GB2116169A - Anthracycline glycosides - Google Patents

Abstract

<IMAGE> The anthracycline glycosides I (X = H or OH, R1 = H or CH3, R2 = H or OCH3, R3 = H or OCH3, R2 = R3) and their pharmaceutically acceptable acid addition salts have antitumour properties. They are prepared from 4-demothoxy-daunomycinone or 2,3-dimethyl-4-methoxy-daunomycinone by condensation with an appropriate N-protected hexopyranosyl chloride and removal of the protecting group and, in the case X = OH, conversion of the 9-acetyl group to a 9-hydroxyacetyl group.

Description

SPECIFICATION Anthracycline glycosidos The invention relates to anthracycline glycosides, to methods for the preparation thereof and to pharmaceutical compositions containing them.

The invention provides anthracycline glycosides of the general forumul I

wherein X represents a hydrogen atom or a hydroxy group, R, represents a hydrogen atom or a methyl group, one of R2 and R3 represents a methoxy group and the other a R2 and R3 represents a hydrogen atom; and further provides pharmaceutically acceptable acid additions salts of such compounds.These compounds are named as follows: 4-demethoxy-4'-0-methyl-daunorubicin (la: R, = R3 = X = H, R2 = OCH3), 4-demethoxy-4'-epi-4'-O-methyl-dau norubicin (Ib: R, = R2 = X = H, R3 = OCH3), 4-demethoxy-2, 3-dimethyl-4'-0-methyj-daunorubicin (Ic: R, = CH3, R2 = OCH3, R3 = X = H), 4-demethoxy-2, 3-dimethyl-4'-epi-4'-O-methyl-daunorubicin (Id: R, = OH3, R2 = X = H, R3 = 0CH3), 4-demethoxy-4'-0-methyl-doxorubicin (le: R, = R3 = H, R = OCH3, X = OH), 4-demethoxy-4'-epi-4'-0-methyl-doxorubicin (Ig: R, = R2 = H, R3 = OCH3X = OH), 4-demethoxy-2,3-dimethyl-4'-0-methyl-doxorubicin (Ig:R, = CH3, R2 = SOCH2, R3 = H, X = OH), and 4-demethoxy-2, 3-dimethyl-4'-epi-4'-0-methyi-doxorubicin (Ih: R1 = OH3, R2 = H, R3 = OCH3, X = OH).

The anthracycline glycosides of the general formula I may be prepared by a process according to the invention which process comprises condensing 4-demethoxy-daunomycinone or 2, 3-dimethyl-4-demethoxy-daunomycinone (United States Patent Specification No. 4046878) with 2,3,6-trideoxy-3-trifluoroacetamido-4-0-methyl-L-lyXo- hexopyranosyl chloride or 2,3, 6-trideoxy-3-trifluoroaceta m ido-4-0-methyl-L-arabino-hexopyrano- syl chloride (United States Patent Specification No. 4183919) to give one of the protected aglycosides of the formula Ia to lid,

Ila: R, = R3 = H; R2 = OCH3 llb: R, = R2 = H; R3 = OCH3 lic: R1 = CH3; R2 = OCH3; R3 = H lid:R, = CH3; R2 = H; R3 = OCH3 removing the N-trifluoroacetyl protecting group by mild alkaline hydrolysis to give the corresponding daunorubicin derivative la to Id, and optionally converting the daunorubicin derivative to the corresponding doxorubicin derivative le to Ih by bromination and treatment of the resulting 1 4-bromo-derivative with aqueous sodium formate.

The conversion of the daunorubicin derivatives la to Id to the corresponding doxorubicin derivatives le to Ih follows the method described in United States Patent Specification No.

3803124.

The invention further provides a pharmaceutical composition comprising an anthracycline glycoside of the general formula I as herein defined or a pharmaceutically acceptable salt thereof in admixture with a pharmaceutically acceptable diluent or carrier.

The invention is illustrated by the following Examples and biological data.

Example 1 4-demethoxy-4'-0-methyl-daunorubicin (la) A solution of 3.68 g of 4-demethoxy-daunomycinone in 400 ml of anhydrous methylene dichloride containing 1.49 of 1-chloro-4-O-methyl-N-trifluoroacetyl-daunosamine was vigorously stirred in the presence of molecular sieve (30 g, 4A Merck) and 1.39 of silver trifluoromethanesulphonate. After 10 minutes at room temperature, the reaction mixture was neutralized with 0.55 ml of symcollidine. After 40 minutes the suspension was filtered and the organic phase was washed with 0.01 N aqueous solution of hydrochloric acid, with water, with a saturated aqueous solution of sodium bicarbonate and finally with water to neutrality.The residue, obtained by evaporating off the solvent under vacuum, was dissolved in 1 50 ml of acetone, treated with 600 ml of 0.2 N aqueous sodium hydroxide and diluted with 450 ml of water.

After 5 hours at 0 C the solution was adjusted to pH 8.5 and extracted with chloroform until the chloroform extracts were no longer coloured. The organic extracts were combined and acidified to pH 5 with 0.1 N methanolic hydrogen chloride. By evaporation under vacuum to a small volume (150 ml) pure 4-demethoxy-4'-O-methyl-daunorubicin (0.69) crystallized. The mother liquor was purified on a column of silica gel buffered at pH 7 with phosphate buffer M/15, using the solvent system chloroform:methanol: water (10:2:0.2) by volume). The eluate, containing the pure compound: was diluted with water and the organic phase was separated off, washed with water and evaporated to a small volume and thereafter acidified to pH 5 with 0.1 methanolic hydrogen chloride.A further amount (0.4 g) of 4-demethoxy-4'-0-methyldaunorubicin hydrochloride was obtained: m.p. 189-190 C (with decomposition). TLC on Kieselgel plates (Merck F 254) solvent system chloroform:methanol:water (10:2:0.2 by volume).

H PLC: experimental analysis: column microbondapack C1 ; mobile phase: water:acetonitrile (69:31 by volume) at pH 2 with 10% orthophosphoric acid: flux rate 1.5 ml/min., retention time 22 min.

FD-MS: m/z 511 (M + -) Example 2 4-demethoxy-4'-0-methyl-doxorubicin (le) A solution of 0.5 g of 4-demethoxy-4'-0-methyl-daunorubicin, prepared as described in Example 1, in a mixture of methanol (8 ml) and dioxan (20 ml) was treated with bromine to form the 1 4-bromo derivative. Treatment of the 14-bromo derivative with an aqueous solution of sodium formate at room temperature for 18 hours gave 0.310 g of 4-demethoxy-4'-O-methyldoxorubicin which was isolated as its hydrochloride m.p. 164-1 65 C (with decomposition); TLC on Dieselgel plate (Merck F 254) solvent system chloroform:methanol:water (10:2:0.2 by volume); Rf 0. 18.

HPLC: experimental conditions: column microbondapack C,8; mobile phase water:acetonitrile (69:31 by volume) at pH 2 with 10% orthophosphoric acid; flux rate 1.5 ml/min; retention time; 10 min.

Example 3 4-demethoxy-2, 3-dimethyl-4 '-0-methyl-da unorubicin (Ic) The coupling reaction between 4-demethoxy-2,3-dimethyl-daunomycinone and 1-chloro-4-0methyl-N-trifluoroacetyl-daunosamine under the conditions described in Example 1 afforded the title compound.

Example 4 4-demethoxy-2, 3-dimethyl-4'-O-methyl-doxornbicin (lg) The conversion of compound Ic to the title compound was performed using the procedure described in Example 2.

Example 5 4-demethoxy-4 '-epi-4 '-O-meth yldaunorubicin (lb) and 4-demethoxy-2, 3-dimethyl-4 '-epi-4 '-0-me- thyl-daunorubicin (Id) The coupling reactions, as described in Example 1, of 4-demethoxy-daunomycinone and 4demethoxy-2, 3-d i methyl-daunomycinone with 2,3, 6-trideoxy-3-trifluoroacetam ido-4-0-methyl-Larabino-hexopyranosyl chloride afforded, after hydrolysis of the N-protecting group, the title compounds.

Example 6 4-demethoxy-4'-epi-4'-0-methyidodorubicin (If) and 4-methoxy-2, 3-dimethyl-4 '-epi-4 '-0-meThyl- doxorubicin (Ih) The compounds Ib and Id were converted, via their 1 4-bromo-derivatives under the conditions described in Example 2 to compounds If and Ih respectively.

BIOLOGICAL ACTIVITY OF la and le The compounds la and lg have been tested against the parent compounds, respectively daunorubicin (DNR) and doxorubicin (DX), in several experimental systems, in order to ascertain their cytotoxicity, antitumour activity and caridac toxicity in expermental animals.

Data reported in Table 1 show that la is about 5 times more cytotoxic than DNR, and le is about 9 times more cytotoxic than DX.

The primary screening in vivo was carried out in CDF-1 mice bearing P383 ascitic leukemia (108 cells/mouse). Results are reported in Table 2. Both la and le were found to be more toxic and more potent than the parent compounds. Comparison at the Maximal Tolerated Dose (MxTD) shows that la is as effective as DNR (giving similar increase of the mice life span) and le has a good antitumour activity, which is of the same order of magnitude as that of DX.

Several studies have been carried out in C3H mice bearing the Gross luekemia injected i..

(2x106 cells/mouse). Data on la are reported in Table 3. Administeed i.v. on day 1 after the tumour inoculation, la was markedly more toxic and more potent than DNR. At the MxTD of 1.25-1.3 mg/kg, la was more effective than DX at the MxTD of 10 mg/kg. It is well known that DNR is not active when administered by oral route, unless very high dose are given (350 mg/kg). Data reported in Table 3 show that la has a good antitumour activity against Gross leukemia also when given orally. Administered orally on day 1 only, la is active at the dose of 1.25-1.3 mg/kg which is also the optimal dose in the case of the i.v. treatment. This result suggests that absorption of la through the gastrintestinal tract is very efficient. Administered on days 1, 2 and 3 by oral route, la is more active than when administered on day 1 only; at the optimal dose of 0.66 mg/kg/day it has an antitumour activity of the same order of magnitude as that of 4-demethoxy-daunorubicin, which was investigated in parallel at the optimal dose of 1.9 mg/kg/day. The observation that the optimal dose of la is lower than that of 4-demethoxydaunorubicin suggests a more efficient absorption through the gastrointestinal tract also in comparison to this compound.

Data on the antitumour activity of le in comparison with DX against Gross leukemia are reported in Table 4. The compounds were administered on day 1 after the tumour cells inoculum; DX was given i.v.; le was given i.v. and orally. When administered i.v., at the MxTD of 1 mg/kg, le showed a good antitumour activity, similar to that observed after DX treatment.

In addition, le was active also when administered orally at doses from 1.3 mg/kg on.

le was tested against L 1210 leukemia, which was inoculated in CDF-1 mice i.p. (ascitic form) or i.v. (105 cells/mouse). Treatment was performed on day 1 after the tumour cell inoculation, i.p. or i.v., respectively. Data reported in Table 5 show that, in the i.p.-i.p.

experiment, le at the MxTD of 0.83 mg/kg was as active DX at the MxTD of 4.4 mg/kg. In the i.v.-i.v. experiment, le at the tolerated doses of 1 and 1.3 mg/kg showed antitumour activity superior to that of DX.

In order to assess the antitumour activity against a solid tumour, Ig has been tested in comparison with DX, against the mammary carcinoma of C3H female mice. A third generation tumour transplant was utilized. Treatment started 1 5 days after the tumour transplant, and was performed i.v. once a week for 4 weeks. Tumour measurement was performed by caliper every week. Non-tumour bearing mice were treated in parallel, in order to evaluate the general toxicity and the cardiac toxicity, which was investigated in 5 mice, treated with the highest doses of the two compounds, and killed 5 weeks after the last treatment. The results of this experiment are reported in Table 6. DX was very effective, and it inhibited tumour growth by 93-94% (in comparison with the untreated controls) at both the doses tested (6 and 7.5 mg/kg).In the nontumoured mice treated with DX at 7.5 mg/kg, toxic deaths were observed (2/3) and all the mice examined showed histologically detectable heart lesions. le at the dose of 0.4 mg/kg was slightly active; at the doses of 0.6 and 0.75 mg/kg it markedly inhibited the tumour growth. In non-tumoured mice treated with le at 0.75 mg/kg no atrium lesions were observed, and only 2 out of 5 mice showed detectable ventricle lesions, which were less severe than those observed after DX treatment.

The data here presented show that la and le are new anthracycline analogues endowed with very interesting biological properties. In comparison with DNR, la is about 3 times more potent when administered i.p., and about 8 times more potent when administered i.v. At the MxTD it has an antitumour activity against ascitic P388 and systemic Gross leukemia equal to that of DNR. In addition, it is active also when administered by oral route, particularly when treatment is performed for 3 consecutive days, at doses lower than those of 4-demethoxy-daunorubicin.

le is about 10 times more potent than DX in vivo. Comparison at the MxTD shows that it is, in respect to DX, equally active against ascitic P 388 and L 1210 leukemia, systemic Gross leukemia and solid mammary carcinoma, and it is more effective than DX against systemic (i.v.

injected) L 1 210 leukemia. In addition, it active against Gross leukemia also when administered orally, and in a preliminary cardiotoxicity test in C3H mice treated chronically i.v. it caused only minimal cardiac lesions.

Compound la has been studied on P388 leukemia cells resistant to doxorubicin (P388/DX) in vitro and vivo. P388 leukemia cells resistant to doxorubicin (DX) (received by dr. Schabel) are maintained by serial transfer in mice treated with DX i.p. For experimental purpose, BDF-1 mice were injected with 104 leukemia cells, i.p., and treated i.p. on day 1 after the tumour inoculation. Results, reported in Table 7, show that daunorubicin (DNR) was not active against this tumour, while compound la, at the optimal dose of 0.8 mg/kg increased the life span of the treated mice. P388 and P388/DX leukemia cells were harvested from mice ascitic fluid and adapted to grow in suspension in vitro.Cytotoxicity tests were carried out exposing the cells to various drug concentrations for 48 hrs; at the end of the exposure period, cells were counted with a Coulter Cell Counter, and the ID50 (dose which gives 50% reduction of the cell number in comparison with untreated controls) was calculated. Results, reported in Table 8, show that la was about twice as cytotoxic as DNR on P388 leukemia cells, and was very active also on P388/DX leukemia cells, while DNR was 163-152 times less active on the resistant than on the sensitive line. Compound le was further investigated on mammary carcinoma of C3H mice.

Mice bearing measurable tumour (3rd generation transplant) were treated once/week for 4 weeks i.v. with le or with DX. Normal mice were treated in parallel, for evaluation of toxicity.

Results, reported in Table 9, confirm that le was about 10 times more potent than DX, and had a remarkable antitumour activity in this experimental system at non toxic doses, while DX was toxic.

Because of the good antitumour activity against mammary carcinoma, compound le was tested against two solid tumours: the colon 26 and the colon 38 adenocarcinomas, transplanted s.c. in BALB/c mice and in BDF-1 mice, respectively. Treatment started on day 1 after the tumour inoculation (early) or when tumour was already palpable (advanced), and was performed i.v. once/week or every 6 days, for 3 or 4 times. Tumour growth was assessed by caliper measurement.

Non-tumour bearing mice were treated in parallel, for toxicity evaluation, and were observed for 90 days. Results of three experiments are reported in Table 10. Against early colon 26, le at the maximal tolerated dose of 0.9 mg/kg/day was more active than DX at the maximal tolerated dose of 7.5 mg/kg/day. Against advanced colon 26, le at the maximal dose tested of 0.7 mg/kg/day was not toxic,- and gave a higher tumour growth inhibition than DX at the maximal tolerated dose of 6 mg/kg/day. Against advanced colon 38, le at the maximal dose tested of 0.9 mg/kg/day was as active as DX 9 mg/kg/day in inhibiting tumour growth, but produced a higher increase of survival time.

In conclusion, all data here reported confirm that compounds la and le are extremely interesting new anthracyclines, endowed with high potency, activity by oral route, activity against anthracycline resistant tumours (la), activity against solid tumours superior to that of DX and not cardiotoxic (le).

Table 1-Colony inhibition test against Hela cells in vitro (Treatent for 24 hrs) Compound Dose I D50 (ng/ml) .a (ng/ml) DNR 12.5 42 6.2 66 12 3.1 121 la 25 0 6.2 0 2.5 1.5 73 DX 12.5 20 6.2 79 9 3.1 177 le 10 0 2.5 9.9 1 0.62 82 0.15 84 aNo. of colonies; % of untreated controls TABLE 2. Antitumour activity against P388 leukemia Treatment i.p. on Day 1 Compound Dose T/Ca LTSd Toxic (mg/kg) % deathsC DNRd 2.9 154 0/10 0/10 4.4 140,154 0/20 4/20 6.6 109,163 0/20 13/20 la 0.5 127 0/10 0/10 0.64 127 0/10 0/10 0.8 172,127 0/20 1/20 1.0 145 0/10 0/10 1.3 95 0/10 10/10 DX 4.4 180 0/10 0/10 6.6 200 2/10 0/10 10.0" 315 4/10 0/10 19 0.44 180 0/10 0/10 0.66 215 0/10 0/10 1.0 250 0/7 1/7 1.5 245 2/10 4/10 a Median survival time; % over untreated controls b Long term survivors ( > 60 days) Evaluated on the basis of autoptic findings on dead mice d Data of two experiments "This is the Maximal Tolerated Dose of DX in this experimental system TABLE 3.Activity of la against Gross leukemia Treatment T/Cb Toxic Route Schedulea Compound mg/kg/day % deathsc i.v. 1 DNR 10 133,150 0/20 15 175,175,166 3/30 22.5 208,191,216 6/30 i.v. 1 la 0.58 116 0/10 0.76 133 0/10 1.0 158,166 0/20 1.25-1.3 183,208 0/20 1.56 100 9/10 1.95 116 6/10 oral 1 1.1 133 0/10 1.25-1.3 133,166 1/20 1.56-1.6 166,166 3/20 1.95-2.0 175,183 2/20 oral 1,2,3 4- 1.31 125 0/20 demethoxy 1.58-1.46 133,1 50 1/20 DNR 1.9 133,190 1/20 2.5 210 2/10 3.3 140 6/10 oral 1,2,3 la 0.44 133 0/9 0.66 183 0/10 1.0 200,220 9/20 1.25 140 8/10 "Days after tumour inoculation b.c see Table 2.

Table 4. Activity of 19 against Gross leukemia Treatment8 T/Cb Toxic Route Compound mg/kg % deathsc i.v. DX 10 200 2/8 13 200,216 2/18 16.9 250,266 3/18 19 1.0 183,233 0/18 1.3 192,208 4/18 1.7 217,200 7/18 2.2 142 7/10 oral Ig 1.0 125 0/8 1.3 150 0/8 1.7 166 0/8 2.5 166 0/8 On day 1 after tumour inoculation b.c See Table 2.

Table 5. Activity of 19 against L1 210 leukemia L1210 Treatment Compound mg/kg T/Ca LTSb Toxic inoculum % deathsc i.p. i.p. DX 4.4 150 0/9 0/9 6.6 150 0/10 1/10 10.0 162 0/10 1/10 Ig 0.83 162 2/10 0/10 1.0 187 1/10 1/10 1.2 393 4/10 3/10 i.v. i.v. DX 10.0 120 0/10 0/10 13.0 120 0/10 0/10 16.9 133 0/10 0/10 19 1.0 200 0/10 0/10 1.3 173 0/10 0/10 1.7 > 580 5/10 0/10 a.bcSee Table 2.

Table 6. Activity against C3H mammary carcinoma, toxicity and cardiac toxicity of Ig in comparison with DX.

Non tumoured mice Tumoured mice Compound mg/kg/day Heart lesionse Tumour Tumour growth MSTc T/C deaths A V growth %a inhibition % b % n. G n G - - 4845 - 53 - - - - - DX 6 318 93 77 145 n.d.

7.5 277 94 74 140 2/3 4/4 1.3 5/5 1.5 Ig 0.4 1740 64 65 123 n.d.

0.6 748 85 76 143 n.d.

0.75 344 93 81 153 1/4 0/5 0 2/5 0.2 aTumour weight on day 43/tumour weight at beginning of treatment, x 100.

b100-[Tumour growth of treated mice/tumour growth of controls, x 100] cMedian survival time (days) dObserved for 90 days eA=atrium; V=ventricles; n.=number of hearts showing lesion/total; G: lesions grade Table 7. Effect on doxorubicin-resistant P388 leukemia in vivo.

Dose T/C" Toxic Compound (mg/kg) % LTSb deaths" DNR 2.9 93 0/10 0/10 4.4 87 0/10 1/10 6.6 84 0/10 3/10 la 0.53 133 0/10 0/10 0.8 142 1/10 0/10 1.2 106 0/10 4/10 a Median Survival time; % over untreated controls.

bLong-term survivors ( < 60 days) "Evaluated on the basis of autoptic findings on dead mice Table 8. Effect on sensitive and doxrubicin-resistant P388 Leukemia in vitro.

ID50 (ng/mlb) Compound P388" P388/DXd Re DNR 5.8,2,3 950,350 163,152 la 0.35,1.3 1.4,5 4,3.8 aData of two experiments.

bDose giving 50% reduction of cell number in comparison with untreated controls.

cP388 leukemia cells sensitive to DX dP388 leukemia cells resistant to DX eRatio between ID50 on P388/DX and ID50 on P388 Table 9. Activity of le against mammary carcinoma of C3H mice Tumoured mice Compound mg/kg/day Tumour T/Cb MSTC T/Cd Non Tumoured mice weighta % % deaths" (mg) - 651 115.5 0/10 DX 6.0 238 37 149 129 1/10 7.5 209 32 100 87 7/10 le 0.6 422 65 118 103 0/10 0.75 263 40 126.5 110 0/10 aEvaluated 1 week after the last treatment, by caliper measurement.

bTumour weight of treated mice/tumour weight of controls, X 100.

median survival time (days).

dMST of treated mice/MST of controls ( X 100).

BMice were observed for 90 days.

Table 10. Effect of le and DX against two transplanted colon adenocarcinomas in mice Dose %e T/Cd Toxic Tumor Stagea Scheduleb Compd. (mg/kg) inhib. % deathse Colon 26 early q7dX4 DX 6 56 217 0/10 7.5 82 224 0/10 9.3 81 161 9/10 le 0.6 81 258 0/10 0.75 85 227 0/10 0.9 93 246 1/10 advanced q6d X 3 DX 6 34 237 0/9 9 62 237 4/9 le 0.6 48 232 0/9 0.7 52 195 0/9 Colon 38 advanced q7d X 4 DX 6 65 92 0/10 9 83 144 1/10 le 0.6 55 133 0/10 0.75 60 129 0/10 0.9 81 184 0/10 aTime of start of treatment in respect to tumour development.

bDays of i.v. administration.

c% inhibition of tumour growth, as compaed with untreated controls.

dMedian survival time of treated mice/median survival time of controls, X 1 00.

"Evaluation in non-tumoured mice treated in parallel and observed for 90 days.

Claims (11)

1. An anthracycline glycoside having the geneal formula I as herein defined or a pharmaceutically acceptable salt thereof.

2. 4-Demethoxy-4'-O-methyl-daunorubicin.

3. Demethoxy-4'-O-methyl-doxorubicin.

4. 4-Demethoxy-4-'epi-4'-O-methyl-daunorubicin.

5. 4-Demethoxy-4'-epi-4'-O-methyl-doxoru bicin.

6. 4-Demethoxy-2,3-dimethyl-4'-O-methyl-daunorubicin.

7. 4-Demethoxy-2,3-dimethyl-4'-0-methyl-doxorubicin.

8. 4-Demethoxy-2, 3-dimethyl-4'-epi-4'-O-methyl-daunorubic

9. 4-demethoxy-2,3-dimethyl-4'-epi-4'-0-methyl-doxorubicin.

10. A process for the preparation of an anthracycline glycoside having the general formula I as herein defined, the process comprising condensing 4-demethoxy-daunomycinone or 2,3dimethyl-4-demethoxy-daunomycinone with 2,3, 6-trideoxy-3-trifluoroacetamido-4-0-methyl-L- lyxo-hexopyranosyl chloride or 2,3, 6-trideoxy-3-trifluoroacetamido-4-0-methyl-L-arabino-hexopy- ranosyl chloride, and removing by mild alkaline hydrolysis the N-trifluoroacetyl protecting group from the resultant anthracycline glycoside Ila, llb, llc or lid as herein defined, and optionally converting the resultant daunorubicin derivative la, Ib, Ic or Id as herein defined to the corresponding doxorubicin derivative le, If, Ig or Ih as herein defined by bromination and treatment of the resultant 1 4-bromo-intermediate with aqueous sodium formate.

11. A process according to claim 10 in which the condensation of the anthracyclinone with the hexopyranosyl chloride is carried out in a solvent in the presence of a soluble silver salt and molecular sieve.

1 2. A process according to claim 11 in which the solvent is dichloromethane.

1 3. A process according to claim 11 or claim 1 2 in which the soluble silver salt is silver trifluoromethanesulphonate.

1 4. A process according to claim 10 the process being substantially as described herein with reference to any of Examples 1, 3 and 5.

1 5. A process according to claim 10 the process being substantially as described herein with reference to any of Examples 2, 4 and 6.

1 6. A pharmaceutical composition comprising an anthracycline glycoside having the general formula I as herein defined or a pharmaceutically acceptable salt thereof in admixture with a pharmaceutically acceptable diluent or carrier.

1 7. An N-protected anthracycline glycoside having the formula Ila, llb, llc or lid as herein defined.