14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
DOI: 10.1111/hiv.12584
© 2018 British HIV Association HIV Medicine (2018), 19, 355--364
ORIGINAL RESEARCH

The world-wide incidence of Kaposi’s sarcoma in the

HIV/AIDS era
Z Liu ,1,2 Q Fang,1,2 J Zuo,1,2 V Minhas,3 C Wood3 and T Zhang1,2
1
Department of Epidemiology, School of Public Health, Fudan University, Shanghai, China, 2Key Laboratory of Public
Health Safety (Fudan University), Ministry of Education, Shanghai, China and 3Nebraska Center of Virology and the
School of Biological Sciences, University of Nebraska-Lincoln, Lincoln, NE, USA

Objectives
Kaposi’s sarcoma (KS) is a multicentric angioproliferative cancer of endothelial origin typically
occurring in the context of immunosuppression or immunodeficiency. Consequently, KS is one of
the most common cancers in HIV-infected individuals and frequently occurs among transplant
recipients. Nevertheless, its incidence in different populations is not well understood.
Methods
We searched online databases for publications on KS incidence. A random-effect meta-analysis
was performed to combine the KS incidences and incidence rate ratios (IRRs) for associated risk
factors.
Results
Seventy-six eligible studies representing 71 time periods were included. For HIV-infected people, the
overall KS incidence was 481.54 per 100 000 person-years with a 95% confidential interval (CI) of
342.36–677.32 per 100 000 person-years. HIV-infected men who have sex with men (MSM) had the
highest incidence of KS (1397.11 per 100 000 person-years; 95% CI 870.55–2242.18 per 100 000
person-years). The incidence of KS was significantly lower in female than in male individuals (IRR
3.09; 95% CI 1.70–5.62). People receiving highly active antiretroviral therapy (HAART) had a lower
incidence compared with people who had never received HAART (IRR 6.57; 95% CI 1.91–24.69). The
incidence of KS was 68.59 (95% CI 31.39–149.86) per 100 000 person-years in transplant recipients,
52.94 (95% CI 39.90–70.20) per 100 000 person-years in children with HIV infection, and 1.53 (95%
CI 0.33–7.08) per 100 000 person-years in the general population.
Conclusions
Globally, a relatively high incidence of KS was found among HIV-seropositive people and, in
particular, in HIV-infected MSM. The introduction of HAART has largely prevented the
development of KS, but it has not entirely removed the challenge of KS. In Africa, in particular,
KS imposes a very heavy disease burden, which can mainly be attributed to the high prevalence of
KS-associated herpesvirus and poor access to HAART.
Keywords: HIV/AIDS, incidence, Kaposi’s sarcoma, meta-analysis
Accepted 6 December 2017

Introduction characterized by neoangiogenesis, inflammatory infiltra-

Kaposi’s sarcoma (KS), which can be grouped into four epi-
demiological forms, is a multifocal mesenchymal neoplasm
Correspondence: Professor Charles Wood, Nebraska Center of Virology
and the School of Biological Sciences, University of Nebraska-Lincoln,
Lincoln, NE, USA. Tel: 402 472 4550; Fax: 402 472 3323; e-mail: cwoo-
d1@unl.edu or Professor Tiejun Zhang, School of Public Health and The
Key Laboratory of Public Health Safety of Ministry of Education, Fudan
University, Shanghai 200032, China. Tel:/fax: 86 21 54237677;
e-mail: tjzhang@shmu.edu.cn
tion and endothelial-derived, spindle-shaped tumour cells
[1]. Prior to the 1980s, KS had been a relatively rare can-
cer world-wide [2]. Classic KS mainly occurred among
elderly men of Mediterranean or Eastern European Jewish
ancestry [3]. Endemic KS principally existed in parts of
central and eastern Africa [4]. Iatrogenic KS mostly
occurred among immunosuppressed persons of any age
and was caused by autoimmune disease, drugs, or trans-
plantation [5]. From the 1980s onwards, AIDS-associated
KS, a major AIDS-defining malignancy, predominantly

355

356 Z Liu et al.
occurred among HIV-seropositive individuals [6]. The
aetiological agent of KS, known as Kaposi’s sarcoma-
associated herpesvirus (KSHV) or human herpesvirus 8
(HHV8), was identified by Chang et al. in 1994 [7]. The
discovery of KSHV shed new light on the prevention and
treatment of KS and may partly explain the frequent
occurrence of KS in the Mediterranean region and Africa,
where KSHV infection is endemic. Generally, KSHV infec-
tion is necessary but not sufficient for KS development.
Immunosuppression, as a result of coinfection with HIV
or transplantation, is an important factor associated with
the development of KS [8,9].
The introduction of highly active antiretroviral therapy
(HAART) in the 1990s and subsequent continual improve-
ments in its availability, efficacy and tolerability have led
to a substantial decline in the incidence of KS in North
America and Europe [10]. However, in sub-Saharan
Africa, KS is still ranked the first and second most com-
mon cancers in men and women, respectively, and con-
tinues to represent a considerable burden of morbidity
and mortality in people infected with HIV. This is proba-
bly attributable to poor access to, uptake of, and
adherence to HAART [11]. Undoubtedly, KS remains a
nonnegligible public health concern world-wide.
Better understanding of the global KS epidemiology
could benefit the control and prevention of KS. However,
the picture of KS distribution is far from complete and
shows considerable geographical diversity. Therefore, we
conducted a systematic review to synthesize the sparse
data on KS incidence, with the aim of adding to knowl-
edge of the epidemiology of KS world-wide.
Materials and methods
Search strategy and selection criteria
A comprehensive literature search was conducted in
PubMed and Embase, with the restriction of publication
date up to 1 October 2016. The search strategy is detailed in
Box S1. Further studies were identified from the reference
lists, related articles and citation lists of each of the papers
identified in the initial searches. The initial search yielded
1092 relevant publications. All references were exported
and managed using ENDNOTE X7 (Clarivate Analytics, Boston,
MA, USA). Two of the authors (ZL and QF) screened the
abstracts and titles of all references independently.
Inclusion criteria were as follows: (1) a prospective or
retrospective cohort study design; (2) follow-up described
sufficiently to enable calculation of total person-years of
observation; (3) number of study participants ≥ 50 in each
cohort. Exclusion criteria were as follows: (1) participants
© 2018 British HIV Association
14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
who had been diagnosed with KS at baseline; (2) time per-
iod not defined or defined but < 1 year; (3) pooled studies
or synthesized studies from several cohort studies; (4) con-
ferences proceedings. When multiple studies reporting
data from the same cohort study were encountered, the
most comprehensive data were utilized. In cases where the
articles reported on different follow-up periods or sub-
groups, all data from each study were collected. A flow
chart for inclusion of studies is presented in Fig. S1.
Data extraction
Initially, 196 relevant references were identified by title
and abstract screening, with strict application of the
inclusion and exclusion criteria. Subsequently, two authors
(ZL and JZ) conducted the quality assessment and data
extraction independently by reviewing the full text. Any
disagreement was resolved by discussion with another
author (TZ). Seventy-six publications were finally included
for further analyses.
Demographic data were recorded, including the gender,
ethnicity, age and HIV-infected status of participants. We
categorized the participants’ ethnicity as Caucasian, Afri-
can, Asian and other. Data from studies that did not
report the specific races of participants were not used for
calculation of the incidence rate ratio (IRR). Age was cat-
egorized as ≥ 15 and < 15 years. Participants we targeted
in this meta-analysis were HIV-infected individuals,
transplant recipients, and the general population. Among
the participants infected with HIV, data for the specific
subgroups men who have sex with men (MSM) and
injecting drug users (IDUs) were extracted. For each time
period, we extracted the start and end year, and calcu-
lated the mid-calendar year. Moreover, incident cases of
KS during follow-up and the total follow-up time were
collected. If the total number of person-years of follow-
up was not reported, it was calculated by multiplying the
number of participants under follow-up by the mean or
median duration of follow-up.
Statistical analysis
For each study, we computed the crude incidence per
100 000 person-years by dividing the number of incident
cases by total person-years. When the count of cases of
KS was zero, a correction of 0.5 was added to the number
of cases and person-years of follow-up prior to calcula-
tion. The between-study heterogeneity was assessed using
the Cochrane Q statistic and I2 value. A random-effect
model was applied to calculate a pooled estimate of the
overall KS incidence if between-study heterogeneity was
HIV Medicine (2018), 19, 355--364
 14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Global incidence of Kaposi’s sarcoma 357

significant; otherwise, a fixed-effect random model was
applied. The pooled estimates of KS incidence are pre-
sented in forest plots. In addition, subgroup analysis was
conducted with the data stratified by ethnic group, gen-
der, publication year and participant type. In addition,
IRRs with 95% CIs were calculated to measure the associ-
ation between different subgroups. Random-effect meta-
regression using the DerSimonian and Laird method [12]
was performed to identify the sources of heterogeneity.
Publication biases were investigated by examining funnel
plots and by using Egger’s test which applied a linear
regression approach to measure funnel plot asymmetry
[13]. A sensitivity analysis was also conducted to investi-
gate whether there was an excessive influence of each
study on overall estimates. The time trend of KS inci-
dence was analysed at the mid-year of each study period
and was expressed as the percentage change of the crude
incidence rate per calendar year. All studies were grouped
into three categories in terms of their study period [(1)
study ended before 1996; (2) study started in 1996 or
later; (3) study started before and ended in or after 1996]
to calculate the annual change in crude incidence by per-
forming a linear regression, weighted by the sample
counts of each study, between mid-year and log crude
incidence. For all tests, P < 0.05 was deemed to be signif-
icant. All statistical analyses were carried out in R version
3.1 (R core team, https://r-project.org) with the meta and
metafor packages.
Results
Study characteristics
Seventy-six articles with data on 151 184 284 partici-
pants from 26 countries, who were observed for a total of
1 087 688 647 person-years, were included in this meta-
analysis (Table 1; for further details, see Table S1). The
studies were conducted in 71 different periods, ranging
from 1967 to 2013; the mid-year of the study period ran-
ged from 1979 to 2010. Fifty-two studies reported data
on HIV-positive participants, 15 studies on transplant
recipients, and 11 studies on participants from general
populations based on local health registries. All studies
had a cohort design, including 37 prospective and 39 ret-
rospective studies (Fig. S5).
Main findings
KS incidence in HIV-infected populations
A total of 1 616 254 HIV-positive individuals contribut-
ing 12 343 842 person-years were included in this study,
among whom 30 219 cases of KS occurred. Overall, the
crude KS incidence was 481.54 per 100 000 person-years
(95% CI 342.36–677.32 per 100 000 person-years)
(Fig. 1). However, the incidence of KS varied from 18.15
to 6899.81 per 100 000 person-years across the studies
(Table 2). HIV-positive heterosexual men had a KS inci-
dence of 655.09 (95% CI 429.34–999.54) per 100 000
person-years, which was significantly higher than that of
female participants (172.18; 95% CI 125.17–236.84 per
100 000 person-years); the IRR between male and female
participants was 3.09 (95% CI 1.70–5.62) (Figs 2, S2 and
S3). The KS incidence was as high as 1397.11 (95% CI
870.55–2242.18) per 100 000 person-years among HIV-
infected MSM (Fig. S4). The IRR between MSM and non-
MSM was 3.60 (95% CI 2.94–4.41). The KS incidence
among those who received HAART was significantly
lower compared with the incidence in those who had
never received HAART [180.67 (95% CI 88.67–368.11) vs.
1271.86 (95% CI 784.89–2060.94) per 100 000 person-

Table 1 The characteristics of the included studies

Asia Europe North America Oceania Africa

n (%) n (%) n (%) n (%) n (%) Total

All studies 5 (6.6) 32 (42.1) 28 (36.8) 3 (3.9) 8 (10.5) 76

Study design
Prospective 0 (0) 18 (48.6) 16 (43.2) 0 (0.0) 3 (8.1) 37
Retrospective 5 (12.8) 14 (35.9) 12 (30.8) 3 (7.7) 5 (12.8) 39
Study period
Before 1996 1 (10.0) 5 (50.0) 4 (40.0) 0 (0.0) 0 (0.0) 10
After 1996 1 (5.6) 6 (33.3) 5 (27.8) 0 (0.0) 6 (33.3) 18
Spanned 1996 3 (6.3) 21 (43.8) 19 (39.6) 3 (6.3) 2 (4.2) 48
Participant type
HIV positive 1 (2.0) 19 (38.0) 21 (42.0) 2 (4.0) 7 (14.0) 50
General population 2 (16.7) 6 (58.3) 3 (25.0) 0 (0.0) 0 (0.0) 12
Transplant recipient 2 (13.3) 7 (46.7) 4 (26.7) 1 (6.7) 1 (6.7) 15
Age group
Adults 5 (6.8) 31 (42.5) 27 (37.0) 3 (4.1) 6 (9.6) 73
Children 0 (0) 1 (25.0) 1 (25.0) 0 (0.0) 2 (50.0) 4

© 2018 British HIV Association HIV Medicine (2018), 19, 355--364

14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

HIV Medicine (2018), 19, 355--364

Fig. 1 The overall incidence of Kaposi’s sarcoma (KS) in people infected with HIV, obtained from the random-effects model. CI, confidence
interval; IRR, incidence rate ratio; W, Weight.

© 2018 British HIV Association

Z Liu et al.
358
 14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Global incidence of Kaposi’s sarcoma 359

Table 2 The pooled Kaposi’s sarcoma (KS) incidence stratified by 52.94 (95% CI 39.90–70.20) per 100 000 person-years
different explanatory variables (Fig. S11).
KS incidence* 95% CI*
KS incidence in the general population
Region We next focused on HIV-free populations. More than
Africa 318.70 224.62, 452.19
Asia 291.40 252.97, 335.06
140 million HIV-uninfected individuals had been fol-
Europe 589.84 320.37, 1085.96 lowed up, retrospectively or prospectively, in 13 different
North America 695.04 451.14, 1070.80 studies, including 11 studies using data derived from
Oceania 590.04 432.65, 804.70
Study design
health registries. The overall incidence of KS was 1.53
Prospective 519.99 311.32, 868.50 (95% CI 0.33–7.08) per 100 000 person-years among the
Retrospective 436.99 269.84, 707.68 HIV-negative general population (Fig. S9).
Participant type
HIV positive 481.54 342.36, 677.32
General population 1.53 0.33, 7.08 KS incidence in transplant recipients
Transplant recipient 68.59 31.39, 149.86 Fifteen studies, including 416 243 people and contribut-
Gender†
Female 172.18 125.17, 236.84
ing 1 824 647 person-years, reported data on KS inci-
Male, non-MSM 655.09 429.34, 999.54 dence among transplant recipients. In total, 553 patients
MSM 1397.11 870.55, 2242.18 developed KS. The overall incidence of KS was 68.59
Age group
Adults 539.32 389.78, 865.43
(95% CI 31.39–149.86) per 100 000 person-years (Fig. 3).
Children 52.94 39.90, 70.20 Transplantations of four organs were listed, including the
Ethnicity kidney, liver, heart and lung. The KS incidence was 95.79
Asian 19.71 0.48, 816.96
African 341.88 154.65, 755.80
(95% CI 42.81–214.31), 44.25 (95% CI 4.78–409.20),
Caucasian 219.48 71.10, 677.52 49.25 (95% CI 2.48–977.84) and 10.97 (95% CI 4.12–
Other 2090.00 1166.52, 3744.54 29.23) per 100 000 person-years, respectively (Fig. S15).
CI, confidence interval; MSM, men who have sex with men. Only two studies provided data for comparisons among
*Per 100 000 person-years.
†
individuals with different organ transplantations, and no
HIV-positive only.
significant IRR was detected.

years, respectively] (Fig. S6). The IRR was 6.87 (95% CI
1.91–24.69). We further separated the patients according
to their therapy status (receiving HAART for ≥ 6 months
and < 6 months). The crude KS incidence for these two
groups was 377.96 (95% CI 274.48–520.64) and 192.90
(95% CI 115.94–320.95) per 100 000 person-years,
respectively. The IRR was 1.72 (95% CI 1.55–1.92). For
people who had a CD4 count > 200 cells/lL at baseline,
the KS incidence was 384.31 (95% CI 364.97–404.66) per
100 000 person-years, which was much lower than that
of people who had a CD4 count < 200 cells/lL (2050.28;
95% CI 1995.71–2106.35 per 100 000 person-years)
(Fig. S7). The IRR was 3.79 (95% CI 1.79–8.02). The inci-
dence for people who had a CD4 count > 350 cells/lL
was 127.51 (95% CI 120.87–134.81) per 100 000 person-
years, and the incidence for those with a CD4 count
< 350 cells/lL was 891.83 (95% CI 860.06–924.77) per
100 000 person-years; the IRR was 2.39 (95% CI 1.07–
5.36).
Five studies reported data on HIV-positive IDUs. The
crude KS incidence was 115.53 (95% CI 64.92–205.54)
per 100 000 person-years among this population
(Fig. S8). Meanwhile, only four studies reported the KS
incidence among HIV-infected children. The pooled esti-
mate of the KS incidence for HIV-positive children was
KS incidence across ethnicities
Forty-two papers provided original data on participants’
ethnicity. Four race categories were used in the data
extraction: Caucasian, African, Asian and other. The KS
incidence for the race categories listed above was 219.48
(95% CI 71.10–677.52), 341.88 (95% CI 154.65–755.80),
19.71 (95% CI 0.48–816.96) and 2090.00 (95% CI
1166.52–3744.54) per 100 000 person-years, respectively
(Fig. S10). Only nine studies provided data for further IRR
calculation. The IRR between the Caucasian and African
groups was 1.62 (95% CI 0.93–2.83); the IRR between the
Caucasian and ‘other’ groups was 1.29 (95% CI 1.17–1.43).
The IRR between the Caucasian and Asian groups could
not be calculated because of the scarcity of original data.
KS incidence across geographical regions
Eight studies were carried out in Africa, 32 in Europe, 28
in North America, five in Asia and three in Oceania. The
crude overall KS incidence for each continent was 318.70
(95% CI 224.62–452.19), 589.84 (95% CI 320.37–
1085.96), 695.04 (95% CI 451.14–1070.80), 291.14 (95%
CI 252.97–335.06), and 590.04 (95% CI 432.65–804.70)
per 100 000 person-years, respectively (Fig. S12). No
study in Africa was conducted before 1996, but half of
the studies on children were conducted in Africa. When

© 2018 British HIV Association HIV Medicine (2018), 19, 355--364


360 Z Liu et al.
Fig. 2 The crude Kaposi’s sarcoma (KS) incidence and incidence rate ratios among different subgroups.
the study start time and participants were confined to
post-1996 and adults, respectively, the KS incidence in
Africa increased to 343.53 (95% CI 227.95–517.71) per
100 000 person-years, while that in Europe and North
America decreased to 241.13 (95% CI 182.97–315.26) and
171.86 (95% CI 104.00–284.00) per 100 000 person-
years, respectively.
Although Europe had a lower KS incidence than Africa,
the Mediterranean area was an exception. Eleven studies
were conducted in the Mediterranean region, including
10 in Italy and one in Greece. In this region, the overall
KS incidence in HIV-infected people, in people free of
HIV infection and in those who had had a transplant was
524.10 (95% CI 445.91–615.99), 5.30 (95% CI 1.65–17.06)
and 146.06 (95% CI 101.56–210.05) per 100 000 person-
years, respectively.
Time trends of KS incidence
To observe time trends of KS incidence, we concentrated
the analysis on prospective cohort studies. Generally, the
KS incidence showed a significantly decreasing trend,
© 2018 British HIV Association
14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
with annual percentage changes of 12.3% (P < 0.001)
(Fig. 4). The time trend of KS incidence varied across dif-
ferent periods. The pooled KS incidence increased by
10.9% annually for those studies conducted before 1996
(P = 0.007). However, for studies conducted after 1996,
the pooled KS incidence decreased by 3.6% annually
(P = 0.029). A decreasing trend of KS incidence was also
detected for those studies that spanned the year 1996 (the
annual percentage change was 3.9%), although this was
nonsignificant (P = 0.101).
Publication bias and meta-regression
No significant publication bias was detected among the inc-
luded studies. In addition, we performed a meta-regression
to explore the sources of heterogeneity (Figs S13 and S14).
The variables region, study design, gender, mid-year, eth-
nicity and participant type were included. Finally, those
variables accounted for 44.65% of total heterogeneity. The
meta-regression coefficients and the modelled effect
estimates are shown in Table S2.
HIV Medicine (2018), 19, 355--364
 14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Global incidence of Kaposi’s sarcoma 361

Fig. 3 The overall Kaposi’s sarcoma (KS) incidence among transplant recipients, obtained from the random-effects model.

Fig. 4 The annual change in Kaposi’s sarcoma (KS) incidence derived from studies with different study periods. (Pooled: all included studies;
Post-HAART: study started in 1996 or later; Pre-HAART: studies that commenced and ended before 1996; Spanned 1996: studies that started
before 1996 and ended in or after 1996.) HAART, highly active antiretroviral therapy.

low incidence in Asia, North America and Western Eur-

Discussion
In this meta-analysis, the incidence of KS was found to be
very low in the general population. However, the incidence
was found to have risen sharply among transplant recipi-
ents and HIV-infected individuals. Remarkably, the KS
incidence varied across the world. KS had a comparatively
ope, but a high incidence in parts of Africa and the
Mediterranean region. HIV-positive patients with a higher
CD4 cell count had a lower risk of KS than patients with
a lower CD4 cell count. Moreover, the findings suggest
that HAART has had a significant positive effect in
reducing the occurrence of KS. Those who had received

© 2018 British HIV Association HIV Medicine (2018), 19, 355--364


362 Z Liu et al.
HAART for > 6 months had a lower incidence compared
with people who had received HAART for < 6 months,
which is consistent with the previous result integrated 21
European cohort studies [14]. MSM, a population with a
disproportionate burden of KSHV and HIV infection, had
a KS incidence of 1397.11 per 100 000 person-years in
this study [15–17]. Moreover, non-MSM male participants
showed a higher susceptibility to KS than female partici-
pants. This may be mainly explained by the higher sero-
prevalence of KSHV infection in male than in female
individuals [18]. Children seem to have a lower risk of KS
development regardless of HIV status. The crude inci-
dence was 52.94 (95% CI 39.90–70.20) per 100 000 per-
son-years for HIV-infected children, which is similar to
the incidence found in another collaborative study [19].
However, the incidence of KS in children continues to
increase despite the drive to make HAART available
nationwide in South Africa [20]. The risk of KS among
people of different ethnic origins is a subject of debate;
Jones et al. [21] found that white men had a higher KS
incidence than black men, while the opposite result was
reported in another study [22]. In the present study, no
significant disparity was found between Caucasian and
African participants, while comparisons between African
participants and those of other races were not conducted
because of the scarcity of data.
To the best of our knowledge, KS was rare before the
pandemic of HIV/AIDS. In 1981, an unusual cluster of KS
cases in gay men from New York City and California was
a harbinger of the outbreak of AIDS [23]. Subsequently,
the incidence of AIDS-associated KS, the most aggressive
form [6], rose dramatically in conjunction with the rise of
the AIDS epidemic. It is speculated that a synergistic
interaction between HIV and KSHV facilitates KSHV
replication, which may lead to an increased risk of KS
[24–26]. Similarly, KSHV has been suggested to activate
HIV replication [27].
With the advent of the HAART era, a considerable
decline in KS incidence has been witnessed in developed
Western countries. In this meta-analysis, we found an
annual decrease of 3.6% in KS incidence after the introduc-
tion of HAART world-wide. Nonetheless, a huge morbidity
and mortality burden is still imposed by KS [28,29]. KS
remains the second most frequent malignancy in HIV-
infected patients world-wide and it has become the most
common cancer in sub-Saharan Africa [11,30], where KS is
definitely a nonnegligible health concern, as a result of the
highest prevalence of both HIV and KSHV and the poorest
access to HAART of all regions of the world [31].
In addition, previous reports showed that transplant recip-
ients were much more susceptible to post-transplantation
cancers, including KS, than the general population [32–34].
© 2018 British HIV Association
14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
In the present study, a significantly elevated incidence of KS
among transplant recipients (68.59 per 100 000 person-
years) was observed compared with the general population.
This result indicates a high risk of post-transplant KS among
transplant recipients as a consequence of immunosuppres-
sive therapy. Such patients therefore need to receive close
clinical follow-up to prevent KS occurrence as well as
alleviate the negative impact of KS on allograft outcome
and survival of transplanted organs [35,36].
A strength of the present study is the comprehensive-
ness of the literature search and evaluation of data for
inclusion, but several limitations of this meta-analysis
should be mentioned. First, the differential diagnosis of
KS was not well documented in all included studies,
which may have introduced bias to the overall estimates.
Secondly, age was previously described to have an influ-
ence on the risk of KS development [37–39], but the age
did not consider in this study as a consequence of the
shortage of original data. Thirdly, incidence data from
South America were not available.
In conclusion, KS was found to occur frequently
among HIV-infected individuals, especially MSM. The
prevention of KSHV and HIV infection and the utilization
of HAART may be crucial for prevention of KS. However,
for patients with KS, there seems to be no curative ther-
apy available to prevent devastating progression of the
disease. Therefore, vaccines and much more potent treat-
ments are urgently needed in the near future.
Acknowledgements
This study was funded by the Natural Science Foundation
of China (Grant No. 81772170 to TZ) and the Doctoral
Fund of the Ministry of Education of China (Grant No.
20120071120050 to TZ).
Conflicts of interest: The authors declare that they have
no potential conflicts of interest.
Author contributions
ZT and CW contributed to the study design; LZ and FQ
contributed to the data collection; LZ and ZJ contributed
to the data analyses and interpretation of results; LZ, ZJ
and VM contributed to the writing of the manuscript.
References
1 Mesri EA, Cesarman E, Boshoff C. Kaposi’s sarcoma and its
associated herpesvirus. Nat Rev Cancer 2010; 10: 707–719.
2 Rebecca R, Mark B. Kaposi sarcoma herpesvirus. In: Bhushan
K, Somesh G, eds. Sexually Transmitted Infection. 2nd ed.
Delhi, Elsevier, 2012. pp. 418–428.
HIV Medicine (2018), 19, 355--364

3 Vano-Galvan S, Moreno C. Classic Kaposi sarcoma. CMAJ
2011; 183: E1040.
4 Gallo RC. The enigmas of Kaposi’s sarcoma. Science 1998;
282: 1837–1839.
5 Radu O, Pantanowitz L. Kaposi sarcoma. Arch Pathol Lab
Med 2013; 137: 289–294.
6 Robey RC, Bower M. Facing up to the ongoing challenge of
Kaposi’s sarcoma. Curr Opin Infect Dis 2015; 28: 31–40.
7 Chang Y, Cesarman E, Pessin MS et al. Identification of
herpesvirus-like DNA sequences in AIDS-associated Kaposi’s
sarcoma. Science 1994; 266: 1865–1869.
8 Barozzi P, Luppi M, Facchetti F et al. Post-transplant Kaposi
sarcoma originates from the seeding of donor-derived
progenitors. Nat Med 2003; 9: 554–561.
9 Siegel JH, Janis R, Alper JC, Schutte H, Robbins L, Blaufox
MD. Disseminated visceral Kaposi’s sarcoma. Appearance
after human renal homograft operation. JAMA 1969; 207:
1493–1496.
10 Rubinstein PG, Aboulafia DM, Zloza A. Malignancies in HIV/
AIDS: from epidemiology to therapeutic challenges. Aids
2014; 28: 453–465.
11 Parkin DM, Sitas F, Chirenje M, Stein L, Abratt R, Wabinga
H. Part I: cancer in indigenous Africans–burden, distribution,
and trends. Lancet Oncol 2008; 9: 683–692.
12 DerSimonian R, Kacker R. Random-effects model for meta-
analysis of clinical trials: an update. Contemp Clin Trials
2007; 28: 105–114.
13 Egger M, Davey Smith G, Schneider M, Minder C. Bias in
meta-analysis detected by a simple, graphical test. BMJ
1997; 315: 629–634.
14 EuroCoord CPWGftCoOHERECsi. Changing incidence and risk
factors for kaposi sarcoma by time since starting
antiretroviral therapy: collaborative analysis of 21 European
Cohort Studies. Clin Infect Dis 2016; 63: 1373–1379.
15 Liu Z, Fang Q, Zuo J et al. High seroprevalence of human
herpesvirus 8 and herpes simplex virus 2 infections in men
who have sex with men in Shanghai, China. J Med Virol
2017; 89: 887–894.
16 Dukers NH, Renwick N, Prins M et al. Risk factors for human
herpesvirus 8 seropositivity and seroconversion in a cohort
of homosexual men. Am J Epidemiol 2000; 151: 213–224.
17 Rohner E, Wyss N, Heg Z et al. HIV and human herpesvirus
8 co-infection across the globe: systematic review and meta-
analysis. Int J Cancer 2016; 138: 45–54.
18 Begre L, Rohner E, Mbulaiteye SM, Egger M, Bohlius J. Is
human herpesvirus 8 infection more common in men than in
women? Systematic review and meta-analysis. Int J Cancer
2016; 139: 776–783.
19 Pediatric AIDS-Defining Cancer Project Working Group for
IeDEA Southern Africa, TApHOD, and COHERE in EuroCoord;
Rohner E, Schmidlin K, Zwahlen M et al. Kaposi sarcoma
risk in HIV-infected children and adolescents on combination
© 2018 British HIV Association
14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Global incidence of Kaposi’s sarcoma 363
antiretroviral therapy from sub-Saharan Africa, Europe, and
Asia. Clin Infect Dis 2016; 63: 1245–1253.
20 Davidson A, Wainwright RD, Stones DK et al. Malignancies
in South African children with HIV. J Pediatr Hematol Oncol
2014; 36: 111–117.
21 Jones JL, Hanson DL, Dworkin MS, Jaffe HW. Incidence and
trends in Kaposi’s sarcoma in the era of effective antiretroviral
therapy. J Acquir Immune Defic Syndr 2000; 24: 270–274.
22 Hsieh MC, Wu XC, Andrews PA, Chen VW. Racial and ethnic
disparities in the incidence and trends of soft tissue sarcoma
among adolescents and young adults in the United States,
1995-2008. J Adolesc Young Adult Oncol 2013; 2: 89–94.
23 Centers for Disease Control (CDC). Kaposi’s sarcoma and
Pneumocystis pneumonia among homosexual men–New York
City and California. MMWR Morb Mortal Wkly Rep 1981;
30: 305–308.
24 Campbell TB, Borok M, White IE et al. Relationship of Kaposi
sarcoma (KS)-associated herpesvirus viremia and KS disease
in Zimbabwe. Clin Infect Dis 2003; 36: 1144–1151.
25 Ensoli B, Gendelman R, Markham P et al. Synergy between
basic fibroblast growth factor and HIV-1 Tat protein in
induction of Kaposi’s sarcoma. Nature 1994; 371: 674–680.
26 Ensoli B, Sturzl M, Monini P. Reactivation and role of HHV-
8 in Kaposi’s sarcoma initiation. Adv Cancer Res 2001; 81:
161–200.
27 Caselli E, Galvan M, Cassai E, Caruso A, Sighinolfi L, Di Luca D.
Human herpesvirus 8 enhances human immunodeficiency virus
replication in acutely infected cells and induces reactivation in
latently infected cells. Blood 2005; 106: 2790–2797.
28 Calabresi A, Ferraresi A, Festa A et al. Incidence of AIDS-
defining cancers and virus-related and non-virus-related non-
AIDS-defining cancers among HIV-infected patients compared
with the general population in a large health district of
Northern Italy, 1999-2009. HIV Med 2013; 14: 481–490.
29 Hleyhel M, Belot A, Bouvier AM et al. Risk of AIDS-defining
cancers among HIV-1-infected patients in France between
1992 and 2009: results from the FHDH-ANRS CO4 cohort.
Clin Infect Dis 2013; 57: 1638–1647.
30 Wabinga HR, Parkin DM, Wabwire-Mangen F, Mugerwa JW.
Cancer in Kampala, Uganda, in 1989-91: changes in
incidence in the era of AIDS. Int J Cancer 1993; 54: 26–36.
31 Muhula SO, Peter M, Sibhatu B, Meshack N, Lennie K. Effects
of highly active antiretroviral therapy on the survival of
HIV-infected adult patients in urban slums of Kenya. Pan Afr
Med J 2015; 20: 63.
32 Hosseini-Moghaddam SM, Soleimanirahbar A, Mazzulli T,
Rotstein C, Husain S. Post renal transplantation Kaposi’s
sarcoma: a review of its epidemiology, pathogenesis,
diagnosis, clinical aspects, and therapy. Transpl Infect Dis
2012; 14: 338–345.
33 Moosa MR. Kaposi’s sarcoma in kidney transplant recipients:
a 23-year experience. QJM 2005; 98: 205–214.
HIV Medicine (2018), 19, 355--364

364 Z Liu et al.
34 Penn I. Kaposi’s sarcoma in transplant recipients.
Transplantation 1997; 64: 669–673.
35 Aseni P, Vertemati M, Minola E et al. Kaposi’s sarcoma in
liver transplant recipients: morphological and clinical
description. Liver Transpl 2001; 7: 816–823.
36 Mbulaiteye SM, Engels EA. Kaposi’s sarcoma risk among
transplant recipients in the United States (1993-2003). Int J
Cancer 2006; 119: 2685–2691.
37 Luu HN, Amirian ES, Chiao EY, Scheurer ME. Age patterns of
Kaposi’s sarcoma incidence in a cohort of HIV-infected men.
Cancer Med 2014; 3: 1635–1643.
38 Lodi S, Guiguet M, Costagliola D, Fisher M, de Luca A, Porter
K. Kaposi sarcoma incidence and survival among HIV-
infected homosexual men after HIV seroconversion. J Natl
Cancer Inst 2010; 102: 784–792.
39 Yanik EL, Napravnik S, Cole SR et al. Incidence and timing
of cancer in HIV-infected individuals following initiation of
combination antiretroviral therapy. Clin Infect Dis 2013; 57:
756–764.
Supporting Information
Additional Supporting Information may be found in the
online version of this article at the publisher’s web-site:
© 2018 British HIV Association
14681293, 2018, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/hiv.12584 by National Medical Library The Director, Wiley Online Library on [21/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Fig. S1. A flow chart of the literature searches for the
systematic review of studies on the incidence of KS.
Fig. S2. The KS incidence in HIV seropositive male.
Fig. S3. The KS incidence in HIV seropositive female.
Fig. S4. The KS incidence in HIV seropositive MSM.
Fig. S5. The KS incidence stratified by study design.
Fig. S6. The KS incidence stratified by HAART using.
Fig. S7. The KS incidence stratified by CD4 count.
Fig. S8. The KS incidence in IDUs.
Fig. S9. The KS incidence in general population.
Fig. S10. The KS incidence stratified by race.
Fig. S11. The KS incidence in children.
Fig. S12. The KS incidence stratified by continent.
Fig. S13. The funnel plot of all studies focused on HIV
seropositive participants.
Fig. S14. The funnel plot of all studies with prospective
study design focused on HIV seropositive participants.
Fig. S15. The KS incidence in transplant recipients strati-
fied by organ.
Table S1. Basic information of included studies.
Table S2. The meta-regression coefficients and the
modeled effect estimates.
Box S1. Search strategies.
HIV Medicine (2018), 19, 355--364