Background: Practical HIV diagnostics are urgently needed in resource-limited settings. Accurate CD4+ T cell count is a critical clinical indicator for HIV disease staging and treatment monitoring. To address the limitations of flow cytometry, we recently developed a simple CD4 counting device based on immunoaffinity cell isolation in a single microfluidic channel functionalized with anti-CD4 antibodies. By isolating target cells from 10 uL of unprocessed whole blood, we could distinguish clinically relevant thresholds of 200, 350, and 500 cells/uL with high sensitivity (>85%) and specificity (>94%). However, this device is incapable of exact CD4 counting, especially for CD4 counts <200 cells/uL, due to monocyte contamination. To improve its accuracy, we added a monocyte depletion chamber upstream of the CD4 capture channel. Performance of the cascaded device is presented in this work.

Methods: We use microfluidic cell affinity chromatography to first deplete monocytes and then specifically isolate CD4+ T cells with high efficiency directly from 10 uL of unprocessed whole blood. CD4 counts are obtained afterwards under an optical microscope in a rapid, simple and label-free fashion. We compared CD4 counts from dual-chamber (a monocyte depletion chamber cascaded with a CD4 T cell capture chamber) and single-channel (with anti-CD4 alone) devices and conventional flow cytometry among HIV+ subjects over a wide range of absolute CD4 counts. We analyzed data via Bland-Altman comparison.

Results: Compared to the single channel devices, CD4 counts determined in the cascaded devices better correlate with the measurements by flow cytometry (R2=0.90 in dual chamber vs. R2=0.77 in single chamber devices). A Bland-Altman comparison demonstrates a bias of +69 cells/uL for the single chamber device, with limits of agreements of -66 to +204 cells/uL. Double chamber devices show a much closer agreement with flow cytometry, with a bias of +12 cells/uL and limits of agreements of -55 to +79 cells/uL respectively.

Conclusions: The dual chamber microfluidic device with a monocyte depletion module followed by a CD4+ T cell isolation channel greatly improves CD4 counting accuracy compared to the single channel device. Our technique establishes the ability to obtain CD4 counts directly from unprocessed whole blood with accuracy across the dynamic range of CD4 cell counts, and is suitable for simple, rapid and affordable HIV management in point-of-care and resource-limited settings.