EP0581906A1 - ABROGATION OF VIRAL RESISTANCE TO NUCLEOSIDE ANALOGUES BY DOUBLE-STRANDED RNAs - Google Patents

Abstract

The rapidity of viral resistance developed during treatment with antiviral nucleoside analogues is reduced by administering double-stranded RNA during the first stage of treatment of the infection or during later stages, when the viral genetic mutation has appeared, in order to restore the susceptibility of the virus to antiviral agents which would otherwise be ineffective. Delay and / or decrease in the appearance of retroviruses resistant to nucleoside analogs, in particular HIV, is achieved by means of poorly paired RNAs, in particular in peripheral mononuclear blood cells, in particular CD4 lymphocytes.

Description

ABROGATION OF VIRAL RESISTANCE TO NUCLEOSIDE ANALOGUES BY DOUBLE-STRANDED RNAs

Nucleoside analogues are commonly employed antiviral agents, particularly against retroviruses. Viruses undergo genetic changes, or mutations, leading to relative resistance to these antiviral agents. When resistance occurs, the viruses multiply more quickly and the underlying disease accelerates. By deploying dsRNAs relatively early in the

infection, the rate of emergence of viral resistance is reduced. Even later in the infection when genetic mutation has already occurred, dsRNAs restore

susceptibility of the viruses to otherwise

ineffective antiviral agents. This can be seen clinically by the unexpected result that long term use of the two modalities, in combination, causes greater recovery of host immune function, and less detectable virus, than is seen with either class of antiviral agent applied by itself.

Prolonged therapy of viral diseases,

particularly retroviral disorders, is associated with emergence of viral resistance (references 1 and 2). Ideally, therapeutically employed the nucleoside analogues are incorporated into viral genetic information which thereby becomes faulty or

incomplete, leading to a reduction in efficiency of the viral growth cycle. Defective, or incomplete, viral progeny are formed of reduced infectivity potential. However, by modifying its genetic makeup, the virus may emerge relatively resistant, thereby generating infectious progeny even in the presence of nucleoside analogues. Typical genetic changes occur in the polymerase gene (i.e., the viral component which directs incorporation of the antiviral

nucleoside in the first place) allowing the virus to escape from these forms of antiviral blockade. The best studied case to date is the interaction between retroviruses, e.g., HIV (human immunodeficiency virus) and 3'-azido-3'-deoxythmidine, also called AZT or zidovudine. The ιrtutation(s) leading to resistance to AZT often confers simultaneous resistance to various other drugs, such as (but not limited to), dideoxyinosine (DDI) and dideoxycytidine (DDC).

Disclosed are procedures for delaying, reducing or both delaying and reducing the appearance of nucleoside analogue resistant virus in a patient having an HIV infection (HIV positive) by

administering an effective amount over a suitable time of a mismatched dsRNA prior to therapy with a nucleoside analogue. These procedures serve to sensitize the patient to the later, in terms of the course of the HIV infection, administration of a nucleoside analogue antiretroviral agent if and when it is required.

Also described are therapeutic procedures for ameliorating the morbidity of the peripheral blood mononuclear blood cells (PBMC), notably the T4 or CD 4 lymphocytes, of a patient infected with a retrovirus that has become resistant to nucleoside analogues by administering an effective amount of a mismatched dsRNA to the patient.

The dsRNA may be a complex of a polyinosinate and a polycytidylate containing a proportion of uracil bases or guanidine bases, e.g., from 1 in 5 to 1 in 30 such bases (poly I · poly(C_4-29×>U or G)).

The dsRNA may be of the general formula

rI_n·r(c_11-14,U)n or rI_n.r(C₁₂,U)n. other suitable examples of dsRNA are discussed below.

By "mismatched dsRNA" are meant those in which hydrogen bonding (base stacking) between the

counterpart strands is relatively intact, i.e., is interrupted on average less than one base pair in every 29 consecutive base pair residues. The term "mismatched dsRNA" should be understood accordingly.

The mismatched dsRNAs preferred for use in the present invention are based on copolynucleotides selected from poly (C_n,U) and poly (C_n,G) in which n is an integer having a value of from 4 to 29 and are mismatched analogs of complexes of polyriboinosinic and polyribocytidilic acids, formed by modifying rI_n.rC to incorporate unpaired bases (uracil or guanidine) along the polyribocytidylate (rC_n) strand. Alternatively, the dsRNA may be derived from poly(I).poly(C) dsRNA by modifying the ribosyl backbone of polyriboinosinic acid (rl_n), e.g., by including 2'-O-methyl ribosyl residues. The mismatched complexes may be complexed with an

RNA-stabilizing polymer such as lysine and

cellulose. These mismatched analogs of rl_n·rC , preferred ones of which are of the general formula rI_n·(C_11-14'U)_n or rI_n1·r(C₂₉'G)_n, are described by Carter and Ts'o in U.S. Patents 4,130,641 and

4,024,222 the disclosures of which are hereby incorporated by reference. The dsRNAs described therein generally are suitable for use according to the present invention. The preferred mismatched dsRNA is rl_n·(C_11-14,U)_n or AMPLIGEN^® of HEM

Research, Inc. of Rockville, MD, USA, available as a lyophilized powder.

Other examples of mismatched dsRNA for use in the invention include: - poly (I) · poly (C₄,U)

poly (I) · poly (C₇,U)

poly (I) · poly (C₁₃,U)

poly (I) · poly (C₂₂,U)

poly (I) · poly (C₂₀,G)

poly (I) · poly (C₂₉ , G) and

poly (I) · poly C_p23 G>p

Another class of dsRNAs suited to the practice of this invention are short dsRNAs of defined structure, for example oligonucleotides of the formula: 5'lock-(I)_n-lock 3'

lock-(C)_m- lock 5' where m and n are each more than 5 and less than 100, I is inosine monophosphate, C is cytidine

monophosphate, and where the locks in one strand are complementary to locks in the opposite strand, or an oligonucleotide of the structure:

5'lock-[(I)_xA]_j-lock 3'

3'lock-[(C)_yU]_k-lock 3' where x and y are each more than 5 and less than 25, j and k each at least 1 and less than 10, I and C are as identified above, A is a nucleotide which is not I, and U is a nucleotide which base pairs with A.

Alternatively, the short oligonucleotide may have the structure:

5'(I)_n-hinge-(C)_m3' where n, m, I and C are as defined above.

These oligonucleotides may have substitutions in one strand not complementary to nucleotides in the opposite strand. Preferably these oligonucleotides are stabilized by internal registers of complementary heteropolymer and desirably the lock or hinge or both contain regions of complementary heteropolymer.

These oligonucleotides desirably have single-stranded tails. These oligonucleotides are described in

PCT/US89/02172.

Patients are treated with intravenous infusions of 200 to 700 mg of rI·r(C_11-14'U) as required, e.g., once a week to as often as daily in accordance with their clinical improvement. The amount of dsRNA administered and the frequency of administration will provide a level of from 0.01 to 1,000 micrograms of dsRNA per milliliter of the patient ' s systemic blood circulation immediately following administration measured at a point distal from the point of infusion.

Illustrative nucleoside analogue antiviral agents include Zidovudine (azidothymidine, RETROVAR^® or AZT as commonly used herein) which is

3'-azido-3'-deoxythymidine, a nucleoside analogue antiviral for the systemic treatment of acquired immunodeficiency syndrome (AIDS) and AIDS-related complex (ARC) caused by human immunodeficiency virus (HIV, HTLV-I, HTLV-II, HTLV-III, LAV, ARV and the like designators for various strains). The usual adult dose is 200 mg every four hours, around the clock. For a 70 kg patient, the corresponding dose is 2,9 mg per kg of body weight every four hours.

Doses up to 60 mg per kg of body weight daily have been used. Usually less than the normal and

customary amounts of the nucleoside analog retroviral is used when coadminstered with a mismatched dsRNA, as illustrated further in the discussion that follows.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a table showing the time in months to death or "full blown" AIDS for 298 patients ins terms of proportion of the patients free of critical

HIV-related events comparing with AZT treatment

"early" (square boxes) before symptoms occurred or "late" (circles) after symptoms occurred.

Fig. 2 compares the early use of a dsRNA or its concurrent use with AZT in controlling retroviral growth as compared to rapid retroviral growth in the case of AZT alone.

Fig. 3 compares the relative effects of a dsRNA and AZT as monotherapies with placebo and dsRNA and AZT as a combinational therapy in long term

maintenance of CD4 cells in HIV disease in median change in CD4T lymphocytes over the indicated number of weeks.

Fig. 4 relates the number of days of

combinational (AZT and dsRNA) therapy to the percent change in mean T4 level showing that the dsRNA

Ampligen^® extends the period of T4 cell stabilization in HIV disease over that expected with AZT alone.

Median does for each member' of the combination are indicated. Fig. 5 is a graph relating the number of days on the combined Ampligen^® and AZT regimen to the mean percent change (increase) in T4 cells showing that Ampligen^® increases and/or stabilizes T4 cell levels in HIV disease beyond the time period in which AZT is effective.

Fig. 6 is a graph relating the proportion of HIV patients free of critical events over a period of 12 months for placebo, AZT alone, Ampligen^® monotherapy and the combinational therapy of Ampligen^® and AZT.

In Figures 4 and 6 the number of patients is indicated near the relevant data point in

parenthesis, in Figures 3 and 5 it is indicated by N=.

In Figure 2, HIV co-culture was performed as described in reference 3 using peripheral blood mononuclear (PBMC) cells from cases 1 and 3. After 4, 7 and 14 days of co-culture, PBMC were harvested and dissolved in a guanidine cyanate solution used to dissolve cells and release viral materials. HIV RNA was measured by molecular hybridization as described in more detail below.

Figure 1 shows the development of critical events (development of full blown AIDS or death) in a well-publicized Veterans' group of HIV infected subjects treated either "early" (e.g., before

symptoms occurred such as weight loss, night sweats, etc.) or "late" (i.e., after such symptoms occurred including HIV-associated infections caused by fungi or bacteria) with AZT. It is apparent from Fig. 1 that the "early" taking of AZT does not prolong life. Samples of blood taken from individuals who developed critical events show an enrichment in concentration of AZT resistant virus (hereafter referred to as AZT^R) whereas patients who remained free of critical events during the approximately 40 month observation

interval generally had more AZT sensitive (AZT⁵) virus.

Below, I characterize more fully the relative

AZT sensitivities of typical HIV isolates taken from individuals whose disease progresses despite AZT therapy and compare these results with viral isolates including hepatitis virus from patients whose disease is under relatively better control. It is apparent that emergence of AZT^R virus contributes a

foreshortened life span. Typical human retroviruses including HTLV-1, HTLV-2 and HTLV-3 and viruses which multiply by similar mechanisms include certain hepatitis viruses.

I describe herein an invention which ameliorates this presently unsolvable problem. The invention can be practiced in multiple ways including: (a) reducing the emergence of AZT ^R by utilizing specific classes of dsRNAs before exposure of the virus to AZT or other analogues and (b) overcoming the lethal

properties of AZT^R HIV (or other virus resistant analogues), by adding back dsRNA to the regimen. The dsRNA/nucleoside analogue regimen shows unexpected therapeutic synergy against AZT^R HIV without a corresponding synergistic toxicity and thus is a truly unique and unexpectedly useful combination of drugs which becomes life saving when used correctly.

Insight as to how combined Ampligen^®-AZT

treatment reduces HIV-viral burden was obtained by my studies in which cell free virus obtained from patients receiving either Ampligen^® or AZT

monotherapy or Ampligen^®-AZT combinational treatment were used to infect fresh human peripheral blood cells which were previously exposed to Ampligen®, AZT, or Ampligen^® and AZT. Virus from the patient who received only AZT monotherapy for more than one year was insensitive to AZT (0.5μM AZT = 4%

inhibition; ED₅₀ < 5μM) but was as sensitive to

Ampligen^® alone as wild type HIV-virus (5μg/ml.

Ampligen^® = 80% inhibition; ED₅₀ < 0.5μg/ml.). In addition, the Ampligen®-AZT combination produced greater inhibition of HIV (93%) than Ampligen^® alone. Virus from the patients who received

Ampligen^® alone or the Ampligen^®-AZT combination showed relative resistance to AZT sensitivity to Ampligen^® and even greater sensitivity to the

Ampligen^®-AZT combination. When results with many different viral isolates were compared from patients on different schedules of therapy, it became apparent that virus from patients who received Ampligen^® before exposure to AZT were more sensitive to AZT (and other nucleoside analogues as well) than the viruses derived from patients who received AZT first. TABLE 1 - DRUG RESISTANCE OF HIV ISOLATES

Ampligen^® AZT

HIV ED₅₀ ED₅₀ Combination AZT

Isolate ( ug/ml ) (ug/molar) Index^a Phenotype

H112-2 2 .03 0.5 sensitive

6910-6 3 2 0. 2 resistant

Case #1 1 0.6 0.4 partially resistant

Case #2 1 0.2 0.2 partially resistant

Case #3 .5 >5 0.2 resistant

Case #4 1 >5 0.8 resistant

Case #5 .5 >5 0.2 resistant

Case # 6 2 >5 0.3 resistant a = AZT-Ampligen^® interaction was determined by

isobole analysis (reference 4). The Combination

Index (CI) is defined as CI + (A _C/A_S) + (B_c/B_s),

where A and B are the concentrations of the drugs used in the combination treatment, and A_s and B_s are the concentrations of the drugs which, when used

alone, give the same effect as the combined

treatment. When CI < 1, the drugs are synergistic, when CI = 1, the drugs are additive, and CI > 1, the drugs are antagonistic. ED₅₀ is the in vitro dose which inhibits virus multiplication by 50%. In Fig. 2, I show a typical embodiment of the invention whereby the early use of Ampligen®, or its use concurrently with AZT, in fact results in substantial benefit in terms of lowering the bodily concentration of harmful virus such as retrovirus. The rapidly increasing production curve of viral RNA (ribonucleic acid) shown with the open circles is indicative of uncontrolled viral growth, whereas the two flat lines illustrate good control of virus production by the combinational approach. The open circles in Fig. 2 illustrative of rapidly developing AZT virus, are thus symbolic of the typical patient in Fig. 1 who, when given AZT alone, proceeds to advance into the more terminal stages of disease.

Certain dsRNAs, notably mismatched dsRNAs, showed synergistic inhibition of retroviruses, when combined with AZT (or other nucleoside analogues) regardless of the drug resistance phenotype (Table 1). H112-2 and 6910-6 are well characterized HIV prototype clones displaying typical sensitivity and resistance, respectively. Whereas, cases #1 through #6 are viruses isolated from specific patients exposed to various regimens (e.g., patients # 1, 2, and 3 are the same as those in Fig. 2, and patients 4, 5 and 6 were treated initially with AZT alone as suggested by Fig. 1).

When I utilized the combination of drugs together clinically, the results were again

unexpected and not at all predictable from the observed clinical results of the two drugs given singly (monotherapy). For example, AZT (and other analogues) given alone results in a transitory increase in certain immune cells (reference 5), termed CD4 cells, which are among the immune cells most favored for attack by certain retroviruses and certain herpes viruses (especially HHV 6). After the transitory increase (seen at around 12-16 weeks), the immune cell number deteriorates as HIV proliferates and AZT^R HIV emerges (Fig. 3). The number in

parentheses refers to number of subjects studied. By comparison, Ampligen^® alone causes a horizontal

"line" of CD4 cell number over time (Fig. 30, e.g., the immune cell number neither increases nor

decreases (in other words, cell number is stabilized) at Ampligen® doses of approximately 400 to 500 mg given twice weekly by IV infusion.

Patients receiving AZT monotherapy took 200 mg. orally every 4 hours, daily (1200 mg./day). Patients in the Ampligen^® monotherapy group received a minimum of two IV infusions weekly (mean dose range of

463-555 mg. weekly). [Patients receiving the

combinational treatment shown in Fig. 4 were also infused with typical Ampligen^® doses of 400 mg. twice weekly with an average concomitant AZT daily dose ranging between 300 and 540 mg].

The trends in the median change in T4

lymphocyte for the placebo and AZT monotherapy treatment (after 16 weeks) groups are consistent with evidence that the course of the disease is inexorably downhill. Without effective therapeutic intervention, the median change in placebo patient lymphocyte counts declines. After an initial increase, even patients receiving AZT monotherapy experienced a deterioration in their median T4 cell counts and AZT^R HIV appears. This decline from Week 12 onward appears to parallel that seen with the placebo patients.

Comparison of the slopes by the statistical method of linear regression lines constructed from the change in median T4 cells from week 12 onward for the AZT monotherapy and placebo groups showed they were significantly different (p < 0.01 and p < 0.01, respectively) than a slope of 0 which represents "no change" in median T4 cell levels. "No change" typically represents a stabilization of immunologic disease, a "negative slope" represents disease deterioration, but the most desirable outcome is a positive slope suggestive of disease (immune) recovery, and this could not be achieved by the

monotherapy regimens studied.

The Ampligen® alone abrogates the expected CD4 decline after twelve weeks of AZT Monotherapy of Placebo. The slope of the serial median CD4 level regression line during Ampligen® treatment is not statistically different from a horizontal line reflecting no CD4 cell loss over time

Slope Comparison With

Horizontal (no change) p Value

Ampligen^® not significant p > 0.2 AZT significant p < 0.01

Placebo significant p < 0. 01

Ampligen® monotherapy (Fig. 3) abrogates the severe decline in median T4 lymphocytes seen in both placebo patients and in AZT treated patients (after a transitory twelve week rise) but the average T4 level did not increase. Moreover, in the Ampligen^® group, the small (statistically insignificant) decline in., median T4 lymphocytes observed at one year (Fig. 3) can be readily reversed thereby bringing about an even more durable T4 stabilization by increasing the dose to above 200 mg. twice weekly (data not shown). The slope of a regression line constructed from the change in median T4 cells observed from week 12 onward in the Ampligen^® monotherapy group was not significantly different from O thereby indicative of disease stabilization. The Ampligen^® monotherapy, AZT monotherapy, or placebo treated groups shown in Fig. 3 had approximately equivalent median and mean absolute T4 (also called CD4) lymphocyte levels. I then compared them with a new group that received combined Ampligen® and AZT treatment. The new group had in fact lower (approximately 33%) median and absolute levels of this prognostic indicator of the progression of HIV infection, thus indicating they were at greater risk of death or other "critical events". The relative effects of Ampligen® and AZT in long term maintenance of CD4 cells per mm³ in HIV disease is shown in the following table.

TABLE 2

Baseline Median Absolute

Number of CD4 Cell Change

Treatment Patients Median Mean at 1 Year

Ampligen^® + AZT 11 201 248 +15

Ampligen^® alone 54 311 350 -15 AZT alone 260 350* 355 -28* Placebo 253 350* 356 -59*

*Estimate based on summary data in Ref. 5

As noted in Fig. 3, on completion of 1 year of treatment, median absolute T4 lymphocyte counts declined 59 cells (approximately 4.9 cells per month) in the patients who received placebo; declined 28 cells (approximately 2.3 cells per month) in patients who received AZT monotherapy; and declined slightly, 15 cells (approximately 1.25 cells per month), in patients who received Ampligen^® monotherapy.

However, the group that received the combinational Ampligen^® and AZT therapy, which had the lowest pretreatment median and mean absolute T4 lymphocyte levels (therefore being at the greatest risk) actually experienced an increase in median absolute T4 cell level of 15 cells even after one year of combination regimen. These results are summarized in Table 2 and provided in more detail in Fig. 4.

Figure 4 demonstrates the serial changes in median T4 lymphocyte levels during the 1 year of combination treatment which should be compared with Ampligen^® monotherapy, AZT monotherapy, and placebo effects of Fig. 3. The baseline for comparison is calculated from the average of the patients' serial absolute T4 lymphocyte counts measured during the three month period immediately before starting the combined therapy.

The increase observed in the first ninety day period is consistent with that previously reported for the initiation of AZT monotherapy as observed in Figure 3 above. However, the subsequent increases, seen from 180 to 540 days of the combined therapy, are statistically significant and are not observed in patients receiving AZT monotherapy. (A

non-parametric analysis confirmed the statistically significant (p<0.05) increase in T4 counts at 1 year. This analysis was performed because of the possibility that the T4 levels were not normally distributed due to the relatively small sample size.) Thus, the duration of the T4 lymphocyte increase is longer than the 3-6 month transient increase expected from AZT monotherapy and must be attributed to the presence of dsRNA. In addition, this effect was seen with relatively low doses of both dsRNA and AZT. This observation is consistent with previous findings that at a dose as low as 200 mg. IV twice weekly for 1 to 4 months, Ampligen^® monotherapy stabilized the T4 cell decline expected in untreated HIV infected patients (T4 = 60-300 cells/mm³), and that this effect was sustained when patients continued this treatment for 5-8 months (Fig. 3).

A regression line (Figure 5) was constructed from the mean percentage change in T4 Lymphocyte counts over 135 to 630 days of combined dsRNA and AZT therapy. The average of each patient's serial T4 lymphocyte counts during 91-180 days of the combined treatment regimen served as the baseline for

determining the percentage changes. This baseline was selected as a point after which AZT's anticipated effect on T4 levels typically has dissipated.

Using statistical methods, the positive slope of the regression line and the slopes of the 95% confidence limits indicate that the combined

Ampligen^® and AZT therapy successfully abrogated the long-term T4 cell decline expected in untreated patients (4-6 cells/month) or those on long term (> 6 month) AZT monotherapy. Thus the combination

treatment in fact sustained T4 levels indefinitely and for much longer time periods than previously observed with AZT monotherapy and this illustrates a basic utility of my invention on nucleoside resistant viruses, especially retroviruses. Finally, Figure 6 corroborates the utility of my invention by presenting the proportion of patients progressing to an AIDS-defining opportunistic

infection/lymphoma ("critical event") while receiving placebo, AZT monotherapy, Ampligen^® monotherapy, or Ampligen^® and AZT combinational therapy.

As indicated in Figure 6, approximately 10% of the patients receiving placebo experienced a critical event during the 12-month observation period. The number of patients at the beginning, month 6 and month 12 of the study are indicated parenthetically. Of the patients receiving AZT, approximately 4% experienced a critical even during that period. Of particular note, was the avoidance of the expected tumor formation seen with AZT given alone (reference 6) by the combined treatment. Further no patient in this study with an absolute T4 lymphocyte count greater than 115 cells/mm³ on entry experienced a critical event while receiving Ampligen^® with

concomitant AZT, even though they were at great risk to do so due to markedly deteriorated immune systems before beginning the combinational regimen having a median CD4 level of only 201 at baseline (Table 2).

In concert with practicing this invention, lymphokines such as interleukins and interferons may be judiciously added after the foundation of control of viral replication is in place. REFERENCES CITED

1. Larder, B. A. et al pp. 436-441 Antimicrobial Agents and Chemotherapy. Vol. 34, March 1990.

2. Larder, B. A. et al. pp. 1731-1734, Science, Vol. 243, 1989.

3. Jackson, J.B. et al. pp. 1416-1418. Journal Clin. Microbiology, Vol. 216, 1990; Thompson, J. D. et al. pp. 371-378, Analytical Biochem. Vol. 182, 1989.

4. Berenbaum, M.C. pp. 269-333. Advances in Cancer Research Vol. 35, 1981.

5. Fischl, M. et al. p. 727-733, Annals of

Internal Medicine Vol. 112, 1990.

6. Pluda, J. M. et al pp. 276-282

Annals of Internal Medicine, Vol. 113 (number 4, 1990

Claims

WHAT IS CLAIMED IS:

1. A combinational regimen which ameliorates the appearance of nucleoside analogue resistant virus and/or ameliorates the morbidity, especially on immunological cells, of previously existent

nucleoside resistant virus.

2. A method of delaying, reducing or both delaying then reducing the appearance of nucleoside analogue resistant virus in a patient having an HIV infection, comprising administering to s aid patient an effective amount of a mismatched dsRNA prior to therapy with a nucleoside analogue antiretroviral agent.

3. A method of ameliorating the morbidity of the peripheral blood mononuclear cells of a patient infected with a nucleoside analogue resistant HIV, comprising administering to said patient an effective amount of a mismatched dsRNA.

4. The method of claim 2, wherein the patient is administered a nucleoside analogue antiretroviral agent subsequent to therapy with said mismatched dsRNA.

5. The method of claim 3, wherein said cells are CD4 lymphocytes.

6. The method of claim 2 or 3, in which the mismatched dsRNA is a polyadenylic acid complexed with polyuridylic acid.

7. The method of claim 6, in which the mismatched dsRNA is a complex of polyinosinate and polycytidylate containing from 1 in 5 to 1 in 30 uracil or guanidine bases.

8. The method of claim 7, in which the mismatched dsRNA is rI_n.r(C_11-14,U) or the

mismatched dsRNA contains reg-ions of bond breakage and exhibits the favorable therapeutic ratio property of rI_n.r(C_11-14,U)_n.

9. The method of claim 6, in which the amount of mismatched dsRNA administered results in a level of from 2 to 1,000 micrograms of the mismatched dsRNA per milliliter of the patient's systemic blood circulation.

10. The method of claim 2 or 3, in which the dsRNA is a short oliglnucleotide of defined structure of the formula:

5'lock-(I)_n-lock 3'

3'lock-(C)_m-lock 5' where m and n are each more than 5 and less than 100, I is inosine monophosphate, C is cytidine monophosphate, or

5'lock-[(I)_xA]_j-lock 3'

3*lock-[(C)_yU]_k-lock 3' where x and y are each more than 5 and less than 25, j and k each at least 1 and less than 10, I and C are as identified above, A is a nucleotide which is not I, and U is a nucleotide which base pairs with A, or

5'(I)_n-hinge-(C)_m3' where n, m, I and C are as defined above, provided that the locks in one strand are

complementary to locks in the opposite strand.

11. The method according to claim 10, in which the oligonucleotide is stabilized by internal registers of complementary heteropolymer and the lock or hinge or both contain regions of complementary heteropolymer.

12. The method of claim 2, 3 or 4, wherein said nucleoside analogue antiretroviral agent is zidovudine, dideoxyinosine, dideoxycytidine or combinations thereof.

13. The method of claim 2 or 3, wherein said mismatched dsRNA is rl_n.(C_11-14,U)_n.