The Evolution of Diabetes Care
Vinay Sharma
Principal - Intellectual Asset Management
Iota Analytics
Ishani Sharma
Manager - Intellectual Asset Management
Iota Analytics
Diabetes care is undergoing a fundamental shift — from reactive treatment to predictive, autonomous management. And the IP landscape is moving just as fast.
Phase 1: Reactive
For decades, diabetes management meant fingerstick blood glucose meters, episodic HbA1c lab tests, and manual insulin injections. The patient was always responding reactively — testing, monitoring, calculating doses, administering insulin, and making treatment decisions based on intermittent data rather than real-time feedback.
Phase 2: Proactive
Then came CGM. Continuous Glucose Monitoring — led by Medtronic’s Minimed, Dexcom G7, and Abbott's FreeStyle Libre, which shifted the paradigm from "what happened" to "what is happening right now." Real-time glucose data, trend arrows, and alerts gave patients and clinicians the ability to act before a crisis, not after one.
The CGM market alone is projected to reach $31.4 billion by 2031 (15.4% CAGR). And with Dexcom's Stelo — the first OTC CGM at $89, now available on Amazon and integrated with OURA ring — glucose monitoring is expanding beyond diabetes into the consumer wellness and prediabetes space entirely.
Phase 3: Predictive (this is where it gets interesting)
The predictive era is not just about ketones, though ketones remain an important part of the picture. It's a convergence of multiple technologies that, together, aim to anticipate metabolic events before they happen:
𝗠𝘂𝗹𝘁𝗶-𝗕𝗶𝗼𝗺𝗮𝗿𝗸𝗲𝗿 𝗦𝗲𝗻𝘀𝗶𝗻𝗴: Abbott has advanced dual glucose-ketone sensing, including Libre Duo systems. Sava Health's wearable patch is being developed to track glucose, ketones, lactate, cortisol, and histamine from a single device. Emerging sweat-based biosensors can detect electrolytes, metabolites, and hormones non-invasively — the wearable sweat sensor market alone is projected to hit $13.47B by 2034.
Why does this matter? Because glucose alone measures the present. Ketones warn of DKA risk. Lactate signals metabolic stress. Cortisol reflects physiological strain. Together, they paint a predictive metabolic picture that glucose alone could never achieve.
𝗔𝗜-𝗗𝗿𝗶𝘃𝗲𝗻 𝗣𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝘃𝗲 𝗔𝗻𝗮𝗹𝘆𝘁𝗶𝗰𝘀: Transformer-based and LSTM deep learning models are now predicting glucose trajectories 30-60 minutes ahead. Digital twin technology is creating virtual patient models that simulate disease progression and optimize treatment in real time. DreaMed Diabetes is deploying AI-driven decision support. Welldoc holds multiple patents for its AI-powered platform and has FDA clearances.
Reinforcement learning algorithms are even personalizing insulin dosing recommendations based on individual HbA1c, BMI, and activity patterns — achieving an 88% match with physician-prescribed doses.
𝗚𝗹𝘂𝗰𝗼𝘀𝗲-𝗥𝗲𝘀𝗽𝗼𝗻𝘀𝗶𝘃𝗲 "𝗦𝗺𝗮𝗿𝘁" 𝗜𝗻𝘀𝘂𝗹𝗶𝗻: Novo Nordisk's NNC2215 — a glucose-responsive insulin conjugate — activates only when blood sugar rises and deactivates when it drops. This is insulin that ‘thinks for itself’ at the molecular level. Earlier candidates such as Merck’s MK-2640 illustrate both the promise and the technical challenges of glucose-responsive insulin development, while newer candidates continue to advance across the field.
𝗔𝗜 𝗳𝗼𝗿 𝗖𝗼𝗺𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗣𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝗼𝗻: Autonomous AI systems (Eyenuk's EyeArt, Digital Diagnostics' LumineticsCore) are screening for diabetic retinopathy with 95.5% sensitivity — predicting complications before symptoms appear.
The Closed-Loop Convergence: Intelligent AID
Today's hybrid closed-loop systems — Medtronic's MiniMed 780G, Tandem's Control-IQ+, Insulet's Omnipod 5 — already auto-adjust basal insulin based on CGM data. But the next frontier is fully autonomous:
Beta Bionics' iLet Bionic Pancreas targets bi-hormonal delivery (insulin + glucagon), with a feasibility trial expected in 2025-26
Insulet reported that EVOLUTION 2 feasibility results showed 68% time-in-range with no boluses, and subsequently initiated the EVOLVE pivotal study for fully closed-loop AID in adults with Type 2 diabetes
A novel "DuoLoop" dual closed-loop system pairs CGM-driven control with glucose-responsive insulin for layered safety
PharmaSens' niia signature combines an insulin patch pump and CGM in a single wearable device
The IP Landscape
From a patent strategy perspective, this is one of the most IP-dense medtech spaces:
Abbott: has key patents in ketone/multi-analyte monitoring; many directly claim DKA-related analyte monitoring
Dexcom: has multiple patents spanning CGM to integrated medication delivery, plus the 10-year cross-license settlement with Abbott
Bi-hormonal pumps: Beta Bionics is the dominant assignee with multiple patents filed
Interoperability standards (iCGM, ACE pumps) are creating SEP-like dynamics in AID, similar to how FRAND licensing shaped the telecom industry
AI/digital health: Welldoc alone holds multiple patents; the AI-in-diabetes IP space is accelerating
The Way Forward
The future of diabetes care isn't a single technology — it's the convergence of five pillars:
Advanced Multi-biomarker wearables (glucose + ketone + lactate + cortisol + other biomarkers)
AI/digital twin predictive engines that learn from each patient
Fully autonomous closed-loop delivery/ AID systems
Glucose-responsive "smart" insulin at the molecular level
Regenerative medicine: Moving toward cell-based therapies (Sernova's iPSC beta cell implants, Bio-hybrid organs) as the long-horizon play toward a functional cure
All-in-One Device Architecture, (PharmaSens & SiBionics niia signature)
The companies building the deepest IP moats across these five pillars today will define diabetes care for the next decade.
