A single-center study of 194 elective ascending aortic and root operations compared transaxillary access with full and partial sternotomy. After propensity matching, transaxillary surgery required longer bypass time, but early clinical outcomes did not differ significantly from partial sternotomy. No transaxillary patient died in hospital or had a stroke. Careful anatomical selection, a small cohort, and residual confounding limit conclusions about comparative safety.
Extending Sternal-Sparing Access to Aortic Surgery
Daniel-Sebastian Dohle and colleagues at University Medical Center Mainz examined whether transaxillary access could support elective ascending aortic and root surgery across a range of procedures. Their retrospective study compared full sternotomy, partial upper sternotomy, and a lateral transaxillary approach. The main finding was that selected transaxillary patients had favorable early outcomes, with longer cardiopulmonary bypass exposure but no statistically significant difference in clinical outcomes compared with propensity-matched partial-sternotomy patients.
The distinction between feasibility and comparative safety is central to interpreting this work. A small observational cohort can show that a procedure was completed successfully in selected patients, but it cannot establish that the approach is equivalent or superior to alternatives. For perfusion teams, the study is particularly relevant to preoperative vascular assessment, cannulation planning, bypass duration, and the relationship between surgical exposure and extracorporeal support.
Study Population and Comparison Strategy
The investigators included 194 consecutive adults undergoing elective, isolated ascending aortic or root surgery between January 2022 and December 2025. Full sternotomy was used in 52 patients, partial upper sternotomy in 106, and transaxillary access in 36. The procedural spectrum included supracoronary ascending aortic replacement, Wheat, Bentall, and David procedures. Emergency surgery, aortic dissection, aortic arch replacement, and planned coronary bypass for known coronary artery disease were excluded.
Access was selected by the surgeon according to anatomy, disease complexity, and technical feasibility. Full-sternotomy patients had a higher baseline risk profile and were retained as a descriptive reference rather than an adjusted comparator. The more informative comparison matched 34 transaxillary patients to 68 partial-sternotomy patients using age, sex, body mass index, EuroSCORE II, preoperative creatinine, and root replacement.
Two transaxillary patients with elevated preoperative creatinine could not be matched and were excluded from that analysis. Creatinine remained imbalanced after matching, with a standardized mean difference of 0.30 and lower values in the transaxillary group. This is particularly relevant when interpreting the numerical differences in postoperative acute kidney injury.
Anatomical Selection and Cannulation Planning
Selection relied on team review of ECG-gated CT angiography extending from the thoracic inlet to the femoral bifurcation. Assessment included aortic position, instrument reach, distal disease extent, clamp-zone calcification, coronary anatomy, and suitability of peripheral vessels. The team did not use a fully standardized quantitative selection model, and the study did not maintain a screening log that would establish what proportion of all candidates were eligible.
Patients requiring an open distal anastomosis were directed toward partial or full sternotomy with antegrade cerebral perfusion. The authors describe a distal ascending or proximal arch diameter of at least 45 mm, substantial caliber change, and aneurysms above approximately 6 cm as important considerations in their institutional selection process. These are descriptions of the study’s approach, not new thresholds established by its results. The extent of aortic repair was determined by pathology rather than by a desire to preserve the smaller incision.
For transaxillary procedures, femoral arterial and venous access was generally used. The left subclavian artery was selected in three patients when femoral access or retrograde perfusion was unsuitable. The authors note that positioning with the right arm elevated restricted access to the right subclavian artery. Thus, the surgical access route and the arterial cannulation site should not be treated as interchangeable concepts: transaxillary surgery did not mean routine axillary arterial perfusion.
Why Bypass Time Was Longer
The transaxillary technique involved a small anterior axillary incision and direct-vision work with long-shafted instruments. Cardiopulmonary bypass was established before opening the pericardium, dissecting the ascending aorta, and identifying the clamp site. In sternotomy procedures, these steps generally occurred before bypass. The authors attribute part of the additional extracorporeal support time to this difference in operative sequence.
In the matched comparison, median bypass time was 196 minutes with transaxillary access versus 160 minutes with partial sternotomy, a statistically significant difference. Median cross-clamp times were 132 and 120 minutes, respectively, without a statistically significant difference. Skin-to-skin operating time was also longer with transaxillary access: 300 versus 275 minutes. Similar cross-clamp times should not be interpreted as proving that the added bypass exposure is harmless; the cohort was too small to exclude clinically meaningful differences in uncommon complications.
Matched Early Outcomes
| Outcome | Transaxillary, n=34 | Partial sternotomy, n=68 | P value |
|---|---|---|---|
| CPB time, median minutes | 196 | 160 | <0.001 |
| Cross-clamp time, median minutes | 132 | 120 | 0.16 |
| Major complications | 5.9% | 11.8% | 0.49 |
| Acute kidney injury, any stage | 2.9% | 10.3% | 0.26 |
| Hospital stay, median days | 9 | 9 | 0.52 |
| In-hospital deaths / strokes | 0 / 0 | 0 / 0 | Not applicable |
Intensive care stay, time to extubation, rethoracotomy, new-onset atrial fibrillation, and wound-healing disturbances also did not differ significantly between matched groups. The major-complication composite broadly followed the Society of Thoracic Surgeons framework, but included wound-healing disturbances of any depth because the database did not distinguish superficial from deep infection. Acute kidney injury was assessed using KDIGO serum-creatinine criteria.
Across all 36 transaxillary patients, the planned access was completed in 35. One patient underwent a precautionary conversion to full sternotomy after separation from bypass because bleeding at the right coronary ostium could not be securely controlled through the smaller access. Hemostasis was achieved without reported sequelae. No cannulation-related complications were reported in the transaxillary group. These observations support feasibility while underscoring the importance of readiness to convert when exposure limits safe control.
Transfusion and Coagulation Findings
Matched red-cell transfusion requirements did not differ significantly. Prothrombin complex concentrate use was higher with transaxillary surgery, averaging 515 versus 321 IU. The authors explain that prophylactic PCC administration after protamine was increasingly adopted for lateral thoracotomy procedures during the study period. This practice difference is relevant to interpreting bleeding and transfusion outcomes; the study does not establish an indication for routine prophylactic PCC.
The unadjusted full-sternotomy group had more complications, longer stays, and greater transfusion requirements. However, that group also had higher risk scores and more complex accompanying procedures. Its outcomes cannot be used to claim that transaxillary access reduces mortality, kidney injury, or other complications relative to full sternotomy.
Evidence Strength and Perfusion Relevance
The study adds a useful comparative dataset to a field previously dominated by smaller case series. Its strengths include consecutive inclusion, defined outcomes, a detailed description of anatomical selection, and propensity matching against partial sternotomy. Important limitations remain: retrospective single-center design, only 36 transaxillary patients, residual anatomical selection bias, incomplete balance in renal function, and the inclusion of a learning-curve period after introduction of the approach in September 2023.
There was no adjustment for multiple comparisons, and the study was not powered for rare outcomes such as death or stroke. Absence of statistically significant differences is not evidence of equivalence. The results support further prospective evaluation of transaxillary ascending aortic and root surgery in carefully selected patients, with particular attention to perfusion exposure, cannulation strategy, conversion planning, and longer-term recovery and aortic outcomes.
Attribution: © AME Publishing Company. The unchanged original article is distributed under CC BY-NC-ND 4.0. This independently written IPA summary accompanies the source article. Cover artwork is conceptual.
Original Article
Dohle DS, Pfeiffer P, Keller C, Zancanaro E, Ghazy A, Masseli F, Probst C, Treede H. Transaxillary access for ascending aortic and root surgery: a single-center comparison with full and partial sternotomy. Ann Cardiothorac Surg. 2026;15(5):65. Read the original article.
