBC, whereas AMB-lipid emulsions had essentially no effect on the K + leakage at AMB concentration of up to 100µg/ml. In conclusion, the AMB-containing lipid emulsions made by 18 hours of agitation may be advantageous in clinical practice as they are efficient, stable, nontoxic and can be produced at low cost from commercially available ingredients approved for clinical use [117]. Furthermore, the reduced acute toxicity of the drug permits administration of higher, more effective doses. 17

35 The results obtained in our previous in vitro studies and overview of literature led us to more elaborate studies, focusing particularly on evaluation of efficacy of the AMB-IL preparations in vivo against experimental murine systemic candidiasis. Furthermore, we tried to explain the reduction of the toxic effects of AMB-Il by evaluation of availability and biodistribution of AMB-Il and by the immune response elicited by it. 18

36 AIMS OF THE STUDY The main goal of this study is the evaluation of AMB-IL admixtures in vivo against experimental murine systemic candidiasis. The specific aims include: 1. Evaluation of in vitro activity of AMB-IL admixtures against Candida species that were not tested in our previous study and against species known as resistant to AMB. 2. Assessment of the in vivo activity of AMB-IL admixtures against experimental systemic candidiasis induced by C.albicans in naive (non-compromised) animals. 3. Assessment of the in vivo activity of AMB-IL admixtures against experimental systemic candidiasis induced by C. albicans and non-albicans species in compromised animals. 4. Analysis of biodistribution and pharmacokinetics of AMB administered as AMB-IL admixtures. 5. Analysis of expression of cytokines as a result of AMB-IL therapy. 19

37 MATERIAL AND METHODS Antifungal compounds Fungizone a yellow, fluffy powder, containing Amphotericin units (50 mg) and sodium desoxycholate 41 mg was a product of Bristol-Meyers Squibb Pharmaceuticals Ltd., Dublin, Ireland. AMB for IV injection was prepared according to the instructions of manufacturers and diluted prior injection with 5% dextrose (d-glucose, FRUTAROM, Laboratory Chemicals, Israel). AmBisome (liposomal AMB) (NeXstar Pharmaceuticals Ltd. Blackrock, County Dublin, Ireland) was prepared according to the instructions of manufacturers and diluted prior injection with 5% dextrose. Fat emulsion Intralipid 20% is a fat emulsions for IV use, made by Kabi Pharmacia, Stockholm, Sweden. Organisms Several species of Candida and several isolates per species were used throughout the study. C.albicans CBS 562 (ATCC 18804), which is the type species strain, was obtained from the Centraalbureau of Schimmelcultures, Delft, Netherlands; C.glabrata was obtained from Hy-Labs (Hy Laboratories Ltd., Rehovot, Israel) and C.tropicalis was obtained from Dr F. Odds s collection. 20

38 Growth of fungi Fungi were stored in 70 0 C in yeast extract broth (Difco Laboratories, Detroit, Michigan, USA) containing 10% glycerol (Merck, Darmstadt, Germany). Prior to the experiments inocula were transferred to Sabouraud-dextrose agar plates (Difco), and grown for 48 hrs at 28 0 C. After that, inoculum from the plates was transferred to a yeast extract broth and incubated overnight at 37 0 C. The fungal cells were harvested by centrifugation, washed three times with sterile saline (0.9% NaCl) and counted in a Neubauer chamber. For final inoculum cells were diluted in sterile PBS. Preparation of AMB-Intralipid admixtures A stock solution of standard Fungizone (5 mg/ml) was prepared in 10 ml 5% dextrose. If uncontaminated this solution may be stored at -70 C for long periods [118]. After being thawed and diluted in 5% dextrose it may be used for 1 week if the solutions are stored in the dark at 4 C. AMB-Intralipid (AMB-IL was prepared by a 25 fold dilution of the stock Fungizone in the lipid emulsions Intralipid 20% [117]. The final concentration of AMB in these preparations was 0.2 mg/ml. These preparations were agitated vigorously at 24 o C for 18 hrs on an Orbit Environmental Shaker, Lab Line (orbital diameter=4 cm) at 280 rpm. In vitro Activity of AMB-IL (Partially published-ref. 117) Efficacy of the AMB-formulation was determined in terms of their MIC, i.e. the lowest concentration of drug preventing growth of fungal colonies, using the Agar dilution technique [95, 118]. The MIC values of AMB either in micellar form 21

39 (Fungizone ) or in lipid emulsions (AMB-IL) were determined for different Candida species: C.albicans (strains CBS 562 and MAS), C. tropicalis, C.glabrata and C.parapsilosis. AMB-IL was stored at 4 C and 24 C, and MIC was determined immediately after preparation and after 2 weeks of storage. Preparation of inocula The test organisms: C. albicans (CBS 562 and MAS), C. tropicalis, C. glabrata and C. parapsilosis, were grown for 48 hours at 28 C on Sabouraud agar plates [118], transferred to yeast extract solution and incubated overnight at 37 C. Fungal suspensions were then prepared in sterile saline to provide a starting inoculum of 1-5x10 5 yeasts/ml, determined by microscopic counts in a Neubauer chamber. Drug dilution A stock Fungizone solution (5 mg/ml) was diluted (two fold dilutions) in 5% dextrose to AMB final concentrations ranging between 200 µg/ml to 4 µg/ml. AMB- IL formulations containing 2 mg/ml AMB in 10 ml lipid emulsions (0.4 ml stock Fungizone plus 9.6 ml lipid emulsions) were also diluted serially in 5% dextrose to same AMB concentrations (200 µg/ml to 4 µg/ml). Performance of test Appropriate dilutions of the drug were made in sterile 5% dextrose to a final volume of 1 ml. 0.1 ml of drug solution was added to bottles with 100 ml of sterile molted YNB. The final concentrations of AMB in bottles were 20µg/ml to 0.04µg/ml. The agar was dispensed into sterile petri dishes and yeasts were added up to final inoculum of 1-5x10 4 yeasts/ml. Some of the yeast suspension was added to a plate 22

40 without drugs to serve as control. The test included a pair of drug-free controls; one contained the solubilizing solution for the drug and one was a growth control. The inoculated plates were incubated 48 hrs at 28ºC. The MIC was defined as the lowest concentration of the drug inhibiting growth of colonies. Pinpoint colonies were not considered as an indication of growth [117]. II. Normal (naïve) animal model (Partially published-ref. 114) Animals Female 4 weeks old ICR mice, mean weight 30 grams, were used in this study. Animals were kept under conventional conditions and were given food and water ad libidum. All preparations were administered IV into the tail vein. The ethics committee of the Faculty of Medicine granted permission for the animal experiments described in the work. Animal model Experimental systemic candidiasis obtained by IV inoculation of mice is a wellrecognized, often used model [61, 79, 104, 106, 111], with which our laboratory has previous experience [79, 104, 111, 112]. It is desirable to use inocula that will result in an infection which kill untreated control mice within 7-14 days [61, 106]. Previous experiments in our laboratory with C. albicans CBS 562 indicated that this type of model could be achieved by a fungal inoculum (prepared as described above) of 1x10 4-5x10 4 yeasts/mouse (determined by microscopic counts in a hemocytometer) [79, 111, 112]. In the present study murine systemic candidiasis was obtained by IV 23

41 inoculation of C.albicans CBS 562 with 5x10 4 yeasts/mouse (0.2ml volume). Infection was followed up for 6 weeks and evaluated by mortality and morbidity. Mortality Mortality data was evaluated in terms of mortality rate and mean time to death during a follow up period of up to 42 days. Morbidity Morbidity was assessed by fungal colonization of viscera. The kidney is a target organ in systemic candidiasis, so special emphasis was given to this organ. Fungal colonization was demonstrated by histopathology of tissue sections from the organs removed from mice sacrificed at various time intervals post fungal inoculation. Isolation of Candida colony forming units (CFU) from tissue homogenates of the organs of the infected animals provides a quantitative determination of colonization at various stages of the infection, starting from the early stages post fungal inoculation. Determination of Fungal Elements in Tissues Fungal colonization was demonstrated qualitatively by histopathology of tissue sections from the kidneys removed from mice sacrificed at various time intervals post fungal inoculation. Tissues were fixed in 10% formalin (Merck) (at least 24 hours), embedded in paraffin, sections cut and stained with Periodic Acid Schiff (PAS) for light microscopy. In addition, unstained or calcoflour (Sigma) stained tissue homogenates were also used for direct examination under light or fluorescent microscopes for detection of fungal elements. (Calcoflour is a flurochrom with 24

42 affinity for chitin and used therefore as a method for demonstration of fungal elements). Candida CFU enumeration Isolation of Candida colony forming units (CFU) from tissue homogenates of the organs of the infected animals provided a quantitative determination of colonization. Viscera (kidneys and spleen) of sacrificed animals were excised aseptically and homogenized in 1 ml sterile saline using a homogenizer (Kinematica GmbH, Littau- Luzern, Drehzahlregler, Switzerland). The homogenates were diluted in phosphatebuffered saline (PBS) and plated on Sabouraud agar (Difco). Culture plates were incubated for 48 hrs at 28 0 C and the enumerated colonies were expressed as CFU/ml/organ. Treatment of normal mice with AMB formulation Infected mice were treated based on the report of Polachek et al [78] with either Fungizone (as control) or AMB-IL at different doses (ranging from 0.4mg/kg to 2mg/kg). Treatment began 48hrs after inoculation of Candida and consisted of 5 consecutive daily injections of AMB (Fungizone or AMB-IL). A control group with infected and sham-treated mice (injected with buffer-pbs) was included. Survival was followed up for 42 days. Assessment of activity of the AMB-IL was based on the parameters described above for characterization of systemic candidiasis model and evaluated by mortality and/or morbidity in comparison to untreated animals and to animals treated with Fungizone (expressed as survival rate and mean survival time- MST) 25

43 Evaluation of In Vivo Toxicity of AMB-IL Acute toxicity was determined for AMB-IL in comparison to Fungizone, in Candida infected animals. Mice were injected during 5 consecutive days through the tail vein with various doses of AMB (from 0.4mg/kg to 2mg/kg) either as Fungizone or AMB-IL and surveyed for mortality. III. Compromised animal model (Partially published-ref. 115) Induction of a compromised state Four weeks old female ICR mice were used in all experiments. A compromised state was induced as described by Atkinson et al [4] and in a previous study of our laboratory [111], by intraperitoneal (ip) injection of the immunosuppressive agent cyclophosphamid (CY) at a dosage of 200mg/kg. Animals were monitored on various days post treatment by determination of total blood leukocyte count (WBC) and by body weight. A decrease in the number of WBC and reduction in the body weight were indications for a compromised state. Induction of experimental candidiasis Experimental systemic candidiasis was induced in CY compromised mice by IV inoculation of 0.2 ml of a suspension of C.albicans and the non-albicans species C.glabrata and C.tropicalis into the tail vein, similar to the procedure used in the naive mice [114]. C.albicans and the non-albicans species were injected at different concentrations for each species, as determined to be optimal for causing systemic candidiasis leading to death within 5 to 17 days [61, 106]. This type of model is best suitable for assessment of the efficacy of antimicrobial drugs. 26

44 Animals were inoculated IV with the fungi on the day that the debilitated state was demonstrated by the most prominent decrease in the number of WBC and loss of body weight (generally, day 4 after pretreatment with CY). Evaluation of systemic candidiasis was performed during a follow-up period of 42 days using similar criteria as in the naive animal model [114]. Specifically, criteria included assessment of morbidity and mortality, as assessed by percentage of dead mice at the end of the follow-up period and by the mean survival time (MST). Morbidity was assessed by determination of fungal colonization in kidneys by macroscopical observation and by microscopical examination of kidney homogenates. Treatment of candidiasis with AMB formulations in compromised mice. Infected mice were treated based on the similar protocol as in the animal model of naive mice [114] with either Fungizone (as control) or AMB-IL at different doses (ranging from 0.4mg/kg to 2mg/kg). Treatment began 48hrs after fungal inoculation and consisted of 5 consecutive daily IV injections of AMB (Fungizone or AMB-IL). A control group with infected and sham treated mice (injected with buffer-pbs) was included. Survival was followed-up for 42 days. Assessment of activity of the AMB- IL was based on the parameters described above for characterization of the systemic candidiasis model and evaluated by mortality in comparison to untreated animals and to animals treated with Fungizone (expressed as survival rate and MST). IV. Bioavailability and pharmacokinetics of AMB-IL (Partially published-ref. 116) Drugs and drug preparations A stock solution of Fungizone was prepared in a solution of 5% dextrose. AMB-IL admixture was prepared as described in a previous study [117] by a 25-fold dilution of 27

45 stock solution of Fungizone in Intralipid 20% to an AMB final concentration of 0.2 mg/ml and then agitated vigorously at 24ºC for 18 hrs on a controlled environment incubator shaker, oscillation rate of 280 rpm. AmBisome (liposomal AMB) was prepared for use according to the instructions of manufacturers and diluted prior injection with 5% dextrose. Animals Female 4 weeks old ICR mice, mean weight 30 grams, were used in this study. Animals were kept under conventional conditions and were given food and water ad libidum. All preparations were administered IV into the tail vein at doses of 1mg/kg. AMB concentrations in serum and organs were determined using HPLC [105]. Determination of AMB concentration Blood samples were obtained from orbital vein at 5, 15, 30minutes, 1, 2, 4, 6, 12, 24, 48, 72 hrs and 10 and 14 days after drug injection. Blood was pooled from two mice, separated by centrifugation, and serum was used for HPLC determination. Mice were then sacrificed by cervical dislocation, and kidneys, liver, spleen, lungs and heart were removed, washed with PBS, weighed, dried with filter paper, and kept frozen in 20 0 C until extraction. We performed two independent experiments, each including two mice for time point. Extraction of AMB from blood AMB was extracted from serum with methanol in mixtures of serum and methanol in a 1:2 ratio. Samples were mixed by vortexing for 20 s; for completion of extraction and protein precipitation, the mixtures were stored at room temperature for 1hr. The 28

46 mixtures were centrifuged at g for 12 min. The supernatants were used for the HPLC assay. Extraction of AMB from organs Organs were homogenized with cold methanol using the ratio of 1-gr. tissue for 3 ml of methanol. Homogenates were stored at RT for 2hr for complete extraction and then centrifuged at g for 12 min. AMB concentration was determined in the supernatants by HPLC. HPLC assay The level of AMB in the blood and tissue extracts was determined by a reverse phase HPLC technique using the Waters HPLC system: Waters 600 Controller, Waters 996 Photodiode Array Detector and Waters 616 Pump. Separation was obtained using Econosphere C x4.6mm column (3 µm particle size) plus guard column (Alltech, Defiled, IL) and a mobile phase of acetonitrile (Merck): 0.02M EDTA (Merck) ph 4.5 in a volume ratio 55:45, based on the method described by Barenholz & Cohen [8]. An injection volume of 25 µl was used with a flow rate of 1ml/min. AMB was eluted at 1.5 min and was detected at 405 nm using Water s spectrophotometer. Quantitation was done by measurement of the peak area calculated by linear regression. Concentrations in samples were calculated by interpolation from the calibration curve using the Pearson s correlation of co-efficient. 29

47 V. Cytokines expression We performed RT-PCR assays to determine the level of different cytokines: the proinflammatory cytokines TNF-ά and IL1-β and the non-inflammatory cytokines IL-2 and IL-6, in animals treated either with conventional AMB, commercially available liposomal AMB-AmBisome or AMB-IL. Animals Female 4-week-old ICR mice were used. Animals were kept under conventional conditions and were given food and water ad libidum. All preparations were administered IV into the tail vein at doses of 1 mg/kg. Drugs Fungizone was prepared in a solution of 5% dextrose. AMB-IL admixtures were prepared as described previously [117]. AmBisome (liposomal AMB) was prepared according to the instructions of the manufacturers and diluted prior to injection with 5% dextrose. Total RNA isolation Spleens were removed from sacrificed normal and compromised mice 48 hrs after drug administration, immediately frozen in liquid nitrogen and stored at 70 C until further processing. Total RNAs were isolated from spleen samples using TRI REAGENT according the recommended protocol of the manufacturer (TRI REAGENT, M.R.C. TALRON). Tissue explants were homogenized with TRI reagent 30

48 solution (1ml per mg of tissue), mixed with adjusted volume of chloroform (MERCK) (0.2 ml per 1ml of initial TRI-Reagent), and then centrifuged 15 min at 13000g. The upper aqueous phase was transferred and mixed with adjusted aliquot of isopropanol (MERCK) (0.5ml per 1ml of initial Tri-Reagent volume) for further RNA precipitation, and then centrifuged for 10 min at 12000g. The obtained pellet was washed with 75% ethanol (MERCK), centrifuged for 5 min at 10500g, 15 min airdried and then dissolved in 100µl of DEPC (diethyl pyrocarbonate) treated water (Biological Industry, Beit Ha-Emek, Israel), followed by incubation at 60ºC for min and stored on ice. The RNA concentrations in the samples were determined by measuring the optical density (OD) at 260 nm (1 OD at 260 nm equals to 40 µg RNA/ml). Purity of RNA was determined by ratio of OD values at 260 to 280. Only samples at 260:280 OD ratios of have been included in the analysis. The samples were stored in 70 C before further use. RT-PCR system RT-PCR analysis of the mrna was performed using one step RT-PCR system (Titan One Tube RT-PCR System, Roche Molecular Biochemicals) according to the recommended protocol of the manufacturer. RNA samples (1mg of total RNA) were amplified using specific mouse cytokine primers (Bio Technology General Ltd, Rehovot, Israel) for the following cytokines: TNF-α, IL-1β, IL-2, IL-6, and β-actin as a control. 31

49 Type of Primer Sequence Yielding product TNF-α FW (5 to 3 ): TCT CAT CAG TTC TAT GGC CC 274 bp RV (5 TO 3): GGG AGT AGA CAA GGT ACA AC IL-1β FW (5 to 3 ): AAG TTC TCC CCT TTC CTG GTT GTC A 234 bp RV (5 TO 3): TGG TGG GCA CAA ACT TGT CCT TCA C IL-2 FW (5 to 3 ): GAC ACT TGT GCT CCT TGT CA 280 bp RV (5 TO 3): TCA ATT CTG TGG CCT GCT TG IL-6 FW (5 to 3 ): GTT CTC TGG GAA ATC GTG GA 310 bp RV (5 TO 3): TGT ACT CCA GGT AGC TAT GG β -ACTIN FW (5 to 3 ): ATG GAT GAC GAT ATC GCT RV (5 TO 3): ATG AGG TAG TCT GGTC AGG T 603 bs Thermocycling was performed in Perkin Elmer GeneAmp 9600 Thermocycler, accordingly to the recommended protocol of the manufacturer (Titan One Tube RT- PCR System Roche), and accordingly to annealing time of primers. Thermocycling Procedures temperature ºC Time Number of Cycles Reverse transcription at min 1 Denaturate template at 94 2 min 1 Denaturation at sec Annealing at 56 Elongation at 68 Denaturation at 94 Annealing at 56 Elongation at sec 2 min 30 sec 30 sec 2 min* x 10 x 25 Prolonged elongation at 68 7 min 1 4 Pause *elongation at 68ºC of 2 min + cycle elongation of 5 sec for each cycle 32

50 Separation of the PCR Products After thermocycling, the samples were analyzed on a 2% agarose gel (Gibco BRL, Lita Technologies, Paisley, Scotland) containing ethidium bromide 10ng/ml (Sigma), using the dye bromophenol blue as the front-runner. The DNA molecular size markers ladder was used (ΦΧ 174 DNA/BsuRI Marker, Fermentas). The bands were photographed and further analyzed semiquantitatively and qualitatively in comparison to rat β-actin expression (β-actin a house keeping gene). To compare between results we used the TINA program (Raytest Isotopenme geriilte, GmbH) based on determination of intensity of bands of the image, using intensity of β-actin as a control (100% expression). The intensity is normally measured in units of PSL (photo stimulated luminescence) and defined by the instrument manufacturer. 33

51 VI. Statistical Analysis. Statistical analysis will be carried out with the assistance of the Laboratory of Statistics of the Tel-Aviv University. The MIC values were analyzed by independent two population Student s t-test The influence of the inoculum concentration was determined by Log Rank test The MST data of each group of treated mice was compared to those from untreated controls using the one-way analysis of variance (ANOVA) test. The data of the end survival rates was analyzed by the χ 2 test Confidence levels of at least p<0.05 will be considered significant 34

52 RESULTS General In our recent study (described in my M.Sc thesis work) we have demonstrated that admixtures of Fungizone and lipid emulsions used clinically for parenteral nutrition, such as Intralipid and Lipofundin prepared under vigorous and prolonged agitation constitute stable preparations. These preparations were characterized regarding chemical and physical properties. They exhibited in vitro antifungal activity towards several pathogenic Candida species comparable to that of AMB, but were significantly less toxic to human RBC. Based on these data, in the present study we assessed antifungal activity of AMB-Intralipid emulsion in vivo in two experimental animal models: non-compromised, naïve animals and compromised animals. Our study included: Development of experimental murine candidiasis: in normal mice in compromised mice Treatment of animals with AMB and AMB-IL at various doses Evaluation of efficacy of treatment in regard to morbidity and mortality In vivo toxicity Bioavailability and pharmacokinetics of the AMB-IL in mice Immune response (cytokine expression) 35

53 At the first step of our study prior to the in vivo investigations, we decided to evaluate and enlarge the study on the in vitro antifungal activity of AMB-IL preparation including testing of C. albicans and non-albicans species. I. In vitro Activity of AMB-IL. Based on the in vitro data obtained in our previous study [117] and in order to assess further the validity of our findings, we enlarged the tested sample, increasing both the number of isolates per species and adding additional nonalbicans species. Table 1 summarizes the data of assessment of the in vitro activity of AMB- IL and Fungizone against 16 Candida isolates of 7 species: C.albicans (4 strains), C.tropicalis (4 strains), C.glabrata (3 strains), C.krusei (2 strains), C.parapsilosis (one strain), C.lusitaniae (one strain), C.rugosa (one strain). The tests revealed that the mean minimal inhibitory concentration (MIC) values for AMB-IL against most of the tested strains of Candida were comparable with those of Fungizone. Moreover, the activity of AMB-IL against some of the tested Candida such as C.krusei, C.rugosa and C.lusitaniae was increased (decreased MIC). These results with the non-albicans species are particular significant in view of the problems associated with the treatment of mycoses caused by these fungi. 36

54 Table 1. MIC values of Candida strains Strain AMB (µg/ml±sd) AMB-IL (µg/ml±sd) 1. C.albicans CBS ± ± C.albicans MAS 0.3± ± C.albicans CBS ± ± C.albicans 116/ ± ± C.tropicalis (clinical isolate) 0.68± ± C.tropicalis (clinical isolate) 0.78± ± C.tropicalis 75/ ± ± C.tropicalis 73/ ± ± C.glabrata (clinical isolate) 0.35± ± C.glabrata ± ± C.glabrata ± ± C.krusei (clinical isolate) 1.25± ±0.02* 13. C.krusei 74/ ± ±0.02* 14. C.parapsilosis (clinical isolate) 0.68± ± C.lusitaniae (clinical isolate) 0.47± ±0.02* 16. C.rugosa HD-CL8 1.25± ±0.02* *MIC values are significantly different (p<0.05). Data represent mean values from 4 independent experiments, each consisted of duplicate samples tested for each strain. 37

55 II. Normal (naïve) animal model Experimental Animal Model As the main objective of this study was to evaluate the AMB-IL preparations in vivo, the initial experiments were devoted to the adaptation of a suitable animal model. Experimental murine candidiasis obtained by direct injection of the fungi into the blood stream (by IV inoculation) is a well-recognized, often used model. Our model was based on previous research in our laboratory with C.albicans CBS 562 by injecting mice IV into the tail vein. It was aimed to induce a lethal infection with not a too rapid course of mortality, starting within 7 to 10 days post fungal inoculation. Our experience indicated that this type of model could be achieved by a fungal inoculum of 5x10 4 organisms (counted microscopically) per mouse. The experiments revealed that using 4-6 weeks old female ICR mice (20 mice) injected with 5x10 4 organisms/mouse resulted in 100% of mortality within 5-10 days. Systemic candidiasis was determined by demonstration of fungal elements in visceral organs: kidneys, liver and spleen. The organs were observed macroscopically for pathological signs before further processing. In these organs micro-abscesses were found (Picture 1). Presence of fungal elements-hyphae and yeasts was shown by calcoflour stained tissue homogenates from organs removed from sacrificed mice and by histopathological observations in PAS stained tissue sections. Picture 2 shows kidneys homogenates stained by calcoflour, revealing fungal elements, including hyphae and budding yeasts. 38

56 Picture 1. Infected visceral organs from C.albicans infected normal mice A. Kidneys B.Liver C. Spleen 39

57 Picture 2. Calcofluor stained kidney homogenates from C.albicans infected normal mouse yeasts hyphae Magnification: x

58 Sections of kidneys stained by PAS, showed large fungal lesions (Picture 3A), containing yeast cells and hyphal elements (picture 3B). Picture 3. Images of Pas stained paraffin sections of kidney from C.albicans infected normal mouse A. Large fungal lesions Magnification: x 10 B. Yeast cells and hyphae Magnification: x

59 Quantitative organ colonization by the fungus was determined by enumeration of candidal colony forming unit (CFU) in tissue homogenates. In order to determine a possible quantitative correlation between fungal colonization of the kidney (a target organ in systemic candidiasis), and death of the animals, we enumerated CFU in kidneys of 20 mice on the day of death. We found that upon the death of the animals their kidneys are colonized with C.albicans with over 10 6 yeasts/kidney (range 1.6x10 6-3x10 7 ; the mean number was 8.6±0.86x10 6 ). Treatment of Systemic Murine Candidiasis with Low Doses of AMB Following the adaptation of a suitable animal model we initiated the treatment experiments. The experiments involved naive ICR mice inoculated IV with 5x10 4 C.albicans cells and treated 48 hrs later with either AMB-IL or Fungizone at a concentration of 0.4mg AMB/kg in comparison to sham (buffer) treated controls. Treatment was administered during 5 consecutive days by IV injections of the drug. Animals were surveyed up to 42 days as described above. Data were obtained from 4 experiments with 108 mice and are summarized in Fig The data show that both formulations (Fungizone and AMB-IL) significantly increase the survival of the mice as compared with the sham treated controls. Furthermore over 50% of mice survived when treated with AMB-IL (Fig. 1). The mean percentage of surviving mice at day 42 was 0, 24 and 52 for the untreated, Fungizone and AMB-IL treated groups, respectively (Fig. 1). The follow-up of the course of infection as shown in Fig. 2 indicates that the Fungizone and AMB-IL increase the survival time of the treated mice. Thus, the mean survival time (MST) in those mice that succumbed to infection, as shown in Fig. 3, indicates an MST of 7.38±0.57; 25±1.77 and 30±2.09 days for the untreated control group, Fungizone and AMB-IL groups, respectively. 42

60 Fig ure 1. M ean survival rate (% ) at day 42 of norm al m ice treated w ith lo w d o se o f A M B fo rm u lation s % survival control group Fungizone group AM B-IL group n=36 n=36 n=36

61 F igure 2. C ourse of infection in norm al m ice treated w ith low doses of A M B form ulations % survival d a y s K e y : control, C. a l b i c a n s (C.A ) C. A + A M B C. A + A M B - I L

62 F igure 3. M e a n su rv iv a l tim e (fo llo w u p 4 2 d a ys) o f n o rm a l m ic e tre a te d w ith low dose of AM B form ulations Days control group Fungizone group AM B-IL group n=36 n=36 n=36

63 The data obtained in these experiments showed that AMB-IL is more effective than Fungizone in the treatment of murine candidiasis, demonstrating a significant difference between the survival rate of Fungizone treated vs. AMB-IL treated animals. However, the AMB-IL treatment did not save almost half of the animals from the experimental candidiasis. Treatment of Systemic Murine Candidiasis with High Doses of AMB In order to improve the efficacy of the AMB-IL preparations we planned experiments with higher doses of AMB. Towards this aim we first had to establish the maximal doses of AMB to which the mice would be tolerant. We found that the maximum tolerated dose was 1 mg/kg/day x 5 for Fungizone and 2 mg/kg/day x 5 for AMB-IL (Table 2). Higher doses of Fungizone (1.2 mg/kg) caused immediate death in 100% of the mice. AMB-IL at 10mg/kg (total) did not cause death during an observation period of six weeks. Table 2. Survival of mice (at day 42), following IV administration of Fungizone or AMB-IL Dose (mg/kg x 5) Fungizone No. dead/injected 0/8 0/4 0/4 0/8 8/8 n.d n.d n.d n.d AMB-IL No. dead/injected 0/8 0/4 0/4 0/8 0/8 0/4 0/4 0/4 0/8 46

64 Based on these results we then performed 4 additional treatment experiments with 104 mice, using the same animal model, but with higher doses of AMB-IL. We had 4 groups in each experiment: sham (buffer) treated controls, Fungizone treated group (1mg AMB/kg/day x 5) and 2 groups treated with AMB-IL (1mg AMB/kg x 5 and 2mg AMB/kg/day x 5). AMB-IL at the higher concentration of 2mg AMB/kg/day x 5 significantly increased the survival of the mice, almost up to 100% (Fig. 4). In addition, AMB-IL preparations at the higher dose prolonged the survival time (Fig. 5 and Fig. 6). Hence, data obtained at this stage (partially published-ref. 114) showed that AMB-IL preparations at the higher doses are more efficient in the treatment of experimental systemic candidiasis in mice. 47

65 F igure 4. M ean survival rate (% ) at day 42 of norm al m ice treated w ith high doses of A M B form ulations % survival C ontrol group Fungizone (1m g/kg x 5) AM B -IL(1m g/kg x 5) AM B -IL(2m g/kg x 5) n=17 n=29 n=29 n=29

66 F igure 5. C ourse of infection in norm al m ice treated w ith high doses of A M B form ulations % survival d a y s K e y : control, C.albicans (C.A ) C.A + A M B ( 1 m g /k g x 5 ) C.A + A M B - I L ( 1 m g /k g x 5 ) C.A + A M B -IL (2 m g /k g x 5 )

67 Figure 6. M ean survival tim e (follow up 42 days) of norm al m ice treated w ith high doses of AM B form ulations Days Control group Fungizone (1m g/kg x 5) AM B-IL(1mg/kg x 5) AM B-IL(2m g/kg x 5) n=17 n=29 n=29 n=29

68 III. Compromised animal model Induction of compromised state. Previous studies in our laboratory involved induction of compromised animals by ip injection of the immunosuppressive agent cyclophosphamide (CY). The compromised state in the experimental animals was most evident 4 days post CY pretreatment, as demonstrated by decreased numbers of white blood cells (WBC) and/or loss of body weight. In the present research, 4-week-old female ICR mice were treated i.p with CY at a dosage of 200mg/kg. The compromised state was determined after various days post treatment by total blood leukocyte count and by measurement of body weight. As shown in Figure 7 A, the treatment resulted in a decrease in the number of WBC, with the lowest level reached on day 4 (from 8x10 6 /ml to 9.1x10 5 /ml), after which the count gradually increased. Also, the greatest decrease of body weight from 26.2 gr. to 21.6gr was observed on day 4-post treatment (Figure 7 B). All further experiments were performed following 4 days of CY pretreatment, at which time WBC counts were at their lowest and weight loss was maximal. 51

69 Figure 7. Effect of CY pretreatment on mice. 16 A. WBC count WBC x10 6 /ml B. Body weight 32 gr days

70 Animal model of systemic candidiasis in compromised mice. Systemic candidiasis was induced in CY compromised female ICR mice (6 weeks old) by IV inoculation of C.albicans and non-albicans species (C.glabrata and C.tropicalis) into the tail vein, similar to the procedure used in the naive mice [114]. C.albicans and non-albicans species were injected at a different concentration for each species, as determined to be optimal for causing systemic candidiasis leading to death within 5 to 17 days (table 3). Table 3. Inoculum of fungi that caused 100% mortality Candida species Inoculum (organisms/mouse) C.albicans 5x10 4 C.tropicalis 5x10 5 C.glabrata 5x10 6 Compromised animals were inoculated with the fungi following 4 days of CY pretreatment. Evaluation of systemic candidiasis was performed during a follow-up period of 42 days using criteria similar to those as for the naive animal model. Criteria included assessment of morbidity and mortality. Macroscopic observation of kidneys removed from compromised mice inoculated with C.albicans or C.tropicalis or C.glabrata revealed numerous yeasts colonies (Picture 4). Sections of kidneys stained by PAS showed yeast cells and hyphal elements (Picture 5). Furthermore, as an in normal mouse, presence of fungal elements-hyphae and budding yeasts in the calcoflour kidneys homogenates also indicated that systemic candidiasis had occurred. 53

71 Picture 4. Kidneys from CY compromised mouse infected with Candida spp A. Infected with C.albicans B. Infected with C.glabrata C. Infected with C.tropicalis 54 lesions

72 Picture 5. Pas stained section of kidney from C.albicans infected CY compromised mouse Yeast cells hyphae Magnification: x

73 As an initial step in the studies with compromised mice it was necessary to establish the effect of CY pretreatment on the mortality of animals inoculated with C.albicans. As can be seen from Table 4, CY pretreatment increased the mortality rate elicited by C.albicans. For example, inoculum of 5x10 4 /mouse caused systemic candidiasis and led to death in compromised animals within a very short period-4 day. Therefore, the optimal inoculum for C.albicans was determined as 1x10 4 /mouse. This concentration caused 100% mortality within 7-16 days (MST was 9.36± 0.4). In addition to C.albicans infection, we attempted to induce murine systemic candidiasis by inoculating mice with non-albicans species: C.glabrata, C.tropicalis, C.krusei and C.lusitaniae. However, as shown in Table 4, all our attempts to elicit systemic candidiasis with non- albicans species in naive mice failed. Only in CY compromised animals we succeeded to induce systemic candidiasis with non-albicans species: C.tropicalis and C.glabrata, that resulted in 100% mortality within 5-17 days (Table 4). The experiments revealed that 5x10 5 /mouse of C.tropicalis and 5x10 6 /mouse of C.glabrata are suitable inoculum for induction of systemic candidiasis in CY treated mice. 56

74 Table 4. Murine systemic candidiasis in naive and compromised animals Candida species Normal (naive) mice CY compromised mice and infecting inoculum/mouse % mortality a MST (days) b % mortality MST (days) C.albicans 5x (0/10) c (6/10) x (7/10) (34/34) x (36/36) (10/10) 2.3 C.glabrata 5x (0/5) 42 0 (0/5) 0 1x (0/5) 42 0 (0/5) 0 5x (0/5) 42 0 (0/5) 0 1x (0/5) (4/5) x (0/5) (16/16) 10 1x (0/5) (5/5) 4.6 5x (0/5) 42 N.D d N.D 1x (0/5) 42 N.D N.D 2x e 0 N.D N.D C.tropicalis 5x (0/5) 42 0 (0/5) 0 1x (0/5) (3/5) x (0/5) (14/14) x (0/5) (10/10) 4.1 5x (0/5) 42 N.D N.D 1x (0/5) 42 N.D N.D 5x (0/5) 42 N.D N.D 1x (0/5) 42 N.D N.D 2x e 0 N.D N.D C.krusei & C.lusitaniae 5x (0/5) 42 N.D N.D 1x (0/5) 42 N.D N.D 5x (0/5) 42 N.D N.D 1x (0/5) 42 N.D N.D 5x (0/5) 42 N.D N.D 1x (0/5) 42 N.D N.D 5x (0/5) 42 N.D N.D 1x (0/5) 42 N.D N.D 2x e 0 N.D N.D a % mortality at day 42; b Mean survival time follow up 42 days; c number in parentheses indicate number of dead mice/total; d N.D - not done; e immediate death- during 1 hour. 57

75 Treatment of C.albicans infection in CY compromised mice with low doses of AMB. Following the establishment of an experimental systemic candidiasis model in CY compromised mice treatment experiments with AMB preparations were initiated. The treatment protocol was similar as in the animal model of naive mice [114]. Efficacy of treatment with AMB-IL was evaluated by comparison to that of untreated mice and those treated with Fungizone. The experiments involved CY compromised ICR mice infected IV with 1x10 4 organisms of C.albicans and 48hrs later treated with either Fungizone (as control) or AMB-IL at a concentration of 0.4mgAMB/kgx5 days. This concentration was also used in the model of systemic candidiasis in naive mice. Systemic candidiasis was followed up during 6 weeks and evaluated by mean survival rate and mean survival time (MST) for those animals that succumbed to infection. Data were obtained from 4 experiments with 94 mice and are summarized in Figure 8 A, B, C. The data show that both preparations (AMB and AMB-IL) increased the survival of infected mice as compared to the untreated control group. For example: percentage of survival at day 42 was 0, 13.3±2.4 and 43.3±1.48 for the control, AMB and AMB-IL treated groups, respectively (Figure 8 A). The follow up of the course of the infection (Figure 8 C), shows that these preparations increase the survival time of the treated mice. The MST was 9.36±0.49; 18.73±2.45 and 26.9±0.68 for the control, AMB and AMB-IL groups, respectively (Figure 8 B). These data show that even at low doses, AMB-IL admixtures were more effective than conventional AMB. 58

76 C. Course of infection Days Figure 8. Treatment of C.albicans infection in CY compromised mice with low doses of AMB formulations A. Survival rate (day 42) B. Mean survival time control n=34 AMB n=30 AMB-IL n=30 0 control AMB AMB-IL n=34 n=30 n= day Key: control, C.albicans (C.A) C.A + AMB C.A + AMB-IL % survival % survival 59

77 Treatment of C.albicans infection in CY compromised mice with high doses of AMB. Although treatment with low doses of AMB-IL preparations was more effective than conventional AMB, nevertheless, only 43.3% of thus treated mice survived. At this stage we decided to increase the doses of AMB. Towards this aim we performed additional experiments involving 79 mice using similar animal and treatment models, that included 4 experimental groups: control (untreated); AMB 1mg/kg; AMB-IL 1mg/kg and AMB-IL 2mg/kg. Treatment was administered during 5 consecutive days by IV injection. The previous study [114] in naive mice revealed that the conventional AMB could not be used at doses higher than 1mg/kg, since it caused immediate death in the animals. The data in Figure 9 show that all AMB formulations at high doses increase significantly the survival of the mice. Furthermore, AMB-IL at the higher concentration of 2mg/kg x 5 was very effective, 100% of treated mice survived during the experimental period (Figure 9 A). Hence, obtained data showed that AMB-IL admixtures at a high dose are more efficient in the treatment of experimental C.albicans systemic infection in CY compromised mice, as was the case in the non-compromised animals. 60

78 day Figure 9. Treatment of C.albicans infection in CY compromised mice with high doses of AMB formulations A. Survival rate (day42) B. Mean survival time C. Course of infection control AMB1mg/kg AMB-IL1mg/kg AMB-IL2mg/kg control AMB1mg/kg AMB-IL1mg/kg AMB-IL2mg/kg Key: control, C.albicans (C.A) C.A+ AMB 1mg/kg C.A+ AMB-IL 1mg/kg C.A+ AMB-IL 2mg/kg % survival % survival 61

79 Treatment of C.glabrata and C.tropicalis infections in CY compromised mice with AMB preparations. Following the development of candidiasis by non-albicans species in CY compromised mice we initiated treatment experiments. A total of 127 CY compromised mice were injected with 5x10 5 organisms of C.tropicalis and a total of 130 mice with 5x10 6 organisms of C.glabrata. For the treatment of infections caused by non-albicans species we used similar treatment protocols, as for C.albicans. The data presented in Figures 10-13, which summarize results of experiments with C.tropicalis and C.glabrata infected mice, show that treatment of candidiasis in CY compromised mice with AMB-IL admixtures was more effective than with conventional AMB. As shown in Figures 11 and 13, and as expected, the high doses of AMB were more effective in the treatment of candidiasis in CY compromised mice. Specifically, about 80% survival rate for C.tropicalis and over 90% for the C.glabrata. It is of interest that AMB-IL admixtures were more effective in the treatment of C.glabrata than in C.tropicalis infection. For example, survival rate for C.glabrata infected mice was 0%, 58.3% and 75% for control, AMB and AMB-IL groups respectively, when AMB doses were low (Fig. 10A), while in the case of C.tropicalis infected mice the survival rate was 0%, 27% and 41.7% for control, AMB and AMB- IL groups respectively (Fig. 10A). Furthermore, treatment with higher doses of AMB- IL (total 10mg/kg) of C.glabrata infected mice was significantly more effective than that of C.tropicalis infected mice. Namely, survival rate of mice infected with C.glabrata was 93.3% vs. 76.7% for C.tropicalis infection (Fig. 11A and Fig. 13 A). 62

80 Figure 10. Treatment of C.tropicalis infection in CY compromised mice with low doses of AMB formulations A. Survival rate (day 42) B. Mean survival time C.Course of infection Control AMB AMB-IL Control AMB AMB-IL %survival day %survival 63 Key: control, C.albicans (C.A) C.A+ AMB C.A+ AMB-IL

81 Figure 11 Treatment of C.tropicalis infection in CY compromised mice with high doses of AMB formulations A. Survival rate (day 42) B. Mean survival time C. Course of infection Control AMB 1mg/kg AMB-IL 1mg/kg AMB-IL 2mg.kg Control AMB 1mg/kg AMB-IL 1mg/kg AMB-IL 2mg.kg % survival day % survival Keys: Control; AMB 1mg/kg; AMB-IL 1mg/kg; AMB-IL 2mg/kg

82 Figure 12 Treatment of C.glabrata infection in CY compromised mice with low doses of AMB formulations A. Survival rate (day 42) B. Mean survival time C.Course of infection Control AMB AMB-IL Control AMB AMB-IL %survival Day %survival Key: control, C.albicans (C.A) C.A+ AMB C.A+ AMB-IL

83 Figure 13 Treatment of C. glabrata infection in CY compromised mice with high doses of AMB formulations A. Survival rate (day 42) B. Mean survival time C. Course of infection day % survival Control AMB 1mg/kg AMB-IL 1mg/kg AMB-IL 2mg/kg Control AMB 1mg/kg AMB-IL 1mg/kg AMB-IL 2mg/kg Keys: Control; AMB 1mg/kg; AMB-IL 1mg/kg; AMB-IL 2mg/kg % survival

84 Data of these investigations (partially published-ref. 115) show that also in compromised animals the AMB-IL admixtures had a significant effect on systemic murine candidiasis induced by C.albicans and the non-albicans species, C.glabrata and by C.tropicalis. The latter findings are of particular relevance as the compromised state is in analogy to clinical situations with high risk for development of candidiasis and in view of the increase in the rate of candidiasis due to non-albicans species. IV. Bioavailability and pharmacokinetics of AMB-IL AMB standard curve In the next stage of our study we aimed to compare levels of AMB in blood and various organs following administration of conventional and lipid formulations of AMB. Towards this aim the initial step consisted of creation of a calibration curve of AMB concentration in serum of mice supplemented with various known concentrations of AMB (Fungizone) as determined by HPLC measurements. The obtained calibration curve is shown in Figure

85 Figure 14. Calibration curve of AMB Area Linear Regression for Area: Y = A + B * X Parameter Value A 7517 B R P < Concentration µg/ml

86 AMB concentrations in serum Figure 15 presents the data of AMB concentrations in serum of mice injected IV either with 1mg/kg of Fungizone, AmBisome, or AMB-IL, at various time intervals post injection. The HPLC measurements showed that concentrations of AMB in serum of mice treated with AMB-IL or AmBisome are higher than in those treated with Fungizone (P< 0.05). Furthermore, we found that the levels of AMB in the serum of mice injected with AMB-IL were similar to those of mice treated with AmBisome. The data also reveal that the levels of AMB in serum decreased relatively rapidly, independent of the formulation. However, at 48 hrs no detectable levels of AMB in animals administered with Fungizone were noted, while those treated with AMB-IL or AmBisome revealed presence of AMB up to 72-hrs post treatment. 69

87 Figure 15. AMB concentration in serum Time (hr) after administration Key: Fungizone 1mg/kg AMB-IL 1mg/kg AmBisome 1mg/kg concentration (ug/ml)

88 In addition, table 5 presents regarding data of pharmacokinetic parameters of the different preparations. Table 5. Pharmacokinetics of AMB preparations administered IV to mice. Parameters AMB 1 ķ (h) 2 V d (L) 3 Clearance 4 AUC 5 C max preparations (L h) (mg/l h) (mg/l) AMB 1mg/kg AMB-IL 1mg/kg AmBisome 1mg/kg AMB-IL 2mg/kg ķ = elimination rate constant = LnC 1 -LnC 2 t 2 -t 1 2 V d = volume of distribution = Dose C 0 3 Clearance = ķ V d 4 AUC = area under the plasma drug concentration-time curve = Dose Clearance 5 C max = highest drug concentration observed in blood. 71

89 Levels of AMB in organs In addition to bioavalability measurements of AMB in blood, HPLC measurements of levels of AMB were carried out in extracts of kidneys, liver, spleen, lungs and heart from mice injected IV with Fungizone, AMB-IL or AmBisome at the same time intervals as for determination of AMB levels in blood. The data are summarized accordingly in Figures The measurements in the kidneys (Fig. 16) showed that the highest levels of AMB during all measurement-points were in animals injected with Fungizone, while the lowest values were seen in mice administered with AmBisome. The levels of AMB in animals treated with AMB-IL were lower than in the Fungizone treated, but higher than in the AmBisome treated. It was also noted that the AMB in kidneys of mice treated with lipid formulations was found up to 24 hrs, while in those treated with Fungizone it was detected up to the 72 hrs. Figure 17 presents the determinations of AMB in liver, the organ in which the majority of AMB was found. The pattern is different from that in kidneys. In the liver the highest levels of AMB are in animals receiving lipid formulations, albeit the level of AMB-IL is lower than that of AmBisome. While the peaks of AMB in liver of mice treated with Fungizone appear early after the administration of the drug, the peaks of the lipid formulations are delayed. Figure 18, which demonstrate the AMB levels in spleen, has generally a similar pattern as the liver. The lipid formulations appear at higher levels than Fungizone. Moreover, AMB in animals treated with AmBisome can be detected in spleen even after 96 hrs. In the lungs (Fig. 19) treatment with Fungizone yields higher levels of AMB than the lipid formulations. In the heart (Fig. 20) the highest levels were noted at the first time-interval measured (5 min), which was followed by a gradual decrease in the level of AMB. 72

90 Figure 16. AMB concentration (ug/gr) in the kidneys Time (hrs) after administration key: Fungizone 1mg//kg AMB-IL 1mg/kg AmBisome 1mg/kg Concentration of AMB (ug/gr)

91 Figure 17. AMB concentration (ug/gr) in the liver Time (hrs) after administration Key: Fungizone 1mg//kg AMB-IL 1mg/kg AmBisome 1mg/kg Concentration of AMB (ug/gr)

92 Figure 18. AMB concentration (ug/gr) in the spleen Time (hrs) after administration Key: Fungizone 1mg//kg AMB-IL 1mg/kg AmBisome 1mg/kg Concentration of AMB (ug/gr)

93 Figure 19. AMB concentration (ug/gr) in the lungs Time (hrs) after administration Key: Fungizone 1mg//kg AMB-IL 1mg/kg AmBisome 1mg/kg Concentration of AMB (ug/gr)

94 Figure 20. AMB concentration (ug/gr) in the heart Time (hrs) after administration Key: Fungizone 1mg//kg AMB-IL 1mg/kg AmBisome 1mg/kg Concentration of AMB (ug/gr)

95 Bioavailability of AMB in animals treated with higher doses of AMB-IL As in the course of evaluation of efficacy of AMB-IL it was noted that AMB-IL at higher doses (2 mg/kg x 5) yielded better results, we evaluated the level of AMB in blood and organs of mice treated with AMB-IL at the dose of 2mg/kg. Fungizone cannot be administered IV to mice at higher doses than 1 mg/kg, as this is the maximal non-lethal dose for mice. The results of the experiment with the AMB-IL injected at the dose of 2mg/kg are presented in Table 6. The data indicate that the pattern of bioavailibility of AMB in blood and organs is similar to that obtained in animals receiving 1mg/kg of AMB-IL. However, as expected, the levels are, accordingly, higher (data partially published-ref. 116). 78

96 Table 6. AMB concentration in blood (µg/ml) and organs (µg/gr) after treatment of mice with AMB-IL at the dose of 2mg/kg Mean AMB concentration a ± SD b Time (hours) after administration Tissues (5 min) 0.25 (15 min) 0.5 (30 min) Blood 1.18± ± ± ± ± ± ± ± ± ± c 0 0 Kidneys 1.12± ± ± ± ± ± ± ± ± Liver 7.95± ± ± ± ± ± ± ± ± ± ± ±0.8 0 Spleen 5.93± ± ± ± ± ± ± ± ± ± Lung 1.27± ± ± ± ± ± ± ± Heart 0.71± ± ± ± ± ± ± ± a blood and tissues samples from two mice were pooled b standard deviation from data of 2 experiments c not detectable 79

97 V. Cytokines expression In this stage of the study we investigated the expression of pro-inflammatory cytokines (TNF-α and IL1-β) and non-inflammatory cytokines (IL-2 and IL-6) in animals treated with either conventional AMB, commercially available liposomal AMB-AmBisome or AMB-IL. Total RNA was extracted from spleens and analyzed on a 2% agarose gel (Picture 6). We observed no expression of cytokines in untreated (control) mice whereas β-actin was expressed at moderate levels (Pictures 7). In contrast, treatment with AMB preparations in normal mice, stimulated the production of cytokine mrna. Treatment with conventional AMB induced expression of high levels of TNF-α and IL-1β mrna (Picture 8), however treatment with AMB-IL or AmBisome induced only low-level expression of TNF- α and IL-1β mrna as compared to AMB treated animals. No IL-2 and IL-6 mrna production was identified (Picture 8). We also examined expression of cytokine mrna in CY compromised animals. As shown in Picture 9, no cytokine mrna expression was found in untreated CY compromised animals (only β-actin expression, as in the normal animals). In contrast, in CY mice, treatment with Fungizone induced expression of high levels of TNF- α and IL-1β mrna (Picture 10), whereas AMB-IL treated mice expressed only a low level of IL-1β mrna. Our results show that treatment with AMB increased the production of proinflammatory cytokines in comparison to non-treated control animals. In mice treated with lipid formulations, both AMB-IL and AmBisome, the expression of these cytokines was lower than in AMB treated animals. 80

98 Picture 6. Mouse spleen RNA bp , , Total RNA from normal untreated mice; 2 Total RNA from compromised untreated mice 3 Total RNA from mice treated with AMB; 4 Total RNA from mice treated with AMB-IL; 5 Total RNA from mice treated with AmBisome 81

99 Picture 7. Expression of cytokines in untreated mice (2% agarose gel) bp β-actin TNF-α IL-1β IL-2 IL

100 Picture 8. Expression of cytokines in normal mice treated with AMB preparations A. 2% agarose gel bp β-actin TNF-α IL-6 IL-1β IL-2 β-actin TNF-α IL-6 IL-1β IL-2 β-actin TNF-α IL-6 IL-1β IL AMB AMB-IL AmBisome B. Expression of cytokines (%) relative to expression of β-actin % TNF-α -IL-1β -IL-2 -IL-6 0 AMB AMB-IL AmBisome 83

101 Picture 9. Expression of cytokines in CY untreated mice bp β-actin TNF-α IL-1β IL-2 IL

102 Picture 10. Expression of cytokines in CY compromised mice treated with AMB preparations A. 2% agarose gel bp β-actin TNF-α IL-1β IL-2 IL-6 β-actin TNF-α IL-1β IL-2 IL-6 β-actin TNF-α IL-1β IL-2 IL AMB AMB-IL AmBisome B. Expression of cytokines (%) relative to expression of β-actin % TNF-α -IL-1β AMB AMB-IL AmBisome 85 -IL-6 -IL-2