Tropical Medicine and International Health
doi:10.1111/tmi.12429
volume 00 no 00
Systematic Review
Serological diagnosis of canine visceral leishmaniasis in Brazil: systematic review and meta-analysis Henry Maia Peixoto1, Maria Regina Fernandes de Oliveira1,2 and Gustavo Adolfo Sierra Romero1,2 1 Center for Tropical Medicine, University of Brasılia, Brasılia, Brazil 2 National Institute for Science and Technology for Health Technology Assessment, Porto Alegre, Brazil
Abstract
objective To evaluate the quality and accuracy of serological diagnosis of canine visceral leishmaniasis in the Americas. methods A systematic review found original studies in the databases MEDLINE, EMBASE and LILACS up to November 2012 and in complementary sources up to February 2013. Studies were evaluated in accordance with QUADAS 2 and STARD parameters and recommended in accordance with GRADE parameters. Meta-analysis was carried out with Meta-DiSc software, using the randomeffect model. results Two hundred and eighty-four studies were identified, of which 25 met the inclusion criteria, comprising the final synthesis. All but one was conducted in Brazil, and only two were judged to be of good quality. 15 studies involving immuno-enzymatic tests with crude antigens (cELISA), 11 studies on indirect immunofluorescence tests (IFAT) and three on the immunochromatographic dual-path platform (DPP) test were meta-analysed. The combined results for sensitivity and specificity were cELISA: 0.89 (CI 95% 0.87–0.91) and 0.87 (CI 95% 0.86–0.88); IFAT: 0.88 (CI 95% 0.85–0.91) and 0.63 (CI 95% 0.61–0.65); and DPP: 0.83 (CI 95% 0.78–0.88) and 0.73 (CI 95% 0.70–0.75). conclusion Enzyme-linked immunosorbent assay with crude antigens and DPP tests have moderate accuracy for the diagnosis of canine visceral leishmaniasis, and the quality of the design, implementation and analysis of validation studies on diagnostic tests for this disease urgently require improvement. The recommendation for use of the evaluated tests is based on evidence that is scarce and restricted to Brazil. keywords visceral leishmaniasis, diagnosis, dogs, meta-analysis, Leishmania infantum, serology
Introduction Visceral leishmaniasis (VL) in the Americas is a zoonotic disease caused by the protozoan Leishmania infantum (sin. Leishmania chagasi), whose main reservoir is the domestic dog. It is principally transmitted by the bite of female sandflies of the genus Lutzomyia (Dantas-Torres & Brand~ao-Filho 2006; Bauzer et al. 2007; Dantas-Torres 2007). Human VL is characterised by fever, hepatosplenomegaly and pancytopenia. Among the factors associated with higher fatality rate, the following stand out: severe anaemia; haemorrhagic manifestations; cardiac abnormalities; jaundice; diarrhoea; fever for more than 60 days; oedema and bacterial infection and coinfection with HIV (Madalosso et al. 2012; Lima et al. 2013).
© 2014 John Wiley & Sons Ltd
Human VL has become a serious emerging problem in worldwide public health, with approximately 200 000– 400 000 new cases per year (Alvar et al. 2012). An estimated 98 countries are considered endemic, and 90% of cases are concentrated in India, Bangladesh, Sudan, Ethiopia and Brazil (Alvar et al. 2012). On the American continent, it is estimated that most cases occur in Brazil, with 4200–6300 cases per year, followed by Paraguay (100–200 new cases per year), Colombia (70–110 new cases per year) and Venezuela (50–70 new cases per year) (Alvar et al. 2012). In Brazil, human cases have been described in all five regions of the country, with a mean lethality of 5.8% (Braga et al. 2013; MS 2011a; de Ara ujo et al. 2012). To reduce dissemination of the disease, strategies to control and monitor canine visceral leishmaniasis (CVL)
1
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
are essential. In Brazil there are approximately 20– 30 million dogs (MS 2013; Reichmann et al. 1999; WHO 1992). In CVL-endemic regions, seroprevalence varies in the general population, including symptomatic and asymptomatic dogs, between 3.4% and 40%, revealing the potential for surveillance of canine infection as a marker of the transmission and emerging challenges (Almeida et al. 2009; MS 2011b; Dantas-Torres et al. 2006; Prado et al. 2011). Euthanasia of infected dogs has been one of the main interventions implemented in Brazil for controlling VL. The detection of canine infection is based on serological tests whose accuracy has not been systematically evaluated in the region. In terms of public health, the intervention has not been successful and a critical review raised the possibility that this may be due to insufficient accuracy of the serological tests used (Romero & Boelaert 2010). An ongoing debate on abolishing dog euthanasia as a public health intervention to control VL is far from over; however, detection of canine infection will remain critical for disease surveillance and a key issue for animal health care. Thus, the choice of an accurate and reproducible diagnostic test, capable of supporting control and surveillance strategies in canines, is fundamental. However, gaps in knowledge make this choice difficult (Grimaldi et al. 2012; Solca et al. 2012; de Souza et al. 2012). Romero and Boelaert (2010) demonstrated that although current tests present high sensitivity among symptomatic dogs, they were not consistently evaluated among asymptomatic dogs. The lack of a reliable reference standard, the long period between taking the test and euthanasia, migration of infected animals and the inability to differentiate between positive serology observed in samples obtained from infected and vaccinated dogs all pose additional challenges for developing appropriate tests to achieve effective control of the disease. The systematic review by Quinnell et al. (2013), focused on rapid tests based on k39 antigen, concluded that their sensitivity is too low for supporting control measures. Both reviews (Romero & Boelaert 2010; Quinnell et al. 2013) lack standardised quality evaluation of the evidence. Hence, the present systematic literature review with meta-analysis on the accuracy of serological tests for diagnosis of CVL in the Americas includes a thorough evaluation of the quality of the evidence base. Methods We selected indexed articles from the databases Excerpta Medica Database (EMBASE); Medical Literature Analysis and Retrieval System Online (MEDLINE) and Latin American and Caribbean Literature in Health Sciences
2
(LILACS), published up to November 2012. Eligibility criteria included original studies evaluating any serological diagnostic test for CVL as target condition; adequate classification by a reference standard; absolute numbers of true-positive, true-negative, false-positive and false-negative observations available or derivable from the data presented. Compound parasitological tests (diagnosis based on more than one technique and/or the same technique in more than one sample) or a simple parasitological test (one diagnostic technique applied to a single sample) was considered as adequate reference standard. Accuracy measures were summarised as sensitivity and specificity, likelihood ratio (LR) and diagnostic odds ratio (DOR). Additional articles and documents were identified up to February 2013, by tracking the references cited in selected articles, review articles, technical documents produced in meetings of specialists in VL at the Brazilian Ministry for Health, master’s dissertations, doctoral theses and unpublished articles. Research conducted outside the Americas was excluded. Research in the databases was organised using the descriptors and MeSH terms (Box 1). Each abstract identified was read by a researcher supervised by a second researcher, and the possible divergences in their evaluation were resolved with the participation of a third researcher. Those that fulfilled the inclusion criteria and did not present the exclusion criteria were selected for complete reading. Abstracts from which no conclusion could be drawn on whether the criteria were fulfilled were also selected for complete reading. Articles selected for complete reading were again evaluated for inclusion and exclusion criteria. Data on sensitivity and specificity, including point and interval estimates, sample size and origin (endemic or non-endemic), use of screening procedures previous to validation process, blinding and the use of reference standards were extracted into standardised tables. The selected articles and documents were evaluated for their quality by means of QUADAS 2 (Quality Assessment of Diagnostic Accuracy Studies) (Whiting et al. 2011), complemented by three questions from the checklist of STARD (Standards for Reporting of Diagnostic Accuracy) (Bossuyt et al. 2003) as suggested by Oliveira et al. (2011). The three criteria were as follows: the sampling process is described; sensitivity and specificity results are reported with their respective confidence intervals (CI); and clinical and demographic characteristics of patients are reported. QUADAS 2 is constituted by four key domains which discuss patient selection, the index test, the reference standard and the flow and timing; all four are judged in terms of the risk of bias, and the first three in terms of applicability (Whiting et al. 2011). Studies were consid-
© 2014 John Wiley & Sons Ltd
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
Box 1 Descriptors and MeSH terms used for literature search MEDLINE via PubMed with syntax adapted from guidelines proposed by Deville et al. (2002): (((((((((((((((((((((((((((‘sensitivity and specificity’[All Fields] OR ‘sensitivity and specificity/standards’[All Fields]) OR specificity’[All Fields]) OR ‘screening’[All Fields]) OR ‘false positive’[All Fields]) OR ‘false negative’[All Fields]) OR ‘accuracy’[All Fields]) OR ‘predictive value’[All Fields]) OR ‘predictive value of tests’[All Fields]) OR ‘predictive value of tests/standards’[All Fields]) OR ‘predictive values’[All Fields]) OR ‘predictive values of tests’[All Fields]) OR ‘reference value’[All Fields]) OR ‘reference values’[All Fields]) OR ‘reference values/standards’[All Fields]) OR ‘roc’[All Fields]) OR ‘roc analyses’[All Fields]) OR ‘roc analysis’[All Fields]) OR ‘roc and’[All Fields]) OR ‘roc area’[All Fields]) OR ‘roc auc’[All Fields]) OR ‘roc characteristics’[All Fields]) OR ‘roc curve’[All Fields]) OR ‘roc curve method’[All Fields]) OR ‘roc curves’[All Fields]) OR ‘roc estimated’[All Fields]) OR ‘roc evaluation’[All Fields]) OR ‘likelihood ratio’[All Fields]) AND ((‘dogs’[MeSH Terms] OR ‘dogs’[All Fields] OR ‘canine’[All Fields]) AND (‘leishmaniasis, visceral’[MeSH Terms] OR (‘leishmaniasis’[All Fields] AND ‘visceral’[All Fields]) OR ‘visceral leishmaniasis’[All Fields] OR (‘visceral’[All Fields] AND ‘leishmaniasis’[All Fields]))). EMBASE: ((((((canine AND visceral AND (‘leishmaniasis’/exp OR ‘leishmaniasis’ OR leishmaniasis/exp OR leishmaniasis)) OR l.infantum OR l.chagasi OR l.donovani OR ‘leishmania’/exp OR ‘leishmania’ OR leishmania/exp OR leishmania) AND infantum) OR ‘leishmania’/exp OR ‘leishmania’ OR leishmania/exp OR leishmania) AND chagasi) OR ‘leishmania’/exp OR ‘leishmania’ OR leishmania/exp OR leishmania) AND donovani AND ((((((((diagnostic AND (‘accuracy’/exp OR ‘accuracy’ OR accuracy/exp OR accuracy)) OR diagnostic) AND (‘performance’/exp OR ‘performance’ OR performance/exp OR performance)) OR sensitivity OR specificity OR validation OR false) AND positive) OR false) AND negative) OR roc) AND (‘Americas’/exp OR ‘Americas’ OR Americas/exp OR Americas). LILACS: canine AND leishmaniasis AND visceral AND diagnosis.
ered applicable if they selected infected and/or noninfected dogs, coming from endemic or non-endemic areas; if they evaluated any serological test and if they used parasitological tests in the composition of the reference standard. The STARD questions broadened the perception of the sampling process; the results for sensitivity, specificity and confidence intervals (CI); and the characteristics of the dogs. Recommendations were made using GRADE (Grading of Recommendations Assessment, Development and Evaluation Guidelines) (Horvath 2009), which evaluates the magnitude of the outcome measures, checked in terms of true-positive and false-negative values and the complications arising from the test. It took into consideration the quality of the evidence, the uncertainty with regard to identified values and advantages of the tests being evaluated (Guyatt et al. 2008; Horvath 2009). Studies were classified according to the following evaluation phases: phase 1: early exploratory studies; phase 2: case–control design; and phase 3: prospective studies validating the test in the target population (Boelaert et al. 2007). The statistical software Meta-DiSc supported the statistical analysis which combined data for sensitivity, specificity, LR and DOR (Zamora et al. 2006) as well as construction of SROC curves. The model of random effects was used to include the heterogeneity observed
© 2014 John Wiley & Sons Ltd
between the various studies, evaluated by the Cochran Q-test, chi-squared test and the inconsistency index (I2), (Higgins et al. 2003; Zamora et al. 2006; Sousa & Ribeiro 2009; Riley et al. 2011). 95% confidence intervals were calculated for the accuracy measurements in the individual studies and for the synthesised measures. Data from studies that analysed more than one similar test on the same sample set were considered, but just data from one test were used in the meta-analysis being prioritised the commercially available tests followed by in-house tests. Homologous antigens were prioritised over those using heterologous antigens. Among the phase 2 studies, data were prioritised when included symptomatic and asymptomatic dogs for constituting the case group and included healthy dogs plus dogs affected by other diseases than VL. Subgroup analysis was performed to compare the summarised accuracy measures in samples obtained from general vs. previously screened population, from dogs residing inside vs. outside the endemic area, symptomatic vs. asymptomatic dogs and all studies vs. those conducted with commercially available tests. Lastly, outliers were individually analysed and excluded from meta-analysis if their accuracy measures were clearly different from most of the studies and the summarised measure were not greatly affected by their exclusion. 3
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
Additional records identified (ARI) through other source (n = 7) (Published papers: 4; unpublished papers: 2; Dissertation: 1)
Records identified through database searching (n = 277) (MEDLINE = 218; EMBASE = 21; LILACS = 38)
Records after duplicates removed (n = 271)
Records excluded: MEDLINE = 188 EMBASE = 14 LILACS = 25 ARI = 1
Records screened (n = 271) (MEDLINE = 218; EMBASE = 17; LILACS = 30; ARI = 6)
Full-text articles assessed for eligibility (n = 43)
Full-text articles excluded, with reasons (n = 18) ·Inappropriate gold standard = 4 ·Developed outside of America = 3 ·Paper does not address validation of diagnostic test = 11
Studies included in qualitive synthesis (n = 25)
Figure 1 Fluxogram of the selection process for the evidence on canine serological diagnosis for visceral leishmaniasis.
Results Selecting the evidence base Figure 1 describes the process of selecting the evidence base during different phases of the review. Initially, 284 studies were retrieved, of which 277 came from a systematic search in the databases and seven from complementary sources. After checking for duplicates, 271 articles were left, of which 228 were excluded after reading the title and abstract because they did not meet the inclusion criteria. Therefore, 43 studies were selected for complete reading. After complete reading, 18 articles were excluded and 25 remained for the final synthesis. Quality of the evidence In Table 1, the main characteristics of the 15 diagnostic tests are described, as identified in the 25 studies selected. Most of the studies used phase 2 design and just four reports corresponded to phase 3 studies. Table 2 shows the result of the quality evaluation of the 25 studies selected in accordance with the QUADAS 2 domains
4
(Whiting et al. 2011), with regard to risk of bias and to applicability, as well as the complementary questions obtained from STARD (Bossuyt et al. 2003). The final evaluation of the studies is described in Table 2, in accordance with the risk of bias (low or high risk), concern about applicability (low or high risk) and the quality of the evidence (weak or strong). The evaluation of the risk of bias related to the appropriate use of the reference standard was conducted applying the two signalling questions for this specific criterion and just studies using a compound parasitological test interpreted without knowledge of the results of the index test were classified as having low risk of bias. As shown in Table 2, only two studies (MS 2011b; de Arruda et al. 2013) did not present any risk of bias in any of the evaluated domains. Of the 25 articles evaluated, three did not present risks of bias regarding selection of individuals, eight with regard to the test being evaluated, four with regard to the reference standard and 12 regarding flow. As for the three questions adapted from STARD, only one study (MS2011b) described the sampling calculation adequately, 10 presented the 95% CI for sensitivity and
© 2014 John Wiley & Sons Ltd
© 2014 John Wiley & Sons Ltd
Brazil/2012/ Solca et al. (2012)
ELISA tests Brazil/1996/ Paranhos-Silva et al. (1996) Brazil/1996/ Garcez et al. (1996) Brazil/2005/ Rosario et al. (2005) Brazil/2005/ Rosario et al. (2005) Brazil/2007/ Ferreira et al. (2007) Brazil/2007/ Porrozzi et al. (2007) Brazil/2008/ Lemos et al. (2008) Brazil2009/ Andrade et al. (2009) Brazil2010/ Camargo et al. (2010) Brazil/2010/ Figueiredo et al. (2010b) Brazil/2010/ Figueiredo et al. (2010a) Brazil/2011/Faria et al. (2011)
Country/year/Ref.
2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and (HD or OD) 2 groups: ID and (HD and/or OD) 2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and HD 1 group: SDGCP
1 Group: SD (seronegative by IFAT-eluate) 2 groups: ID and HD
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 3
Phase 2
Phase 1
Phase 2
1 Group: SD (euthanised)
2 groups: ID and HD
Phase 1
Phase 2
2 groups: ID and HD
Recruitment
Phase 2
Phase of evaluation
cELISAk Promastigote L. major cELISANR/NR
cELISA¶ Promastigote L. chagasi cELISA¶ Promastigote L. chagasi cELISA¶/Eluate Promastigote L. chagasi cELISA¶ L. donovani complex cELISA¶ Promastigote L. chagasi cELISA¶ Promastigote L. chagasi cELISAk Promastigote L. major cELISAk Promastigote L. major cELISAk Promastigote L. major cELISANR NR
cELISANR NR
Test, parasite form and species
Parasitological†
Parasitological†
Parasitological†
Parasitological†
Parasitological†
Parasitological‡ and serology
Parasitological†
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Reference standard
18 (ASY/SY/OL)
62 (ASY/SY/OL)
5 (ASY/SY)
9 (NR)
45 (NR)
64 (ASY/SY/OL)
112 (ASY/SY/OL) – – 50 (SY/OL) 50 (ASY) 100 (ASY/SY/OL) 76 (ASY/SY/OL)
106 (ASY/SY/OL)
106 (ASY/SY/OL)
21 (ASY)
46 (NR)
With CVL (clinical condition)
83.3
8.0
100
100
93.3
95
96 – – 88 30 59 95
100
98
71
98
Sensitivity %
88
100
86
99
Specificity %
27 (DE)
20 (HDE)
141 (DE)
296 (DE)
100 (HDNE)
59.2
100
96.5
96.6
99
100
(HDE) 100 (OD) 60 (HDE +OD 80 (HDNE) 100 (OD) 64 (HDNE + OD) 87 (HDNE) 100
25 (HDE)
20 20 40 25 14 39 33
25 (HDE)
25 (HDE)
14 (HDNE)
102 (HDNE)
Without CVL (origin)
Table 1 Main characteristics of the articles published on validation of the diagnostic tests for detection of canine visceral leishmaniasis in the Americas
Tropical Medicine and International Health volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
5
6
Brazil/2012/ Alves et al. (2012) Brazil/2012/ Alves et al. (2012) Brazil/2013/ de Arruda et al. (2013) Brazil/2013/ de Arruda et al. (2013) Brazil/2008/ C^andido et al. (2008) Brazil/2009/ Pinheiro et al. (2009) Brazil/2009/ Pinheiro et al. (2009) Brazil/1995/ Dietze et al. (1995) Brazil/2011/ Marcondes et al. (2011) Brazil/2009/ Pinheiro et al. (2009) Brazil/2012/ de Souza et al. (2012) Brazil/2005/ Rosario et al. (2005) Brazil/2005/ Rosario et al. (2005)
Country/year/Ref.
Table 1 (Continued)
2 groups: ID and (HD and OD) 1 group: SDGCP
1 group: SDGCP
2 groups: ID and HD 2 groups: ID and (HD and OD) 2 groups: ID and (HD and OD) 2 groups: SDGCP (DNE and DE) 2 groups: ID and (HD and OD) 2 groups: ID and (HD and OD) 2 groups: ID and (HD and OD)
Phase 2
Phase 3
Phase 3
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
2 groups: ID and HD 2 groups: ID and HD
Phase 2
Phase 2
Phase 2
2 groups: ID and HD
Recruitment
Phase 2
Phase of evaluation
Rk26 ELISA¶ Eluate
ELISA rLic-N TPDase-2 antigen¶ Rk26 ELISA¶
ELISA rLdccys1¶
Rapid Test ELISAk
cELISAk Promastigote L. major cELISA¶ Promastigote L. chagasi cELISA¶ Promastigote L. chagasi cELISA¶ Promastigote L. chagasi cELISA¶ Amastigote L. chagasi Dot-ELISA¶
cELISAk Promastigote L. major cELISA¶ (NR) L. chagasi
Test, parasite form and species
Parasitological‡
Parasitological‡
Parasitological§ and serology
Parasitological†
Parasitological‡ and serology
Parasitological†
Parasitological†
Parasitological†
Parasitological†
Parasitological‡
Parasitological‡
Parasitological†
Parasitological†
Reference standard
(SY) (OL) (SY/OLY) (ASY/SY)
106 (ASY/SY/OL)
106 (ASY/SY/OL)
48 (ASY/SY/OL)
209 (ASY/SY)
283 (NR)
37 (ASY/SY)
209 (ASY/SY)
30 30 60 209
98 (ASY/SY/OL)
98 (ASY/SY/OL)
39 (NR)
39 (NR)
With CVL (clinical condition)
100
99.1
100
98.58
94.7
97
89.7
90 86.7 88.5 86.4
89.8
91.8
100
100
Sensitivity %
82.7
83.7
93.6
84.6 68.0
Specificity %
25 (HDE)
25 (HDE)
74 (HDE+OD)
68 (HDNE+OD)
117 (HDNE+OD)
93 (DNE+DE)
68 (HDNE+OD)
96
100
100
96.65
90.6
100
68.2
30 (HDNE) 93.3 30 (HDNE) 100 – – 68 (HDNE+OD) 68.2
1327 (DE)
1327 (DE)
78 (HDE+OD)
39 (HDE) 78 (HDE+OD)
Without CVL (origin)
Tropical Medicine and International Health volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
© 2014 John Wiley & Sons Ltd
Brazil/2007/ Porrozzi et al. (2007) Brazil/2007/ Porrozzi et al. (2007) Brazil/2005/ Rosario et al. (2005) Brazil/2005/ Rosario et al. (2005) Brazil/2007/ Porrozzi et al. (2007) Brazil/2007/ Porrozzi et al. (2007) Brazil/2007/ Porrozzi et al. (2007) Brazil/2007/ Porrozzi et al. (2007) Brazil/2008/ C^andido et al. (2008) Brazil/2008/ C^andido et al. (2008) Brazil/2008/Faria et al. (2011) IFAT tests Brazil/1996/ Paranhos-Silva et al. (1996) Brazil/1996/ Garcez et al. (1996)
Country/year/Ref.
Table 1 (Continued)
2 groups: ID and HD 2 groups: ID and OD 2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and OD 2 groups: ID and HD 2 groups: ID and OD 2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and HD 2 groups: ID and HD
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 2
Phase 1
Recruitment
Phase 2
Phase of evaluation
© 2014 John Wiley & Sons Ltd
Parasitological‡
Parasitological‡
IFATNR NR IFAT¶ Amastigote L. chagasi
Parasitological†
Parasitological†
Parasitological†
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Parasitological‡
Reference standard
ElisaMix10¶
FML–ELISA¶
FML–ELISA¶
Ra2 ELISA¶
Ra2 ELISA¶
Rk39 ELISA¶
Rk39 ELISA¶
Rk39 ELISA¶ Eluate
Rk39 ELISA¶
Rk26 ELISA¶
Rk26 ELISA¶
Test, parasite form and species
21 (ASY)
46 (NR)
62 (ASY/SY/OL)
30 (OL)
30 (SY)
50 (ASY)
50 (SY)
50 (ASY)
50 (SY)
106 (ASY/SY/OL)
106 (ASY/SY/OL)
50 (ASY)
50 (SY)
With CVL (clinical condition)
100
78
75.81
90
86.7
88
70
66
100
97.2
98.1
66
94
Sensitivity %
14 (HDNE)
102 (HDNE)
20 (HDE)
30 (HDNE)
30 (HDNE)
14 (OD)
25 (HDNE)
14 (OD)
25 (HDNE)
25 (HDE)
25 (HDE)
14 (OD)
25 (HDNE)
Without CVL (origin)
100
100
95.0
93.3
96.7
93
100
71
100
100
100
57
100
Specificity %
Tropical Medicine and International Health volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
7
8 2 groups: ID and HD 1 group: SDGCP
1 group: SDGCP
1 group: SD (seronegative by IFAT-eluate) 1 group: SDGCP
Phase 2
Phase 3
Phase 3
Phase 2
Brazil/2012/ Alves et al. (2012) Brazil/2012/ Alves et al. (2012) Rapid tests Brazil/2008/ Lemos et al. (2008) 2 groups: ID and HD
Phase 2
Phase 2
2 groups: ID and HD or (HD and OD) 2 groups: ID and (HD and OD)
Phase 2
Phase 3
Phase 2
2 groups: ID and seronegative dog 2 groups: ID and (seronegative dog) 1 group: SD (euthanised)
Phase 2
Brazil/2009/ Troncarelli et al. (2009) Brazil/2010/ Camargo et al. (2010) Brazil/2010/ Figueiredo et al. (2010b) Brazil/2010/ Figueiredo et al. (2010b) Brazil/2010/ Figueiredo et al. (2010a) Brazil/2011/MS (2011b)
2 groups: ID and HD
Phase 2
Phase 2
1 group: SD (SY)
Recruitment
Phase 3
Phase of evaluation
Brazil/2009/Silva et al. (2009)
Brazil/2006/ da Silva et al. (2006) Brazil/2007/ Ferreira et al. (2007) Brazil/2009/Silva et al. (2009)
Country/year/Ref.
Table 1 (Continued)
rK39 ICTk
IFATk Promastigote L. major IFATk Promastigote L. major IFAT¶ NR L. chagasi
Parasitological†
Parasitological†
Parasitological†
Parasitological‡
Parasitological†
Parasitological†
Parasitological†
Parasitological†
Parasitological‡
Parasitological† and serology
76 (ASY/SY/OL)
39 (NR)
39 (NR)
90 (ASY/SY/OL)
5 (ASY/SY)
9 (NR)
9 (NR)
45 (NR)
66 (NR)
50 (ASY/SY/OL)
50 (ASY/SY/OL)
Parasitological† and serology
36 (SY)
112 (ASY/SY/OL)
Parasitological‡ and PCR
IFATk Promastigote L. major IFAT¶ Promastigote L. chagasi IFAT¶ Promastigote L. chagasi IFATk Promastigote L. major IFAT¶ Promastigote L. major IFATNR Promastigote L. major IFATk/(Eluate) Promastigote L. major IFATk Promastigote L. major IFATNR NR
With CVL (clinical condition)
Parasitological‡
Reference standard
Test, parasite form and species
83
100
100
90
100
100
22.2
97.78
83.3
100
98
72
100
Sensitivity %
33 (HDNE)
78 (HDE+OD)
39 (HDE) 78 (HDE+OD)
1302 (DE)
141 (DE)
296 (DE)
296 (DE)
100 (HDNE)
100
61.5
89.7 70.5
52.3
64.7
65.5
97.0
100
92.5
50.0
44 (DE +DNE)
134 (DE+DNE)
95.4
100
74
Specificity %
44 (DE +DNE)
20 (HDE)
67 (DE)
Without CVL (origin)
Tropical Medicine and International Health volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
© 2014 John Wiley & Sons Ltd
© 2014 John Wiley & Sons Ltd
2 groups: ID and HD 1 group: SD (SY) 2 groups: ID and HD 1 Group: SDGCP 2 groups: ID and HD 2 groups: ID and (HD and OD)
2 groups: ID and HD
Phase 2
Phase 3
Phase 1
Phase 3
Phase 2
Phase 2
Phase 2
2 groups: ID and HD
Recruitment
Phase 1
Phase of evaluation
FC-AFPA-IgG¶
DPPk
DPPk
DPPk
FD-DAT¶
FD-DATk
DATk
DAT¶
Test, parasite form and species
Parasitological‡ and serology
Parasitological‡
Parasitological†
Parasitological‡
Parasitological§
Parasitological‡ or PCR
Parasitological‡
Parasitological‡
Reference standard
64 (ASY/SY/OL)
60 (SY) 60 (ASY) 120 (ASY/SY)
39 (NR)
90 (ASY/SY/OL)
26 (SY)
36 (SY)
112 (ASY/SY/OL)
21 (ASY)
With CVL (clinical condition)
97
98 47 72.5
100
91
84
100
93
71
Sensitivity %
100 97.5
70.2
100
91
100
71
Specificity %
25 (HDE)
100
70 (HDNE + OD) 96 – – – –
39 (HDE) 78 (HDE+OD)
1302 (DE)
16 (HDE)
67 (DE)
20 (HDE)
14 (HDNE)
Without CVL (origin)
Recruitment: ID, Infected dogs previously screened for CVL group; HD, Healthy dogs previously screened for non-CVL group; OD, Dogs with other diseases previously screened for non-CVL group; SDGCP, Selected dogs from general canine population; SD, Selected dogs (selection criterion). Test: cELISA, enzyme-linked immunosorbent assay with crude antigens; Dot-ELISA, dot-enzyme-linked immunosorbent assay; Rapid Test ELISA, SNAPâ CLATK Canine Leishmania Antibody Test Kit; ELISA rLdccys1, L. chagasi cysteine proteinase by ELISA; ELISA rLic-NTPDase-2 antigen, Elisa recombinant Leishmania infantum ectonucleoside triphosphate diphosphohydrolase NTPDase-2; rK26 ELISA, recombinant antigens rK-26 ELISA; rK39 ELISA, recombinant antigens rK-39 ELISA; rA2 ELISA, L. donovani rA2 protein; FML–ELISA, purified fucose-mannose ligand ELISA; ElisaMix10, peptides mixed into a single solution; IFAT, indirect fluorescent antibody test; rK39 ICT, rK39 antigen in an immunochromatographic format; DAT, direct agglutination test; FD-DAT, freeze-dried DAT; DPP, dual-path platform; FC-AFPA-IgG, flow cytometry antifixed Leishmania infantum chagasi promastigotes IgG. Parasite: NR, Not reported. Clinical condition: ASY, asymptomatic; SY, symptomatic; OL, oligosymptomatic. Origin: HDNE, healthy dog from non-endemic area; HDE, healthy dog from endemic area; DE, dog from endemic area unrated clinical status; DNE, dog from nonendemic area unrated clinical status. †Simple parasitological test (only one diagnostic technique in a single sample). ‡Compound parasitological test (test carried out by more than one diagnostic technique and/or in more than one sample). §Parasitological test without information about the diagnostic technique or tissue evaluated. ¶In-house test. kCommercial test.
Brazil/1996/ Garcez et al. (1996) Brazil/2007/ Ferreira et al. (2007) Brazil/2006/ da Silva et al. (2006) Venezuela/2007/ Teran-Angel et al. (2007) Brazil/2011/MS (2011b) Brazil/2012/ Alves et al. (2012) Brazil/2012/ Grimaldi et al. (2012) Others Brazil/2009/ Andrade et al. (2009)
Country/year/Ref.
Table 1 (Continued)
Tropical Medicine and International Health volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
9
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
Table 2 Evaluation of the quality of articles on the diagnosis of canine visceral leishmaniasis in terms of risk of bias and applicability following the criteria from QUADAS 2 and STARD Risk of bias QUADAS 2
Reference Dietze et al. (1995) Garcez et al. (1996) Paranhos-Silva et al. (1996) Ros ario et al. (2005) da Silva et al. (2006) Ter an-Angel et al. (2007) Ferreira et al. (2007) Porrozzi et al. (2007) Lemos et al. (2008) C^ andido et al. (2008) Silva et al. (2009) Andrade et al. (2009) Troncarelli et al. (2009) Pinheiro et al. (2009) Camargo et al. (2010) Figueiredo et al. (2010b) Figueiredo et al. (2010a) Faria et al. (2011) Marcondes et al. (2011) MS (2011b) Alves et al. (2012) Solc a et al. (2012) de Souza et al. (2012)
10
Final evaluation of the quality of articles
STARD
Patient selection
Index Reference test standard
Flow & timing
Sampling process†
SE & SP 95% CI‡
Demographic/ clinic§
Risk of bias
Concern regarding applicability
Qualification of the evidence
Low
Low
High
Low
No
No
No
High
Low
Weak
High
Low
High
High
No
No
No
High
Low
Weak
High
High
Low
High
No
Yes
No
High
Low
Weak
High
Low
High
High
No
Yes
No
High
Low
Weak
High
High
High
High
No
No
No
High
Low
Weak
High
High
High
High
No
Yes
No
High
Low
Weak
High
Low
Low
High
No
Yes
Yes
High
Low
Weak
High
High
High
High
No
No
No
High
Low
Weak
High
Low
High
High
No
Yes
No
High
Low
Weak
High
High
High
Low
No
No
No
High
Low
Weak
High
High
High
Low
No
Yes
No
High
Low
Weak
High
Low
High
High
No
No
Yes
High
Low
Weak
High
High
High
Low
No
No
No
High
Low
Weak
High
High
High
High
No
No
No
High
Low
Weak
High
High
High
Low
No
No
No
High
Low
Weak
High
High
High
Low
No
Yes
No
High
Low
Weak
High
High
High
Low
No
No
No
High
Low
Weak
High
High
High
Low
No
No
No
High
Low
Weak
High
High
High
Low
No
No
No
High
Low
Weak
Low High
Low High
Low High
Low High
Yes No
Yes Yes
No No
Low High
Low Low
Strong Weak
High
High
High
Low
No
No
No
High
Low
Weak
High
High
High
High
No
No
No
High
High
Weak
© 2014 John Wiley & Sons Ltd
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
Table 2 (Continued) Risk of bias QUADAS 2
Reference Grimaldi et al. (2012) de Arruda et al. (2013)
Final evaluation of the quality of articles
STARD
Patient selection
Index Reference test standard
Flow & timing
Sampling process†
SE & SP 95% CI‡
Demographic/ clinic§
Risk of bias
Concern regarding applicability
Qualification of the evidence
High
High
High
High
No
No
No
High
Low
Weak
Low
Low
Low
Low
No
Yes
No
Low
Low
Strong
Strong, Strong recommendation for the use of test; Weak, Weak recommendation for the use of intervention. †Description of the sampling process (sampling calculation). ‡Presentation of the confidence intervals for the parameters of sensitivity (SE) and specificity (SP). §Description of the demographic and clinical characteristics of the participants.
Table 3 Combination of the data of sensitivity, specificity, diagnostic odds ratio (DOR) and likelihood ratio, considering all the studies selected for the cELISA, IFAT and DPP tests with their respective heterogeneity measure Likelihood ratio Test
Sensitivity (95% IC)
Specificity (95% IC)
DOR (95% IC)
Positive (95% CI)
Negative (95% CI)
cELISA†
0.890 (0.869–0.909) P < 0.001; I2 = 87.4% 0.881 (0.849–0.907) P < 0.001; I2 = 86.1% 0.835 (0.783–0.879) P < 0.001; I2 = 92.8%
0.870 (0.856–0.883) P < 0.001; I2 = 92.4% 0.630 (0.610–0.649) P < 0.001; I2 = 97.2 0.729 (0.705–0.752) P < 0.001; I2 = 96.9%
119.58 (45.90–311.51) P < 0.001; I2 = 80.7% 109.12 (33.364–356.89) P < 0.001; I2 = 72.2% 80.263 (14.562–442.40) P = 0,011; I2 = 78.0%
8.714 (5.340–14.220) P < 0.001; I2 = 86.6% 4.836 (2.681–8.723) P < 0.001; I2 = 96.4% 11.484 (0.486–273.264) P < 0.001; I2 = 97.7%
0.096 (0.050–0.187) P < 0.001; I2 = 90.4% 0.117 (0.063–0.228) P < 0.001; I2 = 75.6% 0.125 (0.031–0.504) P < 0.001; I2 = 89.9%
IFAT‡ DPP§
†Fifteen studies: Paranhos-Silva et al. (1996), Garcez et al. (1996), Rosario et al. (2005), Ferreira et al. (2007), Lemos et al. (2008), Andrade et al. (2009), Camargo et al. (2010), Solca et al. (2012), Figueiredo et al. (2010a,b), Alves et al. (2012), de Arruda et al. (2013), Porrozzi et al. (2007), C^andido et al. (2008) and Pinheiro et al. (2009). ‡Eleven studies: Paranhos-Silva et al. (1996), Garcez et al. (1996), da Silva et al. (2006), Ferreira et al. (2007), Silva et al. (2009), Troncarelli et al. (2009), Camargo et al. (2010), Figueiredo et al. (2010b), Alves et al. (2012), MS (2011b) and Figueiredo et al. (2010a). §Three studies: MS (2011b), Alves et al. (2012) and Grimaldi et al. (2012).
specificity and two described the clinical and demographic characteristics for the selected dogs. Meta-analysis and synthesised measures Table 3 presents the combined results for 15 studies on immuno-enzymatic tests with crude antigens (cELISA), 11 on the indirect immunofluorescence test (IFAT) and three on the immunochromatographic dual-path platform (DPP). In Figure 2, the exact and interval-based estimates of sensitivity and specificity are demonstrated, with the corresponding measurement synthesised from the tests evaluated. The meta-analysis of the 15 studies that evaluated the accuracy of cELISA presented precise estimates
© 2014 John Wiley & Sons Ltd
on the parameters of sensitivity (0.89, CI 95%: 0.87– 0.91) and specificity (0.87, CI 95%: 0.86–0.88), although the analysis demonstrated high values for heterogeneity between studies. One study (Faria et al. 2011) was excluded from the final analysis because it was considered an outlier reporting 8% sensitivity – clearly different from all other studies, which reported sensitivity higher than 70%. The combined results of the 11 studies on IFAT, despite having high heterogeneity, supplied precise results about sensitivity with 0.88 (CI 95%: 0.85–0.91) and specificity with 0.63 (CI 95%: 0.61–0.65). In Figure 3, the distribution of the studies in the ROC space can be observed, as well as the respective areas under the curve (AUC) and the summarised measures, for the sero11
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
(a)
Sensitivity (95% Cl) Paranhos-Silva et al. 1996 Garcez. et al. 1996 Rosário et al. 2005 Ferreira et al. 2007 Porrozzi et al. 2007 Lemos et al. 2008 Cândido et al. 2008 Pinheiro et al. 2009 Andrade et al. 2009 Camargo et al. 2010 Figueiredo et al. 2010b Figueiredo et al. 2010a Solcà et al. 2012 Alves et al. 2012 Arruda et al. 2013
0
0.2
0.4
0.6
0.8
1
0.98 0.71 0.98 0.96 0.59 0.95 0.88 0.87 0.95 0.93 1.00 1.00 0.83 1.00 0.92
(b)
Specificity (95% Cl) Paranhos-Silva et al. 1996 Garcez. et al. 1996 Rosário et al. 2005 Ferreira et al. 2007 Porrozzi et al. 2007 Lemos et al. 2008 Cândido et al. 2008 Pinheiro et al. 2009 Andrade et al. 2009 Camargo et al. 2010 Figueiredo et al. 2010b Figueiredo et al. 2010a Solcà et al. 2012 Alves et al. 2012 Arruda et al. 2013
(0.88 – 1.00) (0.48 – 0.89) (0.93 – 1.00) (0.90 – 0.99) (0.49 – 0.69) (0.87 – 0.99) (0.77 – 0.95) (0.81 – 0.91) (0.87 – 0.99) (0.82 – 0.99) (0.66 – 1.00) (0.48 – 1.00) (0.59 – 0.96) (0.91 – 1.00) (0.85 – 0.96)
Pooled sensitivity = 0.89 (0.87 to 0.91) Chi-square = 111.13; df = 14 (P = 0.0000) Inconsistency (I-square) = 87.4 %
0
0.2
Sensitivity Sensitivity (95% Cl) Paranhos-Silva et al. 1996 Garcez et al. 1996 da Silva et al. 2006 Ferreira et al. 2007 Silva et al. 2009 Troncarelli et al. 2009 Camargo et al. 2010 Figueiredo et al. 2010b Figueiredo et al. 2010a MS. 2011 Alves et al. 2012
0.2
0.4
0.6
0.8
1
0.78 1.00 1.00 0.72 1.00 0.83 0.98 1.00 1.00 0.90 1.00
Pooled sensitivity = 0.88 (0.85 to 0.91) Chi-square = 71.87; df = 10 (P = 0.0000) Inconsistency (I-square) = 86.1 %
MS. 2011 Alves et al. 2012 Grimaldi et al. 2012
0.4 0.6 Sensitivity
1
Pooled specificity = 0.87 (0.86 to 0.88) Chi-square = 184.13; df = 14 (P = 0.0000) Inconsistency (I-square) = 92.4 %
(d)
Specificity (95% Cl) Paranhos-Silva et al. 1996 Garcez et al. 1996 da Silva et al. 2006 Ferreira et al. 2007 Silva et al. 2009 Troncarelli et al. 2009 Camargo et al. 2010 Figueiredo et al. 2010b Figueiredo et al. 2010a MS. 2011 Alves et al. 2012
1.00 1.00 0.75 1.00 0.50 0.93 1.00 0.65 0.64 0.52 0.71
(0.96 – 1.00) (0.77 – 1.00) (0.63 – 0.84) (0.83 – 1.00) (0.35 – 0.65) (0.87 – 0.96) (0.96 – 1.00) (0.59 – 0.70) (0.56 – 0.72) (0.49 – 0.55) (0.59 – 0.80)
Pooled specificity = 0.63 (0.61 to 0.65) Chi-square = 356.65; df = 10 (P = 0.0000)
0
0.2
0.4
0.6
0.8
1 Inconsistency (I-square) = 97.2 %
Specificity
(e)
0.2
0.8
(0.64 – 0.89) (0.84 – 1.00) (0.90 – 1.00) (0.63 – 0.80) (0.93 – 1.00) (0.72 – 0.91) (0.88 – 1.00) (0.66 – 1.00) (0.48 – 1.00) (0.82 – 0.95) (0.91 – 1.00)
Sensitivity
0
0.6
(0.95 – 1.00) (0.57 – 0.98) (0.86 – 1.00) (0.64 – 0.91) (0.73 – 0.96) (0.89 – 1.00) (0.88 – 1.00) (0.55 – 0.78) (0.86 – 1.00) (0.95 – 1.00) (0.94 – 0.98) (0.92 – 0.99) (0.39 – 0.78) (0.56 – 0.78) (0.82 – 0.86)
Specificity
(c)
0
0.4
0.99 0.86 1.00 0.80 0.87 1.00 1.00 0.68 1.00 0.99 0.97 0.97 0.59 0.68 0.84
0.8
1
Sensitivity (95% Cl) 0.91 (0.83 – 0.96) 1.00 (0.91 – 1.00) 0.73 (0.64 – 0.80)
Pooled sensitivity = 0.84 (0.78 to 0.88) Chi-square = 27.61; df = 2 (P = 0.0000) Inconsistency (I-square) = 92.8 %
(f) MS. 2011 Alves et al. 2012 Grimaldi et al. 2012
Specificity (95% Cl) 0.70 (0.68 – 0.73) 0.97 (0.91 – 1.00) 0.96 (0.88 – 0.99)
Pooled specificity = 0.73 (0.71 to 0.75) Chi-square = 64.80; df = 2 (P = 0.0000)
0
0.2
0.4 0.6 Specificity
0.8
1 Inconsistency (I-square) = 96.9 %
Figure 2 Forest plots of the sensitivity and specificity of the cELISA (a and b), IFAT (c and d) and DPP (e and f) tests with their respective summarised measures.
logical tests cELISA (AUC = 0.9677), IFAT (AUC = 0.9654) and DPP (AUC = 0.9566) in accordance with the data in Table 3. Subgroup analyses were carried out in accordance with characteristics relative to selection of dogs, clinical status and the availability of tests as commercial kits, for the cELISA and IFAT methods. Subgroup analysis demonstrated the influence of the pre-screening recruitment strategy that overestimates the specificity of cELISA and IFAT in samples of healthy dogs from non-endemic areas. The influence of clinical status was also evident, worsening the sensitivity of the cELISA tests estimate in asymptomatic dogs. Finally, a slightlbetter sensitivity of commercial kits was evident. Although heterogeneity of the estimates between studies diminished in the analysis of specific subgroups, it remained relevant in all the 12
analysis with the exception of the sensitivity of cELISA in symptomatic dogs (Figure 4) and the specificity of IFAT in healthy dogs from non-endemic areas (Figure 5). Discussion Our research presents data that state one of the most challenging problems in relation to the CVL diagnosis, with the consequent concern about the human population exposed to risk of infection by L. infantum. The scope or our research question was restricted to the Americas aiming at the production of an evidence to be useful for VL control programmes in the region and the final results reflected the situation of the serological diagnostic tests in Brazil, bearing in mind that despite an ample search in the literature, only one study outside Brazil was found.
© 2014 John Wiley & Sons Ltd
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
(a)
(b) SROC curve
Sensitivity 1
SROC curve
Sensitivity 1 Symmetric SROC AUC = 0.9677 SE(AUC) = 0.0119 Q* = 0.9162 SE(Q*) = 0.0188
0.9 0.8
Symmetric SROC AUC = 0.9654 SE(AUC) = 0.0157 Q* = 0.9126 SE(Q*) = 0.0241
0.9 0.8
0.7
0.7
0.6
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1 0
0 0
0.2
0.4
0.6
0.8
1
0
1-specificity
(c)
0.2
0.4
0.6
0.8
1
1-specificity
SROC curve
Sensitivity 1
Symmetric SROC AUC = 0.9566 SE(AUC) = 0.0278 Q* = 0.8996 SE(Q*) = 0.0393
0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0
0.2
0.4
0.6
0.8
1
1-specificity
Figure 3 Summarised ROC curve (SROC) of the cELISA (a), IFAT (b) and DPP (c) tests. The points represent the individual studies used in the meta-analyses.
This finding may be a result of the higher incidence and broader distribution of CVL in Brazil, which has the continent’s heaviest burden of the disease as well as of human VL (Alvar et al. 2012). In spite of the limitations of this geographically restricted approach, the potential inclusion of evidence from the Mediterranean basin could introduce heterogeneity from other sources such as the L. infantum diversity in the Old World (A€ıt-Oudhia et al. 2011; Gouzelou et al. 2013) compared with the apparently more homogeneous parasite population in the New World (Ferreira et al. 2012), the peculiarities of transmission cycles involving different vectors, seasonality patterns (Mazeris et al. 2010; Ntais et al. 2013), the sympatric circulation of other Leishmania species and
© 2014 John Wiley & Sons Ltd
other trypanosomatids and current prevention practices that could have an impact on exposure and specific immunological response to infection (Dantas-Torres et al. 2012). Brazil presents a heterogeneous situation with regard to the prevalence of CVL (Dantas-Torres et al. 2006; Almeida et al. 2009; Prado et al. 2011; Belo et al. 2013; Costa et al. 2013), creating an additional challenge, especially for areas with low prevalence, given that the synthesised measures identified moderate accuracy in the available tests and that in low-prevalence circumstances, they would certainly present a less-thandesirable performance, with a low positive predictive value. 13
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
(a) Previously screened Infected dogs
0
0.2
0.4
0.6
Specificity (95% CI) Paranhos–Sliva et al. 1996 Garcez. et al. 1996 Rosário et al. 2005 Ferreira et al. 2007 Porrozzi et al. 2007 Cândido et al. 2008 Lemos et al. 2008 Andrade et al. 2009 Camargo et al. 2010 Alves et al. 2010
(0.88 – 1.00) (0.48 – 0.89) (0.93 – 1.00) (0.49 – 0.69) (0.90 – 0.99) (0.77 – 0.95) (0.87 – 0.99) (0.81 – 0.91) (0.87 – 0.99) (0.82 – 0.99) (0.91 – 1.00)
0.98 0.71 0.98 0.59 0.96 0.88 0.95 0.87 0.95 0.93 1.00
Pooled sensitivity = 0.89 (0.86 to 0.91) Chi-square = 106.34; df = 10 (P = 0.0000) 1 Inconsistency (I-square) = 90.6%
0.8
Healthy dogs
Sensitivity (95% CI)
Paranhos–Sliva et al. 1996 Garcez. et al. 1996 Rosário et al. 2005 Porrozzi et al. 2007 Ferreira et al. 2007 Cândido et al. 2008 Lemos et al. 2008 Pinheiro et al. 2009 Andrade et al. 2009 Camargo et al. 2010 Alves et al. 2010
0
0.2
0.4
0.6
0.8
0.99 0.86 1.00 1.00 1.00 1.00 1.00 1.00 0.99 0.85
(0.95 – 1.00) (0.57 – 0.98) (0.86 – 1.00) (0.83 – 1.00) (0.86 – 1.00) (0.88 – 1.00) (0.89 – 1.00) (0.86 – 1.00) (0.95 – 1.00) (0.64 – 0.94)
Pooled specificity = 0.98 (0.96 to 0.99) Chi-square = 26.76; df = 9 (P = 0.0015) 1 Inconsistency (I-square) = 66.4%
Specificity
Sensitivity
(b)
General population
Sensitivity (95% CI)
Specificity (95% CI) Figueiredo et al. 2010b 0.97 (0.94 – 0.98) Arruda et al. 2013 0.84 (0.82 – 0.86)
Figueiredo et al. 2010b 1.00 (0.66 – 1.00) Arruda et al. 2013 0.92 (0.85 – 0.96)
0
0.2
0.4
0.6
Pooled sensitivity = 0.93 (0.86 to 0.97) Chi-square = 1.46; df = 1 (P = 0.2263) 1 Inconsistency (I-square) = 31.7 %
0.8
0
0.2
0.4
0.6
0.8
Pooled specificity = 0.86 (0.84 to 0.88) Chi-square = 43.65; df = 1 (P = 0.0000) 1 Inconsistency (I-square) = 97.7 %
Specificity
Sensitivity
(c)
All dogs
Rosário et al. 2005 Ferreira et al. 2007 Andrade et al. 2009 Figueiredo et al. 2010b Figueiredo et al. 2010a Solcà et al. 2012 Alves et al. 2012 Arruda et al. 2013
Endemic area
Healthy dogs
Specificity (95% CI) 1.00 (0.86 – 1.00) 0.80 (0.64 – 0.91) 1.00 (0.86 – 1.00) 0.97 (0.94 – 0.98) 0.97 (0.92 – 0.99) 0.59 (0.39 – 0.78) 0.68 (0.56 – 0.78) 0.84 (0.82 – 0.86)
Rosário et al. 2005 Ferreira et al. 2007 Andrade et al. 2009 Alves et al. 2012
0
0.2
0.4
0.6
Specificity (95% CI) 1.00 (0.86 – 1.00) 1.00 (0.83 – 1.00) 1.00 (0.86 – 1.00) 0.85 (0.69 – 0.94)
Pooled specificity = 0.94 (0.88 to 0.98) Chi-square = 12.97; df = 3 (P = 0.0047) 1 Inconsistency (I-square) = 76.9 %
0.8
Specificity
0
0.2
0.4
0.6
Pooled specificty = 0.86 (0.84 to 0.88) Chi-square = 106.46; df = 7 (P = 0.0000) 1 Inconsistency (I-square) = 93.4 %
0.8
Specificity
(d)
All dogs
Non-endemic area
Healthy dogs
Specificity (95% CI) Paranhos–Sliva et al. 1996 0.99 (0.95 – 1.00) Garcez. et al. 1996 0.86 (0.57 – 0.98) Porrozzi et al. 2007 0.87 (0.73 – 0.96) 1.00 (0.89 – 1.00) Lemos et al. 2008 1.00 (0.88 – 1.00) Cândido et al. 2008 0.68 (0.55 – 0.78) Pinheiro et al. 2009 0.99 (0.95 – 1.00) Camargo et al. 2010
0
0.2
0.4
0.6
Pooled specificty = 0.92 (0.89 to 0.94) Chi-square = 66.39; df = 6 (P = 0.0000) 1 Inconsistency (I-square) = 91.0 %
0.8
Specificity
(e)
Symptomatic and asymptomatic Sensitivity (95% CI) Rosário et al. 2005 Porrozzi et al. 2007 Ferreira et al. 2007 Lemos et al. 2008 Andrade et al. 2009 Pinheiro et al. 2009 Figueiredo et al. 2010a Solcà et al. 2012 Arruda et al. 2013
0
0.2
0.4
0.6
0.98 0.59 0.96 0.95 0.95 0.87 1.00 0.83 0.92
(0.93 – 1.00) (0.49 – 0.69) (0.90 – 0.99) (0.87 – 0.99) (0.87 – 0.99) (0.81 – 0.91) (0.48 – 1.00) (0.59 – 0.96) 0 (0.85 – 0.96)
Specificity (95% CI) Paranhos–Sliva et al. 1996 0.99 (0.95 – 1.00) Garcez. et al. 1996 0.86 (0.57 – 0.98) Porrozzi et al. 2007 1.00 (0.86 – 1.00) Lemos et al. 2008 1.00 (0.89 – 1.00) Cândido et al. 2008 1.00 (0.88 – 1.00) Camargo et al. 2010 0.99 (0.95 – 1.00)
0
0.2
0.4
0.6
Pooled specificty = 0.99 (0.97 to 1.00) Chi-square = 8.67; df = 5 (P = 0.1230) 1 Inconsistency (I-square) = 42.3 %
0.8
Specificity
Clinical condition Symptomatic Porrozzi et al. 2007 Cândido et al. 2008
0.2
0.4
0.6
0.8
Asymptomatic Sensitivity (95% CI) 0.88 (0.76 – 0.95) 0.88 (0.77 – 0.95)
Pooled sensitivity = 0.88 (0.81 to 0.94) Chi-square = 0.00; df = 1 (P = 0.9570) 1Inconsistency (I-square) = 0.0%
Garcez. et al. 1996 Porrozzi et al. 2007
0
Sensitivity
0.2
0.4
0.6
0.8
Sensitivity (95% CI) 0.71 (0.48 – 0.89) 0.30 (0.18 – 0.45)
Pooled sensitivity = 0.42 (0.31 to 0.55) Chi-square = 10.50; df = 1 (P = 0.0012) 1 Inconsistency (I-square) = 90.5 %
Sensitivity
Pooled sensitivity = 0.88 (0.86 to 0.90) Chi-square = 88.13; df = 8 (P = 0.0000) 1 Inconsistency (I-square) = 90.9%
0.8
Sensititvity
(f)
Commercial test (L.major crude antigen)
Sensitivity (95% CI) Andrade et al. 2009 0.95 (0.87 – 0.99) Camargo et al. 2010 0.93 (0.89 – 1.00) Figueiredo et al. 2010b 1.00 (0.66 – 1.00) Alves et al. 2012 1.00 (0.91 – 1.00) Arruda et al. 2013 0.92 (0.85 – 0.96)
0
0.2
0.4
0.6
0.8
Pooled sensitivity = 0.95 (0.91 to 0.97) Chi-square = 6.80; df = 4 (P = 0.1469) 1 Inconsistency (I-square) = 41.2%
Sensitivity
Andrade et al. 2009 Camargo et al. 2010 Figueiredo et al. 2010b Alves et al. 2012 Arruda et al. 2013
0
0.2
0.4
0.6
0.8
Specificity (95% CI) 1.00 (0.86 – 1.00) 0.99 (0.95 – 1.00) 0.97 (0.94 – 0.98) 0.68 (0.56 – 0.78) 0.84 (0.82 – 0.86)
Pooled specificity = 0.86 (0.85 to 0.88) Chi-square = 90.17; df = 4 (P = 0.0000) 1 Inconsistency (I-square) = 95.6%
Specificity
Figure 4 Forest plots of the sensitivity and specificity of the cELISA test, in accordance with the subgroup analysis.
Diagnostic tests for CVL that have high accuracy take on a crucial role, given that the strategies that propose control of CVL as a measure to reduce human VL incidence depend on appropriate identification of 14
infected dogs in situations where the elimination of seropositive dogs already tends to be imperfect (Costa et al. 2013). This could partly explain the failure of control measures in containing geographical dispersion
© 2014 John Wiley & Sons Ltd
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
(a) Previously screened Infected dogs
Healthy dogs Specificity (95% Cl)
Sensitivity (95% Cl) Paranhos-Silva et al. 1996 Garcez et al. 1996 Ferreira et al. 2007 Silva et al. 2009 Camargo et al. 2010 Alves et al. 2012
0
0.2
0.4
0.6
0.8
0.78 1.00 0.72 1.00 0.98 1.00
Paranhos-Silva et al. 1996 Garcez et al. 1996 Ferreira et al. 2007 Camargo et al. 2010 Alves et al. 2012
(0.64 – 0.89) (0.84 – 1.00) (0.63 – 0.80) (0.93 – 1.00) (0.88 – 1.00) (0.91 – 1.00)
Pooled sensitivity = 0.87 (0.82 to 0.90) Chi-square = 56.91; df = 5 (P = 0.0000) 1 Inconsistency (I-square) = 91.2 %
0
0.2
0.4
0.6
0.8
1
1.00 1.00 1.00 1.00 0.90
(0.96 – 1.00) (0.77 – 1.00) (0.83 – 1.00) (0.96 – 1.00) (0.76 – 0.97)
Pooled specificity = 0.99 (0.96 to 1.00) Chi-square = 15.99; df = 4 (P = 0.0030) Inconsistency (I-square) = 75.0 %
Specificity
Sensitivity
(b) General population Specificity (95% Cl)
Sensitivity (95% Cl) Figueiredo et al. 2010b MS. 2011
0
0.2
0.4
0.6
1.00 0.90
(0.66 – 1.00) (0.82 – 0.95)
Pooled sensitivity = 0.91 (0.83 to 0.96) Chi-square = 1.80; df = 1 (P = 0.1793) 1 Inconsistency (I-square) = 44.5 %
0.8
Figueiredo et al. 2010b MS. 2011
0
0.2
0.4
0.6
0.8
0.65 0.52
(0.59 – 0.70) (0.49 – 0.55)
Pooled specificity = 0.54 (0.52 to 0.57) Chi-square = 16.36; df = 1 (P = 0.0001) 1 Inconsistency (I-square) = 93.9 %
Specificity
Sensitivity
(c) Endemic area All dogs
(d) Non-endemic area Healthy dogs Specificity (95% Cl)
Specificity (95% Cl) Ferreira et al. 2007 Silva et al. 2009 Troncarelli et al. 2009 Figueiredo et al. 2010b Figueiredo et al. 2010a MS. 2011 Alves et al. 2012
1.00 0.50 0.93 0.65 0.64 0.52 0.71
Paranhos-Silva et al. 1996 Garcez et al. 1996 Camargo et al. 2010
(0.83 – 1.00) (0.35 – 0.65) (0.87 – 0.96) (0.59 – 0.70) (0.56 – 0.72) (0.49 – 0.55) (0.59 – 0.80) 0
0.2
0.4
0.6
0.8
1
1.00 1.00 1.00
(0.96 – 1.00) (0.77 – 1.00) (0.96 – 1.00)
Pooled specificity = 1.00 (0.98 to 1.00) Chi-square = 0.00; df = 2 (P = 1.0000) Inconsistency (I-square) = 0.0 %
Specificity
0
0.2
0.4 0.6 Specificity
0.8
Pooled specificity = 0.59 (0.56 to 0.61) Chi-square = 136.51; df = 6 (P = 0.0000) 1 Inconsistency (I-square) = 95.6 %
(e) Commercial test (Promastigote L. major) Specificity (95% Cl)
Sensitivity (95% Cl) da Silva et al. 2006 Silva et al. 2009 Figueiredo et al. 2010b MS. 2011 Alves et al. 2012
0
0.2
0.4
0.6
0.8
1.00 1.00 1.00 0.90 1.00
da Silva et al. 2006 Silva et al. 2009 Figueiredo et al. 2010b MS. 2011 Alves et al. 2012
(0.90 – 1.00) (0.93 – 1.00) (0.66 – 1.00) (0.82 – 0.95) (0.91 – 1.00)
Pooled sensitivity = 0.96 (0.93 to 0.98) Chi-square = 16.98; df = 4 (P = 0.0020) 1 Inconsistency (I-square) = 76.4 %
Sensitivity
0
0.2
0.4
0.6
0.8
0.75 0.50 0.65 0.52 0.71
(0.63 – 0.84) (0.35 – 0.65) (0.59 – 0.70) (0.49 – 0.55) (0.59 – 0.80)
Pooled specificity = 0.56 (0.53 to 0.58) Chi-square = 35.58; df = 4 (P = 0.0000) Inconsistency (I-square) = 88.8 % 1
Specificity
Figure 5 Forest plots of the sensitivity and specificity of the IFAT test, in accordance with the subgroup analysis.
of CVL on the continent (Salom on et al. 2009; Gould et al. 2013). Among the methodological problems identified in the studies, it is notable that the majority use dogs that were previously screened in relation to their infection or noninfection status using other serological tests, affecting the accuracy of the tests which could be overestimated (Whiting et al. 2011). A number of studies do not explain whether the test under evaluation and the reference standard were interpreted without previous knowledge of one
© 2014 John Wiley & Sons Ltd
another, as well as the fact that some do not use the same reference standard for all the dogs, substantially raising the risk of bias (Whiting et al. 2011). Unlike other recent studies (Maia et al. 2012; Quinnell et al. 2013), we systematically evaluated the evidence base, including aspects of quality and recommendability as proposed in QUADAS 2, STARD and GRADE. Of 25 studies evaluated, only two (MS 2011b; de Arruda et al. 2013) presented simultaneously a low risk of bias, a low risk for applicability and good performance in the 15
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
questions proposed by STARD, indicating that the results identified by these studies constitute strong evidence. However the low risk of bias related to the reference standard should be taken with precaution because of the limitations of the compound parasitological test. Certainly, considering the magnitude of the canine disease and the impact that the canine form has on human health, the number of reasonably conducted studies is very small. Although various studies that use a diversity of serological techniques to diagnose CVL were identified, the scarcity of studies for some techniques prevented us from carrying out meta-analysis. Only the cELISA, IFAT and DPP tests could be combined, although they presented a high heterogeneity in the general combination, justifying the stratified analysis presented in Figures 4 and 5. There was ample variation in the sensitivity of the cELISA test, situated between 8.0% and 100%, with specificity between 60% and 100% (Table 1). The study that presented the best evidence (de Arruda et al. 2013) demonstrated sensitivity and specificity for cELISA (L. major) of 91.8% (CI 95%: 86.42–97.26) and 83.75% (CI 95%: 81.76–85.74). Unfortunately, the evaluation of other immuno-enzymatic tests was compromised by the scarcity and low quality of the studies. For cELISA, lower heterogeneity was observed in the stratified evaluation for some subgroups, but it remained relevant for all with the exception for sensitivity in symptomatic dogs (Figure 4). The good performance of the test could be observed in the data that summarise the performance in the SROC curve (Figure 3a), with the AUC and the Q index corresponding to 0.97 and 0.92. The heterogeneity observed could be attributed to the variations in the antigens used and the peculiarities in the running of the assays, many of which had antigens produced in house (Garcez et al. 1996; Lemos et al. 2008; Alves et al. 2012). However, the variations in the strategy for selecting the non-cases seem to have had a greater impact, mainly in the evaluation of the specificity, as shown in the stratified analysis. The results in Table 1 demonstrate that the sensitivity of IFAT varied between 83.3% and 100% and specificity between 50% and 100%. However, the best study evaluated (strong evidence) presented sensitivity of 90% (CI 95%: 82.38–95.1) and specificity of 52.27% (CI 95%: 49.5–55.0) for the IFAT which used antigens from L. major (MS 2011b). In contrast, the IFAT in a sample obtained on filter paper evaluated in another study (Figueiredo et al. 2010b) presented low sensitivity 22.2% (3.9–59.8) and high specificity corresponding to 97.0% (94.1–98.5), demonstrating the lack of consistency in the results obtained among different studies. 16
For IFAT, more homogeneous meta-analysed results were identified in Figure 5, with prominence given to: sensitivity in dogs from the general population and specificity among dogs from the non-endemic area. The value of the AUC (0.96) and of the Q index (0.91) presented in Figure 3b indicates the excellent performance of the test. In the evaluation of performance, the DPP received a strong recommendation in only one study (MS 2011b) (Tables 1 and 2), with a sensitivity of 91% (CI 95%: 83.60–95.80) and a specificity of 70.2% (CI 95%: 67.68–72.65). Although a satisfactory performance in the evaluation of quality was not obtained, two other studies (Alves et al. 2012; Grimaldi et al. 2012) on the rapid DPP test demonstrated similar results, albeit with a reduction in sensitivity among asymptomatic infected dogs (Grimaldi et al. 2012). The meta-analysis of the DPP presented in Table 3 was carried out with evidence arising from three studies (MS 2011b; Alves et al. 2012; Grimaldi et al. 2012). However, the impossibility of stratifying the result due to the small number of studies, the heterogeneity presented and the high imprecision observed in the SROC curve (Figure 3c) indicate a precautious use of the meta-analysed measures. Thus, the results from primary studies, with strong recommendability, can supply more appropriate measures of accuracy than the combined data. Studies carried out with immuno-enzymatic assays with the recombinant antigens used in the rapid tests showed satisfactory results, such as those obtained by Ros ario et al. (2005), in the rK26 ELISA and rK39 ELISA tests, with sensitivity of 99.1% (CI 95%: 94.1–100.0) and 98.1% (CI 95%: 92.7–99.7) and specificity of 100% (CI 95%: 83.4– 100.0) and 100% (CI 95%: 83.4–100.0). And again, the study by Porrozzi et al. (2007) showed lower sensitivity when evaluating the ELISA tests with the aforementioned recombinant antigens among asymptomatic dogs. In this context, our review identified evidence produced in the last 2 years that used more rigorous methodological designs. These were a study to validate a rapid test based on recombinant antigens, an evaluation of two ELISA assays, one with crude antigen of L. major and one with L. infantum, and an evaluation of IFAT; the last two of these are used as routine in laboratories in Brazil. These studies overcame most of the methodological limitations raised in the review by Romero and Boelaert (2010), especially those linked to the sample size and selection biases, thus establishing more precise estimates of sensitivity and specificity for these tests (MS 2011b; de Arruda et al. 2013). Schubach (2011) evaluated a subgroup of samples from the large multicentric Brazilian study (MS 2011b) demonstrating the reproducibility of the DPP in blood and in serum, which qualifies it for use both in the field and
© 2014 John Wiley & Sons Ltd
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
under laboratory conditions. Another study highlighted the performance of the DPP due to its potential to reduce the time between diagnosis and canine euthanasia, a fact that may increase the effectiveness of the strategy based on euthanasia for infected dogs (Grimaldi et al. 2012). Conclusions The meta-analysis demonstrated that cELISA and DPP tests have moderate accuracy for the diagnosis of CVL, and the evaluation of the quality of evidence demonstrated that there is an urgent need to improve the quality of the design, implementation and analysis of validation studies on diagnostic tests for this disease. The recommendation for use of the evaluated tests is mainly based on evidence that is scarce and restricted to the Brazilian experience. Future validation studies should include samples from dogs representative of affected areas in countries other than Brazil. Also, there is the need to validate algorithms for the combined use of nowadays available diagnostic tests, which will allow for an increase in diagnostic accuracy and at the same time will make their use under field conditions easier. The definition of the reference standard for the diagnosis of CVL is an important challenge that continues to affect validation studies. In this respect, the role of molecular techniques should be thoroughly studied to evaluate the possibility of overcoming the limitations of the conventional parasitological diagnosis. Also, phase 3 studies as well as those on the reproducibility of techniques should be encouraged, because it is this aspect that has received the least attention and consequently presents little evidence in the literature. Finally, the recording of relevant data on other potential sources of heterogeneity such as vaccination status or migration history could improve the quality of future evidences and their appropriateness for further meta-analytical studies. Acknowledgements The authors thank Francisco Edilson Ferreira de Lima J unior, from the Unit for Surveillance of Vector Transmitted Diseases, of the Brazilian Ministry of Health, for supporting the development of this research. GASR received a visiting fellowship from the Strategic Program for Science, Technology & Innovation of FAPEAM (PECTI-SAUDE). This study was partially funded by the Brazilian Ministry of Health and the Pan American Health Organization (Contract PF BR/CNT/1201280.001 -TC74).
© 2014 John Wiley & Sons Ltd
References A€ıt-Oudhia K, Harrat Z, Benikhlef R, Dedet JP & Pratlong F (2011) Canine Leishmania infantum enzymatic polymorphism: a review including 1023 strains of the Mediterranean area, with special reference to Algeria. Acta Tropica 118, 80–86. Almeida ABPF, Faria RP, Pimentel MFA, Dahroug MAA, Turbino NCMR & Sousa VRF (2009) Inquerito soroepidemiol ogico de leishmaniose canina em areas end^emicas de Cuiaba, Estado de Mato Grosso. Revista da Sociedade Brasileira de Medicina Tropical 42, 156–159. Alvar J, Velez ID, Bern C et al. (2012) Leishmaniasis worldwide and global estimates of its incidence. PLoS One 7, e35671. Alves S, Mouta-Confort E, Figueiredo FB, Oliveira RVC, Schubach O & Madeira MF (2012) Evaluation of serological cross-reactivity between canine visceral leishmaniasis and natural infection by Trypanosoma caninum. Research in Veterinary Science 93, 1329–1333. Andrade RA, Silva Ara ujo MS, Reis AB et al. (2009) Advances in flow cytometric serology for canine visceral leishmaniasis: diagnostic applications when distinct clinical forms, vaccination and other canine pathogens become a challenge. Veterinary Immunology and Immunopathology 128, 79–86. de Ara ujo VEM, Morais MHF, Reis IA, Rabello A & Carneiro M (2012) Early clinical manifestations associated with death from visceral leishmaniasis. PLoS Neglected Tropical Diseases 6, e1511. de Arruda MM, Figueiredo FB, Cardoso FA et al. (2013) Validity and reliability of enzyme immunoassays using Leishmania major or L. infantum antigens for the diagnosis of canine visceral leishmaniasis in Brazil. PLoS One 8, e69988. Bauzer LGSR, Souza N, Maingon RDC & Peixoto A (2007) Lutzomyia longipalpis in Brazil: a complex or a single species? A mini-review. Mem orias do Instituto Oswaldo Cruz 102, 1–12. Belo VS, Struchiner CJ, Werneck GL et al. (2013) A systematic review and meta-analysis of the factors associated with Leishmania infantum infection in dogs in Brazil. Veterinary Parasitology 195, 1–13. Boelaert M, Bhattacharya S, Chappuis F et al. (2007) Evaluation of rapid diagnostic tests: visceral leishmaniasis. Nature Reviews Microbiology 11, 30–39. Bossuyt PM, Reitsma JB, Bruns DE et al. (2003) The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Annals of Internal Medicine 138, 1–12. Braga ASC, Toledo Junior ACC & Rabello A (2013) Factors of poor prognosis of visceral leishmaniasis among children under 12 years of age. A retrospective monocentric study in Belo Horizonte, State of Minas Gerais, Brazil, 2001–2005. Revista da Sociedade Brasileira de Medicina Tropical 46, 55–59. Camargo J, Langoni H, Troncarelli M, Machado J, Lucheis S & Padovani C (2010) Performance of IFAT, ELISA, direct parasitological examination and PCR on lymph node aspirates for canine visceral leishmaniasis diagnosis. Journal of Venomous Animals and Toxins Including Tropical Diseases 16, 414–420.
17
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
C^ andido TC, Perri SV, Gerzoschkwitz TDO, Luvizotto MCR & Lima VMF (2008) Comparative evaluation of enzyme-linked immunosorbent assay based on crude and purified antigen in the diagnosis of canine visceral leishmaniasis in symptomatic and oligosymptomatic dogs. Veterinary Parasitology 157, 175–181. Costa DNCC, Codecßo CT, Silva M & Werneck GL (2013) Culling dogs in scenarios of imperfect control: realistic impact on the prevalence of canine visceral leishmaniasis. PLoS Neglected Tropical Diseases 7, e2355. Dantas-Torres F (2007) The role of dogs as reservoirs of Leishmania parasites, with emphasis on Leishmania (Leishmania) infantum and Leishmania (Viannia) braziliensis. Veterinary Parasitology 149, 139–146. Dantas-Torres F & Brand~ao-Filho SP (2006) Visceral leishmaniasis in Brazil: revisiting paradigms of epidemiology and control. Revista do Instituto de Medicina Tropical de S~ao Paulo 48, 151–156. Dantas-Torres F, de Brito MEF & Brand~ao-Filho SP (2006) Seroepidemiological survey on canine leishmaniasis among dogs from an urban area of Brazil. Veterinary Parasitology 140, 54–60. Dantas-Torres F, Solano-Gallego L, Baneth G, Ribeiro VM, de Paiva-Cavalcanti M & Otranto D (2012) Canine leishmaniosis in the Old and New Worlds: unveiled similarities and differences. Trends in Parasitology 28, 531–538. Deville WL, Buntinx F, Bouter LM et al. (2002) Conducting systematic reviews of diagnostic studies: didactic guidelines. BMC Medical Research Methodology 2, 1–13. Dietze R, Falqueto A, Valli L, Rodriques TP, Boulos M & Corey R (1995) Diagnosis of canine visceral leishmaniasis with a dot-enzyme-linked immunosorbent assay. The American Journal of Tropical Medicine and Hygiene 53, 40–42. Faria AR, Costa MM, Giusta MS et al. (2011) High-throughput analysis of synthetic peptides for the immunodiagnosis of canine visceral leishmaniasis. PLoS Neglected Tropical Diseases 5, e1310. Ferreira EDC, de Lana M, Carneiro M et al. (2007) Comparison of serological assays for the diagnosis of canine visceral leishmaniasis in animals presenting different clinical manifestations. Veterinary Parasitology 146, 235–241. Ferreira GE, dos Santos BN, Dorval ME et al. (2012) The genetic structure of Leishmania infantum populations in Brazil and its possible association with the transmission cycle of visceral leishmaniasis. PLoS One 7, e36242. Figueiredo FB, Madeira MF, Menezes RC et al. (2010a) Efficacy of an indirect immunofluorescence test in the diagnosis of canine leishmaniosis. The Veterinary Journal, 186, 123–124. Figueiredo FB, Madeira MF, Nascimento LD et al. (2010b) Canine visceral leishmaniasis: study of methods for the detection of IgG in serum and eluate samples. Revista do Instituto de Medicina Tropical de S~ao Paulo 52, 193–196. Garcez LM, Shaw JJ & Silveira FT (1996) Direct agglutination tests in the serodiagnosis of visceral leishmaniasis in the state of Par a. Revista da Sociedade Brasileira de Medicina Tropical 29, 165–180. Gould IT, Perner MS, Santini MS et al. (2013) Visceral leishmaniasis in Argentina. Cases notification and distribution of vectors (2006–2012). Medicina 73, 104–110.
18
Gouzelou E, Haralambous C, Antoniou M et al. (2013) Genetic diversity and structure in Leishmania infantum populations from southeastern Europe revealed by microsatellite analysis. Parasites and Vectors 6, 342. Grimaldi G, Teva A, Ferreira AL et al. (2012) Evaluation of a novel chromatographic immunoassay based on Dual-Path Platform technology (DPPâ CVL rapid test) for the serodiagnosis of canine visceral leishmaniasis. Transactions of the Royal Society of Tropical Medicine and Hygiene 106, 54–59. Guyatt GH, Oxman AD, Vist GE et al. (2008) GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ (Clinical Research Ed.) 336, 924–926. Higgins JPT, Thompson SG, Deeks JJ & Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ (Clinical Research Ed.) 327, 557–560. Horvath AR (2009) Grading quality of evidence and strength of recommendations for diagnostic tests and strategies. Clinical Chemistry 55, 853–855. Lemos EM, Laurenti MD, Moreira MAB et al. (2008) Canine visceral leishmaniasis: performance of a rapid diagnostic test (Kalazar Detect) in dogs with and without signs of the disease. Acta Tropica 107, 205–207. Lima IP, M€ uller MC, Holanda TA, Harhay M, Costa CHN & Costa DL (2013) Human immunodeficiency virus/Leishmania infantum in the first foci of urban American visceral leishmaniasis: clinical presentation from 1994 to 2010. Revista da Sociedade Brasileira de Medicina Tropical 46, 156–160. Madalosso G, Fortaleza CM, Ribeiro AF, Cruz LL, Nogueira PA & Lindoso JAL (2012) American visceral leishmaniasis: factors associated with lethality in the state of S~ao Paulo, Brazil. Journal of Tropical Medicine 2012, 1–7. Maia Z, Lırio M, Mistro S, Mendes CMC, Mehta SR & Badaro R (2012) Comparative study of rK39 Leishmania antigen for serodiagnosis of visceral leishmaniasis: systematic review with meta-analysis. PLoS Neglected Tropical Diseases 6, e1484. Marcondes M, Biondo AW, Gomes AA et al. (2011) Validation of a Leishmania infantum ELISA rapid test for serological diagnosis of Leishmania chagasi in dogs. Veterinary Parasitology 175, 15–19. Mazeris A, Soteriadou K, Dedet JP et al. (2010) Leishmaniases and the Cyprus paradox. The American Journal of Tropical Medicine and Hygiene 82, 441–448. Ministerio da Sa ude (MS) (2011a) Leishmaniose visceral: reco~es clınicas para reducß~ao da letalidade. Ministerio da mendacßo Sa ude, Brasılia. http://www.paho.org/bra/index.php? option=com_docman&task=doc_view&gid=1464&Itemid=614 (accessed 10 September 2014). Ministerio da Sa ude (MS) (2011b) Reuni~ao de validacß~ao dos kits de diagn ostico da leishmaniose visceral. Technical report. Ministerio da Sa ude do Brasil, Brasilia, Distrito Federal, Brasil. Ministerio da Sa ude (MS). Estimativas populacionais: 2013. Brasılia. http://tabnet.datasus.gov.br/cgi/deftohtm.exe?ibge (accessed 1 November 2013). Ntais P, Sifaki-Pistola D, Christodoulou V et al. (2013) Leishmaniases in Greece. The American Journal of Tropical Medicine and Hygiene 89, 906–915.
© 2014 John Wiley & Sons Ltd
Tropical Medicine and International Health
volume 00 no 00
H. Maia Peixoto et al. Serological diagnosis of canine visceral leishmaniasis
Oliveira MRF, Gomes AC & Toscano CM (2011) QUADAS and STARD: evaluating the quality of diagnostic accuracy studies. Revista de Sa ude P ublica 45, 416–422. Paranhos-Silva M, Freitas LA, Santos WC, Grimaldi G Jr, Pontes-de-Carvalho LC & Oliveira-dos-Santos AJ (1996) A crosssectional serodiagnostic survey of canine leishmaniasis due to Leishmania chagasi. The American Journal of Tropical Medicine and Hygiene 55, 39–44. Pinheiro PHDC, Pinheiro AN, Ferreira JHL, Costa FAL, Katz S & Barbieri CL (2009) A recombinant cysteine proteinase from Leishmania (Leishmania) chagasi as an antigen for delayedtype hypersensitivity assays and serodiagnosis of canine visceral leishmaniasis. Veterinary Parasitology 162, 32–39. Porrozzi R, Santos da Costa MV, Teva A et al. (2007) Comparative evaluation of enzyme-linked immunosorbent assays based on crude and recombinant leishmanial antigens for serodiagnosis of symptomatic and asymptomatic Leishmania infantum visceral infections in dogs. Clinical and Vaccine Immunology: CVI 14, 544–548. Prado PF, Rocha MF, Sousa JF, Caldeira DI, Paz GF & Dias ES (2011) Epidemiological aspects of human and canine visceral leishmaniasis in Montes Claros, State of Minas Gerais, Brazil, between 2007 and 2009. Revista da Sociedade Brasileira de Medicina Tropical 44, 561–566. Quinnell RJ, Carson C, Reithinger R, Garcez LM & Courtenay O (2013) Evaluation of rK39 rapid diagnostic tests for canine visceral leishmaniasis: longitudinal study and meta-analysis. PLoS Neglected Tropical Diseases 7, e1992. Reichmann MLAB, Pinto HBF & Nunes VFP (1999) Vacinacß~ao contra a raiva de c~aes e gatos. Instituto Pasteur, S~ao Paulo, Brasil. Riley RD, Higgins JPT & Deeks JJ (2011) Interpretation of random effects meta-analyses. BMJ (Clinical Research Ed.) 342, 942–967. Romero AGS & Boelaert M (2010) Control of visceral leishmaniasis in Latin America-a systematic review. PLoS Neglected Tropical Diseases 4, e584. Ros ario EY, Genaro O, Franca-Silva JC et al. (2005) Evaluation of enzyme-linked immunosorbent assay using crude Leishmania and recombinant antigens as a diagnostic marker for canine visceral leishmaniasis. Mem orias do Instituto Oswaldo Cruz 100, 197–203. Salom on OD, Quintana MG, Bruno MR, Quiriconi RV & Cabral V (2009) Visceral leishmaniasis in border areas: clustered distribution of phlebotomine sand flies in Clorinda, Argentina. Mem orias do Instituto Oswaldo Cruz 104, 801–804.
Schubach EYP (2011) Validacß~ao da tecnica de imunocromatografia rapida de duplo percurso para o diagn ostico da leishmaniose visceral canina em amostras de sangue total e soro. Universidade de Brasılia, Brasılia. Silva RM, Laurenti MD, Gomes AC & Nogueira YL (2009) TGROC analysis of immunofluorescence assays in canine visceral leishmaniasis diagnosis. Revista de Sa ude P ublica 43, 1044– 1053. da Silva ES, Van der Meide WF, Schoone GJ, Gontijo CMF, Schallig HDFH & Brazil RP (2006) Diagnosis of canine leishmaniasis in the endemic area of Belo Horizonte, Minas Gerais, Brazil by parasite, antibody and DNA detection assays. Veterinary Research Communications 30, 637–643. Solca MDS, Guedes CES, Nascimento EG et al. (2012) Qualitative and quantitative polymerase chain reaction (PCR) for detection of Leishmania in spleen samples from naturally infected dogs. Veterinary Parasitology 184, 133–140. Sousa MR & Ribeiro ALP (2009) Systematic review and metaanalysis of diagnostic and prognostic studies: a tutorial. Arquivos Brasileiros de Cardiologia 92, 229–238. de Souza RF, dos Santos YL, De Souza VR et al. (2012) Recombinant Leishmania (Leishmania) infantum Ecto-Nucleoside Triphosphate Diphosphohydrolase NTPDase-2 as a new antigen in canine visceral leishmaniasis diagnosis. Acta Tropica 125, 60–66. Teran-Angel G, Schallig H, Zerpa O, Rodrıguez V, Ulrich M & Cabrera M (2007) The direct agglutination test as an alternative method for the diagnosis of canine and human visceral leishmaniasis. Biomedica: Revista Del Instituto Nacional de Salud 27, 447–453. Troncarelli MZ, Camargo JB, Machado JG, Lucheis SB & Langoni H (2009) Leishmania spp. and/or Trypanosoma cruzi diagnosis in dogs from endemic and nonendemic areas for canine visceral leishmaniasis. Veterinary Parasitology 164, 118–123. Whiting PF, Rutjes AWS, Westwood ME et al. (2011) QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Annals of Internal Medicine 155, 529–536. WHO (1992) Guidelines for the Dog Population Management. World Health Organization, Geneva. Zamora J, Abraira V, Muriel A, Khan K & Coomarasamy A (2006) Meta-DiSc: a software for meta-analysis of test accuracy data. BMC Medical Research Methodology 6, 1–12.
Corresponding Author Gustavo Adolfo Sierra Romero, Center for Tropical Medicine, University of Brasılia, Brasılia, Federal District 70904-970, Brazil. Tel.: +556131070085; Fax +556131070081; E-mails:
[email protected];
[email protected]
© 2014 John Wiley & Sons Ltd
19
