• No results found

Evaluation of non-cultural diagnostic methods for detection of Candida albicans–a systematic literature review

N/A
N/A
Protected

Academic year: 2021

Share "Evaluation of non-cultural diagnostic methods for detection of Candida albicans–a systematic literature review"

Copied!
37
0
0

Loading.... (view fulltext now)

Full text

(1)

Örebro University School of Medicine Degree project, 15 ECTS January 2018

Evaluation of non-cultural diagnostic methods for detection of

Candida albicans – a systematic literature review

Version 2

Author: Karolina Baranowska Supervisor: Karolina Prytz, specialist physician, Department of Infectious diseases

(2)

Abstract

Introduction: Candida albicans is a common pathogen and the amount of invasive candida infectionsishigh, especially in immunocompromised patients. Blood culture is the gold standard diagnostic method and have a high specificity but time-consuming and can in some cases lead to unnecessary use of antifungal therapy or even a higher mortality.

There are other non-cultural diagnostic methods that can be used such as BDG, CAGTA, PCR, Mannan antigen and Anti-mannan antibodies.

Method: Fourteen articles evaluating non-cultural diagnostic methods were included. They were found in the database PubMed and later evaluated in full-text reading, both their relevance, study quality and results.

Results: BDG had a high sensitivity and a high NPV which is preferred when excluding invasive candidiasis but lacked the specificity that PCR have which can detect several species. CAGTA had a high specificity, high sensitivity and a high PPV and together with Mannan antigen the specificity became higher, even higher than CAGTA/BDG.

Mannan antigen had a high specificity but low sensitivity and together with Anti-mannan the sensitivity increased but the specificity decreased, this was common in all cases when Anti-mannan was used with another method.

CAGTA/Mannan antigen and CAGTA/BDG can be used as help in deciding terminating of antifungal therapy however BDG/Mannan antigen can be used as an indicator for pre-emptive therapy.

Conclusion: Some methods alone had high sensitivity and/or high specificity and combined with other methods the sensitivity or specificity could increase or decrease.

Keywords: Candida albicans, diagnosis, diagnostic test, PCR, BDG, CAGTA, Mannan, Anti-mannan

(3)

Abbreviations

• ICU – Intensive care unit • BDG – (1,3)-β-d-glucan

• CAGTA – Candida albicans Germ-Tube Antibodies • PCR – Polymerase chain reaction

• PPV – Positive predictive value • NPV – Negative predictive value • MeSH – Medical Subjects Headings • DSC – Deep-seated candidaemia • IC – Invasive candidiasis

• IFD – Invasive fungal disease • IFI – Invasive fungal infection • IA – Invasive Aspergillosis • BSI – Blood stream infection • AFT – Anti-fungal therapy

• BSAT – Broad-spectrum antibiotics therapy • AML – Acute myeloid leukaemia

(4)

Table of contents

1. Introduction ... 1

1.1 About Candida species ... 1

1.2 Candida colonization and invasive candidiasis ... 1

1.3 Incidence of invasive fungal infections ... 1

1.4 Risk factors ... 2 1.5 Cost ... 2 1.6 Diagnostics ... 3 1.6.1 BDG ... 3 1.6.2 PCR ... 3 1.6.3 CAGTA ... 4

1.6.4 Mannan antigen and Anti-mannan antibodies ... 4

2. Material and methods ... 5

2.1 Literature search ... 5

2.1.1 Keywords ... 5

2.1.2 Inclusion criteria ... 5

2.1.3 Exclusion criteria ... 5

2.1.4 The search process ... 5

2.2 Relevance assessment ... 8 2.3 Quality assessment ... 9 2.4 Ethical consideration ... 10 3. Results ... 10 3.1 Study selection ... 10 3.2 Patient characteristics ... 11

3.3 Comparison of diagnostic methods ... 11

3.3.1 BDG ... 11

3.3.2 PCR ... 12

3.3.3 CAGTA ... 13

3.3.4 Mannan antigen and Anti-mannan antibodies ... 13

4. Discussion ... 15

4.1 Result discussion ... 15

4.1.1 Detection of Candida albicans ... 16

4.1.2. Confirmation of colonization and distinguishing it between invasive candidiasis 17 4.1.3 Indication of pre-emptive therapy or termination of antifungal therapy ... 18

4.1.4 Cost ... 18

4.1.5 Further research in the future about diagnostic methods ... 18

4.2 Method discussion ... 19

5. Conclusion ... 20

6. Acknowledgement ... 21

7. References ... 22

8. Attachments ... 25

8.1 Overview of articles included in this systematic review ... 25

8.2 Overview of articles excluded in this systematic review ... 26

8.3 Questions evaluated in the quality assessment ... 27

8.4 Overview of the scoring in the quality assessment ... 27

(5)

1. Introduction

1.1 About Candida species

Candida species are common organisms found in blood of hospitalized patients which can spread widely when entering the blood. Candida infections are associated with health-care and these organisms can infect nearly every organ and almost all types of prosthetic materials. These organisms are yeasts, which are small, thin-walled, ovoid cells that can be propagated by budding. They can grow well in blood culture bottles that are ventilated or on agar plates which do not require special fungal media for cultivation. Candida organisms form smooth, creamy white shining colonies that may look like staphylococcal groups [1].

The four most common Candida species that cause invasive candida infections are Candida albicans, Candida parapsilosis, Candida glabrata and Candida tropicalis [2].

1.2 Candida colonization and invasive candidiasis

Candida colonization and infection are two associated events which lead to the progression to invasive candidiasis. Candida colonization can be classified into two grades: low or high. The grade is based on the number of the areas of the body which are colonized and on the duration and intensity of the colonization. Colonization of at least three body sites at two or more sequential occasions is defined as a high-grade of colonization. Only 5-15% of patients on ICU are colonized by Candida species but as the hospital visit get prolonged, the exposure for risk factors increase and can result in 50-80% colonized patients. Comparing with bacterial infections the time between exposure to the colonization and the risk factors leading to the progress of an invasive candidiasis is for about 7-10 days [3].

Invasive candidiasis includes candidaemia (presence of candida in the blood) and/or DSC which is an infection present underneath mucosal surfaces. DSC can develop by blood

dissemination or when a sterile site becomes exposed directly to candida. These infections can stay localized or can spread to adjacent sites and progress to secondary candidaemia [4].

1.3 Incidence of invasive fungal infections

The incidence has increased the past years and Candida albicans is the most common pathogenic Candida species causing these infections. At the same time it is a part of the human microbiota and is present normally on the skin, mucosal surfacesand in the intestinal [2,5].

(6)

25-50% of all candidaemia cases at the hospital are patients at ICU and this kind of ward is a known risk factor for developing an invasive fungal infection [5].

Inadequate antimicrobial treatment is an independent cause for hospital mortality and fungal blood stream infections are one of the most common infections with wrong initial treatment. But also delaying antifungal therapy has a significant impact on the mortality therefore a rapid diagnosis is necessary [6].

Pre-emptive therapy is another way of preventing invasive candidiasis in the case of patients with a high-graded colonization and it also reduces the amount of antifungal therapy.

1.4 Risk factors

The mortality rate for invasive candidiasis is especially high in the cases of patients that are critically ill [5].

Invasive candidiasis is also a common cause to morbidity in ICU patients and patients with hematological malignancy [7].

There are also other risk factors like prolonged use of broad-spectrum antibiotics, hospital stays longer than two weeks, presence of central venous catheter, gastrointestinal surgical procedures, multiple organ injury, chronic renal failure, invasive urinary tract manipulation, pancytopenia, diabetes mellitus [8], total parenteral nutrition, long use of corticosteroids, age <1 month or above 65 years, chemotherapy and multi-site candida colonization. Patients with highest risk for an infection are those who are exposed for several risk factors at the same time [9].

1.5 Cost

Prolonged hospitalization because of invasive candidiasis leads to increased costs of care [9] therefore rapid diagnosis isimportant. But it is also important for timely administration of antifungal therapy and removal of unnecessary treatment in the case of suspected patients. The costs of treatment can also decrease if the species-specific identification is faster which is crucial for the patients care [2].

(7)

1.6 Diagnostics

There are different diagnostic methods that variate in sensitivity, specificity and duration. Other measurements as NPV and PPV are also of consideration, NPV means the probability that an individual with a “healthy” result really is healthy and PPV means the probability that an individual with an “sick” result really is sick.

Blood culture is the gold standard diagnostic method for invasive candidiasis and has a specificity at 100% [10] even though it has a low sensitivity and detects only 50% of cases with invasive candidiasis and time-consuming. It takes about 48 hours to obtain a positive result [11,12].

The method consists of different phases, first overnight agar medium subculture then identification and antifungal susceptibility testing after a bottle is confirmed as positive for candida [7].

There are other microbiological non-culture based methods like detection of (1,3)-β-D-glucan, PCR for detection of fungal DNA, Candida albicans germ-tube antibodies (CAGTA),

Mannan antigen and Anti-mannan antibodies [5].

1.6.1 BDG

Most pathogenic fungi express (1,3)-β-D-glucan which is a cell-wall component and a fungal marker for detection of invasive infections caused by different fungal pathogens. BDG has a high sensitivity and a significant high NPV, it can be useful for excluding invasive fungal infections [5].

The recommended positive cut-off value is 80 pg/mL [3] but BDG which can be detected in amounts as low as 1 pg ml-1 in serum [5].

However it is known that surgical gauze products can contribute to elevated BDG levels [13,14]. Also some antimicrobials can lead to false-positive results after administration [15] which have to be held in mind when BDG levels are elevated.

1.6.2 PCR

PCR is a molecular method that amplifies gene sequences unique to fungi and is a rapid way of detecting fungal DNA. In a recent study PCR showed a 90.9% sensitivity and 100%

(8)

1.6.3 CAGTA

CAGTA is an immunofluorescence assay that detects antibodies against an antigen that is expressed during the mycelial phase of Candida albicans. This antigen is essential for biofilm development and tissue invasion. But there are other Candida species which are able to express this antigen [16]. The CAGTA test has a sensitivity of 77-89% and specificity of 91-100%. The levels of CAGTA usually decrease as antifungal therapy begins and therefore CAGTA can be used as a marker that monitors the treatment [5].

1.6.4 Mannan antigen and Anti-mannan antibodies

During invasive candidiasis, there is circulating Mannan antigen in the blood which is a polysaccharide component in the candida cell wall [3] and which can be measured in serum. Anti-mannan antibodies can also be measured because Mannan antigen induces a strong antibody response. Combining these two methods result in a higher sensitivity (60-89%) rather than using them alone, where each test has individually a sensitivity of less than 50% [5].

The purpose of this systematic literature review was to evaluate multiple different non-cultural diagnostic methods efficacy for detection of invasive candida infections.

(9)

2. Material and methods

2.1 Literature search 2.1.1 Keywords • Candida albicans • Diagnostic test • Diagnosis • Beta-glucan • PCR • Anti-mannan • Mannan • CAGTA 2.1.2 Inclusion criteria • Humans • English • Last 5 years 2.1.3 Exclusion criteria • Animals

• Other fungal species • Bacteria

• Prophylactic interventions • Resistance development

2.1.4 The search process

Searches were made in PubMED, Cochrane Library and Web of science during the period of 2017-11-06 to 2017-11-22 with different keywords related to the subject.

Different databases were used to achieve a wide search on this subject and to not miss any articles that could be useful. The searches were made in PubMED both as separate keywords but also using Medical Subject Headings (MeSH) to include articles that have been

(10)

Initially the searches consisted of single keywords and the number of articles hits are registered in a search matrix.

The second search was made with the same single keywords with the inclusions criteria to exclude irrelevant articles which is also registered in the search matrix.

To make the searches more specific the advanced search builder was used and more keywords were added. Boolean terms were also used with carefulness. The term “AND” was used to give a more narrow result [17], the term “NOT” was not used in order to not decrease the sensitivity of the search and that result in that relevant articles could be missed [18]. The search was first made without inclusions criteria and the number of article hits are registered in the search matrix. After that the search was made with the inclusions criteria and the result of article hits is registered.

It was only possible in the database PubMED to define the search with all those inclusions criteria which were supposed to be used. In Web of science is was only possible to define the search to “English” language and “last 5 years”. In the Cochrane Library is was only possible to define the search to the “last 5 years”.

In order to obtain less than about 100 articles hits there have been used additional keywords. The title and the abstract were evaluated and relevant articles were included for further reading.

This type of procedure was made with all the different keywords and is shown in Table 1. The same Table 1. shows the final searches of all the different keywords that were made for each diagnostic method.

(11)

Table 1. Matrix of the different searches that were made in the different databases, the

number of article hits and the number of articles that were considered as relevant after reading their title and abstract.

Database Search terms and Boolean operator Number of references

Number of relevant references

PubMed Candida albicans AND diagnostic test 45 8

PubMed (“Candida albicans” [MeSH]) AND diagnostic test

11 1

PubMed (“Beta-glucan” [MeSH]) AND Candida albicans

42 6

PubMed Beta-glucan AND Candida albicans AND diagnosis

31 7

PubMed (“PCR” [MeSH]) AND Candida albicans AND diagnostic test

9 3

PubMed PCR AND Candida albicans and diagnostic test

9 3

PubMed (“Candida albicans” [MeSH]) AND anti-mannan

2 2

PubMed Anti-mannan AND diagnosis 9 6

PubMed (“Candida albicans” [MeSH]) AND mannan 25 2

PubMed Mannan AND diagnosis AND Candida albicans

13 4

PubMed (“Candida albicans” [MeSH]) AND CAGTA 4 3

PubMed CAGTA 8 5

Web of science

Candida albicans AND diagnostic test 106 9

Cochrane library

Candida albicans AND diagnostic test 7 3

(12)

2.2 Relevance assessment

A relevance assessment was made with the purpose to exclude those articles which were irrelevant for the aim of this study. The assessment consisted of two steps. First a screening of all articles were the titles and abstracts were examined. Secondly, all articles which passed the first step were read in full text to evaluate their relevance.

From the 321 articles where titles and abstracts were evaluated, 62 articles were selected and proceeded to step two.

From the 62 articles, there were many duplicates which were excluded and the remaining number of articles were 33.

These 33 articles were read in full text and evaluated for their relevance and fulfilled the inclusions criteria for this study. Fourteen articles were included.

The 19 articles which were excluded are listed in Attachment 8.2 and reasons why are shown in Box 1.

(13)

2.3 Quality assessment

The 14 included studies were reviewed based on a review template from the Swedish agency for health technology assessment and assessment of social services called “Template for quality assessment of diagnostic studies” [19].There eleven questions were answered for each study. Four questions were chosen to be more important and each article got a score

depending on if the answer of the question was yes, no or unclear. If the answer was “yes” the study got one point with a “no” or “unclear” it got zero points.

A total of four points equals a high quality, three points equals a moderate quality, two points equals a low quality and zero-one a very low quality.

The four questions are presented in Attachment 8.3 and the scores for each article are presented in Attachment 8.4.

Depending on which total score they got based on the different questions, they were rated as high-, moderate, low- or very low quality and the results are shown in Table 2.

Table 2. Quality assessment of the studies.

Quality Point range (points) Number of articles

High quality 4 3

Moderate quality 3 9

Low quality 2 2

Very low quality 0-1 0

• Fungal rhinosinusitis • Breakthrough candidaemia • Candida meningitis

• Animals

• Comparing Mannan antigen to candida colonization index • Other fungi

• Other pathogens

• Comparing different types of PCR • Pneumonia

• Evaluating different blood culture methods • Double publications

(14)

2.4 Ethical consideration

This is a systematic literature review and had no intention of intervention. No individual data was presented. Therefore, no ethical consideration had to be taken.

3. Results

3.1 Study selection

The different searches in PubMED, Cochrane Library and Web of Science resulted in 321 articles. Of them 259 did not fulfil the criteria and were excluded and 62 remained.

Of those 62, several were duplicates which resulted in 33. These 33 were read in full text and estimated if they were relevant for this study, of these 19 were excluded.

Fourteen articles fulfilled the inclusion criteria and were included, they are listed in

Attachment 8.1. According to the previously described quality assessment template, of these 14 articles, 3 [10,21,26] had high quality, 9 [11,12,14,16,20,23–25,28] had moderate quality and 2 [22,27] low quality.

These 14 articles were further evaluated in study design, patient characteristics, exposure, what the exposure was compared with, efficacy, result and quality. All the results are listed in Attachment 8.5.

Table 3. Overview of the included articles, references, author, publication year and quality.

Article: reference Author Year Quality

Article 1: [20] Yi Guo 2016 Moderate

Article 2: [10] J. Fortún 2014 High

Article 3: [21] M. Carmen Martinéz-Jimenez 2015 High Article 4: [16] M. Carmen Martinéz-Jimenez 2014 Moderate

Article 5: [22] M. Mikulska 2016 Low

Article 6: [14] E. Martin-Mazuelos 2015 Moderate

Article 7: [23] Y. X. Liew 2015 Moderate

Article 8: [24] C. León 2016 Moderate

Article 9: [25] W. Duettmann 2016 Moderate

Article 10: [12] J. Held 2013 Moderate

Article 11: [26] M. Carmen Martinéz-Jimenez 2015 High

Article 12: [27] H. Xafranski 2013 Low

Article 13: [11] J. Poissy 2014 Moderate

(15)

3.2 Patient characteristics

The main group of patients involved in the different studies were ICU-patients [10,11,14,20– 24,26]. Of these ICU-patients, many had different underlying conditions like hematologic disease, gastrointestinal disease, solid tumor [16,20] and abdominal surgery [23].

One study involved only hematological patients who were admitted to the hematology- or bone marrow transplantation ward [25] and another study involved only one study yeast-positive blood culture bottles [27].

The majority of studies involved adult patients but some had also pediatric patients [10,28].

3.3 Comparison of diagnostic methods

The high quality studies evaluated several diagnostics methods like CAGTA, BDG [10,21,26], PCR [10], Mannan antigen and Anti-mannan antibodies [21].

Some of the moderate quality studies evaluated one method [16,20,23,28] and some evaluated several methods [11,12,14,23–25]. The low quality studies [22,27] evaluated only one

separate method.

The studies were further categorized by which type of method or methods they evaluated.

3.3.1 BDG

BDG had the highest sensitivity compared to blood culture at detecting Candida albicans [22] and the method was the most reliable at detecting candidaemia due to Candida albicans. There were no specific correlation between BDG levels at the time of diagnosis correlated to different patient characteristics, sites of infections or previous antifungal use [28].

Two consecutive BDG levels at cut-off value could differentiate between invasive candidiasis and high-grade colonization [14] even though the specificity was only 57.2% [24].

BDG had a much higher NPV at 87% compared to a PPV at less than 20% and for that it could be useful in excluding invasive candidiasis [23].

Higher BDG levels were more common in patients with a high-grade colonization [11,14,24] rather than those with a low one, or those that were not colonized or infected at all [14]. The levels were also high in patients during the first month after an abdominal surgery [12]. Combining BDG/CAGTA resulted in both high sensitivity and NPV [21,24,26] and a moderate specificity [24]. This combination had a 100% sensitivity at detecting Candida albicans, Candida tropicalis and Candida parapsilosis and only a combination with BDG could detect Candida krusei [21]. There were no significant difference between the BDG

(16)

levels and different Candida species [12]. A positive result of both BDG/CAGTA in a single sample or at least one of them in two serial tests could indicate invasive candidiasis and distinguish it from candida colonization [14,24].

The combination of low BDG/CAGTA levels could be used as a safe way to indicate

terminating of antifungal treatment both in ICU-patients and non-ICU patients [24,26], that is also cost-effective [21].

Combining BDG/Mannan antigen could increase the specificity compared to only BDG [12] but the combination also had a high sensitivity [12,24] and a high NPV [24].

A rise in levels of BDG/Mannan antigen could indicate a relapse of candidaemia. The rise of BDG levels could last up to seven weeks but the natural turnover for Mannan is much faster and the levels don't persist as high for a longer time. This combination is not favourable as a marker to follow up during an antifungal treatment. However an increase in the BDG cut-off value could help to indicate preventive treatment [11].

3.3.2 PCR

PCR had a high specificity [20] and sensitivity compared with blood culture [10,20] and had especially high sensitivity at detecting DSC [10,24].

It was more common with positive PCR tests in the cases of patients with invasive candidiasis and high-grade colonization [24].

There was a high concordance between PCR and blood culture at identifying species [10,27]. It could also detect more than the five most common Candida species [20,27] and infections caused by evolving fungal pathogens. PCR could detect mixed infections while blood culture only was able to detect one species [27].

This method could be used as an accurate and rapid one [10,20]. It could contribute with an accurate result within 24 hours compared to blood culture which took at least 72 hours [27]. PCR was also a cost saving method [10,20,27].

The mentioned method had a significant higher PPV and specificity compared to BDG and CAGTA [10].

PCR/Mannan antigen/Anti-mannan antibodies resulted in the best specificity compared to BDG/CAGTA and BDG/Mannan antigen. The results from PCR/Mannan antigen/Anti-mannan antibodies separately and combined could not distinguish between an invasive candidiasis and a candida colonization [24].

(17)

3.3.3 CAGTA

This method’s sensitivity to detect different species was high, but lower sensitivity for Candida tropicalis [16] and the highest for Candida albicans [21].

CAGTA had both a high PPV and specificity particularly in detecting DSC. One positive CAGTA result could indicate the presence of this condition [16].

Positive CAGTA tests were also more common in the cases of patients with a high-grade colonization [14].

The combination of CAGTA/BDG had a high sensitivity, ahigh NPV [21,24,26] and a moderate-high specificity [21,24]. This combination had the highest sensitivity and NPV in ICU-patients but could be used in both ICU- and non ICU-patients [26] to indicate ending of antifungal therapy in a safe way [21,24,26]. It could also be used to differentiate between colonization and infection or between DSC and catheter-related candidiasis [26]. That was also a relative fast and more cost-effective method [16,21,26].

CAGTA/Mannan antigen had a high sensitivity, high NPV and an even higher specificity than CAGTA/BDG together [21].

3.3.4 Mannan antigen and Anti-mannan antibodies

Mannan antigen had a high specificity [11,12,25] and a low sensitivity [11,12] but when combined together with Anti-mannan antibodies the sensitivity became higher [12] and specificity became also higher comparing to the Anti-mannan antibodies alone [25].

The Anti-mannan antibodies method led to a high amount of false positive results [21,25] and could decrease the specificity of a result when combined with other methods [21].

Levels of Mannan antigen, Anti-mannan antibodies and BDG were higher in patients with candidaemia [12]. Mannan antigen levels could also be higher in patients with

intra-abdominal candidiasis and high-grade colonization [24] however they were lower in patients with candidaemia due do Candida parapsilosis and Candida guillermondii [12].

Mannan antigen/BDG had a high sensitivity and high NPV but a low specificity [24]. High levels could indicate candidaemia relapse and could be helpful when considering pre-emptive therapy [11].

(18)

Combination of Mannan antigen/CAGTA had a high sensitivity, high specificity and a high NPV and could lead to end unnecessary use of empiric treatment in patients with suspected candidaemia [21].

PCR/Mannan antigen/Anti-mannan antibodies had the best specificity compared to

BDG/Mannan antigen. A positive test of PCR combined with high levels of Mannan antigen were more common in patients with invasive candidiasis and high-grade candida colonization [24].

(19)

4. Discussion

4.1 Result discussion

The purpose was to evaluate multiple non-cultural methods efficacy for detection of invasive candida infections.

This review included 14 studies which evaluated at least one of the five non-cultural methods mentioned in the introduction. The results showed that the methods had many common characteristics, both strengths and weaknesses and when combining two methods the characteristics could change. To evaluate and conclude the methods efficacy in detecting invasive candida infections, the different characteristics and the purpose wished to achieve, had to be held in mind.

Different strengths that were underlined in all studies were: • Sensitivity in detecting presence of candida.

• Specificity in being able to exclude candida.

• Accuracy in confirming a high-grade colonization or an invasive candidiasis but also to distinguish them from each other.

• Excluding an invasive candida infection. • Indicating a relapse.

• Reliability in indicating a pre-emptive therapy or terminating an antifungal therapy. • Time of duration for the result and the cost of the method.

These strengths are further discussed both correlating to the method but also their importance of clinical use and in further research.

Both BDG, CAGTA, BDG/CAGTA and CAGTA/Mannan antigen are considered to have a high sensitivity [20–22] however CAGTA, Mannan antigen and Anti-mannan antibodies separately have a high specificity [21].

Nevertheless PCR have both a high sensitivity and specificity [20,27] at detecting more than the five common Candida species, evolving fungal pathogens and mixed infections [27]. Combining BDG with other methods like BDG/Mannan antigen and BDG/CAGTA increase the specificity but CAGTA/Mannan antigen still have a higher specificity [21].

However combining Anti-mannan antibodies with Mannan antigen [12], CAGTA/Mannan antigen and CAGTA/BDG decreased the specificity [21].

(20)

Positive PCR tests and PCR/Mannan antigen, where the Mannan antigen levels are high is common in patients with high-grade colonization and invasive candidiasis [24].

Other methods that also could indicate a high-grade colonization was BDG, CAGTA and Mannan antigen [14,23,24]. PCR, CAGTA and Mannan antigen were also especially good at indicating the presence of candida in DSC [10,16,24].

BDG, BDG/CAGTA can be used to differentiate between high-grade candida colonization and infection [24,26] but also between DSC and catheter-related candidiasis [26]. However, even if PCR/Mannan antigen/Anti-mannan antibodies had a high specificity they cannot distinguish between a high-grade candida colonization and invasive candidiasis [24]. BDG can also be used to exclude invasive candidiasis because of the high NPV [23] while BDG/Mannan antigen can be used to indicate a candidaemia relapse where the BDG levels can stay high for up to seven weeks. [11].

BDG/CAGTA and CAGTA/Mannan antigen can be used as biomarkers to terminate antifungal therapy in a safe way [21,24,26] while BDG/ Mannan antigen can be used as indicators for pre-emptive therapy [11].

BDG/CAGTA, CAGTA, PCR are considered ascost-effective methods [10,16,20,21,26] and PCR and CAGTA as faster methods [10,16,20,21,26].

4.1.1 Detection of Candida albicans

Candida albicans is both an organism in the human microbiota and a common pathogen in the hospital environment [1] therefore it is of huge importance that the methods have both high sensitivity and specificity at detecting this organism.

The sensitivity is a measure of the methods efficacy in detecting the presence of fungi, in these studies focus was on the presence of candida. All methods could detect Candida albicans but had different sensitivity in doing that and for some methods the sensitivity was significant low for one type of species. CAGTA had a low sensitivity at detecting Candida tropicalis while Mannan-antigen levels were low in candidaemia due to Candida parapsilosis and Candida guillermondii. This can be a weakness because sometimes the infections are mixed.

The only method, of the different evaluated methods, that can detect multiple species is PCR. On the other hand, the article which confirms that, is rated as a low-quality study because it was not based on a group of patients but on yeast-positive bottles [27].

(21)

The advantages in having a high sensitivity method are many. It is helpful in choosing the treatment and avoiding development of resistance which is important because invasive fungal infections are mostly common in immunocompromised patients.

Patients included in these 14 articles were mostly immunocompromised patients at ICU-wards or other ICU-wards. Being at the ICU-ICU-wards means exposure to many risk factors and being at other wards as an immunocompromised patient mean decreased protection against risk factors.

Comorbidity is common in these patients and therefore they can be on multiple drugs that have to be given in other ways than orally. That can result in that prosthetic materials are colonized and yeast growth begins [1].

Being able to detect exactly which kind of organism it is and even which kind of species is of importance. It can have a strong impact on the treatment and if possible avoidance of

compromising the patients defence for future infections. To have the opportunity to give a narrow-spectrum antibiotic is the goal because prolonged use of broad-spectrum antibiotics is a risk factor [8] and to have this kind of treatment multiple times have a negative impact on resistance development.

Another important aspect is if the patient already is resistant against a species. Then it is valuable to know which species the patient is colonized with and which other choice of treatment is available for the patient. From this aspect, the specificity is also significant for the diagnostics because it can establish that the patient is not colonized with some kind of species.

The specificity varied between the methods. Combining a method with lower specificity with a method with higher specificity could increase the combined specificity however it could decrease when combining with a method with an even lower specificity.

4.1.2. Confirmation of colonization and distinguishing it between invasive candidiasis To be able to confirm if a patient has a high-grade candida colonization is important because it is a risk factor for invasive candidiasis and it can be prevented with adequate treatment. Even though high-grade candida colonization and invasive candidiasis in the end have the same outcomes and can lead to sepsis and later death, it can be valuable to know in which stage they are in the disease regarding treatment decisions.

It is medically a strength for a method to be able to confirm that a patient does not have an invasive candidiasis, because if the patient is treated with antifungal therapy the treatment can

(22)

Several methods were able to confirm a high-grade colonization or an invasive candidiasis but only BDG and BDG/CAGTA could differentiate between those two conditions.

4.1.3 Indication of pre-emptive therapy or termination of antifungal therapy It is clear that a rapid diagnosis of invasive candida is preferred in the use of antifungal therapy because delayed therapy is associated with a higher mortality [6].

A method that can indicate ending of an antifungal therapy has to have a high NPV, which means it can be able to indicate that a healthy person according to a test is really healthy, or in this case of not having any signs of an invasive candidiasis. Methods that can help in deciding in terminating antifungal therapy are BDG/CAGTA and CAGTA/Mannan antigen.

It is also important to treat a high-graded colonization that can progress to an invasive infection which in the worst scenario can progress to a sepsis and further complications can arise and the length of the hospital stay and care can be prolonged which results in further costs. Mannan antigen/BDG can indicate pre-emptive therapy which can reduce the risk for an invasive candidiasis but also the risk for resistance development. This combination is also a useful biomarker to prevent overuse of drugs by being able to indicate an eventual relapse.

4.1.4 Cost

Lower cost of a method is favourable for hospitals but it is equally important that the method is reliable and helpful in diagnostics which can lead to less use of antifungal therapy and less unnecessary treatment which indirectly leads to lower costs.

It is also useful if a method has a faster duration for diagnosis because the gold standard method which is blood culture, has a long duration for a result. PCR is a method that it considered to be really fast but at the same time it cannot be used as an indicator for ending a treatment but can confirm that there is a presence of Candida species.

4.1.5 Further research in the future about diagnostic methods

It would be useful to be sure which method/ methods should be used and considered as the best for a special group of patients, or in a specific condition or as an intervention.

It would make it easier and faster to diagnose, treat the patient and reduce the risk for further complications. That can also reduce the resistance prevalence which is important for the future. If new species develop there have to be methods than can detect them so that the treatment will be as specific as it can be.

(23)

4.2 Method discussion

All 33 articles where relevance estimated by only one person which causes subjectivity and can be reason that relevant articles were not included.

It resulted in 14 articles that were further evaluated and quality estimated also by one person. Three different databases were used for the different searches, but the 14 articles were all chosen from PubMed mostly because of duplicates and because of the inclusions criteria in the other databases were very restricted. This could have led to that articles of relevance were not found.

(24)

5. Conclusion

All the different methods that was evaluated have different strengths and weaknesses. Depending on the approach wanted to be achieved different methods can be recommended. BDG has a high sensitivity and a high NPV and is the only method that could exclude invasive candidiasis but at the same time it lacks the high specificity that PCR and blood culture has. PCR itself has a better specificity than BDG and CAGTA.

PCR is good at detecting different pathogens even those which are more uncommon. It is important to know which specific species the patient is infected with from the treatment aspect. However, PCR cannot distinguish between high-graded colonization and invasive candidiasis. Using BDG and PCR in combination or at least after each other, can lead to a specific result with higher specificity and higher sensitivity.

Another way to increase the BDG specificity is to combine it with Mannan antigen which has a high specificity. Still this combination has a lower specificity than using CAGTA/Mannan antigen. Nevertheless, BDG/Mannan antigen can be used to indicate pre-emptive therapy. Mannan antigen has high specificity but a low sensitivity, together with Anti-mannan antibodies the specificity becomes lower but and the sensitivity higher. Anti-mannan antibodies often result in false-positive results.

CAGTA has a high PPV and specificity however combining it with Mannan antigen results in an even higher specificity than combining it with BDG. The combination has also a high sensitivity and a high NPV.

Combining CAGTA/Mannan antigen and BDG/CAGTA can help in deciding the terminating of antifungal therapy.

(25)

6. Acknowledgement

I want to dedicate a special thanks to my supervisor Karolina Prytz for all the help and guidance in this project!

(26)

7. References

1. Bennett JE, Dolin R, Blaser MJ. Mandell, Douglas, and Bennett’s principles and practice of infectious diseases. Volume 2. Philadelphia: Saunders; 2015. p. 2879.

2. Ahmad S, Khan Z. Invasive candidiasis: a review of nonculture-based laboratory diagnostic methods. Indian J Med Microbiol. 2012 Sep;30(3):264–9.

3. León C, Ostrosky-Zeichner L, Schuster M. What’s new in the clinical and diagnostic management of invasive candidiasis in critically ill patients. Intensive Care Med. 2014 Jun;40(6):808–19.

4. Clancy CJ, Nguyen MH. Finding the ‘missing 50%’ of invasive candidiasis: how

nonculture diagnostics will improve understanding of disease spectrum and transform patient care. Clin Infect Dis Off Publ Infect Dis Soc Am. 2013 May;56(9):1284–92.

5. Pemán J, Zaragoza R. Current diagnostic approaches to invasive candidiasis in critical care settings. Mycoses. 2010 Sep;53(5):424–33.

6. Garey KW, Rege M, Pai MP, Mingo DE, Suda KJ, Turpin RS, et al. Time to initiation of fluconazole therapy impacts mortality in patients with candidemia: a multi-institutional study. Clin Infect Dis Off Publ Infect Dis Soc Am. 2006 Jul 1;43(1):25–31.

7. Oz Y, Gokbolat E. Evaluation of direct antifungal susceptibility testing methods of Candida spp. from positive blood culture bottles. J Clin Lab Anal. 2017 Jul 13;

8. Alam FF, Mustafa AS, Khan ZU. Comparative evaluation of (1, 3)-beta-D-glucan, mannan and anti-mannan antibodies, and Candida species-specific snPCR in patients with candidemia. BMC Infect Dis. 2007 Sep 4;7:103.

9. Kratzer C, Graninger W, Lassnigg A, Presterl E. Design and use of Candida scores at the intensive care unit. Mycoses. 2011 Nov;54(6):467–74.

10. Fortún J, Meije Y, Buitrago MJ, Gago S, Bernal-Martinez L, Pemán J, et al. Clinical validation of a multiplex real-time PCR assay for detection of invasive candidiasis in intensive care unit patients. J Antimicrob Chemother. 2014 Nov;69(11):3134–41. 11. Poissy J, Sendid B, Damiens S, Ishibashi KI, Francois N, Kauv M, et al. Presence of Candida cell wall derived polysaccharides in the sera of intensive care unit patients: relation with candidaemia and Candida colonisation. Crit Care. 2014;18(3):R135.

12. Held J, Kohlberger I, Rappold E, Busse Grawitz A, Häcker G. Comparison of (1->3)- β-D-glucan, mannan/anti-mannan antibodies, and Cand-Tec Candida antigen as serum biomarkers for candidemia. J Clin Microbiol. 2013 Apr;51(4):1158–64.

13. Kanamori H, Kanemitsu K, Miyasaka T, Ameku K, Endo S, Aoyagi T, et al.

Measurement of (1-3)-beta-D-glucan derived from different gauze types. Tohoku J Exp Med. 2009 Feb;217(2):117–21.

(27)

14. Martín-Mazuelos E, Loza A, Castro C, Macías D, Zakariya I, Saavedra P, et al. β-D-Glucan and Candida albicans germ tube antibody in ICU patients with invasive candidiasis. Intensive Care Med. 2015 Aug;41(8):1424–32.

15. Marty FM, Lowry CM, Lempitski SJ, Kubiak DW, Finkelman MA, Baden LR. Reactivity of (1→3)-β-d-Glucan Assay with Commonly Used Intravenous Antimicrobials. Antimicrob Agents Chemother. 2006 Oct;50(10):3450–3.

16. Martínez-Jiménez MC, Muñoz P, Guinea J, Valerio M, Alonso R, Escribano P, et al. Potential role of Candida albicans germ tube antibody in the diagnosis of deep-seated candidemia. Med Mycol. 2014 Apr;52(3):270–5.

17. Forsberg C, Wengström Y. Att göra systematiska litteraturstudier : värdering, analys och presentation av omvårdnadsforskning. Stockholm: Natur & kultur; 2016.

18. Systematiska översikter | Karolinska Institutet Universitetsbiblioteket [Internet]. [cited 2017 Nov 24]. Available from: https://kib.ki.se/soka-vardera/systematiska-oversikter 19. Swedish agency for health technology assessment and assessment of social services. Quality assessment of diagnostic studies; Stockholm; 2017-10-26. [Internet]. [cited 2017 Dec 5]. Available from:

http://www.sbu.se/globalassets/ebm/metodbok/mall_diagnostiska_studier.pdf

20. Guo Y, Yang J-X, Liang G-W. A Real-Time PCR Assay Based on 5.8S rRNA Gene (5.8S rDNA) for Rapid Detection of Candida from Whole Blood Samples. Mycopathologia. 2016 Jun;181(5–6):405–13.

21. Martínez-Jiménez MC, Muñoz P, Valerio M, Alonso R, Martos C, Guinea J, et al.

Candida biomarkers in patients with candidaemia and bacteraemia. J Antimicrob Chemother. 2015 Aug;70(8):2354–61.

22. Mikulska M, Giacobbe DR, Furfaro E, Mesini A, Marchese A, Del Bono V, et al. Lower sensitivity of serum (1,3)-β-d-glucan for the diagnosis of candidaemia due to Candida parapsilosis. Clin Microbiol Infect Off Publ Eur Soc Clin Microbiol Infect Dis. 2016 Jul;22(7):646.e5-8.

23. Liew YX, Teo J, Too IA-L, Ngan CC-L, Tan AL, Chlebicki MP, et al. Candida

Surveillance in Surgical Intensive Care Unit (SICU) in a Tertiary Institution. BMC Infect Dis. 2015 Jul 3;15:256.

24. León C, Ruiz-Santana S, Saavedra P, Castro C, Loza A, Zakariya I, et al. Contribution of Candida biomarkers and DNA detection for the diagnosis of invasive candidiasis in ICU patients with severe abdominal conditions. Crit Care Lond Engl. 2016 May 16;20(1):149.

25. Duettmann W, Koidl C, Krause R, Lackner G, Woelfler A, Hoenigl M. Specificity of mannan antigen and anti-mannan antibody screening in patients with haematological malignancies at risk for fungal infection. Mycoses. 2016 Jun;59(6):374–8.

(28)

26. Martínez-Jiménez MC, Muñoz P, Valerio M, Vena A, Guinea J, Bouza E. Combination of Candida biomarkers in patients receiving empirical antifungal therapy in a Spanish tertiary hospital: a potential role in reducing the duration of treatment. J Antimicrob Chemother. 2015 Nov;70(11):3107–15.

27. Xafranski H, Melo ASA, Machado AM, Briones MRS, Colombo AL. A quick and low-cost PCR-based assay for Candida spp. identification in positive blood culture bottles. Bmc Infect Dis. 2013 Oct 7;13:467.

28. Angebault C, Lanternier F, Dalle F, Schrimpf C, Roupie A-L, Dupuis A, et al. Prospective Evaluation of Serum beta-Glucan Testing in Patients With Probable or Proven Fungal

(29)

8. Attachments

8.1 Overview of articles included in this systematic review

Study Title Author Year

1. A real-time PCR assay based on 5.8S rRNA gene (5.8S Rdna) for rapid detection of candida from whole blood samples

Yi Guo, Jing-xian Yang, Guo-wei Liang 2016

2. Clinical validation of a multiplex real-time PCR assay for detection of invasive candidiasis in intensive care unit patients

J. Fortún, Y. Meije, M.J Buitrago, S. Gago, L. Bernal-Martinez, J. Peman, M. Perez, E. Gomez-G Pedrosa, N. Madrid, V. Pintado, P. Martin-Davila, J. Cobo, G. Fresco, S. Moreno, M. Cuenca-Estrella

2014

3. Candida biomarkers in patients with candidaemia and bacteraemia

M. Carmen Martinez-Jimenez, Patricia Munoz, Maricela Valerio

2015 4. Potential role of Candida albicans germ tube antibody in

the diagnosis of deep-seated candidemia

M. Carmen Martinez-Jimenez, Patricia Munoz, Jesus Guinea, Maricela Valerio, Roberto Alonso, Pilar Escribano, Emilio Bouza

2014

5. Lower sensitivity of serum (1,3)-Beta-d-glucan for the diagnosis of candidaemia due to Candida parapsilosis

M. Mikulska, D.R Giacobbe, E. Furfaro, A. Mesini, A. Marchese, V. Del Bono, C. Viscoli

2016 6. Beta-d-glucan and Candida albicans germ tube antibody

in ICU patients with invasive candidiasis

Estrella Martin-Mazuelos, Ana Loza, Carmen Castro, Desirée Macias, Ismail Zakariya, Pedro Saavedra, Sergio Ruiz-Santana, Elena Marin, Cristobal Leon

2015

7. Candida surveillance in surgical intensive care unit (SICU) in a tertiary institution

Yi Xin Liew, Jocelyn Teo, Irene Ai-Ling Too, Cecilia Cheng-Lai Ngan, Ai Ling Tan, Maciej Piotr Chlebicki, Andrea Lay-Hoon Kwa, Winnie Lee

2015

8. Contribution of Candida biomarkers and DNA detection for the diagnosis of invasive candidiasis in ICU patients with severe abdominal conditions

Cristobal Leon, Sergio Ruiz-Santana, Pedro Saavedra, Carmen Castro, Ana Loza, Ismail Zakariya, Alejandro Ubeda, Manuel Parra, Desiree Macias, Jose Ignacio Tomas, Antonio Rezusta, Alejandro Rodriguez, Frederic Gomez, Estrella Martin-Mazuelos and the Cava Trem Study Group

2016

9. Specificity of mannan antigen and anti-mannan antibody screening in patients with haematological malignancies at risk for fungal infection

Wiebke Duettmann, Christoph Koidl, Robert Krause, Gertrude Lackner, Albert Woelfler, Martin Hoenigl

2016

10. Comparison of (1=>3)-Beta-d-glucan, Mannan/ Anti-mannan antibodies and Cand-Tec Candida antigen as serum biomarkers for Candidemia

Jürgen Held, Isabelle Kohlberger, Elfriede Rappoid, Andrea Busse Grawitz, Georg Häcker

2013

11. Combination of Candida biomarkers in patients receiving empirical antifungal therapy in a Spanish tertiary hospital: a potential role in reducing the duration of treatment

M. Carmen Martinez-Jimenez, Patricia Munoz, Maricela Valerio, Antonio Vena, Jesus Guinea, Emilio Bouza

2015

12. A quick and low-cost PCR-based assay for Candida spp. Identification in positive culture bottles

Hemilio Xafranski, Analy SA Melo, Antonia M Machado, Marcelo RS Briones, Arnaldo L Colombo

2013 13. Presence of Candida cell wall derived polysaccharides in

the sera of intensive care units’ patients: relation with candidaemia and Candida colonisation

Julien Poissy, Boualem Sendid, Sebastien Damiens, Ken Ichi Ischibashi, Nadine Francois, Marie Kauv, Raphael Favory, Daniel Mathieu, Daniel Poulain

2014

14. Prospective evaluation of serum Beta-glucan testing in patients with probable or proven fungal diseases

Cecile Angebault, Fanny Lanternier, Frederic Dalle, Cecile Schrimph, Anne-Laure Roupie, Aurelie Dupuis, Aurelie Agathine, Anne Scemia, Etienne Paubelle, Denis Caillot, Benedicte Neven, Pierre Frange, Felipe Suarez, Christophe d’Enfert, Olivier Lortholary, Marie-Elisabeth Bougnoux

(30)

8.2 Overview of articles excluded in this systematic review

Title Author Years

SeptiFast for diagnosis of sepsis in severely ill patients from a Brazilian hospital

Roberta Sitnik, Alexandre Rodrigues Marra, Roberta Cardoso Petroni, Ozires Pereira Santos Ramos, Marines Dalla Valle Martino, Jacyr Pasternak, Oscar Fernando Pavao dos Santos, Cristovao Luis Pitangueira Mangueira, Joao Renato Rebello Pinho

2014

Comparing immunological and molecular tests with conventional methods in diagnosis of acute invasive fungal rhinosinusitis

Parisa Badiee, Mohsen Moghadami, Hossain Rozbehani 2016

Improved detection of deeply invasive candidiasis with DNA aptamers specific binding to (1 =>3)-Beta-d-glucans from Candida albicans

X.-L. Tang, Y. Hua, Q. Guan, C.-H Yuan 2016

Serum (1,3)-Beta-d-glucan is an inefficient marker of breakthrough candidemia

Masahiro Abe, Muneyoshi Kimura, Hideki Araoka, Shuichi Taniguchi, Akiko Yoneyama

2014 Candida meningitis in an immunocompetent patient

detected through (1=>3)-Beta-d-glucan

Mark K. Farrugia, Evan P. Fogha, Abdul R. Miah, Joel Yednock, H. Carl Palmer, John Guilfoose

2016 Mannose-binding lectin levels and variation during

invasive candidiasis

Sebastien Damiens, Julien Poissy, Nadine Francois, Julia Salleron, Samir Jawhara, Thierry Jouault, Daniel Poulain, Boualem Sendid

2012 What’s new in the clinical and diagnostic management of

invasive candidiasis in critically ill patients

Cristobal Leon, Luis Ostrosky-Zeichner, Mindy Schuster 2014 Finding the “missing 50%” of invasive candidiasis: how

nonculture diagnostics will improve understanding of disease spectrum and transform patient care

Cornelius J. Clancy, M. Hong Nguyen 2013

Mannosylation in Candida albicans: role in cell wall function and immune recognition

Rebecca A. Hall, Neil A. R. Gow 2013

Characteristic and clinical relevance of Candida mannan test in the diagnosis of probable invasive candidiasis

Bernabe F. F. Chumpitazi, Bernadette Lebeau, Odile Faure-Cognet, Rebecca Hamidfar-Roy, Jean-Francois Timsit, Patricia Pavese, Anne Thiebaut-Bertrand, Jean-Louis Quesada, Herve Pelloux, Claudine Pinel

2014

Evaluation of a digital microfluidic real-time PCR platform to detect DNA of Candida albicans in blood

W. A. Schell, J. L. Benton, P. B. Smith, M. Poore, J. L. Rouse, D. J. Boles, M. D. Johnson, B. D. Alexander, V. K. Pamula, A. E. Eckhardt, M. G. Pollack, D. K. Benjamin jr, J. R. Perfect, T. G. Mitchell

2012

Laboratory diagnostics of invasive fungal infections: an overview with emphasis on molecular approach

Lenka Baskova, Vladimir Buchta 2012

Polymerase Chain Reaction-Based assays for the diagnosis of invasive fungal infections

Themistoklis K. Kourkoumpetis, Beth Burgwyn Fuchs, Jeffrey J. Coleman, Athanasios Desalermos, Eleftherios Mylonakis

2012 Rapid diagnosis of infections in the critically ill, a

multicenter study of molecular detection in bloodstream infections, pneumonia and sterile site infections

Jean-Louis Vincent, David Brealey, Nicolas Libert, Nour Elhouda Abidi, Michael O’Dwyer, Kai Zacharowski, Malgorzata

Mikaszewska-Sokolewicz, Jacques Schrenel, Francois Simon, Mark Wilks, Marcus Picard-Maureau, Donald B. Chalfin, David J. Ecker, Rangarajan Sampath, Mervyn Singer, the rapid diagnosis of infections in the critically ill team

2015

Is it feasible to diagnose catheter-related candidemia without catheter withdrawal?

Ana Fernandez-Cruz, Pablo Martin-Rabadan, Marisol Suarez-Salas, Loreto Rojas-Wettig, Maria Jesus Perez, Jesus Guinea, Maria Guembe, Teresa Pelaez, Carlos Sanchez-Carrillo, Emilio Bouza, on the behalf of the COMIC (Colaboracion en Micologia) study group

2014

DNAemia detection by multiplex PCR and biomarkers for infections in systemic inflammatory response syndrome patients

Catherine Fitting, Marianna Parlato, Minou Adib-Conquy, Nathalie Memain, Francois Phillipart, Benoit Misset, Mehran Monchi, Jean-Marc Cavaillon, Christophe Adrie

2012

Persistent spontaneous fungal peritonitis secondary to Candida albicans in a patient with alcoholic cirrhosis a review of the literature

Aneela Majeed, Waqas Ullah, Umar Zahid, Mayar Al Mohajer 2016

Evaluation of three different bottles in BACTEC 9240 automated blood culture system and direct identification of Candida species to shorten the turnaround time of blood culture

Egemen Gokbolat, Yasemin Oz, Selma Metintas 2017

Non-cultural methods for the diagnosis of invasive fungal disease

(31)

8.3 Questions evaluated in the quality assessment

The fours questions chosen from the quality assessment template for diagnostic studies by the Swedish agency for health technology assessment and assessment of social services:

About avoidance of selection bias:

1. Was the composition of the patient group (spectrum) representative of the patients who will get the test in practice?

About avoidance of partial verification bias:

2. Did all patients or a random sample of patient undergo the intended reference test? About avoidance of information bias:

3. Was the results from the index test interpreted without knowing the results of the reference test?

4. Was the loss of patients from the study explained?

8.4 Overview of the scoring in the quality assessment

Study: Question 1: Question 2: Question 3: Question 4: Total: Quality:

1. Yi Guo 2016 1 1 0 1 3 Moderate

2. J. Fortun 2014 1 1 1 1 4 High

3. M. Carmen Martinez-Jimenez 2015 1 1 1 1 4 High

4. M. Carmen Martinez-Jimenez 2014 1 1 0 1 3 Moderate

5. M. Mikulska 2016 1 1 0 0 2 Low 6. E. Martin-Mazuelos 2015 1 1 0 1 3 Moderate 7. Yi Xin 2015 1 1 0 1 3 Moderate 8. C. Leon 2016 1 1 0 1 3 Moderate 9. W. Duettmann 2016 1 1 0 1 3 Moderate 10. J. Held 2013 1 1 0 1 3 Moderate

11. M. Carmen Martinez-Jimenez 2015 1 1 1 1 4 High

12. H. Xafranski 2013 0 1 0 1 2 Low

13. J. Poissy 2014 1 1 0 1 3 Moderate

(32)

8.5 Table of results

Publication Study design Patient characteristics Exposure Compared with Efficacy Results Quality

1. Yi Guo 2016

Prospective study

82 adult patients with suspected BSI from the ICU of Respiratory and Geriatric department in Aerospace Center Hospital in Beijing, China. From these patients 328 blood samples were tested for PCR and compared with results from blood culture.

Average length of hospital stay for all patients were 21 days.

31 patients were under ICU admission, 11 post-operative and 15 patients had chemotherapy.

59 patients had BSAT and 12 had AFT.

Polymerase chain reaction (PCR)

Blood culture Sensitivity and specificity.

PCR had a 100% sensitivity using as gold standard blood culture and the specificity was 98,4%.

7 samples from 1 patient were culture positive while Candida DNA was detected in 12 samples, these were from 5 different patients including those 7 samples.

The other 316 samples were culture negative with both methods.

Conclusion that PCR can be used as an accurate and rapid screening test for the Candida from whole blood.

Moderate

2. J. Fortún 2014

Prospective study

63 ICU patients with suspected IC and no prior antifungal treatment and 40 healthy controls. The case group were later divided into having IC and non-IC.

The age range for the IC-group was 15-84 years (median 66) and for the non-IC group 1-86 years (median 60).

30% respectively 32% of the patients in the different groups was at ICU for a medical reason, 63% respectively 55% for a surgical reason and 6% respectively 12% for a pediatric reason. 93% in the IC group had antibiotics for the last 7 days and 89% in the non-IC group.

Multiplex quantitative real time PCR (MRT-PCR), BDG, CAGTA

Blood culture Sensitivity, specificity, PPV, NPV

27 patients were confirmed with IC from the study group. MRT-PCR had a high sensitivity (96,3%), specificity (97,3%), PPV (92,8%) and NPV (98,7%) at detecting IC. MRT-PCR had a significant higher specificity and PPV than BDG and CAGTA.

Comparing to blood culture the sensitivity for MRT-PCR to detect DSC was much higher. A sensitivity at 45.5% for blood culture comparing it to a sensitivity at 90.9% for MRT-PCR. High 3. M. Carmen Martinez-Jimenez 2015 Prospective observational study

31 adult patients with candidaemia and 50 adult patients with bacteraemia as a control group at Gregorio Maranon Hospital, Madrid, Spain.

CAGTA, Mannan antigen, Anti-mannan antibodies, BDG

Blood culture Sensitivity, specificity, NPV

Low sensitivity for all biomarkers when used alone (58%-84%) but a high specificity (65.8%-92%).

(33)

From the 31 candidaemia patients 15 patients were on surgical ward of admission, 10 on medical, 4 on onco-hemotological and 2 under intensive care. In the bacteraemia group 10 patients were on a surgical ward, 34 on a medical ward, 2 on a onco-hematological ward and 4 under intensive care. 18 patients in the candidaemia group had

gastrointestinal disease as an underlying condition, 16 solid tumors and 2 had hematological disease. All 31 patients with candidaemia was under BSAT which is a risk factor for IC and all were

colonized/ infected with Candida species.

Best combination was CAGTA/BDG with a sensitivity of 96.8% and specificity 84% and CAGTA/Mannan with a sensitivity of 93.5% and specificity of 86%.

The sensitivity of both combinations was 100% for

Candida albicans, Candida. tropicalis and Candida. parapsilosis but only combinations including BDG

detected Candida krusei.

The NPV of both combinations were respectively 97.7% and 95.6%.

Anti-mannan antibodies added to CAGTA/Mannan antigen or CAGTA/BDG decreases the specificity but the

sensitivity remains the same. 4. M. Carmen

Martinez-Jimenez 2014

Retrospective analysis

50 adult patients with candidaemia (29 with DSC and 21 with non-DSC) and a control group of 50 healthy individuals.

24 patients were at a surgical ward when admitted, 15 a medical, 9 a onco-hematological and 2 at intensive care.

The most common underlying disease was gastrointestinal disease in 26 patients, solid tumor in 20 and 7 had a hematologic disease.

48 patients had an intravenous catheter which is a risk factor for IC.

CAGTA Blood culture Specificity,

PPV

CAGTA tests were positive in 1/21 non-DSC patients (4.76%) and 20/20 DSC patients (68.96%).

CAGTA had a high specificity and PPV for detecting DSC. Lower sensitivity for detecting Candida tropicalis but still a high specificity.

Conclusions that the presence of a positive CAGTA test in a sample from a patient with candidaemia suggests DSC.

Moderate

5. M.

Mikulska 2016

Retrospective study

107 adult patients with candidaemia. 46 patients had infections caused by Candida albicans, 37 by

Candida parapsilosis and the rest by other species.

Sampling for BDG testing was performed within (before or after) 48 h from onset of candidaemia. Ward of admission at candidaemia was ICU/ surgery for 35 patients and medical/ hematology for 33 patients.

BDG Blood culture Sensitivity and

levels of BDG due to different Candida species

BDG sensitivity and levels were the highest for Candida

albicans candidaemia, sensitivity was 72% and lowest for Candida parapsilosis with a sensitivity at 41%.

(34)

37 patient had recent major surgery, 7 had a hematopoietic stem cell transplantation, 18 hematological malignancies, 18 solid tumors. 90 patients had presence of central venous catheter at the time of BDG test.

37 patients had Candida parapsilosis and 35 patients had other species.

6. Estrella Martin- Mazuelos 2015 Prospective cohort and observational study

107 adult unselected non-neutropenic critically ill patients admitted for at least 7 days to medical-surgical ICU in Sevilla, Spain.

Patients were classified as neither colonized or infected (n=29), Candida spp. Colonization (n=63) “low grade n=32 and high-grade n=31” and invasive candidiasis (n=15).

BDG and CAGTA

Blood culture Sensitivity, specificity

BDG were higher in the patients with IC and high-grade colonization than in the remaining groups, and two consecutive measurements >80 pg/ml discriminated IC from the remaining groups (sensitivity 80% and specificity 75.7%).

A positive CAGTA test was significant higher in patients with a high-grade colonization than in the remaining groups. Moderate 7. Yi Xin Liew 2015 Prospective cohort and observational study

28 adult patients who were admitted to Singapore general hospital (SGH) surgical intensive care unit (SICU) for at least 48 hours.

They got classified into three groups; 1) Neither colonized nor infected (n=3) 2) Candida spp. colonization without IC (n=23) 3) Candida spp. colonization with proven IC (n=4) 19 patients underwent emergency operations, 3 underwent elective operations while 6 patients did not have any operation.

14 patients had abdominal surgery while the remaining 8 had non-abdominal related surgeries.

BDG, Candida score and Colonization index

Blood culture Sensitivity, specificity, PPV, NPV

All test had a good sensitivity (75-100%) but a low specificity (15.38-38.46%), BDG had the best of them three with a specificity at 38.46% and Candida score had the lowest specificity at 15.38%.

All test had a high NPV (>87%) but a PPV at <20%. Conclusion that they are helpful in excluding IC based on their high NPV. Moderate 8. Cristobal Leon 2016 Prospective, cohort, observational and multicenter study

233 adult non-neutropenic patients with SAC (Severe abdominal conditions) on ICU admission with an expected stay of >7 days.

Patients were classified as neither colonized or infected (n=48), Candida spp. colonization (n=154) “low grade n=130, high grade n=24” and

BDG, CAGTA, Mannan-Ag, Anti-mannan antibodies and C-PCR.

Blood culture Sensitivity, specificity, NPV

CAGTA/BDG positivity in a single sample or at least one of them positive in two following samples showed 90.3% sensitivity, 42.1% specificity and 96.6% NPV.

BDG positivity in two following samples had 76.7% sensitivity and 57.2% specificity.

(35)

IC (n=31) “intra-abdominal candidiasis n=20 and candidaemia n=11).

The group of Intra-abdominal candidiasis and high-grade colonization showed significant differences comparing to the remaining groups in Candida score, length of ICU stay, number of surgical procedures and more frequent antifungal treatment.

In the colonization and infection group (n=185) risk factors were very common, 184 patients had Broad spectrum antibiotics, 183 had a urinary catheter, 183 had a central venous catheter, 165 an arterial catheter, 156 mechanical ventilation and 155 parenteral nutrition.

Mannan-Ag, Anti-mannan antibodies and Candida DNA individually or combined presented a low discriminating ability.

Conclusion that CAGTA/BDG positivity in a single sample or BDG positivity in two following samples can

discriminate between IC and Candida species colonization in critically ill SAC patients.

9. Wiebke Duettmann 2016 Clinical prospective study

67 adult patients admitted to the hematology or bone marrow transplantation ward with a risk for invasive fungal infection at the Department of Hematology, Medical university Graz, Austria. Samples were taken twice a week and a total of 650 samples were obtained.

37 patients had Acute myeloid leukaemia which was the most common underlying disease.

Mannan antigen and Anti-mannan antibodies

Blood culture Sensitivity, specificity

66 patients had no evidence for IC. The one patient with candidaemia represented 10 samples, but from the 640 samples 153 were positive for Anti-mannan antibodies (23.9%) and 9 for Mannan antigen (1.4%).

Most false positive Anti-mannan antibodies result were observed among 375 samples from patients without hematopoietic stem cell transplantation (34.9%). Specificity for Mannan-antigen/Anti-mannan antibodies was 74.8%.

Of 10 samples obtained in the single patient with candidaemia 5 were positive for Ag and none for Ab. Conclusion that Anti-mannan antibodies yielded a high number of false positive results while Mannan Ag was found to be highly specific and may have potential for diagnostic driven testing.

Moderate

10. Jurgen Held 2013

Case-control study

56 adult patients with candidaemia confirmed by blood-culture. Serum samples from day 0 until day

BDG, Mannan antigen with/

Blood culture Sensitivity, specificity

BDG, Mannan antigen and Anti-mannan antibody levels were all separately significant elevated in patients with

(36)

2 after blood culture sampling at the University Medical Centre Freiburg, Germany.

100 patients with bacteraemia and 100 patients with sterile blood cultures served as negative controls.

Underlying diseases were hematologic malignancy/ hematopoietic stem cell transplantation (n=9), solid tumor (n=12), gastrointestinal disease (n=6), major surgery (n=4), abdominal surgery (n=16) and miscellaneous disease (n=9).

mannan antibody and Cand-Tec Candida antigen

With a sensitivity at (58.9-87.5%) and specificity (65.0-97.5%).

For the Mannan antigen/Anti-mannan antibodies the sensitivity was 89.3% and specificity 63.0% and for BDG/Mannan antigen 80.3% sensitivity and 85.0% specificity.

Candi-Tec Candida antigen had a sensitivity at 13.0% and specificity at 93.9%.

Of the four tests compared, BDG and Mannan antigen are the superior biomarkers, depending on whether a sensitivity-driven or specificity-driven approach is used. 11. M. Carmen Martinez-Jimenez 2015 Prospective observational study

63 adult ICU-patients and 37 adult non-ICU patients starting empirical antifungal treatment for suspected IC at Gregorio Maranon Hospital, Madrid (Spain).

High-risk gastrointestinal surgery and sepsis in non-surgical patients where the main indications for empirical treatment.

Later classified as no IC (58%), proven IC (30%) and probable IC (12%).

Common underlying disease was solid tumors (22 patients), intestinal perforation (20), severe cardiovascular disorders (19).

CAGTA and BDG

Blood culture Sensitivity, PPV, NPV

Sensitivity and NPV of CAGTA/BDG was 97% for the entire population.

Best sensitivity and NPV was obtained in ICU-patients (100%).

In the patient group without IC all biomarkers were negative in 31 patients.

CAGTA/BDG could be used to safely stop antifungals in patients receiving empirical antifungal therapy, both in the ICU and in the non-ICU wards.

High 12. Hemilio Xafranski 2013 Evaluation study

160 yeast-positive blood culture bottles.

Negative controls were DNA samples from blood-cultures that represented sepsis due to Gram-positive and Gram-negative bacteria as well ten negative blood cultures.

As a positive control a genomic DNA from the reference strain Candida albicans was used.

ITS-PCR (analyzing fungal DNA in yeast positive blood culture bottles to identify common and cryptic Candida species)

Blood culture Sensitivity 97% concordance for ITS-PCR with the conventional culture method in the identification of species at the molecular level. ITS-PCR also detected species not common in human infections.

This PCR method provides accurate results within 24 hours, in contrast to at least 72 hours required by the automated conventional culture method, it also allows identification of mixed infections as well as infections due to emergent fungal infections.

Also 60% cost savings compared to commercial method.

References

Related documents

For the following extraction steps of DNA the QIAamp tissue kit was compared with two automated extraction robots: BioRobot M48 and NucliSens easyMAG, to determine their

Renal abnormal- ity, bladder dysfunction and breakthrough urinary tract infection were found in many study patients and were also shown to be three strong independent nega-

Renal abnormality, bladder dysfunction and breakthrough urinary tract infection were found in many study patients and were also shown to be three strong

Då den här uppsatsen handlar om hur cloud computing kan vara till nytta för olika företag och vilka för- och nackdelar som finns med det ser vi en rad möjliga intressenter:..

Results published previously by our own group looking at the same patient population as in the current study showed that patients with type 1 diabetes had elevated

If ICC is the goal of teaching culture, the textbooks analyzed in the research reviewed usually do not aid in that. The question is posed as to what this means for teaching culture

17 Other findings where LightCycler FastStart DNA Master SYBR Green I kit (Roche Diagnostics) showed higher sensitivity than Mastermix 16S Basic Kit (Molzym) included higher

(Prior to January, 2005, the series was published under the title &#34;Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and