HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:
Guest Access | Sign In via User Name/Password

This Article Abstract Full Text (PDF) Free Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Article Archive Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (26) Citing Articles via Google Scholar Google Scholar Articles by Aboulafia, D. M. Search for Related Content PubMed PubMed Citation Articles by Aboulafia, D. M.

The Epidemiologic, Pathologic, and Clinical Features of AIDS-Associated Pulmonary Kaposi’s Sarcoma^*

^* From the Virginia Mason Medical Center, Seattle, WA.

Correspondence to: David M. Aboulafia, MD, Section of Hematology/Oncology, Virginia Mason Medical Center, 1100 Ninth Ave, PO Box 900 (H14-HEM), Seattle, WA 98111

	Abstract

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
AIDS-related Kaposi’s sarcoma (KS) occurs principally in homosexual or bisexual men infected with the newly identified human herpes virus-8, also called KS-associated herpes virus. Unlike classical forms of the disease, AIDS-associated KS is a multicentric entity that frequently involves lymph nodes and the GI tract. KS may also occur in the lung, commonly in the setting of extensive mucocutaneous disease and very rarely as an isolated event. The exact incidence of intrathoracic KS in patients with AIDS is unknown. Before the advent of highly active antiretroviral therapy (HAART), pulmonary KS had been reported in approximately 10% of patients with AIDS, 25% of patients with cutaneous KS, and in roughly 50% of postmortem examinations of patients with AIDS, KS, and respiratory infections. In the HAART era, the incidence of KS has declined precipitously in North America and Europe but not in third world countries where HAART is largely unavailable. Pulmonary KS may cause radiographic infiltrates and respiratory symptoms that mimic a variety of other infectious and neoplastic processes. An aggressive diagnostic evaluation of patients who have this condition is essential because chemotherapy and radiation therapy may provide significant palliation, particularly if used in conjunction with HAART. This review briefly explores the changing epidemiology of KS. The pathology and pathogenesis of KS is also reviewed, along with the clinical and radiographic presentation, diagnosis, and management of pulmonary KS.

Key Words: AIDS • HIV • pulmonary Kaposi’s sarcoma

	Introduction

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
Before the AIDS epidemic and the success of solid organ transplantation, Kaposi’s sarcoma (KS) was a rare tumor among Western populations, occurring in only 0.02 to 0.06% per 100,000 people.¹ In a typical dermatology practice, it was unusual for a busy practitioner to see more than one such case every 5 years. However, in June and July of 1981, reports from California and New York described large numbers of homosexual men who were suddenly afflicted with pigmented skin lesions of KS, often in the setting of opportunistic infections such as oral candidiasis and Pneumocystis carinii pneumonia.^{2 3 4} Since then, approximately 15 to 20% of men infected with HIV have received a diagnosis of KS. Typically, these tumors involve skin and lymph nodes and, less frequently, visceral organs. Although uncommon in the thorax, KS has accounted for approximately one third of respiratory episodes in patients with AIDS and KS who have required evaluation for pulmonary complaints.⁵

Dermatologic manifestations of KS can be alarming, but it is visceral involvement that is most commonly life threatening. Pulmonary KS may be difficult to differentiate from other infectious or neoplastic conditions, yet the distinction is essential, for without treatment, patients with pulmonary KS have had a median survival of only a few months.^{6 7 8 9} With therapy, and in particular with the use of highly active antiretroviral therapies (HAART), these patients may achieve significant palliation from debilitating symptoms and an improvement in survival. This article briefly reviews the changing epidemiology of KS in the HAART era. The pathology and pathogenesis of KS will also be discussed, followed by a focus on the radiographic and clinical findings of pulmonary KS. Recent advances in the treatment of KS, including the potential to modulate the natural history of this tumor with HAART, will also be discussed.

	Epidemiology

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
The first cases of KS were described by Moritz Kaposi in 1872.¹⁰ Classical KS, as it is now called, was characterized as a slowly progressive disease involving the cutaneous surfaces of the lower extremity; particularly, the condition was found to be more common among elderly men from Eastern Europe (Jewish) or Mediterranean countries. KS is also common in equatorial Africa. As early as 1971, KS accounted for roughly 3 to 9% of all reported malignancies in Uganda.¹¹ The African (endemic) form of KS is also seen more commonly among male patients, but differs from classical KS in that it may affect a younger population and is more likely to spread to visceral organs and lymphatics.¹² Iatrogenic KS occurs approximately 6% of the time in the setting of immunosuppression after solid organ transplantation and, once again, male patients are affected more commonly than female patients.¹³ In male patients of European, Jewish, or African ancestry, the rate of KS after renal transplantation is 500-fold greater than in other populations who undergo this procedure.¹⁴ Until recently, among patients infected with HIV, the rate of KS was 100,000-fold greater than that of the general population.^{15 16} This risk appeared most strongly concentrated among men who acquired HIV infection through unprotected sex with men. Specifically, KS is 20 times more prevalent in men who have sex with men than among heterosexual HIV-infected hemophiliacs. The disparity between the sexes is reflected by a male to female ratio of > 20:1 among HIV transmission groups.¹⁷

Reasons for the higher incidence of KS among certain groups have long attracted the attention of epidemiologists, infectious disease experts, virologists, dermatologists, and members of the medical oncology community. Investigators have speculated on the roles that differing sexual practices, exposure to various viruses, hormonal milieu, and class II human leukocyte antigen-DR5 antigens might have in promoting KS growth.^{16 17} More recently, a newly identified herpes virus, termed human herpes virus-8, or KS-associated herpes virus (KSHV), has been noted in KS tissues.¹⁸ In contrast to other viruses previously linked to KS pathogenesis (including Epstein-Barr virus, cytomegalovirus, and human papilloma virus), KSHV has been consistently detected in all forms of KS.^{19 20 21} In addition, KSHV DNA is present in the lymphoid system, peripheral blood mononuclear cells, saliva, and semen of patients with KS.^{22 23 24} In HIV-infected individuals, the presence of antibodies to this virus is predictive of KS development.^{25 26}

In the United States, up to 40% of homosexual men who received an AIDS diagnosis in the early 1980s presented with KS at the time of their initial AIDS diagnosis. Ten years later, the percentage of people with HIV infection who had KS as their initial AIDS-defining illness had decreased to approximately 10 to 20%.¹⁷ Several hypotheses have been offered to explain this phenomenon, including expansion of the AIDS case definition to encompass conditions that may be diagnosed earlier than KS, a decrease in identification or reporting of relatively minor KS lesions, and a decline in exposure to environmental factors associated with KS.^{16 17} The interpretation of these and other KS epidemiologic studies was limited insofar as many patients with AIDS developed KS in the months and years following their initial AIDS diagnosis, yet the surveillance data sets from which many of the investigations were derived included only those cases of KS diagnosed at or near the time of an AIDS-defining condition. Also, as some clinicians became more confident in diagnosing small KS lesions that did not always have important clinical implications for their patients, they may have become more apathetic in filling out paperwork and reporting such cases to local and regional monitoring agencies. Consequently, results from these studies may not have fully reflected the spectrum of AIDS-related KS.

Among clinicians who treat HIV-infected patients, there has been a widely held presumption that the incidence of KS has been declining precipitously since the mid-1990s. This belief is supported by several large epidemiologic studies that indicate that the incidence of KS in the United States and Europe is diminishing, particularly with the introduction of HAART combinations typically consisting of two, three, or more drugs that inhibit HIV replication.^{27 28 29}

Data collected by the Centers for Disease Control and Prevention for > 30,000 HIV-infected persons indicate that the incidence of KS declined an estimated 10% per year between 1990 (observed incidence, 4.8/100 person-years [PY]) and 1997 (observed incidence, 1.5/100 PY).²⁹ The Centers for Disease Control and Prevention data analysis indicates that the use of antiretroviral therapy is associated with a reduced risk for the development of AIDS-KS, ranging from a 13% reduction with monotherapy or dual therapy to a 59% reduction with triple therapy (Table 1 ).

Washington State has also witnessed a dramatic decline in KS among men who have sex with men. In 1990, the reported rate of KS was 60/1,000 HIV-infected individuals; by the second half of 1997, this number had declined to approximately 20/1,000 HIV-infected persons.³¹ Preliminary information from 1998 and 1999 indicates that the number of reported cases continues to decline at a precipitous rate.³² What impact the decline in KS incidence may have in accruing a sufficient pool of such patients into novel clinical protocols has been left largely unexplored.³³

	Incidence

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
Even before the introduction of HAART, the prevalence of pulmonary KS in patients with AIDS was uncertain because radiographic changes due to the tumor were often attributed to pulmonary infection.^{8 34 35 36} Definitive antemortem diagnosis requires lung biopsy, but this must be counterbalanced by other factors such as the clinician’s differential diagnosis and the patient’s willingness to undergo an invasive procedure that may be associated with such iatrogenic complications as pain, hemorrhage, or pneumothorax. When postmortem examinations have been conducted on HIV-infected patients already known to have cutaneous disease, the prevalence of pulmonary involvement has ranged from 47 to 75%.^{37 38 39 40} In other clinical studies, pulmonary KS has been diagnosed by fiberoptic bronchoscopy in 8 to 14% of patients with AIDS and respiratory symptoms, in 6 to 32% of patients with AIDS who had cutaneous KS, and in 21 to 49% of HIV-infected persons with known mucocutaneous KS and respiratory symptoms.^{6 7 40 41 42 43}

Pulmonary KS in the absence of mucocutaneous involvement is generally considered a rare event.⁴⁴ The incidence of such cases is unclear but most assuredly has been underestimated because clinicians rarely consider this diagnosis unless mucocutaneous disease is also present. The frequency of isolated pulmonary KS has ranged from 0 to 15.3%.^{6 36 39 40 45} Like the incidence of KS in general, the actual frequency of pulmonary KS in North America and Europe will continue to decline in the short term. Nevertheless, with a prevalence of 20,000 cases in the United States alone, KS is a problem that is not likely to quickly go away.⁴⁶ Factors that may ultimately contribute to the durability of KS control will likely be interrelated to the changing clinical expression and natural history of AIDS, and the long-term effectiveness of HAART in the backdrop of emerging trends in HIV viral resistance.⁴⁷ In the absence of an effective HIV vaccine strategy, impoverished countries will likely continue to experience large case loads of pulmonary KS.²¹

	Pathology

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
KS is an angioproliferative tumor. It is characterized by slit-like neovascular processes and the presence of proliferating endothelial cells, fibroblasts, infiltrating leukocytes, and a population of spindle-shaped tumor cells.⁴⁸ Although the clinical expression of disease may vary, KS histopathology does not differ among the various HIV risk groups.⁴⁹ In the very early stages, cutaneous KS is characterized by inflammatory cell infiltration, extravasation of RBCs, endothelial cell activation, and angiogenesis. This is followed by the appearance of the typical spindle-shaped cells that represent a heterogeneous population dominated by activated endothelial cells mixed with macrophages and dendritic cells (Fig 1 ).^{50 51} In advancing lesions, spindle cells tend to become the predominant cell type, although angiogenesis remains always a prominent feature (Fig 2 ).⁵²

View larger version (154K): [in this window] [in a new window] [Download PPT slide]   Figure 1.. Spindle tumor cells with slit-like small vascular spaces replace smooth muscle around this bronchiole.

View larger version (82K): [in this window] [in a new window] [Download PPT slide]   Figure 2.. KS following lymphangitics, around vessels (left, A) and in bronchi (right, B); in each location, it replaces the smooth muscle with spindle tumor cells and small vascular spaces, often filled with blood.

When viewed under the microscope, pulmonary KS lesions may be subtle, particularly when focal.⁵⁵ Attention must be focused on the areas of expected lymphatic routes. In the more solid regions, spindle cells are in loose fascicles, which may interdigitate. Slit-like spaces often do not have identifiable endothelial cells or lining tumor cells, but do possess extravasated RBCs. The smooth muscle of the bronchioles (Figs 1 and 2 , right, B) and pulmonary arteries (Figs 2 , left, A and 3 ) may be infiltrated by tumor that may mimic granulation tissue. Fairly extensive acute intra-alveolar hemorrhage may also be present, especially at time of death. Nodular forms of the disease possess abundant spindle cells and vascular clefts. Mitoses are not prominent, but tumor nuclei are elongated, moderately dark, and not greatly enlarged, although rarely possess very anaplastic features. Necrosis of the neoplastic cells in lung is rarely caused by KS, but rather the result of coexistent lung infection.⁴⁴

Grossly, the surface of the lungs may show flat to slightly raised disk-shaped red to violaceous plaques confined to the visceral pleura (Fig 4 ). The most striking parenchymal changes include lymphangitic thickening of tumor causing a red to red-blue discoloration about interlobular septa and bronchopulmonary rays (Fig 5 ).³⁷ Similar bright to dark red to violaceous lesions may be seen in the mucous membranes of the transbronchial tree by bronchoscopists. Occasional red to purple to gray nodules of various sizes may coalesce to form larger tumor densities. The nodes may be involved with spongy red to gray material replacing the usual translucent tan architecture. Pleural effusions typically are exudative and often contain reactive mesothelial cells without evidence of neoplasm. At autopsy, effusions are rarely an isolated event; in addition to KS (which infrequently involves the pleura), there is also a mixture of other disease events such as opportunistic infections and respiratory distress syndrome.^{55 56}

View larger version (70K): [in this window] [in a new window] [Download PPT slide]   Figure 4.. Pleural involvement. Left, A: grossly vascular lakes appear on pleura as dome-shaped smooth purple mounds. Right, B: microscopically, these often are telangiectatic dilatations of tumor in pleura.

View larger version (110K): [in this window] [in a new window] [Download PPT slide]   Figure 5.. Interlobular septal involvement. Top left, A: KS is seen following lymphatics, grossly darkening interlobular septa (arrows) and microscopically doing same (top right, B: low power; bottom left, C: medium power), and again note blood-filled spaces.

	Pathogenesis

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

In its early stages, KS behaves more like a hyperplastic proliferative disease than like a true cancer.^{61 62} Basic fibroblast growth factor and VEGF mediate spindle cell growth, angiogenesis, and edema of KS. The abnormal cytokine milieu of HIV infection and the mitogenic effects of HIV-tat protein may further act synergistically to stimulate KS spindle cell growth through autocrine and paracrine loops, leading to an increased frequency and aggressiveness of KS.^{61 63 64 65 66} However, the recognition that KS lesions are clonal and that KS-like cells can be detected in the peripheral blood of patients with AIDS-associated KS provides evidence for the malignant potential of KS spindle cell proliferation.^{67 68 69}

The role that KSHV plays in stimulating KS growth remains uncertain, but a number of clues are emerging. The genetic sequence of KSHV has largely been determined, and portions of its genome are analogous to DNA sequences believed to have oncogenic potential.⁷⁰ For example, the bcl-2 family of proteins is known for its ability to modulate apoptosis (ie, programmed cell death) and KSHV DNA codes for a functional bcl-2 homolog.⁷¹ Dysregulation of bcl-2 may contribute to neoplastic cell expansion via an anti-apoptotic effect that enhances cell survival rather than by accelerating rates of cellular proliferation. The KSHV-encoded G-protein coupled receptor (GPCR) may also act as a viral oncogene. In conjunction with VEGF, it appears capable of inducing angiogenesis in transformed mouse kidney cells containing the KSHV GPCR gene.⁷² In nude mice, the introduction of KSHV-GPCR transformed kidney cells results in tumor formation. KSHV may also code for proteins that mimic human cytokine and cytokine response pathways (including IL-6 and macrophage inhibitory protein-1) or stimulate supporting cells to produce angiogenesis factors.^{73 74}

	Natural History

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
Patients with pulmonary KS face an uncertain future. It has long been assumed they have a shorter life expectancy than other AIDS patients, yet important factors such as comorbid illness and immune profile have not consistently been taken into account.^{75 76 77} Gill and colleagues⁷ reported a median CD4+ cell count of 94 cells/µL in 20 patients with pulmonary KS. Gruden and associates⁷⁸ reported the median CD4+ count was 34 cells/µL; 84% of patients had counts < 100 cells/µL, and 96% had counts < 150 cells/µL. In the largest series describing the clinical presentation of pulmonary KS, investigators from San Francisco reported a median CD4+ count of 19 cells/µL among 168 patients.⁴⁵ These and other studies imply that pulmonary KS is confined primarily to patients with marked immune impairment, but in each of these analyses, patients had radiographically and bronchoscopically advanced disease and may have had early KS at a time when CD4+ cell counts were at a higher level. As immune impairment worsens, so too may the clinical expression of KS.⁷⁹

A few studies suggest that patients with KS as their first AIDS-defining illness may actually have a longer survival than patients without KS.^{80 81} Early in the AIDS epidemic, the median CD4+ T-lymphocyte count was higher at diagnosis of KS than at diagnosis of other AIDS-defining opportunistic infections. This factor may have accounted for the observed survival differences. However, most studies indicate that patients with KS have a diminished survival. For example, in a retrospective analysis of 241 HIV-infected homosexuals with KS compared to 241 HIV-infected control subjects, the KS group had a greater number of opportunistic infections (5.99 vs 3.88 infections) and an increased risk of death (odds ratio, 1:28), even when adjusting for age, previous opportunistic infection, baseline CD4+ cell count, and antiretroviral therapy.⁸²

In an effort to better define the natural history of AIDS-related KS, various staging systems have been proposed. In the United States, the schema originally developed by the AIDS Clinical Trials Group (ACTG) Committee on Malignancies has been the one most utilized by AIDS oncologists.⁸³ It stratifies patients into good- or poor-risk groups based on three main factors: (1) tumor extent (T); (2) immune profile status (I) measured by CD4+ T-lymphocyte count; and (3) evidence for HIV-associated systemic illness (S; the TIS classification).

Recently, this classification schema was used to evaluate survival among HIV-infected patients with KS who were recruited into various ACTG KS therapy studies between 1989 and 1995.⁸⁴ The analysis demonstrated that patients with KS confined to the skin and/or lymph nodes, and/or who possessed minimal (nonnodular) oral KS lived significantly longer than patients with visceral KS, bulky oral KS, or tumor-associated edema (27 months vs 15 months; p < 0.001). A change in the CD4+ count from 200 to 150 cells/µL was the only modification needed to best distinguish between good and poor-risk immune system categories (Table 2 ). Patients with a CD4+ cell count > 150 cells/µL had a median survival of 39 months, whereas those with < 150 cells/µL survived a median of only 12 months (p < 0.001).

Appropriately, Krown and colleagues⁸⁴ were cautious not to overinterpret the relatively good prognosis they noted among patients receiving treatment for pulmonary KS. They point to the small number of patients with documented pulmonary disease in their analysis and recognize that this limits the power to detect survival differences if they exist. Furthermore, the data may be skewed either because of selection bias or underascertainment of pulmonary KS. Patients with the most severe pulmonary KS may have been excluded from participation in trials for a number of reasons: low performance status, prior treatment, inadequate pulmonary function making it difficult to enroll into clinical trials, or a perceived need to treat immediately with noninvestigational agents before trial entry evaluations could be completed.⁸⁴ Some patients with early or subtle forms of pulmonary KS may have gone undetected even after undergoing radiographic and scintigraphic scans and bronchoscopy.

	Clinical Features

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
As with most pulmonary processes in patients with AIDS, the clinical presentation of intrathoracic KS is nonspecific and may be indistinguishable from pneumonia (Table 3 ). Dyspnea and cough are the most common presenting symptoms.^{6 7 8 9 40 41 42 43 44 45 85 86 87 88 89} Fever and night sweats may be present and usually suggest a concomitant infection. Hemoptysis may also occur, although its rate varies from series to series. Among 30 AIDS patients with symptomatic pulmonary KS, 100% had dyspnea, 80% had cough, 47% had chest pain, and 30% had hemoptysis; fever > 38.5°C was seen initially in 4 of 30 patients (13%).⁹ In medically underserved or economically disadvantaged areas such as rural Africa, hemoptysis may be a more frequent event, perhaps reflecting the severity of disease and its late presentation, and fundamental differences in host/virus interactions.^{21 90 91} When hemoptysis does occur, its origin is not always obvious. Meduri and colleagues⁸ were able to identify the site of bleeding via bronchoscopy in only a minority of patients with pulmonary KS.

Some investigators believe that wheezing, hemoptysis, pleuritic chest pain, and stridor are more likely due to KS than to opportunistic infections, and that pleural effusions (which in the setting of KS typically develop over 1 to 2 weeks) are more likely due to an infectious etiology.^{8 43} Others contend that hemoptysis presages a worse prognosis.^{91 92} In contrast, Huang and colleagues⁴⁵ found that fever 38.5°C and elevated serum lactate dehydrogenase concentration were the only clinical or laboratory parameters that distinguished patients with isolated pulmonary KS from those with superimposed opportunistic infection.

Three recurring themes emerge when reviewing the intrathoracic manifestations of HIV-related KS: (1) respiratory involvement nearly always follows rather than precedes the development of mucocutaneous lesions; (2) the frequency of pulmonary involvement that can be documented at autopsy (50%) is higher than that detected clinically (33%); and (3) approximately two thirds of patients with known KS who present with new pulmonary findings actually have a coexisting, usually treatable, opportunistic infection.^{78 93} These observations underscore the importance of evaluating each patient with KS and pulmonary complaints, not only for intrathoracic spread of the malignancy, but for respiratory tract infections as well.

Likewise, when evaluating a patient with AIDS and respiratory symptoms, it is worth reemphasizing that although most patients with pulmonary KS will have cutaneous, mucocutaneous, or endobronchial involvement, this is not invariably true.^{9 91 94} Among the first descriptions of AIDS-associated pulmonary KS was the case of a male homosexual with fever, weight loss, diarrhea, no skin lesions, and a chest roentgenogram that revealed bilateral nodular interstitial infiltrates.⁹⁵ Isolated pulmonary KS may present in a variety of fashions ranging from a slow-growing and asymptomatic peripheral nodule without accompanying adenopathy or pleural effusions⁹⁶ to a rapidly progressive interstitial infiltrate culminating in acute respiratory failure.⁹⁷ It may also be the sole cause of persistent fever in a patient with AIDS.⁹⁸

	Radiologic Findings

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
The radiologic findings of pulmonary KS were recently summarized and are outlined in Table 4 .^{99 100} Usually, plain chest roentgenograms reveal thickening along bronchovascular bundles, often emanating from a perihilar origin. As the tumor grows, a reticulonodular infiltrate appears, mainly in the lower lobes (Fig 6 , top left, A and bottom, left, C ).^{78 101} With continued growth, nodules become irregular and confluent, and this, along with interstitial infiltrates, leads to dense airspace consolidation.⁴⁰ Pleural effusions are present in as many as 50% of cases, and hilar or mediastinal lymphadenopathy is evident in 10 to 16%.^{78 99}

View larger version (148K): [in this window] [in a new window] [Download PPT slide]   Figure 6.. Early and progressive pulmonary KS in AIDS. Top left, A: chest radiograph reveals subtle bronchial wall thickening of the infrahilar regions. Top right, B: CT image through the infrahilar regions demonstrates peribronchovascular thickening characteristic of KS (arrow), and irregular thickening of the interface between the vessel and lung (curved arrows). Bottom left, C: chest radiograph demonstrates advanced pulmonary KS with more nodular thickening of the bronchovascular perihilar mid and lower lung. Confluence of nodules in the right lower lung leads to airspace consolidation (arrow). Bottom right, D: CT image through the mid lungs reveals peribronchial flame- or plume-shaped soft-tissue densities (arrows), extending into the lung from the hilar regions, and creating thickening of bronchial walls. Peripheral nodules have become confluent in some areas (open arrow). Interlobular septal thickening is seen in the lung periphery (curved arrows). Reproduced with permission from McGuinness.99

The striking radiographic characteristics of intrathoracic KS provide important clues to the radiologist. In a study of 102 HIV-infected patients with documented thoracic disease and 20 patients without thoracic complications, radiologists achieved a 90% accuracy score when asked to interpret the radiologic findings while not knowing the clinical diagnosis.¹⁰¹

CT scans of the chest will often identify bronchial wall thickening and spiculated lesions in patients with pulmonary KS (Fig 6 , top right, B and bottom right, D). In a series of 53 patients with pulmonary KS, Khalil and colleagues¹⁰² found these abnormalities occurring in 66% and 79% of cases, respectively. The distinct pattern of pulmonary KS following bronchovascular pathways, whereby poorly marginated nodular infiltrates radiate out from both pulmonary hilum along the bronchovascular structures into surrounding interlobular septa, is readily apparent on CT scan.¹⁰³ With this imaging modality, enlarged lymph nodes are more readily visualized than on chest roentgenogram and are apparent in 15 to 53% of cases.^{102 103 104} CT scans also provide additional clues regarding the presence of intrapulmonary chest disease. Eight of 15 patients (53%) with pulmonary KS also had sternal, rib, thoracic spine, and/or subcutaneous tissue involvement incidentally detected by CT scanning.¹⁰⁴

CT scans offer additional benefits over conventional chest roentgenograms, not only in terms of identifying KS, but by providing greater detail regarding other lung diseases.¹⁰⁵ However, this difference is usually modest, and in most patients, a chest roentgenogram provides adequate information. CT scan is not well suited for the detection of endobronchial tumors; only the very largest lesions are readily visualized by this modality.

The technique of spiral CT scanning is particularly useful for evaluating dyspneic patients because of its shorter imaging time and higher-resolution images.¹⁰⁰ Although not extensively evaluated in the setting of HIV infection, one study did compare this modality to gallium scanning in patients with AIDS and pulmonary symptoms and normal or equivocal chest radiographs. High-resolution CT scanning was most helpful in guiding the method of biopsy and directing the bronchoscopist to the diseased lung segment that would maximize diagnostic yield.¹⁰⁶

Gallium-thallium radionuclide imaging is also used in patients with AIDS to distinguish KS from other processes.¹⁰⁷ KS is thallium avid, but unlike other infectious or neoplastic AIDS-related complications, it does not take up gallium. The combination of focal KS pulmonary involvement and concurrent infection can occasionally lead to a negative thallium and a positive gallium pattern scan.¹⁰⁸ Like gallium, indium-111 labeled polyclonal Ig also localizes to infection but does not accumulate in KS or lymphoma.¹⁰⁹ Radiologists who have experience in interpreting these scans can sometimes provide useful information when evaluating a patient with an uncertain diagnosis. In practice, only a few medical centers in the United States use these studies to establish the diagnosis of pulmonary KS. This is due to the high cost of such tests, and the accuracy of CT scan and bronchoscopy in the identification of characteristic lesions of KS.

The MRI features of pulmonary KS have not been extensively detailed. In one study of patients with established pulmonary KS, characteristic findings of MRI included an increased signal intensity on T1-weighted images, markedly reduced signal intensity on T2-weighted images, and strong lesional contrast enhancement after administration of gadolinium.¹¹⁰ This pattern of signal abnormalities, particularly when seen in a peribronchial vascular distribution, was most suggestive of KS. Further analyses will better define how specific this finding is for pulmonary KS and what role MRI may have in evaluating pulmonary disease in patients with HIV infection.

	Diagnosis

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
Pulmonary KS can involve the tracheobronchial tree, the pulmonary parenchyma, and the pleura. It is the most common endobronchial lesion associated with HIV and has a characteristic red or purple macular or papular appearance often located at airway bifurcations.¹¹¹ It is very rare to identify KS lesions at bronchoscopic examination below the level of the carina and not detect parenchymal lesions on chest radiography.¹¹² In a retrospective analysis of 76 patients with bronchoscopically proven KS, a correlation was noted between the extent of endobronchial and radiographically documented parenchymal disease.⁹⁴ A strong correlation between endobronchial lesions and underlying parenchymal KS has also been observed in a small number of patients at autopsy.^{8 92 113} As a result, most authorities assert that the presence of characteristic tracheobronchial KS is sufficient to make a presumptive diagnosis of pulmonary KS.⁴⁵

Direct inspection of lesions at bronchoscopy may be the most sensitive technique available for establishing a diagnosis of pulmonary KS. Yet, even with this procedure, only 45 to 73% of cases have readily identifiable endobronchial lesions.^{41 42} Parenchymal lesions may occur in the absence of tracheobronchial lesions, and they too may be missed at bronchoscopy.¹¹¹ Bronchoscopic evaluation of the airways may fail to yield a diagnosis of pulmonary KS for several reasons: (1) the bronchoscope may not be advanced far enough to detect distal airway disease; (2) KS may not involve the bronchi or adjacent tissue; (3) the lesions may not extend into the submucosal space where they can be visualized; and (4) the interstitial involvement of KS may be microscopic.^{9 93 102}

Pulmonologists rarely are obliged to obtain biopsy specimens of endobronchial KS lesions with a typical appearance. However, in the presence of atypical lesions, or in rare cases in which there are no cutaneous manifestations of KS, bronchial biopsy may be indicated, although one series reported a 30% incidence of significant hemorrhage.⁸ This risk may be greatest when lesions located in the central airway are sampled.^{5 52}

Both endobronchial and transbronchial biopsy have a diagnostic yield of only 26 to 60% because of the patchy submucous nature of the lesions.^{6 44 56} Histologic identification is difficult because the lesions are composed of spindle cells and blood vessels, some of which may appear normal. Because of the paucity of malignant features, biopsy findings may be misinterpreted as reactive fibrous tissue.¹¹⁴

Alveolar hemorrhage (detected in BAL fluid) is sometimes associated with pulmonary KS. This is a nonspecific finding and is associated with a variety of infective and noninfective HIV-related pulmonary diseases.^{111 115 116} Pleural effusions may be frankly bloody or serosanguinous transudate or exudate. In a recent study of the radiographic features of pulmonary KS, pleural involvement occurred only in the presence of parenchymal abnormalities.¹¹² Thoracentesis rarely leads to a diagnosis of KS, but is important to perform to exclude the possibility of malignancy or pyogenic, mycobacterial, or fungal infection.^{100 113}

Open lung biopsy has a diagnostic yield of approximately 50%, but it is rarely performed due to pain and other potential complications associated with this procedure.^{6 7 44 113} A successful thoracoscopic biopsy obviates the need for diagnostic thoracotomy, although results in the setting of HIV-infection have not been extensively published.¹¹⁷

A promising approach to assist in the diagnosis of pulmonary KS involves polymerase chain–reaction-based localization studies for KSHV. In one investigation of nine patients with KS, normal skin and lung did not reveal KSHV infection, but diseased tissue showed KSHV-specific infection of endothelial cells and KS tumor cells, as well as the epithelioid pneumocytes.¹¹⁸ KSHV has also been detected in the BAL-recovered cells from 7 of 12 patients with endobronchial KS and in only 2 of 39 samples from HIV-infected patients without KS.¹¹⁹

	Treatment

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
Chemotherapy
Active cytotoxic agents for patients with advanced KS with the longest track records include vinca alkaloids, anthracyclines, bleomycin, and etoposide.¹²⁰ The highest response rates have been achieved when these agents are used in combination. Using a variety of chemotherapy regimens incorporating these agents, Garay and colleagues⁶ and Meduri and associates⁸ achieved median survivals ranging between 3.8 and 6.0 months in patients with pulmonary KS. A recurring theme in these and other early trials of pulmonary KS was that the tumor was a direct cause of or contributed to death in the majority of patients.

Due to the disappointing results achieved with single-agent chemotherapy for pulmonary KS, clinicians have tended to offer combination drug therapy, much like they would for other severe manifestations of cutaneous and visceral disease. The most widely used regimen for the treatment of pulmonary KS has been adriamycin, bleomycin, and vincristine or vindesine (ABV) combination chemotherapy. Although criteria used to document responses have varied from trial to trial, response rates of 30 to 50% are typical and usually occurred within a median of 2 months.¹²¹ With ABV and other comparable chemotherapy regimens, responding patients may achieve dramatic improvements in pulmonary symptoms even before radiographic changes are appreciated.

It is gratifying to see patients suddenly breathe more comfortably after they receive cytotoxic chemotherapy. Unfortunately, such benefits are usually short lived, and even with combination therapy, time until relapse is usually brief. In a representative study, Gill and colleagues⁷ treated 20 patients with ABV or bleomycin and vincristine or vindesine (BV). Twelve patients (60%) showed a favorable response to therapy. The median survival for responders was 12 months vs 6 months for the nonresponders (range, 1 to 17 months).

Treatment can lead to significant bone marrow suppression, alopecia, mucositis, GI symptoms, cardiac dysfunction, and peripheral nerve dysesthesias. In one study, 30 patients with respiratory symptoms and bronchoscopically confirmed KS received adriamycin (30 mg/m²), bleomycin (10 U/m²), and vincristine (2 mg) every 4 weeks.⁹ Sixty-four percent had radiographic responses and improvements in respiratory symptomatology. Yet, the median survival in this study was only 6.5 months (range, 1 to 23 months) after the first course of chemotherapy, 9 months after the diagnosis of pulmonary KS, and 11.4 months after the onset of respiratory symptoms. Roughly half of patients died from progressive KS and respiratory failure. Neutropenia and infection were common complications.

In an attempt to achieve better results for patients with pulmonary KS, investigators have used more myelosuppressive chemotherapies coupled with granulocyte colony stimulating factor support. Such strategies have proven dangerous to patients and have not improved outcomes. For example, Sloand and colleagues¹²² used a complex chemotherapy regimen consisting of ABV, actinomycin-D, and dacarbazine with concurrent antiretroviral therapy and granulocyte colony stimulating factor. Fifteen of 18 patients (83%) had improvements in their pulmonary infiltrates and resolution of dyspnea and cough. Responses were rapid, and skin activity mirrored pulmonary disease, but in this group with a median CD4+ cell count of 73/µL, the median survival was only 9 months and maintenance therapy with etoposide was not beneficial.

The use of targeted liposomes as carriers of cytotoxic drugs may reduce the toxic effects to normal tissue, while increasing drug concentrations and effectiveness at the tumor site.¹²³ Recently, the US Food and Drug Administration approved the use of liposomal daunorubicin and liposomal doxorubicin for treatment of AIDS-related KS.¹²⁴ These drugs are effective and are generally associated with less alopecia, mucositis, and cardiac toxicity than their nonliposomal anthracycline counterparts.^{125 126} Extravasation of these agents can cause tissue damage, although early experience suggests this is not as significant as the soft tissue injuries induced following administration of other forms of anthracyclines.¹²⁷ Contrary to initial assumptions,¹²⁸ these agents may also offer a cost-effective alternative to ABV or BV.¹²⁹

Among the first descriptions of liposomal daunorubicin for treatment of AIDS-related pulmonary KS was the case of a patient with severe pulmonary symptoms and advanced immunosuppression.¹³⁰ After receiving chemotherapy, he obtained a radiographic partial response and survived an additional 12 months before succumbing to further HIV-related complications. More recently, in a phase II study, 11 patients with a median CD4+ count of 52 cells/µL received liposomal daunorubicin (40 mg/m² every 2 weeks) and concurrent antiretroviral therapy.¹³¹ Toxicities included mild nausea, fatigue, and modest granulocytopenia. After 3 months of therapy, only one patient had obvious tumor progression. In another study, liposomal anthracyclines were associated with a median survival of 11 month, compared to 4 months for patients who received other forms of chemotherapy.¹³² In an open-label phase II clinical trial of liposomal daunorubicin (60 mg/m² every 2 weeks), the vast majority of patients had significant improvements in pulmonary symptoms and achieved a median survival of 7.1 months.⁸⁷ In this group of 53 patients with a median CD4+ cell count of only 13/µL, side effects were evident and included fever (54%), nausea (53%), and fatigue (51%), and, less commonly, alopecia (11%) and mucositis (6%). The most common laboratory abnormalities were neutropenia (85%), anemia (32%), and thrombocytopenia (17%).

Many oncologists experienced in HIV-care now consider liposomal anthracyclines as the chemotherapy drugs of choice for those patients who require cytotoxic options. However, combining liposomal anthracyclines with BV does not appear to offer additional benefits to patients with cutaneous KS.¹³³ By extension, it is not likely that such a strategy would benefit patients with pulmonary KS.

Paclitaxel has significant anti-KS activity. At doses ranging between 100 to 175 mg/m² every 2 to 3 weeks, response rates of 53 to 65% have been reported among patients with poor prognosis disease.^{134 135} Treatment is associated with alopecia, neutropenia, and peripheral neuropathy, but is generally well tolerated when given at the more conservative dose of 100 mg/m² every 14 to 21 days. Experience with paclitaxel for pulmonary KS is limited, but in one study, all five patients with pulmonary involvement responded.¹³⁴

For individuals with pulmonary KS, the goals of therapy need to be clearly delineated. Despite dramatic advancements in the care of HIV-infected patients, neither KS nor AIDS are yet considered curable conditions. Options need to account for the patient’s level of activity, comorbid illness, immune reserves, and the input of other members of the multidisciplinary team.¹²¹

Radiation Therapy
The role of radiation therapy in the treatment of KS has long been appreciated; the first significant analysis was published in 1962.¹³⁶ In 1985, the first report describing the use of external beam radiation therapy to treat a patient with HIV-associated pulmonary KS was published.¹³⁷ Since then, a scattering of case reports have described patients with advanced pulmonary disease often not responding to systemic chemotherapy who were treated with radiation therapy. Typically, patients achieved short-term palliation before succumbing to pulmonary embarrassment and superimposed opportunistic infections. In one analysis, among 11 patients with pulmonary KS who were treated with radiotherapy, 2 died while receiving therapy; the remaining 9 patients had significant relief of symptoms until death.¹³⁸ Similarly, among four patients with advanced disease who were treated at Walter Reed Hospital, pulmonary symptoms were transiently alleviated, although median survival was only 1.5 months.¹³⁹ Meyer⁸⁸ treated 25 patients with whole lung irradiation for symptomatic pulmonary KS. Treatment was given 4 days per week, 150 cGy per fraction. Eighteen of the patients who presented with dyspnea reported improvements, although there were no long-term survivors.⁸⁸

Whether treatment of pulmonary KS at an earlier stage of disease would offer longer-term palliation is not certain. Among 25 patients with pulmonary KS who failed to respond to chemotherapy, their median survival was only 3.2 weeks. For those who had not yet received chemotherapy, roughly two thirds survived > 3 months.⁸⁸

Thoracic irradiation has been poorly tolerated among HIV-infected patients with lung cancer.¹⁴⁰ Fortunately, KS is a very radiosensitive tumor. At the modest doses that patients with pulmonary KS typically receive, side effects are modest and radiation pneumonitis does not occur. However, for patients with oropharyngeal KS, care must be taken when radiating the oral mucosa. In this setting, significant mucositis, thrush, or reactivation of oral herpes simplex may supervene.¹⁴¹ Patients should receive prophylactic medications to prevent oropharyngeal candidiasis or reactivation of latent herpes simplex.

HAART
Recent studies indicate that patients who respond to HAART typically consisting of two nucleoside analogs and a protease inhibitor achieve decreased plasma levels of HIV, decreased incidence of opportunistic infections, increased circulating CD4+ T cells, and decreased short-term mortality.^{142 143 144 145} Reports further suggest that conditions previously deemed intractable, such as cytomegalovirus retinitis, progressive multifocal leukoencephalopathy, azole-resistant mucocutaneous oral candidiasis, and intestinal cryptosporidiosis and microsporidiosis may stabilize or even diminish after increases in CD4+ cell counts or significant reductions in HIV plasma viral RNA loads.^{146 147 148} What effect potent combination antiretroviral therapy will have in altering the clinical course of AIDS-related neoplasms is uncertain and necessitates long-term study.¹⁴⁹

Recently, I described a patient with symptomatic pulmonary KS, a CD4+ cell count < 50/µL, and an elevated HIV viral load.¹⁵⁰ He declined recommendations to receive chemotherapy, but reluctantly agreed to begin HAART. Over the next month, his pulmonary symptoms resolved; over the ensuing year, his chest roentgenogram gradually improved. His clinical and radiographic improvements corresponded to suppression of his HIV viral load to nondetectable levels and a progressive rise in his CD4+ cell count.

Support for the assertion that effective suppression of HIV replication has important clinical implications comes from a preliminary report involving 13 patients with AIDS-related KS.¹⁵¹ Before initiation of HAART, these patients received one or more systemic therapies for severe KS for a median of 8 months. After the initiation of effective antiretroviral therapy, their median HIV viral load was reduced from 43,000 copies/mL to nondetectable levels, and their median CD4+ cell count increased from 79 to 180/µL. None of the 13 patients experienced progression of KS despite discontinuing their systemic KS therapy for a median of 10 weeks (range, 0 to 41 months). Similar results were reported among eight patients in Italy with KSHV antibodies and documented KS. The initiation of HAART led to tumor shrinkage and a decline in KSHV viral loads.¹⁵² In London, effective suppression of HIV-viral RNA levels among patients with previously treated KS considerably prolonged time to treatment failure.¹⁵³ A report from Chanan-Khan and colleagues¹⁵⁴ described six patients with pulmonary KS who required ongoing chemotherapy. Shortly after beginning HAART, HIV viral levels fell to undetectable levels and the patients were successfully weaned off maintenance chemotherapy without progression of their pulmonary disease. With a median follow-up of 78 weeks off chemotherapy (range, 40 to 100 weeks), all six patients continued to do well. Such an approach suggests that HAART alone may be an effective maintenance therapy for patients with pulmonary KS.

The mechanisms by which HAART influences growth of KS lesions remains incompletely understood, although over the past 5 years it has become apparent that KSHV induces changes in endothelial cells, which result in them both producing and becoming more susceptible to the effects of inflammatory cytokines and angiogenesis-inducing cyto- kines.¹⁵⁵ Furthermore, HIV-1tat also can activate expression of tumor necrosis factor-, IL-6, and adhesion molecules such as E selectin, intercellular adhesion molecule-1 and vascular cell adhesion molecule-1, and can synergize with other inflammatory cytokines to stimulate endothelial cells and KS-derived spindle cells to promote further endothelial cell proliferation. The antiviral drugs that are typically utilized in HIV drug regimens appear to have little if any intrinsic inhibitory activity against KSHV. Rather, by down-regulating HIV expression while promoting some component of immune reconstitution, it appears that HAART may play an active role in regression of KS tumors.¹⁵⁶

Although preliminary clinical reports concerning the impact of HAART on modifying both the incidence and clinical expression of KS are encouraging, enthusiasm must be tempered by several mitigating factors: our uncertain knowledge of the length of time that HAART can effectively suppress viral replication; the inability of antiviral therapies to restore completely impaired immunity; and our imperfect knowledge of how best to maintain patient compliance in a setting in which medications must be taken multiple times a day, have unpleasant side effects, and interact unpredictably with numerous other medications.^{157 158 159 160}

	Conclusion

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion References

 
Pulmonary KS may cause significant morbidity and mortality in the setting of HIV infection. Fortunately, AIDS-defining illnesses that have previously been deemed progressive and poorly responsive to conventional therapies may be controlled with HAART. Pulmonary KS can now be added to this list. Additional studies will continue to investigate the effect of various combinations of anti-HIV medications on the development and progression of KS. New treatment strategies and staging schemas must consider not only the tumor bulk, history of opportunistic infections, and CD4+ cell count, but also the HIV plasma viral RNA level (and perhaps KSHV plasma levels) and antiretroviral options. Modulating KS growth by incorporating potent antiretroviral combinations into treatment programs is dramatically altering the natural history of this most common AIDS-related malignancy.

View larger version (156K): [in this window] [in a new window] [Download PPT slide]   Figure 3.. KS is seen here filling lymphatic spaces halfway around this pulmonary artery. Small vascular lakes are at right and represent the telangiectatic variant of tumor.

	Acknowledgements

	Footnotes

 
Abbreviations: ABV = adriamycin, bleomycin, vincristine or vindesine; ACTG = AIDS Clinical Trials Group; BV = bleomycin and vincristine or vindesine; GPCR = G-protein coupled receptor; HAART = highly active antiretroviral therapy; IL = interleukin; KS = Kaposi’s sarcoma; KSHV = Kaposi’s sarcoma-associated herpesvirus; PY = person-years; VEGF = vascular endothelial growth factor

Received for publication August 4, 1999. Accepted for publication October 21, 1999.

	References

TOP Abstract Introduction Epidemiology Incidence Pathology Pathogenesis Natural History Clinical Features Radiologic Findings Diagnosis Treatment Conclusion