Pioneering work in small-vessel surgery using a surgical microscope was described by Julius Jacobson in 1960.1 This led to the first free tissue transfer by McLean and Buncke in 19722 and the birth of microsurgery. Almost 60 years later, the surgical microscope remains the predominant method for visual magnification to enable successful microsurgical anastomoses in free tissue transfer.
Recently, however, microsurgical anastomoses performed with loupe magnification have been demonstrated to be safe and efficient, with outcomes similar to those of free flaps performed with the operating microscope.3–9 The majority of these studies, however, have focused primarily on abdominally based free tissue transfer for breast reconstruction. Breast free flap reconstruction has been refined to allow for a very favorable microsurgical environment. These vessels are typically larger than 2.5 mm in diameter, and anastomoses are performed end-to-end with a broad flat exposure to work in without concern for a zone of injury.
Traumatic lower extremity salvage presents the reconstructive surgeon with a significantly more hostile microsurgical environment compared with breast free flaps and can be viewed as a true proving ground for different microsurgical techniques. Despite the notoriously higher degree of microsurgical difficulty and significant complication rates in traumatic lower extremity reconstruction, our experience has found loupe-only microsurgery to be a viable option for free flap limb salvage. Considering that loupe-only free flap limb salvage has not been described, this study critically evaluated the safety and efficacy of the technique by comparison to a matched cohort of traditional free tissue transfers using the operating microscope for traumatic lower extremity limb salvage.
PATIENTS AND METHODS
After institutional review board approval, free flap databases at two large metropolitan institutions were retrospectively queried for patients who underwent lower extremity reconstruction. Perioperative data were available for patients who underwent free tissue transfer at the University of Pennsylvania Health System between 2002 and 2016. The New York University Langone Health free flap registry extends from 1976 to 2016; however, only flaps performed between 1996 and 2016 were included, as this corresponds to the period when negative-pressure wound therapy was available for use and ensures similar treatment modalities were available.
Charts were reviewed for patient demographics, injury characteristics, arterial runoff, free tissue transfer operative data, and perioperative free flap outcomes. Arterial injury was defined as either abnormal preoperative angiography related to the trauma (not related to preexisting peripheral vascular disease) or intraoperative identification of arterial injury. Major complications were considered to be take-backs and partial or total flap failures. Take-backs were defined as emergent return to the operating room for suspected vascular compromise. Partial flap failure was defined as flap loss requiring return to the operating room for flap débridement during the index hospitalization. Total flap failure was defined as complete flap loss requiring additional coverage procedures or resulting in amputation.
In both groups, the free flap elevation and recipient vessel dissection portions of the operation were performed using loupe magnification. The decision to use loupes versus microscope magnification to perform the microvascular anastomoses was dependent on surgeon preference and intraoperative discretion. Anastomoses in the loupe-only group were performed using surgical loupes between 3.5 and 4.0× magnification. Arterial anastomoses were performed using 8-0 or 9-0 nylon suture, and a venous coupler was used in the majority of cases at both institutions.
Statistical Analysis and Matching Description
Data were analyzed using Pearson chi-square or Fisher’s exact tests to examine categorical variables and t test for continuous variables to assess for significant differences in patient demographics, injury details, arterial anatomy, and flap characteristics between the two microsurgical technique cohorts. Patients with missing data points were not included in univariate analysis, and reported p values reflect all available data within the cohort. Results with a value of p < 0.05 were considered to be statistically significant. Analyses were performed using STATA IC 11.0 (StataCorp., College Station, Texas).
Optimal full matching10,11 between loupe and microscope cohorts before outcomes analysis was performed by a biostatistician from the University of Pennsylvania Department of Biostatistics, Epidemiology, and Informatics. A distance matrix was applied that combined a caliper on an estimated propensity score with a Mahalanobis distance.12 Strata pairs were matched according to age, wound location (leg or foot), time from injury (flap coverage performed within 30 days from injury or after 30 days from injury), type of free flap (muscle or fasciocutaneous), and arterial injury (present or absent). Perioperative free flap outcomes were then compared between the loupe and microscope groups using adjusted odds ratios obtained from conditional logistic regression models with matching strata.
Strict inclusion criteria honed our cohort to 373 soft-tissue free flaps performed for coverage of traumatic defects below the knee (Table 1). Microsurgical anastomoses were performed with loupes in 150 flaps, and the remaining 223 flaps were performed with the operating microscope. Loupe-only flaps were performed solely by surgeons at the University of Pennsylvania Health System, and the majority of microscope flaps (83.9 percent) were performed at New York University Langone Health. The major complication rate for both groups combined was 15.3 percent (n = 57). Take-backs for free flap vascular compromise occurred in 29 flaps (7.8 percent). Any flap failure was observed in 49 flaps (13.1 percent), and consisted of 29 partial flap failures (7.8 percent) and 20 total flap losses (5.4 percent). Direct comparison of free flap outcomes according to institution (Table 2) demonstrated no significant differences between the University of Pennsylvania Health System (n = 186) and the New York University Langone Health (n = 187) groups with regard to major perioperative complications (p = 0.682), take-backs (p = 0.341), any flap failure (p = 0.894), partial flap failure (p = 0.572), or total flap failure (p = 0.637).
Table 1. -
Patient Demographics and Flap Characteristics According to Microsurgical Technique
| Penn 2002–2016
| NYU 1996–2016
|Tibial defect location†
|Acute trauma (<30 days)†
|Arterial injury present†
| AT injury
| PT injury
| Peroneal injury
| 3 vessel
| 2 vessel
| 1 vessel
| 0 vessel
|Recipient artery injured
|Arterial anastomosis type
|Two venous anastomoses
|Vein graft used
enn, University of Pennsylvania Health System; NYU, New York University Langone Health; AT, anterior tibial; PT, posterior tibial.
Univariate analysis before optimal cohort matching for outcomes analysis.
Covariates used for optimal matching between loupe and microscope groups.
Table 2. -
Comparison of Free Flap Outcomes by Institution
|Any flap failure
|Partial flap failure
|Total flap failure
enn, University of Pennsylvania Health System; NYU, New York University Langone Health.
Comparison of patient characteristics before matching demonstrated the loupe-only cohort to be older (47 versus 39; p = 0.001), with a higher proportion of tibial defect sublocation (75.3 percent versus 63.7 percent; p = 0.018) and fewer flaps performed within 30 days of injury (42.7 percent versus 56.3 percent; p = 0.011). Loupe-only flaps were predominantly fasciocutaneous (70.0 percent), whereas the microscope flaps were more often muscle-based (71.8 percent; p = 0.001). The posterior tibial artery was the most common recipient vessel in both groups; however, the loupe-only group used the posterior tibial artery more frequently than the microscope group (64.0 percent versus 49.3 percent; p = 0.020).
The majority of loupe arterial anastomoses were performed in end-to-side fashion (72 percent), whereas the microscope group had an even distribution of end-to-end (49.3 percent) and end-to-side (50.7 percent) anastomoses (p = 0.001). A significantly higher number of microscope flaps used an injured artery as the recipient vessel (15.9 percent versus 5.3 percent; p = 0.002). There was no difference in vein graft use between groups (p = 0.279). The microscope group trended toward performing a second venous anastomosis (30.9 percent versus 22.0 percent); however, this did not reach significance (p = 0.056).
Angiography data were available for 208 patients (56 percent), and demonstrated arterial injury in 92 cases. There was no difference between groups in terms of arterial injury presence (p = 0.067), which was used as a surrogate marker for degree of injury severity.13 The anterior tibial artery was the most commonly injured vessel in both groups, with a similar incidence (p = 0.274). The posterior tibial injury, however, was more common among the microscope group (11.6 percent versus 4.7 percent; p = 0.025) and may be related to the observed recipient vessel differences. Arterial runoff patterns were similar between groups (p = 0.128), but there were more Gustilo type IIIC injuries in the microscope group (n = 11 versus n = 1).
Optimal matching was successfully performed between the two cohorts (Table 3) according to age, time from injury to coverage, defect sublocation, flap type, and presence of arterial injury. Comparison of matched groups demonstrated equivalent perioperative free flap outcomes (Table 4). Between the loupe and microscope groups, major complication rates (14.0 percent versus 16.1 percent), take-backs (5.3 percent versus 9.4 percent), partial flap failures (7.3 percent versus 8.1 percent), and total flap losses (6.0 percent versus 4.9 percent) were independent of the method of visual magnification used to perform the arterial anastomosis.
Table 3. -
Microsurgical Technique Covariates Balance before and after Matching
|Mean age ± SD, yr
||46.2 ± 16.3
||39.1 ± 17.1
||41.5 ± 16.5
||41.0 ± 16.5
|Tibial defect, %
|Flap coverage <30 days from injury, %
|Fasciocutaneous flap, %
|Arterial injury present, %
The standardized difference is the ratio of the mean to the standard deviation of the difference between the loupe and microscope groups and measures the effect size or magnitude of difference between two groups. If the standardized difference for a covariate after matching is <0.1, it indicates that the matching balances the covariate.
Table 4. -
Matched Perioperative Free Flap Outcomes According to Microsurgical Technique
||Adjusted OR (95% CI)*
|Any flap failure
|Partial flap failure
|Total flap failure
OC, receiving operating characteristic.
Adjusted OR obtained from conditional logistic regression models with matching identifications as strata.
Adjusted area under the ROC curve.
Although free flaps are typically raised with loupe magnification, the field of microsurgery derives its very name from the operating microscope. Traditionally, this has been the preferred method of visual magnification to perform the vascular anastomoses required for free tissue transfer. Microvascular anastomoses performed with loupes, because of the lower power magnification, are often considered less precise and therefore less safe by proponents of the microscope.14
Loupe-only microsurgery, however, provides a number of potential advantages over the operating microscope.3,9 Although the magnification power is significantly higher with an operating microscope (6 to 40× versus 3.5 to 4.5×), it provides this within a fixed viewing angle. Loupes enable viewing angle and depth-of-field adjustments in real time through changes in head position, which improves vessel lumen visualization with minimal need for vessel manipulation or microscope repositioning. In addition, set-up time and operative equipment bulk is decreased compared to the microscope, which may decrease overall operative time.9 Despite traditional concerns about decreased precision, loupe-only microsurgery has been shown to be effective in breast reconstruction,3–5,9 with anastomotic revision rates and free flap outcomes similar to published literature with the operating microscope.
Although the loupe-only technique is a safe alternative to the microscope in breast reconstruction, its use in other anatomical areas has not been thoroughly evaluated. Traumatic lower extremity free flap reconstruction remains notorious for high complication rates, with a microsurgical environment that is considerably more hostile and unfavorable than breast reconstruction, demanding the highest level of microsurgical precision. This challenging population provides a robust proving ground for the safety and efficacy of loupe-only microsurgery compared to the operating microscope.
The results of our matched, multi-institution comparison demonstrated loupe-only microsurgery to be a safe and effective alternative to the operating microscope in soft-tissue free flap reconstruction of traumatic lower extremity defects. Overall microsurgical perioperative complication rates were 14 percent among the loupe-only group compared with 16 percent for the microscope group, with total flap failure rates of 4.9 percent and 6.0 percent, respectively. Unrecognized technical errors often result in take-backs for vascular compromise; however, there was no difference between the groups, and actually a trend toward lower take-back rates among the loupe flaps was observed (5.3 percent versus 9.4 percent).
Considering the nearly identical free flap outcomes demonstrated, the decision of which microsurgical technique to use is left to the surgeon’s discretion. In our collective opinion, the optimal visualization technique is the method that the operating surgeon is most comfortable with that will lead to the best possible outcomes. It is certainly advantageous to suture with higher magnification; however, the value of changing viewing angles in real time cannot be overstated, particularly when an anastomosis must be performed at a difficult angle or in a deep operative field.
There are technical limitations to loupe-only microsurgery with decreasing vessel size, however, with evidence that the operating microscope should be used for vessels smaller than 1.0 to 1.5 mm in diameter.8 The microscope has also been recommended for pediatric patients, digital replantation distal to the palmar arch, major peripheral nerve repairs, and digital nerve repairs.4,5,8,15–17 Unfortunately, vessel diameter was not available for the majority of cases to allow for direct comparison between microsurgical techniques and determine whether a size cutoff was present. Anecdotally, the surgeons included in this study who routinely perform loupe-only anastomoses use 1.5 mm as their typical cutoff, below which they use the operating microscope.
Loupe-only microsurgery has the potential to shorten operative times and improve operating room efficiency compared with use of the operating microscope, which requires a relatively large amount of space and more time to set up.3,9,15 These factors may translate to potential cost savings by reducing operative time and avoiding a microscope bottleneck if there are multiple rooms requiring its use. Operative times were not reliably captured within our data set, however, and we were unfortunately unable to investigate differences in operative times between the two microsurgical techniques.
The retrospective nature of this study leads to inherent limitations. Lower extremity salvage patients are heterogeneous, and coverage decisions are largely surgeon dependent, which can make comparisons between groups difficult. To account for this, only soft-tissue free flaps performed for coverage of a defect related to trauma below the knee were included in our study. Although arterial injury was used as a surrogate marker for injury severity, it would have been preferable to evaluate additional variables related to injury severity such as bone gap presence, defect size, and polytrauma status; however, these variables were unfortunately not reliably captured among our cohort to allow for adequately powered and unbiased analysis. The use of two large separate metropolitan institutions captured multiple different reconstructive surgeons to further mitigate operative bias within the cohort. Although all loupe flaps were performed at the University of Pennsylvania Health System, five different surgeons are included within that subgroup, and none of the perioperative flap outcomes were significantly different between institutions, thus reducing the potential influence of institutional or surgeon-specific bias. In addition, matching according to relevant lower extremity reconstructive variables (i.e.,age, defect location, time from injury to coverage, flap type, and presence of arterial injury) generated the most unbiased comparison possible between the two surgical techniques within the limitations of our available data. Given that we were evaluating microsurgical technical outcomes, long-term limb salvage outcomes beyond flap success rates were not evaluated in this study.
This multicenter, matched comparison of perioperative microsurgical outcomes demonstrated loupe-only microsurgery to be a safe and effective technique for soft-tissue free flap coverage of traumatic lower extremity defects. The optimal method of microvascular visualization remains entirely up to surgeon’s discretion and should be the modality with which the operating surgeon is most comfortable.
The authors sincerely appreciate the contribution of Jesse Y. Hsu, Ph.D., assistant professor of biostatistics, Department of Biostatistics, Epidemiology, & Informatics at the University of Pennsylvania Perelman School of Medicine, for advanced statistical analysis and optimal matching technique that made this work possible.
1. Rickard RF, Hudson DA. A history of vascular and microvascular surgery. Ann Plast Surg. 2014;73:465–472.
2. McLean DH, Buncke HJ Jr.. Autotransplant of omentum to a large scalp defect, with microsurgical revascularization. Plast Reconstr Surg. 1972;49:268–274.
3. Pannucci CJ, Basta MN, Kovach SJ, Kanchwala SK, Wu LC, Serletti JM. Loupes-only microsurgery is a safe alternative to the operating microscope: An analysis of 1,649 consecutive free flap breast reconstructions. J Reconstr Microsurg. 2015;31:636–642.
4. Serletti JM, Deuber MA, Guidera PM, et al. Comparison of the operating microscope and loupes for free microvascular tissue transfer. Plast Reconstr Surg. 1995;95:270–276.
5. Shenaq SM, Klebuc MJ, Vargo D. Free-tissue transfer with the aid of loupe magnification: Experience with 251 procedures. Plast Reconstr Surg. 1995;95:261–269.
6. Ashworth DR, Whear NM, Fan V. Radial free flaps using loupe magnification: Audit of 97 cases of orofacial reconstruction. Br J Oral Maxillofac Surg. 2004;42:36–37.
7. Ross DA, Ariyan S, Restifo R, Sasaki CT. Use of the operating microscope and loupes for head and neck free microvascular tissue transfer: A retrospective comparison. Arch Otolaryngol Head Neck Surg. 2003;129:189–193.
8. Pieptu D, Luchian S. Loupes-only microsurgery. Microsurgery 2003;23:181–188.
9. Ehanire T, Singhal D, Mast B, Leyngold M. Safety of microsurgery under loupes versus microscope: A head-to-head comparison of 2 surgeons with similar experiences. Ann Plast Surg. 2018;80(Suppl 6):S340–S342.
10. Hansen BB. Full matching in an observational study of coaching for the SAT. J Am Stat Assoc. 2004;99:609–618.
11. Hansen BB, Klopfer SO. Optimal full matching and related designs via network flows. J Comput Graph Stat. 2006;15:609–627.
12. Rosenbaum PR. Constructing matched sets and strata. In: Observational Studies
(Springer Series in Statistics). 2002:New York: Springer; 295–331.
13. Stranix JT, Lee ZH, Jacoby A, et al. Not all Gustilo type IIIB fractures are created equal: Arterial injury impacts limb salvage outcomes. Plast Reconstr Surg. 2017;140:1033–1041.
14. Tomaino MM. Routine use of loupe magnification for microvascular anastomoses: At what price? Plast Reconstr Surg. 1996;97:248–249.
15. Stanbury SJ, Elfar J. The use of surgical loupes in microsurgery. J Hand Surg Am. 2011;36:154–156.
16. Hallock GG. Macrovascular surgery and the microsurgeon. J Reconstr Microsurg. 1997;13:563–570.
17. Bernstein DT, Hamilton KL, Foy C, Petersen NJ, Netscher DT. Comparison of magnification in primary digital nerve repair: Literature review, survey of practice trends, and assessment of 90 cadaveric repairs. J Hand Surg Am. 2013;38:2144–2150.