CHANGES IN DONOR SITE SELECTION IN LOWER LIMB FREE FLAP RECONSTRUCTIONS BY INTEGRATING DUPLEX ULTRASONOGRAPHY IN THE PREOPERATIVE DESIGN

INTRODUCTION

The introduction of microsurgical techniques for soft tissue reconstruction of complex lower limb trauma has revolutionized the management of these injuries.  The former in conjunction with improved techniques for bony fixation were reported to result in a high percentage of limb salvage (1–3). The selection of an appropriate soft tissue donor site is a key decision in the algorithm of the reconstruction, since it has a profound effect on the outcome (4). This decision was based largely on clinical criteria, including the wound location, the size of the defect, the type of tissue missing and the functional requirements. Also, the presence of contamination or infection, the length of vascular pedicle required and the donor site morbidity will delineate the decision-making (5, 6). Optimal functional and aesthetic reconstruction (7) could be achieved by matching the volume and the surface area of the defect with a free flap of an appropriate size and volume.

For decades, the prevailing theory consisted of the notion that muscle was an essential component of the lower extremity wound coverage (8), but harvesting is associated with functional donor site morbidity. Perforator flaps provided novel perspectives in reconstructive microsurgery as function at donor site is preserved, while achieving equal or superior reconstructive outcomes to musculocutaneous flaps (9). The preference of the majority of reconstructive surgeons dealing with lower limb trauma has gradually moved towards perforator flaps, but three-dimensional defects and weight-bearing areas may still require a muscle transfer as a better option; thus, preoperative planning is mandatory (10). Imaging modalities such as digital subtraction angiography, computed tomography angiography and magnetic resonance angiography revolutionized the planning of free tissue transfer and ensured low complication rate. Nevertheless, the above modalities are expensive, require professional expertise and provide mainly anatomic information. In contrast, duplex ultrasonography (DU), despite its inherent limitations, provides both anatomic and hemodynamic information for the donor and recipient sites and is a valuable diagnostic and decision-making tool in reconstructive microsurgery (11).

The end-point of this study was to investigate whether the decision-making for the choice of a free flap donor site in microvascular lower extremity reconstruction was affected by integrating DU in the preoperative design.

PATIENTS AND METHODS

Forty-eight patients (32 males and 16 females, mean age 43 years) with complex lower limb defects who underwent microvascular soft-tissue reconstruction in a single institution between January 2004 and January 2013 by the same surgeon (AG) were enrolled to this study.

Ultrasound investigation was focused on recipient vessels (diameter, flow) and pedicle length requirements (Fig. 1A). DU was used to investigate the donor site focusing on the pedicle and perforator’s characteristics, as well as, flap dimensions (length, width, thickness) to match the three-dimensional limb defect (Fig. 1B). Patient data, defect, ultrasound findings, surgical technique, and outcome data were obtained from a prospectively maintained database and medical records.

DU was employed in the preoperative assessment of patients treated in the period from January 2009 to January 2013, whereas clinical criteria alone were used for donor site selection in patients that were treated in the period from January 2004 to December 2008. Therefore the patients were separated into two groups. Group A included 20 patients, who underwent preoperative assessment of the recipient and potential donor sites with DU. The decision on the donor site and the level of vascular anastomosis were based on clinical criteria, assisted by the additional information provided by the DU. Group B included 28 patients in whom the choice of free flap donor site and level of anastomosis were based on clinical criteria only. The variables analyzed included measurements of the lower limb defects, ultrasound findings from the donor site and the recipient site, surgical method and outcome measures.

Study protocol

The surface area and depth of the defect were recorded in all cases. The type of tissue defect (skin, muscle, and tendon) and the presence of contamination and/or infection were taken into account. In Group B, the choice of free flap donor site and the level of anastomosis were based on clinical criteria only. In Group A, DU was integrated in the preoperative design of the microsurgical reconstruction.

Specifically, DU measurements were performed by a single observer (DK), under the guidance of the microsurgeon, using a HD11 XE ultrasound machine (Philips, Andover, MA, USA) equipped with a high-resolution linear 12–20 MHz transducer. The transducer was covered with ultrasonic gel and wrapped in an intraoperative sterile sheath (Microtec medical intraoperative probe cover, 12 cm x 244 cm) to minimize possible wound contamination. First, the vessels adjacent to the injury zone were located by DU and blood flow in the arterial conduit of the recipient site was evaluated by usual two-dimensional, color and standard Doppler measurements as described in detail elsewhere (11). The vessels were scanned on the longitudinal axis, while the operator carefully adjusted the transducer to minimize the angle between the Doppler beam and the long axis of the artery and also to ensure that the sampling volume was located within the vessel lumen for as much of the cardiac cycle as possible. Thereafter, the level where the vessel achieved normal flow patterns was noted. This was defined as continuous uninterrupted laminar flow in the arterial conduit of the recipient site without evidence of aliasing in color mode, which was further confirmed by pertinent normal Doppler velocities profiles. The latter were assessed by the time-velocity integral in diastole (VTID) and systole (VTID) and its ratio (d/s), which were calculated real-time by tracing the contour profile of Doppler velocities using a computerized analysis system (Q-LAB, Philips, Andover, USA), as previously described (11). This level of recorded normal blood flow was designated as the preferred site for the microsurgical anastomosis, while the distance from the defect determined the length of the free flap pedicle required. Thereafter, blood flow in the arterial conduit of the donor site was evaluated by usual two-dimensional, color and standard Doppler measurements accordingly. Finally, color-Doppler techniques were used to evaluate the diameter and patency of the venous conduit at the recipient and donor sites, respectively. Preoperative ultrasound measurements and confirmatory intraoperative measurements preceding the microsurgical operation were performed in all cases. Doppler measurements were obtained for 10 consecutive cardiac cycles and average values were included in the statistical analysis. The perivascular anatomy was assessed by usual two-dimensional ultrasound techniques in all cases. In perforator skin flaps the thickness of the flap was assessed. If muscle was used as a donor tissue, its width, length, thickness and volume were measured, accordingly.

Statistical analysis

Summary data are presented as means ± SD. The Student’s t test for independent means, χ² analysis, or Fisher’s exact test where appropriate were used to identify differences between the two groups. In the DU group no comparisons of pertinent ultrasound variables between perforator and muscle flaps were performed as the small number of cases prevented a meaningful subgroup analysis. A p-value (two-sided in all tests) of <0.05 was considered significant. Analysis was performed with commercially available software package SPSS 11.0 (SPSS Inc. Chicago, IL, USA).

RESULTS

In a total of 48 microsurgical lower limb reconstructions there were three flap failures. All of the failures were recorded in group-B. This difference in the flap survival rate was statistically significant (p<0.05, χ²). In the DU group (group-A) 9 perforator skin flaps, 10 muscle flaps and one chimeric vastus lateralis flap were used. All of the perforator flaps were anterolateral thigh (ALT) flaps, while the muscle flaps employed for reconstruction included 4 vastus lateralis, 2 gracilis, 2 latissimus dorsi and one rectus abdominis muscle flaps, respectively. In the clinical criteria group (group-B), 19 skin flaps were used for the reconstruction, 64% of the total. These included 18 ALT perforator flaps and 1 radial forearm flap. In this group, 9 muscle flaps were used, including 7 gracilis, one latissimus dorsi and one rectus abdominis, comprising 32% of the total flap reconstructions in this group of patients. Comparing the 2 groups, there was a significant decrease in perforator flap (45% over 64%, p< 0.05, Fisher’s test), and a significant increase in muscle flap preference (55% over 32%, p< 0.05, Fisher’s test) in the DU group (Fig. 2A). Nevertheless, the significant decrease in perforator flap use was associated with the more frequent use of muscle flaps with low morbidity such as gracilis and vastus lateralis (Fig. 2B).

In 5 out of 20 cases (25%) in Group A, vessels with normal flow patterns were identified within the zone of injury and they were used for the anastomosis. Doppler flow measurements have shown that the time-velocity integral in diastole and systole and its ratio at the designated preferred level of the anastomosis located at the recipient site were: VTIS= 7.1±0.3, VTID= 0.9±0.3 in cm, and d/s=0.12±0.06, respectively, for the sum of cases in Group A. In the latter, mean distance of the anastomosis from the injury level was 2.8 cm (95% confidence intervals= 0 to 5 cm). The requirements for pedicle length in Group A exhibited mean pedicle length of 5.8 cm (95% confidence intervals= 4 to 7 cm). In Group B, anastomosis were performed at a mean distance of 5.6 cm (95% confidence intervals= 5 to 7 cm) above the level of injury. Flap pedicle length in this group was 11.5 cm (95% confidence intervals= 9 to 15 cm). Differences between the 2 groups concerning mean distance of anastomosis and pedicle length were both statistically significant (p< 0.05, t-test). The venous conduits of all chosen recipient and donor sites in Group A were identified as patent and competent by DU measurements. However, the sonoanatomic investigation performed in group A have shown 3 out of 20 cases (15%) of chronic perivascular inflammation with associated scar tissue, microthrombosis and fat necrosis adjacent to the area of injury; thus these areas were avoided in the selection as potential sites for anastomosis. Finally, the wound healing time was significantly reduced in the DU group of patients (21±3 days) compared to Group B (37±3 days; p< 0.05, t-test).

DISCUSSION

In attempts to salvage massive lower extremity injuries, soft tissue coverage is no longer a limiting factor due to the recent advances in microvascular composite tissue transfer (10). The extremity that requires free tissue transfer typically has sustained a high-energy injury and is severely damaged or mangled. This type of injuries is associated with major vessel injury and the extent of the vascular trauma is difficult to estimate. Increased friability of the vessels and increased perivascular scar tissue can contribute to a higher rate of microvascular thrombosis (12). Thus, the assessment of the level of healthy vessels that can be used as recipients is of outmost importance in planning free tissue transfer (13). Park et al stated that the type of flap used, method, and site of microvascular anastomosis are less important factors in determining the recipient vessels (13). On the other hand, the distance of healthy recipient vessels from the defect would influence the decision for the most appropriate flap to transfer. Commonly, flaps with long pedicle are used in order to ensure anastomoses are located outside the zone of injury; therefore perforator flaps such as ALT (14) and TDAP (15) flap have gained popularity.

Traditionally, free muscle flaps have been generally considered superior in obliterating dead space and combating osteomyelitis in the lower limb because of better pliability and vascularity (8, 10). Nevertheless, there has been a tendency toward the use of more cutaneous and perforator flaps over the past few years (14, 16). Perforator flaps have been proved very reliable with long and wide vascular pedicle that facilitates anastomoses far from the injury zone and ensures high success rate (14, 15). Moreover, stable wound coverage and bony union were achieved with perforator skin flaps even in the presence of osteomyelitis (17, 18).

Taking into account all of the above, perforator flaps have been our first choice for soft tissue lower extremity reconstruction in the period from January 2004 to December 2008. By using anterolateral thigh perforator flap as a workhorse we achieved high success rate and satisfactory healing (18). Apart from the first enthusiasm and trend with perforator flaps, the long vascular pedicle of ALT that facilitated the flap inset and microvascular anastomosis far from the zone of injury was the main advantage in our decision to ensure flap success and wound healing. Nevertheless, grossly contaminated wounds, weight-bearing areas and three-dimensional defects achieved less favourable results with perforator flaps, in the long-term.

This retrospective study revealed a significant shift in our practice, with a decreasing trend in the use of perforator flaps and a compensatory increase in muscle flaps. This trend is attributable to the change of our flap preference in grossly contaminated wounds and weight-bearing areas. Although the change of preference was stimulated by clinical observations, it was further established by the integration of DU in our preoperative design. The length of the vascular pedicle could not restrict our decision anymore as DU indicated precisely the pedicle length requirements.  The shorter pedicle length that was required in this group of patients permitted the use of muscle flaps, such as gracilis and vastus lateralis. Among the various donor sites, the DU indicated the flap with the most suitable dimensions and shape to confront the three-dimensional defect. For instance in the DU-group, patients with heel and plantar defects were reconstructed with ALT chimeric flap (Fig. 3), vastus lateralis (Fig. 4), latissimus dorsi musculocutaneous and integra-grafted latissimus dorsi muscle flap (19).

Apart from the most suitable tissue to reconstruct a particular defect, major concern is the donor site morbidity especially in young and active patients. Perforator flaps despite their limitations in three-dimensional defects and grossly contaminated wounds ensure negligible donor site morbidity. On the other hand, it has been proved that the harvesting of gracilis (20) and vastus lateralis (21) muscle flaps is associated with minimal morbidity. In our experience, nearly every size of soft tissue defect in the lower extremity can be covered with these 2 muscle flaps. Occasionally, latissimus dorsi (23) or rectus abdominis (23) that are associated with considerable morbidity would be the flap of choice for a complex defect, due to its dimensions and pliability. In the DU group, the significant decrease in perforator flap use was compensated with the more frequent use of muscle flaps with low morbidity such as gracilis and vastus lateralis. However, the use of flaps with considerable morbidity was the same in the 2 groups of patients.

Even if perforator flaps have become more popular during the last years, there still are and will be indications for muscle flaps. In the early days of microsurgical lower limb reconstruction, flap survival and defect coverage was the main goal, with less attention paid to the long-term functional outcome. By performing refinements in flap selection by the use of DU, the functional results could be increased and donor-site morbidity decreased. In other words, the preoperative use of Duplex Ultrasound guided our decision to the more frequent use of muscle flaps with low morbidity that may conform better to the three-dimensional wounds, obliterate the dead spaces, and decrease the risk of infection by improving vascularity and oxygen delivery to the contaminated wound. Instead, shallow defects, such as those on the distal third of the tibia and ankle, should be managed by free fasciocutaneous or perforator flaps where underlying bone may be exposed, but no massive bone or soft tissue defects exist.

Surely, this retrospective study exhibits several limitations. The overall number of patients was small, while a meaningful subgroup analysis between perforator and muscle flaps used in Group A in terms of their microvascular characteristics was not feasible due to the limited number of cases. Also, DU exhibits limitations in measuring flow volume especially in vessels with diameters smaller than 2 mm (24). However, there is an increasing evidence that ultrasound remains useful in the hemodynamic studies of free flaps (11, 18). Despite the aforementioned limitations, this study clearly demonstrated that the integration of DU in the preoperative design of free flap reconstruction for complex lower limb defects facilitates the implementation of efficient microsurgical strategies favoring thus the chances of flap survival and reducing the wound healing time. 

Successful microsurgical lower extremity reconstruction today means optimal functional and aesthetic repair with minimal donor site morbidity. It can be achieved by an individual approach, appropriate evaluation of reconstructive requirements with Duplex Ultrasound, and adequate free flap selection, design, tailoring and utilization.

Address for correspondence:

Andreas Gravvanis, MD, PhD, FEBOPRAS,

10 Patroklou Str., Agia Paraskevi,

15343 Athens, Greece

E-mail: gravvani@yahoo.com