Clinical Pharmacology: PAUL
S.
LIETMAN,
M.D.,
Ph.D.,
Foscarnet
Baltimore, Mary/and
Foscarnet exerts its antiviral effects via reversible inhibition of viral pblymerases. Pharmacodynamic data indicate that herpesvirus and human immunodeficiency virus replication is inhibited by therapeutically achievable concentrations of foscarnet; however, the concentrations of foscarnet required for such inhibition have been found to vary widely. Pharmacokinetic data indicate that foscarnet is eliminated via the renal route, undergoes negligible metabolism, and appears to be distributed widely from the circulation. However, the available data indicate that the pharmacokinetics of the drug varies among patients and within the individual patient, particularly with regard to plasma drug levels; furthermore, such factors as the intracellular kinetics of the drug have yet to be well characterized. It is thus difficult to formulate optimal dosing regimens on the basis of what is known of foscarnet pharmacodynamics and pharmacokinetics. Nevertheless, dosages that produce clear-cut therapeutic benefits without unacceptable toxicity have been identified in clinical trials of foscarnet in acquired immunodeficiency syndrome (AIDS) patients with cytomegalovirus (CMV) retinitis.
From the Division of Clinical Pharmacology, The Johns Hopkrns Unlverslty School of Medicine, Baltimore, Maryland. Requests for reprints should be addressed to Paul S. Lletman, M.D., Ph.D., Division of Clinical Pharmacology, Osler 527, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, Maryland 21205.
2A-8s
February 14, 1992
The American Journal of Medicine
F
oscarnet (trisodium phosphonoformate) is an antiviral agent with in vitro activity against herpes simplex viruses 1 and 2, varicella zoster virus, and human cytomegalovirus (hCMV), as well as human immunodeficiency virus (HIV) and other viruses [ 11. In principle, the rational use of foscarnet should be based on an understanding of both its pharmacodynamics (the relationship between the concentration of drug at its site of action and the biochemical and biologic effects it produces) and its pharmacokinetics (the concentration achieved at the site of action and/or in the plasma as a function of time as the drug traverses the body). That is, in principle, some knowledge of the pharmacokinetics of foscarnet should allow the identification of rational doses and dosing regimens that achieve goals based on the pharmacodynamics of the drug.
PHARMACODYNAMICS Oberg [l] has compiled an extensive and useful review of data on foscarnet pharmacodynamics. Foscarnet is a pyrophosphate analogue capable of inhibiting the herpesvirus DNA polymerases and the HIV reverse transcriptase. Its potency against these viral enzymes is considerably greater than its potency against host cell polymerases, thus accounting for its selective toxicity for viral replication as contrasted with host cell replication. Foscarnet is a reversible inhibitor of the viral DNA polymerases and HIV reverse transcriptase; consequently, the extent of inhibition of these viral enzymes fluctuates with rises and falls of drug concentrations at the intracellular locations of the enzymes. The reversible inhibition of the target viral polymerases has been shown not to be competitive with the enzyme substrates (i.e., deoxynucleoside triphosphates) or with templates (i.e., DNA or RNA) [l]; therefore, the degree of inhibition of the viral enzymes is not dependent on either intracellular deoxynucleoside triphosphate levels or intracellular template levels. This is in contrast with nucleoside analogues, such as acyclovir, ganciclovir, and zidovudine, the activity of each of which is influenced, sometimes in an advantageous manner and sometimes in a disadvantageous manner, by the intracellular levels of endogenous nucleoside triphosphates.
Volume 92 (suppl 2A)
SYMPOSIUM
At an enzymologic level, the relationship between the concentration of foscarnet adjacent to the viral enzymes and the extent of inhibition of these enzymes can be described by measuring the inhibitory constant (KJ of foscarnet with respect to each enzyme. This value defines the concentration of drug providing 50% inhibition of the relevant enzyme. The Ki for foscarnet and at least one hCMV DNA polymerase has been reported to be approximately 0.3 PM [l], whereas the Ki for foscarnet for several HIV reverse transcriptases has been reported to be 0.1-l pM [l]. The relationship between foscarnet concentration and the inhibition of viral replication can also be defined at a biologic level by measuring viral replication in infected cells in vitro that are exposed to constant drug concentrations over time. Such measurements usually are expressed in terms of an effective or inhibitory dose for 50% inhibition of viral replication (EDs,, or ID&. Since individual viral isolates vary considerably with respect to EDs0 values, inhibitory capacity is often expressed in terms of the concentration of drug that provides at least 50% inhibition of 90% (or another arbitrarily chosen percentage) of viral isolates tested. For freshly isolated clinical strains of hCMV, the ED, values for individual isolates vary widely, approximately 25-800 pM, and the concentration that provides an EDso for 90% of the isolates may be as high as 400 pM [l]. For HIV isolates, EDs,, values vary in the range of 32-150 pM, with 90% of isolates being inhibited by concentrations within this range [ll. It must be recognized, however, that the measured EDs0 is quite variable and is dependent on cell type, medium, and multiplicity of infection, as well as the viral isolate. Indeed, repeated measurements of one isolate with the same assay conditions can yield variable results; for example, according to a recent report [2], seven measurements of a control strain of the varicella zoster strain Ellen yielded greater than twofold variation. Since there is no standardized in vitro system for defining an EDs,, and since no in vitro system to date has been shown to correlate with clinical effectiveness or failure, one is at a considerable disadvantage in terms of setting truly rational goals for effective drug levels. It should be noted that both enzymologic measurements and biologic measurements are nearly always made with a constant concentration of drug over time; since both the enzymologic inhibitions and the biologic antiviral effects of foscarnet have been shown to be reversible, it is rational to assume that concentrations such as those derived from Ki and ED5c values represent concentrations that provide the inhibitory effects only so long as those levels are maintained adjacent to the enzymes or cells.
ON HERPESVIRUS INFECTlONS
IN AIDS / LIETMAN
It is clear, however, that levels of foscarnet fluctuate as a function of time in humans unless the drug is given as a constant infusion, which is inconvenient. The pharmacokinetics of foscarnet describes these fluctuations of drug levels over time.
PHARMACOKINETICS The pharmacokinetics of foscarnet can be dissected into the traditional components of absorption, distribution, metabolism, and excretion, along with half-life, clearance, and volume of distribution. Absorption After oral administration, foscarnet is absorbed to a variable but always limited extent (~22%) in humans [33. The limited absorption may be associated with a relative inefficiency of a carrier-mediated transport system that is shared with inorganic phosphate and is dependent on both sodium and pH gradients [4]. Interspecies variations in this system may account for the vast differences in bioavailability among species. Rabbits, for example, absorb 95% of orally administered foscarnet [l]. Although foscarnet has been shown to be unstable at pH 2, with a half-life of hydrolysis of approximately 1 hour [5], we have found that effective neutralization of gastric acidity with ranitidine increases the absorption of foscarnet in humans by only approximately 50% (from an average of 6.2-9.9%) and in a somewhat erratic fashion [6]. Thus, chemical instability probably accounts for only a small part of the drug’s limited absorption. As a result of its poor bioavailability, foscarnet is administered only as an intravenous solution. Distribution Foscarnet appears to be distributed rather widely throughout the body, with a volume of distribution that has been reported at 1.3 liters/kg after 3 days and 5.1 liters/kg after 8-21 days of continuous infusion of 0.14 mg/kg/minute and 0.19 mgl kg/minute, respectively [3,71. Few data exist with respect to the tissue distribution of foscarnet in humans, although cerebrospinal fluid drug levels have been measured at 13-68% of simultaneous plasma drug levels at steady state [71. The relationship between cerebrospinal fluid drug levels and brain drug levels has not been described in humans or experimental animals. In mice, drug levels in kidney and lung appear to be about lo-fold higher than simultaneous blood and heart levels, and the levels in each of these tissues appear to decrease over time in a parallel fashion [ll. The distribution of foscarnet into bone, however, is a dramatic exception. In mice, bone foscarnet concentrations are several orders of magnitude higher than blood levels, and elimination from bone is exceedingly slow
February 14, 1992
The American Journal of Medicine
Volume 92 (suppl 2A)
2A-9s
SYMPDSIUMON HERPESVIRUS INFECTlONSIN AIDS/ LIETMAN
following an intravenous dose [l]. It has been esti- cumulate either more or less foscarnet than uninmated that approximately lo-30% of systemically fected cells, in contrast with the preferential accuavailable foscarnet is deposited in bone in mice in a mulation of acyclovir and ganciclovir by herpesdose-independent process [ 11. The accumulation of virus-infected cells. foscarnet is in bone matrix, not marrow [l], and it is presumed that deposition of the drug is similar in Metabolism principle to deposition of inorganic phosphate as a No metabolism of foscarnet has been detected in component of the bone matrix. The elimination of animals or humans [l]. In this regard, foscarnet diffoscarnet from mouse bone has a half-life of many fers from many nucleoside analogues, which must months; it is assumed that this is also true in hu- be anabolically converted intracellularly to active mans, explaining the observation of an exceedingly nucleoside triphosphate analogues in order to be long terminal half-life [l]. The kinetics of bone dep- effective. As a result, the activity of foscarnet is osition and release, however, has not been fully elu- independent of herpesvirus thymidine kinases, and cidated. It is at least plausible to hypothesize that thymidine kinase-deficient herpesvirus mutants there is a rapidly exchangeable pool of foscarnet in remain sensitive to foscarnet. It also follows that bone, as well as a slowly exchangeable pool, thus foscarnet is unlikely to accumulate in infected cells, accounting for some of the variability in plasma lev- as do acyclovir and ganciclovir. els on a short-term (hourly) basis that is observed even with constant intravenous infusion of the Elimination drug. Foscarnet is eliminated as unchanged drug priThe intracellular distribution of foscarnet has not marily, if not exclusively, by the renal route been completely defined, although this is of obvious [3,‘7,10]. The renal elimination of foscarnet clearly importance since the site of action of foscarnet is is a function of glomerular filtration, since the renal intracellular. In a human fibroblast cell line, the clearance varies as a function of the glomerular filtime course of foscarnet association with cells (pre- tration rate [2]. In addition, however, foscarnet is sumably uptake into the intracellular space) was assumed to be secreted into the urine, since the considered to be slow, requiring approximately 4 renal clearance exceeds the clearance expected hours to reach equilibrium, although the time re- from glomerular filtration alone [2]. It has been esquired to reach 50% of the equilibrium value ap- timated that glomerular filtration accounts for appears to be
S.
LIETMAN,
M.D.,
Ph.D.,
Foscarnet
Baltimore, Mary/and
Foscarnet exerts its antiviral effects via reversible inhibition of viral pblymerases. Pharmacodynamic data indicate that herpesvirus and human immunodeficiency virus replication is inhibited by therapeutically achievable concentrations of foscarnet; however, the concentrations of foscarnet required for such inhibition have been found to vary widely. Pharmacokinetic data indicate that foscarnet is eliminated via the renal route, undergoes negligible metabolism, and appears to be distributed widely from the circulation. However, the available data indicate that the pharmacokinetics of the drug varies among patients and within the individual patient, particularly with regard to plasma drug levels; furthermore, such factors as the intracellular kinetics of the drug have yet to be well characterized. It is thus difficult to formulate optimal dosing regimens on the basis of what is known of foscarnet pharmacodynamics and pharmacokinetics. Nevertheless, dosages that produce clear-cut therapeutic benefits without unacceptable toxicity have been identified in clinical trials of foscarnet in acquired immunodeficiency syndrome (AIDS) patients with cytomegalovirus (CMV) retinitis.
From the Division of Clinical Pharmacology, The Johns Hopkrns Unlverslty School of Medicine, Baltimore, Maryland. Requests for reprints should be addressed to Paul S. Lletman, M.D., Ph.D., Division of Clinical Pharmacology, Osler 527, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, Maryland 21205.
2A-8s
February 14, 1992
The American Journal of Medicine
F
oscarnet (trisodium phosphonoformate) is an antiviral agent with in vitro activity against herpes simplex viruses 1 and 2, varicella zoster virus, and human cytomegalovirus (hCMV), as well as human immunodeficiency virus (HIV) and other viruses [ 11. In principle, the rational use of foscarnet should be based on an understanding of both its pharmacodynamics (the relationship between the concentration of drug at its site of action and the biochemical and biologic effects it produces) and its pharmacokinetics (the concentration achieved at the site of action and/or in the plasma as a function of time as the drug traverses the body). That is, in principle, some knowledge of the pharmacokinetics of foscarnet should allow the identification of rational doses and dosing regimens that achieve goals based on the pharmacodynamics of the drug.
PHARMACODYNAMICS Oberg [l] has compiled an extensive and useful review of data on foscarnet pharmacodynamics. Foscarnet is a pyrophosphate analogue capable of inhibiting the herpesvirus DNA polymerases and the HIV reverse transcriptase. Its potency against these viral enzymes is considerably greater than its potency against host cell polymerases, thus accounting for its selective toxicity for viral replication as contrasted with host cell replication. Foscarnet is a reversible inhibitor of the viral DNA polymerases and HIV reverse transcriptase; consequently, the extent of inhibition of these viral enzymes fluctuates with rises and falls of drug concentrations at the intracellular locations of the enzymes. The reversible inhibition of the target viral polymerases has been shown not to be competitive with the enzyme substrates (i.e., deoxynucleoside triphosphates) or with templates (i.e., DNA or RNA) [l]; therefore, the degree of inhibition of the viral enzymes is not dependent on either intracellular deoxynucleoside triphosphate levels or intracellular template levels. This is in contrast with nucleoside analogues, such as acyclovir, ganciclovir, and zidovudine, the activity of each of which is influenced, sometimes in an advantageous manner and sometimes in a disadvantageous manner, by the intracellular levels of endogenous nucleoside triphosphates.
Volume 92 (suppl 2A)
SYMPOSIUM
At an enzymologic level, the relationship between the concentration of foscarnet adjacent to the viral enzymes and the extent of inhibition of these enzymes can be described by measuring the inhibitory constant (KJ of foscarnet with respect to each enzyme. This value defines the concentration of drug providing 50% inhibition of the relevant enzyme. The Ki for foscarnet and at least one hCMV DNA polymerase has been reported to be approximately 0.3 PM [l], whereas the Ki for foscarnet for several HIV reverse transcriptases has been reported to be 0.1-l pM [l]. The relationship between foscarnet concentration and the inhibition of viral replication can also be defined at a biologic level by measuring viral replication in infected cells in vitro that are exposed to constant drug concentrations over time. Such measurements usually are expressed in terms of an effective or inhibitory dose for 50% inhibition of viral replication (EDs,, or ID&. Since individual viral isolates vary considerably with respect to EDs0 values, inhibitory capacity is often expressed in terms of the concentration of drug that provides at least 50% inhibition of 90% (or another arbitrarily chosen percentage) of viral isolates tested. For freshly isolated clinical strains of hCMV, the ED, values for individual isolates vary widely, approximately 25-800 pM, and the concentration that provides an EDso for 90% of the isolates may be as high as 400 pM [l]. For HIV isolates, EDs,, values vary in the range of 32-150 pM, with 90% of isolates being inhibited by concentrations within this range [ll. It must be recognized, however, that the measured EDs0 is quite variable and is dependent on cell type, medium, and multiplicity of infection, as well as the viral isolate. Indeed, repeated measurements of one isolate with the same assay conditions can yield variable results; for example, according to a recent report [2], seven measurements of a control strain of the varicella zoster strain Ellen yielded greater than twofold variation. Since there is no standardized in vitro system for defining an EDs,, and since no in vitro system to date has been shown to correlate with clinical effectiveness or failure, one is at a considerable disadvantage in terms of setting truly rational goals for effective drug levels. It should be noted that both enzymologic measurements and biologic measurements are nearly always made with a constant concentration of drug over time; since both the enzymologic inhibitions and the biologic antiviral effects of foscarnet have been shown to be reversible, it is rational to assume that concentrations such as those derived from Ki and ED5c values represent concentrations that provide the inhibitory effects only so long as those levels are maintained adjacent to the enzymes or cells.
ON HERPESVIRUS INFECTlONS
IN AIDS / LIETMAN
It is clear, however, that levels of foscarnet fluctuate as a function of time in humans unless the drug is given as a constant infusion, which is inconvenient. The pharmacokinetics of foscarnet describes these fluctuations of drug levels over time.
PHARMACOKINETICS The pharmacokinetics of foscarnet can be dissected into the traditional components of absorption, distribution, metabolism, and excretion, along with half-life, clearance, and volume of distribution. Absorption After oral administration, foscarnet is absorbed to a variable but always limited extent (~22%) in humans [33. The limited absorption may be associated with a relative inefficiency of a carrier-mediated transport system that is shared with inorganic phosphate and is dependent on both sodium and pH gradients [4]. Interspecies variations in this system may account for the vast differences in bioavailability among species. Rabbits, for example, absorb 95% of orally administered foscarnet [l]. Although foscarnet has been shown to be unstable at pH 2, with a half-life of hydrolysis of approximately 1 hour [5], we have found that effective neutralization of gastric acidity with ranitidine increases the absorption of foscarnet in humans by only approximately 50% (from an average of 6.2-9.9%) and in a somewhat erratic fashion [6]. Thus, chemical instability probably accounts for only a small part of the drug’s limited absorption. As a result of its poor bioavailability, foscarnet is administered only as an intravenous solution. Distribution Foscarnet appears to be distributed rather widely throughout the body, with a volume of distribution that has been reported at 1.3 liters/kg after 3 days and 5.1 liters/kg after 8-21 days of continuous infusion of 0.14 mg/kg/minute and 0.19 mgl kg/minute, respectively [3,71. Few data exist with respect to the tissue distribution of foscarnet in humans, although cerebrospinal fluid drug levels have been measured at 13-68% of simultaneous plasma drug levels at steady state [71. The relationship between cerebrospinal fluid drug levels and brain drug levels has not been described in humans or experimental animals. In mice, drug levels in kidney and lung appear to be about lo-fold higher than simultaneous blood and heart levels, and the levels in each of these tissues appear to decrease over time in a parallel fashion [ll. The distribution of foscarnet into bone, however, is a dramatic exception. In mice, bone foscarnet concentrations are several orders of magnitude higher than blood levels, and elimination from bone is exceedingly slow
February 14, 1992
The American Journal of Medicine
Volume 92 (suppl 2A)
2A-9s
SYMPDSIUMON HERPESVIRUS INFECTlONSIN AIDS/ LIETMAN
following an intravenous dose [l]. It has been esti- cumulate either more or less foscarnet than uninmated that approximately lo-30% of systemically fected cells, in contrast with the preferential accuavailable foscarnet is deposited in bone in mice in a mulation of acyclovir and ganciclovir by herpesdose-independent process [ 11. The accumulation of virus-infected cells. foscarnet is in bone matrix, not marrow [l], and it is presumed that deposition of the drug is similar in Metabolism principle to deposition of inorganic phosphate as a No metabolism of foscarnet has been detected in component of the bone matrix. The elimination of animals or humans [l]. In this regard, foscarnet diffoscarnet from mouse bone has a half-life of many fers from many nucleoside analogues, which must months; it is assumed that this is also true in hu- be anabolically converted intracellularly to active mans, explaining the observation of an exceedingly nucleoside triphosphate analogues in order to be long terminal half-life [l]. The kinetics of bone dep- effective. As a result, the activity of foscarnet is osition and release, however, has not been fully elu- independent of herpesvirus thymidine kinases, and cidated. It is at least plausible to hypothesize that thymidine kinase-deficient herpesvirus mutants there is a rapidly exchangeable pool of foscarnet in remain sensitive to foscarnet. It also follows that bone, as well as a slowly exchangeable pool, thus foscarnet is unlikely to accumulate in infected cells, accounting for some of the variability in plasma lev- as do acyclovir and ganciclovir. els on a short-term (hourly) basis that is observed even with constant intravenous infusion of the Elimination drug. Foscarnet is eliminated as unchanged drug priThe intracellular distribution of foscarnet has not marily, if not exclusively, by the renal route been completely defined, although this is of obvious [3,‘7,10]. The renal elimination of foscarnet clearly importance since the site of action of foscarnet is is a function of glomerular filtration, since the renal intracellular. In a human fibroblast cell line, the clearance varies as a function of the glomerular filtime course of foscarnet association with cells (pre- tration rate [2]. In addition, however, foscarnet is sumably uptake into the intracellular space) was assumed to be secreted into the urine, since the considered to be slow, requiring approximately 4 renal clearance exceeds the clearance expected hours to reach equilibrium, although the time re- from glomerular filtration alone [2]. It has been esquired to reach 50% of the equilibrium value ap- timated that glomerular filtration accounts for appears to be
