SEARCH

SEARCH BY CITATION

Keywords:

  • Kaposi sarcoma;
  • acquired immunodeficiency syndrome;
  • natural history;
  • highly active antiretroviral therapy

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

The objective of the current study was to evaluate the impact of highly active antiretroviral therapy (HAART) on clinical characteristics of presentation and the natural history of Kaposi sarcoma (KS) in patients already receiving HAART at the time of KS diagnosis.

METHODS

The authors conducted a retrospective cohort study comparing epidemiologic, clinical, and outcome data for 160 patients who were naive to HAART at the time of KS diagnosis (KS-naive) with the corresponding data for 51 patients already receiving HAART at the time of KS diagnosis (KS-HAART). The analysis included all patients with a diagnosis of KS since January 1996 within two Italian cohorts of patients with human immunodeficiency virus.

RESULTS

Immunologic and virologic status at the time of KS diagnosis were significantly more favorable in the KS-HAART group than in the KS-naive group. The frequency of cutaneous involvement was similar in both groups, but cutaneous disease was more indolent among KS-HAART patients, with 1 anatomic site of involvement in 9 patients (21%) and less than 10 lesions in 26 patients (60%), compared with 16 patients (12%; P = 0.06) and 47 patients (34%; P = 0.01), respectively, in the KS-naive group. A smaller proportion of KS-HAART patients presented with visceral disease (24% vs. 39%; P = 0.06); in particular, gastrointestinal tract involvement was significantly less frequent among KS-HAART patients (14%) compared with KS-naive patients (28%; P = 0.05). Median survival was not reached in either group, and the 3-year survival rates of KS-HAART patients (64%) and KS-naive patients (78%) were not significantly different.

CONCLUSIONS

The data from the current study indicate that KS exhibits a less aggressive presentation in patients already receiving HAART compared with patients who are naive to HAART at KS diagnosis. Natural history and outcome do not appear to be influenced by the initiation of HAART before development of KS. Cancer 2003. © 2003 American Cancer Society.

Kaposi sarcoma (KS), a life-threatening multifocal neoplasm, has been the most common acquired immunodeficiency syndrome (AIDS)-associated malignancy since the beginning of the human immunodeficiency virus (HIV) epidemic.1 The introduction of effective antiretroviral therapies in the 1990s has been associated with significant epidemiologic changes—in particular, a sharp decline in the incidence of KS—and has had a favorable impact on the therapeutic management of KS.2–10 Nonetheless, to our knowledge, there are no published studies evaluating the effects of highly active antiretroviral therapy (HAART) on the clinical characteristics at presentation and the natural history of KS in patients already following a stable HAART regimen at the time of KS diagnosis. Considering the direct (e.g., tat gene, cytokine network alterations) and indirect roles (e.g., in causing immunodeficiency) of HIV in the pathogenesis of AIDS-related KS and the possible anti–human herpesvirus 8 and antiangiogenic activity of protease inhibitors,11 it is likely that the development and growth of KS during HAART is profoundly modified and that its clinical presentation and natural history also are altered.

With the aim of addressing this issue, we have collected epidemiologic, clinical, and outcome data on 211 individuals with AIDS-related KS from 2 Italian cohorts of patients with HIV: the GICAT (Italian Cooperative Group on AIDS and Tumors) and the ICONA (Italian Cohort of Naive Antiretroviral Patients).

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

GICAT has recruited HIV-positive patients with malignant disease from a number of centers throughout Italy since 1986; many of these patients have been enrolled in clinical trials. ICONA is a multicenter observational study that has recruited HIV-positive patients who are naive to antiretroviral drugs at enrollment from 65 infectious disease centers in Italy since 1997. We considered all consecutive patients with an initial diagnosis of KS made in January 1996 (the time at which HAART became widely available in Italy) or later; the following were inclusion criteria for the analysis: histologically confirmed KS, serologic evidence of HIV infection, at least one follow-up visit after diagnosis of KS, and treatment with HAART before and/or after diagnosis of KS (combination therapy with two reverse transcriptase inhibitors was considered to be HAART).

Two hundred eleven patients met the inclusion criteria and entered the study. Among them, 2 groups of patients were identified: 51 patients who already were receiving HAART at the time of KS diagnosis (KS-HAART) and 160 patients who were naive to HAART at KS diagnosis and who initiated therapy after diagnosis (KS-naive). At least 3 months of HAART before diagnosis of KS was required for inclusion in the KS-HAART group. A comparison of presenting features and outcomes between the two groups was performed.

Epidemiologic and HIV-related clinical data (age at time of KS diagnosis, gender, race, HIV exposure category, HIV stage at KS diagnosis, immunologic and virologic data at KS diagnosis, antiretroviral therapy history, vital status, and eventual cause of death) were available from the ICONA database and from GICAT case report forms. A questionnaire was sent to participating ICONA and GICAT centers to retrospectively collect the following additional information: KS stage (according to the AIDS Clinical Trials Group [ACTG] criteria); clinical presentation (i.e., skin extension, mucosal and visceral involvement, and atypical sites); diagnostic procedures used to assess visceral involvement; type of KS treatment; and, if necessary, follow-up update.

Although staging procedures probably were not uniform among the various centers, the following criteria were used to define gastrointestinal (GI) tract involvement of KS: positive endoscopic evaluation, or, when endoscopy was not performed, oropharyngeal KS with signs and symptoms of GI disease and without microbiologic evidence of GI tract infection. Furthermore, pulmonary involvement of KS was defined by the following criteria: a positive bronchoscopic evaluation or evidence of interstitial infiltrates on chest radiograph with positive thallium scan and without microbiologic evidence of pulmonary infection.

Patients were staged according to the ACTG criteria, which are based on the evaluation of tumor extension (T), CD4 cell count (I), and a patient's systemic status (S).12

Information on survival was obtained via active follow-up on verification of vital status up to April 15, 2002, and survival was measured from the date of diagnosis to the date of death. Survival analysis was performed using the Kaplan–Meier method,13 and the log-rank test was used to detect differences between groups (KS-HAART vs. KS-naive). Differences were analyzed using the chi-square test. In all cases, statistical significance was claimed when P ≤ 0.05 (two-sided).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

General Characteristics

One hundred fifty-nine patients were recruited from GICAT, and 52 patients were recruited from ICONA. One hundred ninety patients were male, and 123 patients (59%) were older than age 35 years; the median age at KS diagnosis was 37 years (range, 20–80 years). The majority of patients (52%) were homosexual or bisexual, while 25% of patients were heterosexual and 17% were drug users. Fifty-one patients (24%) had experienced an AIDS-defining disease before diagnosis of KS. Most patients had severe impairment of the immune system and uncontrolled HIV viremia at KS diagnosis, with a median CD4 cell count of 86/μL (range, 0–1117/μL) and a median HIV viremia level of 89,000 copies/mL (range, < 50 to > 1,000,000 copies/mL). With regard to KS staging, 136 patients had poor-risk tumor extension (T1), 148 patients had less than 200 CD4 cells/μL (I1), and 125 patients had poor-risk systemic disease (S1). Only 17 patients had no poor-risk features (T0I0S0), whereas most patients (n = 48) had all poor-risk features (T1I1S1). Sixty-nine patients received chemotherapy; most of these 69 were enrolled in clinical trials. The most common regimens were paclitaxel-vinorelbine (n = 33), liposomal daunorubicin (n = 19), and doxorubicin-bleomycin-vincristine (n = 14). Twenty patients received radiotherapy.

After a median follow-up period of 24 months (range, 1–72 months), median survival had not been reached. The survival rate at 3 years was 75%. Progression of KS was the most common cause of death, accounting for 70% of all cases (n = 33), while 30% of deaths (n = 14) were caused by opportunistic infections.

Antiretroviral Therapy

All patients received HAART. Fifty-one patients already were receiving HAART at the time of KS diagnosis (triple combination therapy with protease inhibitors [PI; n = 31], dual combination therapy with nucleoside reverse transcriptase inhibitors [NRTI; n = 14], or triple combination therapy with nonnucleoside reverse transcriptase inhibitors [NNRTI; n = 6]), whereas 160 patients began receiving HAART at or after diagnosis of KS (triple combination therapy with PI [n = 148] or NNRTI [n = 12]). The median duration of HAART (before diagnosis of KS) among the 51 KS-HAART patients was 11 months (range, 3–61 months); median durations were 7, 18, and 13 months for patients receiving PI, NRTI, and NNRTI, respectively. Among the 51 KS-HAART patients, 36 changed therapy regimens and began receiving novel HAART combinations. Of these 36 patients, 28 (14 of 14 patients who previously received dual combination therapy with NRTI, 6 of 6 patients who previously received triple combination therapy with NNRTI, and 8 patients who previously received triple combination therapy with PI) began following a different triple combination therapy regimen involving PI and 8 (all of whom previously received triple combination therapy with PI) began receiving triple combination therapy with NNRTI. Fifteen patients did not change HAART regimens at diagnosis of KS. Data on the response of KS to HAART were available for 130 patients; 83 patients achieved a complete or partial response, and 29 experienced disease progression.

Clinical Data

A comparison of demographic, immunologic, and virologic characteristics between KS-HAART and KS-naive patients can be found in Table 1. Gender distribution, median age, and mode of HIV infection transmission were similar in both groups. Patients in the KS-HAART group had a significantly more favorable immunologic situation at KS diagnosis compared with patients in the KS-naive group (median CD4 cell count, 129 × 106/L [range, 3–1117 × 106/L] vs. 61 × 106/L [range, 0–864 × 106/L]; P = 0.02). Virologic status was significantly more favorable in KS-HAART patients compared with KS-naive patients, with HIV RNA viremia levels less than 30,000 copies/mL in 27 patients (63%) and 21 patients (18%), respectively (P = 0.001). No patient had undetectable HIV RNA viremia levels in the KS-naive group, whereas HIV RNA viremia was undetectable in 10 patients (20%) in the KS-HAART group (P = 0.001).

Table 1. General Characteristics of 211 Patients with Acquired Immunodeficiency Syndrome–Related Kaposi Sarcoma: Comparison of Patients Who Were and Were Not Naive to Highly Active Antiretroviral Therapy at the Time of Diagnosis
General characteristicKS-HAART n (%)KS-naive n (%)P
  1. KS-HAART: patients who were receiving highly active antiretroviral therapy at the time of diagnosis of Kaposi sarcoma; KS-naive: patients who were naive to highly active antiretroviral therapy at the time of diagnosis of Kaposi sarcoma; IVDU: intravenous drug use; AIDS: acquired immunodeficiency syndrome; HIV: human immunodeficiency virus; KS: Kaposi sarcoma; T: classification of tumor extension; I: classification of CD4 cell count; S: classification of systemic disease status.

All patients51160 
Gender   
 Male47 (92)143 (89) 
 Female4 (8)17 (11) 
Age (yrs)   
 Median3837 
 Range22–8122–73 
Risk factor   
 IVDU9 (18)27 (17) 
 Homosexuality29 (59)77 (49) 
 Heterosexuality8 (16)44 (28) 
Previous diagnosis of AIDS23 (45)34 (21)≤ 0.001
CD4 cell count (× 106/L)   
 Median12961≤ 0.02
 Range3–11170–864 
HIV RNA   
 ≤ 30,000 copies/mL27 (63)21 (18)≤ 0.001
 Undetectable10 (45)0 (—)≤ 0.001
KS classification   
 T0/123/28 (45/55)52/108 (33/67) 
 I0/122/29 (43/57)48/109 (31/69) 
 S0/121/30 (41/59)104/56 (65/35)≤ 0.003

The distribution of disease stages was similar in the two groups, with the majority of patients exhibiting poor-risk tumor extension (T1) and immunologic variables (I1) in both cases. With respect to the HIV- and KS-related S classification, the KS-HAART group had a higher rate of poor-risk cases compared with the KS-naive group, with 30 patients (59%) and 56 patients (35%), respectively, having S1 disease (P = 0.003). These data presumably are associated with the higher rate of previous AIDS-defining illness observed in the KS-HAART group. Twenty-three patients (45%) in the KS-HAART group had a previous diagnosis of AIDS at the time of KS diagnosis, compared with 34 patients (21%) in the KS-naive group (P = 0.001).

Table 2 shows a comparison of the clinical presentation features between the two groups. There was a similar frequency of cutaneous involvement in both groups (92% in the KS-HAART group vs. 93% in the KS-naive group), but localized and indolent cutaneous disease were more common in the KS-HAART group, with 1 anatomic site of involvement in 9 patients (21%) and less than 10 lesions in 26 patients (60%), compared with 16 patients (12%) and 47 patients (34%), respectively, in the KS-naive group (P = 0.06 and P = 0.01, respectively). The prevalence of confinement of KS to the skin was similar in both groups, with 20 patients (39%) in the KS-HAART group and 45 patients (28%) in the KS-naive group exhibiting only cutaneous involvement. No difference was found between the two groups with respect to the prevalence of lymphedema (17% in the KS-HAART group vs. 15% in the KS-naive group), the frequency of mucosal involvement (31% in the KS-HAART group vs. 41% in the KS-naive group), and the proportion of patients with lymph node involvement (25% in the KS-HAART group vs. 24% in the KS-naive group). A smaller proportion of patients in the KS-HAART group presented with visceral disease (24% vs. 39%; P = 0.06); in particular, GI tract involvement was significantly less common in the KS-HAART group (14%) than in the KS-naive group (28%; P = 0.05). Pulmonary disease also occurred less frequently in the KS-HAART group (12% vs. 16%), but the difference was not statistically significant.

Table 2. Clinical Characteristics at Presentation with Kaposi Sarcoma: Comparison of Patients Who Were and Were Not Naive to Highly Active Antiretroviral Therapy at the Time of Diagnosis
Clinical characteristicKS-HAART n (%)KS-naive n (%)
  • KS-HAART: patients who were receiving highly active antiretroviral therapy at the time of diagnosis of Kaposi sarcoma; KS-naive: patients who were naive to highly active antiretroviral therapy at the time of diagnosis of Kaposi sarcoma.

  • a

    P = 0.06 for the difference between the two groups.

  • b

    P = 0.01 for the difference between the two groups.

  • c

    P = 0.05 for the difference between the two groups.

Skin involvement47 (92)148 (93)
 Localized (1 body site)a 9 (21) 16 (12)
 Diffuse (> 2 body sites)a34 (79)120 (88)
 Indolent (< 10 lesions)b26 (60) 47 (34)
 Aggressive (> 100 lesions)b 4 (9) 15 (11)
 Skin as only disease site20 (39) 45 (28)
Lymphedema 8 (17) 23 (15)
Mucous membrane involvement15 (31) 63 (41)
 Conjunctival 1 (1)  2 (1)
Lymph node involvement12 (25) 37 (24)
Visceral involvementa12 (24) 61 (39)
 Gastroentericc 7 (14) 44 (28)
 Pulmonary 6 (12) 25 (16)
Involvement of unusual sites 3 (6)  3 (2)

A comparison of treatment strategies and responses between the two groups is shown in Table 3.

Table 3. Results of Therapy for Kaposi Sarcoma: Comparison of Patients Who Were and Were Not Naive to Highly Active Antiretroviral Therapy at the Time of Diagnosis
 KS-HAART n (%)KS-naive n (%)
  • KS-HAART: patients who were receiving highly active antiretroviral therapy at the time of diagnosis of Kaposi sarcoma; KS-naive: patients who were naive to highly active antiretroviral therapy at the time of diagnosis of Kaposi sarcoma; CR: complete response; PR: partial response; NC: no change; PRO: progressive disease.

  • a

    Paclitaxel + vinorelbine in 48% of patients, liposomal anthracycline in 28% of patients, and adriamycin, bleomycin, vindesine (ABV) in 20% of patients.

  • b

    P = 0.05 for the difference between the two groups.

  • c

    Patients switched to different highly active antiretroviral therapy regimens involving 2 nucleoside reverse transcriptase inhibitors + protease inhibitors (n = 28) or 2 nucleoside reverse transcriptase inhibitors + nonnucleoside reverse transcriptase inhibitors (n = 8).

  • d

    Patients received first-line therapy involving 2 nucleoside reverse transcriptase inhibitors + protease inhibitors (n = 148) or 2 nucleoside reverse transcriptase inhibitors + nonnucleoside reverse transcriptase inhibitors (n = 12).

  • e

    Twenty-six patients were evaluable in the KS-HAART group, and 100 patients were evaluable in the KS-naive group.

Chemotherapya  
 Yes15 (29) 54 (34)
 No36 (71)106 (66)
 Response  
  CR + PRb 7 (47) 37 (70)
  NC + PROb 8 (53) 16 (30)
HAART  
 Yes36 (71)c160 (100)d
 No15 (29)  0 (0)
 Responsee  
  CR + PR17 (65) 64 (64)
  NC + PRO 9 (35) 36 (36)

After a median follow-up period of 24 months for KS-HAART patients and 22 months for KS-naive patients, the death rate was similar in both groups (27% and 21%, respectively), with KS progression being the major cause of death in both groups (72% and 70%, respectively). Opportunistic diseases occurred with similar frequency in both groups during follow-up (20% in the KS-HAART group and 16% in the KS-naive group) and were the cause of death for 28% of KS-HAART patients and 30% of KS-naive patients. Median survival was not reached in either group. The 3-year survival rates of patients in the KS-HAART group (64%) and patients in the KS-naive group (78%) were not significantly different (Fig. 1).

thumbnail image

Figure 1. Survival of patients with acquired immunodeficiency syndrome–related Kaposi sarcoma (KS) who had (solid line; n = 51) or had not (dashed line; n = 160) been receiving highly active antiretroviral therapy at the time of KS diagnosis. Log-rank test: χ2 = 0.74; P = 0.39.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

To our knowledge, the current study is the first to evaluate the impact of HAART on the clinical presentation and natural history of AIDS-related KS. The most relevant data emerging from this large series indicate that patients already receiving HAART at the time of KS diagnosis present with a more favorable immunologic and virologic status compared with patients with KS who are not receiving HAART at the time of diagnosis; this improved status appears to translate to a less aggressive clinical presentation. Specifically, patients already receiving HAART at the time of KS diagnosis have a more indolent disease presentation, with fewer than 10 cutaneous lesions in 60% of patients and only 1 site of cutaneous involvement in 21% of patients, compared with fewer than 10 cutaneous lesions in 34% of patients (P = 0.01) and only 1 site of involvement in 12% of patients (P = 0.06) in the KS-naive group. Furthermore, a smaller proportion of patients already receiving HAART at the time of KS diagnosis presented with visceral disease (24% vs. 39%; P = 0.06); in particular, patients in the KS-HAART group had significantly less frequent GI tract involvement (14%) compared with patients in the KS-naive group (28%; P = 0.05).

We decided to treat dual NRTI regimens as HAART for the purposes of the current analysis. Although this choice could be questioned, we believe that it is reasonable, because the decision to consider a dual regimen appropriate as an effective mode of antiretroviral therapy was made by each physician based on the virologic and immunologic status of the patient in a time when other options (i.e., IP and NNRTI) were available. In fact, we included only patients who followed stable, dual NRTI regimens, and we excluded patients who either received intermittent therapy or shifted from triple therapy to dual therapy due to toxic effects and thus were not receiving optimal treatment. It is unfortunate that in these patients, the specific impact of IPs on the development of KS could not be evaluated.

We believe that the findings of the current study are noteworthy, in that they provide novel insight into the relation between KS and HAART. The interaction between KS and HAART has been widely investigated since the introduction of HAART in 1995, and many aspects of the interaction are becoming progressively clearer. To date, the majority of studies on HAART have focused on its epidemiologic and therapeutic impact. In particular, there are abundant data indicating that the incidence of KS has declined sharply since the introduction of HAART,2, 3 that the use of HAART prolongs the overall survival of patients with KS and is associated with an 80% reduced risk of death,14 that KS responds to HAART in more than 50% of patients,4–6 and that HAART is associated with prolonged time to treatment failure in patients with KS who receive systemic treatment.15

The current study focused on certain unknown aspects of HIV-related KS in the HAART era. We observed that KS also occurs in patients receiving stable HAART (24% of the current series) and that these patients exhibit good virologic control and a favorable immunologic response to HAART: at KS diagnosis, 63% of these patients had less than 30,000 copies/mL of HIV RNA and 37% had a CD4 count greater than 200 cells/μL. In addition, we found that although outcome is not influenced by whether HAART is initiated before diagnosis of KS, KS appears to be characterized by a more indolent course with less aggressive clinical presentation characteristics in patients who begin receiving HAART before diagnosis. It is likely that these findings are associated with the increased CD4 cell count at diagnosis of KS. It is difficult to determine whether antiretroviral drugs (e.g., IPs) had a direct impact on the clinical presentation of KS.

It is noteworthy that the majority (70%) of patients already receiving HAART at the time of KS diagnosis switched to a new antiretroviral combination after diagnosis, with good control of KS observed in 65% of these patients. This response rate was comparable to the rate (64%) among patients who began receiving HAART for the first time at diagnosis of KS. It is unfortunate that we do not have data on the immunologic and virologic responses to HAART and that we therefore are unable to compare the responses of HIV and KS to HAART. Chemotherapy was required with similar frequency in the two patient groups (29% in the KS-HAART group and 34% in the KS-naive group), but the response rate was more favorable in the KS-naive group (70%) compared with the KS-HAART group (47%; P = 0.05). We are unable to provide an explanation for this finding; however, the number of patients evaluable for response to chemotherapy is too small to allow us to draw any conclusions.

In conclusion, the data from the current study indicate that KS also occurs with significant frequency in patients already receiving HAART and that in these patients, KS exhibits a more indolent presentation compared with KS in patients who are naive to HAART at diagnosis. Response to HAART, natural history, and outcome do not appear to be influenced by the initiation of HAART before the development of KS.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES