A Phase I dose-escalation study of third-line regorafenib with trifluridine/tipiracil in metastatic colorectal cancer.

Aim: To determine a recommended Phase II dose of the oral fluoropyrimidine trifluridine/tipiracil (FTD/TPI) combined with the multi-kinase inhibitor regorafenib (REG) in refractory metastatic colorectal cancer patients. Materials & methods: A conventional 3 + 3 dose finding design was used. FTD/TPI was administered on days 1-5 and 8-12 of a 28-day cycle, REG on days 2-22. Two dose levels were used: FTD/TPI 25 mg/m2 b.i.d. + REG 120 mg/d, then escalated to FTD/TPI 35 mg/m2 b.i.d. + REG 120 mg/d. Results: In total, 12 patients were treated at two dose levels. Three dose-limiting toxicities were observed; all were grade 3 hypertension causally attributed to REG. Recommended Phase II dose is FTD/TPI 25 mg/m2 b.i.d. + REG 120 mg/d. Median progression-free survival was 3.81 months (95% CI: 1.51-5.29), median OS 11.1 months (95% CI: 2.3-18.2). Conclusion: The combination of REG and FTD/TPI is feasible and safe. Efficacy signals exceed that of the single agents at acceptable toxicity levels and are clinically meaningful.

Lay abstract Many patients with metastatic colorectal cancer need a sequence of different treatments over time. Regorafenib and trifluridine/tipiracil (also called TAS-102) are two drugs which are both used late in this sequence of treatments, but there is no rule as to which should be used first. Both drugs have very different mechanisms of action, and it might be beneficial to patients to administer them both at the same time as a combination treatment, instead of sequential treatment. We therefore conducted a Phase Ib study with a small number of patients to investigate whether this combined treatment would be feasible and safe. The study was designed to test the drug combination at different doses, and we found that treatment with trifluridine/tipiracil at 25 mg/m² twice daily combined with regorafenib at 120 mg daily had acceptable side effects and is likely to be safe for use in future clinical trials. Efficacy results suggest that combined treatment with both drugs may extend patient's life span. However, these observations are preliminary and need testing in further clinical trials. Clinical trial registration: EudraCT 2016-001968-11; NCT03305913 (ClinicalTrials.gov).

An in vitro study of ophthalmic antiviral agent toxicity on rabbit corneal epithelium.

Using an in vitro system we measured the corneal epithelial cytotoxicity and the antiviral activity of the antiviral agents idoxuridine (IDU), trifluridine (TFT), ethyldeoxyuridine (EDU), and (E)-5-(2-Bromovinyl)-2'-deoxyuridine (BVDU). Confluent rabbit corneal epithelial cell cultures were established, and the antiviral agents were added for 5, 30, or 60 min at a range of concentrations including that used clinically (IDU 0.1%, TFT 1.0%, BVDU 0.1%, EDU 2.0%). Twelve hour [3H]thymidine incorporation then was measured and expressed as % inhibition of control cultures. In separate experiments confluent corneal epithelial cell monolayers were inoculated with 10(4) plaque forming units (PFU) of HSV type 1 (McKrae strain) for 1 h, and IDU 0.1%, TFT 1.0%, and BVDU 0.1% were added to the culture for determination of PFU inhibition. Significant dose-, but not time-dependent, toxicity was observed at the clinical concentrations of IDU, TFT, and EDU. Toxicity was absent for BVDU. TFT and IDU were the most toxic, and EDU was of intermediate toxicity. IDU, TFT, and BVDU showed significant antiviral activity in this corneal epithelial cell culture system (TFT greater than BVDU greater than IDU). The results of this in vitro study paralleled the findings of previous in vivo corneal epithelial toxicity studies of IDU, TFT, and BVDU. Our data, however, suggest that EDU has a potential for clinical toxicity and further studies are recommended. Our model may be useful in the future toxicologic study of new antiviral agents.

Analysis of trifluorothymidine-resistant (TFTr) mutants of L5178Y/TK+/- mouse lymphoma cells.

Three classes of TFTr variants of L5178Y/TK+/- -3.7.2C mouse lymphoma cells can be identified--large colony (lambda), small colony (sigma), and tiny colony (tau). The sigma and lambda mutants are detectable in the routine mutagenesis assay using soft agar cloning. The tau mutants are extremely slow growing and are quantitated only in suspension cloning in microwells. Variants of all three classes have been analyzed in the process of evaluating the usefulness of the thymidine kinase locus in L5178Y/TK+/- mouse lymphoma cells for detecting induced mutational damage. 150 of 152 variants from mutagen treated cultures and 163 of 168 spontaneous mutants were TFTr when rechallenged approximately 1 week after isolation (3 weeks after induction). All of the 41 mutants assayed for enzyme activity were TK-deficient. The sigma and tau phenotypes were found to correlate with slow cellular growth rates (doubling time greater than 12 h), rather than from effects of the TFT selection or mutagen toxicity. Cytogenetic analysis of sigma mutants approximately 3 weeks after induction shows an association between the sigma phenotype and readily observable (at the 230-300 band level) chromosomal abnormalities (primarily translocations involving that chromosome 11 carrying the functional TK gene) in 30 of 51 induced mutants studied. Using an early clonal analysis of mutants (approximately 2 weeks after induction) 28 of 30 sigma mutants showed chromosome 11 rearrangements. All lambda mutants studied (17 of 17 evaluated 3 weeks after induction and 8 of 8 evaluated 2 weeks after induction) showed normal karyotypes (at the 230-300 band resolution level), including the chromosome 11s. These observations support the hypothesis that sigma (and likely tau) mutants represent chromosomal mutations and lambda mutants represent less extensive mutations affecting the TK locus. The inclusion of sigma mutants in the total induced mutant frequency, as well as distinguishing them as a separate subpopulation of TK-deficient mutants, is, therefore, essential in obtaining maximum utility of the information provided by the L5178Y/TK+/- mouse lymphoma assay.

Toxicity and mutagenicity of low dose rates of ionizing radiation from tritiated water in human lymphoblastoid cells.

In order to study the toxic and mutagenic effects of low-dose-rate exposure to ionizing radiation, human lymphoblast cells were grown continuously in tritiated water (3H2O) for up to 8 days. Dose rates ranged from 0.0054 to 0.064 rad/min. Mutation to trifluorothymidine resistance (TK locus) and 6-thioguanine resistance (HGPRT locus) was measured; comparable results were observed at both loci. The mutant fraction as a function of total absorbed dose was independent of dose rate over the range studied. At the lower doses, the dose-response curve was linear, with no indication of a threshold. Overall, it appeared to be slightly biphasic with a diminished slope at higher total doses. These data are discussed in relation to earlier studies utilizing high-dose-rate X-irradiation and incorporated [3H]TdR; 3H2O and [3H]TdR were more mutagenic per rad than X-rays, but 3H2O was less cytotoxic than either X-rays or [3H]TdR.

Mutagenic activity of fluorides in mouse lymphoma cells.

The L5178Y mouse lymphoma cell forward-mutation assay was used to test for the mutagenic activity of sodium and potassium fluoride at the thymidine kinase locus. Mutants were detected by colony formation in soft agar in the presence of trifluorothymidine. Mutagenic and toxic responses were observed in the concentration range of 300-600 micrograms/ml with both sodium and potassium fluoride. Approximately 3-fold increases in mutant frequency were observed for concentrations in the 500-700 micrograms/ml range that reduced the relative total growth to approximately 10% in the absence or presence of a rat-liver S9 activation system. A sample of 30% sodium fluoride-70% sodium bifluoride (NaHF2) induced a similar mutagenic response but was more toxic with respect to the fluoride concentration. A specificity for fluoride ions in causing mutagenesis was indicated by the fast that much higher concentrations of sodium or potassium chloride were necessary to cause toxicity and increases in the mutant frequency. The possible involvement of chromosomal changes was signaled by the predominant increase in the small colony class of mutants.

In situ analysis of trifluorothymidine-resistant (TFTr) mutants of L5178Y/TK+/- mouse lymphoma cells.

TFTr mutants of L5178Y/TK+/- mouse lymphoma cells are analyzed as they appear in situ following cloning and incubation for 9-11 days in soft agar cloning medium. These TFTr mutants can be divided by colony size into sigma, small colony, and lambda, large colony, mutants. The use of a size discriminator on an automatic colony counter allows the production of histograms to evaluate the size distribution of colonies on a plate. The evaluation of these size distribution curves provides insight into the properties of sigma and lambda mutants. From these analyses several conclusions may be drawn. The sigma phenotype is preferentially associated with the TFTr subpopulation of a treated culture. The sigma phenotype is not an artifact of delayed toxicity following treatment. The frequency of quantifiable sigma mutants is not affected by agar concentrations between 0.20% and 0.45% in the cloning medium. TFTr sigma mutants are produced spontaneously and can be induced by a variety of mutagens. The decline in overall detectable mutants frequency observed for some mutagens with increasing time after treatment is due to the decline in sigma mutant frequency. The quantitation of both sigma and lambda mutants is thus useful in obtaining maximum utility of the information provided by the L5178Y/TK+/- mouse lymphoma assay.

Chromosome analysis of small and large L5178Y mouse lymphoma cell colonies: comparison of trifluorothymidine-resistant and unselected cell colonies from mutagen-treated and control cultures.

Mutagenesis assays at the thymidine kinase (TK) locus in L5178Y mouse lymphoma cells frequently yield mutant colonies with a bimodal size distribution. The objectives of this study were to determine whether a relationship exists between mutant colony size and chromosomal aberrations and whether the colony-size distributions obtained from this assay can indicate the clastogenic activity of a test chemical. Cells from 8 different types of L5178Y mouse lymphoma cell colonies were examined for chromosomal abnormalities within 10 cell generations after colony isolation. The colonies included small (sigma) and large (lambda) unselected cell (UC) and trifluorothymidine-resistant (TFTr) colonies derived from TK +/- cell cultures treated with the solvent dimethyl sulfoxide (DMSO) or hycanthone methanesulfonate (HYC). Chromosome abnormalities were present in cells from 12% (7/60) of the UC colonies, but there was no apparent relationship between colony diameter and the presence of chromosomal abnormalities. Abnormalities affecting chromosome 11, which is believed to be the site of the TK gene, were not observed in cells from UC colonies. Abnormalities affecting chromosome 11 were observed only in cells from sigma-TFTr colonies irrespective of whether they were spontaneous (5/15 colonies) or induced by HYC (4/15 colonies). Overall, 30% (9/30) of sigma-TFTr colonies had cells with an abnormal chromosome 11 and 10% (3/30) had abnormalities affecting other chromosomes. Abnormalities affecting chromosome 11 were not observed in cells from lambda-TFTr colonies (0/30 colonies). The observation of only 30% of sigma-TFTr colonies with chromosome damage affecting chromosome 11 indicates that other mechanisms, in addition to chromosome damage at the level of resolution used in this study (i.e., 200-300 chromosome bands). contribute to small TFTr colony size.

The mutagenicity of 2-amino-N6-hydroxyadenine to L5178Y tk +/- 3.7.2C mouse lymphoma cells: measurement of mutations to ouabain, 6-thioguanine and trifluorothymidine resistance, and the induction of micronuclei.

2-Amino-N6-hydroxyadenine (AHA) was tested in the mouse lymphoma L5178Y tk +/- assay using the microtitre cloning technique over concentrations from 0.005 micrograms/ml-1 (100% viability) to 6 micrograms/ml (10% viability) as measured by cloning efficiency immediately after treatment. At low, non-toxic concentrations (0.005-0.25 micrograms/ml) a dose-related linear increase in the frequency of ouabain-resistant mutants was seen, in addition to an increase in 6-thioguanine- and trifluorothymidine-resistant mutants. No consistent induction of micronucleated cells was observed in this concentration range. Toxic concentrations (20-90% kill) induced a dose-related increase in micronuclei, while the frequency of ouabain-resistant mutants fell (although it was still highly significantly above the control value). These results suggest that the mechanism of action of AHA depends on the concentration, with point mutations being induced at low, non-toxic doses and detectable chromosome breakage occurring only at higher doses. Both large-colony and small-colony trifluorothymidine-resistant mutants were induced at all concentrations. The utility of using multiple genetic end-points in one cell line and the importance of dose range selection for risk assessment and an understanding of the mode of action of test substances is underlined.

A comparison of the soft agar and microtitre methodologies for the L5178Y tk +/- mouse lymphoma assay.

The L5178Y tk +/- mouse lymphoma assay (MLA) has been validated as a sensitive and specific test system for the detection of mutagens/clastogens. There are currently two methodologies for performing the MLA: the original soft agar procedure and the newer microtitre procedure. While both procedures are considered acceptable, a limited amount of comparative information exists for the two methods. In this report the two methods were compared with regard to: (1) spontaneous and induced mutant frequencies; (2) cloning efficiencies; and (3) colony size distributions for mutants. In addition, small and large mutant colonies from microtitre wells were rechallenged for trifluorothymidine (TFT) resistance. In a majority of the cases, cloning efficiency values were higher for the microtitre as were the spontaneous and induced mutation frequency (MF) values. Nevertheless, when responses were compared according to mutation index (fold increase over background MF) the results from the two systems were often similar. More spontaneous small colonies were observed in the microtitre assay. While colony size distribution for induced mutant colonies was compound specific, generally, more small colonies were counted in microtitre. All mutant clones that were rechallenged with TFT demonstrated resistance. Aside from the differences mentioned above, both the microtitre and the soft agar procedures appear equally capable of identifying mutagenic agents.

Mutational and pseudomutational effects of 5-bromodeoxyuridine in human lymphoblasts.

We have studied the effects of 5-bromodeoxyuridine (BrdUrd) at two genetic loci in diploid human lymphoblast cells. In thymidine kinase heterozygotes (tk +/-), a 2-h dose of BrdUrd induced a transient, non-heritable resistance to the thymidine analogue, trifluorothymidine (F3TdR). We have called this phenomenon pseudomutation and have shown that affected cells acquire the ability to survive in the presence of F3TdR and then, after degradation of F3TdR in the medium, return to an apparently normal wild-type state. Our data suggest that BrdUrd incorporation into DNA as a thymidine analogue is responsible for the effect, which we interpret as a temporary loss of thymidine kinase activity. This effect is not seen in tk +/+ homozygotes. In contrast, at the hypoxanthine-guanine phosphoribosyl transferase locus in tk +/- heterozygotes, BrdUrd did not induce a permanent, heritable resistance to 6-thioguanine (gene locus mutation). We detected such mutations only in the tk +/+ homozygote and only at external BrdUrd concentrations considerably higher than those which saturate the uptake of BrdUrd into DNA as a thymidine analogue. We postulate that the reduced TK enzyme levels (30%) in the heterozygote prevent the build-up of a sufficiently high intracellular BrdUrd triphosphate pool to promote the misincorporations as deoxycytidine triphosphate which may be responsible for gene locus mutation.

An in situ protocol for measuring the expression of chemically-induced mutations in mammalian cells.

The generation of expression curves and the evaluation of mutagenic responses of mammalian cells using standard mutagenesis assays can be inaccurate because mutant and wild-type cells are usually mixed during the expression phase. If some mutant progenitors or mutants grow more slowly than the wild-type cells during the expression period, there will be a decrease in the mutant to wild-type ratio with time and the mutant fraction will not accurately represent the number of mutational events that occurred. The mutant fraction may also inaccurately assess the number of mutations if these mutations are expressed over a number of generations during the time before selection. We previously showed that recovery of L5178Y mouse cell mutants is not complete when mutations are allowed to express in suspension because slowly growing mutants and/or mutant progenitors are diluted out during this time (Rudd et al., 1990). In order to more accurately quantitate the mutagenic response of the cells, we developed an in situ procedure which segregates and immobilizes cells during expression. Because of this immobilization, slowly growing mutant progenitors and mutants expressed at different times will have an equal probability of being scored as mutants. Thus, one mutation leads to one mutant colony and the measurement of the mutagenic response of the cells to the chemical accurately reflects the mutational events that occurred. We plated L5178Y tk+/- mouse cells in semisolid medium immediately after treatment. As the cells grew and formed microcolonies, the selective agent TFT was added as an overlay at specified times, permitting only TFTr cells to survive. In this procedure, each mutation was captured as an individual colony; consequently, the measured mutation fraction accurately reflected the mutational events that occurred at the selected locus. In addition, the induced mutant colonies arising in the agar are the result of independent mutational events. We previously described the in situ protocol for L5178Y cells and showed that the spontaneous mutation rate measured was 50-fold greater than when the cells expressed the phenotype in suspension (Rudd et al., 1990). From this we concluded that the slow growth phenotype was expressed before TFT resistance. In the present paper, we evaluate the effect of chemical treatment on the mutation fraction as a function of the time to TFT addition. Using the in situ protocol, we generated expression curves for three nucleotide analogs, 5-azacytidine, TFT and AraC. The numbers of TFTr colonies produced at various times after treatment indicated that chemically-treated cultures had higher mutation fractions than the solvent controls.(ABSTRACT TRUNCATED AT 400 WORDS)

Anti-recombinogenic and convertible co-mutagenic effects of (E)-5-(2-bromovinyl)-2'-deoxyuridine (BVDU) and other 5-substituted pyrimidine nucleoside analogs in S. cerevisiae MP1.

In experiments using yeast, without addition of an external metabolic activation system, (E)-5-(2-bromovinyl)-2'-deoxyuridine (BVDU) was co-mutagenic and showed an insignificant anti-recombinogenic effect in combination with triethylene melamine (TEM). In the presence of activating S9-mix, the anti-recombinogenicity and co-mutagenicity could clearly be seen. At higher concentrations the co-mutagenic effect was converted into anti-mutagenicity. The other three 5-substituted pyrimidine nucleoside analogs were tested only in the presence of activating S9-mix and showed similar effects. As TEM is a direct alkylating agent that is inactivated by liver microsomes, the higher activity in presence of S9-mix can be interpreted as resulting from metabolic activation of the 5-substituted pyrimidine nucleoside analogs. In previous experiments using yeast bacteria, Drosophila or mice, tumor promoters were co-recombinogenic/anti-mutagenic, and co-carcinogens were co-mutagenic/anti-recombinogenic. Thus, there is not only an operational difference between tumor promoters and co-carcinogens but a real difference in respect to their genetic effectiveness. As up to now only co-carcinogens have shown co-mutagenic and anti-recombinogenic effects, it is perhaps possible that, within a certain concentration range, 5-substituted pyrimidine nucleoside analogs may have co-carcinogenic activity in carcinogenicity tests. At higher concentrations the co-carcinogenic effect may be converted into an anti-carcinogenic one.

Retinal toxicity of combination antiviral drugs in an animal model.

We investigated the retinal toxicity of various combinations of four antiviral agents in intravitreal injection and vitrectomy infusion solutions in the rabbit. The nontoxic concentration of each individual agent for intravitreal injection and for vitrectomy infusion has previously been described. Our results support the concept that concentrations of 200 micrograms/0.1 mL of trifluridine, 400 micrograms/0.1 mL of hydroxyacyclovir, 200 micrograms/0.1 mL of acyclovir and 30 micrograms/0.1 mL of vidarabine can be combined for intravitreal injection without toxic retinal effects. Moreover, 60 micrograms/mL of trifluridine, 20 micrograms/mL of hydroxyacyclovir, 40 micrograms/mL of acyclovir and 8 micrograms/mL of vidarabine can be combined in vitrectomy infusion solutions without retinal damage. Further study is needed to determine the full potential of such combinations.

[Antiviral agents and corneal epithelium cicatrisation: study in rabbits (author's transl)]. / Antiviraux et cicatrisation de l'épithélium cornéen. Une étude expérimentale sur le lapin.

Five clinically-used antiviral drugs (3% adenine arabinoside ointment; 3% acycloguanosine ointment; 0.24% idoxuridine ointment; 1% trifluorothymidine drops) were compared with a control (petrolatum base) to determine their toxic effects on rabbit corneal epithelium after injury by iodine vapors: --Only trifluorothymidine significantly retarded healing of epithelial erosions. --Histopathologic examination after seven-day treatment showed that all five drugs, except vidarabine and to a lesser degree acycloguanosine, caused toxic changes in the regenerating epithelium.

Mode of action of trifluorothymidine (TFT) against DNA replication and repair enzymes.

Trifluorothymidine (TFT) is well known to be converted to TFT-monophosphate by thymidine kinase and to inhibit thymidylate synthase. In addition, TFT-triphosphate (TFT-TP) is also incorporated into DNA, resulting in cytocidal effects. However, the precise mechanism of TFT-induced DNA damage is still unclear. Therefore, we investigated the modes of action of TFT against DNA replication and repair enzymes, as compared with those of 5FU and FdUrd. When HeLa cells were treated with TFT at a concentration of 1 µM (IC50 value), the concentration of TFT in the DNA was calculated as 62.2±0.9 pmol/1x106 cells for 4 h. On the other hand, following treatment of the cells with FdUrd (0.5 µM) and 5FU (10 µM) at their IC50 doses, the drug concentrations in the DNA were 7.53, and 0.17 pmol/1 x 106 cells for 4 h, respectively. These results show the markedly greater degree of incorporation of TFT into the DNA of the HeLa cells compared with that of 5FU (approximately more than 300-fold for 4 h) or FdUrd (approximately more than 8-fold for 4 h). The primer extension assay demonstrated that TFT-TP was also incorporated into the T-sites of the growing DNA strand, however, it competed only weakly with thymidine triphosphate. The DNA glycosylase assay was performed using commercially available DNA glycosylase and fractionated HeLa cell extracts obtained by gel filtration. There was no detectable excision of the TFT pairing to adenine by uracil DNA glycosylase (UDG), thymine DNA glycosylase (TDG), methyl-CpG binding domain 4 (MBD4) or the fractionated HeLa cell extracts, however, TDG and MBD4 were able to excise the TFT pairing to guanine. Additional data indicate that small-interfering RNA-mediated knockdown of TDG or MBD4 significantly increased the resistance to the cytotoxic effects of FdUrd, but not to that of TFT. These studies show the greater degree of incorporation of TFT into the DNA than that of 5FU or FdUrd, and that such a high degree of incorporation of TFT residues into the DNA might be related to exhibit potent cytotoxic activity to be refractory to cleavage by these DNA glycosylases; thus, the DNA-directed cytotoxic effect of the compound is quite different from that of 5FU.

Intravitreal trifluorothymidine and retinal toxicity.

Intraocular trifluorothymidine was administered to one group of albino rabbits as an intravitreal injection, and to a second group of albino rabbits in the infusion solution during pars plana vitrectomy. The eyes of both populations were evaluated with preoperative and postoperative indirect ophthalmoscopy, electroretinography, and histologic examination. From these data, it was determined that an intravitreal injection of 200 micrograms/0.1 ml and a vitrectomy solution containing a concentration of 60 micrograms/ml trifluorothymidine were nontoxic to the rabbit eye.

Toxicity and clearance of a combination of liposome-encapsulated ganciclovir and trifluridine.

New Zealand albino rabbits were given intravitreal injections of liposome-encapsulated ganciclovir and trifluridine. Preoperatively and postoperatively, the eyes were evaluated by indirect ophthalmoscopy at different time intervals up to 14 days after injection. At intervals up to 14 days postinjection, the animals were sacrificed, the eyes enucleated, vitreous antiviral levels determined in one group of eyes, and histopathological examination conducted in the other group. The results of this study demonstrated prolonged intravitreal drug levels above the mean inhibitory dose for many strains of virus belonging to the herpes simplex virus family. No evidence of gross retinal toxicity was found by clinical or light microscopic examination of the treated eyes.

Intensive chemotherapy for metastatic neuroblastoma: a Southwest Oncology Group study.

Thirty-three children with Evans stage IV neuroblastoma were treated with an intensive chemotherapy regimen reported by Helson to be highly effective. The purpose of the study was to determine whether the toxic regimen was manageable by different investigators and to increase the sample of patients. Remission induction therapy consisted of courses repeated every four weeks: Cyclophosphamide (CTX) 80 mg/kg IV, with IV fluids, and furosemide on days 1 and 2; vincristine (VCR) 0.03 mg/kg IV 12 hours after cyclophosphamide; trifluoro-methyl-2-deoxyridine (F3TdR) 45 mg/kg IV push, and papaverine (PAP) 45 mg/kg (12-hour infusion) under cardiac monitoring on days 3 and 4. Initially during maintenance, courses of therapy were reduced to two days. Because this was found to be ineffective therapy, the courses were extended to four days. Some of the patients who achieved response were removed from the protocol and placed on different maintenance therapy. Seventeen of 21 children newly diagnosed and 6/12 children previously treated for metastatic neuroblastoma achieved partial or complete remission. Eight of 16 newly diagnosed patients achieving response are still alive, six without evidence of disease for periods of time ranging from 20 to 41 months. The median of the administered drug dosages was 100% of the recommended dosages. Seventy percent of the 229 courses given were initiated at correct interval. Therapy had to be delayed on the others because of toxicity. The value of the four-drug combination is limited because of side effects related to myelosuppression which resulted in severe complications and frequent hospitalizations.

Inhibition of human cytomegalovirus by trifluorothymidine.

The antiviral activity of trifluorothymidine (TFT) singly and in combination with other antiviral agents against human cytomegalovirus (HCMV) was evaluated by using an infectious center plaque reduction assay. The 50% inhibitory dose of TFT against six different patient HCMV strains was 0.57 (+/- 0.24, standard deviation) microM and ranged from 0.32 to 0.97 microM. The 50% inhibitory dose for the laboratory-adapted HCMV strain, AD-169, was 2.1 microM. When TFT (0.17 microM) was combined with human fibroblast interferon (25 U/ml), the combination was additive against all four HCMV isolates evaluated. Synergism was observed when TFT (0.17 microM) was combined with phosphonoformic acid (25 microM) for all strains studied or with acyclovir (20 microM) for three of the four clinical HCMV strains tested. Each of the three antiviral agents, when combined with TFT, exhibited additive effects against strain AD-169. TFT at concentrations of 0.5, 1.7, and 3.5 microM had an increasing inhibitory effect on uninfected human embryonic lung fibroblast (HEL) cell growth over 72 h, with 16% growth inhibition at 3.5 microM after 3 days. There was no increased toxicity to growing HEL cells when the paired antiviral agent combinations were evaluated. These findings suggest that TFT may be useful singly or in combination with other antiviral agents in treating HCMV infections.