Dengue fever in renal transplant patients: a systematic review of literature

Background Dengue fever in renal transplanted patients has not been studied well, and we review all the literature about episodes dengue fever in renal transplant patients. Methods The aim was to describe clinico-pathological characteristics, immunosuppressive protocols, need renal outcome and mortality. PubMed, LILACS, Google Scholar and Research Gate were searched for “Dengue” and “Renal/Kidney Transplantation” with no date limits. Hits were analyzed by two researchers separately. Results Fever, myalgia, arthralgia and headache was significantly lower than normal population, while pleural effusions and ascites were observed more. Incidence of severe dengue is significantly higher among transplant patients in this review, as well as they had a significantly higher mortality (8.9% vs 3.7%, p = 0.031). Age, period after transplantation and immunosuppressive profile had no effect on disease severity, mortality or graft out come. Presence of new bleeding complications and ascites was associated with more severe disease (p < 0.001 and p = 0.005), death (p = 0.033) or graft loss (p = 0.035). Use of tacrolimus was associated with new bleeding complications (p = 0.027), and with ascites (p = 0.021), but not with thrombocytopenia. 25% of patients with primary disease fail to mount an IgG response by 15 weeks of the illness. 58.9% had graft dysfunction during illness. Postoperative transplanted patients were at risk of severe disease and unfavorable outcome. Conclusions The physical and laboratory findings in dengue fever in renal transplanted patients differ from the general population. Some degree of graft dysfunction is common during the illness, but only a minority develops graft failure.


Background
Dengue is the most rapidly spreading mosquito-borne viral disease in the world. In the last 50 years, incidence has increased 30-fold with spread to new countries, and from urban to rural settings. An estimated 50 million dengue infections occur annually and approximately 2.5 billion people live in dengue endemic countries [1]. The illness ranges from being asymptomatic or a mild flu like illness (dengue fever, DF) associated with a rash, to a more severe form with plasma leakage, hemorrhage (dengue hemorrhagic fever -DHF), shock and multi organ failure (dengue shock syndrome-DSS). Liver failure, cardiac, neurological or hematological complications may occur asatypical complications, termed as extended dengue syndrome [2][3][4][5].
With the advancement of immunosuppression, renal transplant has become the management of choice for end stage renal disease (ESRD). Some transplant recipients are living in hyper-endemic and endemic areas for dengue, and are at risk of developing the disease. The affected countries are places of great tourist attraction, and travelling transplant recipients are at risk too. Dengue is described as a mild disease in renal transplant patients [6], but severe morbidity (graft failure needing dialysis and graft nephrectomy) [7], and mortality [4], had been reported. This paper reviews demographical, clinic-pathological and immunosuppression profiles of the renal transplant patients with confirmed dengue infection.

Methods
We searched PubMed, LILACS, Google Scholar and ResearchGate for publications by key words / MeSH terms "dengue" AND ("renal transplant" OR "kidney transplant") in title, abstract or full text, with no date limits. The search was performed in December, 2015. The two investigators (RMW and JAP) independently gone through the abstracts to identify relevant papers. We selected all the papers that had information on patients with renal transplant who suffered from dengue infection. There were no randomized trials; all the papers were cross sectional studies or case series / reports. Due to the rarity of the condition, we included all of them in this review. We also searched related publications and papers referenced in review articles. Full papers of selected studies were read through by the two investigators, who extracted relevant data. Duplicate publications of the same data were excluded. The results of the larger studies are presented as it is, and findings of all the case reports and the case series (composite group) were summarized and analyzed. Data of two patients from an unpublished work from RMW was added to the review. Post operative period was defined as first 14 days after the transplant. Whenever possible the combined statistics across all the cross sectional studies / case reports was calculated. SPSS 16.0 was used for the analysis. For summary data the Fisher's exact value calculator from the web site GraphPad Software® was used [8], while for calculating pooled averages and standard deviations, tools from University of Baltimore web site [9] was utilized. Forest plots were created using RevMan 5.3® from Cochrane Reviews [10]. Mann Whitney U test and Fishers exact tests was used to compare groups as normality assumption was violated by the continuous variables, and of small values in cross-tabs. Risk of bias analysis was performed by two the authors independently, using ROBINS-I tool. Inter-rater agreement was 100%.

Results
The search of PubMed, Google Scholar, LILACS and ResearchGate resulted in 33 hits. Once the duplicates were removed we were left with 13 records. Three records were responses to previous case reports, and hence were excluded (See Fig. 1). We reviewed two large cross sectional studies, three case series and four case reports. In one study only the abstract was available [11], and in the other immunosuppression protocol was absent [12], however we used data on symptomatology, lab investigations and mortality. The total number of subjects described in literature was 168. Table 1 gives a summary of the reviewed studies. Most of the studies had information on basic demographics (age and gender), time since transplant, details of immunosuppression, symptomatology, Nasim et al. [6] has done the largest study on DF in RT involving 102 subjects (Male 73.7%) in city of Karachchi for a period of 2 years. The conclusions of their study are summarized in Table 2. Mean increase of creatinine was 66 ± 48 umol/L. Azavedo et al. [2]. Reported a cross sectional of 27 (Male 67%) patients from Brazil, and found that, the symptoms are similar to that of nontransplant population other than for arthralgia. The outcome of transplanted patients was similarto the normal dengue population. Immunosuppression did not affect the outcome. Mean increase of creatinine was 35.7%.
We have analyzed the risk factors for signs and symptoms as well as mortality in the composite group. Out of the large number of comparisons only the outcomes described in Nasim et al. and Azavedo et al. was analyzed. Gender and immunosuppressive profile was not associated with disease severity or the outcome of the disease. Presence of new bleeding complications and ascites was associated with more severe disease (Fishers' Exact, p < 0.001 and p = 0.005) as well as with adverse outcome of death or graft loss (Fishers' Exact, p = 0.033 and p = 0.035). Use of tacrolimus was associated with new bleeding complications (Fishers exact test, p = 0.027), and with ascites (Fishers exact test, p = 0.021), but not with thrombocytopenia. We could not run a multiple logistic regression with new bleeding manifestations, severity of disease, use of tacrolimus and presence of • Fever is commoner in patients taking low dose steroids to patients on high-dose steroids (p = 0.013)

•
The anti-mitotic agents (azathioprine (AZA) or mychophenolate mofetil (MMF)) have no effect; on the severity or duration of thrombocytopenia;leucopenia; and occurrence of gastro-intestinal symptoms.
•  ascites as covariates, due to small number of records (n = 13).

Synthesis of results
In the synthesis of the results, we tried to pool the common parameters to achieve single values of central tendency and proportion. The combined results included 168 patients (Male 67.9%) with a mean age of 38.9 ± 13.8y (n = 66), and a mean time since transplant 53.6 ± 60.8 months (n = 66). The duration of transplant did not have effect on the presentation or the outcome. Mean increase of creatinine was 67.1% (n = 131) from the baseline. Table 5 summarizes the combined results across all the studies. We found new bleeding complications or diarrhea to show no difference in frequency compared to the normal population [15]. Fever, myalgia, arthralgia and headache were significantly lower than normal population, while pleural effusions and ascites were observed in a greater frequency. Incidence of severe dengue is significantly higher among transplant patients in this review, as well as they had a significantly higher mortality. The mortality was significantly higher than that of severe dengue fever (8.9% vs 3.7%, t = 2.27, p = 0.031). The forest plot for the mortality results are shown in Fig. 2.
In the study by Nasim et al., patients who are on tacrolimus had a higher mortality (n = 102, p = 0.031, Fishers' Exact), but in the final analysis showed no association (n = 131, v0.28, Fishers' Exact). Being in the post operative period, increased the risk of dengue hemorrhagic fever (n = 31, p = 0.001, Fisher's Exact) as well as, ending up dead or with a failed graft (n = 31, p = 0.044, Fisher's Exact).

Risk of bias across studies
All of these studies are either case reports or cross sectional surveys, and as a result have selection bias. Patients who have severe disease are more likely to turn up at their physicians' office rather than a person having a mild flu like illness. Hence all the studies are biased towards severe manifestations of the disease, as a result higher death rates and morbidity. A formal risk of bias analysis was performed using ROBINS-I tool, and the results are described in Table 4. Publication bias is obvious, due to the rarity of the clinical circumstances.

Discussion
Renal transplanted patients with Dengue Fever exhibit different clinical features to that of general population. Fever, headache, myalgia and arthralgia are less common in transplant population, while incidence of ascites and pleural effusions is higher. Transplanted patients had a higher incidence of severe dengue fever and a higher mortality. Nasim et al. report a significantly lower rate of severe primary dengue infection, in patients taking low dose steroids. Additionally, they report a lower mortality from secondary infection, in patients whom on cyclosporine based regimens. We suggest that deciding whether a dengue infection is primary or secondary Table 4 Risk of bias summary for the studies. The colored bullet signifies the risk of bias for each study, under sections described in ROBINS I checklist. Light green -low, dark green -moderate, orange -serious, red -critical, ash -risk of bias cannot be assessed due to lack of information.
should be done in care in transplant population. Nasim et al. reports that 10 out of 21 of their patients who were IgG negative and IgM positive −22.7% of patients diagnosed of primary infection-failed to mount an IgG response after 15 weeks of the illness. This raises the validity of the diagnosis of primary dengue infection in transplanted population in previous studies. We suggest that the distinction between primary and secondary infections in transplanted patients should be disregarded, given the unreliability of the antibody response following the infection. The higher mortality associated with tacrolimus [6] was no longer significant when composite group was added to the analysis. Serological diagnosis is still the most popular form of diagnosis, but RT-PCR and NS1 antigen which provide speedy confirmatory diagnosis are gaining popularity. Nearly sixty percent developed graft dysfunction, with a mean creatinine rise of 61.7%, during the course of illness and from most of them recovered. Mortality was 8.9% and graft loss was about 6.5%, and is not due to direct effects of dengue. Mortality is way higher than for the normal population (0.062%) [1], but as discussed above, the samples were subjected to heavy selection bias as only the patients with most severe symptoms would have undergone confirmation of the dengue infection, while others would have been treated as for a non-specific viral illness. The results show post operative patients being more vulnerable to severe disease as well as reaching the end point of death or losing the graft.

Limitations
Most of the studies used to in this review are cross sectional studies and case series / reports. Hence the quality of evidence is not as high as that of a randomized trial. The selection bias as described earlier makes the generalization of the results difficult to the whole population. Since ROBINS-I was developed mainly for case control and cohort studies, we cannot get and accurate estimate of bias as well. Still the evidence is helpful in the management of inward patients with dengue fever.

Conclusions
The physical and laboratory findings in dengue fever in renal transplanted patients do not differ from the