Sage Journals: Discover world-class research

Abstract

Background

This multicentre study evaluated the prevalence, distribution, and risk factors of cancer among people living with HIV (PLWH) in Turkey over two decades. Mortality rates and factors were also investigated.

Methods

This retrospective cohort study included 154 adults diagnosed with cancer and HIV between 2008 and 2024. Data were collected from the medical records. The demographics, clinical characteristics, cancer types, treatments, and outcomes were evaluated. The risk factors for ADCs and mortality were defined using multivariate logistic regression and Cox regression. The estimated survival time was determined using Kaplan-Meier analysis.

Results

Among 8708 PLWH, cancer prevalence was 1.8% (95% CI:1.5-2). ADCs (66.9%) were twice as frequent as NADCs (33.1%). Most prevalent cancers were non-Hodgkin lymphoma (35.1%), Kaposi sarcoma (29.9%), and lung cancer (7.8%). Age ≤54 years, CD4 + T lymphocyte count ≤94/mm³, and HHV-8 co-infection were risk factors for ADCs. Mortality rate was 32%. Non-Hodgkin lymphoma and CD4 + T lymphocyte count ≤232/mm³ at treatment completion predicted mortality. The 3-years overall survival rate was 67.6%.

Conclusions

This study offers insights into cancer burden among PLWH in Turkey. High rates of ADCs and mortality highlight the importance of early HIV diagnosis and antiretroviral therapy access to reduce cancer burden in this population.

Keywords

HIV acquired immunodeficiency syndrome malignancy cancer mortality

Introduction

People living with HIV (PLWH) have an elevated risk of developing cancer due to HIV-related immunosuppression and its oncogenic potential through the tat gene.^1,2 Inflammation from HIV infection, immunosenescence, oncogenic viruses such as Epstein-Barr virus (EBV), herpes virus 8 (HHV-8), and human papillomavirus (HPV), and environmental factors such as smoking and alcohol consumption are associated with increased cancer rates.³

Antiretroviral therapy (ART) has lowered the incidence of AIDS-defining cancers (ADCs) including Kaposi sarcoma, non-Hodgkin lymphoma, and invasive cervical cancer. However, non-AIDS-defining cancers (NADCs) have increased more than two-fold compared to the pre-ART era.⁴ Common NADCs subtypes include anal carcinoma, hepatocellular carcinoma, Hodgkin lymphoma, leiomyosarcoma, oropharyngeal carcinoma, and nasopharyngeal carcinoma.⁵

Despite ART, all types of cancer are associated with elevated mortality rates among PLWH.^6,7 Factors that affect mortality include the income level of the country, presence of widespread vaccination programs, timely initiation of ART, access to cancer screening programs and cancer treatment, and financial problems.⁸

While HIV incidence has declined globally, Turkey has shown rising HIV rates, particularly among young people and women.⁹ This rising trend, combined with Turkey’s pricing policies that limit access to oncology medications, raises concerns about the burden of cancer on PLWH.¹⁰

Our study aimed to evaluate the prevalence and types of cancer among PLWH in Turkey over the past two decades. We investigated the distribution of ADCs and NADCs and the risk factors affecting this distribution.

Furthermore, our objective was to evaluate mortality rates and examine preventable factors contributing to cancer-related fatalities, including restricted access to ART and novel cancer therapies. Understanding these factors may guide interventions to reduce mortality rates and improve cancer outcomes.

Methods

Participants were recruited from various provinces across Turkey for this multicentre, observational, retrospective cohort study. The study followed the protocols of the Ethics Committee of the Kocaeli University Faculty of Medicine and the guidelines of the Declaration of Helsinki. The project received the identifier 2024/38 and approval under the code GOKAEK-2024/04.06. This study was registered in ClinicalTrials (registration number NCT06774495) on 4 December 2024.

Informed consent was not obtained because of anonymised patient data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.¹¹

The study population comprised adult patients of diverse socioeconomic levels and ethnicities. Individuals aged 18 years and older living with HIV who were diagnosed with cancer histopathologically between 2008 and 2024 were included in the study. Patient data were retrospectively assessed using the hospital’s electronic medical record system. Demographic information, vaccination history for hepatitis B virus (HBV) and HPV, risk factors for HIV acquisition, comorbid conditions (obesity, hypertension, diabetes mellitus, chronic kidney, liver, lung, and dermatological diseases, congestive heart failure, rheumatological disease, use of steroids, chronic hepatitis B and C), viral co-infections (EBV, HPV, HHV-8), CD4 + T and CD8 + T lymphocytes counts/mm³ and HIV RNA copies per millilitre of plasma, both at the time of cancer diagnosis and at recovery or death, were investigated. CD4 + T lymphocyte counts were categorised as follows: >500 cells/mm³, 201-500 cells/mm³, and ≤200 cells/mm³. Similarly, HIV viral loads were grouped as either negative or positive.

Antiretroviral therapy (ART) regimens were grouped according to major drug classes. Two nucleoside reverse transcriptase inhibitors (NRTIs) plus an integrase strand transfer inhibitor (INSTI) were classified as two NRTIs + INSTI. Two NRTIs plus a non-nucleoside reverse transcriptase inhibitor (NNRTI) were classified as two NRTIs + NNRTI, and two NRTIs plus a ritonavir-boosted protease inhibitor (PI/r) were classified as two NRTIs + PI/r. One NRTI plus an INSTI (lamivudine plus dolutegravir) formed a separate category. Integrase inhibitor–based regimens, including those boosted with cobicistat, were collectively analysed under the INSTI category.

Dates of cancer diagnosis, type, and anatomical location were recorded. Cancers were categorised as either ADCs or NADCs. The effects of patient hospitalisation requirements and treatments, including surgery, chemotherapy, radiotherapy, immunotherapy, targeted therapy, and bone marrow transplantation, on patient prognosis, in conjunction with other variables, were examined. Prognosis was evaluated across three categories: recovered, passed away, and lost to follow-up, with inaccessible data. Loss to follow-up was defined as absence from the clinic for more than 12 months without recorded death.

In our survival analysis, the time of origin was defined as the date of cancer diagnosis among people living with HIV. The primary event of interest was the all-cause mortality. The survival time was calculated from the date of cancer diagnosis to the date of death. Participants who were alive at the end of the follow-up or lost to follow-up were censored at their last known contact date. Missing data were mainly due to the inability to contact some individuals. As the proportion of missing values for the key variables was modest and the mechanism of missingness was likely to be non-informative, we conducted a complete case analysis instead of performing imputation.

Statistical analysis

All statistical analyses were performed using IBM SPSS 29.0 (IBM Corp., Armonk, NY, USA). Kolmogorov-Smirnov and Shapiro-Wilk tests were employed to evaluate the assumption of normality. Continuous variables are expressed as mean ± standard deviation or median and interquartile range (IQR). Categorical variables were summarised as numbers and percentages. To make comparisons between groups, the independent sample t-test was used for normally distributed variables and the Mann-Whitney U test was used for non-normally distributed variables. The Chi-square test with Bonferroni correction was used to analyse the associations between categorical variables. Receiver operating characteristic (ROC) curve analysis was used to calculate the cut-off, area under the curve (AUC), sensitivity, and specificity. A binary logistic regression model was used for multivariate analysis to investigate factors associated with ADCs and NADCs.

The Kaplan-Meier method with a Log-rank test and Cox regression model were used for survival analysis. The variable selection for the Cox regression model was based on clinical plausibility and univariate analyses. Individuals who experienced the event (death) and those who survived were compared with respect to the clinically relevant parameters. Variables that were statistically significant in univariate analyses were subsequently included in the multivariate Cox regression model. We formally tested the proportional hazards assumption using the cox. zph () function in R (version 4.3.0; R Foundation for Statistical Computing, Vienna, Austria).¹² Because the median survival could not be reliably estimated for all groups because of incomplete events. Therefore, the restricted mean survival time (RMST) was calculated in R using the survRM2 package with a tau value of 60 months.¹² The RMST provides a valid and interpretable estimate even in the presence of censoring. All statistical tests were two-sided, and a P-value less than 0.05 was considered statistically significant.

Results

This study included 103 patients with ADC and 51 patients with NADC, for a total of 154 cancer patients living in 26 different provinces. A total of 8708 PLWH were followed in the medical centres, and the prevalence of cancer was found to be 1.8% (95% CI: 1.5-2). As only eight of these patients were diagnosed with cancer before 2015, analyses were conducted for the entire cohort without stratification by calendar year or era of diagnosis. Supplemental Table 1 presents the types of cancer observed in the patients included in the study and Table 1 details the demographic and clinical characteristics of the population.

Table 1.

Baseline demographic characteristics and clinical features.

Characteristics	Total cohort	ADCs	NADCs	P-value
Characteristics	n = 154	n = 103 (66.9%)	n = 51 (33.1%)	P-value
Age in years, mean ± SD	46.94 ± 13	43.51 ± 12	53.84 ± 13	<0.001^a
Age ≤ 54 years, n (%)	n = 106	84 (81.6)	22 (43.1)	<0.001^b
Sex, n (%)				0.94^b
Female	17 (11)	12 (11.7)	5 (9.8)
Male	137 (89)	91 (88.3)	46 (90.2)
BMI (kg/m²), mean ± SD	23.63 ± 3.61	23.32 ± 3.38	24.27 ± 4.01	0.16^a
Occupation, n (%)				0.009^b
Service sector worker	46 (34.1)	31 (34.8)^a	15 (32.6)^a	<0.001^b
Freelancer	34 (25.2)	25 (28.1)^a	9 (19.6)^a
Retired	25 (18.5)	9 (10.1)^a	16 (34.8)^b
Unemployed	19 (14.1)	14 (15.7)^a	5 (10.9)^a
Artist	6 (4.4)	5 (5.6)^a	1 (2.2)^a
Student	5 (3.7)	5 (5.6)^a	0 (0)^a
Non-citizens, n (%)	9 (5.8)	6 (5.8)	3 (5.9)	>.99^b
Single household, n (%)	44 (28.6)	31 (30.1)	13 (25.5)	0.69^b
Regular exercise, n (%)	12 (8.8)	7 (7.6)	5 (11.1)	0.53^b
Smoking, n (%)	93 (60.4)	62 (60.2)	31 (60.8)	>.99^b
Alcohol consumption, n (%)	52 (34.9)	37 (37.4)	15 (30)	0.48^b
HBV vaccination, n (%)	69 (44.8)	43 (41.7)	26 (51)	0.36^b
MSM, n (%)	45 (29.2)	30 (29.1)	15 (29.4)	>.99^b
Comorbidities, n (%)
Obesity	8 (5.2)	4 (3.9)	4 (7.8)	0.44^b
Hypertension	15 (9.7)	7 (6.8)	8 (15.7)	0.09^b
Diabetes mellitus	10 (6.5)	5 (4.9)	5 (9.8)	0.30^b
Chronic kidney disease	8 (5.2)	5 (4.9)	3 (5.9)	>.99^b
Chronic lung disease	12 (7.8)	6 (5.8)	6 (11.8)	0.21^b
Chronic dermatological disease	8 (5.2)	4 (3.9)	4 (7.8)	0.44^b
Steroid usage	20 (13)	16 (15.5)	4 (7.8)	0.28^b
Chronic hepatitis B	10 (6.5)	8 (7.8)	2 (3.9)	0.50^b
EBV seropositivity	52 (38.5)	37 (41.1)	15 (33.3)	0.49^b
HPV Infection	10 (7.9)	9 (10.5)	1 (2.4)	0.17^b
HHV-8 infection	46 (31.5)	45 (45.9)	1 (2.1)	<0.001^b
ART-naive at cancer diagnosis, n (%)	61 (39.6)	46 (44.7)	15 (29.4)	>.99^b
CD4 cell count on cancer diagnosis, median (IQR)	157 (32-328)	92.5 (21.25-258.25)	250 (130-450)	<0.001^c
CD4 cell count 0-200, n (%)	86 (57)	65 (65)^a	21 (41.2)^b	0.012^b
CD4 cell count 201-500, n (%)	47 (31.1)	27 (27)^a	20 (39.2)^a
CD4 cell count >500, n (%)	18 (11.9)	8 (8)^a	10 (19.6)^b
CD4 cell count on cancer diagnosis ≤ cutoff: 94, n (%)	n = 58	52 (52)	6 (11.8)	<0.001^b
CD8 cell count on cancer diagnosis, median (IQR)	692 (448-1123)	774 (456-1206)	610 (425-894)	0.10^c
HIV Viral load on cancer diagnosis (copy/mL), median (IQR)	121 900 (1130-967 016)	208 000 (5771.5-968 008)	46 400 (140.40-932 393)	0.09^c
HIV Viral load negative, n (%)	13 (9.1)	4 (4.3)	9 (18)	0.012^b
HIV Viral load positive, n (%)	130 (90.9)	89 (95.7)	41 (82)
Access to the cancer therapies, n (%)	141 (92.8)	94 (93.1)	47 (92.2)	>.99^b
Hospitalisation for treatment, n (%)	107 (69.5)	71 (68.9)	36 (70.6)	0.98^b
Surgical treatment, n (%)	24 (15.6)	8 (7.8)	16 (31.4)	<0.001^b
Chemotherapy, n (%)	96 (62.3)	66 (64.1)	30 (58.8)	0.65^b
Numbers of chemotherapy cycles, median (IQR)	6 (3-8)	6 (3-8)	5 (3-8.5)	0.94^c
Radiotherapy, n (%)	18 (11.7)	9 (8.7)	9 (17.6)	0.18^b
Number of radiotherapy sessions, median (IQR)	6 (2.5-20)	3.5 (2-9)	30 (4.5-35)	0.07^c
Death, n (%)	48 (31.6)	32 (31.7)	16 (31.4)	>.99^b
Recovery, n (%)	44 (30.1)	29 (29.9)	15 (30.6)	>.99^b
Lost to follow-up	18 (12.2)	14 (14.1)	4 (8.2)	0.44^b
Switch of ART during cancer treatment, n (%)	21 (13.6)	15 (14.6)	6 (11.8)	0.82^b
CD4 cell count on recovery or death, median (IQR)	342.5 (121-641.25)	263 (71-485)	431 (272.5-697)	0.006^c
CD4 cell count 0-200, n (%)	41 (32.5)	37 (43.5)^a	4 (9.8)^b	<0.001^b
CD4 cell count 201-500, n (%)	50 (39.7)	28 (33)^a	22 (53.7)^a
CD4 cell count >500, n (%)	35 (27.8)	20 (23.5)^a	15 (36.6)^a
CD8 cell count on recovery or death, median (IQR)	693 (388.75-1186.25)	718.50 (277.75-1240.50)	626 (440.25-920.75)	0.73^c
HIV Viral load on recovery or death (copy/mL), median (IQR)	0 (0-229)	45 (0-306.50)	0 (0-79.75)	0.028^c
HIV Viral load negative, n (%)	67 (53.6)	40 (47.1)	27 (67.5)	0.05^b
HIV Viral load positive, n (%)	58 (46.4)	45 (52.9)	13 (32.5)

ADCs: AIDS-defining cancers; NADCs: non-AIDS-defining cancers; SD: Standard deviation; IQR: Interquartile range; BMI: Body mass index; MSM: Men who have sex with men; ART: Antiretroviral therapy; HBV: Hepatitis B virus; HPV: Human papillomavirus; HHV-8: Human herpes virus 8; EBV: Epstein-Barr virus.

^aIndependent t-test.

^bChi-square test.

^cMann-Whitney U Test. A statistically significant difference was identified between column proportions with different uppercase letters in the same row (p < 0.05, chi-square test with Bonferroni correction).

A total of 103 patients (66.9%) had ADCs. The most prevalent cancer types were non-Hodgkin lymphoma (35.1%), Kaposi sarcoma (29.9%), lung cancer (7.8%), Hodgkin lymphoma (5.8%), and gastrointestinal system cancers (5.2%). Among gastrointestinal system cancers, colon cancer was found in four patients, gastric cancer in two patients, ampulla vateri cancer in one patient, and anal canal cancer in one patient.

Nine of the patients had a family history of cancer. The hepatitis B vaccination rate was 44.8%, and the HPV vaccine was administered to two patients. It is also important to note the low frequency of the comorbidities. Heart failure was found in five patients and rheumatologic disease in one. Among the viral co-infections, hepatitis C virus was identified in one patient, while hepatitis delta virus was not detected in any patient.

At the time of cancer diagnosis, 75 individuals (48.7%) were receiving two NRTIs plus an INSTI, 61 (39.6%) were not receiving ART, 9 (5.8%) were receiving two NRTIs plus a PI/r, 6 (3.9%) were receiving one NRTI plus an INSTI, and 3 (1.9%) were receiving two NRTIs plus an NNRTI. After cancer diagnosis, all individuals except two were receiving ART regimens as recommended by the current guidelines. Of the two exceptions, one discontinued ART, and the other was diagnosed with HIV and cancer simultaneously but could not be initiated on ART and died before treatment could start. During cancer treatment, the most common regimen was two NRTIs plus an INSTI, received by 128 individuals (83.1%). 14 individuals were receiving one NRTI plus an INSTI, 8 were on a PI/r-based combination, and 2 were receiving two NRTIs plus an NNRTI.

When the ADCs and NADCs groups were compared, significant differences were found in terms of retirement status, age, HHV-8 co-infection, CD4 + T lymphocyte count at the time of cancer diagnosis, and negative HIV viral load (Table 1).

Receiver operating characteristic (ROC) curve analysis was performed to determine the age and CD4 + T lymphocyte count thresholds for AIDS-related cancers (Supplemental Figure 1). Multivariate logistic regression analysis using the threshold values revealed that being 54 years of age or younger (aOR: 4.62, 95% CI: 1.81-11.78), having a CD4 + T lymphocyte count CD4 + T lymphocyte count is 94/mm³ or below (aOR: 7.40, 95% CI: 2.4-22.87), and the presence of HHV-8 co-infection (aOR: 24.46, 95% CI: 3.09-193.93) were statistically significant in terms of ADCs development (Supplemental Table 2).

48 patients in the study cohort died. Surgery, chemotherapy, and radiotherapy were the most commonly used cancer treatments. Targeted therapy was applied in six patients, immunotherapy in three, hormone therapy in three, and bone marrow transplantation in five. Table 2 compares the demographic and clinical characteristics of deceased and surviving patients. Statistically significant factors were a history of HBV vaccination, HHV-8 co-infection, CD8 + T lymphocyte count at the time of cancer diagnosis, presence of NHL and Kaposi sarcoma, access to appropriate oncologic treatment, number of chemotherapy cycles, CD4 + and CD8 + T lymphocyte counts, and HIV viral load (copies/mL) at the time of recovery or death. A CD4 + T lymphocyte count ≤232/mm³ and HIV viral load >641 copies/mL at the end of cancer treatment were statistically significant predictors of mortality (Supplemental Figure 2).

Table 2.

Comparison of clinical and demographic characteristics between deceased and survived patients.

Characteristics	Survived	Deceased	P-value
Characteristics	n = 104 (68%)	n = 48 (32%)	P-value
Age in years, mean ± SD	45.56 ± 13.32	49.73 ± 12.45	0.07^a
Sex, n (%)			0.75^b
Female	12 (11.5)	4 (8.3)
Male	92 (88.5)	44 (91.7)
BMI (kg/m²), mean ± SD	23.92 ± 3.51	23.1 ± 3.85	0.24^a
HBV vaccination, n (%)	57 (54.8)	12 (25)	0.001^b
HHV-8 infection, n (%)	36 (36.4)	8 (17.8)	0.04^b
ART-naive at cancer diagnosis, n (%)	42 (40.4)	19 (39.6)	>.99^b
CD4 cell count on cancer diagnosis, median (IQR)	177 (42.5-350.25)	157 (24-269)	0.25^c
CD8 cell count on cancer diagnosis, median (IQR)	750 (478-1230)	607 (382.5-871)	0.043^c
HIV viral load on cancer diagnosis (copy/mL), median (IQR)	130 962 (990.45-877 809)	81 455 (892-1 120 000)	0.84^c
Type of cancer, n (%)
Non-hodgkin lymphoma	29 (27.9)^a	25 (52.1)^b	0.022^b
Kaposi sarcoma	37 (35.6)^a	7 (14.6)^b
Hodgkin lymphoma	8 (7.7)^a	1 (2.1)^a
Gastrointestinal cancer	6 (5.8)^a	2 (4.2)^a
Lung cancer	7 (6.7)^a	5 (10.4)^a
Other cancer	17 (16.3)^a	8 (16.7)^a
Access to the cancer therapies, n (%)	103 (99)	37 (80.4)	<0.001^b
Hospitalisation for treatment, n (%)	67 (64.4)	38 (79.2)	0.10^b
Surgical treatment, n (%)	20 (19.2)	4 (8.3)	0.14^b
Chemotherapy, n (%)	69 (66.3)	27 (56.3)	0.31^b
Numbers of chemotherapy cycles, median (IQR)	6 (4-8.25)	4 (2-6)	0.002^c
Radiotherapy, n (%)	11 (10.6)	7 (14.6)	0.66^b
Switch of ART during cancer treatment, n (%)	17 (16.3)	4 (8.3)	0.28^b
CD4 cell count on recovery or death, median (IQR)	415 (219-696.3)	133 (37.8-325.5)	<0.001^c
CD4 cell count ≤ cutoff: 232/mm³, n (%)	21 (25.6)	29 (69)	<0.001^b
CD4 cell count > cutoff: 232/mm³, n (%)	61 (74.4)	13 (31)	<0.001^b
CD8 cell count on recovery or death, median (IQR)	844 (520-1338)	425 (183.5-785.5)	<0.001^c
HIV Viral load on recovery or death (copy/mL), median (IQR)	0 (0-115)	68 (0-2509)	0.003^c
HIV Viral load, ≤ cutoff: 641, n (%)	76 (92.7)	27 (65.9)	<0.001^b
HIV Viral load > cutoff: 641, n (%)	6 (7.3)	14 (34.1)	<0.001^b

BMI: Body mass index; HBV: Hepatitis B virus; HHV-8: Human herpes virus 8; SD: Standard deviation; IQR: Interquartile range.

^aIndependent t-test.

^bChi-square test.

Table 3 presents the results of the univariate and multivariate Cox regression analyses used to predict mortality. Multivariate analyses revealed that NHL (adjusted hazard ratio [aHR]: 2.94, 95% CI: 1.40-6.20, p = 0.005) and a CD4 + T lymphocyte count ≤232/mm³ at the end of cancer treatment (aHR: 4.1, 95% CI: 1.65-10.15, p = 0.002) were factors that increased mortality.

Table 3.

Univariate and multivariate Cox regression analysis for mortality outcomes.

Variable	Univariate		Multivariate
Variable	HR (95% CI)	P	aHR (95% CI)	P
HBV vaccination		0.002		0.15
Yes	3.01 (1.52-5.99)		1.86 (0.80-4.31)
No	1.00 (reference)		1.00 (reference)
Type of cancer				0.005
Non-hodgkin lymphoma	2.82 (1.54-5.17)		2.94 (1.40-6.20)
Other types of cancers	1.00 (reference)		1.00 (reference)
CD4 cell count on recovery or death (cutoff: 232/mm³)		<0.001		0.002
≤232	6.81 (3.19-14.56)		4.1 (1.65-10.15)
>232	1.00 (reference)		1.00 (reference)
HIV viral load on recovery or death, copy/mL (cutoff: 641)		<0.001		0.19
>641	5.52 (2.73-11.16)		1.75 (0.76-4.04)
≤641	1.00 (reference)		1.00 (reference)
Access to the cancer therapies		<0.001		0.15
No	5.07 (2.32-11.09)		1.96 (0.78-4.91)
Yes	1.00 (reference)		1.00 (reference)

HR: Hazard ratio; aHR: Adjusted hazard ratio; CI: Confidence interval.

The proportional hazards assumption was met in our model, as both the individual variable tests and the global test showed no evidence of violation, with all p-values exceeding 0.05 (see Table 4 for details).

Table 4.

Cox proportional hazards assumption results for individual variables and the global model (cox.zph, R 4.3.0).

Parameters	Chi-square (χ²)	P Value
Access to the cancer therapies	0.00596	0.94
HBV vaccination	0.92396	0.34
HIV viral load on recovery or death, copy/mL (cutoff: 641)	0.93720	0.33
CD4 cell count on recovery or death (cutoff: 232/mm³)	0.22575	0.63
Type of cancer	0.89223	0.34
GLOBAL	2.12532	0.83

According to Kaplan-Meier survival analysis, the overall survival rate for all cancer types was 76.7% (Standard Error [SE]: 0.036) at 1 year, 70.7% (SE: 0.040) at 2 years, and 67.6% (SE: 0.044) at 3 years. The restricted mean survival time of the study population was 43.2 months (SE: 0.71).

When the cancer types were grouped as NHL, Kaposi sarcoma, and other cancers, the estimated restricted mean survival time was 50.36 months (SE: 1.03) for NHL, 33.40 months (SE: 2.92) for Kaposi sarcoma, and 45.66 months (SE: 2.30) for other cancers. Figure 1 presents the Kaplan-Meier survival curves for the entire study population and grouped cancer types.

Figure 1.

Kaplan Meier survival plots (restricted to 60 months) show A. overall survival, and B. cancer type specific survival outcomes.

The Log-Rank (Mantel-Cox) test was used to assess the survival disparities among the study groups. The difference in survival between NHL and other cancers was statistically significant (p = 0.01). Additionally, a significant difference in survival was observed between patients with NHL and Kaposi sarcoma (p = 0.002). However, no statistically significant difference was observed between Kaposi sarcoma and other cancers (p = 0.334).

Discussion

This multicentre study provided information on the prevalence, characteristics, and outcomes of cancer patients living with HIV in Turkey. The cancer prevalence rate of 1.8% in our cohort was slightly lower than the 2.6% reported by Altuntaş Aydın et al. in a previous Turkish study of 1872 individuals living with HIV covering the period from 1998 to 2016, but was similar to the rate reported in a U.S. study.^13,14

Koroukian et al. calculated an adjusted cancer prevalence of 1.84% in 82,495 male HIV patients using Medicaid data from the USA. The rate increased to 2.74% when patients were symptomatic.¹⁴ Zhou et al. analysed 72,508 women with HIV using Medicaid data and found that the prevalence of all cancer types was 4.79%.¹⁵

Globally, there is a consistent trend of decreased ADCs and increased NADCs as HIV/AIDS patients are treated with ART.¹⁶ In our study, ADCs (66.9%) were observed twice as frequently as NADCs (33.1%) were. A study by Mendoza-Mori et al. in Peru examined 276 HIV patients with cancer and found that 80% had ADCs, with Kaposi sarcoma and NHL being the most common types. In regions with high HHV-8 co-infection, such as Latin America and Africa, Kaposi sarcoma is the most common ADC malignancy.¹⁷ Atwine et al. conducted a study in Uganda involving 386 patients. The rate of ADCs was 77.5%, with Kaposi sarcoma being the most common (90%), followed by cervical cancer (8%). An excessive rate of ADCs was associated with non-concordance with ART and health follow ups.¹⁸ In a study conducted by Achieng in Uganda on 200 individuals living with HIV and cancer, the ART adherence rate was found to be 55%.¹⁹

The distribution of ADC types in our study aligns with that reported in literature. NHL and Kaposi sarcomas were the most prevalent.⁵ Our study identified several variables that were associated with ADCs. These included younger age (≤54 years), lower CD4 + T lymphocyte count (≤94 cells/mm³), and HHV-8 co-infection. The higher prevalence of ADC in younger patients was an indicator of delayed HIV diagnosis in this demographic group because our patients had access to ART and adherence was ensured. Eight patients had been diagnosed with cancer before 2015. At the time of cancer diagnosis, 61 individuals (39.6%) were ART naive. After cancer diagnosis, only two individuals were not receiving ART, while 83.1% of the population were on two NRTIs plus an INSTI, as recommended since 2015.^20,21 This highlights the importance of screening for HIV and its co-infections.

The distribution of NADCs varies between countries. In Italy, Cattelan et al. found that hepatocellular carcinoma (12.7%) and anal cancer (10.8%) were the most common cancers among 166 cases of NADCs.⁵ In a national study by Tanaka et al. in Japan, the most common types of NADCs were lung (14%), colorectal (11.9%), stomach (10.5%), and liver cancer (8%).²²

A study by Nicolau et al. of 1275 cancer patients living with HIV in Canada between 1997 and 2020 examined the distribution of NADCs by sex. Among men, the most prevalent types of cancer were prostate cancer (10.4%), lung cancer (8%), anal (8.8-9.4%), and colorectal cancer (5.9%). Among women, the most prevalent cancers were breast (18%), lung (10%), thyroid (6.7%), and upper genital tract (6.3%).⁴

According to Shiels et al., the prevalent types of NADCs among 79,657 cancer patients with HIV between 1991 and 2005 were lung cancer, anal cancer, and Hodgkin lymphoma. A total of 4140 cases of lung cancer and 2540 cases of anal cancer were identified. Hodgkin lymphoma was observed in 2004 patients. During 2004-2007, in 15,884 patients, lung cancer was detected in 454 patients and showed a decreasing trend.²³

In our study, the most prevalent type of NADCs was lung cancer. The smoking rate of 60.4% in the study population was also a contributing factor. According to cancer statistics in Turkey, lung cancer is the most common cancer in men, whereas breast cancer is the most common cancer in women.²⁴ The fact that 89% of the study population consisted of male patients indicates that the data obtained were consistent with the cancer distribution in the country.

Our study found that the hepatitis B vaccination rate was low (44.8%), and the HPV vaccination rate was very low (only two patients) in this population. This concerns us due to increased risk of hepatitis B and HPV-related cancers in PLWH. Efforts must be made to improve the vaccination coverage.

The low rate of comorbidities, including cardiovascular diseases, was indicative of a relatively young age in the study cohort. Although our patients were young and had no comorbidities, the mortality rate was 32%. This mortality rate was higher than the 17.5% rate reported by Horner et al. for 31,611 patients with cancer-attributable mortality among PLWH. In the same study, when cancer subtypes were analysed, the highest mortality rate was 4.3% in the NHL.⁷

Among our patients, 92.8% received the appropriate cancer treatment. In Turkey, oncology care is mainly provided through public hospitals under a universal health coverage scheme. However, national pricing and reimbursement policies, along with regional differences, can delay or limit access to some cancer medicines, especially high-cost targeted therapies and immunotherapy.¹⁰ In our cohort of 154 individuals, only six patients received targeted therapies and three received immunotherapies, reflecting this limited access to advanced cancer treatments.

In multivariate Cox regression analyses, mortality was significantly associated with NHL and a low CD4 + T lymphocyte count (≤232 cells/μL) at the end of cancer treatment. This suggests that our patients had received a late HIV diagnosis and had already developed a relatively advanced disease.

Survival analysis showed worse outcomes for non-Hodgkin lymphoma versus Kaposi sarcoma and other cancers. This aligns with studies showing that NHL is particularly aggressive in PLWH.⁷ The 3-years overall survival rate of 67.6% shows that despite advances in HIV and cancer treatment, significant cancer-related mortality persists in this population.

A summary of these key studies, including their study periods, populations, cancer prevalence, ART characteristics, and mortality, is presented in Table 5.

Table 5.

Summary of studies reporting cancer prevalence among people living with HIV.

Author, year of publication	Country	Study period	Total cases of PLWH	Cancer cases	ADC n (%)	NADCs n (%)	ART	Mortality n (%)
Anastasia 2025¹	Italy	1997-2022	846	153	60 (39.2)	93 (60.8)	NA	46 (30.1)
Hernández-ramírez 2017²	USA	1996-2012	448 258	21 294	6384 (30)	14 344 (67.4)	NA	NA
Nicolau 2022⁴	Canada	1997-2020	19 403	1275	386 (30.3)	889 (69.7)	NA	NA
Cattelan 2024⁵	Italy	1996-2023	NA	289	166 (57.4)	120 (41.5)	NA	78 (27.1) (10 years)
Horner 2020⁷	USA	2001-2015	521 623	31 611	12 315 (39)	19 296 (61)	NA	17.5% of deaths preceded by cancer^a
Altuntaş aydın 2020¹³	Turkey	1998-2016	1872	48	35 (72.9)	13 (27.1)	NA	15 (31.3)
Nkwonta 2023¹⁶	USA	2005-2020	11 238	704	250 (35.5)	454 (64.5)	NA	NA
Mendoza-mori 2021¹⁷	Peru	2000-2018	NA	276	223 (80.8)	53 (19.2)	NA	NA
Atwine 2025¹⁸	Uganda	2012-2021	NA	386	299 (77.5)	87 (22.5)	ART, first line n = 362 second line n = 10	26% for ADCs
Achieng 2023¹⁹	Uganda	2018	NA	200	142 (71)	58 (29)	Adherent to ART n = 110	NA
Shiels2011²³	USA	1991-2005	413.080	79 657	1991–95 (34 587) 1996-2000 (13 439) 2001-2005 (10 325)	1991–95 (3193) 1996-2000 (6011) 2001-2005 (10 059)	NA	NA

ADCs: AIDS-defining cancers; NADCs: non-AIDS-defining cancers; ART: Antiretroviral therapy; NA: Not available; PLWH: People living with HIV.

^aData reflect cancer-attributable mortality.

The limitations of our study include that due to the retrospective nature of this study and irregular attendance of patients for follow-up, as well as the variability of hospital registration systems, some patient data were not available. In addition, because of the limited size of the cohort and the fact that only eight patients were diagnosed with cancer before 2015, it was not statistically feasible to stratify or adjust by calendar year or era of diagnosis. This reduced the ability to assess potential temporal trends. Another limitation of this study is the lack of systematically collected data on injection drug use (PWID). Owing to the sensitive nature of this behaviour and related regulatory and ethical constraints, this information was not obtained, which may limit the generalisability of the findings to PWID populations. Moreover, because one-third of the cohort was ART-naive and ART duration data were unavailable, an optimal analysis of ART duration and immune reconstitution was not possible.

In conclusion, this multicentre study provides valuable data on the epidemiology and outcomes of PLWH in Turkey. The findings of this study, including high ADCS and mortality rates, offer valuable insights for healthcare providers and researchers. This study underscores the potential for reducing the cancer burden through earlier HIV diagnosis, improved vaccination coverage, and access to antiretroviral therapy.

Supplemental Material

Supplemental material - Cancer prevalence and prognosis among individuals living with HIV and AIDS in Turkey: Multicenter BUHASDER study

Supplemental material for Cancer prevalence and prognosis among individuals living with HIV and AIDS in Turkey: Multicenter BUHASDER study by Özlem Güler, Sıla Akhan, Atahan Çağatay, Ayşe Batirel, Elif Sargin Altunok, Sibel Bolukçu, Hülya Özkan Özdemir, Emine Ilay duman, Müge toygar Deniz, Derya Özyiğitoğlu, Kemalettin Özden, Handan Alay, Derya Seyman, Melda Türken, Şebnem Çalik, Özge Çaydaşi, Derya Öztürk Engin, Özgür Günal, Şeyma Topal, Oğuz Karabay, Figen Sarigül, Hanife Nur karakoç parlayan, Ayşin Kilinç toker, Özlem Aydin, Emre Bayhan, Pınar Ergen, Nevin İnce, Şafak Kaya, Selcan Arslan Özel, Aybegüm Özşahin, Esra Gürbüz, Ahmet Şahin, Sibel Balci, Şükran Köse in International Journal of STD & AIDS.

Footnotes

Acknowledgements

We would like to thank BUHASDER, a non-profit civil society organisation, for enabling access to health centres and participation in the study.

ORCID iDs

Özlem Güler

Sibel Bolukçu

Müge Toygar Deniz

Hanife Nur Karakoç Parlayan

Ethical considerations

The authors declare that the work described has been carried out in accordance with the Declaration of Helsinki of the World Medical Association revised in 2013 for experiments involving humans as well as with the protocols of the Ethics Committee of Kocaeli University Faculty of Medicine. The project was assigned number 2024/38 and received approval under the code GOKAEK-2024/04.06.

Consent to participate

The authors declare that this report does not contain any personal information that could lead to the identification of the patients and/or volunteers.

Author contributions

Ö. Güler and Ş. Köse conceptualised the study methodology. Ö. Güler wrote the original draft of this manuscript. S. Balcı analysed data. S. Akhan and Ş. Köse reviewed the manuscript as mentor. Other authors have curated data from patient files. All authors have read and approved the final manuscript.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Kocaeli University has an open access agreement with Sage for article processing charges.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

The data will be shared via email upon request.*

Clinical trial registration

NCT06774495, 04 December 2024.

Supplemental Material

Supplemental material for this article is available online.

References

Anastasia

Mazzucco

Pipitò

, et al. Malignancies in people living with HIV: a 25-years observational study from a tertiary hospital in Italy. J Infect Public Health 2025; 18: 102661.

Hernández-Ramírez

Shiels

Dubrow

, et al. Spectrum of cancer risk among HIV-Infected people in the United States during the modern antiretroviral therapy era: a population-based registry linkage study. Lancet HIV 2017; 4: e495–e504.

Hessol

Whittemore

Vittinghoff

, et al. Incidence of first and second primary cancers diagnosed among people with HIV, 1985–2013: a population-based, registry linkage study. Lancet HIV 2018; 5: e647–e655.

Nicolau

Antoniou

Brooks

, et al. The burden of cancer among people living with HIV in Ontario, Canada, 1997–2020: a retrospective population-based cohort study using administrative health data. CMAJ Open 2022; 10: E666–E674.

Cattelan

Mazzitelli

Presa

, et al. Changing prevalence of AIDS and Non-AIDS-Defining cancers in an incident cohort of people living with HIV over 28 years. Cancers 2024; 16: 70.

Coghill

Pfeiffer

Shiels

, et al. Excess mortality among HIV-Infected individuals with cancer in the United States. Cancer Epidemiol Biomarkers Prev 2017; 26: 1027–1033.

Horner

M-J

Shiels

Pfeiffer

, et al. Deaths attributable to cancer in the US human immunodeficiency virus population during 2001–2015. Clin Infect Dis 2021; 72: e224–e231.

Yuan

Zhou

, et al. Incidence and mortality of non-AIDS-defining cancers among people living with HIV: a systematic review and meta-analysis. eClinicalMedicine. 2022; 52: 101613.

Derin

. Gender and age trends in HIV incidence in Turkey between 1990 and 2021: joinpoint and age–period–cohort analyses. Medicina 2024; 60: 1357.

10.

Vural

Kaskati

Vural

İM

, et al.

Are Turkish pharmaceutical pricing strategies an access barrier to oncology medicines for Türkiye?

Front Pharmacol. 2024; 15: 1364341.

11.

von Elm

Altman

Egger

, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370: 1453–1457.

12.

R Core Team . R: a language and environment for statistical computing. : R Foundation for Statistical Computing, 2023. https://www.R-project.org

13.

Aydin

Gunduz

Sargin

, et al. Prevalence and mortality of cancer among people living with HIV and AIDS patients: a large cohort study in Turkey. East Mediterr Health J 2020; 26: 276–282.

14.

Koroukian

Zhou

Navale

, et al. Excess cancer prevalence in men with HIV: a nationwide analysis of medicaid data. Cancer 2022; 128: 1987–1995.

15.

Zhou

Koroukian

Navale

, et al. Cancer burden in women with HIV on medicaid: a nationwide analysis. Womens Health 2023; 19: 17455057231170061.

16.

Nkwonta

Zhang

Chen

, et al. Prevalence and trend of AIDS-Defining cancers and non-AIDS-defining cancers and their association with antiretroviral therapy among people living with HIV in South Carolina: a population-based cohort study. AIDS Care 2023; 35: 753–763.

17.

Mendoza-Mori

Valenzuela-Medina

Gotuzzo

, et al. Cancer in people living with HIV-AIDS at a referral hospital in Lima, Peru. Rev Peru Med Exp Salud Publica 2021; 38: 278–283.

18.

Atwine

Yekosani

Birungi

, et al. Spectrum and trends of cancer among HIV patients in Southwestern Uganda. PLoS One 2025; 20: e0317222.

19.

Achieng

Bunani

Kagaayi

, et al. Adherence to antiretroviral and cancer chemotherapy, and associated factors among patients with HIV–Cancer co-morbidity at the Uganda cancer institute: a cross sectional study. BMC Public Health 2023; 23: 1451.

20.

2015_eacsguidelines_8_0-english_rev-20160124.pdf. https://www.eacsociety.org/media/2015_eacsguidelines_8_0-english_rev-20160124.pdf (accessed 26 September 2025).

21.

EACS Guidelines . EACS guidelines. https://eacs.sanfordguide.com (2024), accessed 26 September 2025.

22.

Tanaka

Oshima

Kawano

, et al. Nationwide longitudinal annual survey of HIV/AIDS referral hospitals in Japan from 1999 to 2021: trend in Non–AIDS-defining cancers among individuals infected with HIV-1. J Acquir Immune Defic Syndr 2024; 96: 1–10.

23.

Shiels

Pfeiffer

Gail

, et al. Cancer burden in the HIV-infected population in the United States. J Natl Cancer Inst 2011; 103: 753–762.

24.

Kanser İstatistikleri. https://hsgm.saglik.gov.tr/tr/kanser-istatistikleri (accessed 29 March 2025).

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.26 MB