Dengue among immunocompromised patients: a systematic review and meta-analysis

Abstract

Background

Although there is a rising trend in both dengue cases and immunocompromised conditions, there is limited research on how common severe dengue is in immunocompromised individuals. This data is key for those advising the ever-increasing numbers of immunocompromised travellers.

Methods

We conducted a systematic review and meta-analysis of studies reporting dengue frequency or outcomes in immunocompromised populations. Non-human and review articles were excluded. Risk of bias was assessed using the ROBINS-E tool.

Results

Eighty-five studies were included; 63 had a very high risk of bias. Frequency of dengue among different immunocompromised patient cohorts varied from 0.3% to 6.3%. Of 1182 dengue cases, 664 had autoimmune diseases, 388 were post-solid organ transplant (SOT), 20 post-stem cell transplant (HSCT), 28 had haematological malignancies, 24 non-haematological malignancies and 58 were HIV-positive. Severe dengue and mortality were estimated at 0.27 [0.22–0.33] and 0.14 [0.10–0.18], decreasing to 0.16 [0.09–0.27] and 0.04 [0.03–0.05] when very high risk or small-sample studies were excluded. Twenty-three (5.6%) of post-transplant dengue patients were considered as donor-related. Mortality reached 66.7% in HSCT and 10% in SOT. Dengue RNA was detectable up to four months in blood and up to two years in urine; viable virus was isolated from urine at nine months.

Conclusions

Dengue in immunocompromised, especially HSCT, is associated with high severity and mortality. It also has the potential for prolonged viral persistence.

Introduction

Dengue has increased worldwide, with a 10-fold rise in reported cases from 2000 to 2019,¹ including ~60 million symptomatic cases across 130 countries.² Its global spread is expected to expand,³ further reinforcing dengue as a major public health concern. Dengue is the leading cause of fever in returning travellers from most tropical areas⁴ with a seroconversion rate ranging from <1% to >20% depending on destination, season and duration of travel and activities during travel.⁵ Most infections are asymptomatic or mild. In symptomatic cases, the illness typically progresses through febrile, critical and convalescent phases. According to the World Health Organization (WHO) 2009 criteria,⁶ dengue is classified into three categories: without warning signs, with warning signs and severe dengue. Severe dengue involves shock, fluid accumulation with respiratory distress, severe bleeding and severe organ involvement such as severe hepatitis and is associated with increased mortality.⁷ In travellers, symptomatic dengue with seroconversion is estimated to occur in >1 per 1000 person months of travel.⁸

Pre-existing conditions such as diabetes, chronic kidney disease and cardiovascular disease are known to be risk factors for severe dengue outcomes.^9,10 A recent scoping review suggested that individuals with sickle cell disease had an increased risk of severe disease and fatality.¹¹ Whether immunocompromised status is a risk factor has not been systematically studied. Use of biologics, immunosuppressants and transplants has increased, with immunocompromised prevalence in the USA rising from 3% to 6% over the past decade.¹² International travel among immunocompromised individuals is also rising, with similar patterns to immunocompetent travellers.¹³ Thus, dengue incidence in this population may increase, but current estimates are lacking.

Reports on dengue severity among immunocompromised individuals are inconsistent: some describe severe cases,¹⁴ while others report mild symptoms¹⁵ or even reduced severity.¹⁶ Donor-derived dengue infection has also been reported in organ transplant recipients in endemic countries. Therefore, we conducted a systematic review to examine the prevalence, source, clinical course and severity of dengue infection in immunocompromised individuals.

Methods

This systematic review was conducted and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Supplementary file 1). The study protocol was registered on the PROSPERO platform (CRD42024584266).

The study was guided by the following research questions: ‘How common is dengue infection among immunocompromised individuals?’, ‘What are the modes of acquisition of dengue infection among immunocompromised individuals?’, ‘Is the pattern of clinical manifestation typical i.e., febrile phase, critical phase and recovery phase?’ and ‘What is the severity of dengue infection among immunocompromised individuals?’

Search strategy and study selection

An experienced librarian searched EMBASE and MEDLINE for articles published between 1946 and August 2024 using terms related to the review questions (Supplementary file 1). Eligible studies reported on prevalence, clinical presentation or outcomes of dengue infection in immunocompromised humans. Immunocompromised conditions included: solid organ or haematopoietic stem cell transplant, haematological or active non-haematological malignancy, autoimmune diseases on immunosuppressants and human immunodeficiency virus (HIV) infection. Non-English studies were excluded. Clinical trials, observational studies, cohort, case–control, case series and case reports were included; reviews without original data were excluded. Articles were uploaded to Covidence for independent title and abstract screening by two reviewers (J.C. and V.L.). Full texts were assessed for eligibility; disagreements were resolved by discussion. Reference lists were also reviewed.

Data extraction

Two reviewers (J.C. and V.L.) independently extracted data using a predefined spreadsheet. Variables included participant characteristics, immunocompromising conditions, prevalence, clinical presentation, severity and outcomes.

Outcomes of interest

We were interested in reporting the prevalence, severity, clinical course, source of infection and duration of viremia/viruria. For dengue severity, WHO 2009 classification was applied, the diseases ranged from dengue without warning signs, dengue with warning signs, to severe dengue.⁶ Clinical course was classified into typical and atypical courses. A typical course is characterized by a febrile phase lasting 2–7 days, a critical phase marked by plasma leakage usually around days 3–7 of illness, and a recovery phase in the following 48–72 hours.⁶ Clinical presentation that was not consistent with the described typical course of dengue infection was considered atypical. The source of infection was divided into donor-derived, travelling to an endemic area and unknown.

Study quality assessment

Risk of bias was assessed using the risk of bias in non-randomized follow-up studies of exposure effects (ROBINS-E) tool¹⁷ by two reviewers (J.C. and V.L.).

Data analyses

The extracted data were summarized in tables. Data on prevalence and dengue severity outcomes were pooled for meta-analyses using R software package. Proportion meta-analyses were conducted using a random-effects model. Risk ratios were calculated when comparators were available. Sensitivity analyses excluded studies with very high bias or sample size < 10. Publication bias was assessed with funnel plots. Evidence certainty was graded using the Grading of Recommendations Assessment, Development and Evaluation tool (Supplementary file 1).

Results

The search identified 2515 studies, 593 duplicates were removed, leaving 1922 studies to be screened by title and abstract. Following the first screening, 1837 were excluded and the full text of 83 studies was assessed for eligibility. Of these, nine were excluded, leaving 74 studies. Eleven additional studies were identified during extraction, resulting in 85 included studies (Supplementary Fig. 1).

Study and patient characteristics

Of 85 included studies, 75 (88.2%) studies reported 1116 immunocompromised adults, and 14 (16.5%) studies reported 46 immunocompromised children under 18 (Supplementary Tables 1 and 2, Table 1). Solid organ transplantation (SOT) was the most common condition: kidney transplants were reported in 28 studies, liver in seven and multiple organs in three (total: 38 studies, 388 patients). Autoimmune disease was the second most common (total 19 studies, 664 patients). Eleven studies reported 58 adults living with Human Immunodeficiency Virus (PLWH). Ten and nine studies reported dengue in 20 post-haematopoietic stem cell transplant (HSCT) patients and 28 with haematological malignancy, respectively. Two paediatric studies reported 24 cases with non-haematological malignancy. Three studies (Bahl 2011, de Souza Pereira 2017, Machado 2017) reported dengue in multiple immunocompromised groups.

Table 1.

Study characteristics and risk of bias of all included studies

Study design (study)	Country (study)	Years	Type of immunocompromised (case)	Immunocompromised with dengue (case)	Risk of bias (study)
Solid organ transplantation (n = 38)
Case report or series: 31 Retrospective cohort: 5 Prospective cohort: 2	Asia: 25 South America: 10 Non-endemic: 3	2005–19	Kidney: 372 Liver: 10 Multiple organs: 6	388	Low risk: 1 Some concerns: 7 High risk: 6 Very high risk: 24
Haematopoietic stem cell transplantation (n = 10)
Case report or series: 7 Retrospective cohort: 1 Prospective cohort: 2	Asia: 3 South America: 6 Non-endemic: 1	2001–18	HSCT: 20	20	Low risk: 1 Some concerns: 0 High risk: 0 Very high risk: 9
Haematological malignancy (n = 9)
Case report or series: 7 Retrospective cohort: 1 Prospective cohort: 1	Asia: 6 South America: 2 Non-endemic: 1	2010–16	ALL: 13 AML: 1 APL: 2 Lymphoma: 2 Not report: 10	28	Low risk: 0 Some concerns: 1 High risk: 1 Very high risk: 7
Non-haematological malignancy (n = 2)
Case report or series: 1 Retrospective cohort: 1 Prospective cohort: 0	Asia:2 South America: - Non-endemic: -	2010–13	Primitive neuroectodermal tumour: 3 Not report: 21	24	Low risk: 0 Some concerns: 1 High risk: 0 Very high risk: 1
Autoimmune diseases (n = 19)
Case report or series: 17 Retrospective cohort: 2 Prospective cohort: 0	Asia: 5 South America: 7 Non-endemic: 7	1997–2021	SLE: 77 RA: 561 MS: 17 AS: 4 Others: 5	664	Low risk: 0 Some concerns: 1 High risk: 0 Very high risk: 1
HIV infection (n = 11)
Case report or series: Retrospective cohort: Prospective cohort:	Asia: 5 South America: 5 Non-endemic: 1	2001–22	CD4 < 350: 11 CD4 ≥ 350: 47	58	Low risk: 0 Some concerns: 1 High risk: 0 Very high risk: 1

ALL, acute lymphoblastic leukaemia; APL, acute promyelocytic leukaemia; SLE, systemic lupus erythematosus; RA, rheumatoid arthritis; MS, multiple sclerosis; AS, ankylosing spondylitis.

Full citation of each included study is listed in Appendix D of supplementary file

Most cases were from dengue-endemic countries, but 25 (2.2%) cases were reported from non-endemic countries (e.g. France, Italy, USA) in 12 studies (Marinelli 2024, Lecadieu 2021, Kooijmans 2019, Punzel 2014, Deligny 2014, Park 2008, Strobel 2001, DiAscia 2024, Guerra 2021, Sharp 2014, Kanitkar 2020, Kumar 2010).

The median time from transplant to dengue infection was 14.2 months (range: immediate to 17 years) for SOT, and 1.27 months (range: 2 days to 1.6 years) for HSCT. CD4 counts in 58 PLWH were >350 cells/μl, except in 11 cases (range: 5–303 cells/μl; mean: 134.3 ± 112.9 cells/μL) (Siong 2008, Pang 2015, Espinoza-Gomez 2017, Marinelli 2024).

Of 664 patients with autoimmune disease, 294 had a detailed description of concomitant medications and 199 (67.7%) were on some form of immunosuppression or disease-modifying anti-rheumatic drugs (DMARDs). The largest study with n = 370 reported by de Abreu 2018, concomitant medications were not clearly described but most patients were reported to be on DMARDs.

Study quality assessment

There are four categories of risk of bias; low, some concern, high risk and very high risk. Three prospective studies (Thomas 2019, de Oliviera 2023, Hottz 2019) were classified as low risk of bias. Nine studies were classified as some concern, and 10 were very high risk. Most studies (63, 74.1%) were very high risk based on their nature of case report/case series. Funnel plots indicated moderate to strong evidence of publication bias in the studies reporting dengue with warning signs and mortality, whereas no evidence of publication bias was observed in studies reporting severe dengue (see Supplementary file 1).

Frequency of dengue infection

Ten studies (Batista 2011, Chavan 2015, Deligny 2014, de Oliveira 2023, DiAscia 2024, Fernandes-Charpiot 2019, Jain 2014, Machado 2009, Pinsai 2019, Renaud 2007) reported dengue frequency in SOT (5), HSCT (2), haematological malignancy (1), autoimmune (1) and PLWH (1) populations (see Supplementary file 1). The reported frequency varied widely—from 0.1% in a cohort of 1754 SOT recipients in Brazil between 2001 and 2006 (Batista 2011) to 6.3% in 126 patients with haematological malignancy in India in 2014 (Jain 2014). Notably, variability was observed even within the same subgroup. For example, among the five studies of SOT recipients, dengue frequency ranged from 0.1% (Batista 2011) to 3.5% (Fernandes-Charpiot 2019). Of the 10 studies, 2 were prospective: de Oliveira (2023) reported five cases (5.4%) in 93 HSCT recipients during the period from transplant to engraftment (2017–18, Brazil) and Chavan (2015) reported one case (0.5%) in 185 PLWH in India; however, the study period was not specified.

Source of dengue infection

Most studies did not specify the source. Interestingly, 25 immunocompromised patients were reported from non-endemic countries (e.g. Australia, France, Germany, Italy, Korea, Netherlands, UK and USA). Eight of 25 were patients in Reunion Island, France, which can be considered a dengue endemic area. These cases had early post-transplant dengue infection, and most were donor derived (Lecadieu 2021, DiAscia 2024). Five others acquired infection through travel (Kooijmans 2019, Park 2008, Guerra 2021, Kanitkar 2020, Kumar 2010), and another patient in Germany received stem cells from a donor who had returned from Sri Lanka (Punzel 2014). Of the 11 remaining cases, one was an untreated PLWH with a very low CD4 count of 30 cells/μl, who presented with encephalitis 3 months after having migrated from Nigeria to Australia. Exhaustive investigation was unrevealing. However, the dengue virus was found in the brain tissue by the polymerase chain reaction method (Marinelli 2024). Interestingly, there was a fatal hemophagocytic lymphohistiocytosis (HLH)–associated with locally acquired dengue infection in Texas, USA (Sharp 2014). The source in the remaining nine cases was not reported (Deligny 2014, Strobel 2001).

Donor-derived dengue infection

Fourteen studies reported 23 donor-derived (Table 2 and Supplementary file 1): 3 HSCT and 11 SOT. All HSCT recipients developed severe dengue, and two died. Of 20 SOT recipients, 2 died and 6 (30%) had typical dengue. Both SOT fatalities were liver transplants.

Table 2.

Dengue infection outcomes in donor-derived dengue infection

Type of IC population	Number of cases	Age range	Severe dengue	Mortality
HSCT	3	2–51	3/3, 100%	2/3, 66.7%
SOT Kidney Liver Heart	13 6 1	34–64 19–58 41	3/13, 23.1% 5/6, 83.3% 1/1, 100%	0/13, 0% 2/6, 33.3% 0/1, 0%
Total	23	2–64	12/23, 52.2%	4/23, 17.4%

Dengue clinical outcomes

Of 85 studies, 83 (537 cases) reported dengue outcomes; 92 cases were severe. Pooled severe dengue proportion [95% CI] was 0.27 [0.19–0.37]. This proportion varied across different types of immunocompromising conditions (Table 3 and Supplementary file 1). It was most frequent in HSCT recipients (0.38 [0.16–0.66]) and least in PLWH (0.22 [0.09–0.42]). Excluding very high risk of bias and with small studies, the proportion decreased to (0.16 [0.09–0.27]) (Table 3).

Table 3.

Summary of dengue severity by immunocompromised population according to WHO 2009 definition

Proportion (95% CI)	Dengue with warning signs	Severe dengue infection	Dengue mortality
All IC populations	0.37 [0.32; 0.43]	0.27 [0.22; 0.33]	0.14 [0.10; 0.18]
Autoimmune	0.30 [0.19; 0.45]	0.29 [0.17; 0.45]	0.15 [0.07; 0.31]
Non-haematological cancer	0.51 [0.00; 1.00]	0.32 [0.00; 1.00]	0.10 [0.02; 0.34]
Haematological cancer	0.27 [0.18; 0.38]	0.30 [0.15; 0.51]	0.15 [0.09; 0.25]
HSCT	0.42 [0.21; 0.65]	0.38 [0.16; 0.66]	0.32 [0.15; 0.58]
SOT	0.36 [0.29; 0.44]	0.27 [0.19; 0.37]	0.12 [0.08; 0.18]
HIV	0.40 [0.29; 0.52]	0.22 [0.09; 0.42]	0.12 [0.06; 0.22]
Excluding studies with a very high risk of bias and studies with n < 10
All IC populations	0.42 [0.21; 0.66]	0.16 [0.09; 0.27]	0.04 [0.03; 0.05]
Non-HIV	0.40 [0.11; 0.78]	0.16 [0.10; 0.26]	0.04 [0.03; 0.06]
HIV	0.41 [0.04; 0.92]	0.17 [0.00; 1.00]	0.03 [0.00; 0.95]

Due to limited data on studies without a very high risk of bias, HSCT, haematological and non-haematological malignancy studies were pooled. The proportion of severe dengue in non-HIV vs. HIV population excluding studies with very high risk of bias and studies with n < 10 was 0.16 [0.10–0.26] vs. 0.17 [0.00–1.00] (Table 3 and Supplementary file 1).

Dengue with warning signs: 0.37 [0.32–0.43]; reduced to 0.42 [0.21–0.66] after exclusions. Dengue mortality: 0.14 [0.10–0.18]; reduced to 0.04 [0.03–0.05] (Table 3). Three studies compared severity between immunocompromised and non-immunocompromised groups. Pooled risk ratio: 1.95 [0.51–7.49] (Fig. 1). In PLWH, of the 11 patients with CD4 counts <350 cells/μl, 6 (54.5%) had severe dengue (Marinelli 2024, Pang 2015).

Figure 1.

Meta-analysis comparing risk of severe dengue between immunocompromised and non-immunocompromised populations. Each included study is listed in the left column, along with the number of dengue cases and total sample size in the immunocompromised and the non-immunocompromised. Data are presented as risk ratios (squares) and 95% confidence intervals (error bars). The weight of each study, calculated using the random-effects model, is shown in the rightmost column. The pooled estimate is represented by a diamond, where the width reflects the 95% confidence interval.

Clinical course of dengue infection

Sixty-eight studies reported clinical presentation sufficiently for dengue clinical course evaluation. Forty-two (61.8%) studies reported the typical course of dengue infection. Atypical presentation from the remaining 26 studies involved 30 patients: 9 with underlying autoimmune diseases, 10 post-SOT, 6 post-HSCT, 4 with haematological malignancies and 1 with HIV infection (Supplementary file 1). Atypical dengue presentations included prolonged illness with prolonged viremia (n = 5), encephalitis (n = 3), HLH (n = 3), absence of fever (n = 3), rapid deterioration to critical phase within 1–2 days of fever onset (n = 2), prolonged febrile phase > 7 days (n = 2), prolonged thrombocytopenia (n = 2), profound thrombocytopenia (3 × 109/L) without a change in haematocrit (n = 1), rhabdomyolysis (n = 1), hepatic veno-occlusive disease (n = 1), cardiac tamponade (n = 1), chylous ascitic fluid (n = 1), choroiditis (n = 1), parotid abscess (n = 1), bone marrow failure (n = 1) and concurrent fever and rash mimicking SLE flare (n = 1).

There were two cases whose new diagnosis of acute lymphocytic leukaemia (ALL) was made around the same time as the dengue diagnosis (Kooijmans 2019, Ganesan 2022). In some cases, symptoms overlapped with SLE flare, which delayed the diagnosis of dengue infection (Souza 2010).

Duration of dengue viremia or viruria

Eight studies reported viremia (Kanitkar 2020, Souza Pereira 2017, Machado 2017, Yadav 2021, Sim 2021, Pinsai 2019, Ng 2019, DiAscia 2024) and/or viruria (Machado 2017, Sim 2021, Ng 2019, DiAscia 2024). Median viremia duration: 30 (IQR 16.25–39.38); viruria: 157.5 (41.25–414.75) days. The longest dengue RNA detection: 4 months in blood (Ng 2019), over 2 years in urine (DiAscia 2024)—both in kidney transplant patients without dengue warning signs but lymphopenia. In addition to dengue RNA being detected in the urine, Ng et al. also demonstrated that dengue virus could be isolated from a urine sample at 9 months post-infection.¹⁸ Viral antigens were found in urinary podocytes at months 6–7, suggesting that podocytes may serve as viral reservoirs.

Discussion

Severe dengue was common, with an overall rate of 0.27 [0.22–0.33] decreasing to 0.16 [0.09–0.27] after excluding small, very high-bias studies. Compared to non-immunocompromised individuals, the risk of severe dengue was higher but not statistically significant (1.95 [0.51–7.49]). Notable findings include atypical dengue presentation, donor-derived dengue infection in transplant recipients and prolonged viral persistence.

Global dengue prevalence varied significantly by regions, and exact figures are difficult to determine due to differences in reporting practices and case definitions. In South America, which reports a large number of cases, the cumulative incidence was 419 cases per 100 000 persons in 2023.¹ Our review suggests that dengue is probably not more common in immunocompromised individuals, as incidence ranged from 0.1% to 6.3% in the included studies. However, this interpretation warrants caution, as most available data do not capture asymptomatic cases. Without active surveillance, the true burden of dengue among immunocompromised individuals remains unclear.

Our reported rate of severe dengue among immunocompromised individuals (16%–27%) exceeds that of the general population, which ranges from 0.25% to 1%.^4,19 However, our estimates may be inflated due to reporting bias, as severe cases are more likely to be reported than non-severe ones. A key limitation of this review is that the risk of severe dengue compared to non-immunocompromised individuals cannot be reliably determined, as only three studies (n = 3) reported severity outcomes for both groups. While our meta-analysis suggests a possible increase in the risk of severe dengue among immunocompromised individuals, the evidence remains inconclusive due to the small number of events and limited sample sizes in the included studies (Fig. 1). Therefore, although immunocompromised individuals may be at higher risk of developing severe dengue, further research is needed to confirm this association. This review did not investigate the underlying causes of severe disease, but we hypothesize that a combination of co-morbidities and impaired viral clearance may contribute to the increased severity observed in this population.

Sohail et al. reported a higher odds ratio of dengue complications (1.87 [1.04–3.35]) but a lower likelihood of severe dengue (0.83[0.69–1.00]) in immunocompromised individuals in a recently published systematic review.²⁰ The key limitation of that review was the inclusion of many studies involving malnutrition, which was not considered an immunocompromised condition in our analysis. Additionally, most studies on transplant recipients, patients on biologics and those with cancer were excluded from their meta-analysis due to the absence of control groups.

Mortality among dengue-infected immunocompromised individuals in all included studies was notably higher than in those used for the sensitivity analysis, which excluded studies with a very high risk of bias and small sample sizes (n < 10). This suggests that true mortality in immunocompromised populations is likely below 10%. However, most studies in the sensitivity analysis involved SOT recipients. High mortality was reported in HSCT and haematological malignancy cases, but these were largely low-quality case reports. Therefore, we cannot conclude that mortality in HSCT or haematological malignancies are as low as in SOT. Overall, mortality in immunocompromised individuals is likely higher than in the general population at 0.05%–0.52%.¹

When assessing all PLWH regardless of their CD4 counts, the proportion of severe dengue was not as high as in other immunocompromised subgroups. This was likely because most PLWH included in the review had CD4 > 350 cells/μl and were on stable anti-retroviral therapy. In fact, when we reviewed PLWH with CD4 < 350 cells/μl, 6 of 11 (54.5%) cases had severe dengue. Thus, we speculate that PLWH whose CD4 cell counts are low are at a greater risk of severe dengue.

Pre-transplant dengue screening is not universal. Our review revealed that 23 of 408 transplant recipients with dengue infection acquired dengue from their donors (Table 2). This is notable given the mortality: 66.7% in HSCT and 20% in SOT. Some patients had atypical presentations, which may delay diagnosis. For instance, the absence of fever may not prompt clinicians to suspect dengue. This study urges transplant teams to remain vigilant and consider implementing screening in endemic countries, as done in Reunion Island, France, where four early post-kidney transplant cases occurred before screening was introduced.²¹ Among SOT recipients, particular attention may be warranted in liver transplantation, as both fatal donor-derived cases occurred in liver recipients. This may reflect the role of Kupffer cells or liver-resident macrophages, which support viral replication, contributing to viremia and disease progression.^22,23

Further, dengue diagnosis can be challenging. Several studies reported that patients may present with symptoms like their underlying autoimmune conditions such as SLE, which led to inappropriate management of dengue infection initially.^24,25 In addition, concurrent diagnosis of ALL and dengue was also reported.^26,27 There is currently no proof of causation of dengue-induced ALL or whether this was only a coincidence.

The dengue incubation period is typically under 2 weeks; however, one study described possible chronic and neurotropic infection in a severely immunocompromised patient (PLWH with CD4 of 30 cells/μl) who presented with encephalitis three months after emigrating from Nigeria to Australia.²⁸ Other studies also reported persistent viremia and viruria, suggesting prolonged viral presence in immunocompromised hosts. In addition to neurons, podocytes have been identified as reservoirs for the dengue virus.¹⁸ While dengue RNA is generally undetectable beyond 2 weeks in blood^29,30 and 3 weeks in urine,³¹ we identified studies reporting detectable dengue virus up to 4 months in blood and 2 years in urine. However, it is important to note that the viral strain detected at 2 years post-infection was not confirmed to be the same strain as the original infection.

The mechanism of persistence was partially eluded in the two studies. Ng 2019 reported viral resolution coinciding with CD8+ T-cell recovery.¹⁸ T-cell immunity appears central to viral control: murine and in human studies show that dengue-specific CD4+ and CD8+ T cells are more frequent in individuals with subclinical infections,^32,33 and more robust, polyfunctional memory CD8+ T-cell responses are linked to protection from severe disease.³⁴ Future studies should confirm if lymphopenia predisposes to persistent dengue infection.

Dengue prevention strategy through vaccination has become available in recent years. Among the two currently licensed vaccines (Qdenga and Dengvaxia), Qdenga is available in many endemic and non-endemic countries.³⁵ It is recommended for individuals aged 4 years and older. However, some experts have expressed concern about a theoretical risk of severe infection with dengue virus serotypes 3 and 4 in dengue-naïve individuals.³⁶ Limited post-marketing surveillance data exist,³⁷ but there are no reports of severe dengue following vaccination in travellers. Both Qdenga and Dengvaxia are live attenuated vaccines and therefore contraindicated in immunocompromised individuals. For these individuals, education about dengue and mosquito-borne illnesses, along with mosquito bite prevention strategies, must be emphasized.

This study has several limitations. Since there were only a few studies comparing severe dengue rate between immunocompromised and non-immunocompromised, we were not able to conclude with certainty if immunocompromised individuals are at a higher risk of severe dengue. In addition, most studies about dengue in immunocompromised individuals were case reports that had inherent biases. We included a broad range of immunocompromised populations to enhance generalizability; however, this heterogeneity limits the precision of our conclusions. Subgroup analyses by type of immunocompromised condition were limited due to the small number and low quality of relevant studies. Furthermore, dengue virus serotypes were not reported in most included studies, which may be a source of confounding if different serotypes influence disease severity in immunocompromised individuals. It is also possible that excluding studies published in languages other than English from endemic countries leads to inaccurate estimates. Nonetheless, this review provides an estimate of severe dengue among the immunocompromised population that clinicians can use when counselling immunocompromised individuals and/or travellers. Additionally, these estimates may inform future studies, such as sample size calculation for vaccine or antiviral therapeutic trials in this patient population.

Conclusions

Dengue infection in immunocompromised individuals may not be more common than non-immunocompromised individuals, but once infected, severe outcomes could be more frequent. Illness may be prolonged, and the virus itself may persist for months. Clinicians should be aware of donor-derived and other atypical presentations of dengue infection for prompt diagnosis. This study provides insights that are clinically relevant. Future studies should further elicit factors associated with these adverse outcomes.

Prospero: CRD42024584266

Funding

None declared.

Author contributions

J.C. and V.L. conceptualized the study, screened the study, collected and analysed the data, and wrote the first draft of the manuscript; S.G., P.R. and P.P. provided critical input for data collection, data analysis and data interpretation; all authors critically edited and approved the final version of the manuscript.

Juthaporn Cowan (Conceptualization, Data curation, Methodology, Writing—original draft, Writing—review & editing [equal], Formal analysis, Visualization [lead]), Viravarn Luvira (Conceptualization, Data curation, Methodology, Writing—original draft, Writing—review & editing [equal], Formal analysis [supporting]), Sivaporn Gatechompol (Conceptualization, Data curation, Formal analysis, Writing—review & editing [supporting]), Pinyo Rattanaumpawan (Conceptualization, Data curation, Formal analysis, Writing—review & editing [supporting]), Punnee Pitisuttithum (Conceptualization, Data curation, Formal analysis, Writing—review & editing [supporting])

Conflict of interest

J.C. and P.R. received honoraria from Takeda not related to this work. V.L. and P.P. are site investigators for dengue vaccine trials sponsored by Merck and Takeda. S.G. declares no conflict of interest.

Data availability

The data that support the findings of this study are available in the supplementary data file. Additional data can be requested from the corresponding author.

Ethical approval statement

This study is a systematic review of previously published data and does not involve human participants or identifiable data. Therefore, ethics approval was not required.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work, the authors used ChatGPT to assist with spelling, grammar and clarity, as well as to support the development of R scripts used for meta-analysis. All authors reviewed and edited the content and take full responsibility for the final content.