CA2223874A1 - Methods and compositions for inhibiting cytomegalovirus replication - Google Patents

Abstract

Methods and compositions for use in inhibiting CMV replication in mammalian patients, wherein an effective amount of TPFA is administered to a patient in need of such treatment. TPFA is demonstrated to be at least as effective as PFA in inhibiting CMV viral replication, while having substantially improved bioavailability relative thereto.

Description

W O 96/39148 PCTAUS96i'~fin29 METHODS AND COMPOSITIONS FOR INHIBITING
CYTOMEGALOVIRUS REPLICATION
Back~lùulld of the I~v~n~ioll The present invention relates generally to the fields of biology-S and medicine. More particularly, the ~l~S~,lIL invention relates to composit~and methods for use in treating a ~ n patient infectecl with particular viruses, such as cytomegalovirus (CMV).
Acylu~alhiclelluvil u ,, human i.. , .. ~ efl'c;~ ;y vil u, (HIV~
has been linked to p~thogene~i~ of the fatal immuno~u~lcssive disease known 10 as Acquired Immune Deficiency Syndrome (AIDS). Repliç~ti~n of the virus is dependent on an RNA-directed DNA polymerase (r~ c ll~l.s~ lasc, RT). Compounds that selectively inhibit HIV RT relative to human DNA
pûlylllelascs provide a basis for anti-viral chemotherapy of pre-AIDS and AIDS
p~fient~
AIDS is believed to have been first ol)selvGd as a m~-lic~l phenom~non in the ~ l of 1981 in Los ~ng~l~s The plilllaly cause of AIDS is well established to be infection with HIV. It is e~ tecl that as many as 1,000,000 ~ Oll5 are infecte-l ("HIV-~osiliv~") in the USA alone. After inf~ction, the average progression to :~y~lylOllls of the disease syndrome is 7 20 years or longer; however, death usually follows within 18-24 months of ~ gnosi~ [Stine, G. J. Acquired Immune Deficiency Syndrome: Biological, Medical, Social and Legal Issues; Prentice-Hall: Englewood, New Jersey, 1993;
Vol. 462].
HIV is princir~lly lln~ e-i by sexual contact, IV drug use 25 withcn~ te~lneedles,andtlall;,ru~ionwithco"L~i~te~1blood. Thevirus attacks human T4 cells, which are essenlial to normal function of the immune system. As the T4 cell count ~limini~hes, loss of the immune response leads to ncontrolled b~cttor~ fungal, parasitic and/or viral infections, known as OppolLulli~Lic Infections (OI).
Chemotherapy of AIDS involves drugs dil~ cLly talgelillg HIV, such as AZT (Zidovudine ), and drugs targeting various oppolLulli~Lic infection~. Ideally, drugs to treat any disease caused by an infectious pathogen WO 96/39148 PCT~US96/08029 should be effective against the target organism at re~on~hly low collcellll~Lion(i.e., have high potency). They should also have low toxicity and minim~l side effects, exhibit good stability under standard storage cnn~itinnc, be readily bioavailable when taken orally (avoiding discolllro, I and cost from injection),5 and be ineApel.sive to m~mlf~rtllre and ~-lminicter. These considerations are particularly important with respect to AIDS chemotherapy. Because infection is perm~ne~t, incurable and lethal if untreated, medication must be taken for the life of the patient; this enh~nres the importance of toxicity and cost. For the same reasons, the path of drug ~ tion is especially im~olLall~.
Because AIDS patients must be treated for HIV infection and simlllt~neously for one or more oppolluni~Lic infectionc, drug compatibility is also a key issue. This is a matter of particular importance where the OI is viral in origin. For eY~mple, AZT (to treat HIV) and g~nrirlQvir (Cytovene, an anti-viral used to treat CMV) are only compatible because they reillru~;e each 15 other's toxic effects.
CMV causes blin~ln~cc in a s.,bsl~ l portion of AIDS p~ti~ntc (CMV reLilliLi~). One estim~te [Stine 1993, supra] su~ L~ that 46% of AIDS
patients who have reached the latter stages of the disease suffer from CMV
rGli lilis; other estim~tÇs range from 20% to 76% [Keijer, W. J. et al. Ocular 20 complic~tion~ of the acquired immunodeficiency syndrome. Focus on the treatment of ~;y~o~eg~l()virus leli.li~is with ganciclo-vir and foscarnet. PharmWorld Sci 1993,15, 56-67] and 11% to 40% ~Hansen, L L. {Retinal ~iice~CPs in AIDS}. Ophthalmologe 1993, 90, 239-49]. A CMV vaccine is still in the early stages of develo~ l and ~ l agents will colllillue to play a major role 25 in CMV il,r~cLioll management for the Çor~secable future [Sasadeusz, J. J.
Sacks, S. L. Systemic a-lLivil~ls in herpesvirus infectinnc. Dermatologic Clinics 1993,11, 171-85]. CMV pnS~e,~es a virus-specific DNA polymerase.
One drug recently a~luved by the FDA for tre~tment of AIDS-related CMV-reli lilis is foscarnet (Astra Pharmaceuticals), also known 30 as Foscavir and PFA (phosphonoformic acid) [Stine 1993, supra]. Like another drug a~pruved for the same cnn-litinn, ganciclovir, PFA does not cure the condition, but it can signific~ntly delay progression to blindness and W O 96/39148 PCT~US96/08029 increase patient survival time [Gumbel, H. et al. {Therapeutic altc,naLive or 2dchoice drug. Trisodium phosphonorc,lllla~e in cytomegalovirus rG~hli~is}.
Fo~bch~ e der Ophthalmologie 1991, 88, 731-4; Brockmeyer, N. H. et al.
Foscarnet treatment in various cytomegalovirus infections. Intema~ional Journal 5 of Clinical Pharmacology, Therapy, and Toxicology 1993, 31, 204-7; Polis, M. A.
et al. Increased survival of a cohort of patients with acquired illllllunodeficiency syndrome and cytom~g~lovirus l elil-ilis who received sodium phosphonoro""~te (foscarnet). Amencan Joumal of Medicine 1993, 94, 175-80]. Both drugs have an initial efflcacy of induction therapy of 80-90%, and m~int~n~nce therapy is 10 always needed to pr~ ~enL a relapse [Keijer et al. 1993, supra~. About 40% ofAIDS p~ti~ntc with CMV le~ iLis cannot tolerate the side effects of ~nriclovir, which include inhibited production of bone ,.la,l~,w white blood cells.
Although PFA also shows toxic side effects, notably reversible kidney damage and abnormal blood electrolyte levels [Ryrfeldt, A. et al.
15 H~pocalcemia in~ ced by foscarnet ~Foscavir) i-lruSiOI- in d~gs F~ndamental and Applied Toxicology 1992, 18, 12~30; Cea,lla, l, M. O.& Sorg, T. B.
Foscarnet-intillce~l severe hy~ gnr~.. i~ and other electrolyte disorders.
Annals of Pharmacotherapy 1993, 27, 285-9], its IuAiciL~ profile is very di~crellt from that of g~n~ irlQvir. This di~crGllce is i~ u- L~ particularly for p:~tient~
20 also receiving AZT, with which PFA is more coll,~alible than g~nric lovir, and fosc~rnet-AZT combination chemotherapy has been demon~L.ated to prolong the lives of AIDS-CMV patients relative to the lives of ganciclovir-treated hislolical controls [Polis et al. 1993, srlpra]. It has also been reported that concurrent use of ganciclovir and fosca.-let in cases of failure of either alone25 in AIDS CMV infectir)n~ was as effective as st~n-l~rd therapy with single agents and may be of value in cases of clinic~lly-r~c;~l~..L CMV lc~i.lili. in AIDS
[Flores-Aguilar, M. et al. Pathophysiology and treatment of clinically resistant~ cy~o~neg~lovirus lcLilliLiS. Ophthalmology 1993,100, 1022-31].
Foscarnet was further shown to induce remission of CMV
30 ~asl~oi,-te~tin~l disease in 67% of a group of AIDS patients when ganciclovirin~ ction had failed. Fosc~rnet has been shown to penetrate the blood-brain barrier and is stated to be the drug of choice for CMV encephalitis [Hengge, CA 02223874 1997-12-0~

U. R. et al. Foscarnet penetrates the blood-brain barrier: r~tion~lp for therapyof cytomegalovirus encephalitis. Antimicrobial Agents and Chemotherapy 1993, 37, 1010-4]. Efforts to ~ ulllvell~ problems of systemic (infusion) z~.l...i..;~Ll~tion of foscarnet have included ll~ns~cleral iontophoresis [Sarraf, D.
et al. TrRn~cr~--ral iontophoresis of rosca~ L American Joumal of Ophthalmology 1993,115, 748-54]; nonetheless, there remains a need for agents which may be ~l",i";~.ed in a more collvt:"ient m~nnl-r, and ~refe.~bly orally.
U.S. Patents 5,072,032 and 5,183,812 to McKenna, the entire disclosures of which are hereby incorporated by ,erere,lce, disclose the use of TPFA as an anti-HIV drug. A novel, simple ~lllhesis of trisodium TPFA from trimethyl phosphon~rollllate was also reported. The compound has been completely characterized by elemental analysis, 31p, 13C and lH NMR, W, IR
and X-ray crystallographic analysis [see, e.g., McK~nn~, C. E. et al. Design andSynthesis of Organophosphorus Compounds with Antiviral and Other Bioactivities. Phosphorus Sulfur 1990, 49, 183-186; McKenna, C. E. et al.
Sodium Thiophosphonoro"lla~e, a Selective HIV RT Inhibitor: Facile Synthesis and Effects in HIV-Infected Cell Culture. Annals NYAcad. Sci. 1990, 616, 569-572].
TPFA was found to inhibit HIV-1 reverse transcriptase with a very similar potency to that of PFA (ICso near 1 ,~LM) [McKenna et al. 1990, supra]. TPFA was significantly less inhibitory than PFA for DNA polymerases specific for Herpes simplex virus in these experiments and somewhat less inhibitory than PFA for a group of human DNA polymerases. In a co,ll~alison of TPFA vs. PFA for dose-related inhibition of HIV-1 infected human H9 cells in culture (p24 ~ A~l~ssion assay), it was found that the two drugs had virtually identical inhibition curves. An NMR study further showed that under model assay c~ n-lition~, almost all the TPFA was co,l~el~ed in situ into PFA. It is nolewul~,y that TPFA was the only simple pyrophosphate analog so far reported to equal the potency of PFA.
In an HPLC study on the pharmacokinetics of TPFA vs. PFA
in the cat [Straw, J. A. et al. Pharm~ç--kinetics of potential anti-AIDS agents -s thiof~ sc~rnet and fn~ç~rn~t in the cat. Journal of Ac~uired Immune De~ciency Syndromes 1992, 5, 936-42], it was reported that TPFA had a shorter renal ~ clearance than PFA (42 min. vs. 172 min.), and that TPFA was partly co~ led to PFA in vivo; an inactive metabolite, thiophosphonate (TPA) was 5 also detected. The 6-h cumulative urinary excretion was 42% of an inll~/el,ou:, -dose, including 23.5% as TPFA, 14% as PFA and 5% as TPA. A key ~A~ lll cv...~ ed oral bioavailabilitv of the two drugs. TPFA was about 3x more bioavailable than PFA when ~.l...i..i~t.~ed by SSL enteric coated c~rs~ . When ~ erc;d orally by gavage, TPFA also had a higher mean 10 oral bioavailability, 33% (with greater variabilitv). Based on 6-h urinary eA1lGLioll of total drug, the mean oral bioavailability with TPFA was 44~o, similar to that c~ tetl from the plasma (il~Llilvellous) data.
In another study [Straw, J. A. et al. Il-iofosc~--lct (TPFA) and Foscarnet (PFA) Pharm~cokinetics in Beagle Dogs. Proc. Am. Assoc. Cancer 15 R~s. 1992, 33, 530], it was found that 95% of i.v. ~ d TPFA was excreted in the urine, in the form of 78% TPFA, 8% PFA and 9% TPA. Oral bioavailability of TPFA was only 13%, but was increased to 44.5% by cimetidine ~ L (the latter, an anti-ulcer drug, decreases stomach acidity and ~resu"lably thereby increases TPFA stability in the stom~th). In 20 c~ l i .on, the oral bioavailability of PFA, 12%, was only slightly increased by ~nt~ . ell eal-l-ent.
It is an object of the present invention to provide compositions and methods for treating CMV infection which do not suffer from all of the drawbacks of the helelorolc-known methods and col~l~o~ilions.

Summary of the Invention Pur~uallt to the present invention, there are provided methods and compositions for use in inhibiting CMV replication in " ~ n patients, wherein an effective amount of TPFA is ~ elf d to a patient in need of S such treatment. TPFA is demonstrated herein to be at least as effective as PFA in inhibiting CMV viral replication, while having sub~ lly illl~lu~,d bioav~ hility relative thereto.
Brief Description of the Drawin~
The illven~ion may be better understood with reference to the 10 ~cc~---pallying drawing, in which:
Fig. 1 illustrates inhibition of hCMV repli~ tion in human cells by added TPFA and PF~
Detailed Dcs~ Liol, of the Invention Pu,~ua"t to the ~res~"l invention, compositions and methods lS are provided for inhibiting CMV replication in ..,~,.i...~li~n p~ti~nts using an effective amount of TPFA. ReC~use it has been reported (for example, in the aforementioned U.S. Patents 5,072,032 and 5,183,812) that TPFA was sub~a"lially less effective than PFA with respect to viruses other than HIV
(e.g., herpes simpl~Y Types I and II, P.pst-o.in-Barr, Herpes Virus 6), it is quite 20 ~Ul~ illg that s~lbst~nti~l efficacy against CMV has now been found. Given the completely unpredictable nature of the antiviral inhibitory activity of TPFA, it clearly could not have been predicted that TPFA would be efflçzlcious in the m~nner Ai~close~l and claimed herein.
As would be immediately apparent to those skilled in the art, 25 TPFA and/or its ~AAitinn salts (as described, for example, in the aforementioneA U.S. Patents S,072,032 and 5,183,812) may be ~Aministered to rn~mm~ls (inr.luAing humans) in an amount effective to inhibit viral replic~tifln.
The O~UIilllUlll rate of ~fl~u;,-;~ tion for a given formulation (e.g., the freecompound or a physiologically acceptable salt form thereof) and mode of 30 delivery may routinely be dele,mi,led empirically. In general, an e~eclivt;
a~uoulll of the active agent is in the range of about 1 ,uM to about 10 mM per kilogram of patient body weight. The compound may be ~Amini~tered orally, W O 96/39148 PCT~US~6~08029 parenterally, topically or by other standard routes of ~ tion; it is presently ~.~rel-~ d to ~ "i"i~er the compound orally. Typically, compositinn~
for use in accordance with the present invention cu~ ,lise a pharmaceutically acceptable carrier, adjuvant or excipient; details concerning some suitable types 5 of formul~tions may be derived from, e.g., U.S. Patent 4,665,062 to F.rik~son et al., the entire ~ nsllre of which is hereby incorporated by l~;Çelel,ce.
Additionally, TPFA may suitably be ~ eled in accol-lancG with the plcsellt illvelltion in colljullclion with other bioactive colll~oullds, such as AZT, ddC, ddI, antibiotics, etc.
Preliminary eAp~lilnell~ have c~ cd the effects of TPFA
and PFA on cl~ e titers in the animal model. PFA caused significant ~u~ression of cre~tinine levels, which correlate with normal renal function, whereas TPFA at similar dose had little or no effect, c.~n~ ed to a conkol.
This result could signify decreased kidney toAicity by TPFA vs. PFA.
TPFA has already been shown to have similar anti-HIV-1 potency to PFA, but dirr~ pharm~colo~c~l characteristics in animal models which indi~t~ that TPFA might have adv~nt~gPs as a replacement for PFA.
Pursuant to the ~ selll invention, it is delllull~ ted that TPFA also has activity against CMV, further çnh~ncin~ its ~otell~ial as an AIDS-related anti-20 viral agent.
The invention may be better understood with leLercllce to the ~ccolllll~llyillg examples, which are int~onrle~l for purposes of illustration only and should not be construed as in any sense limiting the scope of the invention as defined in the claims appended hereto.
25 Example 1 Synthesis of the target colllpoulld, TPFA, depended upon the ple~alation of the intermediate Me3TPFA, which was effected as described in - the arolGlllellLioned U.S. Patents 5,072,032 and 5,183,812. Thinn~tinn of trimethyl phosphonoformate (Me3PFA) by Lawesson's reagent in acetonitrile 30 under N2 afforded analytically pure Me3TPFA in 75-80% yield on a small (10 mmole) or large scale (1.5 mole). The ~llu~;lulG of Me3TPFA was verified by lH, 13C, and 31p ~nMnR (Table 1).

WO 96t39148 PCTrUS96/08029 Table 1. NMR P~Jr lies of Me,TPFA and Na3TPFA S~ ks Coepound Solvent 1H ~R 13C ~R 3~P ~R
d ~pp ) d Cppo) d ~pp ) He3TPFA CDCI33.85(d, 6H) 52.7~s) 3.88~s, 3H) 54.3~d, 2J~7 Hz) 167~d, ~J~226 Hz) 65.0 ~s) TPF~ D20 no ~i~n~l 183.3~d, 1J~181 Hz) 37.8 (s) Hydrolysis of Me3TPFA with 10 M NaOH gave, after purifir~ti-)n, the drug in 24-30% yield and analytical purity. The structure was verified by lH, 13C, and 31p NMR (Table 1) and elemental analysis (CNa304PS 1.5H20, found: C, 5.24;
H, 1.50; S, 13.68; Na, 29.24; calcd: C, 5.11; H, 1.29; S, 13.64; Na, 29.34).
15 W/vis s~ccL.uscopy (H20, pH=8.8) gave e254"",=0.9x103, e23"",=3.6x103, e~Os,,,,,=6.7xlO3. Titration of TPFA was carried out at both 25~C and 0~C with 0.0837 M standardized HCl using a Corning 125 pH meter. The pKa3 was determined to be ca. 7.5.
TPFA, (LZIII.1.19) (0.1175 g) was dissolved in pure water 20 (HPLC Grade) and diluted to 5.00 mL (volumetric flask). Two 0.100 M
solutions were ~lcpared kl~ntics~lly Aliquots of each solution (2 x 3.00 mL) were employed in the CMV inhibition assay described in E-xample 2. The r~ ;..i..g solutions were stored in a freezer (-20~C), and an equivalent solution of PFA sodium salt (from Alfa products) was also ~r~ared and used as a 25 positive control.
To dele~ e the chemir ~l behavior of TPFA in the assay for inhibition of CMV in infected human cell cultures, it was necessary to separate the drug from the very numerous colllp~llellL~ of the assay llliX~UI~, particularly those collllil)uLed by the DMEM medium which coll~ s 14 amino acids, 9 30 inorganic salts, 8 vil~ , and 3 other colll~ounds. Previous work had used di~ lL sample matrices (blood, urine) [Straw, J. A. et al., Joumal of AcquiredImmune Deffciency Syndromes 1992, 5, 93~42; Straw, J. A. et al., Proc. Am Assoc. Cancer Res. 1992, 33, 530], requiring development of a method coll~alible with the DMEM medium. The goal was to develop a method -CA 02223874 1997-12-0~
W O 96/39148 PCTrUS96/08029 pe~ ilLing qllA"LilAI;ve determinAtion of 10-3 _ 106 M TPFA in drug assay samples. TPFA absorbs W weakly, and PFA is undetectable by UV. Neither compound is directly detectAhle by emi~ion (fluorescence) spe~;Lloscu~y.
These considerations suggested electrochPmi~l detection (ECD) as the most 5 ~lolllisil.g direct-detection method for TPF~ ECD can be highly sclccLive and also very sellsilive for specific analytes. The analyte passes through an electrochemical cell, where it undergoes nYi~lAtion or reduction at the working electrode, and the reslllt~nt current is detected. In these c~e~ i"lPnt~
~xi-l~tion mode was used, corresponding to a positively charged wolhil~g 10 electrode. The earlier work was based on the use of an obsolete t1etectnr andprovided inadequate scllsiLiviLy for present purposes. Cvll~clilive tests demo~ ated that the Waters 464 detector would best suit the re4uilel.lents of these experiment~; in ~d~iti~n to equivalently high sel-silivily and ease of use, the Waters 464 afforded a larger selectir n of sel~silivily ranges, 15 coll~ollding to the need to conduct analyses over a wide span of drug concentraffons.
As expected, TPFA -- with its more easily nYi-li7~-1 C-P=S
bonds -- gave a stronger leS~ )llSe than PF~ To ensure detection of PFA, a relatively high potential of +1.2 V was adopted. In reversed phase HPLC, 20 electrolysis of water limits the working electrode potential at pH 7 to around + 1.2 V. At this potential, the background ~;ullcllL was ca. 1 ,uA, and the mostsel-silivc detection current range was 0.5 - 10 ,uA.
ECD is more sen~iLivc to changes in eluting conditions, including the eluent, than detection based on light absorption or emi~cion To ensure the 25 integrity of the results, such variables as pH, ionic strength and solvent were studied for their effect on llliALulc resolution and detection-Reversed-phase ion-pair chromatography (IPC) has quickly gained widespread acce~lal.ce as a versatile and efflcient method for the separation of ionized and easily ionizable analytes. It is complementary to ion-30 eYrhAnge chlu,.,Atography which is used to separate similar samples. AniLupolLal t adv~ntAges of IPC is its ability to simlllt~neQusly se~al~te samples col,l~i"i,-g neutral and ionized molecules. Method development in IPC is CA 02223874 1997-12-0~
W O 96/39148 PCT~US96/08029 generally more flexible since the type and capacity of the stationary phase for ion interactions can be varied by changes in the composition of the mobile phase. Retention and selectivity in reversed phase IPC are influenced by a large number of c~e~ ,ental variables, including the type and hydrophobicity 5 of the coul-telion, the concc~ tioll of ion-pairing reagent, the type and conce~ tion of the buffer, the pH, ionic strength, concellllation of organic morlifiçr, telll~G,~tulc and the sorptive ~.o~e.lies of the st~ti~n~ry phase.
In the HPLC system used herein, a 25 cm x 4.6 ,um 5 ~m Spherisorb ODS-2 cnlumn (Rainin) was employed. The normal operating pH
10 range for this column is 2.5-8Ø It was found that injection of DMEM culturemedium samples without filtration resulted in rapid reduction in separation efficiency together with increasing back ylei~U~e. ~onsequently, medium s~mpl~s were filtered before injection. For this purpose, samples were celllliruged using Microcon microconce.lll~tol~ model 3, with membrane 15 MWCO (molecular weight cut off in Daltons) at 3000, spun in an Eppendorf 5415~ microcGllllifug~, at 14,000 rpm for 25 min. In spite of this ~lec~ul;~ n, some reduction in column efflri~ncy was seen after two weeks of intensive medium sample injections (around 30 ~mrlç~ per day). ThelerorG, every two weeks the column was legel,~ .ated by ~U~llpillg 30 volumes (125 mL) of HPLC-20 grade water, acetonitrile, chlolorolm, ~rGlc--.iL. ;l~, water and finally the mobile phase each. A standard approach to yrGvGllL column performance delGliolalion is the use of guard columns. However, this led to ~i~nific~nt deterioration in TPFA peak shape (guard column with Ultrapack ODS 10 um).
All compounds and solvents were of high purity (solvents, 25 HPLC grade). For oYidative detection, removal of dissolved air from the mobile phase is necçs~ry to ~ ell~ air bubble rol~alion at the column outlet, which di~lulbs the electrolysis process. Solvents were degassed by vacuum filtration through a Versapor 450 membrane filter (0.45 ,um) and argon sparging. All new st~inl~c steel components (among them a pulse damping 30 system, which was incolyolated between pump and injector for some e~yeli~l~e~lb) of the HPLC system (~Ycl~ ing the detector flow cell) were -, passivated with 6 M nitric acid by pUlll~illg at 1 mL/min, 20 mL water, 20 mL
nitric acid (6M), 20 mL glacial acetic acid, 20 mL water, 20 mL mobile phase.
- With daily use, the detector lost sensitivity after about one month, chiefly due to cont~Tnin~tion of the working glassy-carbon electrode.
5 Glassy-carbon working electrodes were re-activated by immersion in a solution of 1 g ch~o"lic acid in 10 mL reagent grade sulfuric acid for 15 min (background current ~ 0.50 ,uA instead of ~ 1.0/1.2 ~LA). It is also necessary to add rere~ ce electrode filling solution to refelence electrode once monthly.
To m~int~in c~ lç. .cy with data on TPFA in blood and urine, 10 p-hydro~ybenzoic acid (POBA) was ~ltili7.o-1 as the int~rn~l HPLC st~nrl~rd.
Detector output was charted and integrated using a Hewlett-Packard 3390A
integrator. Chrom~tograms and NMR spectra were sc~nn~.~l using a Hewlett-Packard CI~ Scanner and Des~ç~n II sorlw~re. NMR spectra were obtained on a Bruker 360 MHz ill~LlumellL~ using D20 as internal or external lock solvent.
Initially, the phn~lJh~te bu~red eluent at pH 6.8 was used.
This gave good resolution of PFA, the POBA st~n~l~rd~ and TPFA (all at 100 ,uM) in an aqueous sample. A ~llullg. ~ lespol-se from TPFA vs. PFA was evident. In a sepalaLion of DMEM medium at eluent pH 7.02, strikingly few peaks (considering the many medium components) were detected, and 'windows' were observ-ed in the 4-6 min and >7 min regions of the chlulllalogram. The major peak at 6.39 min was i~le~tified using a kit of amino acid standards as l~y~ophan~ a medium component. Medium bach~loulld could be reduced by sample dilution, at a Cull~ n~lin~ cost in total sensi~ivilyTPFA peak broadening was found to be dependent upon the medium concel-ll~lion; ~llullately, it could be abated simply by moderate dilution of samples with H2O. Subsequent samples were diluted by 1: 1 or more, which gave single TPFA peaks.
It was found that a slight reduction of pH broadened the TPFA
peak, which at pH 6.55 split into two peaks. Resolution of the POBA st~n-l~rd from ~ly~lophan was only fair at this pH. At pH 7.02, the st~n~l~rd was nicely resolved, but lly~tophan now co-eluted with TPFA. Optimal resolution was CA 02223874 1997-12-0~

obtained at about pH 6.8-6.9.
Initially, a phosphate buffer concentration of 0.1 N was used, which provides good resolution of p-hytlluAyl~enzoic acid (POBA - standard), TPFA and PFA in aqueous solutions. However, to o~limi~ medium-drug S resolution with respect to the LlypLophan interference, other concentrations of phosphate buffer and the organic modifier (aceLo~ lile) were investig~ted to resolve the st~ntl~rd, TPFA and lly~LOphan peaks s~tic~torily. Optimal separation was achieved using a phosphate concel,l~dtion of 0.07 N, with 7%
acetonhrile.
The concentration of tetrabuLylallllllonium phosphate (TBAP), set at 2.4 mM, significantly influenced resolution of the TPFA, POBA and ~lyp~o~han peaks. As pointed out above, with prolonged usage the resolving power of the column deteriorates. However, it proved possible to re-o~Lill,i~e resolution by modifying the concentration of the ion-pair agent over a range of 15 2.4 -3.2 mM. Pyrophosphoric acid also can be used as a component of the buffer (conc. = 1 mM) to sharpen the PFA peak.
Calibration cuIves were developed using the o~
conditions for HPLC-ECD analysis of TPFA and PFA in water. Good linearity of rcs~ol,se was observed over drug concc"lldlion ranges of 0 to 75-100 ,ILM.
20 The sel,~ilivily ratio for PFA vs. TPFA was 0.0285. A similar response slope was obtained with TPFA in the medium (dilution 1:1) vs. water. The stability of TPFA in medium was monitored by both HPLC (low, medium concentrations) and 31p NMR (high concentration). TPFA was dissolved in the medium and incllh~tP(1 at 37+0.5 ~C on a VWR 400 HPS hot plate with 25 tem~cldture probe. The 31p NMR spectrum of the me-lium alone reveals a single peak, attributed to phosphate, which was easily distinguishable from TPFA and PFA at pH 8.3 (sample concentrations 25 mM).
At high initial concentration in medium, TPFA is slowly col,velled to PFA in a zero order process with a half-life of >150 h. At a 30 consentration of 1 ~LM (HPLC analysis), col,vt;l~ion of TPFA to PFA is much more rapid, but still apparently linear, possibly with some admixed non-linear component; the half-life is ca. 6 h. At 150 ,~LM, the process is seen to be 1st W O 96/39148 PCT~US~6/08~29 order-like, showing a half-life of ca. 3 hr (half life = 161 min). The same behavior was seen at 75 ~M, with a similar but not j~lentics~l half-life (186 rnin).
~ The synthesis described in the aforementioned U.S. PatentsS,072,032 and 5,183,812 provided a TPFA sodium salt sample inrlictin~lich~hle 5 from past ~le~al~Liolls spectroscopically. Freshly ~epared sample differed from a sample ~ulJjceled to prolonged storage in a vial in being virtually free of odor. It was also pure by HPLC. The HPLC methodology for the analysis was successfully developed using a new Waters ECD and ~e~ ;r PY~min~tinn of elution cc n~litionc Despite the complPYity of the assay llliAlU~C, 10 con~litionC were est~bli-ched for the deterrnin~tinn of TPFA and PFA in micromolar concellLl~Lions,pe,lllliLLil.g study of TPFA 'prodrug' behavior in the assay medium.
EYample 2 ~Gli,llents using TPFA and PFA to inhibit replication of 15 hCMV were carried out and the data were analyzed ~ Js~ ly as described in the litelalul. [Angulo, A. et al. Pcetinni-l Activation of ~Ptinojc Acid RCCe~LU.:~ but Not of Retinoid X Rcce~ Promotes Cellular Dirrerc~-Li~tion and Re~lir-~tion of \human C~lo...çg~lnvirus in Embryonal Cells. J. rrolo~gy 199S, 69, 3831-3837]. The results in-iir~te-l that the activities of TPFA and 20 PFA in inhibilil,g hCMV DNA synthesis are similar under the c~-n~itinns used,at least at highly effective concentrations. Additional eYperiments showed that PFA was lmrh~nged during the eA~Glilll~ , whereas TPFA was partly Llall~rc,lllled into PFA and into a metabolite. The metabolite was identified asa compound called thiophosphonic acid, or TPA. TPA was i~lPntifi~cl in 25 reported eYperiments on TPFA metabolism in dogs and cats [Straw et al.
(1992), supra]. TPA was shown herein to have some activity of its own, a finding which has heletorole not been reported.
- ~hPmi~lly pure TPA was synth~ci7p~l as a control; new TPFA, PFA and TPA samples were ~ ared and used in the protocol previously 30 described herein.
The % viral DNA repli~tion for treated (+drug) and ulllleated (-drug) virus-infected cells is plotted as a function of initial drug concentration CA 02223874 1997-12-0~
W O 96/39148 PCT~US~6/0~029 in Fig. 1. The positive control drug, PFA, is seen to ,uyyless viral DNA
synthesis from about 0.1-1 mM, with virtually complete ~uyyre;~sion observed at 1 mM drug. In the TPFA-treated cells, a similar pattern is observed, but ayyalcntly at slightly lower overall concentrations. Several paired points which5 show a variation in the replication value of ~10% or more are included;
however, the data indicate that TPFA has at least co",y~,able activity to that of PFA. The observed activity of TPFA in Fig. 1 is probably due to a combined effect of TPFA, PFA and TPA (i.e., the sllmme~l inhibiLiolls of TPFA
and its two metabolites); no attempt has been made to factor out the individual 10 contributions to activity.
ChPmir~l analysis of TPFA and PFA in assay samples and controls using electrochPmi~l analysis methorl~ as described in Example 1 revealed the presence of a new peak in the HPLC trace, identified by use of authentic st~n~l~rd as thiophosphonic acid (TPA). The TPA was col,filll,ed by 15 NMR c~yelilllGnts. PFAwas also identified in the TPFA-treated, virus-infectedcells. The results are s~ f ~1 in Tables 2-5. Data for two conc~ nLl~Lions of drug are yl~ellLe~l- At 1 mM drug, about 1/2 the ori~in~l TPFA was found in the cell culture (after 24 hours), and a little less in the infected cell cultures.
TPFA was stable in the initially yl~yared medium ("DMEM"). The sum of 20 TPA, rem~ining TPFA and PFA was in good agreement with the initial TPFA
concentration, ruling out a major ~rl~litinn~l metabolite. PFA was stable under the same cnn-litinn~.
From the foregoing description, one skilled in the art can readily fiscel Laill the ess~llLial characteristics of the invention and, without departing 25 from the spirit and scope thereof, can adapt the invention to various conditions Changes in form and substitution of equivalents are COI-tçllllllr~tP-l as ci~ n~çs may suggest or render expedient, and any specific terms employed herein are inten~lerl in a descliyLive sense and not for purposes of CA 02223874 l997-l2-05 W O 96/39148 PCTAJS96~'~80~9 ~_ ~ o o O ~A O O ~ X ~

Nl X O o ou- O ~ ~ X ~ ~ ~
~X
i N
m ~ 2 ~ -~ c ~ ~~~ol ~ o o ~ ~ o o o x o o o ~C ~ ~ ' o ~ I_ ._ .
M
M O ~t ~: _O O X-- O O O _ O O
~x 1~ -- C ~ 1-- C

W O 96/39148 PCT~US96/08029 N 1O N N ~ ~o N r0 ~ ~0 _ X ~ ~ ~ X ~ ~ ~ X

Nl X ~ ~ ~ X ~ ~ ~ X
~ n : *
~; ~

~ ~ ~
Sl ~o ~. O r~7 'o ~ ~ ~ ~O _ ;~ ~ ~~ X o o ~ X ~ ~ ~ X

~ _ _ O O _ . O O _ _ --~ e ~: c ;~

WO 96/39148 PCT~US96108029 ~" O ~ O ~o O C> C O C O
~ x_ , ~ o _ ~ ~ x8 ~ ~
o C o X o o C o C o o ol ~ ~ ~o ~ ~ o X ~ _ ~o r~l _ ~ X ~t _ ~ X
.~

Ul .
g~ ' ' ~ '~ x a o, rl, O ~ ~ ~ x ' ~

4.~ C

~~.
o ~t _ o o X_ oo _ _ o o ~X

-WO 96/39148 PCTAJS96/~gO29 ~ ~ O O O ~ O O ' ~-- ~ O
~, _ ~ O N -- ~ 0 1 --X r~ X ~ X C~
O O O O O C

~ ~ O o r 'r ~ O ~ ~ O
o o C C o C o ~;
,~
~. C

~ O ~ ~ ~ ~' O O ~ O
~ --~ ~ ~C~ ~ --X ~ ~ X CO
o c ~_ ~ C

~: o-- o o o _ o o o _ X X X
-._ -

Claims (2)

WHAT IS CLAIMED IS:

1. A method of inhibiting CMV replication in a mammalian patient in need of such treatment, said method comprising administering to the patient an amount of TPFA effective to inhibit viral replication.

2. A composition for use in inhibiting CMV replication in a mammalian patient, comprising an effective amount of TPFA and a suitable carrier or excipient.