Skip Navigation LinksHome > February 15, 2002 - Volume 16 - Issue 3 > Antitumor activity of oral 9-cis-retinoic acid in HIV-associ...
Clinical Science

Antitumor activity of oral 9-cis-retinoic acid in HIV-associated Kaposi's sarcoma

Miles, Steven A.; Dezube, Bruce J.; Lee, Jeannette Y.; Krown, Susan E.; Fletcher, Mary Ann; Saville, M. Wayne; Kaplan, Lawrence; Groopman, Jerome; Scadden, David T.; Cooley, Timothy; Von Roenn, Jamie; Friedman-Kien, Alvin*; for the AIDS Malignancy Consortium

Free Access
Article Outline
Collapse Box

Author Information

From the AIDS Malignancy Consortium Operations Center, University of Alabama at Birmingham, Birmingham, Alabama, USA. *See the Appendix for members of the AIDS Malignancy Consortium.

Requests for reprints to: Dr J. Y. Lee, 2001 Third Avenue South – Room 1078, Birmingham, Alabama 35223, USA.

Received: 8 June 2001;

revised: 21 September 2001; accepted: 2 October 2001.

Sponsorship: this work was funded in part by support from Ligand Pharmaceuticals Inc. and the National Cancer Institute (Grant nos. U01 CA70079, U01 CA70081, U01 CA70068, U01 CA70075, U01 CA70047, U01 CA70054, U01 CA70080, U01 CA71375, U01 CA70062, and U01 CA70019).

Collapse Box


Objective: To assess the efficacy, safety and tolerance of oral 9-cis- retinoic acid in HIV-infected patients with Kaposi's sarcoma.

Methods: Sixty-six patients with AIDS-related Kaposi's sarcoma were enrolled at 14 centers; 60 received the study medication and were analyzed and, of these, 45 (75%) had received prior therapy for Kaposi's sarcoma. Once daily oral 9-cis-retinoic acid (alitretinoin, Panretin) was administered at doses up to 140 mg/m2. Most patients (72%) received a maximum dose of 100 mg/m2. Response was assessed using AIDS Clinical Trials Group (ACTG) criteria.

Results: The median age was 38 years and the median absolute CD4 cell count was 194 × 106 cells/l (range 6–784 × 106). Despite the use of three- and four-drug antiviral regimens (83%), the median HIV RNA at baseline was 8701 copies/ml [range < 500 (lower limit of detection) to 4.24 × 106]. The tumor response rate was 37% (95% confidence interval 25–49). Tumor response was associated with improved quality-of-life measures. There was a significant increase in interleukin 6 (IL-6) levels from baseline to week 4. Responders had significantly lower baseline soluble IL-6 receptor levels (P = 0.029) than non-responders. The median time to response was 9 weeks (mean, 13 weeks; range, 4–36). HIV RNA levels did not change significantly during therapy nor did they correlate with tumor responses. Study drug was discontinued by 28 patients for adverse events, which included headache (13) and skin toxicity (10).

Conclusion: Oral 9-cis-retinoic acid is an active antitumor drug for AIDS-related Kaposi's sarcoma. Treatment is associated with skin and constitutional toxicity and further studies are needed to improve its long-term tolerance.

Back to Top | Article Outline


Kaposi's sarcoma is the most common malignancy that affects persons with HIV infection. Although there are several effective systemic treatments for Kaposi's sarcoma, including liposomal anthracyclines, paclitaxel and interferon alfa [1–5], the identification of less toxic, orally bioavailable, effective drugs that work by novel mechanisms would be desirable.

Several retinoids have activity in AIDS-related Kaposi's sarcoma [6]. Crude retinoid preparations and all-trans- retinoic acid applied topically can induce clinical responses [7,8]. Recent clinical trials of topical 9-cis- retinoic acid [9,10] and a related retinoid compound [11] have documented objective clinical responses in approximately 40% of individuals treated. These responses occurred without the concomitant administration of potent antiretroviral regimens. However, topical treatment is impractical for large bulky lesions or widespread disease, and it does not prevent the development of new lesions.

Initial attempts to treat Kaposi's sarcoma with oral retinoids such as 13-cis- retinoic acid [12] and all-trans-retinoic acid resulted in excessive toxicity and variable results [13]. This was possibly a consequence of the complex biopharmacodynamics of the retinoids [14]. In the case of all-trans-retinoic acid, the induction of counter-regulatory proteins, cellular binding proteins and increased hepatic metabolism may have limited its effectiveness despite continued administration. In addition, these counter-regulatory mechanisms are highly variable among patients. 9-cis- retinoic acid is a human vitamin [15] and one of the principal active metabolites of all-trans-retinoic acid [16]. In vitro studies demonstrate that 9-cis- retinoic acid is comparable to all-trans-retinoic acid in its ability to inhibit growth of cells derived from AIDS-related Kaposi's sarcoma lesions [17]. Most importantly, 9-cis- retinoic acid has significantly less complex biopharmacodynamics than all-trans- retinoic acid [18–21] and does not induce counter-regulatory processes that lower drug concentrations over time.

This phase II clinical trial evaluates the ability of a once daily oral dose of 9-cis-retinoic acid (alitretinoin, Panretin) to induce antitumor responses in HIV-infected patients with Kaposi's sarcoma.

Back to Top | Article Outline


This multicenter, open-label, safety and efficacy evaluation of oral 9-cis- retinoic acid (Panretin®, Ligand Pharmaceuticals, San Diego, California, USA) in patients with AIDS-related Kaposi's sarcoma was conducted by the AIDS Malignancy Consortium. Human experimentation guidelines of the US Department of Health and Human Services and the participating institutions were followed.

Back to Top | Article Outline
Inclusion and exclusion criteria

HIV-seropositive patients with biopsy-proven Kaposi's sarcoma, no symptomatic organ involvement and at least five mucocutaneous lesions (each of which was required to have a longest diameter of at least 10 mm or to have been present for at least 30 days) were enrolled. Patients could not have received local or topical therapy within 60 days. All patients were required to have serum aspartate transaminase and alanine transaminase ≤ 5 times the upper limits of normal (ULN), bilirubin ≤ 3 times ULN, creatinine ≤ 2 times ULN, hemoglobin ≥ 80 g/l for females and ≥ 70 g/l for males, absolute neutrophil count ≥ 750 × 106 cells/l, platelet count ≥ 50 × 109 cells/l, serum amylase ≤ 1.5 times ULN, serum calcium ≤ 115 mg/l, and fasting trigylcerides ≤ 10 g/l (or ≤ 5 g/l if risk factors for pancreatitis were present). All patients were required to have a Karnofsky performance status ≥ 60%. Patients with chronic AIDS-defining illnesses were allowed on study as long as they were well controlled on a stable regimen of medication with no changes within 30 days. All patients were to use an effective means of contraception during the study and for 3 months after study treatment was discontinued. Women of child-bearing potential were required to have a negative serum human chorionic gonadotropin pregnancy test result 7 days before study entry.

Patients were excluded from this study if they were pregnant or if they had other active malignancies; cytotoxic therapy within 3 weeks; therapy with interferons, megestrol acetate or pentoxifylline within 4 weeks of study; inability to comply with the protocol; a history of myocardial infarction, significant arrhythmias or class III/IV cardiac functional capacity; concurrent systemic treatment with vitamin A or vitamin A derivatives; or previous local or topical therapy to any indicator lesion.

During treatment, patients were not allowed to receive local or topical therapy to any indicator lesion. In addition, patients were prohibited from receiving concurrent systemic anticancer chemotherapy, hormonal therapy and/or immunotherapy.

All patients were required to be on a stable, approved antiretroviral drug regimen for a minimum of 4 weeks prior to study entry and were encouraged to maintain that treatment through the first 16 weeks of study drug treatment. The antiretroviral regimen recommended was some form of multiagent combination therapy.

Back to Top | Article Outline

After eligible patients provided written informed consent, they were treated with once daily oral doses of 60 mg/m2 9-cis- retinoic acid. The daily dose was escalated by 40 mg/m2 at intervals of 2 weeks, as tolerated, to a daily maximum of 140 mg/m2. After 24 patients were enrolled on the study, the protocol team was informed that headache, pancreatitis, hypertriglyceridemia and alopecia, believed to be related to study drug at a daily dosage of 140 mg/m2, had been observed in a similar study. The protocol was, therefore, amended to reduce the maximum daily dose from 140 to 100 mg/m2. It was recommended that 9-cis- retinoic acid be taken with the evening meal along with acetaminophen (paracetamol) or narcotics. The total dose in milligrams was approximated to within the nearest 30 mg or less using a combination of 25 and 60 mg capsules.

Back to Top | Article Outline
Management of dose-limiting toxicity

If a patient developed dose-limiting toxicity and was receiving a daily dose of 100 mg/m2, the treatment was discontinued for up to 2 weeks. When the toxicity resolved to grade 1 or less, 9-cis- retinoic acid treatment was resumed at a daily dose of 75 mg/m2. Similarly, if dose-limiting toxicity occurred at a daily dose of 75 mg/m2, treatment was discontinued until toxicity resolved and re-instituted at a daily dose of 60 mg/m2. If a patient developed dose-limiting toxicity at 60 mg/m2 daily, the patient was withdrawn from study treatment. Any patient with an unresolved dose-limiting toxicity 2 weeks after stopping treatment at any dose was also withdrawn from study treatment. It was recommended that patients with dry skin be given 1000 IU oral vitamin E four times per day.

Back to Top | Article Outline
Duration of treatment

The initial treatment period was 16 weeks. Patients who had shown benefit and had not experienced unacceptable toxicity were allowed to continue on study drug as long as treatment was deemed beneficial by the treating investigator.

Back to Top | Article Outline
Study procedures

At baseline, all patients had a comprehensive history and physical examination performed along with a complete blood count, serum human chorionic gonadotropin pregnancy test (if applicable), urinalysis, and measurement of T lymphocyte subsets, plasma HIV RNA and cytokine levels. Every 2 weeks during treatment, patients underwent a symptom-directed physical examination and laboratory monitoring. Formal evaluations of Kaposi's sarcoma lesions for response were performed every 4 weeks based on AIDS Clinical Trials Group (ACTG) criteria [22]. This assessment included photographs of all indicator lesions. Complete or partial responses were required to be maintained for at least 4 weeks. Response duration was calculated from the date of the first observed response to the date of progression. Toxicities were graded using the National Cancer Institute common toxicity criteria. The toxicity grading for hypertriglyceridemia was grade 1: 2.5–5 times ULN, grade 2: 5–7 times ULN, grade 3: > 7 times ULN without pancreatitis, and grade 4: ≥ 5 times ULN with pancreatitis.

Back to Top | Article Outline
Cytokine and HIV RNA measurements

Whole blood samples were collected for measurement of interleukin-6 (IL-6), IL-6 receptor and C-reactive protein; these markers of immune system activation are frequently elevated in patients with Kaposi's sarcoma. They were analyzed in a batch assay at a single site using commercially available enzyme-linked immunosorbent assays (R & D Systems, Minneapolis, Minnesota, USA). HIV RNA assays were performed in an National Institute of Allergy and Infectious Diseases-funded and ACTG-certified laboratory. The level of sensitivity of the HIV RNA assay was 500 copies/ml.

Back to Top | Article Outline
Quality of life

At each visit, patients were asked about their symptoms, and the impact of pain and their physical appearance on their quality of life in the preceding 4 weeks. On a scale of 1 to 5, with 5 indicating the most extreme negative effect on quality of life, patients rated the degree to which pain interfered with their normal activities, their level of dissatisfaction with their physical appearance, and their level of distress with respect to five symptoms: edema, hair loss, weight loss, diarrhea, and sores in mouth. The symptom score was the sum of the scores for the five symptoms.

Back to Top | Article Outline
Statistical analysis

A two-stage design [23] was used to test the null hypothesis that the objective response rate was < 15% against the alternative hypothesis that it was > 30% at the 0.05 significance level with power of 0.80. If fewer than four responses were observed in the first 19 patients, the trial would be stopped and the null hypothesis would not be rejected. If at least four of the initial 19 patients met the ACTG criteria for response [22], enrollment was to be expanded to up to 55 evaluable patients. If a minimum of 13 responses among 55 patients were observed, the null hypothesis would be rejected. In order to be considered evaluable, the patient must have received 16 weeks of uninterrupted treatment with 9-cis- retinoic acid unless discontinuing before 16 weeks because of dose-limiting toxicity.

Descriptive statistics were used to describe demographic and other patient characteristics. The primary efficacy endpoint was the tumor response rate up to and following 16 weeks of treatment using the ACTG criteria for evaluating Kaposi's sarcoma overall disease [22]. The binomial proportion and its 95% confidence interval (CI) were used to estimate the overall response rate. The response duration was evaluated using the Kaplan–Meier method. The Wilcoxon signed rank test was used to evaluate changes in HIV viral load from baseline to minimum value on therapy using the log transformation.

Because cytokine levels were not normally distributed, statistical analyses were performed using the logarithmic transformation of these values. The Wilcoxon rank sum test was used to compare responders and non-responders with respect to baseline cytokine levels and changes from baseline to week 4. The Wilcoxon signed rank test was used to assess changes in cytokine levels from baseline to week 4. Spearman rank correlation coefficients were used to evaluate the correlation between IL-6, IL-6 receptor and C-reactive protein at baseline.

Baseline values of the following factors were evaluated by univariate analysis using the Wilcoxon rank sum test to determine their association with response: total cholesterol, triglycerides, IL-6, IL-6 receptor, C-reactive protein, and HIV viral load. Fisher's exact test was used to evaluate the relationship between tumor response and CD4 cell count (< 200 × 106 cells/l versus ≥ 200 × 106 cells/l), HIV viral load (undetectable versus detectable), number of lesions (> 50 or ≤ 50), presence of tumor-associated edema, presence of oral lesions and visceral disease. Factors that were statistically significant at the 0.05 level were incorporated into a multivariate logistic regression model for response. Analysis of variance methods were used to evaluate the effect of time on total cholesterol and triglyceride levels using the logarithmic transformation. If a significant time effect was found, Dunnett's test was used to compare measures taken on treatment with baseline values.

For each patient, the baseline and minimum on-treatment levels were determined for each of the quality-of-life measures. The Wilcoxon rank sum test was used to compare responders and non-responders with respect to baseline measures. Within responders and non-responders, the Wilcoxon signed rank test was used to determine if changes from baseline to minimum value on treatment were significantly different from zero.

Back to Top | Article Outline


Patient characteristics

There were 66 patients enrolled on the study between November 1996 and September 1997, and 60 received the study medication (Table 1). The majority (72%) were non-Hispanic white males and the median age was 38 years (range, 22–60). The median CD4 cell count at baseline was 194 × 106 cells/l (range, 6–784). The median percentage CD4 cells at baseline was 14% (range, 1–42). The median HIV RNA at baseline was 8701 copies/ml (range < 500 to 4.24 × 106 copies/ml).

Table 1
Table 1
Image Tools

Most patients (75%) had received prior therapy for Kaposi's sarcoma (Table 2). Thirty patients had received local therapy, 30 patients had received non-investigational systemic chemotherapy or immunotherapy and 14 patients had received some form of investigational treatment.

Table 2
Table 2
Image Tools

Almost all of the patients (95%) were receiving some form of combination antiretroviral therapy at entry. This included 49 patients (82%) who were receiving protease inhibitor-containing regimens.

Back to Top | Article Outline
Extent of disease

At baseline, 25 patients (42%) had more than 50 cutaneous lesions. The median number of lesions per patient was 17. Oral lesions were present in 18 patients (30%), tumor-associated edema in 22 patients (37%) and visceral disease in five patients (8%).

Back to Top | Article Outline

The median duration of therapy was 15 weeks (range < 1 week to ≥ 164 weeks). Two patients remain on therapy. Forty-three patients (72%) received a maximum dose of 100 mg/m2. Six patients received higher doses (two at 115 mg/m2 and four at 140 mg/m2) before the protocol was amended to limit the maximum dose to 100 mg/m2.

Back to Top | Article Outline
Tumor response

The overall response rate was 37% (95% CI, 25–49;Table 3). One patient achieved a complete response that was pathologically documented. This patient, who had a baseline CD4 cell count of 42 × 106 cells/l and a pretreatment HIV RNA viral load of 2267 copies/ml, had 19 Kaposi's sarcoma lesions confined to the skin. Antiretroviral treatment consisted of nevirapine, stavudine, lamivudine and nelfinavir, and no new drugs were added prior to the complete response. The patient achieved a partial response after 4 months, at which time the viral load was 44 000 copies/ml, and a complete response after 5 months, at which time the viral load was 13 857 copies/ml. Disease progression was documented at month 11, when the viral load was 72 120 copies/ml and the CD4 cell count was 39 × 106 cells/l.

Table 3
Table 3
Image Tools

The median time to response was 9 weeks (range, 4–36 weeks). Twenty-one patients achieved a partial response. The median duration of response has not been reached but exceeds 37 weeks (range, 8 to > 137 weeks). Flattening of 50% or more of previously raised lesions was the basis for partial or complete response for the majority of responding patients (86%). Sixteen responders (73%) had at least a 50% decrease in the number of lesions. Improvement or resolution of tumor-related edema was seen in seven of the eight responding patients who had tumor-related edema at baseline. Of the 22 responders, only three had a change in their antiretroviral drug regimen prior to achieving a response. Eight patients (13%) withdrew from study treatment less than 6 weeks after enrollment because of adverse events and were not evaluable for response.

Back to Top | Article Outline
Laboratory abnormalities

Laboratory abnormalities not necessarily attributed to therapy with 9-cis- retinoic acid included 23 patients who experienced hematologic toxicities of grade 3 or 4. Ten patients had grade 3 and four patients had grade 4 neutropenia. Five patients had grade 3 leukopenia. Anemia of grade 3 or 4 was seen in 12 and 11 patients, respectively. Elevated liver function tests were observed in three patients. Seventeen patients (28%) experienced grade 3 hypertriglyceridemia, which led to treatment interruption in five patients and dose reduction in one patient.

Back to Top | Article Outline
Toxicity and adverse events

The most common grade 3 or higher adverse event was headache (n = 18; 30%); all were reported as possibly or definitely related to study drug. Nine patients experienced grade 3 or greater adverse events in the skin and mucous membranes, which included cellulitis, cracking and peeling of skin, desquamation of the soles of the feet, erythematous rash, foot ulcer, onycholysis and alopecia. These toxicities were considered related to study drug by the investigators in five patients and resulted in modification or discontinuation of study drug dosage. Additional grade 3 or 4 toxicities that were considered possibly or definitely related to the study drug occurred in a total of 16 patients. These included grade 4 hypercalciuria, depression, hypotension, and pneumonia, and grade 3 infection, peripheral neuropathy, hypertension, deep vein thrombosis, pneumonia, diarrhea, constipation, myalgias, leg edema, fever, fatigue, malaise, edema, abdominal pain, vomiting, basal cell carcinoma, weakness, osteonecrosis and pancreatitis. Each of these was reported in one, or at most two, patients. One patient was hospitalized during the first month on study with fever and dyspnea, which were thought to represent either a lung infection or heart failure. The symptoms resolved after 4 days with antibiotics and oxygen and were not ascribed to treatment. The study drug was discontinued during the episode and was then restarted without dose reduction and without recurrence of symptoms. One patient was hospitalized for appendicitis and one patient developed a brain lesion from CNS toxoplasmosis. Neither event was believed by the investigators to be related to study drug.

Twenty-eight patients discontinued study drug therapy because of adverse events, which included one or more of the following: headache (13), skin toxicity (10), and depression (2). There were four patients who discontinued treatment because of both headache and skin toxicity. Skin toxicity appeared to be cumulative during study drug administration. The decision to terminate treatment because of headaches was initiated by the patient in eight (62%) cases. One patient was hospitalized and nine were managed with prescription medications, including narcotic analgesics. Gender, race and HIV risk group were not significantly related to discontinuation of therapy because of headache.

Median triglyceride and cholesterol levels increased significantly from baseline (P < 0.05) (Table 4). Most of the changes occurred in the first 4 weeks of treatment. There was one episode of clinical pancreatitis.

Table 4
Table 4
Image Tools
Back to Top | Article Outline
HIV RNA changes

Baseline viral load measurements were available for 44 patients. There were no significant changes in plasma HIV RNA copy number during study drug treatment. Fourteen patients had an undetectable HIV RNA level at baseline. Levels for nine of these patients remained undetectable throughout the period of study drug treatment; one rose to a detectable level, and four did not have follow-up measurements. Of the 30 patients in whom HIV RNA was detectable at baseline, 20 had levels that remained detectable, three had levels that were undetectable after initiation of study drug treatment and seven patients did not have follow-up measurements. Three patients had > 1 log10 reduction in HIV RNA levels during study drug treatment without a change in antiviral therapy. In two of these patients, the levels went from detectable to undetectable.

Back to Top | Article Outline
Cytokine levels

For the 43 patients for whom baseline cytokine levels were available, C-reactive protein levels were significantly correlated with IL-6 levels (r = 0.345;P = 0.023) and IL-6 receptor levels (r = 0.409;P = 0.007). Paired baseline and week 4 measurements were available for IL-6 and IL-6 receptor levels in 25 patients, and for C-reactive protein in 21 patients. There was a significant increase in median IL-6 levels from baseline (6.14 pg/ml) to 4 weeks (22.07 pg/ml) (P = 0.014). There were no significant changes in either IL-6 receptor or C-reactive protein levels.

Back to Top | Article Outline
Quality of life

Baseline levels were comparable for responders and non-responders for all quality-of-life measures. Neither responders nor non-responders showed a significant change in pain score from baseline to the minimum value on treatment. Responders showed a significant decrease in physical appearance score (P < 0.001) and symptom score (P < 0.001), whereas no significant change was seen in non-responders.

Back to Top | Article Outline
Correlates of antitumor response

Potential laboratory correlates of antitumor response were investigated at baseline. For HIV RNA, the response rate was 5/14 (36%) in patients with undetectable HIV RNA at baseline and 13/30 (43%) in patients with detectable levels at baseline (P = 0.748). Of the four patients whose HIV RNA levels decreased more than 1 log10 during study drug treatment and/or went from detectable to undetectable levels, two responded (50%). Although the response rate among patients with baseline CD4 cell counts ≥ 200 × 106 cells/l was almost twice as high as that for patients with CD4 cell counts < 200 × 106 cells/l (50% versus 26%), this difference did not achieve statistical significance. Responding patients showed a significantly lower baseline IL-6 receptor level than non-responders (P = 0.042;Fig. 1). With respect to baseline clinical characteristics, there was no association between response and the number of lesions (> 50 or ≤ 50), the presence of tumor-associated edema, or presence of visceral disease. Gender, race and HIV risk group were not significantly associated with response.

Fig. 1
Fig. 1
Image Tools
Back to Top | Article Outline


Despite the dramatic decline in the incidence of Kaposi's sarcoma with recent advances in antiretroviral therapy, it continues to be a source of significant morbidity and occasional mortality among HIV-infected patients [24]. Although liposomal anthracyclines [1,2,25,26] and paclitaxel [3,4] are highly effective in controlling advanced Kaposi's sarcoma, they may induce significant toxicities including neutropenia, hand/foot syndrome, and, in the case of paclitaxel, neuropathy and hair loss. In addition, all of these agents require intravenous administration. An active, well-tolerated, oral agent could improve the quality of life for patients with this disease.

This trial demonstrates that 9-cis- retinoic acid is an active oral agent in patients with Kaposi's sarcoma. Antitumor responses were seen in various stages of disease as well as in patients who received prior chemotherapy, had low CD4 cell counts or high HIV RNA levels. This suggests that the mechanism of 9-cis- retinoic acid is independent of the level of residual immunological function, HIV replication or tumor burden. Interestingly, the response rate observed in this trial is nearly identical to that observed in a small phase II trial of oral all-trans retinoic acid [7] as well as a large randomized, placebo-controlled trial of topical 9-cis- retinoic acid [10]. This suggests that 9-cis- retinoic acid, whether applied topically or administered orally, could be of value to patients with Kaposi's sarcoma across a broad spectrum of disease severity.

Quality of life improved for patients whose tumors responded, as measured by greater satisfaction with their appearance and a reduction in a variety of disease- or treatment-related symptoms. However, the agent was associated with significant toxicity. Many of the observed toxicities are known to be associated with retinoids. The relationship of some of the observed laboratory abnormalities to the study drug is difficult to assess, however, in individuals with advanced HIV infection receiving multiple concomitant medications. The major initial toxicity was headache, whereas the major long-term toxicities are cumulative mucocutaneous toxicities, including dry pruritic rashes, desquamation and cheilitis, and hematological toxicity. The cutaneous toxicity was observed despite prophylactic administration of oral vitamin E. Interestingly, unlike other retinoids, there was no evidence of tachyphylaxis to drug toxicities. In fact, the levels of plasma lipids and skin toxicity increased over time.

To avoid administering an agent with significant toxicity to patients who are unlikely to benefit, it would be useful to identify baseline or early on-treatment factors that correlate with subsequent response to 9-cis- retinoic acid. We observed no association between Kaposi's sarcoma response and baseline CD4 cell count or HIV viral load, or changes in IL-6, C-reactive protein, tumor burden or prior therapy. We did see increases in IL-6, C-reactive protein and serum lipid levels. These occurred early but were not associated with response. Although baseline levels of the soluble IL-6 receptor were significantly lower in responding patients than in non-responders, there was considerable overlap between the two groups. Thus, this study provides no firm criteria on which to base future patient selection for trials of this agent. In fact, the observed increase in IL-6 levels is difficult to reconcile with the observation that elevated IL-6 levels predict the subsequent development of Kaposi's sarcoma in HIV-infected individuals [27,28].

All trans-retinoic acid downregulates IL-6 expression in a number of human cells [29–31] and cell lines [32,33], including freshly isolated Kaposi sarcoma cells [34]. This is presumably secondary to its antagonism of the IL-6 nuclear factor receptor by the RAR-PPAR heterodimer nuclear receptor [30,33–35]. However, in the present study, we observed just the opposite: treatment with 9-cis- retinoic acid was associated with increased circulating levels of IL-6. While this is similar to the effect of all-trans retinoic acid on IL-6 in a small trial in multiple myeloma [36], it is difficult to explain this result in light of the in vitro data, the antitumor responses observed and the lack of corresponding changes in the level of soluble IL-6 receptors [29]. It is possible that, in the setting of human herpes virus 8 infection, more specific measurements of viral IL-6 may be more germane and should be considered in future studies.

Antiretroviral therapy for HIV infection has improved dramatically since the mid-1990s. In turn, the incidence of Kaposi's sarcoma has declined by over 50% and there are reports of spontaneous regression of Kaposi's sarcoma among patients treated with potent antiretroviral regimens. This has complicated the interpretation of several recent trials of therapy for Kaposi's sarcoma. In our study, the majority of patients had poorly controlled HIV replication despite three- and four-drug regimens that included protease inhibitors. Few patients had a fall in the level of HIV RNA of greater than 1 log10 or to below the limits of detection for the assay, yet 37% had an antitumor response. In addition, the rate of response was the same in those with and without detectable levels of HIV RNA at baseline and in those who had an antiviral response. Therefore, the observed responses are unlikely to have resulted from concomitant HIV therapies.

We conclude that oral 9-cis- retinoic acid has activity in the treatment of Kaposi's sarcoma. Its ease of administration and presumably non-cytotoxic mechanism of action support its further development, but its toxicities, in particular its cumulative cutaneous toxicity, may limit its use. There is also concern that the cholesterol and triglyceride elevations induced by 9-cis- retinoic acid may exacerbate those already observed with protease inhibitor therapy, but only long-term studies can determine whether these changes are associated with adverse cardiac outcomes. Future studies should be directed toward the development of more tolerable treatment regimens, which could include intermittent dosing schedules or the use of lower, better tolerated doses in combination with other active agents, and the identification of biologically relevant factors predictive of response.

Back to Top | Article Outline


The AIDS Malignancy Consortium deeply thanks all the participating patients and all their referring physicians. We would like to thank Richard Yocum of Ligand Pharmaceutical Inc. for his guidance in developing the protocol and medical monitoring during the conduct of the study. We would also like to thank Mortimer Smithberg and Bertha Bradford for their secretarial efforts.

Back to Top | Article Outline


1. Northfelt DW, Dezube BJ, Thommes JA. et al. Pegylated-liposomal doxorubicin versus doxorubicin, bleomycin, and vincristine in the treatment of AIDS-related Kaposi's sarcoma: results of a randomized phase III clinical trial. J Clin Oncol 1998, 16: 2445–2451.

2. Northfelt DW, Dezube BJ, Thommes JA. et al. Efficacy of pegylated-liposomal doxorubicin in the treatment of AIDS-related Kaposi's sarcoma after failure of standard chemotherapy. J Clin Oncol 1997, 15: 653–659.

3. Gill PS, Tulpule A, Espina BM. et al. Paclitaxel is safe and effective in the treatment of advanced AIDS-related Kaposi's sarcoma. J Clin Oncol 1999, 17: 1876–1883.

4. Welles L, Saville MW, Lietzau J. et al. Phase II trial with dose titration of paclitaxel for the therapy of human immunodeficiency virus-associated Kaposi's sarcoma. J Clin Oncol 1998, 16: 1112–1121.

5. Krown SE, Real FX, Krim M. et al. Interferons and other biological response modifiers in the treatment of Kaposi's sarcoma. Front Radiat Ther Oncol 1985, 19: 138–149.

6. Corbeil J, Rapaport E, Richman DD, Looney DJ. Antiproliferative effect of retinoid compounds on Kaposi's sarcoma cells. J Clin Invest 1994, 93: 1981–1986.

7. Saiag P, Pavlovic M, Clerici T. et al. Treatment of early AIDS-related Kaposi's sarcoma with oral all-trans-retinoic acid: results of a sequential non-randomized phase II trial. Kaposi's Sarcoma ANRS Study Group. Agence Nationale de Recherches sur le SIDA. AIDS 2000, 12: 2169–2176.

8. Bonhomme L, Fredj G, Averous S. et al. Topical treatment of epidemic Kaposi's sarcoma with all-trans-retinoic acid. Ann Oncol 1991, 2: 234–235.

9. Duvic M, Friedman-Kien AE, Looney DJ. et al. Topical treatment of cutaneous lesions of acquired immunodeficiency syndrome-related Kaposi sarcoma using alitretinoin gel: results of phase 1 and 2 trials. Arch Dermatol. 2000, 136: 1461–1469.

10. Walmsley S, Northfelt DW, Melosky B, Conant M, Friedman-Kien AE, Wagner B. Treatment of AIDS-related cutaneous Kaposi's sarcoma with topical alitretinoin (9-cis-retinoic acid) gel. Panretin Gel North American Study Group. J Acquir Immune Defic Syndr 1999, 22: 235–246.

11. Leoung GS, Aboulafia D, Millikan L et al. Treatment of Kaposi's sarcoma using Targretin (LGD1069), a topical retinoid gel. XI International Conference on AIDS. Vancouver, July 1996 [abstract We.B. 3241].

12. Rosenthal E, Pesce A, Vinti H, Reboulot B, Hoffman P, Cassuto JP. Isotretinoin plus interferon alpha-2a in AIDS-related Kaposi's sarcoma.IX International Conference on AIDS. Berlin, June 1993 [abstract PO-B12-1607].

13. Von Roenn J, von Gunten C, Mullane M, French S, Blough R, Benson AB. All-transretinoic acid (ATRA) in the treatment of AIDS-related Kaposi's sarcoma. IX International Conference on AIDS. Berlin, June 1993 [abstract PO-B12-1571].

14. Adamson PC, Bailey J, Pluda J. et al. Pharmacokinetics of all-trans-retinoic acid administered on an intermittent schedule. J Clin Oncol 1995, 13: 1238–1241.

15. Heyman RA, Mangelsdorf DJ, Dyck JA. et al. 9-cis-Retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 1992, 68: 397–406.

16. Urbach J, Rando RR. Isomerization of all-trans-retinoic acid to 9-cis-retinoic acid. Biochem J 1994, 299: 459–465.

17. Corbeil J, Rapaport E, Richman D, Badel P, Looney DJ. Inhibition and killing of Kaposi's sarcoma cells by retinoids. Second Conference on Human Retroviruses and Related Infections. Washington, DC, January–February 1995 [p. 79].

18. Kurie JM, Lee JS, Griffin T. et al. Phase I trial of 9-cis- retinoic acid in adults with solid tumors. Clin Cancer Res 1996, 2: 287–293.

19. Rizvi NA, Marshall JL, Ness E. et al. Phase I study of 9-cis-retinoic acid (ALRT1057 capsules) in adults with advanced cancer. Clin Cancer Res 1998, 4: 1437–1442.

20. Weber C, Dumont E. Pharmacokinetics and pharmacodynamics of 9-cis-retinoic acid in healthy men. J Clin Pharmacol 1997, 37: 566–574.

21. Sass JO, Masgrau E, Saurat JH, Nau H. Metabolism of oral 9-cis-retinoic acid in the human. Identification of 9-cis-retinoyl-beta-glucuronide and 9-cis-4-oxo-retinoyl-beta-glucuronide as urinary metabolites. Drug Metab Dispos 1995, 23: 887–891.

22. Krown SE, Testa MA, Huang J. AIDS-related Kaposi's sarcoma: prospective validation of the AIDS Clinical Trials Group staging classification. AIDS Clinical Trials Group Oncology Committee. J Clin Oncol 1997, 15: 3085–3092.

23. Krown SE, Metroka C, Wernz JC. Kaposi's sarcoma in the acquired immune deficiency syndrome: a proposal for uniform evaluation, response, and staging criteria. AIDS Clinical Trials Group Oncology Committee. J Clin Oncol 1989, 7: 1201–1207.

24. Simon R. Optimal two-stage designs for phase II clinical trials. Control Clin Trials 1989, 10: 1–10.

25. Miles SA, Wang H, Elashoff R, Mitsuyasu RT. Improved survival for patients with AIDS-related Kaposi's sarcoma. J Clin Oncol 1994, 12: 1910–1916.

26. Gill PS, Wernz J, Scadden DT. et al. Randomized phase III trial of liposomal daunorubicin versus doxorubicin, bleomycin, and vincristine in AIDS-related Kaposi's sarcoma. J Clin Oncol 1996, 14: 2353–2364.

27. Sturzl M, Zietz C, Eisenburg B. et al. Liposomal doxorubicin in the treatment of AIDS-associated Kaposi's sarcoma: clinical, histological and cell biological evaluation. Res Virol 1994, 145: 261–269.

28. Dourado I, Martinez-Maza O, Kishimoto T, Suzuki H, Detels R. Interleukin 6 and AIDS-associated Kaposi's sarcoma: a nested case control study within the Multicenter AIDS Cohort Study. AIDS Res Hum Retroviruses 1997, 13: 781–788.

29. Sawatsri S, Desai N, Rock JA, Sidell N. Retinoic acid suppresses interleukin-6 production in human endometrial cells. Fertil Steril 2000, 73: 1012–1019.

30. Kozawa O, Tokuda H, Kaida T, Matsuno H, Uematsu T. Retinoic acid suppresses interleukin-6 synthesis induced by prostaglandins in osteoblasts. Prostaglandins Leukot Essent Fatty Acids 1998, 58: 215–219.

31. Ahmed N, Sammons J, Khokher MA, Hassan HT. Retinoic acid suppresses interleukin 6 production in normal human osteoblasts. Cytokine 2000, 12: 289–293.

32. Chen YH, Lavelle D, DeSimone J, Uddin S, Platanias LC, Hankewych M. Growth inhibition of a human myeloma cell line by all-trans retinoic acid is not mediated through downregulation of interleukin-6 receptors but through upregulation of p21(WAF1). Blood 1999, 94: 251–259.

33. Ogata A, Nishimoto N, Shima Y, Yoshizaki K, Kishimoto T. Inhibitory effect of all-trans retinoic acid on the growth of freshly isolated myeloma cells via interference with interleukin-6 signal transduction. Blood 1994, 84: 3040–3046.

34. Nagpal S, Cai J, Zheng T. et al. Retinoid antagonism of NF-IL6: insight into the mechanism of antiproliferative effects of retinoids in Kaposi's sarcoma. Mol Cell Biol 1997, 17: 4159–4168.

35. DiSepio D, Malhotra M, Chandraratna RA, Nagpal S. Retinoic acid receptor-nuclear factor-interleukin 6 antagonism. A novel mechanism of retinoid-dependent inhibition of a keratinocyte hyperproliferative differentiation marker. J Biol Chem 1997, 272: 25555–25559.

36. Niesvizky R, Siegel DS, Busquets X. et al. Hypercalcaemia and increased serum interleukin-6 levels induced by all-trans retinoic acid in patients with multiple myeloma. Br J Haematol 1995, 89: 217–218.

Back to Top | Article Outline

Participating Clinical Centers: Zale Bernstein, Roswell Park Cancer Institute, Buffalo, New York; Michael Caligiuri, Ohio State University, Columbus, Ohio; Tony Cheung, Mount Sinai Medical Center, New York; Timothy Cooley, Boston Medical Center, Boston, Massachusetts; Bruce Dezube and Jerome Groopman, Beth Israel Deaconess Medical Center, Boston, Massachusetts; Margaret Fischl, University of Miami, Miami, Florida; Alvin Friedman-Kien, New York University, New York; Lawrence Kaplan, San Francisco General Hospital, San Francisco, California; Barbara Klencke, Mount Zion Hospital, San Francisco, California; Susan Krown and Ariela Noy, Memorial Sloan-Kettering Cancer Center, New York; Bryan Leigh, University of California, Davis, California; David Lyter, Illinois Masonic Cancer Center, Chicago, Illinois; Robert Mass, Kaiser-Permanente Medical Center, San Francisco, California; Ronald Mitsuyasu, University California, Los Angeles, California; Lee Ratner, Washington University, St Louis, Missouri; M. Wayne Saville, University of California, San Diego, California; David Scadden, Massachusetts General Hospital, Boston, Massachusetts; and Jamie Von Roenn, Northwestern University, Chicago, Illinois. Operations Center: Jeannette Lee, University of Alabama at Birmingham, Birmingham, Alabama. Central Laboratories: Mary Ann Fletcher, University of Miami, Miami, Florida; Parkash Gill, University of Southern California, Los Angeles, California; Judith Lieberman, Center for Blood Research, Boston, Massachusetts; and Brooks Jackson, Johns Hopkins University, Baltimore, Maryland. Cited Here...

Cited By:

This article has been cited 2 time(s).

Current Opinion in Oncology
Targeting angiogenesis for the treatment of sarcoma
Balasubramanian, L; Evens, AM
Current Opinion in Oncology, 18(4): 354-359.
PDF (140) | CrossRef
Current Opinion in Oncology
Update in Kaposi sarcoma
Noy, A
Current Opinion in Oncology, 15(5): 379-381.

PDF (167)
Back to Top | Article Outline

HIV-1; 9-cis-retinoic acid; Kaposi's sarcoma; antitumor therapy

© 2002 Lippincott Williams & Wilkins, Inc.


Article Tools



Article Level Metrics

Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.