Source British Journal of Plastic Surgery Volume: 50 Number: n/a March 1997

The fate of the oblique abdominal muscles after free TRAM flap surgery

Ph. N. Blondeel*, W. D. Boeckxt, G. G. Vanderstraetent, R. Lysens§, K. Van Landoyt*, P. Tonnard*, S. J. Monstrey' and G. Matton'

Departments of 'Plastic and Reconstructive Surgery, University Hospital Gent, and Traumatology and Reconstructive Surgery, University Hospitals Leuven, and Departments of Physical Medicine and Rehabilitation, f University Hospital Gent and University Hospitals Leuven, Belgium

SUMMARY: During recent years, clinical research on the donor site morbidity after free or pedicled transverse rectus abdominis myocutaneous (TRAM) flap surgery has been focusing on the reduced flexion capacity of the abdominal wall. However, the rectus abdominis muscles have close interactions with their synergists and antagonists and collaborate with their neighboring muscles. The purpose of this study was to examine the consequences of partially resecting the rectus abdominis muscle on the different muscle groups of the abdominal wall.

Twenty free TRAM flap patients, 12-61 months (mean 32.1 months) after surgery, were clinically examined, evaluated for curl-up performance and underwent isokinetic dynamometry for flexion, extension and rotation. The patients were compared with 20 non-operated controls. Nineteen patients answered a questionnaire.

Abdominal wall abnormalities occurred in 10 patients: umbilical asymmetry (n = 3), abdominal wall asymmetry (n = 4), lower abdominal bulging (n = 2) and hernia (n = 1). Curl-up performance was less in the TRAM flap patients (P = 0.001, Mann-Whitney). Isokinetic flexion, extension and rotation were also less in the TRAM flap patients (Fisher's exact test).

This study indicates that what has been believed to be 'limited' surgical damage to the abdominal wall leads to an important reduction in flexion strength but to an even more important reduction of rotation strength due to bilateral displacement and damage of the insertion of the oblique muscles. Partial compensation by synergists is variable and unpredictable on an individual basis. These functional disorders can potentially lead to important changes in activities of daily life.

Although several surgeons have tried to limit the damage to the rectus abdominis muscle by selective harvesting of free transverse rectus abdominis myocutaneous (TRAM) flaps, TRAM flap harvesting still involves the resection and denervation of an important part of the rectus muscle above the arcuate line. Besides the structural changes inflicted on the abdominal wall leading to abdominal protrusion, bulging or hernias, the loss of function caused by interrupting the continuity of the muscle is at least equally important. Recent studies3-'0 evaluating the donor site morbidity of TRAM flaps have only been focusing on the decrease of flexion capacity. Interested in the consequences of partially resecting one rectus abdominis muscle on the function of the entire abdominal wall, i.e. flexion and rotation, we extended our investigations to the assessment of both the rectus and oblique abdominal muscles.

In this study, the flexing and rotating capacity of the trunk of free TRAM flap patients was evaluated by their ability to perform curl-ups and by isokinetic dynamometry and compared to a control group who had not had TRAM flaps. Patients' opinions of the results of surgery were assessed by a questionnaire.

Patients and methods

Twenty patients who had undergone a breast reconstruction with a unilateral free TRAM flap at least

12 months prior to this study (mean 32.1 months, range 12-61 months) by two of the authors (PhB, WB) at the University Hospitals Leuven were included in this study. The patients' mean age was 46.8 years (range 29-63 years) at the time of surgery. Their mean postoperative weight of 66.5 kg (range 47-92 kg) did not differ significantly from the mean preoperative weight of 66.2 kg (range 49-90 kg). Patients with tumour recurrence or distant metastasis were excluded. In 11 patients, a 6-8 cm long segment of the left rectus abdominis muscle had been harvested over its entire width above the arcuate line - TRAM (Left) group. In 9 patients the same dissection had been performed on the right side - TRAM (Right) group. Closure of the anterior rectus sheath was reinforced by synthetic mesh in 18 cases. The control group was composed by inviting a sister or female friend of the patient to perform the same tests at the same time. In this way, a group of 20 healthy non-operated women with comparable socio-economic background, physical condition, age (mean 45.4 years; mean 29-68) and weight (mean 68.7 kg; range 52-97) was formed.

Patients and controls were examined supine and upright by a physician other than the surgeons. All patients were inspected for asymmetric positioning of the umbilicus, abdominal wall asymmetry, lower abdominal bulging and hernias. Asymmetry of the abdominal wall was dehmed as a unilateral distension of the lower abdomen. Bulging was defined as a protrusion of a part of the lower abdominal wall with palpable edges but without a defect in the abdominal fascia. A hernia was defined as a defect in the deep abdominal fascia.

A postoperative CT scan or MRI of the Iower abdominal wall between umbilicus and pubis was offered to the patients to evaluate the anatomical changes.

All subjects were asked to perform straight and rotational curl-ups. In this exercise, subjects have to consecutively flex and lift the head, shoulders, thoracicand lumbar spine and finally pelvis. Subjects were placed in a dorsal supine position, knees and hips flexed and feet fixed on the table. A higher score was given to a better curl-up performance following the criteria in Table 1." During the rotational curl-ups, subjects were asked to turn the upper body and bring the elbow to the contralateral knee. Left and right rotation were tested. Sit-ups with a straight upper body were avoided. Evaluations were done by a physician other than the surgeons.

The CybexR II isokinetic dynamometer (Cybex, Division of Lumex, Inc., Ronkonkoma, N.Y.) with stabilization system for trunk muscle force assessment was used to objectively evaluate the dynamic flexion, extension and rotation performance of the trunk. This apparatus has previously proven to give reliable measurements in a test/retest design.'2 In the Trunk Extension-Flexion (TEF) unit, all subjects were positioned according to the trunk stabilization system.'3'4 Measurements were done in a standing position. Knees were positioned in 15° of flexion to avoid hamstring strain when performing trunk flexion. The input axis of the dynamometer's shaft was aligned at the L5-S1 articulation. A testing trial consisted of five consecutive flexion/extension repetitions from 0° to 80° at a speed of 60°/s. In the Torso-Rotation (TR) unit, subjects were tested in 4 sitting position (Fig. 1). Knees and hips were flexed in 90° and hips were additionally abducted 10° to improve stability of the pelvis. Otherwise, fixation was similar to the TEF unit. Here a testing trial consisted of five consecutive trunk rotations from 45° right to 45° left and back, at a speed of 60°/s. The resting time between both trials was approximately 5 minutes. Before testing, each subject received detailed instructions and, once positioned, an opportunity was given to warm up and familiarize with the testing motion. During testing each subject was strongly encouraged by the physical therapist conducting the measurements to maximize her efforts.'5 Peak torque/body weight (pt/bw) and average power/ body weight (ap/bw) provided by the Cybex data reduction computer were the parameters used in this study to compare groups while avoiding the influence of weight changes.

Figure 1—Positioning of the tested subject in the torso-rotation unit of the Cybex II isokinetic dynamometer.

A self-administered questionnaire was sent to all TRAM flap patients to assess their opinion about abdominal strength and complaints, posture, lower back pain and changes in activities of daily life. Details of the questionnaire are published elsewhere.'6

To compare the dynamometric results between groups the Mann-Whitney test was used. To compare the curl-up scores, Fisher's exact test was used. P < 0.05 was considered as statistically significant.

Results

Clinical examination

Three (15%) patients had umbilical asymmetry and an additional four (20%) patients had abdominal wall asymmetry. A: bulging of the lower abdominal wall was found in two (10%) patients and a hernia in one case. Three persons in the control group were found to have a small symptom-free umbilical hernia.

Physical exercise evaluation

A distinct difference in curl-up performance was recorded for straight flexion and rotatory flexion to the left and right in both groups (Fig. 2). Sixteen controls were able to attain a straight curl-up score of 5. Only seven TRAM flap patients were able to obtain the same score. Most patients achieved a maximum score of 3 (5 patients) or 4 (8 patients) for straight curl-ups, significantly more than the number of controls with a maximum score of 3 or 4 (P = 0.001). Even fewer patients were able to perform a score 5 rotational curl-up (4 left, 5 right). Most patients were only able to lift one scapula from the surface (9 patients). The number of patients with a maximum score of 3 or 4 was significantly greater than the number of controls (left P = 0.0004; right P = 0.001).

Figure 2 - Physical examination. Percentages of subjects able to achieve muscle power scores of 3,4 or s. (Straight: Straight curl-up. Left: Left rotational curl-up. Right: Right rotational curl-up.)

Table 2 Isokinetic dynamometry results of TRAM flap patients vs. control group for different movements of the torso at 60°/s and their statistical significance (pt/bw = peak torque per body weight. ap/bw = average power per body weight)

Isokinetic dynamometry

Lower mean pt/bw and ap/bw were observed in the TRAM flap group for extension of the torso but the differences from the control group were not statistically significant (Table 2). The mean pt/bw and ap/bw to flex the upper body were significantly lower for TRAM flap patients compared to the controls. The mean flexion/extension ratios were not significantly different. Left rotation and to a lesser degree right rotation were significantly less powerful in patients than in controls. The weaker rotation to the left was once again demonstrated by the lower mean left/right rotation ratios for TRAM flap patients but statistical significance was considered to be borderline.

In TRAM (Left) and TRAM (Right) patients no significant changes (Mann-Whitney test) were found in the capacity to rotate the upper body to the left or to the right at a speed of 60°/s (Table 3).

CT or MRI of the abdominal wall

Only 6 TRAM flap patients presented themselves for the imaging study. The postoperative follow-up in these 6 patients varied from 19 to 54 months. The rectus muscle had been replaced by a mixed sheet of fascia, mesh and/or scar tissue. The midline was deviated to the operated side and the distance between the midline and the insertion line of the oblique muscles was reduced. The medial border of the external oblique muscle and to a lesser degree the medial border of the internal oblique had shifted medially. Finally, we observed a reduction in diameter of all oblique muscles of about 25% compared to the non-operated side.

Questionnaire

Nineteen of the 20 TRAM flap patients filled out and returned the questionnaire. The results are given in detail in a second paper in this issue of the Journal.'6 In summary, the patients complained about-decreased abdominal power (44%), reduced ability to lift heavy objects (28%), abdominal protrusion (42%), pain in the lower abdominal wall when the intra-abdominal pressure was raised (e.g. coughing, Valsalva manoeuvre) (47%) and difficulties in getting up from a supine position (38%).

One of 14 patients in employment before surgery was not able to continue her job as a housekeeper. Five patients (26%) had to alter the way they did domestic activities. Three (30%) of 10 patients were no longer able to carry on with their favorite preoperative sport. Two others (20%) hac1 to change to a different sport. Two (20%) out of 10 patients with an active hobby preoperatively (gardening), were not able to continue with their gardening.

Discussion

The normal function of the abdominal muscles

The role of the rectus abdominis muscles in flexion of the trunk is generally overestimated. Although they primarily flex the lumbar spine, they are only responsible for the first 30° of flexion of the upper body as an initiator of movement. The iliopsoas muscles then take over and are the strongest flexors responsible for trunk flexion over the largest part of the trajectory. In daily life the rectus muscles are hardly ever used as pure flexors because in an upright position gravity flexes the upper body. The flexing function of the rectus muscles is mostly needed to get up from a supine position, which is done maybe once or twice a day. The rectus abdominis muscles are far more important in daily life for stabilization of the upper body. Not only do they form a dynamic and flexible muscular pillar as a counterpart of the rigid bony spine but they are also an important site of insertion for all the oblique muscles. In this way they assist in rotatory movements and are essential for normal function of the oblique muscles. Together with the transverse muscles they are responsible for raising intra-abdominal pressure, a crucial function for lifting heavy objects, bowel movements, forced expiration (coughing, sneezing), etc.

The vertically orientated muscle fibers of both the internal and external oblique muscles assist in flexing the trunk, synergistic with the rectus muscles. The oblique muscle fibers are mainly responsible for lateral flexion and rotation of the trunk. Unilateral contraction of the external oblique muscle causes rotation to the contralateral side supported by the contralateral internal oblique. The oblique muscles are the strongest rotators of the trunk. The oblique part of the para-vertebral muscles and shoulder musculature only assist in the rotational movement.

Table 3 The mean peak torque and average power of left rotation compared to right rotation in the TRAM (Left) and TRAM (Right) groups at 60°/s (pt/bw = peak torque per body weight, ap/bw = average power per body weight). No differences were statistically significant

Evaluation of the abdominal muscles

The complex interaction of the recti with their surrounding muscles, specifically the oblique abdominal muscles, and the influence of synergistic muscles, e.g. the iliopsoas muscles, and of antagonists, e.g. the hamstrings and paravertebral muscles, make an isolated clinical evaluation of the function and strength of the rectus abdominis muscles very difficult. All reported clinical tests measure the capacity to flex the upper body by joint contraction of the rectus abdominis muscles, iliopsoas muscles and the vertical fibers of the oblique muscles but do not measure rectus muscle activity independently. A reduced flexion capacity can be caused by damage to, or loss of strength of, one or more of these flexing muscles. The specificity of these tests for the rectus muscles can be increased by reducing the synergistic influence of the other muscles, e.g. by flexing the hips during curl-ups. Even so, results have to be interpreted carefully.

1. Clinical examination. Although an asymmetric position of the umbilicus is probably only an aesthetic problem, protrusion of the abdominal wall can become hazardous to abdominal wall function in the long term. Abdominal asymmetry, bulging and hernias are probably three stages in a problem with the same aetiology, which is an increasing laxity and elasticity of the deep abdominal fascia and/or scar tissue. This laxity can either not occur or not be apparent, or remain stable or progress into a hernia over time. Long-term studies will be necessary to prove if the incidence of hernias will increase in these patients.

2. Curl-up performance. A substantial difference has to be made between a 'sit-up’ and a 'curl-up'. In a 'sit-up’ the spine is lifted from the surface but kept straight while the hips are flexed. Flexion is mainly performed by the iliopsoas muscles, while the rectus muscles have an isometric stabilizing function. In the first phase of a 'curl-up', the cervical spine followed by the thoracic spine is lifted up from the surface by contraction of the rectus muscles and the flexing part of the oblique muscles, flexing the lumbar spine for about 30°. Further flexion by rectus muscle activity then becomes impossible as shortening of the rectus muscle has reached its maximum. In the second phase of a 'curlup', that is over 30° to 45°, the iliopsoas muscles take over. The static or stabilizing function of the rectus muscles then slowly decreases as flexion increases."

To test the rectus muscle as a pure flexor while excluding any influence of the iliopsoas muscles, the first phase of the curl-up would probably be the best exercise because at that point the iliopsoas muscles are inactive. But the range of motion of 0° to 30° is limited, making it very difficult to grade performances. In other concentric flexing tests such as a sit-up or a full curl-up, the influence of the contraction of the iliopsoas muscles cannot be avoided. The rectus muscles then act as strong stabilizers during flexion of the hips to keep the upper body in an isometric position. If the rectus muscles were not active and the trunk could not be stabilized, no sit-up or curl-up could be performed because the iliopsoas muscles would only be tilting the pelvis and lower lumbar spine, leaving the upper body immobile.

We preferred using the 'full curl-tip' to clinically test the flexing capacity because the rectus muscles are evaluated in this test as flexors in the first part and stabilizers in the second part. In this test, the influence of the iliopsoas muscles can be reduced by flexing the hips and knees, causing shortening of the iliopsoas muscles and reduced lordosis of the lumbar spine. If additionally the feet are not supported, the test becomes an even more selective test for the rectus abdominis muscles." This is because the ability to perform flexion of the trunk is then directly correlated to the degree of curling up of the upper body and in a consecutive phase stabilizing the upper body maximally. The curling up brings the point of gravity of the upper body as short as possible to the pivot point (the hip joint), thereby reducing the torque necessary to flex the upper body. On the other hand, not having foot support decreases the sensitivity of this test too much. Looking at the general female population, it has been reported that only very few non-operated women are able to perform such a curl-up because the overall strength of their rectus muscles is too low to execute phase one.49 If the majority of persons is not able to do the curl-up preoperatively, it will become very difficult to grade a decrease in strength postoperatively. For that reason we preferred to support the feet.

If the hips are flexed and the feet are supported, almost every woman can perform a curl-up. A decline in this performance reflects either a weakening of the iliopsoas muscles, or reduced rectus strength, or both. The condition of the undamaged iliopsoas muscles can only be affected by an important change in the patient's general condition. We considered both groups to have a comparable general condition because no major differences were found in extension strength during the isokinetic dynamometry testing and back muscles are the first to weaken rapidly if a person's general condition declines. Therefore a decline of curl-up performance with feet supported largely depends on reduced rectus muscle strength. As the rectus muscle is the only one of the flexor group that has been surgically damaged, we may presume that the decreased abdominal wall strength is caused by resection of a part of the rectus muscle. We found that only 35% of TRAM flap patients were able to reach a score 5 whereas 80% did so in the control group. As 75% of TRAM flap patients, compared to 95% in the control group, were able to at least perform a score 4 curl-up, we can deduce that the flexing movement itself still can be executed by contraction of the intact contralateral rectus abdominis muscle and the oblique muscles, but with increasing work load (score 5) performances decrease drastically (Fig. 2).

The rotational curl-up is a very specific exercise for assessing the function of the ipsilateral external oblique muscle, laterally flexing and rotating the vertebrae. Almost half of the TRAM flap patients were not able to complete a full rotational curl-up and only 20% were able to execute the complete exercise (score 5). The poor performances (score 3 in 45% of TRAM flap patients) indicate important damage to the external oblique muscles, although they are not directly surgically injured during surgery. Whereas in a straight curl-up the function of a unilaterally damaged rectus abdominis muscle is compensated by strong synergistic muscles, the function of the apparently undamaged external oblique muscle, dominant in rotational movements, can only be partially compensated in a rotational curl-up by weak synergists. This is why a large number of patients was not able to execute the complete exercise, regardless of the work load, and why differences with the control group are even more significant for the rotatory movements, especially for score 5.

3. Isokinetic dynamometry. This is a method to measure the force or power exerted over a joint by a group of muscles at a fixed angular speed.'8 The testing device will automatically keep the speed constant by increasing the resistance to the movement if the tested person increases their efforts. The calculated peak torque (= force x distance, Nm) represents the maximum muscle capacity while the average power (= work/time, Watts) represents the total amount of work a muscle is able to perform over a certain amount of time. Because movements of the upper body were made in the same direction as the vector of gravity, all results were corrected for the patient's weight. During testing, feet, hips and knees have to be fixed to be able to measure forges in a constant and reproducible manner. This means that the function of the iliopsoas muscles, as in curl-up testing, can never be excluded and results will always be influenced by both iliopsoas and rectus muscles. Isolated testing of the rectus muscles is impossible with this device. We preferred to test our patients in a standing position because this is considered to be the most functional posture, best resembling daily living activities and providing sufficient amplitude for movement of the upper body.

Because the general condition of both groups was comparable, we once again could assume that the strength of the undamaged iliopsoas muscles was similar in both groups and that the statistically significant difference in flexion strength was due to the surgical resection of a part of the rectus abdominis muscle. Differences became even more apparent looking at the average power. The flexion/extension ratio values were lower in TRAM flap patients but did not differ significantly from the control subjects. It is unclear at this moment whether this could be an indicator that the balance between abdominal and back musculature was disturbed and whether this can be correlated with a higher number of patients with postoperative lower back pain as reported in the questionnaire. Despite the statistically significant reduced rotational strength measured in both directions in TRAM flap patients, rotation to the left was slightly weaker than to the right. This difference, represented by a decreased left/right rotation ratio, was not statistically significant and remains unexplained. The side of muscle harvest did not influence the rotational strength in a particular direction. Right rotation was slightly stronger in both the TRAM (Left) and TRAM (Right) groups but differences were not statistically significant.

Implications for abdominal wall function and strength

The resection of a part of one rectus abdominis muscle leads to an interruption of the rectus muscle continuity and thus to the complete loss of function of one of the rectus abdominis muscles. Theoretically, the strength of the central muscular pillar of the abdomen is therefore reduced by 50%. By compensatory mechanisms of the synergistic muscle groups, the functional loss is limited but a statistically significant decrease in flexing strength is unavoidable and permanent. The fact that only a part or the whole muscle is harvested does not seem to make a difference.4

Additionally, the resection of a part of the rectus abdominis muscle causes major functional changes in the oblique muscles. The loss of strength of the oblique muscles can only be attributed to the iatrogenic alterations that took place at its insertion line, because vascularisation and innervation have been left intact. As seen intraoperatively and on postoperative imaging, the contralateral rectus muscle (in case of a unilateral free TRAM) is pulled over the midline by the tension of the partially resected and sutured anterior rectus fascia. The decreased width of the ipsilateral rectus fascia and the decreased distance between the insertion lines of the oblique muscles of both sides causes overstretching of the external and internal oblique muscles on the operated and non-operated side. In the early postoperative period, this can lead to an increased resting length of the sarcomeres. As shown in Brand's'9 sarcomere tension-length curve of a skeletal muscle (Fig. 3), this could result in a reduction of the tension capacity. Extreme tension and overstretching can even lead to areas of permanent fibrosis. In an intermediate phase, the muscles, held in a 1engthened position, can add sarcomeres to increase muscle fiber length, finally restoring the normal resting length. But in the meantime, the elasticity of the mixed meshscar-aponeurosis layer that replaced the rectus muscle and fascia may slowly increase. This can be clinically observed with the different degrees of laxity of the abdominal wall, ranging from abdominal asymmetry to bulging. Increased elasticity is difficult to prove on CT scans or MRI as patients are lying in a relaxed supine position, reducing the tension on the abdominal wall. The pulling of a muscle with increased fiber length on semi-elastic tissue can be responsible for an important loss of strength of the oblique muscles.

Besides these physiological changes in the oblique muscles, the biomechanical changes taking place at the interface of the oblique muscles with the damaged rectus muscles are maybe even more important. Each contraction of the rectus abdominis muscle increases the muscle's diameter temporarily. The tension on, and the medial displacement of, the insertion line of the oblique muscles is necessary for proper function of the oblique muscles. This mechanism is destroyed on the ipsilateral side and is damaged on the contralateral side due to the high passive tension on the anterior and eventually posterior rectus fascia. Reefing of the contralateral anterior rectus fascia to obtain symmetry of the umbilical position will probably only further deteriorate this function.

Figure 3 - Clinical length-tension curve of a skeletal muscle. R1 is the resting length of a normal muscle. R2 is the resting length of a stretched muscle (modified from Brand19).

Additionally, the weakening of the central muscular pillar after partial muscle resection will reduce its dynamic rigidity. In non-operated subjects, a rigid and dynamic central muscular pillar is necessary for the oblique muscles to be able to exert their forces without laxity at the insertion line. If the oblique muscles do not receive sufficient counter-action from a weakened central pillar containing only the contralateral rectus muscle (unilateral TRAM) or no muscle at all (bilateral TRAM), the insertion line will be deflected, resulting in an improper function of the oblique muscles and loss of strength.

These physiological and biomechanical changes can express themselves in a variable degree in each patient. After surgery, 35% of TRAM flap patients were still able to perform a full straight curl-up and 25% a full rotational curl-up. On the other hand, some patients were only able to partially initiate the movement and suffered an important loss of strength. In most patients synergists can take over the functional movement but as the load increases (e.g. when hands are brought to the neck or higher) flexion and rotation performances decrease drastically. Re-examining the questionnaire taught us that patients with the lowest curl-up and dynamometry score had the most complaints in daily living activities. They specifically complained about decreased abdominal power (44%), reduced ability to lift heavy objects (28%), abdominal, protrusion (42%), abdominal pain when the intra-abdominal pressure was raised (47%) and difficulties in getting up from a supine position (38%). Although these complaints were seldom incapacitating, adaptations of activities of daily life were noticed in 31% for professional activities, 30% for sports and 20% for hobbies.

Acknowledgements

We wish to thank Dr K. Depuydt, Mrs L. Vervaet and Mrs 1. Didden for their efforts in physically and clinically evaluating the patients. We also wish to express our gratitude to Mr G Vanmaele for his statistical work.

References

1. Duchateau J, Declety A, Lejour M. Innervation of the rectus abdominis muscle: implications for rectus flaps. Plast Reconstr Surg 1988; 82: 223-7.

2. Hammond DC, Larson DL, Severinac RN, Marcias M. Rectus abdominis muscle innervation: implications for TRAM flap elevation. Plast Reconstr Surg 1995; 96: 105-10.

3. Feller AM. Free TRAM, results and abdominal wall function. Clin Plast Surg 1994; 1: 223-30.

4. Mizgala CL, Hartrampf CR, Bennett GK. Abdominal function after pedicled TRAM flap surgery. Clin Plast Surg 1994; 21: 255-72.

5. Suominen S, Asko-Seljavaara S, von Smitten K, Ahovuo J, Sainio P, Alaranta H. Sequelae in the abdominal wall after pedicled or free TRAM flap surgery. Ann Plast Surg 1996; 36: 629-36.

6. Galli A, Adami M, Berrino P, Leone S, Santi R Long-term evaluation of the abdominal wall competence after total and selective harvesting of the rectus abdominis muscle. Ann Plast Surg 1992; 28: 409-13.

7. Kroll SS, Schusterman MA, Reece GP, Miller MJ, Robb G, Evans G. Abdominal wall strength, bulging, and hernia after TRAM flap breast reconstruction. Plast Reconstr Surg 1995; 96:616 19.

8. Zauner-Dungl A, Resch KL, Herczeg E, Piza-Katzer H. Quantification of functional deficits associated with rectus abdominis muscle flaps. Plast Reconstr Surg 1995; 96: 1623-8.

9. Lejour M, Dome M. Abdominal wall function after rectus abdominis transfer. Plast Reconstr Surg 1991; 87: 1054-68.

10. Mizgala CL, Hartrampf CR, Bennett GK. Assessment of the abdominal wall after pedicled TRAM flap surgery: 5- to 7-year follow-up of 150 consecutive patients. Plast Reconstr Surg 1994; 93: 988-1002.

11. Janda V. Muskelfunktionsdiagnostik. 2nd ed. Berlin: yolk und Gesundheit, 1986.

12. Thompson NN. A reliability analysis of the Cybex II dynamometer and trunk stabilization system. Masters Thesis, University of Wisconsin, La Crosse, 1984. In: Davies GJ, ed. A compendium of isokinetics in clinical usage. 3rd ea., Onalaska Wisconsm: S&S Publishers, 1987: 307 (abstract).

13. Davies GJ, Gould JA. Trunk testing using a prototype Cybex II Isokinetic dynamometer stabilization system. J Orthop Sports Phys Ther 1982; 3: 164-70.

14. Thompson NN, Gould JA, Davies GJ, Ross DE, Price S. Descriptive measures of isokinetic trunk testing. ~ Orthop Sports Phys Ther 1985; 7: 43-9.

15. Matheson L, Mooney V, Caiozzo V et al. Effect of instructions on isokinetic trunk strength testing variability, reliability, absolute value and predictive validity Spine 1992; 17: 914 21.

16. Blondeel Ph, Vanderstraeten GG Monstrey SJ, et al. The donor site morbidity of free DIEP fiaps and free TRAM flaps for breast reconstruction. Br J Plast Surg 1997; 50: 322-30.

17. Dikkeboer B, Jorritsma W. Kinesiologie. Bonn: Scheltema & Holkema, 1983.

18. Newton M, Waddell G Trunk strength testing with isomachines, Part 1: Review of a decade of scientific evidence. Spine 1993; 18: 801-11.

19. Brand PW. Clinical mechanics of the hand. St. Louis: Mosby, 1985.

The Authors

PhiDip N. Blondeel MD, FCCP, Associate Professor, Department of Plastic and Reconstructive Surgery, University Hospital Gent

Willy D. Boeekx MD, PhD, Professor, Department of Traumatology and Reconstructive Surgery, University Hospitals Leuven

Guy G Vanderstraeten MD, PhD, Professor and Chief of the Department of Physical Medicine and Rehabilitation, University Hospital Gent

Roeland Lysens MD, PhD, Professor and Chief of the Department of Physical Medicine and Rehabilitation, University Hospitals Leuven

Koenread Van Landoyt MD, FCCP, Associate Professor, Department of Plastic and Reconstructive Surgery, University Hospital Gent

Patriek Tonnard MD, FCCP, Clinical Assistant Professor, Department of Plastic and Reconstructive Surgery, University Hospital Gent

Stan J. Monstrey MD, PhD, FCCP, Professor and Chief of the Department of Plastic and Reconstructive Surgery, University Hospital Gent

Guido Matton MD, FACS, Emeritus Professor and former Chief of the Department of Plastic and Reconstructive Surgery, University Hospital Gent.

Correspondence to Phillip N. Blondeel, Department of Plastic and Reconstructive Surgery, University Hospital Gent, De Pintelaan 185, B-9000 Gent, Belgium.

Paper received 8 November 1996.
Accepted 19 March 1997, after revision.