At the end of 2020 there were an estimated 38.4 million people living with human immunodeficiency
virus (HIV) worldwide, the majority in low- and middle-income countries.[1] Widespread implementation of effective antiretroviral therapies (ARTs) has transformed
the natural history of HIV such that affected individuals now have a life expectancy
approaching that of the general population. Despite highly effective ART patients
living with HIV remain at increased risk of arterial vascular disease,[2] possibly mediated by chronic inflammation.[3] It is unclear whether they also remain at increased risk of venous thromboembolism
(VTE).[4]
We searched PubMed for observational and randomized studies published since January
1, 2000, involving patients with HIV that reported event rates for VTE, including
deep vein thrombosis (DVT) and/or pulmonary embolism (PE). The Supplementary Material
details the search strategy ([Supplementary Table S1], online only) and the process of study selection ([Supplementary Fig. S1], online only). All 18 studies identified in our search reported VTE, whereas 12
separately reported DVT and 11 separately reported PE. Only 10 of the 18 studies reported
the mean (or median) follow-up period, and use of ART was not consistently reported.
The mean age of patients enrolled in these studies was between 33.5 and 59 years.
In pooled analyses, the crude incidence rates were: VTE 1.77% (interquartile range
[IQR]: 1.40–2.14), DVT 1.44% (IQR: 0.98–1.89), and PE 0.43% (IQR: 0.21–0.64) ([Table 1]). For studies that reported the duration of follow-up (10 studies, 28,139 patients),
the pooled incidence rate for VTE per 1,000 person-years was 2.8 (IQR: 2.5–3.0) ([Supplementary Table S2], online only).
Table 1
Crude incidence rates for venous thromboembolism, deep vein thrombosis, and pulmonary
embolism[a]
|
Study
|
Mean Age[b]
|
Venous thromboembolism
|
Deep vein thrombosis
|
Pulmonary embolism
|
|
Patients
|
Events
|
Crude incidence (%)
|
N
|
Events
|
Crude incidence (%)
|
N
|
PE
|
Crude incidence (%)
|
|
Saif et al 2001[6]
|
38.3
|
131
|
10
|
7.63
|
131
|
6
|
4.58
|
131
|
2
|
1.53
|
|
Saber et al 2001[7]
|
43[c]
|
4,752
|
36
|
0.76
|
4,752
|
33
|
0.69
|
4,752
|
3
|
0.06
|
|
Copur et al 2002[8]
|
–
|
362
|
14
|
3.87
|
362
|
9
|
2.49
|
362
|
5
|
1.38
|
|
Fultz et al 2004[9]
|
46.7
|
37,535
|
745
|
1.98
|
–
|
–
|
–
|
–
|
–
|
–
|
|
Majluf-Cruz et al 2004[10]
|
38.3
|
1,550
|
34
|
2.19
|
1,550
|
31
|
2.00
|
1,550
|
2
|
0.13
|
|
Jacobson et al 2004[11]
|
43
|
650
|
24
|
3.69
|
650
|
14
|
2.15
|
650
|
10
|
1.54
|
|
Lijfering et al 2008[12]
|
41
|
109
|
11
|
10.09
|
109
|
6
|
5.50
|
109
|
5
|
4.59
|
|
Matta et al 2008[13]
|
–
|
2,429,000
|
42,000
|
1.73
|
2,429,000
|
34,000
|
1.40
|
2,429,000
|
10,000
|
0.41
|
|
Jong et al 2010[14]
|
36.4
|
86
|
0
|
0
|
86
|
0
|
0
|
–
|
–
|
–
|
|
Rasmussen et al 2011[15]
|
36.4
|
4,333
|
148
|
3.42
|
–
|
–
|
–
|
–
|
–
|
–
|
|
Arab et al 2017[16]
|
–
|
1,997
|
25
|
1.25
|
–
|
–
|
–
|
–
|
–
|
–
|
|
Borjas-Howard et al 2017[17]
|
35
|
87
|
10
|
11.49
|
–
|
–
|
–
|
–
|
–
|
–
|
|
Howard et al 2019[18]
|
44
|
14,389
|
232
|
1.61
|
14,389
|
99
|
0.69
|
14,389
|
105
|
0.73
|
|
Castilho et al 2019[19]
|
36.9
|
6,206
|
44
|
0.71
|
–
|
–
|
–
|
–
|
–
|
–
|
|
Erbe et al 2003[20]
|
39
|
49
|
6
|
12.24
|
49
|
3
|
6.122
|
49
|
2
|
4.08
|
|
Stellbrink et al 2019[21]
|
33.5
|
657
|
1
|
0.15
|
–
|
–
|
–
|
657
|
1
|
0.15
|
|
Olson et al 2021[22]
|
53
|
110
|
4
|
3.64
|
110
|
3
|
2.727
|
110
|
1
|
0.91
|
|
Zimba et al 2021[23]
|
59
|
58
|
4
|
6.90
|
58
|
4
|
6.897
|
–
|
–
|
–
|
|
Pooled
|
|
2,502,061
|
43,348
|
1.77
(1.40–2.14)
|
2,451,246
|
34,208
|
1.44
(0.98–1.89)
|
2,451,759
|
10,136
|
0.43
(0.21–0.64)
|
Abbreviation: PE, pulmonary embolism.
a References are provided in the Supplementary Material.
b Some studies only reported age ranges.
c Average age reported only for patients with DVT.
Our data have limitations related to the potential for selection and information biases
and confounding. Additionally, estimates of the crude incidence of VTE in patients
with HIV were dominated by a single study that included 2.429 million patients and
accounted for 97% of patients included in our pooled estimates. The pooled rate of
VTE per 1,000 patient-years did not include this study and may be more informative
because it takes account of the risk exposure.
In the general population the incidence rate of VTE is 1 to 2 per 1,000 patient-years,
but is highly age-dependent, ranging from 0.1 per 1,000 patient-years under the age
of 30 to 10 per 1,000 patient-years in those over the age of 80.[4] A Danish nationwide cohort study reported that patients aged 30 to 60 have incidence
rates for VTE ranging from 0.32 to 1.50 events per 1,000 person-years.[5] Our data confirm that even in the era of widespread use of highly active ART, patients
with HIV have a risk of VTE that remains substantially elevated compared with the
general population. For individuals, the risk of VTE will vary according to traditional
risk factors (e.g., inherited hypercoagulable states, hospitalization, surgery) as
well as the severity of HIV (e.g., CD4 count) and disease complications (e.g., Kaposi's
sarcoma, non-Hodgkin lymphoma, tuberculosis) as well as diseases that are more common
in long term survivors of HIV (e.g., cancer). Patients with HIV who have unexplained
chest pain, dyspnea, or hypoxemia should be investigated for PE.
