Radiology Devices; Reclassification of Digital Breast Tomosynthesis System | Proposed Rule (FR Doc. 2026-16209)
91 FR 51406 · Docket No. FDA-2026-N-7630 · Comments Due October 9, 2026
Last Reviewed: August 21, 2026
- Publisher
- U.S. Food and Drug Administration / Federal Register
- Author
- Grace R. Graham, Deputy Commissioner for Policy, Legislation, and International Affairs, FDA
- Edition
- Proposed Rule — 91 FR 51406, 2026
- Pages
- 11 pages
- Added to library
- Added August 21, 2026
Overview
This 11-page Federal Register proposed rule (Vol. 91, No. 152, August 10, 2026) proposes to reclassify DBT systems (prescription devices intended to generate digital cross-sectional x-ray images of the breast for screening and diagnosis of breast cancer) from class III (premarket approval) into class II (special controls), subject to 510(k) premarket notification. Product code OTE. DBT systems are postamendments class III devices automatically classified under §513(f)(1) of the FD&C Act. FDA's proposed reclassification is supported by: (1) PMA data from 4 original PMAs (P080003, P130020, P140011, P160031) and 26 supplements approved since 2011; (2) a Radiological Devices Panel consensus (September 24, 2010) that benefits outweigh risks; (3) a 42-study meta-analysis covering 2,606,269 patients demonstrating CDR and iCDR improvements; (4) the absence of Class I or III recalls and absence of major safety signals in 968 MDR reports; (5) 94% adoption at certified MQSA facilities (9,107 facilities as of July 21, 2026); and (6) significant advances in evaluation methods and consensus standards since initial approval. The proposed reclassification would create new 21 CFR §892.1717 with eight specific special control labeling requirements and four device-level testing requirements. FDA is not proposing to exempt DBT systems from 510(k) requirements. If finalized, manufacturers would submit 510(k)s rather than PMAs for new DBT systems. Proposed effective date: 30 days after publication of a final order. Signed by Grace R. Graham, Deputy Commissioner for Policy, Legislation, and International Affairs.
Key Contents
-Regulatory action: Proposed amendment; proposed order; request for comments
-Published: Federal Register Vol. 91, No. 152 (August 10, 2026), pages 51406–51416
-Docket No. FDA-2026-N-7630 | FR Doc. 2026-16209
-Comment deadline: October 9, 2026 (electronic: regulations.gov; 11:59 p.m. Eastern Time)
-Proposed effective date: 30 days after date of publication of final order
-Signed: Grace R. Graham, Deputy Commissioner for Policy, Legislation, and International Affairs
-Contact: Yanna Kang, CDRH, FDA, 10903 New Hampshire Ave., Bldg. 66, Rm. 3544, Silver Spring MD 20993-0002 | 301-796-6704 | Yanna.Kang@fda.hhs.gov
-Proposed change: DBT systems (product code OTE), class III (PMA) → class II (special controls) + 510(k)
-Proposed new CFR: 21 CFR §892.1717 Digital breast tomosynthesis system
-Not proposing 510(k) exemption: premarket notification required for all DBT systems
-MQSA intersection: DBT is a mammographic modality under MQSA (Pub. L. 102-539). Facilities performing mammography using DBT systems are subject to all MQSA requirements regardless of device classification. Physicians performing DBT clinical image evaluation must qualify under 21 CFR 900.12 to interpret mammography exams.
-Regulatory history:
• February 11, 2011: FDA approved first DBT system, Selenia Dimensions 3D (PMA P080003, Hologic Inc.)
• Radiological Devices Panel consensus (September 24, 2010): benefits outweigh risks
• 4 original PMAs approved: P080003, P130020, P140011, P160031
• 26 PMA supplements approved including 1 panel-track supplement (P080003/S001)
• As of July 21, 2026: 9,107 MQSA-certified facilities; 94% have DBT units; 95% of accredited DBT units also accredited as 2D units
• MDR review (as of July 21, 2026): 968 reports, majority no known clinical impact; 0 Class I recalls, 5 Class II recalls, 0 Class III recalls
-Clinical evidence (Section VI, meta-analysis, Ref. 8, 42 studies, 2,606,269 patients, 13,003 breast cancer cases):
• CDR: DBT + FFDM = 6.36/1000; DBT + s2D = 7.40/1000, significantly higher than FFDM alone (4.68/1000)
• iCDR: DBT + FFDM = 7.40/1000; DBT + s2D = 5.68/1000, significantly higher than FFDM alone (3.42/1000)
• DBT alone (5.20/1000 CDR): not significantly different from FFDM alone
• PPV1: highest in combined DBT + s2D (16.0%) and combined DBT + FFDM (10.0%) vs. FFDM alone (7.0%)
• Recall rate: lowest in combined DBT + s2D (42.3/1000) vs. FFDM alone (78.8/1000)
• Conclusion: DBT + s2D preferred over DBT + FFDM; maintains diagnostic performance while reducing radiation dose and costs
-Proposed 21 CFR §892.1717 special controls:
(b)(1) Performance testing data including:
(i) Imaging characteristics across representative settings: bench testing AND objective task-based diagnostic accuracy assessment (reader study with human subjects, observer study with structured physical phantoms, or in silico testing)
(ii) Detailed description of system hardware and software
(b)(2) Clinical image evaluation data demonstrating sufficient diagnostic quality for screening and diagnosis
(b)(3) Software verification, validation, and hazard analysis
(b)(4) Electrical safety, mechanical safety, thermal safety, and EMC data
(b)(5) Patient-contacting components demonstrated biocompatible
(b)(6) Labeling must include:
(i) Detailed device description: principles of operation, system hardware/software, acquisition workstation, image receptor, technique factors, AEC, image processing/reconstruction, patient/equipment supports, component parts, accessories
(ii) Detailed description of device outputs
(iii) User qualifications and/or clinical training needed for safe use
(iv) Detailed summary of objective task-based diagnostic accuracy assessment: test methods, dataset characteristics, results, sub-analyses stratified by relevant confounders
(v) Detailed summary of bench testing results with graphs or tables
(vi) Detailed summary of clinical image evaluation performed with device
(vii) Description of QC testing including detailed procedures and frequency
(viii) Validated methods and instructions for cleaning and disinfection of reusable patient-contacting components
-10 identified risks to health and special control mitigations (Table 1):
(1) Corrupted or non-diagnostic images → performance testing, software verification/validation/hazard analysis, clinical image evaluation, labeling
(2) Failure to interpret images correctly (false negative/positive) → performance testing, clinical image evaluation, labeling
(3) Inadequate breast coverage → performance testing, clinical image evaluation, labeling
(4) Inappropriate breast compression → performance testing, software verification/validation/hazard analysis, clinical image evaluation, labeling
(5) Device failure or malfunction → performance testing, software verification/validation/hazard analysis, electrical safety/EMC testing, labeling
(6) Use error/improper use → software verification/validation/hazard analysis, labeling
(7) Excessive x-ray radiation exposure → performance testing, software verification/validation/hazard analysis, electrical safety/EMC testing, labeling
(8) Interference with other devices → electrical safety/EMC testing
(9) Adverse tissue reaction → biocompatibility evaluation
(10) Infection → labeling (validated cleaning/disinfection instructions)
-Bench tests mentioned for special controls: spatial resolution, noise analysis, signal-to-noise ratio transfer, detective quantum efficiency (DQE), detector lag, AEC performance, geometric distortion, missed tissue at top/bottom and chest wall side in reconstructed DBT volume, testing of alignment and collimation, radiation dosimetry
-Objective task-based assessment methods: reader studies (human subjects), observer studies (structured physical phantoms), in silico testing (IST / Virtual Clinical Trials, VCT)
-Not applicable to FFDM systems (21 CFR 892.1715, class II) or dedicated breast CT systems (OLQ, remains class III)
-PCCPs (Predetermined Change Control Plans) available to manufacturers to implement future modifications without new 510(k) per each significant change
How This Applies to Your Practice
This proposed rule is the most significant DBT regulatory development since the first system was approved in 2011. For mammography facilities, the outcome affects every DBT system manufacturer's pathway to market; class II + 510(k) is substantially less burdensome than PMA, which should accelerate innovation and potentially increase the number of FDA-cleared DBT systems available. However, the reclassification does NOT change any MQSA obligations. Facilities operating DBT equipment will continue to be inspected and accredited under MQSA regardless of whether their DBT system received a PMA or a 510(k). The new 21 CFR §892.1717 special controls (particularly the labeling requirements for QC testing procedures and cleaning/disinfection instructions) are relevant to facility compliance staff who should verify that equipment labeling meets these requirements. The comment period closes October 9, 2026; facilities and accreditation bodies with views on the proposed special controls may submit comments at regulations.gov docket FDA-2026-N-7630.
Practitioner Implications
Active and immediate implications: (1) OPEN COMMENT PERIOD: comments on the proposed rule are due October 9, 2026 at regulations.gov (docket FDA-2026-N-7630); (2) This is a PROPOSED RULE only; current class III (PMA) requirements for new DBT systems remain in force until a final order is published; (3) Proposed effective date of a final order is 30 days after Federal Register publication; (4) If finalized, new DBT manufacturers submit 510(k) not PMA; existing PMA holders may use PCCPs for future device modifications; (5) MQSA obligations for facilities using DBT (accreditation, inspection, quality standards, personnel qualifications) are unchanged by this reclassification and remain in full force; (6) New §892.1717(b)(6) labeling requirements for QC testing procedures and frequency are relevant to medical physicists and QC programs at facilities operating DBT equipment.
Didn't find what you were looking for?
Our team has supported breast imaging centers since 1995. If it's not here, we probably know the answer.