INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

MUSCLE TRANSFER OPERATIONS are performed frequently to replace skeletal muscles that have become diseased or damaged (13). Although numerous studies have documented the recovery of muscles transferred in the presence of intact synergistic muscles (8, 18, 20, 24), the replacement of an entire muscle group with a muscle transferred in the absence of synergistic muscles is more relevant clinically. Muscle transfers of the latter type have been used in human beings to reestablish function in the upper and lower extremities (3, 9, 22) but have been investigated on a limited basis in animal models (10). This study in rats documented the recovery of nonvascularized soleus muscle transfers replacing the entire plantar flexor muscle group, which, given the mere 6% of the mass of the plantar flexor group that the soleus muscle represents (27), would not be clinically the muscle of choice for this site.

When muscles are present that are synergistic to the muscles transferred, muscle mass, total fiber cross-sectional area (CSA), and maximum isometric tetanic force (P_o) decrease 25-40% (11). Conversely, these variables increase for intact muscles following the ablation of their synergists (27, 28), with the magnitude of these increases being proportional to the relative load on the muscle (27). In the case of muscles transferred in the absence of synergistic muscles, a combination of these two situations exists. Consequently, the first objective of this study was to determine how the recovery of muscles transferred in the absence of their synergists is affected by these conflicting stimuli for adaptation. Through combinations of the lateral gastrocnemius (LGN), medial gastrocnemius (MGN), and plantaris (PLN) muscles in rats, we composed four different muscle transfers to replace the total plantar flexor muscle group. These transfers represented ~20, 40, 60, and 80% of the mass of the original plantar flexor group (27). We hypothesized that, compared with their individual control values for muscle mass, total fiber CSA, and P_o, the four transfers would display a hierarchical array of deficits, proportional to their initial mass and, consequently, inversely proportional to the relative load on the transfers.

Whereas transfer procedures may provide a partial restoration of function at the recipient site, deficits are introduced at the site from which the donor muscle is removed. Previously, a relationship between the percentage of the mass of the muscle group removed from the donor site and the percentage of the P_o restored at the donor site has been proposed for the plantar flexor muscle group in rats (25). Therefore, the second objective of this study was to quantify the compromise between the restoration of function at the recipient site and at the donor site after the transfer procedures. In particular, we sought to determine the transfer mass that would result in an equal recovery of control P_o at the donor and recipient sites.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Muscle transfer operations were performed bilaterally in 40 male Fischer 344 rats that were 11-12 mo of age. Sham operations were performed bilaterally in an additional group of 28 rats, which served as control animals. The structural and functional properties of the muscle transfers and control muscles were evaluated 120 days after the operations, a time point by which transfers in rats are known to have stabilized (24). Rats had body masses ranging from 363 to 497 g and were 15-16 mo of age at the time of data collection. For the transfer operations and muscle function experiments, the animals were anesthetized with an initial intraperitoneal injection of pentobarbital sodium (55 mg/kg). Supplemental doses were administered as required to maintain a depth of anesthesia that prevented a response to tactile stimuli. All operations, postoperative care, and experimental procedures were conducted in accordance with the National Institute of Health "Guide for the Care and Use of Laboratory Animals" [DHEW Publication No. (NIH) 86-23, revised 1985, Office of Science and Health Reports, DRR/NIH, Bethesda, MD 20205].

All operations were performed under aseptic conditions. For the muscle transfer operations, the entire plantar flexor muscle group was excised from each hindlimb by severing the proximal and distal tendons of each component muscle and by sectioning the tibial nerve at a point proximal to its divergence into individual muscular branches. The vasculature remained intact for those plantar flexor muscles used as transfers in the site (5, 18), whereas blood vessels were ligated for those muscles that were removed permanently. Muscles were transferred orthotopically within the plantar flexor site in one of four experimental groups: 1) the LGN and MGN muscles (LGN+MGN), 2) the LGN and PLN muscles (LGN+PLN), 3) the MGN muscle, or 4) the PLN muscle. These groups represent 78, 58, 36, and 16%, respectively, of the mass of the total plantar flexor muscle group (27).

Muscles were sutured to their original proximal tendon stumps (7-0 ethilon), and the tibial nerve was repaired (11-0 ethilon). On the distal end, the PLN and Achilles tendons were sutured together for each transfer to preserve the attachments of those muscles that were removed. In the control animals, sham operations consisted of severing the fascia and overlying muscles in each hindlimb, then separating the individual plantar flexor muscles without severing the tendons or nerve. For all operations, fascia and skin were sutured and closed separately (4-0 ethilon), and Betadine was applied to the incisions after closing. From daily observations of the behavior of experimental and sham-operated control rats, the operative procedure did not cause a significant impairment in mobility after the first 14 days.

Functional evaluations of the muscle transfers and control muscles were conducted in vivo by using a modified version of the apparatus described by Ashton-Miller et al. (2). The apparatus consists of a torque transducer (model QWFK-8M, Sensotec, Columbus, OH), which measures the torque developed by the muscle or muscles about the ankle joint. The torque transducer is attached to the shaft of a custom-built servomotor that is mounted rigidly to a Plexiglas platform such that the torque transducer is oriented vertically. The foot was placed in an aluminum shoe device attached to the torque transducer and secured in the shoe by wires covered with Silastic tubing. The knee was held stationary with a screw clamp at the femoral condyle, and care was taken not to compromise the circulation of the hindlimb. Body temperature was maintained at ~37°C with a heating pad, and injections of saline were administered subcutaneously throughout the experiment to prevent dehydration.

Muscle function experiments were controlled through LabVIEW software (version 2.2.1, National Instruments, Austin, TX) run on a Macintosh computer (model IIVX, Apple Computer, Cupertino, CA). To stimulate muscles in the plantar flexor site, the tibial nerve was isolated through an incision in the lateral surface of the hindlimb. The peroneal and sural nerves were severed to avoid unwanted contraction. Stainless steel electrodes were placed on either side of the nerve, and the muscles were activated by stimulation with 0.2-ms square-wave pulses. With the ankle positioned at 90°, single-stimulus pulses were administered, and the voltage was adjusted until the maximum isometric twitch force was observed. P_o was determined by stimulating the muscles with pulse trains of 300-ms duration at increasing frequencies of stimulation. P_o was obtained from the plateau of the frequency-force relationship and was usually observed between 100 and 120 Hz. Force measurements for individual control muscles were obtained by severing the distal tendons of their synergistic muscles at the time of evaluation.

The isometric force was recorded at different ankle angles throughout the range of motion to determine the optimum length for force development (L_o). After the muscle was positioned at the desired angle, three successive twitches were evoked at 1-s intervals, similarly to the method of Woittiez et al. (33). Three seconds after the third twitch, the passive force was measured. Two seconds after the passive force measurement, a tetanic contraction was evoked and the total force was recorded. The passive force was then subtracted from the total force to yield the active muscle force for the contraction. Muscles were allowed to recover at 90° between measurements taken at ankle angles of <90°, since muscles might be lengthened with respect to L_o in these positions. A period of 1 min was taken between contractions to avoid muscle fatigue.

Data were sampled at 5 kHz and low-pass filtered (cut off frequency = 200 Hz) before storage. The torque recorded by the transducer results from the product of the force developed by the muscle and its moment arm, which is a nonlinear function of joint angle. The relationship between muscle moment arm and ankle angle for the plantar flexor muscle group was determined by using the method described by Miller and Dennis (23), where the moment arm is calculated from the torque produced by the application of a known force along the muscle line of action. The variation of the moment arm (r) with ankle angle (a) was fit to a parametric model of the form r = R sin(a + ), where R is the maximum moment arm and  is the angle at which the maximum occurs as offset from 90°. Values of R = 5.21 mm and  = 6.1° were determined for the plantar flexor muscle group, and through application of the parametric equation the torque recorded was converted into units of force.

At the completion of each experiment, L_o was measured at the ankle angle that corresponded with maximum force development. Muscles were then excised, tendons were trimmed, and the muscles were blotted and weighed. Muscle fiber length (L_f) was calculated by multiplication with predetermined L_f /L_o ratios of 0.43 for the LGN muscle (33), 0.37 for the MGN muscle (27, 33), and 0.37 for the PLN muscle (28, 33), with the assumption that these ratios were not altered as a result of the transfer procedure (5, 17, 18). The total fiber CSA was calculated by dividing muscle mass by the product of L_f (mm) and 1.06 mg/mm³, which is the density of mammalian skeletal muscle. Values of muscle mass and total fiber CSA for individual muscles were added to produce values for the LGN+MGN and LGN+PLN controls and transfers as well as for the total plantar flexor group. Specific P_o was calculated by dividing P_o by total fiber CSA. The animals were euthanized by an intravenous injection of potassium chloride.

Muscles were frozen in cooled isopentane, sectioned at 10 µm in a cryostat, and stained with hematoxylin and eosin. Single-fiber CSAs were measured for 200 fibers from each of the proximal, middle, and distal sections (total = 600 fibers) of every muscle by using an image analyzer and Bioquant software. For each muscle, the frequency distribution of single-fiber CSAs as well as a mean single-fiber CSA were determined. The single-fiber CSA was then divided into the total fiber CSA to provide an estimate of the number of fibers present in each muscle. The mean single-fiber CSAs of individual muscles were weighted by the fiber number for that muscle, then added to yield average values of single-fiber CSA for the LGN+MGN and LGN+PLN controls and transfers. Values of fiber number for these four groups were determined by simple addition.

Data are means ± SE. To determine statistical significance between values for muscle transfers and their corresponding control muscles, an analysis of covariance test was administered, with body mass as the independent variable. Post hoc analyses were made by using the Bonferroni correction for multiple comparisons. The level of significance was set a priori at  = 0.05.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Control values for the total plantar flexor muscle group (n = 8) were 2,294 ± 32 mg for muscle mass, 164 ± 2 mm² for total fiber CSA, and 38.1 ± 0.8 N for P_o. Control values of these variables for the LGN+MGN, LGN+PLN, MGN, and PLN groups represented ~76, 59, 40, and 19% of the value for the total plantar flexor group, in excellent agreement with data from Roy et al. (27). Compared with their individual control values, the muscle mass, total fiber CSA, and P_o were each decreased for the LGN+MGN, LGN+PLN, and MGN transfers, whereas no difference existed between the PLN transfers and their control values for these variables (Table 1). The specific P_o of each muscle transfer did not differ from that of its individual control value. The variation in control values of specific P_o among the different muscles may be a result of errors in the estimation of total fiber CSA arising from the assumption of a constant L_f/L_o ratio for all muscle fibers, rather than the existence of a distribution of fiber lengths.

Values for the L_o of the control LGN, MGN, and PLN muscles were 33.0 ± 0.3, 33.3 ± 0.4, and 33.9 ± 0.5 mm, respectively. The values of L_o measured for these muscles in each of their two transfer groups were compared, and no differences were found; consequently, values were pooled for each muscle. When the LGN muscle was transferred, the L_o was 31.9 ± 0.3 mm and when the PLN muscle was transferred, the L_o was 34.5 ± 0.3 mm. Each of these values was different from the control value. In contrast, no difference was found between the L_o of 32.7 ± 0.2 mm for the MGN muscle after its transfer and its control value.

At the level of the single muscle fiber, frequency distributions of single-fiber CSAs were compiled for the LGN, MGN, and PLN control muscles (Fig. 1, A-C). Because individual LGN, MGN, and PLN muscles were each used in two different transfer groups, the resulting bin values were compared for each muscle. No differences were found; consequently, the data were combined to produce a single distribution for each muscle after its transfer (Fig. 1, A-C). Compared with their individual control values, the mean single-fiber CSA was decreased for each transfer (Table 2). Each muscle transfer recovered its respective control value for the total number of fibers, with the exception of the LGN+PLN transfers for which the total fiber number was decreased slightly.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Frequency distributions of single-fiber cross-sectional areas (CSAs) for lateral gastrocnemius (A; LGN), medial gastrocnemius (B; MGN), and plantaris (C; PLN) control muscles and muscle transfers. Each distribution was compiled from the single-fiber CSAs of 6,000 muscle fibers.

When expressed as a percentage of their individual control values for muscle mass, total fiber CSA, and P_o, the LGN+MGN, LGN+PLN, and MGN transfers recovered equally well, displaying deficits of 30-40%, whereas, by comparison, the recovery of the PLN transfer to 100% of its control value for these variables was substantially greater. For each of the six structural and functional variables measured, the percentage of the control value attained by the muscle transfers from the present study is plotted (Fig. 2A), as is the recovery for these variables documented for muscles transferred in the presence of synergistic muscles (Fig. 2B). Data for the latter were obtained from previous studies of rectus femoris muscle (RFM) transfers in rabbits (5, 16-18) and MGN muscle transfers in rats (24, 31).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Structural and functional properties, expressed as a percentage of control value, for muscles transferred in absence (A) and in presence (B) of synergistic muscles. In A, data are shown for 4 types of muscle transfers in present study: LGN+MGN, LGN+PLN, MGN, and PLN muscles. In B, data are presented from previous studies of rectus femoris muscle (RFM) transfers in rabbits (see Refs. 5, 16-18) and MGN muscle transfers in rats (see Refs. 24, 31).

For each ~20% increment in transfer mass, the P_o developed by the transfers was increased significantly. Values of P_o for the transfers, expressed as a percentage of the P_o of the total plantar flexor group, are shown as a function of the percentage of the mass of the plantar flexor group transferred (Fig. 3A). This relationship was fit with a second-order polynomial of the form f(x) = 0.004x² + 0.929x + 1.743. The inverse nature of the relationship for the donor site compared with the direct nature of the relationship for the recipient site is apparent, when the P_o values of the plantar flexor group restored in each site are plotted against the percentages of the mass of the group (25). The intersection of the two curves signifies the point of equal functional recovery between the donor and recipient sites. For the plantar flexor muscle group in rats, ~45% of control P_o is restored at each site when ~70% of the mass of the group is transferred.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   A: maximum force developed by each muscle transfer, expressed as a percentage of maximum force of plantar flexor muscle group, as a function of mass of plantar flexor group transferred. Data are means ± SE, and relationship is fit with a 2nd-order polynomial. B: percentage of control plantar flexor force restored at recipient site (from A) and percentage of control plantar flexor force restored at donor site (25) as a function of the mass of plantar flexor group transferred. Intersection of the 2 curves, designated by horizontal and vertical dotted lines, signifies point of equal functional recovery between the donor and recipient sites.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

For muscles transferred in the presence of synergistic muscles, deficits of 25-40% in muscle mass, total fiber CSA, and P_o have been documented (11). These deficits appear to be universal, since they occur in orthotopic transfers ranging from 1 g MGN muscles in rats (24, 31) to 80 g gracilis muscles in dogs (8, 20). The structural and functional deficits have been attributed in part to adaptations of muscles synergistic to the transfer, such as increases in mass (24) and changes in recruitment patterns (32), which may repress the need for transfers to recover fully. In the present study, we hypothesized that when muscles were transferred in the absence of synergistic muscles, a hierarchical array of deficits in structural and functional variables, proportional to the initial mass of the transfers and, consequently, inversely proportional to the relative load on the transfers, would result. This hypothesis was rejected. Apparently, with the exception of the small PLN transfers, each of the other transfers was not loaded sufficiently (12) to restore >60-70% of control mass, CSA, and force development. As a consequence, for muscle transfers with masses representing >20% of the mass of the original muscle group, the level of recovery was independent of the mass of muscle transferred and not different from muscles transferred in the presence of synergistic muscles (5, 17, 18, 20, 24, 31).

Given the nearly complete recovery of the PLN transfers in the present study, the deficits that occur in muscle transfers must depend on the functional environment into which the muscle is transferred. The restoration of control values for muscle mass, total fiber CSA, and P_o of the PLN transfers is similar to that reported for nonvascularized soleus muscle transfers in the rat (10), which represent 6% of the mass of the plantar flexor muscle group (27). The difference in the recovery between the PLN and soleus transfers and the LGN+MGN, LGN+PLN, and MGN transfers may lie in the average force requirements of the plantar flexor muscle group. At their preferred walking speed, cage-sedentary rats require only 33% of the P_o of the plantar flexor muscle group for normal locomotion (26). For muscles transferred in the absence of their synergists, this requirement must be met by one muscle, such that those transfers that represent <33% of the mass and force production of the plantar flexor group appear to have a stimulus to recover completely, whereas those that account for larger percentages of group structure and function might not receive an adequate stimulus to recover as fully. This force requirement may explain the abrupt transition between deficits occurring in the PLN and MGN transfers, which represent ~20 and ~40%, respectively, of the mass of the plantar flexor muscle group.

The specific P_o (kN/m²) permits evaluation of the functional recovery of muscle transfers, independent of the deficits in structural variables. The percentage of control-specific P_o restored in muscles transferred in the presence of their synergists ranges from 75% for the RFM muscle in rabbits (5, 18) to 100% for the MGN muscle of rats (24, 31). In the present study, values of specific P_o were not different between muscle transfers and their respective control muscles. Therefore, muscles in rats do not display a deficit in specific P_o whether transferred in the presence or absence of synergistic muscles. Restoration of control values of specific P_o is one indication of the success of the reinnervation process following neural repair. Compared with the smaller absolute masses of muscles in rats, the 9-g mass of the rabbit RFM muscle may prohibit the complete reinnervation of muscle fibers, producing a loss in force with respect to total fiber CSA.

Whereas the restoration of control values of specific P_o provides one indication of successful reinnervation, further confirmation that the muscle transfers in the present study were reinnervated completely is provided by the characteristics of the frequency distributions of single-fiber CSAs. In these distributions, the single-fiber CSAs appear to have decreased nearly uniformly after the transfer procedure, such that the distribution for each transferred muscle is shifted to the left, without a change in general shape, compared with its control muscle. In contrast, the distributions of single-fiber CSAs in rat muscles that have undergone long-term denervation show the development of a considerable population of muscle fibers with extremely small CSAs (30), thus changing the shape of the distribution. The absence of this characteristic in the frequency distributions for the plantar flexor muscle transfers suggests that each transfer was reinnervated completely and that the incomplete recovery of structure and function in the MGN, LGN+PLN, and LGN+MGN transfers, compared with the 100% recovery of the PLN transfers, cannot be ascribed to neural factors.

In the present study, each ~20% increment in the muscle mass transferred resulted in a significant increase in P_o at the recipient site, restoring a maximum of ~45% of original plantar flexor group function. Even though function at the recipient site continues to improve slightly, this benefit may not be worth the loss in P_o introduced at the donor site by the transfer procedure. The initial deficits at the donor site are not necessarily permanent, since the synergistic muscles that remain intact may hypertrophy in response to the relocation of the donor muscle (25). As would be expected, a greater percentage of control P_o is produced at the donor site compared with the recipient site for a given mass of muscle. For example, 52% of plantar flexor group P_o would be restored if a full 100% of the plantar flexor group were transferred into the recipient site, whereas muscles at the donor site produce 52% of plantar flexor group P_o when only 36% of the group remains (25).

Given the disparity between the donor and recipient sites in terms of the ability to regain control values of P_o, a transfer mass must be selected to provide the recovery necessary at each site. For instance, one may desire to balance the restoration of function at the recipient site with the restoration of function at the donor site. Assuming that the original muscle groups at the donor and recipient sites are of equal mass, ~45% of control P_o is restored at each site when ~70% of the mass of the plantar flexor group is transferred. Alternatively, the retention of force production may be more important biomechanically at one site compared with the other, such that a more complete restoration of control P_o can be provided at one site at the expense of a larger deficit at the other. If feasible, surgeons would prefer to select a donor site containing a larger muscle mass than that existing in the recipient site to obtain optimal results at both sites (21). Unfortunately, such a donor site usually does not exist, except for cases involving facial reanimation (19), where the muscle mass required is less than that for the upper and lower extremities.

Although informative, the donor site-recipient site relationship developed in the present study is limited in that it is formulated solely from data for the plantar flexor muscle group in rats. Because of the lesser adaptation of slow-twitch, compared with fast-twitch, muscles after an increase in load (27), the recovery of the predominantly fast-twitch muscles of rats (1) after their transfer may differ from the recovery of transfers in humans, for whom most muscles are of a mixed fiber-type composition (15). Certainly, the force requirements at specific donor and recipient sites are quite varied (14, 29), potentially influencing the recovery of the intact muscles and muscle transfers and altering the relationship between the two sites. In addition, exercise may improve the functional capabilities of intact muscles at the donor site and muscle transfers at the recipient site (12). For singly housed, cage-sedentary rats, only ~10% of each day is spent on locomotor activities (6). An increase in this percentage through use of a conditioning protocol, analogous to a rehabilitation program in humans, may reduce or even eliminate the structural and functional deficits observed in the single-muscle transfers. Alternatively, detrimental changes to the function at the donor and recipient sites may occur with age, since the function of control muscles (4) as well as muscle transfers (7, 31) is known to be decreased, compared with muscles from young animals.

In summary, we have shown that plantar flexor muscles transferred in the absence of synergistic muscles display deficits of 30-40% in structural and functional variables, which are of the same magnitude as those seen in muscles transferred in the presence of their synergists. The exception to this finding is the complete recovery of control values for the PLN transfers, which likely is a result of the average force requirements of the plantar flexor muscle group. In addition, we found that each ~20% increment in transfer mass resulted in a significant improvement in the plantar flexor group P_o at the recipient site. Based on this relationship, as well as on a previous model (25) for the P_o restored at the donor site after the removal of muscles, we determined a relationship for the simultaneous recovery of function at both the donor and the recipient sites. Alterations in muscle function due to the use of different donor and recipient sites, exercise, or aging require further investigation.