Two-Stage Prepectoral Breast Reconstruction with and without Acellular Dermal Matrix: A Cluster-Randomized Noninferiority Trial

Authors: Nelson JA, Shammas RL, Boe LA, Gönen M, Poulton RT, Stern CS, Rochlin DH, McCarthy CM, Allen RJ Jr, Coriddi M, Disa JJ, Mehrara BJ, Matros E

Affiliation: Memorial Sloan Kettering Cancer Center, New York, NY

Journal: Plastic and Reconstructive Surgery, March 2026

PMID: 41811168

Key takeaways

• ADM and no-ADM prepectoral tissue-expander reconstruction had nearly identical 90-day major complication rates: 13.2% vs 13.1%.

• ADM met the prespecified noninferiority criterion; the one-sided 95% upper confidence limit for excess major complications was 6.2%, below the 7% margin.

• Infection, reoperation, explantation, BREAST-Q physical well-being, and pain were not significantly different; seroma trended lower with ADM (20.0% vs 27.4%, p=0.056).

• The trial establishes short-term safety of ADM, not a demonstrated benefit or necessity; long-term contracture, rippling, malposition, aesthetics, and cost were not resolved.

Background

Acellular dermal matrix (ADM) is commonly used in prepectoral implant reconstruction for support and implant coverage, but it adds cost and has been linked to infection and seroma. High-level evidence directly comparing ADM with no ADM in the prepectoral plane has been limited.

Objective

Determine whether immediate two-stage prepectoral reconstruction with ADM is noninferior to reconstruction without ADM for major complications within 90 days of tissue-expander placement.

Methods

• Design: Prospective, single-center, randomized noninferiority trial at Memorial Sloan Kettering Cancer Center. The first 35 patients were individually randomized; after a November 2022 amendment, 333 were cluster-randomized by calendar month. Blinding was not feasible. Level of evidence was not stated in the manuscript.

• Population: 368 women, 623 breasts; median age 44 years, BMI 24.4 kg/m². Mastectomies: 69.3% bilateral, 91.0% therapeutic. Eligible patients were age 21–75 with adequate mastectomy-flap perfusion and a viable prepectoral plane. Current smoking, non-English fluency, prior sternotomy, and direct-to-implant reconstruction were excluded.

• Intervention: Smooth prepectoral tissue expander with anterior perforated ADM versus the same expander secured by tabs without ADM. Skin-flap perfusion was assessed with SPY.

• Primary endpoint: Breast-level occurrence of ≥1 major complication within 90 days—infection, unplanned reoperation under general anesthesia, or expander explantation. Secondary endpoints included clinically significant seroma, BREAST-Q Physical Well-being of the Chest, and postoperative pain.

• Statistics: Prespecified 7% noninferiority margin; planned 320 patients (~480 breasts) for ≥85% power at one-sided α=0.025. Primary analysis was intention-to-treat with adjustment for clustering within bilateral patients; as-treated and per-protocol sensitivity analyses were also performed.

Results

• Primary outcome- Major complications within 90 days: 39/295 breasts (13.2%) with ADM vs 43/328 (13.1%) without ADM; absolute difference 0.11%. The one-sided 95% upper limit was 6.2%, meeting noninferiority.

• Sensitivity analyses were concordant: as-treated 12.7% vs 13.5% (upper limit 5.2%); per-protocol 12.7% vs 13.1% (upper limit 5.8%).

• Individual events: infection 9.2% vs 11.6% (p=0.4); reoperation 2.7% vs 1.5% (p=0.4); explantation 7.5% vs 5.2% (p=0.3); seroma 20.0% vs 27.4% (p=0.056).

• BREAST-Q physical well-being was similar at baseline (median 92 both groups), 30 days (60 vs 55; p=0.2), and 90 days (76 vs 76; p=0.3). Pain trajectories were also similar.

Conclusion

The authors conclude that ADM-assisted two-stage prepectoral reconstruction is noninferior to no ADM for 90-day major complications, with comparable minor complications, physical well-being, and pain. They emphasize that longer follow-up is needed to determine whether ADM provides meaningful long-term clinical benefits.

Strengths

• Prospective randomized design with a prespecified noninferiority margin, power calculation, intention-to-treat analysis, and concordant sensitivity analyses.

• Large breast-level cohort with standardized prepectoral technique and objective skin-flap perfusion assessment.

Limitations

• Single-center, unblinded study in a relatively healthy, nonsmoking population; generalizability to higher-risk patients and other practice settings is uncertain. Note: avg. BMI of 24.4

• Most enrollment used month-level cluster randomization, with treatment known in advance to surgeons and staff; this introduces potential selection and temporal-practice bias despite similar measured baseline characteristics.

• Protocol crossover occurred, although sensitivity analyses were consistent. The 7% noninferiority margin could still permit a clinically meaningful excess complication rate.

• Follow-up stops at 90 days after expander placement. The study does not establish benefit for capsular contracture, rippling, implant malposition/bottoming out, aesthetic outcome, or cost.

Clinical relevance

For surgeons already using ADM, these data are reassuring that ADM does not appear to impose a major short-term complication penalty within the trial’s 7% margin. For surgeons omitting ADM, the nearly identical observed outcomes provide no short-term evidence that routine ADM is necessary; the choice should remain driven by longer-term reconstructive goals and cost until later follow-up is available. This study is limited by its 90 day follow up; Practice-changing evidence will come from the planned post-exchange follow-up, ideally incorporating aesthetic assessment, revision burden, capsular contracture, rippling, and formal cost-effectiveness.