Current Therapeutic Options for Kaposi's Sarcoma

June 2, 1996

Treatment options for patients with acquired immune deficiency syndrome (AIDS)-related Kaposi's sarcoma (KS) have changed very little over the past decade. Research has been hampered by the absence of an accepted

ABSTRACT: Treatment options for patients with acquiredimmune deficiency syndrome (AIDS)-related Kaposi's sarcoma (KS)have changed very little over the past decade. Research has beenhampered by the absence of an accepted staging system for KS andthe lack of a uniform definition of response. New advances inunderstanding the pathogenesis of KS, particularly the role ofangiogenesis and growth factors, may help in the future developmentof additional therapies. At present, available options includeradiation therapy, chemotherapy, and biologic response modifiers.Improved delivery systems for existing chemotherapeutic agentsmay help to target more effective doses to KS lesions and to preventcertain dose-limiting side effects. Our goal should be to recommitourselves to finding ways to improve the survival of patientswith AIDS-related KS, not just to deliver cosmetic or palliativetreatments. [ONCOLOGY 10(Suppl):24-27, 1996]


In the early 1980s, KS was one of the major initial manifestationsof human immunodeficiency virus (HIV) disease in San Francisco.Patients presented with disfiguring lesions in sites other thanthe lower extremities, where the lesions of the classic variantof KS were most commonly seen. Extensive facial lesions clearlymarked patients with AIDS-related KS. Visceral disease was foundin the lymph nodes, gastrointestinal tract, and pulmonary parenchyma.It appeared that patients with widespread, rapidly progressive,or visceral KS lesions had a poorer prognosis than those withlesions confined to mucocutaneous or lymph node involvement.

The lack of a good staging system presented a problem. A recommendedstaging system, proposed by Krown et al [1], was never widelyadopted. It used the tumor, immune status, and systemic illness(TIS) criteria. Under this system, "good-risk" patientsincluded those with disease confined to the skin and/or lymphnodes or those with minimal oral disease. "Poor-risk"patients included those with more extensive oral disease, gastrointestinallesions, other visceral lesions, or tumor-associated edema orulceration.

Immune status remains the most important predictor of responseto therapy for KS: patients with more than 200 CD4 cells/mm³are likely to do better than those with more serious immune disruption.However, this may simply be an illustration of a well-known preceptin oncology: Patients with a good performance status tend to respondbetter to treatment than patients with a poor performance status.

In our earliest attempts to standardize treatment efforts whenfirst confronted with KS in 1981, a simple two-by-two grid wasdeveloped. The four variables were localized vs widespread diseaseand indolent vs aggressive disease. Indolent disease, regardlessof whether it was localized or widespread, was treated with radiotherapy.Aggressive, localized disease was also treated with radiotherapy.Aggressive, widespread disease was treated with chemotherapeuticregimens based on those used in the African or classic KS variants.The nonexperimental treatment regimens currently used (Table 1)have not changed significantly since then.

One important lesson we learned early in the epidemic, after somedisastrous results with early chemotherapy, was that observationalone could also be considered a good option in many patientswith minimal disease. We discovered that we did not need to attackeach individual lesion. We also learned that radiation therapywas useful in many cases.

Difficulty in Rating Response to Treatment

In the absence of an accepted staging system, it is difficultto compare responses to treatment. Certainly, if a patient's lesionsincrease in number and size after treatment, it is easy to classifythat as progressive disease. If lesions do not change much, thecategory of stable disease or minor response seems correct. Eradicationof all previously detectable disease, including lesions on theskin, in the mouth, or in the lymph nodes or visceral organs,indicates a complete response. However, KS has some unusual featuresthat complicate grading of response.

For example, Figure 1 shows a patient with a KS lesion on theeyelid before and after treatment with interferon-alfa. This isa fairly common phenomenon: The raised violaceous lesion flattensout and becomes hyperpigmented, a condition described as a hemosiderintattoo. Although a biopsy specimen does not show the histopathologicindications associated with active KS, the lesion is still visible.Should this be considered a partial response or a complete response?

Another such example is the situation seen in patients with KS-relatededema. This can include extensive periorbital involvement to thepoint where the patient is unable to open his or her eyes. A courseof radiation or chemotherapy may often dramatically improve theswelling, but the KS lesions are still present. The complete resolutionof the edema benefits the patient more than grading of partialresponse would suggest.

Current Treatment Approaches

For patients with one or a few isolated KS lesions, commonly usedlocal interventions include spot radiation or laser therapy, intralesionalchemotherapy, or cryotherapy. Cryotherapy (liquid nitrogen) isused by some clinicians for treating patients with localized KSlesions. However, patients may develop a depressed brown lesionfollowing cryotherapy, which many find less cosmetically acceptablethan the initial KS lesion.

KS-associated lymphedema of the extremities often responds toradiation treatment delivered in a single course or to systemicchemotherapy. Patients who want facial lesions eradicated forcosmetic purposes are also candidates for radiation therapy. Furthermore,radiation therapy has been of use in patients with pulmonary KSlesions.

Pulmonary KS is usually seen in association with extensive cutaneousdisease. In our early experience, two thirds of patients had aconcomitant opportunistic infection at the time of diagnosis ofpulmonary KS [2]. The life expectancy of an untreated patientdiagnosed with pulmonary KS was approximately 6 to 8 weeks, comparedwith an 18-month life expectancy for patients without pulmonaryKS. Pulmonary KS is one of the few pulmonary problems in HIV diseasethat leads to the development of pleural effusion, in additionto the common interstitial reticulonodular infiltrates. Patientswith pulmonary KS are also likely to suffer cough and hemoptysis.Bleeding lesions that can be localized by bronchoscopy may respondto radiation therapy.

Although radiation is localized therapy, KS associated with HIVdisease is a systemic disease, and chemotherapy is often moreappropriate. Pulmonary KS and rapidly progressive disease arebest managed with systemic chemotherapy. Chemotherapeutic agentsin wide use include the vinca alkaloids, bleomycin (Blenoxane),doxorubicin, and etoposide (VP-16 [VePesid]).

Some of the earliest work with chemotherapy for KS was performedby Laubenstein et al [3]. In their original schema, patients withmore indolent disease were treated with intravenous etoposide,whereas patients with more aggressive disease were treated withAdriamycin, bleomycin, and vinblastine (ABV), a combination thathad proved useful in African patients with KS. It was soon observed,however, that patients treated with the more aggressive combinationtherapy tended to develop more opportunistic infections than patientswho received the 3-day, single-agent etoposide infusion. Thisled us to suspect that perhaps we were contributing to the immunecompromise in some patients by using overly aggressive combinationchemotherapy. This is a problematic conclusion, however, becausethe patients who were chosen to receive the combination therapywere those with severe disease, who may have been predisposedto the rapid development of opportunistic infections and may havehad a poor prognosis regardless of the intervention.

A Better Combination

Ultimately, in San Francisco, we chose a regimen of vinca alkaloids,alternating vincristine and vinblastine weekly [4]. This regimenalternates the periods of myelotoxicity from vinblastine and neurotoxicityfrom vincristine, apparently delaying the appearance of theseadverse effects. With this approach, we have been able to achievea response rate of 33% (primarily partial responses) with minimaltoxicity.

Very early in the epidemic, before we knew that KS was relatedto HIV disease, we knew that we were dealing with an immune deficiencyand so began to test biologic response modifiers as a way of supportingimmune function. Interferon-alfa remains the sole biologic responsemodifier/antiviral agent that has demonstrated activity for AIDS-associatedKS. Questions about dosing and schedules persist, despite a decadeof widespread use. Recent studies suggest that there may be asynergistic anti-KS effect in combining interferon-a with zidovudine(Retrovir) [5]. Interferon has antiangiogenic properties as well,which may contribute to its anti-KS effect. The response ratesreported with interferon-alfa have ranged from 24% to 46%, similarto the rates observed in patients treated with alternating weeklydoses of vincristine and vinblastine. One problem, however, isthat the toxicities associated with interferon-alfa are greaterthan those seen with the alternating vinca chemotherapeutic regimen.Other biologic response modifiers have not been promising. Interferon-gammaproduced no response in our patients with KS. Isotretinoin plusinterferon was similarly ineffective [6].

Antiviral agents alone appear to have no impact on KS lesions.The first clinical trial performed by the AIDS Clinical TrialGroup (ACTG)--ACTG 001--was a comparison of zidovudine and placeboin patients with AIDS-related KS. No apparent impact on the KSlesions was seen in either patient group, although the resultsof this trial have never been published.

Zidovudine in combination with interferon-alfa appeared to bemore effective against KS lesions than either agent used alone.The effect was seen mainly in patients with better overall statusat the onset of therapy. However, the response rate for the combinationwas as high as 30% in patients with CD4 counts of less than 200cells/mm³, compared with a response rate of less than 10%for interferon-alfa used alone in this subset of patients. Addinginterferon-alfa to an antiretroviral agent may be helpful, butasymptomatic patients should be warned that the side effects ofinterferon, particularly the flu-like syndrome, may be daunting.

Chemotherapeutic agents should be used with caution in patientswho are also receiving antiretroviral therapy, because both regimensmay be myelosuppressive. In addition, vincristine should be usedwith caution in patients who are also receiving dideoxynucleosides,because both are associated with peripheral neuropathy. A relatedcaveat pertains to combining anti-KS regimens with treatment ofcytomegalovirus (CMV) disease. Ganciclovir (Cytovene) is the first-lineagent for treatment of CMV, and it is often associated with neutropenia.This causes problems if a doxorubicin-containing anti-KS regimenis then added. Some of these difficulties can be ameliorated occasionallyby using colony-stimulating factors, but the cost of this approachcan soon become prohibitive.

When to Treat

Because there is still no approach that clearly extends survivalfrom AIDS-related KS, the decision to initiate treatment mustbe made on other grounds. Asymptomatic patients with minimal KSand relatively high CD4 counts probably should be examined andhave their CD4 levels checked every 3 to 6 months.

The initiation of therapy for palliation and cosmesis is indicatedin a number of situations. Painful or bulky lesions, lesions causingrestricted mobility, and disfiguring lesions may be treated withlocal or systemic interventions. Generally, KS lesions are notpainful, but some patients do have painful lesions, especiallyon the soles of their feet or in areas that span joints. Lesionsthat ulcerate and tend to become infected may benefit from radiationtherapy. Extensive intraoral or pharyngeal lesions, particularlyin the posterior pharynx, can be obstructive and should also betreated with radiation or laser therapy. In these cases, lasertherapy may be preferred because it avoids the mucositis thatoften follows standard radiation therapy for intraoral lesions.The severity of radiation-induced mucositis often seems greaterthan expected from the same amount of radiation delivered to apatient who does not have HIV disease.

Initiation of treatment of extensive, rapidly progressive, cutaneousdisease requires thoughtful assessment; the risks and costs oftreatment should be weighed against the potential benefits. Asan alternative to embarking on a course of conventional chemotherapythat probably will be lifelong, patients may be offered the optionof participating in protocols evaluating new agents, if available.Although treatment of pulmonary KS may frequently require emergencyintervention in previously asymptomatic patients, the decisionof when to begin treatment again becomes more problematic.


Survival for patients with AIDS-related KS in San Francisco was21.5 months for those diagnosed from 1981 to 1983 and 16.4 monthsfor those diagnosed from 1984 through 1987. During this period,the survival of patients initially diagnosed with Pneumocystiscarinii pneumonia improved, whereas survival of those initiallydiagnosed with KS worsened. Aggressive therapeutic interventionshave not necessarily produced the desired outcome.

Prolongation of survival should be a goal of newer KS interventions.Ongoing work with improved delivery forms, inhibitors of viralproteins, and antiangiogenetic agents and similar new compoundsmust all be directed toward this end. In conventional oncologicapproaches, an experimental therapy that improves disease-freesurvival but not overall survival, compared with the current standardtherapy, is generally discarded. Quality of life is important,but, ultimately, we need to return our focus once again to survival.


1. Krown SE, Metroka C, Wernz JC, et al: Kaposi's sarcoma in theacquired immune deficiency syndrome: A proposal for a uniformevaluation, response, and staging criteria. J Clin Oncol 7:1201-1207,1989.

2. Kaplan LD, Abrams DI, Volberding PA: Treatment of Kaposi'ssarcoma in acquired immunodeficiency syndrome with an alternatingvincristine-vinblastine regimen. Cancer Treat Rep 70:1121-1122,1986.

3. 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.

4. Kaplan LD, Hopewell PC, Jaffe H, et al: Kaposi's sarcoma involvingthe lung in patients with the acquired immunodeficiency syndrome.J Acquir Immune Defic Syndr 1:23-30, 1988.

5. Krown SE, Gold JWM, Niedzwiecki D, et al: Interferon-a withzidovudine: Safety, tolerance, and clinical and virologic effectsin patients with Kaposi's sarcoma associated with the acquiredimmunodeficiency syndrome (AIDS). Ann Intern Med 112:812-821,1990.

6. Rosenthal E, Pesce A, Vinti H, et al: Isotretinoin plus interferonalpha-2a in AIDS-related Kaposi's sarcoma (abstract PO-B12-1607).Proceedings of the IX International Conference on AIDS. Berlin,Germany, June 6-11, 1993.