Abstract

Background

While 90% of cancer deaths are associated with metastasis, it is imperative to monitor early-stage cancer patients for the presence of systemic disease to improve overall survival (OS) and progression-free survival (PFS). Despite complete remission, up to 25–50% of colorectal cancer (CRC) stage II–III and early breast cancer cases are known to relapse. Furthermore, the existence of microtumors often remains undetected by radio-imaging tools due to their limited detection thresholds. Following curative-intent therapies, minimal residual cellular disease (MRCD) may persist and is often represented by circulating tumour cells (CTCs). These cells are known for their ability to extravasate and invade distant sites from the primary tumor. They can also evade immune surveillance. Therefore, there is a need to design safer extracorporeal devices for the capture and depletion of CTCs, particularly those overexpressing PD-L1. In this study, we designed an automated device to capture and remove CTCs from whole blood.

Methods

We developed an automated microprocessor-operated fluidic device, OncoMetastat, equipped with cartridges for blood and reagent tubes, along with a 3D-printed biocompatible spiral channel. The controller unit powers peristaltic pumps that circulate blood through the spiral channel (96 mm diameter × 6 mm height). The system incorporates 2 mm glass beads conjugated with antibodies and transferrin. Additionally, four vibrators provide micro-stirring to enhance CTC capture from 5–10 mL of patient blood samples (n = 54). White blood cell (WBC) count, hemolysis, and protein binding were measured. The beads were scanned for CTCs using markers CK18⁺, DAPI⁺, and CD45⁻ through an automated imaging system and compared with the OncoDiscover CTC enumeration platform approved by CDSCO India. We analyzed true positives, false negatives, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy.

Results

Retrospectively, blood samples from 54 pan-cancer patients—including breast, colorectal, prostate, and lung cancer—were analyzed to capture and deplete CTCs. The OncoMetastat platform demonstrated a capture efficiency of over 90% when compared with the OncoDiscover platform. Automated scanning achieved 100% efficiency in CTC imaging compared with manual imaging. Leukocyte adhesion was low with anti-EpCAM and transferrin-coated glass beads (2 ± 1 WBCs per sample, n = 54). WBC counts showed cancer-type-specific trends (mean WBC count/mL: 4.9 × 10⁶ for breast cancer, 3.9 × 10⁶ for rectal cancer, and 3.5 × 10⁶ for prostate cancer), representing a 40% decrease compared with healthy controls (mean 6.9 × 10⁶ WBCs/mL). Clinically insignificant hemolysis (<1%) and minimal protein binding (~1.5%) were observed. Vibration-assisted operation enhanced CTC sequestration, achieving more than 90% cell capture efficiency. The platform demonstrated sensitivity of 94.4%, specificity of 92.9%, PPV of 94.4%, NPV of 92.9%, and overall accuracy of 93.8% for CTC capture.

Conclusion

This study demonstrates efficient and specific depletion of CTCs using the automated device. The platform shows potential as an extracorporeal system capable of removing CTCs from whole blood, thereby offering a promising strategy to enhance cancer therapy outcomes.