710

Combined Treatment of Symptomatic Human Immunodeficiency Virus Type 1 Infection with Native Interferon-a and Zidovudine Ove Berglund, Katarina Engman, Anneka Ehrnst, Jan Andersson, Knut Lidman, Borje Akerlund, Anders Sonnerborg, and Orjan Strannegard

Departments of Infectious Diseases, Roslagstull and Danderyd Hospitals, Karolinska Institute, and Department of Virology, Central Microbiological Laboratory, Stockholm; Bionative AB, Umea, Sweden

Inhibitors of human immunodeficiency virus type 1 (HIV1) nucleic acid synthesis, such as zidovudine (AZT) , have so far shown limited efficacy in the treatment of chronic HIV-I disease [1-3]. Many investigators believe that a combination of drugs with different targets in the virus cycle will be the preferred treatment of HIV infection in the future. A candidate for combination therapy would be interferon-a (lFN-a), which exerts an antiviral activity on retroviruses by inhibiting virus assembly and release rather than nucleic acid replication [4]. IFN-a shows inhibition ofHIV-I replication in vitro [5, 6]. A similar antiviral effect in vivo, as evidenced by decreased HIV p24 antigen levels, has been observed in patients treated with IFN-a for Kaposi's sarcoma [7, 8]. A synergistic inhibition of HIV-I replication in vitro by AZT and recombinant IFN-a has been reported [9]. In a pilot study, Orholm et al. [10] found that a combination oflymphoblastoid IFN-a and low-dose (400 mg) AZT given to HIV-I-infected patients decreased the p24 antigen levels supposedly more than was expected from the simple additive effectsof the two drugs. Similarly, Kovacs et al. [11], who treated Kaposi's sarcoma patients with AZT and IFN-a, concluded that this combined therapy has potential in managing such patients. In the present study, native IFN-a was given to patients who

Received 22 June 1990; revised 29 October 1990. Presented in part: 2nd European Conference on Clinical Aspects of HIV Infection, Brussels, 1990. Informed consent was obtained from all patients; the study was done with the approval of the ethical committee of the Karolinska Institute, Stockholm. Financial support: Swedish Medical Council (16H-8775). Reprints or correspondence: Dr. Ove Berglund, Roslagstull Hospital, Box 5651, S-114 89 Stockholm, Sweden. The Journal of Infectious Diseases 1991;163:710-715 © 1991 by The University of Chicago. All rights reserved. 0022-1899/91/6304-0005$01.00

.had been treated with AZT for long periods. The antiviral effect was measured by determination of HIV p24 antigen levels and titers of cell-free infectious HIV in plasma.

Materials and Methods Patients. We studied 18 patients (cases 1-18) with symptomatic HIV-l infection. The inclusion criteria were HIV-l infection, ongoing treatment with AZT, and p24 antigenemia. Almost all patients who were asked agreed to participate. Seventeen were homosexual men and one was a heterosexually infected woman (mean age, 42 years; range, 32-54). Their initial CD4 cell counts are shown in table 1. Five patients (cases 1, 3, 12, 15, and 17) had AIDS defined by Pneumocystis carinii pneumonia, and one (case 18) had Kaposi's sarcoma. The other patients fulfilled the criteria for Centers for Disease Control group IVa, IVc2, or IVe. The patients had been treated with 400-1200 mg of AZT daily for 4-28 months (median, 17). This treatment wascontinued throughout the 24-week trial period. Interferon. The IFN-a preparation (Interferon Alfanative; BioNative, Urnea, Sweden) was purified from a crude extract prepared from Sendai virus-stimulated human leukocytes [12] by immunoaffinity and conventional chromatography to an IFN-a protein purity of >90 % according to the manufacturer. The preparation contained rv15 different IFN-a subtypes. The specific activity was 2 x 108 IU/mg protein as determined by standard biologic [13] and immunologic methods. The preparation was inactive in Limulus amoebocyte lysate test for endotoxin and contained trace amounts of interleukin-6 (~1.0 ng/ml) and tumor necrosis factor-a (61-95 pg/ml) but no demonstrable IFN-)', interleukin-l, or interleukin-4. The preparation procedure had been shown to bring about a complete (>I()6-fold) inactivation of several viruses, including HIV-l. Virusculture. HIV-l isolations from plasma were carried out as previouslydescribed [14].In addition,virus titrations were performed as follows: 10 ml of plasma was obtained from heparinized blood after centrifugation at 440 g. Tenfold dilutions were made in RPMI 1640 with 10% fetal calf serum. Each dilution was ultracentrifuged

Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on August 9, 2015

Patients with advanced human immunodeficiency virus type 1 (HIV-l) infection who had p24 antigen despite treatment with zidovudine (AZT) for 4-28 months received 3 x 1()6 IV of native interferon-a (IFN-a) daily for 3 months. Infectious HIV was detected in the plasma of all patients, in most cases at high titers, before lPN-a treatment. There was no correlation between HIV titers and p24 antigen levels. Antiviral activity, as measured by significantly decreased levels of infectious virus or p24 antigen, was observed in six of eight completely treated but in only one of nine incompletely treated patients. After termination of lPN-a treatment, there was a significant rise of p24 antigen levels. During IFN treatment, absolute CD4 cell counts showed a tendency toward an increased rate of decline. The side effects were unexpectedly severe. Despite its anti-HIV effect in vivo, lPN-a in the dosages used does not seem to be a viable additional treatment for severely immunodeficient patients in ongoing AZT therapy.

IFN-a and AZT in HIV-l Infection

JID 1991;163 (April)

Table 1. Characteristics of patients before treatment with interferon-a. Case no.

Initial CD4 count

(cells/mml)

8 12 14 16 21 26 21 17 17 22 11 18 4 10 5 17 15 28

80 160 50 190 130 130 120 180 190 320 170 >3 30 60 130 210 3 350

HIV-I plasma titer *

p24 antigen level (pg/ml), day ot

~100

1400 39 44 32 730 69 48 18 104 56 1I8 24 39 83 6 56 2800 ·12

~Ioo ~Ioo

100 I 10 10 100 10 100 ~loo

~100

10 100 10 10 100 100

* Titration of plasma human immunodeficiency virus type I (HIV-I) levels was done by virus isolation from l O-fold dilutions of plasma. t HIV-I antigen detection in sera was done by ELISA (Abbott).

in an angle rotor (R65) for 30 min at 45,000 rpm. The viruscontaining pellet was resuspended in RPMI 1640 and inoculated into a Falcon flask with 2 X 107 phytohemagglutin-stimulated peripheral blood mononuclear cells (PBMC) from HIV-seronegative blood donors. The cultures were maintained for 4 weeks. The culture supernatants were assayed for HIV antigen by a capture ELISA (Abbott Laboratories, North Chicago). The initial titrations were in some cases performed only to a 1:100 dilution. Retesting of thawed samples gave lower titers, indicating loss of viral activity during freeze-thawing. It is therefore possible that the initial titers were in fact >1:100. Therefore, the results given below are reciprocal titer of ;?;100 in these cases. Other methods. Total numbers of PBMC were determined by differential cell counting, and CD4 cells were enumerated by cytofluorography using OKT4 monoclonal antibodies (Ortho Diagnostics, Raritan, NJ). HIV-l antigen in serum was assayed by the antigen-capture ELISA (Abbott) that primarily detects the core protein p24 [15]. Determinations of p24 antigen levels were done in batch (i.e., all sera from each patient were tested simultaneously). Concentrations of IFN-a in serum were determined with a dissociation-enhanced lanthamide fluorescence immunoassay, and IFN -a antibodies were determined by an ELISA (both developed by BioNative). Study design. The 18 patients, while continuing AZT therapy, self-administered by subcutaneous injection 3 x 106 IU of IFN-a daily for 12 weeks. Pharmacokinetics were determined for all patients in connection with initiation of therapy by measuring serum IFN-a before and 4, 8, 12, and 24 h after the first injection. Blood samples for determination of p24 antigen and for standard hematologic and biochemical tests were collected 2-4 weeks before the first injection, at the first injection (week 0), weekly for 4 weeks, and

at weeks 6, 8, 10, 12, 16, 20, and 24. Serum samples were stored at -20°C until tested. Blood plasma samples for virus culture were collected at weeks 0, 6, 12, and, in some cases, 20. Statistics. The significance of the changes of p24 antigen levels in consecutive serum specimens and of changes of CD4 cell counts was determined by regression analysis and by Wilcoxon's test for pair differences.

Results

Withdrawals and dose reductions. In patients 2,3, 8, 12, and 15, IFN-a treatment was discontinued after 8, 8, 4, 6, and 8 weeks, respectively, due to side effects or progression to terminal illness. In another five patients (cases 5, 6, 10, 14, and 17), the IFN-a dose was reduced after 6-8 weeks from daily injections to three injections per week. Thus, only eight patients were treated completely, according to the protocol. All patients were evaluated for safety and for changes in p24 antigen levels and infectious HIV titers, except patient 8, who was not evaluated for possible titer changes. Virus culture. HIV was recovered from undiluted or 10fold dilutions of the plasma of all individuals before initiation ofIFN-a therapy (table 1, figure 1). The reciprocal titer was ~100 in 11 of 18 cases and in 10 of the 17 cases who-were later adequately followed-up, even though all subjects had ongoing treatment with AZT. Repeated virus cultures after 6 and 12 weeks showed a tendency toward declining titers. Thus, the median reciprocal titer of infectious HIV, which was 100 before treatment, was 10 after 6 and 30 after 12 weeks of treatment (figure 1). A significant decline in titer (~100-fold decrease) was seen in patients 1,4, and 18, that is, in three of the eight patients who received a full dose of IFN -a for 12 weeks. None of the nine incompletely treated patients (decreased dosage or premature withdrawal) displayed any significant decrease of HIV-l titer

~ ~ 100

>-

:r:

1 0

iii

CD®@0@ ®@@ @@@@@-@@@@ 00® @@@

0

~@ 13@

0ffiffi~- @~~-

0@ @

0

a.

@

0

0

(1)

0®@

CD

a:
.1) in either case. Still, there was a significant decline (P < .05) in the absolute CD4 values during the interval of . -21 to 0 weeks. However, the difference between the rates of decline before and after initiation of treatment was not statistically significant (P> .1), neither in the interval between -12 and 12 weeks nor between -21 and 24 weeks. The rate of decline of relative number of CD4 cells was similar before, during, and after IFN -a treatment. Relationships between plasma viremia, antigenemia, and CD4 cell count. There was no correlation between initial HIV plasma titers and p24 antigen serum levels (table 1). Fur-

Table 2. Influence of interferon-a (IFN-a) treatment on human immunodeficiency virus type 1 p24 antigen levels. Week of treatment Case no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Median value

-4 to -2

0

1000 39 48 30 430

1400 39 44 32 730 69 48 18 104 56 118 24 39 83 6 56 2800 12 52

72

59 16 102 53 118 26 33 nt 14 52 2400 nt 52.5

2 1000 35 55 30 470 77

21 18 108 41 123 nt 46 36 4 59 3300 10 46

800 30 38 30 600 76 nt 14 65 nt 113 60 39 36 nt 3000 39

3 800 29 39 27 630 64 25 18 44 nt 114 26 44 30 10 nt 2200 34.5

4 1100 26 45 27 520 68 22 15* 32 nt 115 13 57 nt 40

2800 36

6

8

10

12

16

20

24

900 29 49 29 600 53 16 18 30 54 110 13* 65 40 6 nt 2200

1200 30* nt* 37 610 35 13 nt 20 58 115 25 52 nt 6* 25 2800

1200 nt nt 32 880 31 16 nt 22 nt 108 14 nt nt nt 29 nt

40

35

30

1200* nt nt 40* 960* 40* 22* nt 22* nt* 131* nt 44* 48* 5 nt* 1000* 5* 42

3000 60 nt nt 1570 43 96 nt 117 61 141 nt 59 62 nt 45 1300 nt 79

2400 nt nt 42 1630 46 103 nt 193 60 195 nt 28 nt nt 116 2200 nt 116

1900 nt nt 54 nt 54 44 nt 246 82 nt nt 156 nt nt nt nt nt 87

NOTE. Antigen levels are in picograms per milliliter; nt, not tested; -, below detection limit (1'\.13 pg/ml). Decrease in antigen levels was statistically significant in the following patients: case 6 (P < .(01); case 7 (P = .03); case 9 (P = .003); case 16 (P < .01). * Time of discontinuation of IFN-a treatment.

Downloaded from http://jid.oxfordjournals.org/ at University of Iowa Libraries/Serials Acquisitions on August 9, 2015

during the study period. A rough estimate based on the geometric mean of the reciprocal titers of samples from the eight completely treated patients indicated that the IFN -a treatment caused rv70 % net decrease of the amount of virus isolated at both weeks 6 and 12 compared with the pretreatment values. However, it should be noted that these figures may underestimate the true decline in virus titers because of the failure to obtain exact end-point titers for some of the pretreatment samples. From the three patients (cases 1, 4, and 18) in whom IFN -a had a significant antiviral effect, follow-upplasma samples were obtained 2 or 3 months after cessation of treatment. These showed 10- to 100-fold increases in HIV-l titers (figure 1), compared with the titer obtained at week 12. HIV-l p24 antigen concentration. At enrollment, the serum p24 antigen concentrations of the 18 patients varied between 6 and 2800 pg/ml (table 2). In paired serum specimens from each of 16 patients obtained 2-4 weeks apart, the p24 antigen levels showed little intraindividual variation, and the median values of the two groups of sera were similar (52.5 and 52 pg/ml). During IFN-a treatment, the p24 antigen levels remained essentially unchanged in 12 of 18 patients (table 2), but in 6 patients (cases 6, 7, 9, 14, 16, and 18), a >50% decline was observed (table 2, figure 2). Three of the 8 completely treated patients (cases 7, 9, and 16) showed a statistically significant decline in p24 levels, and one other (case 18) showed a decrease from 12 pg/ml to undetectable levels. Only one of the incompletely treated patients (case 6) showed a statistically significant decrease of p24 levels. The overall decline in p24 levels from before treatment to

1ID 1991;163 (April)

lPN-a and AZT in HIV-I Infection

JID 1991;163 (April)

250

z:

§

/

Cl

/

.9:

/

200

s::

/6 # 9

0 /

~

/

C 150 (l)

/

u

/

c 0 o

100

s:: (l)

Cl

C a.

0 0

-4

4

8

12

16

20

24

Week

Figure 2. Human immunodeficiency virus p24 antigen levels in four patients (shown by case number) responding to interferon-a treatment.

1E "

1Il

180-

4i

1Il

~

QJ

1502

Ul

4i

+

u

Combined treatment of symptomatic human immunodeficiency virus type 1 infection with native interferon-alpha and zidovudine.

Patients with advanced human immunodeficiency virus type 1 (HIV-1) infection who had p24 antigen despite treatment with zidovudine (AZT) for 4-28 mont...
581KB Sizes 0 Downloads 0 Views