British HIV Association guidelines for HIV‐associated malignancies 2008

Abstract

Table of Contents 1.0 Introduction 1.1 Reference 2.0 Kaposi's sarcoma 2.1 Diagnosis, staging and prognosis 2.2 Management 2.2.1 Prevention 2.2.2 Treatment 2.2.2.1 Local therapy 2.2.2.2 Radiotherapy 2.2.2.3 Other local therapies 2.2.2.4 Systemic therapy 2.2.2.5 Other systemic therapies 2.3 Summary of recommendations 2.4 References 3.0 Systemic AIDS-related non-Hodgkin's lymphoma 3.1 Introduction 3.2 Diagnosis, staging and prognosis 3.3 Management 3.3.1 Diffuse large B-cell lymphoma 3.3.1.1 First-line chemotherapy for diffuse large B-cell lymphoma in HIV-infected individuals 3.3.1.2 The effect of adding HAART 3.3.1.3 Infusional chemotherapy for diffuse large B-cell lymphoma 3.3.1.4 Rituximab for diffuse large B-cell lymphoma 3.3.1.5 Second-line therapy for systemic AIDS-related non-Hodgkin's lymphoma 3.3.1.6 Recommendations for diffuse large B-cell lymphoma 3.3.2 Burkitt's lymphoma 3.3.2.1 Recommendations for Burkitt's lymphoma 3.3.3 Leptomeningeal lymphoma management: prophylaxis and treatment 3.3.3.1 Recommendations for meningeal lymphoma management 3.3.4 Response evaluation and follow-up 3.4 References 4.0 Primary central nervous system lymphoma 4.1 Introduction 4.2 Diagnosis, staging and prognosis 4.3 Management 4.4 Summary of recommendations 4.5 References 5.0 Cervical intraepithelial neoplasia and cervical cancer 5.1 Introduction 5.2 Key recommendations of BHIVA, BASHH and FFPRHC 2007 guidelines on cervical screening in HIV 5.3 Management of cervical intraepithelial neoplasia 5.3.1 The effect of HAART on progression of cervical intraepithelial neoplasia 5.4 Recommendations for management of abnormal smears in HIV-positive women 5.5 Diagnosis, staging and prognosis of invasive cervical cancer 5.6 Management of invasive cervical cancer 5.6.1 Stage IA 5.6.2 Stage IB/IIA 5.6.3 Stage IIB/III/IVA 5.6.4 IVB/recurrent cancer 5.6.5 Invasive cancer in pregnancy 5.7 Summary of recommendations for management of women with cervical cancer 5.8 References 6.0 Anal cancer 6.1 Introduction 6.2 Key recommendations of BHIVA, BASHH and FFPRHC 2007 guidelines on anal cancer in HIV 6.3 Key recommendations of NICE 2004 guidelines on anal cancer 6.3.1 Primary treatment 6.3.2 Management of relapse 6.4 Diagnosis, staging and prognosis of HIV-associated anal cancer 6.5 Management of anal cancer 6.6 Screening for anal intraepithelial neoplasia 6.7 Summary of guidance 6.8 References 7.0 Hodgkin's lymphoma 7.1 Introduction 7.2 Diagnosis, staging and prognostic factors 7.3 Re-assessment and follow-up 7.4 Management 7.4.1 First-line treatment 7.4.2 Treatment at relapse 7.5 Recommendations 7.6 References 8.0 Multicentric Castleman's disease 8.1 Introduction 8.2 Diagnosis 8.3 Staging 8.4 Prognosis 8.5 Management 8.6 Surgery 8.7 Chemotherapy 8.8 HAART 8.9 Immunotherapy (excluding rituximab) 8.10 Anti-Kaposi's sarcoma herpesvirus therapy 8.11 Rituximab 8.12 Key recommendations 8.13 References 9.0 Other non-AIDS-defining malignancies 9.1 Introduction 9.2 Testicular germ cell cancers 9.2.1 Introduction 9.2.2 Diagnosis, staging and prognostic factors 9.2.3 Management 9.2.3.1 Stage I disease 9.2.3.2 Metastatic disease 9.2.4 Summary 9.3 Non-small cell lung cancer 9.3.1 Introduction 9.3.2 Management 9.3.2.1 Operable disease 9.3.2.2 Locally advanced disease 9.3.2.3 Metastatic disease 9.3.3 Summary 9.4 Hepatocellular cancer 9.4.1 Introduction 9.4.2 Presentation and diagnosis 9.4.3 Management 9.4.4 Screening for hepatocellular cancer in patients with hepatitis and HIV coinfection 9.4.5 Summary 9.5 Other cancers 9.5.1 Summary 9.6 References 10.0 Highly active antiretroviral therapy (HAART) and prophylaxis of opportunistic infections 10.1 Introduction 10.2 HAART 10.2.1 Nucleoside reverse transcriptase inhibitors 10.2.2 Nonnucleoside reverse transcriptase inhibitors 10.2.3 Protease inhibitors 10.3 Opportunistic infection prophylaxis 10.3.1 Bacterial infections 10.3.2 Antifungal prophylaxis 10.3.3 Pneumocystis jirovecii pneumonia prophylaxis 10.3.4 Mycobacterium avium complex prophylaxis 10.3.5 Hepatitis B 10.4 Recommendations 10.5 References HIV infection is associated with three AIDS-defining malignancies (Kaposi's sarcoma, high-grade B-cell non-Hodgkin's lymphoma and invasive cervical cancer) as well as an increased risk of a number of other malignancies. The clinical care of patients with these tumours requires a multidisciplinary approach drawing on the skills and experience of all healthcare professional groups. Moreover, optimal care can only be achieved by the close co-operation of oncologists, haematologists and HIV physicians, and unless all these clinicians are intimately involved in the care of patients it is likely that the outcome will be less favourable. Patients with HIV-associated malignancies should therefore only be managed in a centre dealing with large numbers of patients with these tumours. An audit study in North London confirmed the better management of patients with AIDS-related lymphoma in HIV centres with cohorts of >500 patients (level of evidence IV C) [1]. We recommend that all patients with HIV and malignancy should be referred to centres that have developed expertise in the management of these diseases and serve an HIV cohort of >500. The multidisciplinary medical team managing these patients must include HIV physicians, oncologists, haematologists and palliative care physicians. In line with national cancer waiting times, all patients with suspected cancers must be referred urgently and seen within 2 weeks of referral. Moreover, the NHS Cancer Plan sets out the goal that no patient should wait longer than 1 month from an urgent referral with suspected cancer to the start of treatment. The early chapters of these guidelines consider the three AIDS-defining malignancies, Kaposi's sarcoma, high-grade non-Hodgkin's lymphoma (including primary cerebral lymphoma) and cervical cancer. These chapters are followed by chapters on the non-AIDS-defining malignancies including anal cancer, Hodgkin's lymphoma, multicentric Castleman's disease and other non-AIDS-defining malignancies, whilst the final chapter discusses the role of antiretroviral therapy and opportunistic infection prophylaxis in the management of malignancy in people with HIV infection. These guidelines have used the British HIV Association (BHIVA) standard grading for levels of evidence (see Table 1). Brook MG, Jones K, Bower M, Miller RF. Management of HIV-related lymphoma in HIV treatment centres in North Thames Region. Int J STD AIDS 2004; 15: 765–766. Kaposi's sarcoma (KS) is the most common tumour in people with HIV infection and is an AIDS-defining illness. The cutaneous lesions are characteristic and often diagnosed clinically. The diagnosis can be confirmed histologically and graded into patch, plaque or nodular grade disease. Visceral disease is uncommon, affecting about 10% at diagnosis, and computed tomography (CT) scans, bronchoscopy and endoscopy are not warranted in the absence of symptoms. The AIDS Clinical Trial Group (ACTG) staging system for AIDS-related KS was developed in the pre-highly active antiretroviral therapy (HAART) era to predict survival and unlike most cancer staging schemes includes tumour-related criteria (T), host immunological status (I) and the presence of systemic illness (S) (see Table 2) [1,2]. The ACTG also established uniform criteria for response evaluation in AIDS KS (see Table 3) [1]. In the era of HAART the prognostic value of this staging system has been questioned and one study suggested that only the T and S stages identified patients with a poor survival prognosis [3]. However, a comprehensive evaluation of prognostic factors in 326 patients diagnosed with AIDS-KS in the era of HAART, externally validated on 446 patients from the US HIV/AIDS Cancer Match Study, has established a prognostic score [4]. Having KS as the first AIDS-defining illness (−3 points) and increasing CD4 cell count (−1 for each complete 100 cells/μL in counts at KS diagnosis) improved prognosis, whereas age at KS >50 years (+2) and S1 stage (+3) conveyed a poorer prognosis. On the basis of this index it was suggested that patients with a poor risk prognostic index (score >12) should be initially treated with HAART and systemic chemotherapy together whilst those with a good risk prognostic index (score <5) should be treated initially with HAART alone, even if they have T1 disease. The introduction of HAART was associated with a substantial reduction in the incidence of KS in many large cohorts [5–9]. Some of this decline in incidence appears to have preceded the introduction of HAART [10]. However, cohort studies have demonstrated that HAART protects against the development of KS and that nonnucleoside reverse transcriptase inhibitor (NNRTI)-based regimens are as effective as protease inhibitor (PI)-based regimens in terms of their protection [6]. In contrast, the incidence of KS continues to rise in Africa [11–14]. Specific therapies against human herpesvirus-8, the cause of KS, may also be helpful although these are unlikely to be effective against established lesions which contain mainly latent rather than lytic virus. A UK cohort study of 3688 people with HIV showed that the risk of KS was reduced by ganciclovir and foscarnet exposure but not acyclovir [15]. However, data from a cohort of 935 homosexual men with AIDS found that neither acyclovir, nor ganciclovir, nor foscarnet significantly reduced the risk of KS [16]. 2.2.2.1 Local therapy. Local treatments are most useful for managing localized bulky KS lesions or for cosmesis. However, local therapies are limited by their inability to affect the development of new lesions in untreated areas. 2.2.2.2 Radiotherapy. During the pre-HAART era radiotherapy had an important and established role in the management of low-volume cutaneous KS, including the cosmetic control of skin lesions and the treatment of painful lesions on the soles, on the genitalia, in the oral cavity and on the conjunctiva [17]. An early randomized study of radiation fractionation for cutaneous KS showed that both response rate and duration of local control were better with fractionated regimens (40 Gy in 20 fractions and 20 Gy in 10 fractions) compared with an 8 Gy single fraction, although toxicity and patient convenience were worse [18]. A second nonrandomized study of 57 patients found no significant difference in response rates between 16 Gy in four fractions and 8 Gy in a single fraction [19]. A retrospective study of 80 patients including some with endemic KS treated with a radiotherapy dose of 8 Gy reported an objective response rate of 74% [20]. In another study of 36 patients with KS of the feet, with a schedule of 3 fractions/week at 3.5 Gy/fraction, up to a total dose of 21 Gy, the response rate was 91% with a complete response rate of 80% [21]. However, the side effects of radiotherapy in people with AIDS are often severe [17,22]. This is particularly notable in the oral cavity and on the soles of the feet. Modified fractionated schedules and close attention to skin care including avoidance of friction and sparing use of moisturisers are required to keep toxicity as low as possible. The explanation for this increased toxicity is not clear. Although the use of radiotherapy in the management of KS has declined since the introduction of HAART, it still maintains an important role in the management of KS at specific sites. 90Strontium brachytherapy is an effective and well-tolerated treatment for eyelid and conjunctival lesions [23]. An important large randomized study from Zimbabwe has evaluated treatments for AIDS-KS in 495 patients who were not treated with antiretroviral (ARV) agents. This showed that radiotherapy did not improve either overall survival or quality of life compared with supportive care alone [24]. Although discomfort from radiotherapy is frequent, it usually resolves without intervention within 2 weeks of completion of therapy. 2.2.2.3 Other local therapies. Alitretinoin gel (0.1%) (9-cis-retinoic acid) is a topical, self-administered therapy approved for the treatment of KS in the USA but not licensed in Europe. Two double-blind, randomized placebo-controlled trials, involving a total of 402 individuals, evaluated 12 weeks of twice-daily alitretinoin gel [25,26]. The response rates in the active arm after 12 weeks were 37% [26] and 35% [25] compared with 7 and 18% in the placebo arms analysed by intention to treat. In both studies over 80% of participants were receiving HAART and this did not influence the results. The gel may cause dermal irritation and some undesirable skin lightening at the application site. Responses are typically noted in patients with a wide variety of CD4 cell counts and typically occur 4–8 weeks after treatment. Local problems such as gastrointestinal bleeding, perforation, volvulus and intussusception may be treated surgically, but surgery including amputation is no longer indicated in the routine management of this disease. Intralesional vinblastine is the most widely used intralesional agent and responses of around 70% were reported in the pre-HAART era [27,28]. Treated lesions usually fade and regress although typically do not resolve completely. A randomized study in 16 patients comparing intralesional vinblastine or sodium tetradecyl sulphate in the treatment of oral KS demonstrated partial responses in both groups with no significant differences [29]. Intralesional injections of biological agents such as interferon (IFN)-α have also shown activity, but are infrequently used now. In one early study of 20 patients, complete responses were observed in 80% of lesions treated with cryotherapy, and the duration of the response was more than 6 weeks. Greater than 50% cosmetic improvement of KS was reported in this pre-HAART era study [30]. An alternative experimental approach is photodynamic therapy, which is based upon activation by light of a photosensitizing drug that preferentially accumulates in tumour tissues such as KS [31]. A series of 25 patients with a total of 348 KS lesions received photofrin 48 h prior to light activation. No patients were on HAART and 95% of the lesions responded to therapy (33 and 63% complete and partial responses, respectively) [32]. To a large extent local therapies for KS have been superseded by the introduction of HAART. 2.2.2.4.1 HAART. There are no randomized trials comparing HAART with no HAART as all patients with KS should receive HAART. Many case reports and small series have described the regression of KS with HAART (and individual ARVs), and HAART, alone and in combination with other treatment modalities (local therapy, systemic therapy, immunotherapy, biological therapy and radiotherapy), has been shown to prolong time to treatment failure in KS [33] and to prolong survival in patients who have been treated for KS with chemotherapy [34]. No difference has been demonstrated in time to progression between patients receiving a PI-based HAART regimen and those receiving an NNRTI-based HAART regimen despite the anti-angiogenic effects of PIs observed in the laboratory [33]. The effect of HAART on KS is highlighted by the Swiss cohort study: the relative risk of KS development between 1997 and 1998 (HAART era) compared with the time period between 1992 and 1994 (pre-HAART era) was 0.08 [95% confidence interval (CI) 0.03–0.22] [35]. A further European cohort study reported a significant reduction in the cumulative probability of KS on HAART [36]. 2.2.2.4.2 Cytotoxic chemotherapy. Administration of systemic cytotoxic chemotherapy is warranted in patients with more advanced or rapidly progressive disease. It has been suggested that patients with a poor risk prognostic index (score >12) should be initially treated with HAART and systemic chemotherapy together whilst those with a good risk prognostic index (score <5) should be treated initially with HAART alone, even if they have T1 disease [4]. However, the decision to initiate systemic chemotherapy is usually based on a number of parameters including the prognostic index, response to HAART alone, patient performance status and end organ function, including hepatic and bone marrow reserve. Typical indications for systemic chemotherapy include widespread skin involvement such as more than 20 lesions, extensive KS of the oral cavity, tumour-associated oedema or ulceration, symptomatic visceral involvement and immune reconstitution inflammatory syndrome-induced KS flare [37]. In the pre-HAART era, several chemotherapeutic agents (bleomycin, doxorubicin, vinblastine, vincristine and etoposide) were shown to have activity against KS in case series and small phase II trials using different combinations and doses of these drugs [38–42]. However, liposomal anthracyclines and taxanes have become established as the backbone of current standard systemic cytotoxic therapy against KS. 2.2.2.4.3 Liposomal anthracyclines. Liposome encapsulation of anthracyclines constitutes a considerable advance in the chemotherapy of KS. The advantages of liposomal formulation include increased tumour uptake and hence favourable pharmacokinetics. The trials of liposomal anthracyclines for HIV-associated KS were undertaken in the pre-HAART era but clinicians continue to regard them as the gold-standard first-line chemotherapy for KS. Both liposome-encapsulated daunorubicin (DaunoXome 40 mg/m2 every 2 weeks) and the pegylated liposomal doxorubicin, which is known variously as Caelyx, Doxil or PLD (20 mg/m2 every 3 weeks), have been shown to have good antitumour activity. The toxicity profile is better than for other anthracyclines, with no reported cardiotoxicity even at high cumulative dosages [43] and rarely significant alopecia; however, there remains considerable myelosuppression, and occasional emesis. In addition, infusion-related hypotension and hand/foot syndrome are novel side effects seen with these liposomal formulations. Three sizeable, randomized controlled studies have compared liposomal anthracyclines with conventional combination chemotherapy regimens and all were conducted before the introduction of HAART. A phase III randomized comparison of DaunoXome and ABV (doxorubicin, bleomycin and vincristine) demonstrated equivalent overall response rates (partial and complete responses), time to treatment failure and survival duration [44]. Two randomized phase III trials compared pegylated liposomal doxorubicin with conventional combination chemotherapy [ABV in one study and BV (bleomycin and vincristine) in the other], as first-line therapy for KS in patients not on HAART. Both found response rates were higher in the Doxil arms but responses were often not sustained [45,46] (see Table 4 for details). The three phase III studies may not be directly comparable. In one small randomized study in 79 patients, KS patients were randomized to PLD (20 mg/m2) or DaunoXome (DNX) (40 mg/m2) every 2 weeks for up to six cycles; nonsignificant differences favouring PLD were found, although the study was under-powered and there is insufficient evidence for a recommendation of which liposomal anthracycline to use [47]. Since the widespread introduction of HAART, the duration of responses to treatment for KS have increased [48] and no further randomized trials have compared liposomal anthracyclines with nonencapsulated chemotherapy regimens. The safety and tolerability of these drugs in combination with HAART have been evaluated. In one study of 54 patients, 82% had a response within 8 weeks and the PLD/HAART combination was well tolerated with no evidence of suppression of CD4 cell counts [49]. In a cohort study of 50 patients treated with concomitant HAART and liposomal anthracycline chemotherapy for KS, there was no decline in CD4 cell count or rise in HIV viral load [50]. These findings suggest that standard opportunistic infection prophylaxis guidelines may be followed when treating patients with liposomal anthracycline chemotherapy for KS. Based on the response rates, median response durations and the toxicity profile, liposomal anthracyclines are considered first-line chemotherapy for advanced KS. 2.2.2.4.4 Taxanes. The major mechanism of cytotoxicity of taxanes, including paclitaxel, which is approved for KS treatment, is attributed to binding to β-tubulin polymers, which stabilizes microtubules against depolymerization. Paclitaxel also promotes apoptosis and down-regulates Bcl-2 protein expression in KS cells in vitro and in KS-like lesions in mice [51,52]. In a number of phase II trials, paclitaxel was shown to have single-agent activity against AIDS-KS; furthermore, these studies included a number of patients who had previously received anthracyclines [53–57]. In one phase II study of paclitaxel (135 mg/m2 every 3 weeks) for KS, 28 patients were enrolled and a response rate of 71% was reported. As a whole, this included four patients (14%) who had received anthracyclines but no patients received HAART [54]. A second, larger study of 56 patients included 20 patients (36%) who received a PI at some stage during the study and 40 (70%) who had received prior therapy for KS, which included liposomal anthracyclines in 17 patients (30%). The overall objective response rate was 59% (amended ACTG criteria), and the median response duration was 10.4 months [55]. Subsequently two studies have addressed the role of paclitaxel as second-line chemotherapy. In one open-label multicentre trial, 107 individuals were enrolled who had received prior chemotherapy for AIDS-KS. The previous therapy regimens included ABV (adriamycin, bleomycin and vincristine) in 52 patients, liposomal daunorubicin in 49 patients, and liposomal doxorubicin in 40 patients. Moreover, only 77% were receiving concomitant HAART (all PI-based) and 33% started this treatment at the same time as the taxane chemotherapy. The paclitaxel protocol used was 100 mg/m2 fortnightly. The overall response rate was 56% with no significant difference in response rate when comparing patients on or not on HAART. Less surprising was the finding that patients on HAART had a significantly improved survival. The main side effect reported in these studies was neutropenia – this generally resolved prior to the next cycle [56]. In a second study of 17 patients with anthracycline-refractory AIDS-KS that had progressed during or within 6 months of completing liposomal anthracycline chemotherapy, all patients were receiving a stable HAART regimen to avoid confounding of results. The treatment schedule was again 100 mg/m2 fortnightly. The objective response rate to paclitaxel was 71% (95% CI 60–81); eight (of 17) partial responses and four (of 17) complete responses. There were no significant changes in CD4, CD8, CD16/56 (natural killer cells) and CD19 (B cells) lymphocyte subset cell counts during and for up to 1 year following chemotherapy. Similarly, plasma HIV-1 viral loads did not change significantly during or after treatment, suggesting that the combined use of paclitaxel and HAART reduces the risk of chemotherapy-related immunological decline and opportunistic infections [58]. In contrast, previous trials without concomitant HAART were worrying in this respect; Gill [55] reported 51 AIDS-defining opportunistic infections in the 56 patients treated with paclitaxel (10.5/100 patient-months on paclitaxel), only 36% of whom received HAART, and Welles et al. reported 27 opportunistic infections (8.4/100 person-months on paclitaxel) among their cohort of 28, none of whom received HAART [54]. Thus the concomitant use of HAART and paclitaxel appears to be safe and not detrimental to immune function despite initial concerns about pharmacological interactions [59]. These findings suggest that standard opportunistic infection prophylaxis guidelines may be followed when treating patients with taxane chemotherapy for KS. The higher prevalence rates of alopecia, myalgias and myelosuppression and the need for a 3-h infusion make paclitaxel a less attractive first-line option than PLD. Moreover, the need for corticosteroid administration (typically dexamethasone 10–20 mg intravenously 30 min prior to paclitaxel, or 10 mg orally 12 and 6 h prior) to prevent allergic reactions raises further concerns for some clinicians. The clinical experience with docetaxel in KS is much more limited, although two small studies suggest that this agent can produce meaningful responses when used weekly [60], or in anthracycline pretreated individuals [61]. 2.2.2.4.5 Immunotherapy. The biological response modifier IFN-α was approved for KS treatment before the availability of HAART and liposomal anthracyclines. The ACTG randomized 68 individuals to low- and intermediate-dose IFN-α (1 million and 10 million units daily, respectively) plus didanosine [62]. Response rates and durations were not statistically different although there were more toxicities in the higher dose group. In another randomized study, 108 patients were treated with IFN-α (1 million or 8 million units daily) with zidovudine [63]. The higher dose regimen was associated with a statistically higher response rate and longer time to progression. In a retrospective study of patients with classic KS comparing PLD with low-dose IFN-α, 12 patients received 20 mg/m3 of PLD monthly while six received 3 million units of IFN-α three times per week, and PLD was found to be superior in terms of responses and toxicity [64]. Response to IFN-α frequently requires continued treatment for 6 months or more, as the time to response is typically more than 4 months. It should not be considered for progressive or visceral disease. Toxicity at higher doses including fever, chills, neutropenia and depression is common, and poor responses are observed in the setting of low CD4 cell counts. While it can be considered in those with residual KS who have appropriately reconstituted their immune systems with HAART, it is seldom used. 2.2.2.5 Other systemic therapies. Thalidomide has significant anti-angiogenic activity. A total of 37 patients were enrolled in two phase II studies. Partial responses were recorded for 35 and 47% of evaluable patients, with toxicity including fatigue, neuropathy and depression [65,66]. The importance of the c-kit pathway has been evaluated in 10 patients with previously treated cutaneous KS who received oral imatinib; half achieved a partial response but diarrhoea necessitated dose reduction in 60% [67]. Other therapies are being developed including COL-3, a matrix metalloproteinase inhibitor which in a phase II trial of 75 patients demonstrated partial responses in 41% [68]. Similarly, interleukin (IL)-12 was administered to patients on HAART with KS and the response rate was 71% [69]. Unlike other therapies discussed above, neither COL-3 nor IL-12 has been approved for use in any disease. A number of anti-herpes virus agents have been studied in AIDS-related KS; none has demonstrated significant activity, although they have been shown to prevent KS in one cohort study [15]. A Cochrane Database systematic review which focused on five trials involving 915 individuals concluded that alitretinoin gel or radiotherapy is effective in treating cutaneous KS and PLD is effective treatment for advanced KS. Interestingly, this systematic review found that only radiotherapeutic options were applicable to resource-poor settings [70]. Krown S, 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. 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. Nasti G, Talamini R, Antinori A et al. AIDS-related Kaposi's sarcoma: evaluation of potential new prognostic factors and assessment of the AIDS Clinical Trial Group Staging System in the HAART Era – the Italian Cooperative Group on AIDS and Tumors and the Italian Cohort of Patients Naïve from Antiretrovirals. J Clin Oncol 2003; 21: 2876–2882. Stebbing J, Sanitt A, Nelson M, Powles T, Gazzard B, Bower M. A prognostic index for AIDS-associated Kaposi's sarcoma in the era of highly active antiretroviral therapy. Lancet 2006; 367: 1495–1502. International collaboration on HIV and cancer: highly active antiretroviral therapy and incidence of cancer in human immunodeficiency virus-infected adults. J Natl Cancer Inst 2000; 92: 1823–1830. Portsmouth S, Stebbing J, Gill J et al. A comparison of regimens based on non-nucleoside reverse transcriptase inhibitors or protease inhibitors in preventing Kaposi's sarcoma. AIDS 2003; 17: 17–22. Carrieri MP, Pradier C, Piselli P et al. Reduced incidence of Kaposi's sarcoma and of systemic non-hodgkin's lymphoma in HIV-infected individuals treated with highly active antiretroviral therapy. Int J Cancer 2003; 103: 142–144. Mocroft A, Kirk O, Clumeck N et al. The changing pattern of Kaposi's sarcoma in patients with HIV, 1994–2003: the EuroSIDA Study. Cancer 2004; 100: 2644–2654. Engels EA, Pfeiffer RM, Goedert JJ et al. Trends in cancer risk among people with AIDS in the United States 1980–2002. AIDS 2006; 20: 1645–1654. Selik RM, Byers RH, Dworkin MS. Trends in diseases reported on U.S. death certificates that mentioned HIV infection, 1987–1999. J Acquir Immune Defic Syndr 2002; 29: 378–387. Sitas F, Carrara H, Beral V et al. Antibodies against human herpesvirus 8 in black South Afric