AIDS defining neoplasms prevalence in a cohort of HIV infected patients, before and after highly active antiretroviral therapy

Abstract

Introduction

Malignant disorders have been linked to HIV epidemic from its onset. Implementation of highly active antiretroviral therapy (HAART) has resulted in a dramatic reduction in the HIV/AIDS morbidity and mortality. The present study evaluates the neoplasm prevalence before and after the implementation of HAART.

Methods

A cross-sectional study was performed in 171 adults HIV infected persons followed in Puerto Rico, between May 1992 and December 2005. Neoplasm prevalence was measured and the difference in AIDS and non-AIDS defining neoplasms was analyzed before and after the HAART era. Chi-square, Fisher exact test, ANOVA and student t test were used to explore differences between groups.

Results

Malignant neoplasms were detected in 171 patients (4.8%). Of these, 51.5% were AIDS defining neoplasms and 68% were established before HAART. AIDS defining neoplasms accounted for 62.4% of the neoplasms before the availability of HAART and 25.9% after HAART. Except for cervical carcinoma, the prevalence of AIDS defining neoplasms was decreased after HAART. Non-AIDS lymphomas and prostate neoplasms were more frequent after HAART. Discussion: Our study finds a significant reduction of Kaposi's sarcoma and AIDS related lymphoma in the HAART era of the epidemic. A higher prevalence of non-AIDS defining lymphomas, prostate and cervical carcinoma were seen in the HAART era. These findings suggest that factors other than severe immunosuppression are involved in the neoplasms' pathogenesis. Preventive strategies that include screening tests, vaccination and life style modification should be routinely applied in the HIV infected patients.

INTRODUCTION

Malignant disorders have been associated to clinical AIDS since the onset of the HIV epidemic. Kaposi's sarcoma (KS) and high-grade non Hodgkin's lymphoma (NHL) were included in the first list of AIDS defining condition in 1987.¹ Six years later invasive cervical cancer was included as an additional AIDS diagnosis criteria.² AIDS defining neoplasms are considered as additional manifestation of the high degree of disruption of the immune system. The introduction of highly active antiretroviral therapy (HAART) in 1996 increased the life expectancy among subjects infected with HIV in large part due to an improvement in the immune status of patients and a lowering of the HIV viral load. In Puerto Rico HAART was routinely given to patients as of 1998 with the same combination of Protease inhibitors, Nucleosides Transcriptase Inhibitors and non Nucleosides Transcriptase Inhibitors antiretroviral therapies reported to be highly effective in HIV/AIDS patients. The combination of these drugs partially restores defects in cell-mediated immunity, suppress the HIV viral load and improve the patients' immunological and clinical status.^3,4 As a consequence, AIDS defining illnesses, including Kaposi sarcoma and high-grade NHL declined considerably after the introduction of these drugs. ^4–7 The pathogenesis of invasive cervical carcinoma appears to be more related to papilloma virus infection and HAART will not decrease its incidence.⁸ Eltom et al found a significant decline in the AIDS-related NHL in the general US population between 1996 and 1998, which was mainly attributed to the improvement of the immunological suppression of HIV/AIDS patients after the implementation of HAART.⁹ On the other hand, with the improvement in survival many HIV infected patients have experienced, a number of co-morbid entities are having a major role in the morbidity and mortality of patients. Non AIDS neoplasms are co-morbid conditions that increase the morbidity, change the therapeutic index and worse the mortality profile of patients with HIV infection.^5,10–14

The prevalence of cancer in men in Puerto Rico for the year 2000 was 300 per 100,000; with the highest prevalence seen in prostate cancer with 100 per 100,000 followed by colorectal cancer with 40 per 100,000. The cancer prevalence in women was 250 per 100,000, principally related to breast cancer 85 per 10,000 and followed by the colorectal cancer 30 per 100.000. The principal causes of cancer deaths were prostate cancer 30 per 100,000 and pulmonary cancer 20 per 100,000 in men and breast cancer 18 per 100,000 and colorectal 11 per 100,000 in women.¹⁵ In this paper we measure and evaluate the prevalence of malignant disorders, including the AIDS related and the non-AIDS related neoplasms in a cohort of HIV infected persons before and after HAART came into widespread use in Puerto Rico.

METHODS

Study population

The sample was comprised of a group of 171 patients, selected form 3,576 adults HIV infected patients enrolled in the our institutions Retrovirus Research Center (RRC) cohort between May 1992 and December 2005. Our cohort is composed of adult men and women who are initially seen and followed every six months at the Ramón Ruiz Arnau University Hospital or in the HIV ambulatory clinics of our institution. The cohort is composed of principally medically indigent persons with an income below the US poverty level, 29% are female and 52% have injection drug use as the major HIV risk practice. Once an informed consent was obtained, a baseline questionnaire was completed and appropriate laboratory test were conducted, including CD4+ lymphocytes count, CD8 + lymphocytes count, HIV viral load, HCV viral load, platelets, white and red cells blood counts and a comprehensive panel test that include albumin, glucose and creatinine levels are performed. Data collection and laboratory findings were reevaluated and collected every 6 months. Most of the HIV risk behaviors and socio demographic data were gathered via personal interviews. The information was complemented with medical record abstraction in which clinical manifestation of the infection, laboratory data, therapy-related information and medical complications are abstracted and recorded. Participants were interviewed by an expert case manager, who is fluent in Spanish and English, and all information is kept strictly confidential. All of our participants spoke Spanish as their primary language. The questionnaire included Demographic variables, HIV risk factors, clinical manifestations variables, and mortality related information. AIDS defining illnesses, such as esophageal candidiasis, Pneumocystis jirovencii pneumonia (PJP), cerebral toxoplasmosis, recurrent bacterial pneumonia, pulmonary tuberculosis, Kaposi's sarcoma, high- grade non-Hodgkin lymphoma, invasive cervical carcinoma, and wasting syndrome were recorded. Non-AIDS defining neoplasms were also tabulated and organized into several categories on the basis of the primary organ of tumor origin. The HAART era was defined as the period when HAART was available for the HIV therapy. In Puerto Rico HAART has been routinely given to all qualified patients after 1998; consequently we divided the epidemic in two time periods; the pre HAART era, which ends in 1998, and the HAART era, which begins in 1999. The status of the study participants as of December 2005 was used to measure the mortality trends. Mortality data were obtained from a review of the institutional medical records and from the Puerto Rican AIDS surveillance system. In addition the mortality registry of the Puerto Rican Health Department was reviewed in order to confirm the death status of the participants. The reported causes of death were tabulated and organized into several categories which included: 1) systems or organ failure (cardiovascular, pulmonary, gastrointestinal, renal, neurological and metabolic) and 2) AIDS conditions (Kaposi's sarcoma, cerebral toxoplasma, pulmonary tuberculosis (TB) and wasting syndrome). A subgroup of liver conditions that included liver failure (chronic and acute) and cirrhosis was also evaluated.

Statistical Analysis

SPSS(SPSS, Inc., Chicago, Ill) was used to perform univariate, and bivariate analyses. Univariate analysis described the frequencies of demographic parameters, risk factor, comorbidities, mortality rates and demise causes. Differences between patients groups were analyzed with the Chi-square or Fisher exact test, ANOVA and student t test were used to evaluate means differences. Differences in mortality rates and causes of death were evaluated and analyzed in the HIV study group. The P value used to determine statistical significance was < 0.05.

RESULTS

General findings

Of the initial 3,576 HIV infected cohort 72.5% were male, all were Puerto Rican Spanish speaking persons, with a mean educational level below ninth grade, 53.8% were injecting drug users (IDUs) and 12% reported men sex with men as a HIV risk behavior. Of the entire cohort, 171 had a diagnosis of at least one malignant condition established at some point in their lives, which represent a prevalence of 4.8%, 31.5% participants with malignancies were females, 37.4% were IDUs, 46.1% were men who had sex with men and less < 60% had completed the ninth grade. Approximately 80% of the participants reported having more than two sexual partners in the last year. The malignancy prevalence was higher in men than in females (4.9% vs.4.4%) and higher in non IDUs than in IDUs (6.3% vs. 3.3%) (data not shown). In those persons with neoplasms, 74.9% were male, 51.5% had AIDS defining neoplasm, 48.5% had non AIDS defining neoplasm and 79.5% had died as of December 2005 (Table 1). The mean age at neoplasm report was 41.1 ± 11.4 years. As seen in Table 2, persons with AIDS defining neoplasm in the pre HAART era were younger, were more often men who had sex with men and had a higher prevalence of AIDS related conditions, particularly esophageal candidiasis and wasting syndrome. On the other hand, Table 1 shows that Kaposi's sarcoma (39.8%) and cervical invasive carcinoma (23.3%) were the most prevalent AIDS defining neoplasms in this sample. Breast-ovarian cancer (23.3%), non-AIDS lymphomas (19.3%) and aero-digestive malignancies (5.3%) were the most common non-AIDS defining neoplasm. In 68.5% (117) the neoplasm was reported before the HAART era and in the remaining 31.5% was reported after HAART.

Table 1.

Demographic variables, risk factors, AIDS related antecedents, neoplasm and death rates, among HIV positive Puerto Rican before and after the advent of HAART

Parameter	Pre HAART (n=117)	HAART (n=54)	Total (171)
Male %	77.8	68.5	74.9
Age mean (SD)	38.8±10.2	46.1±12.4*	41.1±11.4
Injecting drug users %	42.7	25.9§	37.4
Men sex with men (n=128)%	53.8	27.0*	46.1
Histoplasmosis antecedent %	5.1	3.7	4.7
Esophageal candidacies antecedent%	29.4	7.4*	22.2
PJP antecedent%	18.8	3.7§	14.0
Recurrent pneumonia antecedent %	11.1	3.7	8.8
Brain toxoplasmosis antecedent %	6.0	1.9	4.7
Wasting syndrome antecedent %	26.5	18.5	24.0
Hepatitis C antecedent %	3.4	14.8*	7.0
AIDS defining neoplasm %	62.4	25.9*	51.5
Kaposi's sarcoma %	52.1	13.0*	39.8
Burkitt's lymphoma %	3.4	0.0	2.3
Immunoblastic lymphoma %	2.6	0.0	1.8
Brain lymphoma %	1.0	1.9	1.2
Cervical invasive cancer (n=43) %	15.4	35.3	23.3
Non-AIDS defining neoplasm%	37.6	74.1*	48.5
Aero-digestive %	4.3	7.4	5.3
Colon-rectal %	1.7	3.7	2.3
Liver %	2.6	3.7	2.9
Skin %	1.7	5.6	2.9
Breast/Ovarian (n=43) %	26.9	17.6	23.3
Anal %	0.0	1.9	0.6
Lymphoma %	14.5	29.6§	19.3
Leukemia %	1.7	1.9	1.8
Kidney %	0.9	0.0	0.6
Prostate (n=128) %	0.0	8.1§	2.3
Mortality by December 2005 %	88.0	61.1*	79.5

^*P value<.01 between HAART groups.

^§P value<.05 between HAART groups.

AIDS= autoimmune disease syndrome; HAART= highly active antiretroviral therapy; PJP= Pneumocystis jiroveccii pneumonia; SD= Standard deviation.

Table 2.

Demographic variables, risk factors, AIDS related co-morbidities by AIDS-defining neoplasm antecedents among HIV positive Puerto Rican before and after the advent of HAART

Parameter	AIDS neoplasm (n=84)	Non AIDS neoplasm (n=87)	Total (171)
Male %	80.5	69.0	74.9
Age mean (SD)	37.6±9.1	44.9±12.4*	41.1±11.3
Injecting drug users %	43.7	31.0	37.4
Men sex with men (n=128) %	58.6	31.0	46.1
Histoplasmosis antecedent %	8.0	1.2	4.7
Esophageal candidacies antecedent %	29.9	14.3§	22.2
PJP antecedent %	16.1	11.9	14.0
Recurrent pneumonia antecedent %	12.6	4.8	8.8
Pulmonary tuberculosis antecedent %	3.4	1.2	2.3
Brain toxoplasmosis antecedent %	6.9	2.4	4.7
Wasting syndrome antecedent %	31.0	16.7§	24.0
Hepatitis C antecedent %	8.0	6.0	7.0
Mortality by December 2005 %	81.6	77.4	79.5

^*P value<.01 between HAART groups.

^§P value<.05 between HAART groups.

AIDS= autoimmune disease syndrome; PJP= Pneumocystis jiroveccii pneumonia; SD= Standard deviation.

Findings by HAART Groups

In Table 1 we present that 68% of the neoplasms were reported in the pre-HAART time. With the exception of the men having sex with men, no other significant demographic differences were seen between HAART groups. Patients in the pre HAART group were more frequently IDUs and had higher prevalence of AIDS defining conditions than did those in the HAART group. AIDS defining neoplasms accounted for 62.4% of the malignancies in the pre HAART group and only 25.9% in the HAART group, this difference was statistically significant. Non-AIDS defining neoplasm and Hepatitis C were significantly higher in the HAART group. Patients in the pre HAART group had a significantly higher mortality, than did those of the HAART group.

We found a significant reduction of Kaposi's sarcoma and a reduction or maintenance of the prevalence of the high grade non Hodgkin's lymphoma after HAART. Uterine cervical carcinoma had a higher prevalence in the post HAART group; however this difference did not reach statistical significance. Conversely, when evaluating the prevalence of non-AIDS neoplasms, we found that with the exception of breast and ovarian neoplasms the remaining malignancies were more prevalent in the post HAART group. These differences were significant in the non-AIDS defining lymphomas and in the prostate neoplasms. Similar trends were found when evaluating the clinical manifestation and neoplasm prevalence according to sex (Table 3). The differences in the post HAART group were significantly higher in the man than in women. When evaluating the group of men who have sex with men we found a significant reduction of Kaposi's sarcoma (71.4% vs. 57.1%) after HAART.

Table 3.

Age, risk factors, AIDS related antecedents, neoplasm and death rates, by sex and HAART group, among HIV positive Puerto Rican before and after the advent of HAART

Parameter	Male	Female
	Pre-HAART/HAART (n=91/n=37)	Pre-HAART/HAART (n=26/n= 17)
Age mean	38.0/47.2*	40.0/43.0
Injecting drug users %	46.2/32.4	30.8/11.8
Men sex with Men (n=128)%	53.8/27.0§	na/na
Histoplasmosis antecedent %	5.5/5.4	3.8/0.0
Esophageal candidacies antecedent %	33.0/8.1*	15.4/5.9
PJP antecedent %	17.6/2.7§	23.1/5.9
Recurrent pneumonia antecedent %	8.8/5.4	19.2/0.0
Brain toxoplasmosis antecedent %	6.6/0.0	3.8/5.9
Wasting syndrome antecedent %	27.5/24.3	23.1/5.9
Hepatitis C antecedent %	4.4/16.2§	0.0/11.2
AIDS defining neoplasm %	69.2/18.9*	38.5/41.2
Kaposi's sarcoma %	61.5/18.9*	19.2/0.0
Burkitt's lymphoma %	3.3/0.0	3.8/0.0
Immunoblastic lymphoma %	3.3/0.0	0.0/0.0
Brain lymphoma %	1.1/0.0	0.0/5.9
Cervical invasive cancer (n=43) %	0.0/0.0	15.4/35.3
Non-AIDS defining neoplasm %	38.5/81.1*	61.5/58.8
Aero-digestive %	5.5/10.8	0.0/0.0
Colon-rectal %	0.0/5.4	7.7/0.0
Liver %	3.3/5.4	0.0/0.0
Skin %	2.2/5.4	0.0/5.9
Breast/Ovarian (n=43) %	0.0/0.0	26.9/17.6
Anal %	0.0/2.7	0.0/0.0
Lymphoma %	15.4/32.4§	11.5/23.5
Leukemia %	1.1/2.7	3.8/0.0
Kidney %	0.0/0.0	3.8/0.0
Prostate (n=128) %	0.0/8.1§	0.0/0.0
Mortality by December 2005 %	87.9/70.3§	88.5/41.2*

^*P value<.01 between HAART groups.

^§P value<.05 between HAART groups.

AIDS= autoimmune disease syndrome; HAART= highly active antiretroviral therapy; PJP= Pneumocystis jiroveccii pneumonia.

Mortality Findings

Of the 171 HIV infected patients 136 (79.5%) had died at the study end. As of December 2005, 88.8% of the pre HAART group and 61.1% of the HAART group had died (Table 1). The five more prevalent causes of death included pulmonary conditions, non-AIDS neoplasm, Kaposi's sarcoma, cardiovascular conditions and gastrointestinal conditions (Table 4). In general, AIDS related causes of death were more frequent in the pre HAART group than in the HAART group. However, Kaposi's sarcoma was significantly more prevalent. Conversely, cardiovascular conditions, gastrointestinal conditions and non-AIDS neoplasm were more frequently reported as cause if death in the HAART group, although only non-AIDS neoplasm was significantly more prevalent.

Table 4.

Causes of death among HIV positive Puerto Rican before and after the advent of HAART

Decease cause	Pre HAART (n=103)	HAART (n=33)	Total (n=136)
Cardiovascular conditions %	17.5	30.3	20.6
Pulmonary conditions %	37.9	33.3	36.8
Gastrointestinal conditions %	8.7	12.1	9.6
Hepatic conditions %	4.9	6.1	5.1
Renal conditions %	5.8	6.1	5.9
Metabolic conditions %	2.9	3.8	3.1
Neurological conditions %	3.9	6.1	4.4
Non-AIDS Neoplasm %	21.4	72.7*	33.8
Brain Toxoplasmosis %	5.8	0	4.4
Wasting syndrome %	6.8	3.0	5.9
Kaposi's sarcoma %	38.8	6.1*	30.9
Pulmonary tuberculosis %	1.9	0	1.5

^*P value<.01 between HAART groups.

AIDS= autoimmune disease syndrome; HAART= highly active antiretroviral therapy.

DISCUSSION

The role of HIV in the process of inducing malignant transformation appears to be an indirect effect of virus likely related to a disruptive state of the immunoregulation of the body.¹⁶ Most malignancies in the AIDS setting appear to be related to the presence of oncogenic virus infections, such as Epstein Barr virus, Kaposi's sarcoma-associated Herpesvirus and Human Papilloma Virus (HPV).¹⁷ HIV chronically stimulaties B lymphocyte which alters the patient's antitumoral immunity. This effect appears to be synergistic with the oncogenic potential of co-infected viruses. The presence of the HIV virus itself may be capable of inducing malignancies by interacting either directly or indirectly with cell cycling and cell growth or by altering oncogene regulation after genomic integration.¹⁸ Our findings, consistent with previous published data demonstrate a significant reduction of AIDS defining neoplasms,^{6, 8, 17, 19} with the exception of cervical cancer with the introduction of HAART. Conversely, during the same time period we found a significant increment in non-AIDS defining neoplasms, particularly non-AIDS lymphomas and prostate cancer. These findings show how the profile of comorbidities in HIV-infected persons is becoming more similar to the chronic conditions of an aging population, such as non AIDS related neoplasms. Consequently, programs directed to early detection and prevention of these chronic conditions will not only reduce the morbidity and mortality of the persons, but also the costs associated to the health care of these conditions.

We believe that the introduction of HAART in 1999 in our community is partially responsible for the changes seen in the prevalence of neoplasms. The drug combination used in HAART suppresses the HIV viral load and improves immune restoration which eventually improves the clinical course of the disease.^3–7 Correction of the immunological dysfunction has been associated with a dramatic impact on the morbidity and mortality of individual living with HIV.^{3, 4, 10–14} Previous studies describe a clear relation between CD4 T cell count and progressive immunosuppression with the risk of developing Kaposi's sarcoma and AIDS defining lymphoma.^{6, 8} In scenarios in which the CD4 cell count can be partially restored by HAART, HIV related neoplasms should decrease. The increasing trend of cervival cancer in women with HIV is explained by the known causative role of HPV on this tumor and the diminished role of immunosuppression as a promoter of the cervical tumor.^{6, 8, 15} In our study, a higher prevalence of cervival cancer was seen in the HAART era as compared to the pre HAART era. Vigorous preventive and therapeutic strategies need to be continued in this high risk population of patients. Furthermore, primary prevention that include HPV vaccine needs to be highly recommended for HIV infected women.

The increasing age in patients with HIV infection has been frequently reported in industrialized countries. In this aging population the presence of non-AIDS malignancies are beginning more relevant. We have seen an increase in Non-AIDS related lymphomas, prostate cancers, aero digestive tumors, skin and gastro intestinal cancers after HAART in HIV infected patients. We are uncertain if the prevalence of these tumors in the HIV infected group is similar to the rest of the population. Nevertheless considering the relevance of the immune system in controlling growth of the malignant transformation, additional factors may be involved in the pathogenesis of these neoplasms in HIV infected patients. Tobacco use, alcohol use, diet and probably co-infections with other viruses may contribute to the malignant transformation. In general our HIV positive patients are more likely to smoke and abuse drugs and alcohol then are their HIV negative counterparts,²⁰ which may increase their risk for developing malignant neoplasms.¹¹ Strategies for reducing these risk factors, including effective preventive interventions that increase screening tests; improve dietary habits and reduce alcohol, tobacco and drugs consumption, should be designed and delivered to this high risk population.

As reported by previous authors, our study detected a lower mortality rate in patients in the post HAART era. The most significant variation of cancer related death was seen in patients with Kaposi's sarcoma. Non-AIDS conditions were more commonly reported as the causes of death in the HAART era. As previously mentioned, HAART's benefits to the immune system should sharply decrease AIDS related mortality, accompanied by an increase in the frequency of death related to non-AIDS conditions, in particular death related to non-AIDS defining malignant conditions.^11–12

Our study has several limitations: The sample was selected from a passive surveillance cohort, in which the patients were recruited from the health care facility. Patients not seen in our health care center for more than 18 months were considered to be lost to followup, which may have result in underreported data. In addition, proximally 22% of the diagnosis of neoplasm was obtained exclusively from death certificate report; however all of these patients had a complete data collection in the RRC regarding their HIV infection. The quality of the data collect in the death certificate depends on how accurately and thoroughly the demise was reported. In very few cases autopsies were performed to confirm the clinical diagnosis, which may also lead to underreport or overreport. Finally, the study measured and evaluated HAART based on the basis of the year this type of therapy was introduced in Puerto Rico and may not reflect whether individual patients received the medication. Despite these limitations, the study will help improve the treatment of HIV-infected persons in the HAAR era. Additional studies should explore in more detail the full effect of HAART on the HIV/AIDS epidemic in Puerto Rico.

Implications for Improving Health Disparities

The present study was performed in a Hispanic HIV infected population and reveals an increase prevalence of neoplasms, including cervical and prostate carcinoma. Improvements to preventive strategies are being included per the study recommendations to decrease this health disparity in HIV population.