Diabeloop’s DBLG2 and a Closed Loop With the Twiist Delivery System for T1D

On December 18, 2025, Diabeloop received 510(k) clearance from the US Food and Drug Administration (FDA) for its DBLG2 algorithm, which is designed to automate and personalize insulin delivery for patients with type 1 diabetes. The app analyzes glucose data in real-time, altering insulin dosage delivery to maintain blood sugar levels within a target range.1

Following this announcement, Diabeloop partnered with Sequel Med Tech to pair the algorithm with Sequel’s automated insulin delivery system Twiist. This collaboration moved Twiist closer to a closed-loop solution, delivering personalized and proven clinical outcomes while maintaining ease of use. Diabetes Dialogue cohosts Diana Isaacs, PharmD, and Natalie Bellini, DNP, sat down to discuss this recent advancement and its indications for insulin monitoring moving forward.2

Check out the full episode on the 2025-2030 Dietary Guidelines for Americans here.

Isaacs opens the discussion of Diabeloop’s newly FDA-cleared insulin delivery algorithm and its implications for the evolving landscape of automated insulin delivery. Although initially hesitant to cover the clearance, given the frequent disconnect between regulatory approval and real-world availability, she and Bellini revisit the topic after the announcement of a confirmed partnership with the Twiist pump. With anticipated readiness by late 2026 and a projected commercial launch in 2027, the discussion shifts from speculation to a more concrete assessment of how this system may enter clinical practice.

A central focus of the conversation is what differentiates Diabeloop from existing hybrid closed-loop systems. The hosts emphasize that Diabeloop is designed to be a truly fully closed-loop algorithm, requiring no meal announcements or manual boluses, though optional meal inputs may further increase time in range. This positions Diabeloop as a step beyond current systems such as Tidepool Loop and even Beta Bionics’ iLet, which, while highly automated, still rely on some degree of user input. The panel frames this as a meaningful advance toward reducing the daily cognitive burden of diabetes management, particularly for individuals who struggle with consistent bolusing or pre-meal timing.

The hosts discuss the potential clinical and behavioral impact of such a hands-off approach, noting how frequently clinicians counsel patients on bolus timing and meal announcements to optimize glycemic outcomes. They suggest that Diabeloop’s design may help more patients reach glycemic targets with fewer interactions, potentially improving adherence and quality of life. At the same time, they consider whether patients’ needs may evolve over time, with some initially preferring minimal interaction and later seeking greater control and customization.

This leads to a broader discussion of interoperability and choice. The group explores whether future pump platforms might allow patients to switch between algorithms as easily as they currently change continuous glucose monitors, while acknowledging the added safety and regulatory complexities inherent to insulin delivery. They highlight that Twiist’s willingness to support multiple algorithms represents a notable departure from the industry norm, where most pump manufacturers remain tightly coupled to proprietary software.

Finally, the hosts reflect on the absence of CamAPS from Twiist’s current lineup, noting that it remains the only FDA-cleared algorithm approved for pregnancy in the United States. While expressing disappointment, they recognize that Diabeloop offers a distinctly different use case compared with more customizable algorithms and may attract a broader audience. The episode concludes with cautious optimism, positioning Diabeloop as a bold and potentially transformative step toward more accessible, low-burden automation, while underscoring the importance of safety, flexibility, and patient-centered choice as closed-loop technologies continue to advance.

Editor’s Note: Isaacs reports disclosures with Dexcom, Abbott, Lilly, Novo Nordisk, Medtronic, Insulet, and others. Bellini reports disclosures with Abbott Diabetes Care, MannKind, Povention Bio, and others.

References

1. Diabeloop. Diabeloop receives FDA 510(k) clearance for DBLG2, an Automated Insulin Delivery Algorithm in an App: a Strategic Turning Point. January 8, 2026. Accessed January 20, 2026. https://www.diabeloop.com/media-press/press-releases/diabeloop-fda-510k-clearance

2. Diabeloop. Diabeloop and Sequel Med Tech Enter Partnership to Integrate FDA-Cleared DBLG2 Into the Twiist Automated Insulin Delivery System. Presswire. January 14, 2026. Accessed January 20, 2026. https://presswire.com/release/diabeloop-and-sequel-med-tech-enter-partnership-to-integrate-fda-cleared-dblg2-into-the-twiist-automated-insulin-delivery-system-2/