Clinical Evidence Supporting US Food and Drug Administration Clearance of Novel Therapeutic Devices via the De Novo Pathway Between 2011 and 2019

Abstract

This cross-sectional study characterizes the premarket clinical evidence supporting clearance of moderate-risk novel therapeutic medical devices by the US Food and Drug Administration.

In 1997, US Congress established the US Food and Drug Administration (FDA) De Novo premarket review pathway for novel low-risk and moderate-risk medical devices.¹ In recent years, the FDA and medical device manufacturers have increasingly used the De Novo pathway; the FDA cleared 65 devices via this pathway between 1997 and 2012 and an additional 187 between 2013 and 2019.^2,3,4 Devices cleared through the De Novo pathway represent novel technologies, and approval may serve as the basis for subsequent clearance of devices through the 510(k) pathway.⁵ The 510(k) pathway requires manufacturers to demonstrate that new devices are “substantially” equivalent to previously cleared devices, typically without clinical studies demonstrating safety or effectiveness. This study characterizes the premarket clinical evidence supporting clearance of moderate-risk therapeutic devices cleared through the De Novo pathway and postmarket experience, including FDA-required postmarket studies, recalls, and subsequent 510(k) clearances.

Methods

In conducting this cross-sectional study, we identified moderate-risk therapeutic devices cleared through the De Novo pathway by the FDA between January 1, 2011, and December 31, 2019, using the FDA De Novo database and FDA-designated product codes.⁴ There were 65 moderate-risk therapeutic devices that met these criteria between 2011 and 2019 (eTable in the Supplement). Using previously described methods,⁶ we reviewed publicly available FDA decision summary documents to identify and characterize all pivotal clinical studies—which generally serve as the primary basis for device clearance—and their primary effectiveness end points.

We searched the FDA’s 522 Postmarket Surveillance Studies and FDA Medical Device Recall databases for each device to identify FDA-required postmarket studies and device recalls, respectively, as of March 22, 2020. To identify subsequently cleared models and competitor products for each device, we searched the FDA 510(k) database on March 13, 2020, using product codes and recorded (as applicable) the dates of (1) the first 510(k) clearance by the device manufacturer after De Novo clearance and (2) clearance by a competing manufacturer.

All information was summarized using descriptive statistics using Microsoft Excel, version 16.40. This study analyzed information made publicly available through FDA databases and therefore did not require institutional review board approval.

Results

Of the 65 moderate-risk therapeutic devices cleared by the FDA via the De Novo pathway between 2011 and 2019, 63 (97%) had publicly available decision summaries. Of these 63 devices, 10 (16%) were implantable and none was life sustaining (Table 1). For 51 devices (81%), FDA clearance was based on 54 pivotal clinical studies; 12 (19%) were not evaluated through pivotal studies. The median number of patients enrolled in pivotal studies was 112.5 (interquartile range [IQR], 73.5-187.0). Nearly half (23; 43%) were blinded, and a majority used randomization (57%) and an active (26%) or sham (35%) comparator. Of 60 studies measuring premarket effectiveness, 36 (60%) had clinical outcomes as primary effectiveness end points (Table 2). However, 17 studies (31%) failed to meet at least 1 primary effectiveness end point.

Table 1. Characteristics of Novel Moderate-Risk Therapeutic Devices Cleared by the US Food and Drug Administration (FDA) Through the De Novo Pathway Between 2011 and 2019.

Characteristic	No. (%) (N = 63)	Examples (if applicable)
Device characteristic
Implantable
Yes	10 (16)	Stents such as a biliary stent system for treatment of benign strictures
No	53 (84)	Weight loss treatments, including software and transient space occupiers
Life sustaining
Yes	0
No	63 (100)
Clinical use
Neurology	18 (29)	Nerve stimulators used for treatment of tremors, headaches, and migraines
Gastroenterology and urology	13 (21)	Treatments for benign prostatic hyperplasia including ultrasound tissue ablation and embolic devices
General and plastic surgery	10 (16)	Hemostatic devices
Otolaryngology	5 (8)	Hearing aids
Ophthalmology	4 (6)	Nerve stimulators to increase tear production
Other	13 (21)	Ventilation devices to treat patients with sleep disorders, carbon monoxide poisoning, and hypoxemia
Postmarket experience
FDA-required postmarket study
Yes	1 (2)	Powered exoskeleton to assist ambulation
No	62 (98)	Transcatheter aortic valve replacement embolic protection system that failed primary end point of reduction in new lesion volume within protected brain territories
Highest recall class
Class I	0
Class II	6 (10)	Endoscopic hemostat recalled due to malfunction that could cause delay in hemostasis
Class III	0
No recall	57 (90)
Used as predicate for 510(k) device
Yes	32 (51)	Catheter for balloon aortic valvuloplasty that served as a predicate for 15 510(k)-cleared devices
No	31 (49)

Table 2. Characteristics of Pivotal Studies and Their Primary Effectiveness End Points Supporting US Food and Drug Administration (FDA) Clearance of Therapeutic Medical Devices Through the De Novo Pathway Between 2011 and 2019.

Characteristic	Pivotal studies, No. (%)
Total No.	54
Total patients, median (IQR)	112.5 (73.5-187.0)
Patients treated, median (IQR)	89.0 (49.0-118.5)
Study arms
Multiarmed	33 (61)
Single armed	21 (39)
Randomized	31 (57)
Blinding
Blinded	23 (43)
Open label	23 (43)
Not specified	8 (15)
Comparator type
Active	14 (26)
Sham	19 (35)
Historical	2 (4)
Pre–post comparison	8 (15)
None	11 (20)
Enrollment centers
Multicenter	41 (76)
Single center	3 (6)
Not specified	10 (19)
Pivotal premarket effectiveness end points
Total No.	60a
End point type
Clinical outcome	36 (60)
Clinical scale	17 (28)
Surrogate marker	7 (12)
Duration of longest primary end point follow-up, median (IQR), mo
Overall	1 (0-6)
Nonimplantable	1 (0-3)
Implantable	10 (4-12)
Primary end points met
Yes	40 (67)
No	18 (30)b
End point success criteria unspecified	2 (3)

Abbreviation: IQR, interquartile range.

^a Does not include 5 pivotal studies supporting 5 devices without a primary effectiveness end point.

^b One pivotal study failed to meet 2 primary effectiveness end points.

Only 1 device, a powered exoskeleton to assist ambulation, was subject to an FDA-required postmarket study. None underwent class I (high-risk) recalls. Of the 63 devices, 32 (51%) served as the basis for devices subsequently cleared through the 510(k) process (median days to first clearance: 202 [IQR, 159.0-302.0]), including 16 (25%) serving as the basis for competitor devices (median days to first clearance: 364 [IQR, 184.5-632.0]).

Discussion

Between 2011 and 2019, the FDA cleared most novel moderate-risk therapeutic devices via the De Novo pathway based on pivotal clinical studies. However, approximately one-fifth of devices were cleared without pivotal clinical studies, and approximately one-third were based on pivotal studies that failed to meet at least 1 primary effectiveness end point. Nevertheless, the FDA rarely required postmarket studies, and devices cleared through the De Novo pathway often served as the basis for new models and competitor products subsequently cleared via the 510(k) process.

Limitations to the present study include that it was restricted to therapeutic devices and is not generalizable to diagnostic devices. This study did not examine non–FDA-required postmarket studies and may therefore underestimate the strength of evidence supporting device safety and effectiveness.

To ensure that better evidence is generated to guide clinical decision-making, the FDA could require that devices cleared through the De Novo pathway meet prespecified effectiveness end points for clearance and that postmarket studies include larger patient populations. Such studies could be particularly informative for devices cleared without pivotal studies and those that failed to meet effectiveness end points.

Supplement 1.

eTable. Moderate-Risk Therapeutic Devices Cleared Through the FDA’s De Novo Pathway Between 2011 and 2019