Important note: Information in this article was accurate in August 2000. The state of the art may have changed since the publication date.| Introduction Epidemiology Kaposi's Sarcoma Developing Therapies HPV in the HIV Infected Anal Cancer |
"It was appalling to me that I could list on one slide more or less everything that we know about HHV-8 prevalence and transmission," Harold Jaffe told the 4th International AIDS Malignancy Conference, which met on the campus of the US National Institutes of Health in Bethesda, Maryland, May 16-18, 2000. The associate director for HIV/AIDS of the US Centers for Disease Control and Prevention (CDC, Atlanta) said, "We don't really know a whole lot." (Conference abstracts were published in JAIDS, Vol.23, No. 3, March 1, 2000.)
The grim news is that Jaffe was speaking about the epidemiology of the best-known AIDS malignancy. Even less could be said with certainty about the incidence and treatment of various other malignancies that afflict people living with HIV.
The purpose of the conference was to facilitate research into these topics. It offered new snippets of information but the outlines of the puzzles that must be solved remain so indistinct that it is difficult to tell if these are signposts to the destination of useful therapy or missteps leading to further confusion.
According to Valerie Bernal (Imperial Cancer Research Fund, London), two cancers in particular have stood out from the beginning of the HIV epidemic: Kaposi's sarcoma (KS) and non-Hodgkin's lymphoma (NHL). Bernal was part of a team presenting preliminary data from the International Collaboration on HIV and Cancer, which was established in 1998 "to try to bring together the whole world evidence" of published studies on that subject. It compared observed incidence of specific types of cancer in people with HIV with what would be expected in comparable sociodemographic populations.
Beyond the AIDS-defining KS and NHL, "There was almost no significant heterogeneity with the exceptions of Hodgkin's disease, liver cancer, and leukemia," said collaboration researcher Robert Newton (also from the Imperial Cancer Research Fund). Those associations came from Western data and often did not correlate with what was seen in Africa.
The higher incidence of liver cancer in the West likely is explained by the greater role that injection drug use plays in transmission in those regions, and a resulting higher coinfection with hepatitis, which can lead to hepatocellular carcinoma. The study did not control for this coinfection.
Small data sets are one element tempering the significance of this study. The set for cancer of the eye was the largest with 133 combined cases, while vulvar cancer was the smallest, with 11 examples from Africa and two from the United States.
Epidemiological data from Africa is spotty. Because access to care and the medical resources to treat a disease are limited, often there is limited ability or incentive to make a diagnosis of a condition, let alone report it to a central authority. Furthermore, HAART allows time for slower-developing malignancies to advance to a point of clinical significance, a luxury that few have in Africa. Thus the occurrence of particular malignancies differs by geographic region.
Bernal compared pre- and post-HAART rates of cancer in the West. "In every single study there was a decrease in Kaposi's sarcoma," generally of about 60 percent. Non-Hodgkin's lymphoma showed a decline of about 40 percent. Only anogenital cancer showed a significant increase.
Robert Biggar and a team at the US National Cancer Institute (NCI) went about the epidemiological task "quite differently" by making a linkage between the cancer registry and the AIDS registry within specific geographic locations. Phase I of the study covered 95,000 AIDS cases from 1988 to 1990, while Phase II expanded to 366,000 AIDS cases over the years 1994-1996.
The sample was chosen to be representative of the US AIDS caseload in general, but with a slight bias toward women to assure that there would be adequate numbers powering the analysis of this subset. Each subset was compared to the non-AIDS norm of that population subset. For example, injection drug users have higher incidence of hepatitis, while gay men smoke more heavily and hence show a higher incidence of lung cancer.
"By knowing when they got their cancer and when they got their AIDS, you could put them into some kind of logical sequence," said Biggar. Researchers were trying to see if it was HIV itself that influenced the development of particular types of cancer, or if it was immune suppression that had the effect.
"We put a lot of weight on the early post-AIDS period [two years after diagnosis], because they are still in follow-up," he said. Over time there is increasing loss to follow-up, either to death, which may not be recorded, or to relocation. "Our observations became more conservative the farther out we went."
Biggar said that the rate of AIDS-related lymphoma "is high to begin with and stays fairly high." But over time, the relative risk declines because the background rate one would expect to see in the population increases over time. He concluded that immunosuppression associated with AIDS does not appear to increase risk for cancer in general. The exceptions are KS, NHL, and perhaps Hodgkin's disease.
The NCI's Morten Frisch used the database to examine relationships between HIV and the development of human papilloma virus (HPV)-associated anogenital cancers [abstract S5]. He found that women had a four- to seven-fold increased risk for both in situ and invasive vaginal-cervical cancer. There was "a huge increase" in anal cancer in men, especially in the younger age cohorts.
"Penile cancer saw an increased risk of 3.7-fold based on 14 cases," he said. Penile cancer was the only malignancy that seemed to correlate with reduced CD4+ cell counts. However, "All of the excess of penile cancer was found in populations who are less likely to be circumcised, blacks and Hispanics," leading Frisch to suggest that the explanatory factor may be lack of circumcision rather than low CD4+ count.
That interpretation is consistent with research by Thomas Quinn (Johns Hopkins University, Baltimore) in Uganda. Quinn found that uncircumcised males were more likely than their circumcised peers to become infected with HIV from their female partners through vaginal sex. It also is consistent with in vitro and animal studies of how HIV can infect penile tissue.
Eric Engels (NCI) reported on the relationships of coinfection with HIV and hepatitis C virus (HCV) to hepatocellular carcinoma and NHL [abstract S4]. In mono-infected patients the risk of HCV progression to cirrhosis and hepatocellular carcinoma is unknown, and is likely influenced by viral subtype and host factors.
Engels stratified 61 cases of hepatocellular carcinoma into six groups based on likely routes of HIV infection. "Relative risk was highest in those with hemophilia, over 50. Relative risk was intermediate, around 10, for the IV drug users, and it was lower still for persons in the other risk groups," said Engels. This indicated that HCV was not the only factor leading to progression.
In the HIV-infected population, those who were coinfected with HCV had a 2.9 greater relative risk of progression to hepatocellular carcinoma. "But importantly, even in the absence of hepatitis C virus infection," Engels noted, "HIV-infected persons appear to have an elevated risk for hepatocellular carcinoma." He estimated that risk at 4.5. Engels speculated that this might be due to higher incidence of exposure to hepatitis B, alcohol abuse, or other risk factors in the targeted populations.
A longitudinal examination of the population from time of infection forward found that there was "no evidence for an effect of immunosuppression on risk for hepatocellular carcinoma."
Studies looking for a relationship between HCV infection and increased risk for NHL have reached conflicting conclusions. Engels' findings on NHL mirrored those on hepatocellular carcinoma: "We do not see any trend in risk within any of the grades of lymphoma."
In subsequent discussion among the participants, Bernal said that HIV increases the risk for cancer "thousands of fold." HAART reduces that, "but the excesses are not disappearing, they are just being reduced." Robert Yarchoan, an NCI researcher, noted that some cancers are more responsive to CD4+ counts than others. He asked what factors other than immunosuppression might be responsible for progression from in situ to invasive cancer.
Frisch said that there might not be other factors. He suggested that perhaps the odds of progression simply increase as the number of cancer cells increases.
Joel Palefsky (University of California, San Francisco [UCSF]) said that with cervical dysplasia "we see a higher and higher proportion of genetic changes" which may simply cascade into aggressive growth of the tumor. He also suggested that there may be factors at the local tissue level that affect the switch from in situ to invasive cancer.
The human herpes virus-8 (HHV-8) is "quite ancient," at least 100,000 years old, said Gary Hayward (Johns Hopkins University, Baltimore). There are "essentially four flavors" of the virus tied to human migration and evolution. Type B predominates in sub-Saharan Africa, type D in South Asia, and types A and C across North Asia, Europe, and the Americas.
HHV-8 has been linked conclusively to development of KS when a declining immune system no longer can hold the virus in check. But how HHV-8 is transmitted remains a mystery. Another mystery is why, in the United States and Europe, gay men are much more likely than the rest of the population to be seropositive for HHV-8.
There are no documented cases of HHV-8 transmission through commercial blood products, despite the fact that several patients have received transfusions from carriers of the virus. A low prevalence of HHV-8 among injection drug users is a further indication that it is not blood-borne, or at minimum has an extremely low transmission efficiency through blood and blood products.
Jeffrey Martin (University of California, San Francisco) is puzzled about why female sex workers have lower rates of HHV-8 infection than do gay men literally living next door.
He reviewed data on bodily fluids that could harbor HHV-8 [abstract S7]. "Any act that involves passage of these bodily fluids between gay men has to be considered fair game for transmission," said Martin. Sampling has found HHV-8 in the semen of 15 percent of gay men, while the saliva of those same men yielded more than twice that rate of virus, 35 percent (Table 1).
It has been "very hard to find" HHV-8 in fecal matter, he said. Still, because initial research in the 1980s found a correlation between oral-anal sexual practices and being seropositive for HHV-8, that possible route of transmission continues to loom large in the minds of many.
"But mere presence of the virus in a bodily fluid does not mean that passing the fluid can transmit the virus," said Martin. Both HIV and hepatitis C are present in saliva, "but we do not believe that kissing is an important route of transmission" for them.
Research in the Republic of Cameroon has shown that rates of HHV-8 infection are low in children, but increase at a constant rate of about 1.2 percent a year. By age 20, about half the population is seropositive for the virus. Patt erns of infection reflect the serostatus of the mother and siblings of the child, but not the father [abstract S8].
Lawrence Corey (University of Washington, Seattle) had a simple explanation for why gay men have such a high rate of HHV-8 infection: "They kiss everybody, their mother, their sisters, their boyfriends." The researcher was being humorous but not facetious.
He found that 35 percent of the gay men he tested were positive for HHV-8 [abstract S10]. In a detailed study of 27 gay men who carried HHV-8, 11 of whom also were HIV-infected, Corey took tissue and fluid samples once a week for four weeks.
More than half of the men shed virus in their saliva at some point during the study. That was significantly higher than shedding found at any other body site. And the amount of virus being shed was 2.5 logs higher in saliva than in the next most active site.
A third group of gay men was asked to collect separate daily swabs from the genital, anal/rectal, and oral areas for 47 consecutive days. Corey and his colleagues were able to isolate HHV-8 DNA from 33.7 percent of the samples of saliva, but from only 1.3 percent of the anal samples and 0.3 percent of the urethral samples.
Those who shed the virus seemed to do so rather consistently. Men coinfected with HIV seemed more likely to shed HHV-8, but the subset was too small to allow Corey to reach a definitive conclusion.
Corey hypothesized from work with tissue cultures, "The wide spectrum of cell types that are infected suggests that cofactors may enhance the replication or transmission of HHV-8." Subsequent experiments in tissue showed that "CMV [cytomegalovirus] markedly upregulates HHV-8 and produces higher titers of infectious virus than we have ever seen before." His lab is now looking to see if active CMV infection in patients correlates with increased shedding of HHV-8 in saliva.
Corey concluded, "The prevalence and titers of HHV-8 are so much higher in saliva that it has to be a major form of transmission." He believes that the virus is much more common than is currently thought. As with many ubiquitous infections, such as Epstein-Barr virus, he theorized, a healthy immune system readily controls it. Problems emerge as that response becomes severely compromised.
"As our knowledge of HHV-8 has evolved, it has become clear that something in addition to infection of the host is necessary for Kaposi's sarcoma" to develop, said Margaret Offermann (Emory University, Atlanta).
One possibility is that the virus must infect endothelial cells. Working with tissue cultures, Offermann found that the cytokine interferon-a did not completely protect a cell from HHV-8 but that it did dramatically reduce infection [abstract 8]. It appears to set off a chain reaction of events that "decreases protein synthesis" and makes it more difficult for the virus to reproduce.
"Interferon protects endothelial cells from infection with HHV-8. It reduces the amount of virus that gets into the cells," Offermann said. When cells are already infected, interferon "enhances elimination of the virus-infected cells." In contrast, the cytokine "TNF-a enhances the rate of infection of endothelial cells with HHV-8, leading to higher levels of viral gene expression," she said. Modulating these, and perhaps other, immune functions offers a promising line of inquiry for developing therapy for KS.
"Angiogenesis is thought to be a critical pathway in the pathogenesis of Kaposi's sarcoma, many other cancers, and complications of diabetes and cardiovascular disease," said Ellen Feigal (NCI). She posed a series of questions associated with evaluating and monitoring antiangiogenesis agents used as therapy.
"From the beginning we've known that KS is a very vascular tumor," said NCI researcher James Pluda. He reviewed the cytokines known to promote angiogenesis and noted that "many of these are inflammatory cytokines as well." Inflammatory cytokines can "increase expression of activated receptors" in endothelial cells and affect KS pathogenesis.
He visually summed up the entire interactive pathogenic process with his "Jerry Garcia slide." It shows "how complicated the process is, where we might be able to interfere in the pathogenic process, and why it may be necessary to interfere in more than one step, because of redundancy within the process."
Pluda listed the angiogenesis inhibitors now in clinical study. He suggested that there is a bell curve response to interferon-a and that we are currently using it improperly. Perhaps lower doses "on the order of 1 million units a day" may have a greater antiangiogenesis effect than currently used higher doses.
Finally, Pluda stressed the need to develop biological assays in preclinical drug development and bring them forward in clinical trials.
Feigal noted that clinical response to therapy can be quite delayed, sometimes by six months or more. "Is there a point earlier than seeing the clinical response [when] we can tell if the therapy is modulating the pathway?" she asked. And if it is modulating a pathway, "Does that have an impact on anything clinical?" The hope is that imaging can point to an earlier identification of such biological markers (Figure 1).
Non-Hodgkin's lymphoma is four-fold more deadly than KS, said David Scadden (Harvard University, Boston) in opening a roundtable on lymphoma. Response rates of HIV patients are generally 10 to 15 percent lower than patients with similar conditions. In general, "one-third to one-half of our patients do not get an adequate response from initial therapy," Scadden said. He asked if therapy should not be modified for these patients.
Joseph Sparano (Montefiore Cancer Center, New York) reviewed data on "standard CHOP regimens" (cyclophosphamide, doxorubicin, vincristine, and prednisone) in HIV patients over 10 years [abstract S15]. He said that patients on didanosine (ddI) experienced fewer side effects of chemotherapy--neutropenia, thrombocytopenia, leukopenia, and fewer red cell and platelet transfusions. Use of saquinavir "was associated with a significant increase in degree of grade 3 or 4 mucositis." CD4+ cells also declined "by about 50 percent."
Sparano's team's more recent trials have added rituximab to the CHOP combination because it has a different mechanism of action and nonoverlapping toxicity, plus "preclinical data indicating at least an additive effect," said Sparano. In 12 patients, they saw "no significant difference in the incidence of infection," while also seeing roughly comparable side effects.
Richard Little (NCI) updated the EPOCH study of 33 patients, which is based on the concept of lower dose infusion over an extended period [abstract S16]. It requires discontinuation of antiviral therapy during the course of chemotherapy. The response rate is not significantly different from CHOP but those who did have a positive response did not have a single relapse.
"Matrix metalloproteinases (MMPs) are a family of enzymes involved in the degradation of extracellular membranes," said Mary Cianfrocca (Fox Chase Cancer Center, Philadelphia). MMPs correlate with tumor aggressiveness, likely facilitate metastasis, and play a role in angiogenesis.
Col-3 is a chemically modified tetracycline that has anti-MMP, or inhibitory activity. The AIDS Malignancy Consortium designed an open-label phase I trial of Col-3 for 18 patients with KS involving the skin, lymph nodes, oral mucosa, or GI tract [abstract 75]. The dosing was 25, 50, and 70 mg/m2 administered orally once a day.
"Partial responses were seen in all dosing cohorts, for an overall response rate of 33 percent," said Cianfrocca. Three patients had a durable response lasting at least six months, and one at least a year. Analysis is currently underway to see if responses correlated with MMP-2 and MMP-9 levels, which would validate use of these as biomarkers to evaluate the effect of potential therapies at an earlier point in a trial.
Richard Ambinder (Johns Hopkins Oncology Center, Baltimore) discussed a novel approach to therapy. "Herpes virus-associated malignancies all carry two potential viral prodrug converting enzymes, thymidine kinase and a phosophorotransferase also known as a protein kinase," he said. These are expressed in lytic infection but not in tumors. Perhaps one could use pharmacologic agents to upregulate their expression in tumors, increase the intracellular production of ganciclovir, and kill the offending cells.
He found that butyrate, phenylbutyrate, and a number of other compounds had a considerable killing effect in various cell lines that harbored virus but had very little effect on uninfected cells. "This was a clue that maybe we had been trying to do things the hard way. Lytic infection itself is pretty good at killing cells," said Ambinder.
He proposed that in Epstein-Barr Virus associated tumors and HHV-8 associated tumors, "Induction of lytic infection may kill cells directly, or may lead to the induction of enzymes that activate suicide substrates." In addition, there is a strong T-cell response to lytic antigens, and little response to tumors. Thus, inducing a lytic response would bring killer T-cells back into the equation.
Ambinder said that butyrates are "agents with which we have many years of experience and are already known to be well tolerated." Thus they have the potential of being moved quickly into human trials.
The REACH [Reaching for Excellence in Adolescent Care and Health] study aims to gain a better understanding of adolescents who become infected with HIV [abstract S26]. The multicenter study has 15 participating institutions in 13 US cities.
The REACH cohort of 364 "is overwhelmingly female, over 75 percent," explained Anna-Barbara Moscicki (University of California, San Francisco). The majority live at home and few report IV drug use. There is a matched cohort of non-HIV-infected adolescents.
The study participants are interesting because their infections are recent, they are biologically immature with possibly different risks of disease progression, and their thymic potential for mounting an immune response may differ from adults. Flow cytometery has shown that at baseline these adolescents have high levels of CD8+ naïve cells compared with adults.
During puberty a process of squamous metaplasia occurs in the vagina, creating a new squamocolumnar junction. "We believe this is a very important piece in the role of HPV in adolescence." Moscicki said the process leaves them more vulnerable to both HPV infection and the expression of lesions because of hormonal changes during adolescence.
Her work last year showed that "for every 10 percent change that we saw in the cervix that represented squamous metaplasia, we had a three-fold [increased] risk of developing a low grade SIL [squamous intraepithelial lesion] lesion within four months."
Moscicki "found absolutely no correlation with CD4+ count or viral load and the rate of anal HPV; only HIV status remained significant" in the multivariate analysis. Men in the cohort tested positive for HPV in the anal canal 55 percent of the time, and women 26 percent. The men had a lower average viral titer.
The absence of chlamydia correlated with SILs. This surprising association led to two alternative hypothesis: That inflammation associated with the infection might obscure lesions and lead to misdiagnosis; or that chlamydia induces Th1 responses, "which may be protective against HPV." Moscicki speculated that Th2 type responses "may either be detrimental or have no effect at all."
HPV is a localized infection, so REACH looked at cytokine responses at localized sites over four-month intervals. It found that women who were positive for the virus on one visit but had cleared it by the next visit all showed a Th1 response."They were interferon gamma positive and IL-4 negative," Moscicki said.
At a plenary session on the therapeutic challenges and opportunities of anal squamous intraepithelial lesions, Joel Palefsky (University of California, San Francisco) presented data suggesting that anal cancer is "a significant problem in HIV-positive men and potentially in HIV positive women." Both the incidence and clinical significance of these cancers are likely to rise in the era of HAART.
"Unlike all of the other cancers that you have heard about at this meeting, anal cancers are potentially completely preventable" through recognition of early signs and appropriate action, said Palefsky. He pointed out that before the adoption of routine Pap smear screening for vaginal-cervical cancer, the incidence was 40 to 50 per 100,000 women in the United States. Now it is 8 to 10 per 100,000.
The incidence of anal cancer in the United States is only 0.9 per 100,000. But among men who have sex with men, before the advent of HIV, incidence of anal cancer "was estimated to be as high as 35 per 100,000." Data indicate that risk is twice as high among those who are HIV-infected, which is "roughly 10 times higher than the current rate of cervical cancer," said Palefsky.
Anal cancer presents much the same as its cervical counterpart. The anorectal junction mirrors the vaginal-cervical transformation zone as the most likely s ite of problems, while lesions and invasive cancer develop in a similar manner.
Testing for the presence of HPV in the anal canal is not an effective screen. A sample of 600 gay men found that virtually all of them had been exposed to the virus, though only a small percent would develop lesions.
Palefsky lamented the fact that "basically nobody" is doing routine anal cytology screening, biopsies, and treatment. He called it "one of the world's easiest screenings" that can be performed. "The key is that the swabs have to be dacron, not cotton, because dacron is much better at giving up the cells in the liquid solution for the slides."
"What we are looking for is big ugly cells with large nucleoside/plasma ratios that are highly suggestive of a high-grade lesion," said Palefsky. Risk increases as the CD4+ count falls. Below 200 cells/mm3, "more than 75 percent had abnormal anal cytology."
High-resolution anoscopy (HRA) should be used to examine suspect tissue. "The most severe dysplasia has an appearance that you can entirely miss if you are doing a standard venoscopic examination with no magnification."
Palefsky said it is essential to use vinegar "to allow for the more subtle lesions to show up." He uses a stick wrapped in gauze and soaked in 3 percent vinegar solution. He inserts it through the anoscope, then withdraws the tube to let the gauze have contact with anal mucosa for about a minute, before withdrawing the swab and reinserting the anoscope.
The same criteria used to evaluate cervical lesions are used to evaluate anal lesions. He urged a detailed examination because "microinvasive cancers" often are found at the base of warts.
Palefsky followed patients prospectively over four years in the pre-HAART era. "Using histology as the outcome, not using cytology to grade disease, what you find is that a very high proportion of these patients develop an anal high-grade lesion."
Most researchers believe that localized HPV infection is acquired soon after initiating sexual activity and is initially suppressed by the immune system. As control of the virus wanes, low-grade lesions develop and progress to high-grade ones. "We suspect it takes years" to progress from low- to high-grade lesions, Palefsky said.
"In the pre-HAART era, a lot of patients died of other [noncancerous] causes," he said. An ongoing study of 800 gay men at UCSF is trying to determine if HAART has an impact on the natural history of anal cancer. Initial data "strongly" suggest that it does not; there is normal progression to disease. "The one little glimmer of hope" is that patients with the highest CD4+ counts when they begin HAART are the least likely to progress. "This data suggests that [anal cancer] is going to become more of a clinical problem, not less."
Survey data from HIV-positive women show rates of anal HPV infection "almost as high as with the men, when categorized by CD4+ levels. That surprised me enormously," said Palefsky, because women are less likely to engage in anal intercourse and do so less frequently than gay men.
Palefsky is urging adoption of anal screening as a cost-effective medical procedure for gay men and people who are HIV positive. Currently, most lesions are not detected until they are advanced beyond the ability to treat in the office, thus requiring the services of a colorectal surgeon using either cold scalpel incision or laser ablation techniques.
He showed a video of a laser burning away portions of a lesion into a charred mass. The audience visibly flinched at the sight. "I think you can understand why patients complain about the pain," he said. One should anticipate the patient being laid up two to three weeks after the procedure.
Palefsky has developed a one-week training program on screening for anal cancer at UCSF, which includes use of high-resolution endoscopy. The initial priority is to train those participating in the AIDS Malignancy Consortium.
Bob Roehr is a medical writer based in Washington, DC (BobRoehr@aol.com).
000810
IA000803
Copyright © 2000 - International Association of Physicians in AIDS Care. All rights reserved. http://www.iapac.org
ÆGiS is made possible through unrestricted grants from Roxane Laboratories, Inc., iMetrikus, Inc., the National Library of Medicine, and donations from users like you. Always watch for outdated information. This article first appeared in 2000. This material is designed to support, not replace, the relationship that exists between you and your doctor.
ÆGiS presents published material, reprinted with permission and neither endorses nor opposes any material. All information contained on this website, including information relating to health conditions, products, and treatments, is for informational purposes only. It is often presented in summary or aggregate form. It is not meant to be a substitute for the advice provided by your own physician or other medical professionals. Always discuss treatment options with a doctor who specializes in treating HIV.
Copyright ©1990, 2000. ÆGiS & the Sisters of Saint Elizabeth of Hungary. All materials appearing on ÆGiS are protected by copyright as a collective work or compilation under U.S. copyright and other laws and are the property of ÆGIS and the Sisters of Saint. Elizabeth of Hungary, or the party credited as the provider of the content.
