Kaposi sarcoma

Appraisal of therapeutic agents


  • Liron Pantanowitz MD

    Corresponding author
    1. Department of Pathology, Baystate Medical Center, Tufts University School of Medicine, Springfield, Massachusetts
    • Department of Pathology, Baystate Medical Center, 759 Chestnut Street, Springfield, MA 01109
    Search for more papers by this author
    • Fax: (413) 794-3195

  • See referenced original articles on pages 1083–8 and 1147–52, this issue.


Despite major advancements, the optimal treatment for Kaposi sarcoma has yet to be defined. Comparable standards employing existing therapies to satisfactorily manage all epidemiologic forms of Kaposi sarcoma, and the utilization of mechanism-based therapeutic strategies with innovative surrogate endpoints to appraise their efficacy, are required.

In this issue of Cancer, there are 2 articles pertaining to the treatment of Kaposi sarcoma (KS): a multicenter phase 1/2 trial investigating the effect of the novel matrix metalloproteinase (MMP) inhibitor BMS-275291 in the treatment of 15 patients with KS who were infected with the human immunodeficiency virus (HIV),1 and a retrospective study examining the role of pegylated liposomal doxorubicin (PLD) as second-line therapy in 20 pretreated HIV-negative men with aggressive, cutaneous (classic) KS.2 These articles bring to light the enormous advancements that have been made to date regarding the treatment of KS. However, they also accentuate the imminent need for us to formulate an effective and tolerable treatment strategy for this intriguing vascular tumor.

Much has been written about KS since Moriz Kaposi first described this entity back in 1872, particularly after the recognition of acquired immunodeficiency syndrome (AIDS)-associated KS in 1981.3 It is not surprising to find that a PubMed search (accessed online October 10, 2007) of MEDLINE (the National Library of Medicine's premier bibliographic database) for KS yielded greater than 10,000 articles related to this vascular tumor. KS is still reported to be the most common cancer among HIV-infected persons, and certainly one of the most widespread tumors in sub-Saharan Africa. Moreover, KS provides investigators with an excellent model with which to study the interplay of the immune system with viruses and cancer. Many recent publications regarding KS elegantly bridge basic and clinical research.

KS is an angioformative lesion characterized histologically by neoangiogenesis and proliferating spindle-shaped cells, admixed with a variable chronic inflammatory infiltrate. The majority of spindled tumor cells express endothelial markers. Both groups of researchers in the current issue investigated different epidemiologic forms of KS.1, 2 Despite their environmental and immunologic variation, the development of all epidemiologic forms of KS (classic/ sporadic, African/endemic, AIDS-associated/epidemic, and immunosuppression/iatrogenic-associated) depends largely on infection with KS herpesvirus/human herpesvirus-8 (KSHV/HHV-8). These different KS subtypes also exhibit a similar histologic appearance. KSHV encodes many gene products that induce cellular proliferation, transformation, cell signaling, cytokine production, immune evasion, antiapoptosis, and angiogenesis.4 Although KSHV is definitely required, the KS tumor microenvironment also plays a critical role in KS tumorigenesis.4, 5 Current data indicate that KS spindled tumor cells do not display a highly transformed phenotype in vitro.4 In addition, investigators have raised questions concerning the clonal nature of in vivo KS. The constant presence of an admixed inflammatory infiltrate has led some authors to believe that KS may result fromreactive hyperproliferation induced by chronic inflammation, and that it is not a true neoplasm.4 It is interesting to note that although some KS lesions are monoclonal expansions of HHV-8-infected spindle cells, more advanced lesions appear to represent distinct oligoclonal proliferations.6

It is well known that KS is usually more frequent, widely disseminated, and presents more aggressively in AIDS than in other immunodeficiency states. Additional factors in this setting, such as the HIV-1 Tat protein, certainly stoke the fire. Fortunately, in areas in which highly active antiretroviral therapy (HAART) is available, there has been a marked decline in the incidence of KS. Until recently, AIDS-related KS was staged according to the AIDS Clinical Trials Group (ACTG) classification system, which originally (prior to HAART) characterized patients as “good” or “poor” risk based on their tumor burden (T) and immune function (I) as measured by their CD4 T-lymphocyte count, and the presence of systemic illness (S).7 In the HAART era, investigators found CD4 cell counts to no longer provide valuable prognostic information.8 However, the decision to initiate systemic chemotherapy is not based solely on the extent of KS disease, but also on other parameters such as patient performance status, end organ function, and concomitant medications. The patient cohort in the AIDS Malignancy Cohort study were not staged, but did require a Karnofsky performance status ≥60 and adequate hepatic, renal, and hematopoietic function for enrollment.1

As pointed out by Di Lorenzo et al. in their article, because of the highly variable clinical course of classic KS, clinicians find it difficult to decide whether and when to treat these elderly patients.2 These authors adopted a “Mediterranean KS” staging system, previously proposed for classic KS to facilitate such therapeutic decision making.9 This system comprises 4 stages: a maculonodular stage with lesions localized to the lower extremity (stage I); an infiltrative stage in which KS involves wide areas on the lower limbs (stage II); a florid stage for exuberant, often ulcerated, lesions involving ≥1 limbs (stage III); and a disseminated stage in which KS extends to cutaneous sites other than the limbs (stage IV). In addition, disease progression in this system is scored as “A” for slow or “B” for rapidly (defined as an increase in the number or total area of lesions in 3 months) evolving KS and complications (eg, ulceration, lymphedema, or pain). However, this system does not incorporate visceral KS. In fact, patients with evidence of visceral involvement were excluded from the PLD study published by Di Lorenzo et al.2

The treatment goals for KS include 1) symptom palliation; 2) shrinkage of tumor to alleviate edema, organ compromise, and psychologic stress; 3) the prevention of disease progression; and 4) perhaps cure.10, 11 The improvement of cytotoxic chemotherapy, enhancement in local approaches, the advent of HAART specifically for AIDS-related KS, and the introduction of molecularly targeted therapies have transformed KS therapy.10 Typical indications for systemic therapy include widespread skin involvement, extensive KS of the oral cavity, symptomatic pedal or scrotal edema, symptomatic visceral involvement, and immune reconstitution inflammatory syndrome-induced KS flare.10, 11 Although several chemotherapeutic agents (eg, bleomycin, vinblastine, vincristine, doxorubicin, and etoposide) were noted to be efficacious against KS in the past, current systemic cytotoxic therapy comprises liposomal anthracyclines (PLD and liposomal daunorubicin) and taxanes (paclitaxel).11 Liposomal anthracyclines exhibit efficacies comparable to those of conventional anthracyclines, but with better safety profiles and less cardiotoxicity. These agents offer a therapeutic advantage over the free (unencapsulated or nonliposomal) drug due to their prolonged circulation time and decreased drug-induced toxicity.12 Currently, this group of liposomal drugs are considered to be the best first-line chemotherapeutic option for the majority of patients with widely disseminated, symptomatic KS.13 Undeniably, most studies evaluating liposomal anthracyclines involved patients with AIDS-associated KS.10 Only limited data exist regarding the use of these agents in patients with classic KS. Therefore, the retrospective study published by Di Lorenzo et al. provides a valuable contribution to the literature, demonstrating that PLD is also associated with clinical improvement in elderly, HIV-negative men with pretreated classic KS, and that such systemic therapy is well tolerated in this group of aged individuals as a second-line treatment. However, further studies are still required to firmly establish first-line and second-line treatments for classic KS, as has been published regarding AIDS-related KS.11

The various steps involved in the pathogenesis of KS offer several targets for novel anticancer drug development.14 For example, by destroying the underlying extracellular matrix, MMPs facilitate angiogenesis and thereby play an important role in KS tumorigenesis.15 These zinc-dependent endopeptidases therefore represent potential targets for future MMP inhibitors. Indeed, this was the rationale for phase 1 and 2 trials in KS patients using the MMP inhibitor known as COL-3 (6-demethyl-6-deoxy-4-dedimethylaminotetracycline), a chemically modified tetracycline.16, 17 In these trials, COL-3 was shown to be active and well tolerated in the treatment of AIDS-related KS. This may be partly related to the finding that COL-3 affects MMPs by multiple mechanisms, including inhibition of MMP activation and expression. BMS-275291, a nonpeptidic MMP inhibitor, was specifically designed to inhibit certain MMPs.1, 18 In their trial, Brinker et al. evaluated the safety and efficacy of this novel MMP inhibiter in patients with AIDS-related KS.1 Unfortunately, BMS-275291 induced unacceptable toxicity (grade 3 fatigue, allergic reactions, and arthralgias) and demonstrated inadequate efficacy. Nevertheless, this trial bears testimony to the promising contemporary, pathogenesis-related, molecularly targeted approach for the treatment of KS.1

Utilizing appropriate and reliable surrogate endpoints for evaluating response is necessary to permit the rapid and accurate clinical evaluation of biologic agents. The development of new anti-KS agents, with mechanisms of action that differ from conventional chemotherapy, necessitates a new look at clinical trial design. The evaluation of endpoints, other than tumor response, in this regard is important. In general, the critical steps that govern successful phase 2 cancer chemoprevention trials include well-characterized agents, suitable cohorts, and reliable intermediate biomarkers for measuring efficacy.19 Surrogate biomarkers (eg, specific MMP levels) must be closely linked to the pathogenesis of the neoplasm being treated, undergo some degree of modulation by the therapeutic agent (eg, selective MMP inhibition), be amenable to easy measurement (eg, quantitative assay), and ideally correlate with decreased cancer incidence.19 Biomarker panels that evaluate an array of pathogenic pathways provide excellent surrogate endpoints. A predetermined biomarker (eg, KSHV) needs to be expressed differently in tumor compared with adjacent normal tissue. This is particularly important when tissue biopsies of KS are procured for analysis. In addition, markers of normal cellular processes that may be increased or expressed during tumorigenesis, but are nonspecific, are best avoided.

To date, investigators conducting clinical trials involving KS have relied on several endpoints for evaluating treatment response. The primary endpoint used by Di Lorenzo et al. was the objective tumor response rate, and their secondary clinical endpoints included a determination of pain intensity, time to disease progression, and overall survival. Brinker et al. based their decision to escalate the dose of BMS-275291 on dose-limiting toxicity, tolerability, objective tumor responses, and the number of patients whose tumor biopsies demonstrated a biologic response, as measured by the percentage change in apoptotic cells by use of the terminal dUTP nick end labeling (TUNEL) assay. These investigators are to be commended for attempting to use a biologic endpoint (apoptosis assay) as a primary endpoint in determining dose escalations. Unfortunately, only 17 of 28 biopsies (61%) evaluated with this assay were found to be informative. In addition, in cases in which apoptosis could be reliably quantified, their assay was shown not to be very helpful in predicting response. Perhaps an alternate biomarker(s) would have been more supportive. In general, apoptosis is an uncommon finding in the evolution of KS,20 particularly in vivo.

In conclusion, although much light has been shed on a unifying KSHV-related pathogenesis, it is apparent that the optimal treatment for KS has yet to be discovered. Comparable standards employing existing therapies are still required to satisfactorily manage all epidemiologic forms of KS. In concert with emerging data regarding the pathogenesis of KS and novel drug targets, we can look forward to investigators sharing their findings after the use of mechanism-based therapeutic strategies and innovative surrogate endpoints to appraise their efficacy.