Pharmacology of Liposomal Daunorubicin and Its Use in Kaposi's Sarcoma

June 2, 1996

In the early 1980s, we first began to see cases of Kaposi's sarcoma (KS) in patients with the lymphadenopathy now known to be associated with infection by the human immunodeficiency virus (HIV). During that period, we

ABSTRACT: In the early 1980s, we first began to seecases of Kaposi's sarcoma (KS) in patients with the lymphadenopathynow known to be associated with infection by the human immunodeficiencyvirus (HIV). During that period, we were still looking at strongcombinations of chemotherapeutic drugs to combat this disease,based on experience in nonimmunosuppressed African patients. Itrapidly became apparent, however, that we needed both better regimensand more efficient ways of targeting antineoplastic agents totumors, because these older regimens were highly toxic. This isa common problem throughout oncology but is particularly troublesomein the treatment of patients with the acquired immunodeficiencysyndrome (AIDS). [ONCOLOGY 10(Suppl):28-33, 1966]

Introduction

The two key elements to improving the efficacy of an antineoplasticdrug--assuming that the drug is active against tumor cells--areto deliver the drug to the tumor and to prevent it from damagingother tissues on the way. Encapsulating cytotoxic agents, suchas daunorubicin (Cerubidine) in liposomes appears to address bothproblems and thus can markedly improve the therapeutic ratio.

Characteristics of Liposomal Daunorubicin

DaunoXome is a combination of the antineoplastic anthracyclineagent daunorubicin with a liposomal carrier system. This typeof phospholipid structure arranges itself into a bilayer thatbecomes spheric when placed in an aqueous solution. Drugs in anaqueous solution, such as daunorubicin, can be trapped in theaqueous center of such a structure. The DaunoXome system (Figure1) is a unilamellar liposome composed of two typical componentsof cell membranes--cholesterol and the phospholipid distearoylphosphatidylcholine--in approximately a 1:2 ratio. This particularlipid composition has been optimized to produce liposomes withremarkable physical stability and a particularly high affinityfor solid tumors in vivo [1]. To maximize tumor targeting further,the DaunoXome liposomes are filtered so the particles are withinthe 35-nm-to-65-nm range; tumor blood vessels have increased permeabilityto particles in this range. Another important benefit of thisformulation is that it can be used concurrently with the morecommonly prescribed antiretrovival agents without increased risk.

Pharmacokinetics of Free vs LiposomalDaunorubicin

A comparison of the plasma levels of daunorubicin in patientsadministered similar doses of daunorubicin in either conventionalor liposomal formulations demonstrated that peak plasma levelsof DaunoXome are 100-fold higher than those for free daunorubicin(Figure 2) [2]. The mean area under the curve (AUC) for plasmaconcentrations of daunorubicin is 36-fold higher with DaunoXomethan with free daunorubicin. This appears to be due to the factthat free daunorubicin is rapidly eliminated, whereas plasma levelsof daunorubicin in patients administered DaunoXome remain highfor many hours. The retention of daunorubicin within circulatingliposomes has been confirmed by studies in which the lipid andaqueous components of the liposomes were radiolabeled separately.These studies revealed that these structures are stable to leakagefor at least 30 hours in the circulation. Because free daunorubicinfluoresces, quenching this fluorescence can be used to monitorthe retention of daunorubicin by liposomes. Fluorescence-quenchingexperiments demonstrated that more than 95% of daunorubicin remainsassociated with liposome for at least 12 hours in the circulation.

The pharmacokinetics of free daunorubicin vs liposome-encapsulateddaunorubicin have also been studied in CD2F1 mice bearing P-1798lymphosarcomas [3]. Both drugs were given intravenously at 20mg/kg. At 1 hour following administration, the plasma level offree daunorubicin was 1.4 mg/mL, whereas the plasma level of daunorubicinwas 268 mg/mL after administration of DaunoXome. The ratios ofthe AUC of free drug relative to DaunoXome were most significantlyincreased (10-fold) in tumor tissue. Similarly, in DaunoXome-treatedmice, the peak tumor level of daunorubicin was 100 mg/g, whichoccurred approximately 8 hours following administration; in freedaunorubicin-treated mice, the peak tumor level of daunorubicinwas 9.6 mg/g and occurred 1 hour after administration.

Tumor Targeting of Liposomal Drugs

Although the specific mechanism by which DaunoXome is preferentiallydelivered to solid tumors in situ has not been established, adramatically higher level of daunorubicin is achieved in tumorswith DaunoXome than after treatment with free daunorubicin [4].Laser-inducedfluorescent video imaging has shown that as liposomes break down,daunorubicin is released only within the tumor mass, not in thesurrounding normal tissue [5]. Autoradiography has shown a specificand dramatic targeting of KS lesions with indium-111-labeled liposomes[6]. Experimental studies of mice bearing J6456 lymphoma celltumors revealed that liposomal formulations of gallium-67-labeleddeferoxamine, containing monosialoganglioside, cholesterol, anddistearoyl phosphatidylcholine, were retained in the blood andaccumulated in tumors at markedly higher levels than were liposomescomposed of other constituents [7].

Therapeutic Efficacy of Free vs LiposomalDaunorubicin

Given that this type of carrier brings the anticancer drug tothe tumor, what evidence is there that this has any therapeuticimportance? Mouse MA16C mammary adenocarcinoma cells were implantedsubcutaneously into mice and were used as a model system for testingthe antitumor efficacy of liposomal daunorubicin [3]. Four daysafter implantation, single intravenous doses of free daunorubicin(2 to 40 mg/kg) or DaunoXome (2 to 50 mg/kg) were given. (Thehigher DaunoXome maximum dose was possible because of reducedtoxicity; see below.)

The median time of survival was 36 days for untreated controlmice bearing MA16C tumors, 70 days for mice treated with 20 to25 mg/kg doses of free daunorubicin, and 114 days for mice treatedwith 15 to 35 mg/kg of DaunoXome. In the free daunorubicin groups,there were 4 long-term survivors of a total of 90, as opposedto 40 survivors of 80 in the DaunoXome groups. Furthermore, treatmentwith the most effective dose of DaunoXome (25 mg/kg) cured all10 mice treated with this concentration of drug. All were long-termsurvivors, with no gross evidence of tumor. In contrast, the mosteffective dose of free daunorubicin (20 mg/kg) cured only 4 of10 treated mice. Of the 80 mice treated with DaunoXome, 15 hadcomplete regression of subcutaneous tumors, which was later confirmedat necropsy. The increased survival observed with DaunoXome treatmentin murine solid tumor models indicates that the strategy of enclosingdaunorubicin in this type of liposome effectively increases thecytotoxicity for the malignant cells.

Toxicity Profiles

Phase I and II clinical trial experiences with patients with KShave demonstrated that, in general, DaunoXome is associated withfewer side effects than free daunorubicin [8,9]. The primary toxicityis myelosuppression, which is much milder than that observed withconventional anthracyclines; this is perhaps due to the decreasein free drug in the circulation. Cardiotoxicity has not been observed,despite cumulative doses of more than 1,000 mg in several patients.The latter finding is supported by data from mouse toxicologystudies that showed little accumulation of liposome-encapsulateddaunorubicin in cardiac tissues [10]. Alopecia and tissue necrosisassociated with drug extravasation at the injection site havenot occurred. Other complications, including nausea, vomiting,and mucositis, occurred infrequently. These studies demonstrateda significant improvement in the therapeutic ratio for the liposomalformulation DaunoXome, and clinical experience has confirmed thatthis is an effective drug for use in the treatment of KS in patientswith AIDS [8,9].

Current Studies of the Use of LiposomalDaunorubicin for Kaposi's Sarcoma

AIDS-related KS often presents as cutaneous disease with nodularlesions that may rapidly progress. Up to 50% of patients haveoral cavity or gastrointestinal tract involvement, which may leadto abdominal pain, early satiety, or gastrointestinal hemorrhage.Pulmonary involvement is a common, serious problem. Even in casesin which KS is not life-threatening, extensive cutaneous diseasecan cause significant morbidity. Patients may become disfiguredand are subject to skin ulceration, bleeding, and edema at a timewhen they are already dealing with the psychological implicationsof AIDS [11,12].

Therapeutic options for KS in these patients are limited. Interferon-ais the only agent approved specifically for this indication; thisagent only benefits the minority of patients with CD4 cell countshigher than 400/mm³. Furthermore, even in cases where interferon-ais efficacious, there are frequently problems with drug toxicity;patients ultimately develop progressive disease despite continuedtreatment [13]. For these patients, as well as for those deemedunlikely to respond to interferon-a, chemotherapy is generallyused. Chemotherapy often makes use of either single agents (suchas bleomycin [Blenoxane], vincristine [Oncovin], vinblastine [Velban],doxorubicin [Adriamycin], or etoposide [VePesid] or combinations(such as Adriamycin, bleomycin, and vincristine [ABV]). Althoughsignificant response rates have been reported with the ABV combination,toxicities often limit the duration of treatment [14].

In this setting, it was considered worthwhile to evaluate DaunoXomefor the management of AIDS-related KS. This approach was chosenfor several reasons: the proven antitumor efficacy of conventionaldaunorubicin, the ability of the liposomal daunorubicin formulationto be targeted to KS lesions [6], and preclinical studies suggestingthat the liposomal form was likely to cause fewer side effectsthan conventional daunorubicin or ABV. Toxic effects with ABVinclude cardiotoxicity from doxorubicin, pulmonary toxicity frombleomycin, and neurotoxicity from the vinca alkaloids. All thesedrugs may increase the risk of infectious complications and gastrointestinaltoxicity seen with each treatment cycle; this can be particularlyhazardous for an immunocompromised population of patients [15].

Phase I Studies

Phase I studies of liposomal daunorubicin were conducted by Gilland colleagues at the University of Southern California, Los Angeles[2]. The studies included 23 patients with extensive disease (25or more mucocutaneous lesions, the development of 10 or more lesionsin a 1-month period, or symptomatic visceral involvement). Multiplepatients with KS were studied at each of 5 dose levels: 10 mg/m²,20 mg/m², 30 mg/m², 40 mg/m², and 60 mg/m².

Documented toxic effects from free daunorubicin include myelosuppression,nausea and vomiting, mucositis, alopecia, cardiotoxicity, andtissue necrosis associated with extravasation from administrationsites. By comparison, the adverse effects with liposomal daunorubicinwere relatively mild and included mild or nonexistent nausea orvomiting; absent or much less frequent occurrence of alopecia;neutropenia or thrombocytopenia only at doses higher than 80 mg/m²;and no morbidity related to extravasation. The phase I data supportedthe use of liposomal daunorubicin at a dose of 40 mg/m² every2 weeks for these patients.

Pharmacokinetic studies performed at the same time again demonstratedthat the pharmacokinetics of the liposomal formulation are quitedifferent from those of free daunorubicin; peak plasma levelsfor equivalent doses are nearly 100-fold higher due to a muchsmaller volume of distribution. Conventional daunorubicin is clearedfrom the circulation in a biphasic fashion, with a rapid alphahalf-life of less than 1 hour followed by a slow terminal half-life.Liposomal daunorubicin does not undergo this early, rapid clearing,and its pharmacokinetics follow a monoexponential decline in theplasma level. The net effect is a 35-fold increase in the plasmaAUC for the liposomal formulation.

Not designed to test for efficacy, the phase I trials neverthelessprovided important data. Although there were no complete remissionsin this group, the majority of patients experienced either partialremissions or stable disease. Patients who withdrew from the studydid so due to either progression of HIV disease or progressiveKS, rather than treatment-related side effects. The phase I trialdata demonstrated that it was possible to administer safely dosesof liposomal daunorubicin that were much higher than the conventionaldoses of free daunorubicin, without causing serious side effects.

Several patients experienced a syndrome of back pain, flushing,and chest tightness following liposomal daunorubicin administration.In general, the back pain commenced about 5 minutes after infusionbegan and was sometimes associated with a feeling of chest tightnessand/or flushing. The symptoms resolved within 5 to 10 minutesof termination or interruption of the infusion. This syndromesometimes involved patients who had previously been treated withliposomal daunorubicin without incident. In other cases, patientswere subsequently retreated without recurrence of these symptoms.The etiology of this syndrome is not clear, but the temporal sequenceand nature of the symptoms suggest it is due to the release ofendogenous vasoactive compounds triggered by interactions betweencell-surface receptors and liposomal components, rather than toan effect of the encapsulated daunorubicin.

Phase II Studies

Phase II studies have evaluated liposomal daunorubicin for responserate, duration of therapy, duration of response, and effect onquality of life. Judging complete and partial responses in patientswith KS, however, can be problematic. A decrease in the numberand/or size of lesions would appear to be the most straightforwardcriterion for judging response; however, this is misleading inpatients with KS. Because a pigmented area is often left behind,these lesions do not always completely "disappear."Therefore, these residual pigmented areas can interfere with resultswhen response is judged by determining the decrease in the sumof the products of the largest perpendicular diameters of thetarget lesions. When such a cutaneous lesion persisted but wasshown on a biopsy specimen to have no malignant cells, it wasalso considered a complete response.

Criteria for judging partial responses (Table 1) were also carefullydesigned. The most important partial responses were flatteningof at least 50% of nodular or plaque-like lesions or flatteningof at least 75% of nodular lesions to indurated plaques [16,17].

The phase II trials of DaunoXome were conducted at six differentresearch centers. These trials were open-labeled studies, withall patients receiving liposomal daunorubicin. Three of thesestudies will be discussed separately and then the conclusionsdrawn from analysis of the pooled results will be presented.

We evaluated results from 22 male patients with AIDS-related KSwho were treated with 40 mg/m² of intravenous liposomal daunorubicinevery 2 weeks and given as an infusion without routine premedication.Eligible patients were HIV-positive and with histologically provenKS. Lesions were evaluated at entry and after two treatment cycles.Evaluation was performed after the second treatment. If patientsdid not show at least a partial response, they were removed fromthe study. Patients with partial responses continued to receiveliposomal daunorubicin until disease progression or significanttoxicity.

The vast majority of the patients in this study (20 of 22) hadextensive disease, with at least 25 measurable cutaneous lesions,adenopathy, lymphoma "B" symptoms (fever, night sweats,weight loss), or visceral involvement. The median CD4 cell countwas 95 cells/mm³. Two patients had minimal disease, one withonly three lesions and a CD4 cell count of 730 cells/mm³.All patients had a Karnofsky performance status higher than 50and a life expectancy of at least 8 weeks. Cardiac ejection fractionshad to be higher than 45%, and all patients received prophylaxisfor Pneumocystis carinii pneumonia (PCP); most patients receiveda nucleoside analogue for antiretroviral therapy, most commonlyzidovudine (Retrovir) or didanosine (Videx). Treatment of opportunisticinfections was allowed as clinically indicated.

Twenty patients had received prior chemotherapy. Prior therapywith anthracycline was allowed, but not within 3 weeks prior tostudy entry and only with a total dose of up to 200 mg/m².Prior therapy with mitomycin C (Mutamycin) was allowed up to 8weeks before study entry.

Among our 22 patients, there was a single complete response ina patient with a CD4 cell count of 730 cells/mm³ who waspreviously identified with minimal disease. Although this patient'sKS lesions resolved, he opted to withdraw from the study afterseveral cycles because he was concerned about a decrease in hiswhite blood cell count. One patient had stable disease after thesecond cycle, and 20 patients had partial responses.

The principal toxicity of concern was neutropenia, in 18% (4 of22) of patients who had an absolute granulocyte count less than1,000 cells/mm³ in 1.6% (6 of 371) of total cycles. Anemia(hemoglobin less than 8 mg/dL occurred in 1.5% of cycles and thrombocytopenia(platelet count less than 50,000) in 2% of cycles. Vomiting anddiarrhea accompanied less than 1% of cycles, with mild nauseain 22% of cycles. Two percent of cycles were accompanied by headacheand 1% by fatigue. No changes in cardiac ejection fraction wereobserved. In all, this profile of side effects is much improvedover that expected with free daunorubicin administration, basedon historical controls.

The effect of liposomal daunorubicin on life expectancy cannotbe determined at this point. We do know, however, that patientswith KS who are treated with this drug experienced much fewerside effects than those who were treated with conventional anthracyclinetherapy.

Another Study

Similar results were reported by Presant and colleagues at theLos Angeles Oncologic Institute [8]. This group treated 25 patientswith HIV-associated KS of poor prognosis, 24 of whom were evaluable.The liposomal daunorubicin dosage was 40 mg/m² every 2 weeks.There were 2 complete remissions (8.3%) and 13 partial remissions(54.2%). The median duration of response to treatment was 12 weeks.

Patient-measured quality of life assessments in terms of physicalperformance and emotional well-being were performed using specifiedcriteria for serial evaluation of quality of life in patientswith cancer [18]. There was considerable improvement in the measuredquality of life parameters, with a response rate of 71% of patientsfor physical performance and 74% of patients for emotional well-being.

Higher doses of liposomal daunorubicin (50 mg/m² or 60 mg/m²)have also been studied [2]. Among 9 patients treated with 50 mg/m²,there were 5 partial responses, and the median number of cyclesto response was 3. Among 13 patients treated with 60 mg/m²,there was 1 complete response and 6 partial responses, and themedian number of cycles to response was again 3. Overall, 55%(12 of 22) of patients treated with 50 or 60 mg/m² had clinicalresponses at these slightly higher doses of liposomal daunorubicin.

Pooled Data From All Phase II Trials

The pooled results of data from 95 patients in the phase II trialsshow that, overall, 97% received important clinical benefits:complete response, partial response, or stable disease lastingfor more than 2 cycles (4 weeks) of therapy (Table 2). The overallresponse rate of 67% compares quite favorably with the single-agentresponse rate for other agents and is analogous to the responserates seen with various combination chemotherapeutic regimens.

The ability to continue therapy over long periods is a major concernin patients with KS. Severely immunocompromised patients withadvanced KS often manifest progressive disease within a few weeksin the absence of effective therapy. Overall, the mean durationof treatment in the phase II trials was 20 weeks (9.6 cycles),with a mean cumulative dose of 378.6 mg/m². This extendedtreatment was not accompanied by declines in quality of life,as indicated by scores on the Karnofsky Performance Status anda quality of life questionnaire. Significantly, there was no changein the Karnofsky Performance Status score at cumulative dosesof up to 1,000 mg/m².

As expected, the primary toxicity was hematologic. At least twothirds of patients were anemic at initiation of liposomal daunorubicintherapy, and more than 50% had leukopenia. The severe antecedentbone marrow dysfunction in this population is attributable toHIV infection, antiretroviral therapy, or other medications received.There was a decline in hemoglobin levels to grade 3 or grade 4(less than8 g/dL after initiation of liposomal daunorubicin therapybut no evidence of cumulative myelosuppression with additionalcycles of treatment.

The problems of cardiac toxicity that often occur when the cumulativedose of free daunorubicin reaches 500 mg/m² were notablyabsent with the liposomal formulation. No impairment of cardiaccontractility or decline in left ventricular ejection fractionwas observed. Seven patients received cumulative doses exceeding1,000 mg/m², an amount that corresponds to approximately1 year of therapy. Of these 7 patients, 5 remained on therapybeyond 32 cycles, and 1 received a cumulative dose exceeding 2,000mg/m². No patient had to discontinue liposomal daunorubicinbecause of cardiac toxicity.

Results in Previously Treated Patients

Separate analysis was performed for the subset of patients fromthe phase I and phase II trials who received prior systemic cytotoxicchemotherapy for AIDS-related KS. In general, these patients wereheavily pretreated, had advanced HIV infection, and had a historyof opportunistic infection (a negative prognostic factor). TheirKS tumor burden was great, with cutaneous lesions too numerousto count, lymphedema, and visceral involvement. Patients whoseprior systemic therapy was limited to interferon-a were excludedfrom this analysis, as were those who received only intralesionalchemotherapy. This left a group of 60 patients: 16 from phaseI studies and 44 from phase II studies.

In the phase I patients, who were the most heavily pretreated,50% (8 of 16) showed evidence of antitumor activity from liposomaldaunorubicin. This included three patients with partial responsesand five patients with clear antitumor activity, as shown by flatteningof the nodular lesions or shrinking of the cutaneous disease.For two additional patients, disease stabilized for longer than4 months.

Of the 44 patients in the phase II trials, 52% (23 of 44) demonstrateda partial response. Disease stabilized in 17 additional patients[38% (17 of 44)], including 7 patients [15% (7 of 44)], with someregression of cutaneous lesions but not enough to qualify as partialremission. Taking into account patients with partial responses,ie, those with some tumor regression or prolonged stable disease,75% of previously treated phase II patients derived benefit fromliposomal daunorubicin therapy. An important point is that eventhe 25% of patients who had become refractory to treatment withdoxorubicin had partial responses to liposomal daunorubicin.

Conclusions

Overall, the clinical studies conducted with this formulationof liposomal daunorubicin indicate that it is an active drug,even in patients whose KS is refractory to doxorubicin, and thatthis efficacy is accompanied by relatively mild toxicity. Encapsulationof daunorubicin in liposomes preferentially targets delivery ofthe drug to tumor cells, spares other tissues, and increases tumor-cellkilling. Like other treatment options, liposomal daunorubicinis not curative for KS, but it does appear to be an effectivepalliative agent. It has arrested progression of advanced KS inmore than 95% of patients and has produced actual remissions in67% of patients.

References:

1. Proffitt RT, Williams LE, Presant CA, et al: Tumor-imagingpotential of liposomes loaded with In-111-NTA. J Nucl Med 24:45-51,1983.

2. Gill PS, Espina BM, Muggia F, et al: Phase I/II clinical andpharmacokinetic evaluation of liposomal daunorubicin. J Clin Oncol13:996-1003, 1995.

3. Forssen EA, Coulter DM, Proffitt RT: Selective in vivo localizationof daunorubicin in small unilamellar vesicles in solid tumors.Cancer Res 52:3255-3261, 1992.

4. Gabizon AA: Selective tumor localization and improved therapeuticindex of anthracyclines encapsulated in long-circulating liposomes.Cancer Res 52:891-896, 1992.

5. Tromberg BJ, Krasieva T, Shimizu S, et al: Non-invasive, laser-indicatedfluorescence imaging of daunorubicin in solid tumors in situ:Comparison of conventional drug with tumor-targeting liposomes(DaunoXome) [Abstract] Proc Am Assoc Cancer Res 33:446, 1992.

6. Presant CA, Blayney D, Proffitt RT, et al: Preliminary report:Imaging of Kaposi sarcoma and lymphoma in AIDS with indium-111-labelledliposomes. Lancet 335:1307-1309, 1990.

7. Gabizon AA, Papahadjopoulos D: Liposome formulations with prolongedcirculation time in blood and enhanced uptake by tumors. ProcNatl Acad Sci USA 85:6949-6951, 1988.

8. Presant CA, Scolaro M, Kennedy P, et al: Liposomal daunorubicintreatment of HIV-associated Kaposi's sarcoma. Lancet 341:1242-1243,1993.

9. Money-Kyrle JF, Bates F, Ready J, et al: Liposomal daunorubicinin advanced Kaposi's sarcoma: A phase II study. Clin Oncol 5:367-371,1993.

10. Fichtner I, Arndt D, Elbe B, et al: Cardiotoxicity of freeand liposomally encapsulated rubomycin (daunorubicin) in mice.Oncology [Switzerland] 41:363-369, 1984.

11. Gill PS, Rarick MU, Espina BM, et al: Advanced acquired immunedeficiency syndrome-related Kaposi's sarcoma. Cancer 65:1074-1078,1990.

12. Laubenstein LJ, Krigel RL, Odajnyk CM, et al: Treatment ofepidemic Kaposi's sarcoma with etoposide or a combination of doxorubicin,bleomycin, and vinblastine. J Clin Oncol 2:1115-1120, 1984.

13. Krown SE: Interferon and other biologic agents for the treatmentof Kaposi's sarcoma. Hematol Oncol Clin North Am 5:311-322, 1991.

14. Gill PS, Rarick MU, Bernstein-Singer M, et al: Treatment ofadvanced Kaposi's sarcoma using a combination of bleomycin andvincristine. Am J Clin Oncol 13:315-319, 1990.

15. de Wit R, Schattenkerk JK, Boucher CA, et al: Clinical andvirological effects of high-dose recombinant interferon-alphain desseminated AIDS-related Kaposi's sarcoma. Lancet 2:1214-1217,1988.

16. Krown SE, Metroka C, Wernz JC: Kaposi's sarcoma in the acquiredimmune deficiency syndrome: A proposal for uniform evaluation,response, and staging criteria. J Clin Oncol 7:1201-1207, 1989.

17. Krown SE: AIDS-associated Kaposi's sarcoma: Pathogenesis,clinical course, and treatment. AIDS 2:71-80, 1988.

18. Presant CA, Wiseman C, Blayney D, et al: Proposed criteriafor serial evaluation of quality of life of cancer patients. JNatl Cancer Inst 82:322-323, 1990.