Simplified Dynabeads Method Using Light Microscopy for Enumerating TCD4+ Lymphocytes in Resource-Limited Settings

Background: We demonstrated the feasibility of implanting the Dynabeads method for CD4+ T lymphocyte enumeration in resource-poor settings (ANRS 1226 study). However, as this technique requires a fluorescence microscope which is not usually available in these settings, WHO has encouraged to simplify the method allowing TCD4+ lymphocyte counting under a light microscope. Methods: TCD4+ lymphocytes enumeration was assessed using Dynabeads after staining cells nuclei with nonfluorescent dyes and readings under light microscope (DLM). A total of 305 triple of values of CD4 cells counts were generated by both Dynabeads method using a light microscopy (DLM), Dynabeads method using a fluorescent microscope (DFM) and the single-platform flow cytometry technique (FCM). The accuracy of DLM was analyzed using 4 fresh blood samples showing 200, 400, 500 and 1000 cells/μl in FCM respectively. Correlations have been studied between the 3 methods. The DLM was then evaluated for its ability to correctly segregate absolute TCD4+ lymphocyte values at the thresholds of 200 cells/μl and 350 cells/μl. Findings: Cells nuclei staining with Sternheimer-Malbin, Turck1, and Giemsa allows TCD4+ lymphocytes enumeration using DLM. FCM has shown the greatest standard deviations and amplitudes. The reproducibility of DLM was better than FCM. The correlation coefficient between FCM and DFM was 0.975 and it was 0.973, 0.972 and 0.969 with DLM using Sternheimer-Malbin, Turck1 and Giemsa, respectively. The ability of DLM to correctly segregate TCD4+ lymphocyte values at the threshold of 200 cells/μl and 350 cells/μl was good. Conclusion: Reliable TCD4+ enumeration can be obtained with DLM. These results will contribute in resourcelimited-settings to further reduce the cost of TCD4+ lymphocytes counting and make it more widely available in peripheral laboratories and even in central laboratories that face problems with maintenance and stock-out of reagents for flow cytometers.


Introduction
Human immunodeficiency virus (HIV) is the causative agent of Acquired Immunodeficiency Syndrome (AIDS). In 2012, there were an estimated 2.3 million new HIV infections, 35.3 million people living with HIV of which 71% reside in sub-Saharan Africa, and 1.6 million AIDS patients have died [1]. The disease progression is seen as a decline in TCD4+ lymphocytes counts and an increase in the plasma HIV/RNA viral load [2]. With the introduction of highly active antiretroviral therapy (HAART), the disease progression is slowed and the HIVinfected population have an improve quality of life leading to decrease in morbidity and mortality [3]. However, one cause of concern remains the monitoring of patients on HAART therapy [4]. Monitoring HIV infection progression, time to start prophylactic treatment of opportunistic infections, initiation of HAART and the response to HAART is traditionally carried out using TCD4+ lymphocytes counts and HIV/RNA viral load. Indeed, according to WHO recommendation (http:// apps.who.int/iris/bitstream/10665/208825/1/9789241549684 _eng.pdf?ua=1) [5], as a priority, antiretroviral therapy (ART) should be initiated in all adults and adolescents with severe or advanced HIV clinical disease (WHO clinical stage 3 or 4), adults and adolescents with a TCD4+ lymphocytes count ≤ 350 cells/µl. ART should be initiated in all children less than 2 years of age or children younger than 5 years of age with WHO clinical stage 3 or 4 or CD4 count ≤ 750 cells/µl and children 5 years of age and older with WHO clinical stage 3 or 4 or CD4 count ≤ 350 cells/µl. Cotrimoxazole prophylaxis is recommend for adults (including pregnant women) with severe or advanced HIV clinical disease (WHO stage 3 or 4) and/or with a CD4 count ≤ 350 cells/µl [5] (WHO 2016).
Unfortunately, conventional methods for TCD4+ lymphocytes enumeration require expensive sophisticated equipment usually inaccessible for many areas in middle and low income-countries which carry the majority of the burden of HIV-infected population [2,4,[6][7][8]. Since the poor and economically weak third world nations cannot afford regular CD4+ T cell counts and HIV-1/RNA viral load testing, cheaper and alternative methods for HIV disease progression were needed [7]. So, researchers, WHO, people living with AIDS and NGO across the globe were actively focused for less expensive tools to measure TCD4+ cell in Low income countries. Manual TCD4+ methods such as Dynabeads (Dynal Biotech, Oslo, Norway) are less expensive alternatives to flow cytometry in resource-limited countries [9]. Previous study demonstrated the feasibility of the implementation of Dynabeads methods as an alternative technique to flow cytometry for TCD4+ lymphocytes counting in resource-limited settings [7]. However, this method requires fluorescent microscope which is expensive and not usually available in the majority of peripheral laboratories in these settings.
This study aimed (1) to evaluate several staining methods allowing an optimal counting of TCD4+ lymphocyte using Dynabeads method with light microscopy (DLM) and (2) to analyze its usefulness for TCD4+ lymphocyte enumeration in comparison to the Dynabeads method using fluorescence microscopy (DFM) and to the flow cytometry single platform method (FCM).

Study design and study population
This work was a cross-sectional comparative study conducted between January and September 2003 in the Microbiology and Immunology Lab of Centre Muraz, a biomedical and public health research institute based in Bobo-Dioulasso, Burkina Faso. The study population consisted of adults and children with an HIV-seropositive and HIVseronegative status who regularly consulted the Centre Muraz for routine laboratory tests or for HIV cohort entry.
An effort has been made to include asymptomatic and pauci-symptomatic HIV-patients. Thus, the Dynabeads® technique using light microscopy could be tested in a wide range of values of CD4+ lymphocytes. In terms of immunoassays of TCD4+ lymphocytes, the inclusion of pediatric subjects is justified by the fact that children are known to have higher absolute lymphocytes than adults [10] (Diagbouga, Traoré, Ledru and Fumoux, unpublished data) and that it is also useful to test new techniques in populations with very high TCD4+ lymphocyte values.

Single platform flow cytometry technique
Blood samples were collected using EDTA-containing tubes and tested for TCD4+ count within 6 hours by flow cytometry using the BD FACScan (Becton Dickinson, San Jose, US), threecolor mAb (TriTEST™ CD3FITC/CD4PE/CD45 PerCP) reagent and BD Trucount tubes according to standard procedures. Briefly, 50 μl of freshly-obtained whole blood were stained with 20 μl of anti-CD45-perCP, anti-CD3-FITC and anti-CD4-PE mAbs for 15 min at room temperature in the presence of a fixed number of fluorochrome-labeled, polystyrene reference beads. Red blood cells were lysed using a FACS Lysing Solution (BDIS) for 15 min at room temperature. Samples were then, analyzed using the FACSCan and the Multiset and Attractor software (BDIS) for calculating absolute values of TCD4+ cells.
Internal quality control was established and systematized by the Centre Muraz microbiology and immunology lab. It consists essentially of the daily calibration and the optimization of the flow cytometer, the daily and weekly maintenance of the devices. The performance parameters of the cytometer are evaluated: analysis window, linearity, background noise, detector sensitivity. All these parameters were systematically analyzed and recorded in Quickcal (Flow Cytometry Standards Corp).
The Centre Muraz participated in the external quality control organized by the QASI Program an international program for quality assessment and standardization for immunological measures relevant to HIV/AIDS (Dr. John Fahey, Dr. Francis Mandy) and the UK National Quality Assessment Scheme for Leukocyte Immunophenotyping (Dr. David Barnett, Dr. JT Reilly, Mr. V. Granger).

Dynabeads technique (Dynal T4 QUANT kit)
The Dynabeads method (Dynal T4 QUANT kit, Dynal Biotech Oslo, Norway) using a staining with acridine orange and readings under a fluorescent microscope (AXIOLAB) for TCD4+ lymphocyte enumeration was done as described previously [7]. The technique used magnetic beads coated with anti-CD4 monoclonal antibodies (mAbs) to capture and isolate TCD4+ lymphocytes from whole blood. One hundred and twenty-five (125) µl of freshly-obtained EDTA-anti-coagulated blood were added to 350 µl of phosphate buffer saline (PBS). Twenty-five (25) µl of suspended magnetic beads coated with anti-CD14 mAb were then added and the mixture incubated for 10 min at room temperature on a Dynal mechanical rotator in order to deplete blood from monocytes. Magnetic separation of monocytes was performed using a magnetic particle concentrator, as recommended by the manufacturer. One aliquot of 200 µl was taken from the supernatant of monocyte depleted blood and dispensed into 200 µl of PBS. Twenty-five microliters of beads coated with anti-CD4 mAb were then added prior to incubation at room temperature for 10 min on the rotator. The beads were separated using the magnetic particle concentrator and washed twice with PBS.    After addition of 50 µl of lysing solution, cells were stained with 50 µl of a solution of acridine orange and layered on a 10band Malassez Cell where they were integrally enumerated using a fluorescent microscopy. Results were expressed as a number of positive cells per microliter of whole blood.

Selection of dyes allowing TCD4+ lymphocytes enumeration under light microscopy
After a bibliographic search on the potential dyes of the lymphocyte nuclei, we have compiled a list of eleven different methods of staining. The procedure for preparing the dyes is shown in the Annex. We then tested the different coloring methods on a number of 15 patients. The various criteria of this first essential qualitative assessment of the ease of nucleus counting under the light microscope included: the appearance of the bottom of the Malassez cell, the appearance of the nucleus, the contrast, etc. An advanced research on three (3) patients among the 15 subjects allowed us to assign one value out of 10 to each of these dyes. Five technicians gave an appreciation (in a completely blind and independent manner of the appreciation of his colleagues). The averages of the five (5) values are presented in Table 1. These various researches led us to retain for the rest of the work the three (3) best dyes which were the Sternheimer-Malbin solution (STMB), the Turck1 solution, and the Giemsa solution.
For the use of Dynabeads method with light microscopy, the same procedure was used and differences appeared after the beads were separated using the magnetic particle concentrator and washed twice with PBS to isolate TCD4+ cells. Then, the procedure was done in respect with the type of dyes as described in the Annex.

Statistical analysis
Statistical analyses were performed using the SAS-PC software (version 8.2, SAS Institute, Cary, North Carolina, USA). The independent variable analyzed was the absolute TCD4+ count. Flow cytometric testing was compared with each Dynabeads method used. Paired data were compared by linear regression, including slopes, intercepts, and Pearson correlation coefficients. Analysis of agreement between the 2 methods was done by the Bland and Altman method, which compares the difference between paired measurements against the mean of the 2 measurements. P-values were the results of two-sided tests and considered as significant if below 5%. Difference between the two techniques was computed on each sample as the subtraction of the FCM result to the Dynabeads result. The correlation coefficient between the two techniques was calculated and a model of regression then applied to the data.
To assess the accuracy of the Dynal technique using light microscopy, 4 fresh blood samples (200, 400, 500 and 1000 TCD4+/µl in FCM) have been analyzed ten times for each dye against the fluorescence microscopy used with acridine orange.
Positive predictive value (PPV) and negative predictive value (NPV) were computed at the level of 200 and 350 TCD4+

Ethical approval
The study protocol was approved by the Centre Muraz Institutional Review Committee Patient's informed consent was obtained prior to inclusion in the study

Qualitative assessment of the dyes used with light microscopy and their classification
The procedure for the preparation of dyes is presented in the Annex. The average of the 5 values gave by the technicians are presented in Table 1. Score of readings with acridine orange used with fluorescent microscopy was considered as the reference of the readings. It was scored 10/10. The best's score were obtained with Sternheimer-Malbin (7/10), Turck1 (7/10) and Giemsa (7/10).
The nuclei colored with the Sternheimer-Malbin solution appeared dark purple in microscopy under white light. The nuclei colored with Turck1 solution appeared purple under white light and nuclei colored by Giemsa solution appeared purple. The comfort of reading is tiny room compared to a staining with acridine orange and, consequently, a fall of sensitivity and reproducibility is to be feared with a reading in microscopy in white light. Turck1 and Sternheimer-Malbin dyes were preferred by the technicians.
After this qualitative evaluation, the three best's dyes were used for the evaluation of Dynal T4 Quant using light microscopy in comparison with flow cytometry and Dynal T4 Quant using fluorescent microscopy.

Overall correlation between Dynabeads method using light microscopy and flow cytometry single platform method
The correlations between the different techniques were studied for the absolute values of the TCD4+ lymphocyte obtained from 55 children (40 HIV+ and 15 HIV-) and 119 adults (99 HIV+ and 21 HIV-) ( Table 3). The correlation coefficient was better between FCM and DFr. The correlation coefficient was better for adults than for children. However, it was acceptable in all cases. The lowest correlation coefficient was obtained with Giemsa. All the correlation coefficients were higher. The overall Pearson correlation coefficient obtained between Dynabeads methods versus FACScan were 0.975, 0.969, 0.972, and 0.973 respectively for acridine orange (AO), Giemsa, Turck1 and Sternheimer-Malbin (STMB). Very best correlation with the flow cytometry was obtained with the Dynabeads technique using acridine orange with fluorescent microscopy. However, correlation obtained with the other dyes used with light microscopy was also good; the weakest correlation was obtained with Giemsa dye.   Figure 1C shows the line of regression between Dynabeads using Turck1 and FACScan where most of the values lay close to the regression line showing that there was a good agreement between Dynabeads using Turck1 and FACScan methods. The correlation between TCD4+ counts assessed by Dynabeads using Turck1 and flow cytometry was 0.972 (y=0.929x + 37.35; R²=0.947). Figure 1D shows the line of regression between Dynabeads CD4 using Giemsa and FACScan where most of the values lay close to the regression line showing that there was a good agreement between Dynabeads using Giemsa and

Discussion
Due to the increased demand of TCD4+ lymphocyte enumeration in resource-limited settings, this study intended to allow the large decentralization of Dynabeads method in central and peripheral laboratories which have the possibility to possess light microscopy for routine biomedical analyzes. Qualitative evaluation of dyes led us to select the three bests which could be used efficaciously with light microscopy. The specificity of these dyes is that they are used in medical laboratory for routine diagnosis. Therefore, the light microscopy is the main tool in the labs at the district levels and the small laboratories. Giemsa solution is used with May-Grünwald in routine for detection of malaria parasites on stained thick and thin blood films in research settings [11] for hematology parameters analyses, diagnosis of malignant cells [12] and for male fertility exam by the detection of spermatogenic [13]. Sternheimer-Malbin protocol is also used in hematology and for the research of leucocytes in urine [14]. Turck's solution is also used in hematology for white blood cells differentiation [15]. Because these dyes are usually used in routine laboratory activities, they will be useful for TCD4+ enumeration of HIV patients in rural areas and peripheral districts with limited logistics. Acridine orange is used for vital staining in hematology, apoptosis, bone marrow staining [16,17] and in parasitological analyses [18]. Dynabeads method using acridine orange with fluorescent microscopy was evaluated in many settings with good correlation compared to the flow cytometry TCD4+ results [7,[19][20][21]. In Japan, modified protocol of the Dynabeads method was assessed by the used of reducing volume of anti-CD14 and anti-CD4 and a significant correlation of 0.91 with flow cytometry was obtained [21]. In this study, the positive and negative predictive values for a TCD4 count less than 350 cells/µl were good, 97% and 83%. The negative predictive values for a TCD4+ less than 350 cells/µl were also good: 94.74%, 90.53%, 98.95% and 95.79% respectively for acridine orange, Giemsa, Turck1 and Sternheimer-Malbin in our study. Sternheimer-Malbin solution with Dynabeads technique and counted under light microscopy was used in Uganda and the correlation coefficient was good (0.85) [22]. In our study, the use of Sternheimer-Malbin solution shows an excellent correlation coefficient of 0.973, while correlations coefficients with Turck1 and Giemsa solutions were also excellent. In our study, at all levels of TCD4+ count, FCM is the technique that presents the greatest standard coefficient of variation and amplitude. Coefficients of variation obtained with the stains were better than the coefficient of variation obtained with flow cytometry technique. Reproducibility was best with Dynabeads techniques than flow cytometry. The Dynabeads method using acridine orange is the technique with the lowest standard deviation, with the exception of the 200 TCD4+ level. For 350 TCD4+/µl or 200 TCD4+/µl thresholds, the concordance between FCM and the numeration with the Turck1 solution is 98% and 95%, respectively. The concordance between FCM and the numeration with the STMB is 97% and 93%, respectively. The use of Giemsa solution generates lower correlations and higher standard deviations.
The main limitations of Dynabeads technique include its requirements several steps, its chronophage nature and its precision that could be rely on operator expertise. The technique could seem complex for a new user but after long time of using the technique become simple and easy to use. Dynabeads technique requires refrigeration of reagents, a microscope with a 40X objective, a hemocytometer, calibrated pipettes, test tubes, and a manual counter [23].
The use of the selected dyes with light microscopy could substantially reduce the price of TCD4+ lymphocyte enumeration in resource limited settings. With the development of alternative flow cytometry and point of care CD4 technologies by the firms and the efforts of international organisms, non-governmental organizations to decentralize monitoring tools for HIV care in resource limited settings Dynabeads techniques using light microscopy could be in the light of these policies to be implemented in rural areas and peripheral districts. In addition, Dynabeads technology would be implemented in the central laboratories also to face up to the reagents out of stock, breakdown of flow cytometers for the continuing care of HIV patients. A study was demonstrated the feasibility of automation of Dynal T4 Quant in hematology automate as a Point of Care CD4 with the delivering in the same time of hematology parameters and TCD4+ counting [4]. To improve on throughput using the manual Dynal T4 Quant method, the Sysmex pocH-100i hematological analyzer was used to count stained nuclei and showed 93% concordance with FCM [4]. The capacity to use the Dynal T4 Quant with hematology automate by the using K-Xpert software was successfully validated [4]. In April 2002, WHO recommended that when TCD4 cell count is not available or is not affordable to be obtained for affected individuals, a total lymphocyte count of less than 1000-1200 lymphocytes/µl in individuals with stage 2 or stage 3 diseases to be used as an indication to initiate antiretroviral therapy [24]. To the best of our knowledge, studies comparing the utility of absolute lymphocyte counts (ALC) as a surrogate for CD4 count in monitoring HIV infected individuals have shown contradictory results. Indeed, many published studies showed low sensitivity of absolute lymphocyte count less than 1200/ul for predicting CD4 cell count less than 200 cells/ul [25][26][27]. Conversely, studies from Karanth et al. and Obirikorang et al. [28,29] shown higher sensitivity and specificity for total lymphocyte count cut-off 1200 cells/µl to predict CD4 cell count less than 200/µl. The differences could be due to different ethnic, racial, epidemiological and socioeconomic factors.

Conclusion
Dynal T4 Quant technique is an affordable and accurate method for TCD4+ counting and it is an effective alternative TCD4+ method for resource-limited settings, despite the development of many news alternatives methods for TCD4+ screening. Following this study, it has been demonstrated that reliable TCD4+ numeration can be obtained with the Dynal T4 Quant Kit, in light microscopy with the Sternheimer-Malbin Turck1 and Giemsa solutions. These results could contribute to further reduce the cost of TCD4+ lymphocytes counting and make it more widely available in peripheral laboratories and even in central laboratories that face problems with maintenance and stock-out of reagents for flow cytometers.

Authors Contributions
SD and LC designed the study, HH, AO recruited the study population, SD, EB, AK, DZ, AO performed the experiments, SD, LC and DZ analyzed the data SD, LC, DZ, EB wrote the manuscript which was approved by all the other co-authors.