Long-term sera storage does not significantly modify the interpretation of toxoplasmosis serologies

Journal of Microbiological Methods 134 (2017) 38–45

Contents lists available at ScienceDirect

Journal of Microbiological Methods journal homepage: www.elsevier.com/locate/jmicmeth

Long-term sera storage does not significantly modify the interpretation of toxoplasmosis serologies C Dard a,b, S Bailly a,c, T Drouet a, H Fricker-Hidalgo a, MP Brenier-Pinchart a,b, H Pelloux a,b,⁎ a b c

Laboratoire de Parasitologie-Mycologie, Institut de Biologie et Pathologie, CHU Grenoble Alpes, France Institute for Advanced Biosciences (IAB), Team Host-Pathogen Interactions and Immunity to Infection, INSERM U1209 - CNRS UMR5309, Université Grenoble Alpes, 38700 Grenoble, France UMR 1137-IAME Team 5-DeSCID, Inserm/Paris Diderot, Université Sorbonne Paris Cité, Paris, France

a r t i c l e

i n f o

Article history: Received 24 October 2016 Received in revised form 3 January 2017 Accepted 4 January 2017 Available online 16 January 2017 Keywords: Toxoplasma gondii Toxoplasmosis Serum Stability Storage Biobank Freezing Parasitology

a b s t r a c t Background: Serological investigation of Toxoplasma gondii can answer many questions about toxoplasmosis in human pathology. Along these lines, studies on serum storage in biobanks need to be performed especially in terms of determining the impact of storage on relevance of sera analysis after freezing. This study assessed the impact of long-term sera storage on the stability of anti-Toxoplasma immunoglobulins. Material and methods: The stability of anti-Toxoplasma IgG and IgM was studied in 244 and 242 sera respectively, stored at −20 °C from one month to ten years. ELISA-immunoassay (Vidas®, bioMérieux) was used for initial and post-storage analyses. Linear models for repeated measures and subgroup analyses were performed to assess the effect of storage duration and sample characteristics on immunoglobulins stability. Results: Until ten years, the variability attributed to storage (maximum 8.07% for IgG, 13.17% for IgM) was below the variations inherent to the serological technique and allowed by quality assurance systems (15%). Subgroup analysis reported no variation attributed to sera storage. Serological interpretation was modified for 3 sera (1.2%) tested for IgM, all stored more than seven years. Conclusion: Anti-Toxoplasma immunoglobulins can reliably be measured for at least up to six years of storage with no modification of interpretation of toxoplasmosis serologies. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Toxoplasmosis is a parasitic disease caused by Toxoplasma gondii, an intracellular protozoan parasite belonging to the Apicomplexan family. Toxoplasmosis is a global health hazard as the worldwide seroprevalence is around 50% (Montoya and Liesenfeld, 2004). Infection with T. gondii can be acquired by the ingestion of T. gondii cysts in undercooked meat or oocysts from the environment (Montoya and Liesenfeld, 2004). Three main clinical entities in humans are generally distinguished: primary infection with T. gondii in immunocompetent individuals, which is generally asymptomatic and benign; mother-to-child transmission, which can cause serious fetal malformations; and reactivation in immunocompromised patients, which can lead to death (Montoya and Liesenfeld, 2004). Serological tests are used to diagnose toxoplasmosis in most clinical contexts (Dard et al., 2016; Robert-Gangneux and Dardé, 2012). Usually, anti-T. gondii antibodies are rapidly analyzed after blood sampling. If immediate analysis is not possible, manufacturer package ⁎ Corresponding author at: Laboratoire de Parasitologie-Mycologie, Institut de Biologie et Pathologie, Département des Agents Infectieux, CHU Grenoble Alpes, CS10217, 38043 cedex 9 Grenoble, France. E-mail address: [email protected] (H. Pelloux).

http://dx.doi.org/10.1016/j.mimet.2017.01.003 0167-7012/© 2017 Elsevier B.V. All rights reserved.

insert allows storage of sera at 2 to 8 °C before testing, usually b 7 days (Bennet, 1980. Carpenter, 1997; National Committee for Clinical Laboratory Standards, 2001). If longer storage is needed after routine laboratory analysis, the leftover sera - usually under 1.5 mL - can be stored frozen in biobanks. As sera should be thawed and utilized only once after thawing according to the manufacturer guidelines, the use of single usage aliquots is recommended with repartition of the leftover of sera in several aliquots. Within the framework of French legislation and to ensure high quality health care, sera collected for toxoplasmosis serology are consistently stored at −20 °C in biobanks for at least one year (https://www.legifrance.gouv.fr/eli/arrete/2005/9/20/ SANS0523407A/jo/texte). The main goal of sera conservation is to permit further analysis later in case of medical needs. This storage process enables us to analyze sequential sera in the same run to assess stability of immunoglobulin. After the legal one year, the frozen sera can be used for various retrospective studies including sero-epidemiological investigations, quality control and methods comparison studies (Murat et al., 2013a, 2013b). A number of studies have examined how storage conditions affect the stability of various biochemical components (Gao et al., 2007; Hsing et al., 1989; Kale et al., 2012; Tanner et al., 2008; Donnelly et al., 1995; Cuhadar et al., 2012; Boyanton and Blick, 2002; Cuhadar et al., 2013). However, very little information exists on the stability of

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

immunoglobulins in human sera and what constitutes optimal sera storage conditions (Ellis et al., 2004; Pappin et al., 1995; Rosa-Fraile et al., 2004). Indeed, few studies have investigated the stability of antiToxoplasma antibodies after short-term freezing but none after longterm storage. Moreover, these studies were based on limited samples and did not consider sera characteristics, related clinical and serological settings. Given the need of the current quality assurance system for robust data and more knowledge about immunoglobulin stability of sera frozen in biobanks, we evaluated the biological relevance of anti-Toxoplasma IgG and IgM analysis in sera stored at −20 °C for several years in biobanks. 2. Materials and methods 2.1. Study design Sera were retrospectively selected from the biobank collection of the Parasitology-Mycology Clinical Laboratory in Grenoble Alpes University Hospital in France. This biobank is registered with the French Ministry of Health number DC-2008-582. The selected sera were stored for toxoplasmosis serological routine analysis between January 1, 2005 and December 31, 2014. All the analyses were performed in the ParasitologyMycology Clinical Laboratory of Grenoble Alpes University Hospital with Vidas® Toxo IgM and Vidas® Toxo IgG reagents (bioMérieux, France). 2.2. Preanalytical phase All the sera were obtained from blood samples collected in 5 or 7 mL polyethylene terephthalate (PET) top BD Vacutainer® tubes with clot activator, either with or without serum-gel separator. Within 72 h of the blood draw, the sera were separated by centrifugation at 2000g for 15 min at room temperature and initial anti-Toxoplasma IgG and IgM determinations were performed on the same day. After initial testing, the remaining sera were first stored at 4 °C until the end of each week to allow additional analysis if needed. Remaining sera were then transferred into 2 mL graduated polypropylene cryotubes (VWR, Radnor, USA) and frozen at − 20 °C in our laboratory biobank for long-term storage. 2.3. Sample selection and characteristics Thirty to 38 sera samples were selected for each year - from 2005 to 2014 - for IgG and/or IgM analysis (Table S1 in Supplementary data). The selected sera vary in terms of volume stored in biobanks, immunoglobulin levels and clinical and immune status of the corresponding patient. All the sera came from different patients (only one sample per patient). Mainly sera with equivocal and positive immunoglobulin levels on initial analysis were selected (Table 1). Only one serum with negative IgM was included for IgM analysis as its value was very close to the equivocal threshold. Sera with IgG values upper limit of the linear range (IgG N 300 IU/mL) were excluded for post-storage IgG titration. For each serum, corresponding patient and sample characteristics age, gender, clinical setting category, Toxoplasma immune status and date of sampling, volume of frozen sample, date of initial analysis, and date of second analysis - were documented (Table 1).

39

Table 1 Descriptive of repartition and groups of sera using quantitative and qualitative variables. Groups Quantitative variables: mean (±std) Storage duration NA (months) Initial aliquot volume NA (mL) Age (years) NA Qualitative variables: N (%) All sera Storage duration b1 year 1 to 2 years (Time of conservation 2 to 3 years between the initial and 3 to 4 years the last analysis) 4 to 5 years 5 to 6 years 6 to 7 years 7 to 8 years 8 to 9 years 9 to 10 years

IgM IgG (Ntotal = 244) (Ntotal = 242) 60.9 (±35.7)

61.2 (±35.2)

0.8 (±0.5)

0.8 (±0.4)

34.7 (±19.1)

35.5 (±13.8)

244 (100) 25 (10.2) 27 (11.1) 25 (10.2) 25 (10.2) 23 (9.4) 27 (11.1) 23 (9.4) 25 (10.2) 23 (9.4) 21 (8.6)

242 (100) 25 (10.3) 23 (9.5) 26 (10.7) 25 (10.3) 25 (10.3) 29 (12.0) 22 (9.1) 25 (10.3) 18 (7.4) 24 (9.9)

166 (68.0) 78 (32.0) 104 (42.6)

188 (77.7) 54 (22.3) 137 (56.6)

58 (23.8)

45 (18.6)

45 (18.4)

44 (18.2)

27 (11.1)

0 (0)

10 (4.1) 108 (44.3)

16 (6.6) 94 (38.8)

Description of groups and subgroups Groups Subgroups Sex F M Pregnant women Clinical settings serological follow-up categoriesa Immunocompetent patients investigation Immunocompromised patients follow-up Infants suspected of congenital toxoplasmosis Undetermined Toxoplasma immune Past immunity b status (N6 months) Recent T. gondii infection (b6 months) Non immunized Undetermined Initial Ig rate category Negative Equivocal Positive Volume of frozen V = 0.10–0.7 mL aliquots V = 0.8–2 mL

54 (22.1)

80 (33.1)

0 (0) 82 (33.6) 0 (0) 37 (15.2) 207 (84.8) 123 (50.4) 121 (49.6)

8 (3.3) 60 (24.8) 1 (0.4) 39 (16.1) 202 (83.5) 118 (48.8) 124 (51.2)

Description of set of reagent Set of reagents A B

147 (60.2) 97 (39.8)

96 (39.7) 146 (60.3)

The quantitative and qualitative variables were respectively expressed by using mean (± standard deviation) and N (%). - Quantitative variables: descriptive of the sera according the mean age, initial volume of sampling and storage duration. - Qualitative variables: repartition of the sera according their storage duration, their initial and final immunoglobulin rates and the characteristics of the related patient with their sex, clinical settings categories and Toxoplasma immune status. a 10 IgG sera and 16 IgM sera non-classified in clinical settings categories due to lack of clinical information in our health records. b 82 IgG sera and 58 IgM sera non-classified in Toxoplasma immune status either because taken from immunocompromised patients or infants suspected of congenital toxoplasmosis, either due to lack of clinical information in our health records.

2.4. Description of sera repartition and categories 2.4.1. Sera selected for IgG analysis Two hundred and forty-four sera that had been stored for up to 10 years were included for anti-Toxoplasma IgG analysis (Table 1). Mean storage duration was 60.9 months (±35.7). Mean sample volume was 0.8 mL (± 0.5). Sera came from patients with a mean age of 34.7 years (±19.1) and mainly from women (68%). The clinical-categories were: pregnancy monitoring (104, 42.6%), immunocompetent

patients (58, 23.8%), immunocompromised patients (45, 18.4%) and suspected congenital toxoplasmosis (27, 11.1%). Ten sera (4.1%) were not classified due to lack of information. As for the Toxoplasma immune status, 108 patients (44.3%) had a past infection, 54 (22.1%) a recent T. gondii infection, whereas 82 (33.6%) could not be classified, as they belonged to immunocompromised patients and suspected congenital toxoplasmosis groups (Table 1).

40

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

2.4.2. Sera selected for IgM analysis Two hundred and forty-two IgM sera were included from 10 years storage period. The mean storage duration was 61.2 months (±35.2) (Table 1). Mean sample volume was 0.8 mL (±0.4). Sera came from patients with a mean age of 35.5 years (±13.8) and mainly from women (77.7%). The main clinical-setting reasons for sampling were pregnancy monitoring for 137 patients (56.6%), followed by investigation for immunocompetent patients for 45 sera (18.6%) and immunocompromised for 44 sera (18.2%) whereas the reason for original testing of 16 patient's sera (6.6%) was not explained in the records. Concerning the Toxoplasma immune status, 94 sera (38.8%) had a past infection, 80 sera (33.1%) had an acute T. gondii infection and 8 sera (3.3%) were attributed to non-infected patients (non-specific IgM). Sixty sera (24.8%) were not classified as they were collected from immunocompromised patients biased with abnormal antibodies secretion. 2.5. Toxoplasmosis serological testing After thawing at room temperature for 30 min, selected samples were gently vortexed and assessed for anti-T. gondii antibodies. The samples were analyzed with ELISA immunoassay (Vidas® Toxo IgG II and Vidas® Toxo IgM, bioMérieux, Marcy l'Etoile, France) (Alvarado-Esquivel et al., 2002; Roux-Buisson et al., 2005) to determine T. gondii antibody levels after storage. Two different sets of reagents were used for IgG and IgM analysis each. Antibody titers for IgG were quantitatively expressed in IU/mL whereas IgM were expressed as an index. The cut offs defined by manufacturer were: (i) IgG (IU/mL): negative b 4.0; 4.0 ≤ equivocal (grey zone) b 8; ≥ 8 positive; (ii) IgM (index): negative b 0.55; 0.55 ≤ equivocal (grey zone) b 0.65; ≥0.65 positive. Positive and negative controls were performed on each series of tests. All testing was done in accordance with the manufacturer's guidelines. 2.6. Statistical analysis Sera and patient characteristics were described by median and interquartile ranges for quantitative data and as frequencies and percent for qualitative data. Variable normality was assessed using Shapiro-Wilk test and normal quantile plots. Non-normal variables were changed into their logarithmic values. Univariate analyses were performed to compare IgG and IgM results separately for each year using the Wilcoxon signed rank sum test. To assess the impact of the conservation interval on Ig variability, a linear mixed model for repeated measures was performed, including random intercept and slope (statistical analysis in Supplementary material). Subgroup analyses were performed in five groups of sera: (i) sex (male or female); (ii) clinical settings categories; (iii) Toxoplasma immune status; (iv) initial immunoglobulin category (equivocal, positive); (v) initial volume of sera (low or high). If information was available, IgM values were classified as residual or non-specific IgM. Four different clinical setting subgroups and three Toxoplasma immune status subgroups were established (statistical analysis in Supplementary material). To subtract the intrinsic internal variability attributed to the Vidas® method to the variability observed after storage, the confidence interval of the correction coefficient and the accuracy of correction factor (Δ) were calculated (statistical analysis in Supplementary material). The intrinsic variability accepted by the assurance quality system is considered as 15% and was therefore selected as the standard value (Le Vacon, 2007; Jacobson, 1998). To compare the variability between the two set of reagents used, the variation for each measure of IgG and IgM was calculated as follows: V = (post-storage value − initial value) / initial value. The absolute values of these variations were considered and the coefficient of variation calculated for each set of reagents. Data management and descriptive analyses were performed using SAS v9.4 (SAS Institute, North Carolina) and linear mixed models were performed using nlme package

(Pinheiro et al., 2007) from R v3.0.2. A p-value of 0.05 was considered significant for statistical analysis. 3. Results A total of 339 serum samples from 339 different patients were included in this study; 147 were tested for anti-Toxoplasma IgM and IgG; 97 were specifically tested for anti-Toxoplasma IgG and 95 for anti-Toxoplasma IgM only (Table S1 in Supplementary material). 3.1. Impact of long term freezing on anti-Toxoplasma IgG: comparison of results between initial and post-storage titers 3.1.1. Description of IgG values IgG concentrations ranged from 4 to 300 IU/mL and from 4 to N300 IU/mL, respectively for initial and post-storage titrations. The mean (median) values were 102.7 IU/mL (71.5 IU/mL) and 89.4 IU/mL (84.0 IU/mL) respectively (Table 2). Initial analyses reported 207 (84.8%) positive and 37 (15.2%) equivocal sera, whereas post-storage analysis reported 223 (91.4%) positive and 21 (8.6%) equivocal sera (Table 3). 3.1.2. Global analysis of IgG variation Regardless of group classification, statistical analysis by Wilcoxon sign rank test showed significant differences between initial and poststorage IgG levels (p b 0.01) (Table 2). However, the variation attributed to the storage is lower than that of the inherent variability of serological method accepted by the quality assurance system (15%) up to 10 years of storage as the accuracy of correction factor Δ is always b15% (Table 2). 3.1.3. Analysis of IgG variation according to groups and subgroups When IgG values were compared before and after storage for each period (years) of storage, statistical significant differences between the first and the second antibody titration were observed for the sera stored for 2, 3, 5 and 10 years (Table 2). There was a statistically significant increase among samples stored for 2 and 10 years and a significant decrease in the IgG value for samples stored for 3 and 5 years (Table 2). However, accuracy of correction factors shows that the variation attributed to the duration of storage is lower than the inherent variability of the method until 10 years of storage (Δ factors b 15%, Table 2). Among the 5 groups of sera including 16 subgroups, a significant effect of storage duration was observed for only four sub-groups (p b 0.05 and significant random slope) (Table 2). These subgroups are sera from male cases (p = 0.05), sera from suspected congenital toxoplasmosis cases (p = 0.04), sera from patients with recent immunity (infection acquired within 6 months) (p = 0.04) and sera with low initial volume of sample (V b 0.7 mL) (p b 0.01). The variation attributed to the storage duration was always lower than that of the inherent variability of the method until 10 years (Δ factors b 15%, Table S2, Supplementary data), apart from the subgroup sera from suspected congenital toxoplasmosis. For this last one, Δ factors were N 15% above 5 years of storage, but the sample size of this subgroup was low (N = 27) and consequently may have lower statistical power. 3.1.4. Comparison of variability between the two set of reagents No significant difference was observed between the two sets of reagents (sets A and B) (Table 2). Indeed, the variability between initial value and post-storage value is the same with the two reagent kits used (14% and 13%) (Tables 1 and 2). 3.1.5. Classification on the basis of biological interpretation The biological interpretation (positive, equivocal, negative) remained unchanged for 226 sera (92.6%) and modified for 18 sera (7.4%) between the initial and the post storage result (Table 3). Among the 207 positive sera, 206 sera stayed positive whereas 1

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

41

Table 2 Comparison of anti-Toxoplasma IgG and IgM levels between initial and post-storage assessment.

IgG Sera classification

All sera

IgM

Initial level (IU/mL)

Post-storing level (IU/mL)

Variation analysis

Initial level (index)

Poststoring level (index)

71.5 [15; 183.5]

84.0 [19.5; 145]

p < 0.01*

2.00 [0.83; 3.62]

1.94 [0.76; 3.69]

NS 0.84%

2.25 [0.85; 3.5]

2.08 [0.79; 3.78]

Δ

Variation analysis

p < 0.01*

Analysis of storage duration impact

Storage duration

<1 year

33 [10; 180]

39 [13; 139]

1 to 2

96 [18; 165]

101 [21; 181]

p = 0.01* Δ 1.67%

2.07 [0.82; 3.62]

2.09 [0.93; 3.66]

Δ

2 to 3

113 [17; 172]

87 [17; 102]

p < 0.01* Δ 2.49%

1.65 [0.71; 3.75]

1.56 [0.71; 3.65]

Δ

3 to 4

69 [16; 187]

66 [18; 131]

NS 3.31%

2.28 [0.93; 3.36]

2.19 [0.76; 3.35]

Δ

4 to 5

112 [8; 221]

87 [11; 149]

p < 0.01* Δ 4.12%

1.92 [0.98; 3.74]

1.72 [0.76; 3.76]

Δ

5 to 6

96 [19; 198]

80 [30; 142]

Δ

NS 4.93%

1.99 [1.02; 3.86]

1.82 [0.92; 3.61]

Δ

6 to 7

72 [14; 192]

82 [22; 125]

Δ

NS 5.72%

1.99 [0.82; 3.24]

1.99 [0.75; 4.06]

7 to 8

83 [15; 164]

99 [19; 153]

Δ

NS 6.51%

2.14 [1.02; 3.15]

1.97 [0.76; 3.32]

Δ

8 to 9

61 [14; 214]

97 [31; 200]

Δ

NS 7.30%

1.55 [0.72; 3.84]

1.48 [0.71; 4.39]

Δ

9 to 10

90 [25; 192]

109 [36; 202]

p = 0.02* Δ 8.07%

2.07 [0.73; 4.03]

1.21 [0.72; 4.2]

Δ

Δ

Δ

NS 1.40% NS 2.79% NS 4.15% NS 5.49% NS 6.82% NS 8.13%

p = 0.03* Δ 9.42% NS 10.69% NS 11.94% NS 13.17%

Analysis of groups and subgroups impact Groups1

Subgroups2 F

69 [17; 179]

81 [21; 142]

NS

2.08 [0.80; 3.61]

1.97 [0.73; 3.68]

NS

M

90 [8; 192]

87 [11; 149]

p = 0.05*

1.92 [1.01; 3.74]

1.80 [0.91; 3.78]

NS

Pregnant women serological followup

88.5 [26.5; 188]

91 [27; 138.5]

NS

2.07 [0.74; 3.59]

1.96 [0.75; 3.67]

p = 0.02*

Immunocompetent patients investigation

109 [22; 198]

113 [26; 158]

NS

1.74 [0.88; 3.24]

1.77 [0.77; 3.66]

NS

Immunocompromise d patients follow-up

50 [11; 172]

45 [14; 135]

NS

2.2 [0.90; 3.89]

1.90 [0.80; 3.91]

NS

Infants suspected of congenital toxoplasmosis

7 [5; 25]

10 [7; 57]

NA

NA

NA

Past immunity (> 6 months)

126 [49.5; 194]

104.5 [53.5; 154]

NS

0.925 [0.61; 2.01]

0.92 [0.62; 2.02]

NS

Recent immunity (< 6 months)

35.5 [14; 179]

33 [16; 142]

p = 0.04*

3.43 [2.26; 5.86]

3.675 [2.24; 5.86]

p = 0.04*

Non immunized

NA

NA

NA

1.14 [0.68; 1.53]

0.93 [0.75; 1.6]

NA

Sex

Clinical settings categories *

Toxoplasm a immune status**

p < 0.01*

42

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

Table 2 (continued)

Initial Ig rate category

Volume of frozen aliquots

Negative

NA

NA

NA

NA

NA

NA

Equivocal

5 [5; 7]

7 [6; 8]

NS

0.59 [0.57; 0.61]

0.61 [0.56; 0.69]

NS

Positive

105 [28; 198]

99 [36; 158]

NS

2.43 [1.31; 3.97]

2.40 [1.08; 4.31]

NS

V = 0.10 - 0.7 mL

55 [10; 165]

63 [13; 139]

p < 0.01*

1.97 [0.93; 3.62]

1.85 [0.78; 3.69]

NS

V = 0.8 – 2 mL

98 [19; 197]

89 [23; 157]

NS

2.07 [0.83; 3.67]

2.00 [0.73; 3.71]

NS

Analysis of impact of different set

Set of reagents

of reagents

Lot A

72 [14; 189]

80 [18; 140]

CV = 14%

2.07 [0.83; 3.35]

1.85 [0.75; 3.52]

CV = 13%

Lot B

70 [16; 169]

87 [20; 158]

CV = 13%

1.97 [0.84; 3.75]

1.94 [0.77; 3.78]

CV = 7%

p-value ≤ 0.05

became equivocal. From the 37 initially equivocal sera, 20 remained equivocal and 17 became positive (Table 3).

significant variation due to storage duration is observed for any of the other subgroups.

3.2. Impact of long term freezing on anti-Toxoplasma IgM: comparison of results between initial and post-storing analyses

3.2.4. Comparison of variability between the two set of reagents A significant variation was observed according to the two sets of reagents (lots A and B) used. For IgM analysis, the variability is higher with the reagent batch A (CV = 13%) than with batch B (CV = 7%) (Table 2).

3.2.1. Description of IgM values The IgM levels ranged from 0.54 to 10.75 (index) and from 0.30 to 12.14 (index), respectively for initial and post-storage titrations (data not shown). The mean (median) values were 2.71 (2.00) and 2.78 (1.94) respectively. Initial analyses reported 202 (83.5%) positive values, 39 (16.1%) equivocal sera and 1 (0.4%) negative serum (Table 3). Post-storage analyses reported 204 (84.3%) positive values, 27 (11.2%) equivocal values and 11 (4.5%) negative values (Table 3). 3.2.2. Global analysis of IgM variation Regardless of sera groups, statistical analysis by Wilcoxon sign rank test showed significant differences between initial and post-storage immunoglobulin levels (p b 0.01) (Table 2). However, the variation attributed to the storage is lower than that of the inherent variability of the serological method accepted by the quality assurance system (15%) up to 10 years of storage as the accuracy of correction factor Δ is always b15% (Table 2). 3.2.3. Analysis of IgM variation according to groups and subgroups When IgM values were compared pre- and post-storage for each period of storage, significant differences between the first and the second antibody titrations were only observed for the sera stored for 7 years (p = 0.03) (Table 2). However, accuracy of correction factors showed that the variation attributed to the time effect of storage (Δ = 9.42%) is lower than that of the inherent variability of the serological method accepted by the quality assurance system (15%) as the accuracy of correction factor Δ is b15% (Table 2). A significant variation after storage was observed for only two subgroups of sera (p b 0.05 and significant random slope) among 15 subgroups (Table 2). These categories included the sera from pregnant women (p = 0.02) and the sera taken during recent T. gondii infection (p = 0.04). Average correction coefficients were calculated for these two subgroups. Accuracy of correction factor showed that the variation attributed to the effect of freezing on storage is lower than that of the inherent variability of the method (Table S2, Supplementary material). No

3.2.5. Classification according biological interpretation The biological interpretation (positive, equivocal, negative) was unchanged for 213 sera (88.0%) and modified for 29 sera (12.0%) between the initial and the post-storage analysis (Table 3). Of the 202 positive sera at initial dosage, 192 (79.3%) sera were positive, 5 (2.1%) were equivocal and 5 (2.1%) were negative after storage (Table 3). Of the 39 equivocal sera, 21 (8.6%) remained equivocal, 12 (5.0%) became positive and 6 (2.5%) became negative (Table 3). The unique negative (at 0.54 just below the equivocal threshold) serum became equivocal (0.57) after the second analysis. 4. Discussion Sera for toxoplasmosis diagnostic purposes are usually stored for one year in laboratories biobanks, except for toxoplasmosis reference centers, which keep the sera stored for longer periods for research purposes (Murat et al., 2013a, 2013b; Gay-Andrieu et al., 2009; Villard et al., 2013). Thus, the main objective of this study was to evaluate the impact of long-term storage at −20 °C on anti-T. gondii IgG and IgM stability and their efficacy after prolonged storage on Toxoplosmosis diagnosis. Prerequisites for this study were to use the same serological technique before and after sera storage and to use sera that have not been thawed since they were first frozen. The ELISA-immunoassay Vidas® was selected to perform this study as this technique has not been modified by the manufacturer bioMérieux for N 10 years and hence give comparable anti-Toxoplasma IgG and IgM results before and after sera storage. Throughout the entire 10-years observation period, sera were not voluntarily or forcibly thawed. Biobank did not encounter any problems - i.e. power or freezer failures – that may have result in thawing of the stored sera. The variability between initial and post-storage analysis has been studied in light with various parameters including (i) the intrinsic variability of the serological method, (ii) the effect of long-term storage, (iii)

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

43

Table 3 Modification of sera classification between the first and second analysis according to the manufacturer cut-offs.

IgG (n = 244) Positive (poststoring)

Equivocal (poststoring)

IgM (n = 242) Negative (poststoring)

Total

Positive (poststoring)

Equivocal (poststoring)

Negative (poststoring)

5 (2.1%) 1 (0.4%)

Positive (initial)

206 (84.4%)

Detailed value (IU/mL)

0 (0%)

207

192 (79.3%)

Detailed values (index) Initial-Final 0.70 0.71 0.77 0.77 1.07

Initial-Final 9 7

0.59 0.60 0.56 0.62 0.60

Total

5 (2.1%) Detailed values (index)* Initial - Final 0.69 0.45 (0.29**) 0.72 0.47 (0.35**) 0.88 0.32 (0.21**) 1.72 0.30 2.44 0.30

202

17 (7.0%)

Equivocal (initial)

Detailed values (IU/mL) InitialFinal 4 8 4 9 5 8 5 8 5 8 5 9 5 9 5 9 6 10 6 10 7 8 7 8 7 8 7 10 7 12 7 14 7 15

12 (5.0%) Detailed values (index) InitialFinal

20 (8.2%)

0 (0%)

37

0.56 0.57 0.58 0.58 0.58 0.59 0.59 0.62 0.63 0.63 0.63 0.63

0.72 0.68 0.69 0.71 1.33 0.70 0.78 0.73 0.68 0.71 0.72 0.74

6 (2.5%)

21 (8.6%)

Detailed values (index) Initial-Final 0.55 0.51 0.56 0.54 0.57 0.50 0.58 0.52 0.60 0.46 0.62 0.52

39

1 (0.4%) Negative (initial)

Total

0 (0%)

0 (0%)

0 (0%)

0

0 (0%)

223

21

0

244

204

the aliquot volumes of sera and (iv) the clinical characteristics and immunity of the patient. Not only the statistical variation of the immunoglobulin levels between initial and post-storage dosage has been considered. The variations between first and post-storage analysis were also evaluated in accordance with their possible clinical impact. Therefore, a distinction has been drawn between the biological interpretations related to immunoglobulins quantitative categories defined by the manufacturer (negative, equivocal, positive) from the broader interpretation of serological results by an expert. Serological interpretation by an expert was done for conflicting results, on the basis of quantitative and qualitative analysis of Ig and clinical features and immunity of each patient, to determine Toxoplasma immune status (e.g. recent T. gondii infection, past immunity and reactivation). Therefore, results have been discussed in one strict statistical part,

Detailed values (index) Initial-Final 0.54 0.57 27

0 (0%)

1

11

242

one part in light with quantitative categories modifications and one based on serological interpretation by an expert. 4.1. Statistical analysis of IgG and IgM levels The statistically significant variation of IgG and IgM between initial and post-storage results was mainly attributed to the intrinsic and usual variability of the Vidas serological method, when considering the 15% accepted variability of a serological technique. The variability due to long-term storage remained b15% for both IgG and IgM (8.07% and 13.17% respectively at 10 years of storage). First, variability according to the time of storage (from 1 to 10 years) has been evaluated. During the first 10 years of storage, the global variability observed between initial and post-storage analysis was not directly correlated to the duration of storage. Indeed, no global trend of

44

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

time effect was observed regarding the increase or decrease of immunoglobulin rates after storage. Then, the impact of three clinical (sex, clinical settings, Toxoplasma immune status of patients) and two biological (initial Ig quantitative category, volume of frozen aliquots) groups of parameters on Ig stability has been evaluated. Each group of clinical and biological parameters has been divided in several subgroups. Statistical analysis showed no major significant impact on different subgroups studied. We observed a significant variation after storage of IgG and IgM from sera of patients with recent immunity for toxoplasmosis (b 6 months), of IgG for sera taken from male patients, sera with low volume and sera from an infant suspected for CT and of IgM only for sera from pregnant women. However, for most of these subgroups, the variation was more explained by the intrinsic variability of the method than the effect of sera storage for up to 10 years. Only sera from infants suspected of congenital toxoplasmosis had a variability N15% attributed to sera storage beyond 5 years of storage. This can be related to the small sample size (Table S2, Supplementary data) but also to the special features of IgG in these sera. Indeed, the provenance and maturity of IgG is uncertain in these sera as IgG can be either transplacentally transmitted, either directly synthetized by the newborn with an immature immune system. 4.2. Quantitative categories modifications Concerning modification of quantitative categories for IgG (7.4% of IgG sera), most of the sera with a modified biological interpretation after second testing had initial equivocal or low positive IgG level. Most of the changed categories (7%) concerned initial equivocal rates that become weakly positive, always around the grey zone (4– 8 IU/mL) and ≤ 15 IU/mL at the post-storage analysis. This can be explained by the intrinsic variability of the method at a level near cutoff. Only one positive serum (0.4%) became equivocal after storage, but had an initial value near the threshold (from 9 IU/mL to 7 IU/mL, positive threshold ≥ 8 IU/mL). This serum comes from a 2-year old child with a symptomatic T. gondii primary infection with detection of IgG at low titer. Despite the change of IgG quantitative categories for these sera, the second interpretation of serological results by an expert has not been modified compared to the initial. The modification of quantitative categories for IgM mainly concerned sera with residual or non-specific IgM, probably because their level is generally low (data not shown). Residual IgM corresponds to levels that remain detectable several months or years after acute infection whereas non-specific immunoglobulins correspond to cross reactions with others IgM, i.e. from viral infections. Among the twelve sera that passed from equivocal to weakly positive (5%), most of them were classified as residual or non-specific IgM and one was taken from a seroconversion during pregnancy. The 6 equivocal sera that became negative and 5 positive sera that became equivocal concerned almost only residual IgM, except for one serum that concerned seroconversion during pregnancy (from 1.07 to 0.6 (index)). The negative serum with residual IgM under the equivocal threshold became equivocal after storage (from 0.54 to 0.57 (index)) due to the inherent variability of the method. 4.3. Serological interpretation by an expert The 7.4% and 12.0% of sera with modifications of quantitative categories for IgG and IgM respectively have been reinterpreted by an expert, considering all the serological, biological and clinical data available. Concerning the sera analyzed for IgG (with or without IgM), no interpretation modification was noticed between the first and post storage analysis, when pre- and post-storage equivocal results were not considered. Among the sera analyzed for IgM (with or without IgG), the five positive sera that became negative (2.1%) after the second analysis were the most critical for our laboratory experts. These sera were stored for 7 to

10 years, showing that sera analysis for anti-Toxoplasma IgM becomes critical beyond 7 years of storage. Post-storage serological results have been confirmed by a second dosage for three of these sera (i, ii, iv) whereas the two others were not reanalyzed because of insufficient volume (iii and v) (Table 3). Finally, a modification of serological interpretation by our laboratory experts was observed for 3 of them (3/242 = 1.2% of total IgM sera), when equivocal results were not considered (Table 3, initial positive IgM measured negative after storing). The first serum (i) concerned an immunocompetent organ donor with residual IgM at 0.72 (positive) at first titration and 0.47 (negative) after 8years of storage. The second serum (ii) concerned an immunocompetent patient with residual IgM at 0.88 (positive) at first titration and 0.32 (negative) after 8-years of storage. For these two sera, the observed variations would not change the biological interpretation of past immunity, as IgM were here considered as residual. The third serum (iii) was taken from a 36 year-old woman with IgG N 300.0 UI/mL in an unknown clinical context. The initial dosage showed IgM = 1.72 (positive) and IgM = 0.3 (negative) after 10 years of storage. The fourth serum (iv), also stored for 10 years, belonged to a pregnant woman with no IgG and apparition of IgM due to recent primary infection. The initial IgM index was 0.69 (positive) and the second 0.45 (negative). The fifth serum (v) stored in a 0.5 mL aliquot was taken from a 21-year-old woman with unknown clinical context. Results reported initial index at 2.44 (positive) and second index at 0.3 (negative). The volume here was about 0.5 mL with unknown context. No explanations were found for the variations observed for these five sera, but all were stored for N7 years. Concerning the comparison of the two sets of reagents, a statistically significant difference between the two sets is observed only for IgM. However, the analysis of each serum two times with both sets of reagents would have allowed a more relevant statistical analysis and a comparable coefficient of variation. Results from all eight hemolyzed sera showed no misclassification even if it has also been reported that hemolyzed serum may not be suitable for some serological analysis (Tate and Ward, 2004). This study is the first to evaluate the impact of long-term freezing on stability of anti-Toxoplasma immunoglobulins, so no comparative data are available. This evaluation has also not been done for long-term storage of samples derived from patients infected with other pathogens. However, the results are in concordance with those obtained on shorter storage periods for CMV, Toxoplasma and auto-antibodies against thyroid (Cuhadar et al., 2013; Constantine and Lana, 2003; Männistö et al., 2007). 5. Conclusion In conclusion, our study provides evidence that anti-Toxoplasma sera can be stored in deep freezers at − 20 °C for up to 7 years and can provide valid results through that time. Indeed, the interpretation of serological results was not modified between the initial and the post-storage analysis for all samples stored b 7 years. Only 1.2% of the sera tested for IgM had a modification of biological interpretation, all after 7 years of storage. Overall variability of anti-Toxoplasma immunoglobulins between pre- and post-storage was mainly due to the intrinsic and normal variability of the serological assays and not proportional to the storage duration. This study comforts the conjecture that anti-Toxoplasma IgG and IgM are stable in time when frozen at − 20 °C. Declaration of interest This work was supported by bioMérieux (IPM16319/ CDE4400121113). The Parasitology-Mycology laboratory of CHU Grenoble-Alpes received research grants from Roche, Abbott and bioMérieux.

C. Dard et al. / Journal of Microbiological Methods 134 (2017) 38–45

Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.mimet.2017.01.003. References Alvarado-Esquivel, C., Sethi, S., Janitschke, K., Hahn, H., Liesenfeld, O., 2002. Comparison of two commercially available avidity tests for toxoplasma-specific IgG antibodies. Arch. Med. Res. 33, 520–523. Bennet, C.W., 1980. Clinical Serology. Thomas Publisher, Springfield, III. Boyanton, B.L., Blick, K.E., 2002. Stability studies of twenty-four analytes in human plasma and serum. Clin. Chem. 48, 2242–2247. Carpenter, A.B., 1997. Enzyme-linked immunoassays. In: Rose, N.R., de Macario, E.C., Folds, J.D., Lane, H.C., Nakamura, R.M. (Eds.), Manual of Clinical Laboratory Immunology, fifth ed. American Society for Microbiology, Washington, D.C., pp. 20–29. Constantine, N.T., Lana, D.P., 2003. Immunoassays for the diagnosis of infectious diseases. In: Murray, P.R., Baron, E.J., Jorgensen, J.H., Pfaller, M.A., Yolken, R.H. (Eds.), Manual of Clinical Microbiology, eighth ed. American Society for Microbiology, Washington, DC, pp. 218–233. Cuhadar, S., Atay, A., Koseoglu, M., Dirican, A., Hur, A., 2012. Stability studies of common biochemical analytes in serum separator tubes with or without gel barrier subjected to various storage conditions. Biochem. Med. 22, 202–214. Cuhadar, S., Koseoglu, M., Atay, A., Dirican, A., 2013. The effect of storage time and freeze– thaw cycles on the stability of serum samples. Biochem. Med. 23, 70–77. Dard, C., Fricker-Hidalgo, H., Brenier-Pinchart, M.P., Pelloux, H., 2016. Relevance of and new developments in serology for toxoplasmosis. Trends Parasitol. 32, 492–506. Donnelly, J.G., Soldin, S.J., Nealon, D.A., Hicks, J.M., 1995. Stability of twenty-five analytes in human serum at 22 degrees C, 4 degrees C, and −20 degrees C. Pediatr. Pathol. Lab. Med J. 15, 869–874. Ellis, V., Charlett, A., Bendall, R., 2004. A comparison of IgG anti-Rubella activity in frozen serum stored in primary gel separation tubes or secondary tubes. J. Clin. Pathol. 57, 104–106. Gao, Y.C., Yuan, Z.B., Yang, Y.D., Lu, H.K., 2007. Effect of freeze-thaw cycles on serum measurements of AFP, CEA, CA125 and CA19-9. Scand. J. Clin. Lab. Invest. 67, 741–747. Gay-Andrieu, F., Fricker-Hidalgo, H., Sickinger, E., Espern, A., Brenier-Pinchart, M.P., Braun, H.B., Pelloux, H., 2009. Comparative evaluation of the ARCHITECT Toxo IgG, IgM, and IgG avidity assays for anti-toxoplasma antibodies detection in pregnant women sera. Diagn. Microbiol. Infect. Dis. 65, 279–287. Hsing, A.W., Comstock, G.W., Polk, B.F., 1989. Effect of repeated freezing and thawing on vitamins and hormones in serum. Clin. Chem. 35, 2145. Jacobson, R.H., 1998. Validation of serological assays for diagnosis of infectious diseases. Rev. Sci. Tech. Int. Off. Epizoot. 17, 469–526.

45

Kale, V.P., Patel, S.G., Gunjal, P.S., Wakchaure, S.U., Sundar, R.S., Ranvir, R.K., Jain, M.R., 2012. Effect of repeated freezing and thawing on 18 clinical chemistry analytes in rat serum. J. Am. Assoc. Lab. Anim. Sci. JAALAS 51, 475–478. Le Vacon, F., 2007. Quality control in serology laboratories: how to guarantee the quality of a result of a semi quantitative analysis. Immunol. Anal. Biol. Spéc. 22, 391–397. Männistö, T., Surcel, H., Bloigu, M., Ruokonen, A., Hartikainen, A., Järvelin, A.L., Pouta, M.R., Vääräsmäki, A.M., Suvanto-Luukkonen, E., 2007. The effect of freezing, thawing, and short- and long-term storage on serum thyrotropin, thyroid hormones, and thyroid autoantibodies: implications for analyzing samples stored in serum banks. Clin. Chem. 53, 1986–1987. Montoya, J.G., Liesenfeld, O., 2004. Toxoplasmosis. Lancet 363, 1965–1976. Murat, J.B., Dard, C., Fricker-Hidalgo, H., Dardé, M.L., Brenier-Pinchart, M.P., Pelloux, H., 2013a. Comparison of the Vidas system and two recent fully automated assays for diagnosis and follow-up of toxoplasmosis in pregnant women and newborns. Clin. Vaccine Immunol. 20, 1203–1212. Murat, J.B., Fricker-Hidalgo, H., Brenier-Pinchart, M.P., Pelloux, H., 2013b. Human toxoplasmosis: which biological diagnostic tests are best suited to which clinical situations? Expert Rev. Anti-Infect. Ther. 11, 943–956. National Committee for Clinical Laboratory Standards, 2001. Specifications for immunological testing for infectious diseases. Clin. Lab. Stand. 21, 22–24. Pappin, A., Grissom, M., Mackay, W., Huang, Y., Yomtovian, R., 1995. Stability of cytomegalovirus antibodies in plasma during prolonged storage of blood components. Clin. Diagn. Lab. Immunol. 2, 25–29. Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., 2007. Linear and nonlinear mixed effects models. R Package Version 3, p. 57. Robert-Gangneux, F., Dardé, M.L., 2012. Epidemiology of and diagnostic strategies for toxoplasmosis. Clin. Microbiol. Rev. 25, 264–296. Rosa-Fraile, M., Sampedro, A., Rodríguez-Granger, J., Camacho, E., Manrique, E., 2004. Suitability of frozen serum stored in gel separator primary sampling tubes for serological testing. Clin. Diagn. Lab. Immunol. 11, 219–221. Roux-Buisson, N., Fricker-Hidalgo, H., Foussadier, A., Rolland, D., Suchel-Jambon, A.S., Brenier-Pinchart, M.P., Pelloux, H., 2005. Comparative analysis of the VIDAS Toxo IgG IV assay in the detection of antibodies to Toxoplasma gondii. Diagn. Microbiol. Infect. Dis. 53, 79–81. Tanner, M., Kent, N., Smith, B., Fletcher, S., Lewer, M., 2008. Stability of common biochemical analytes in serum gel tubes subjected to various storage temperatures and times pre-centrifugation. Ann. Clin. Biochem. 45, 375–379. Tate, J., Ward, G., 2004. Interferences in immunoassay. Clin. Biochem. Rev. 25, 105–120. Villard, O., Breit, L., Cimon, B., Franck, J., Fricker-Hidalgo, H., Godineau, N., Houze, S., Paris, L., Pelloux, H., Villena, I., Candolfi, E., 2013. Comparison of four commercially available avidity tests for Toxoplasma gondii-specific IgG antibodies. Clin. Vaccine Immunol. 20, 197–204.