A biomechanical study of the flexor digitorum superficialis: Effects of digital pulley excision and loss of the flexor digitorum profundus

A Biomechanical Study of the Flexor Digitorum Superficialis: Effects of Digital Pulley Excision and Loss of the Flexor Digitorum Profundus Jeffrey Hamman, BA, Arif Ali, BA, Craig Phillips, MD, Benjamin Cunningham, MD, Daniel P. Mass, MD, Chicago, IL Many reports have been devoted to characterizing the significance of the pulleys for the flexor digitorum profundus (FDP). However, no comparable work has been published on the flexor digitorum superficialis (FDS). This study characterized the FDS in a human cadaver model. Eleven fresh-frozen cadaver hands were used. By using a tensiometer, data were gathered for tendon excursion, tendon load, and work of flexion. Changes in efficiency were caused by excision of annular pulleys A1, A2, A3, and the palmar aponeurotic pulley. We also measured the effect of FDP excision on FDS efficiency. Sectioning of the A2 and A3 pulleys together caused statistically significant losses of efficiency in all three parameters (work, load, and excursion). When the FDP was removed from a finger with an intact pulley system, losses in both work and excursion efficiencies were significant. Removing the FDP while cutting different pulleys caused significant decrease in FDS excursion efficiency. We conclude that A2 and A3 are the most important pulleys for maintaining normal FDS function, and that the presence of the FDP in the digital sheath is essential for optimal FDS excursion efficiency. (J Hand Surg 1997;22A:328-335.)

The flexor digitorum profundus (FDP) is the primary digital flexor during simple finger flexion, whereas the flexor digitorum superficialis (FDS) must be recruited to achieve a power grip. 1-3 The FDS is also needed for independent finger flexion. 4,5 Without the FDS, hyperextension of the proximal interphalangeal joint (PIP) will occur when forceful pinch is attempted. In time, the resting posture of

Fromthe Sectionof OrthopaedicSurgeryand RehabilitationMedicine, Departmentof Surgery,Universityof Chicago, Chicago,IL. Receivedfor publicationJune 9, 1995;acceptedin revisedform July 23, 1996. No benefits in any form have been or will be receivedfrom a commercialpartyrelateddirectlyor indirectlyto the subjectof this article. Reprint requests: Daniel P. Mass, MD, Section of Orthopaedic Surgery and RehabilitationMedicine, Departmentof Surgery,University of Chicago Hospitals, MC 6032, 5841 S. Maryland Avenue, Chicago, 1L60637. 328

The Journal of Hand Surgery

mobile fingers will change, often resulting in a hyperextension deformity of the PIP and a fixed flexion deformity of the distal interphalangeal joint (DIP). 3 Understanding the biomechanics of the FDS and the effects of sheath and FDP excision may allow the physician greater understanding of the extent of repair required in an injured finger to restore optimal FDS function. Although the FDS is a vital component of balanced finger flexion, few biomechanical descriptions of its function have been published. To our knowledge, no articles have addressed the importance of the digital sheath anatomy for optimum FDS function. Here, we describe how the FDS tendon transmits a physiologic pinch force within the context of the normal and pathologic anatomy of the digital sheath. The fibro-osseous annular pulleys keep the flexor tendons close to the axes of the finger joints during

The Journal of Hand Surgery/Vol. 22A No. 2 March 1997

flexion, thus forfeiting mechanical advantage to maximize range of motion (ROM). 3 Various studies have shown that the digital pulley system is essential for optimal FDP function. 6-17 Many of the investigators used tendon excursion, tendon bowstringing, or range of joint motion as parameters for comparing the function of fingers before and after various portions of the digital sheath were excised. 6-12 Other authors added other parameters such as tension in the FDP (load) throughout flexion 13-15 and FDP load at full flexion. 16All of these parameters are valuable barometers of tendon function, but they do not fully characterize the hand at work because the internal forces that resist tendon gliding are not considered. Lane et al. 17 tested the gliding function of tendons after flexor tendon injury by calculating the work of flexion, which is the area under the measured force per excursion curve from initial loading to full flexion. Lane and colleagues included tendon excursion (total tendon excursion at full flexion) and terminal force (force recorded at full flexion) as parameters for evaluating tendon function. To date, one study has used this method to examine the biomechanical changes in finger flexion brought about by resection of the digital sheath. ~8 That article highlighted the basic loss of ROM due to bowstringing and accounted for the forces that resist tendon gliding before and after pulley excision. More recent work by Greenwald et al. 19 has demonstrated the use of a computer-integrated tensiometer unit to characterize dynamic FDP function. This system simultaneously measured tendon excursion, endpoint load, fingertip pinch force, and work of flexion. Our purpose in this study was to use this computer-integrated tensiometer to (1) study how FDS function is affected by excision of annular pulleys A1, A2, A3, and the palmar aponeurotic (PA) in various combinations and (2) examine how FDS biomechanics changes after loss of the FDP.

Materials and Methods Hand Preparation

The biomechanics of the FDS was studied in 30 fingers (index, middle, and ring) from 11 unrelated, fresh-frozen human cadaveric hands. Each hand was sharply severed 2 cm proximal to the wrist crease, placed into an airtight plastic bag, and thawed in a warm-water bath 4 hours before testing. Each defrosted hand exhibited a full ROM at the metacarpophalangeal (MP) and interphalangeal joints. The

329

DIP joint of each finger was fixed at 20 ~, so that DIP hyperextension during fingertip pinch driven by FDS excursion was prevented. A Bruner incision was made along the length of the index, middle and ring fingers and the digital sheath was carefully exposed. The presence of the A 1, A2, A3, and PA pulleys was verified, and the skin was carefully closed with a running suture of 4-0 silk. The digital skin always remained sutured except when reopened locally for a pulley excision; all FDS pulis were performed with the skin closed. To prevent desiccation, we frequently irrigated the internal and external tissues with 0.9% saline. Except for the brief time during a pull, exposed tendons and sutured areas were wrapped in saline-saturated gauze. FDS and FDP tendons were identified proximally in the carpal canal and exposed through a transverse palmar incision. Intertendinous connections and surrounding tissues were dissected away, so that independent tendon gliding was ensured. A loop of suture was woven into the proximal end of each flexor tendon: 2-0 coated Vicryl into each FDP and 0 Prolene into each FDS. The proximal ends of the extensor tendons were sutured together with a long loop of 2-0 coated Vicryl. The FDS suture anchored the tendon to a force transducer. Suture loops woven into the FDP and extensors were drawn over pulleys and counterweighted (15g for each FDP, 100g for extensors), thus eliminating slack or buckling in the tendons. Two 0.062-inch Kirschner wires (K-wires) were drilled into each hand, one laterally through the shafts of the index and middle metacarpals, and one medially through the ring and small metacarpals. The K-wires were bolted to a motor-driven slide tray that rigidly positioned the hand palm side up. A digital pinchmeter (Greanleaf Medical Systems, Palo Alto, CA) was mounted on the slide tray just above the palm using aluminum bars. The pinchmeter mount was adjustable to allow the index, middle, and ring fingers to impinge squarely on the meter when each was tested. Data Collection

A diagram of the experimental design is shown in Figure 1. A high-torque motor powered a screwdriven slide tray. The FDS tendons were distracted at 4 cm/min for all trials. ~9Data for linear displacement (tendon excursion), endpoint tendon load, and fingertip pinch force were gathered with analog-to-digital devices connected to a Macintosh computer system. Linear displacement was measured with a linear variable differential transformer (LVDT) (Lucas

330

Hamman et al. / A2, A3 Pulleys, FDP Needed for FDS Function

Figure 1. Schematic diagram of the computer integrated tensiometer. A hand is mounted on a sliding platform powered by a high-torque motor. Analog signals from linear variable differential transformer (LVDT) force transducer, and EVAL unit are digitally sampled and sent through computer interface for acquisition and analysis. A/D converter, analog-to-digital converter.

Schaevitz, Pennsauken, NJ; accuracy: _+ 0.1% over 10-cm range). Tendon load was measured using a force transducer (Lucas Schaevitz; accuracy: _+ 0.02% over a 450-N range). These units, together with the Greenleaf pinchmeter, simultaneously collected analog data at 10 Hz. An analog-to-digital converter (ADC488/16, IO Tech, Cleveland, OH) transformed the analog data and sent the digital output through a Macintosh interface unit (MacSCSI 488, IO Tech). Data files were then received by custom software for formatting and calculation of work, the area under the load-excursion curve. This integral was calculated from the excursion and load when tension was first recorded in the proximal tendon to a reproducible endpoint: the excursion and load in the proximal tendon at which the fingertip pinch force reached 0.5 kg. This 0.5-kg endpoint, although arbitrary and not a part of the work integral itself, is a physiologically modest pinch force that any normal FDS tendon should be capable of generating without elastic deformity. This reproducible endpoint gave each hand the same objective task, independent of what was happening in the sheath. Because each hand performed the same task, all three experimental parameters related to the digital sheath (tendon excursion, endpoint tendon load, and work of flexion) could be compared within a consistent model of finger function. Preparation for A n a l y s i s

Our experiments had two objectives: (1) to characterize which portions of the digital sheath are most important for preserving optimal FDS function and (2) to measure how much FDS function is compromised by loss of the FDP.

In order to answer the first question, we devised a randomized excision protocol to exhaust all combinations of one, two, and three pulley cuts (choosing from A1, A2, A3, and PA). The second question supposed that FDP excision might affect how the FDS slides through the slackened digital sheath. We tested this possibility at each pulley-cutting combination by pulling the FDS both with the FDP in place (FDP-in) and with the FDP excised from the sheath (FDPout). We accomplished this by drawing the FDP out of the sheath distally through a window cut between A4 and A5. This window allowed the FDP to slide back into the sheath for the next step in the pulley cutting-series. Sliding the FDP into and out of the sheath was done easily and did not affect any pulleys touching the FDS. The absolute values of tendon excursion and load vary widely among cadaveric hands, owing to variations in tendon moment arm, phalanx length, and tissue compliance; therefore, the raw numbers generated for each hand could not be directly compared. Instead, each finger served as its own control. Data from an excision step were compared to the control data from the same finger's intact state and an efficiency number was determined. Hence, relative changes due to pulley excision and FDP removal could be compared across hands as changes in efficiency. Changes in efficiency of FDS function for all three parameters (tendon excursion, endpoint tendon load, and work of flexion) were calculated as follows: Efficiency (of Work, Load, or Excursion) = Control Value x 100% Eperimental Value Each finger was randomly assigned a pulley-cutting protocol that progressed from the intact pulley

The Journal of Hand Surgery / Vol. 22A No. 2 March 1997

system (control state) through a series of cuts of one, two, and three pulleys (experimental states). All four states (control included) were tested with FDP-in and FDP-out; for changes in efficiency due to FDP removal, we used the FDP-in state as control. The raw data for excursion, load, and work for each combination of pulley and FDP state (eg, A2 excised, FDP in place = A2 [-] FDP-in) were collected three times and the results were averaged. These averages were inserted into the efficiency equation. The efficiency data are summarized in Table 1 as averages of the efficiency values collected from independent finger specimens (n). Fingers were considered independent when they came from different hands; only independent fingers were used in statistical analyses, Statistical significance was determined by a Student's t-test for each combination of pulley excision and FDP placement. All p values less than .05 indicated that the observed effect could not be caused by random variation and that the data points were clustered closely enough to permit quantitative judgments. Our quantitative conclusions were drawn from only the values in Table 1 marked with an asterisk, all of which are reported as average efficiencies. These values represent statistically significant findings. These findings are based on two measured quantities, tendon excursion and tendon load, and one calculated quantity, work of flexion. When an efficiency number is less than 1 it means that the amount of excursion, load, or work was increased in the experimental finger when compared to the control, intact finger. These in vitro losses in efficiency represent possible similar losses in vivo. In summary, raw excursion data measure the distance a tendon must travel to flex its finger to a fixed endpoint, a pinchmeter reading of 0.5 kg at full fist. Decreased excursion efficiency indicates that the pathologic finger requires more excursion to flex the finger the same amount compared to the intact finger. This is most likely due to increased bowstringing of the tendon between pulleys. This is a potential problem for finger function because the arm has a limited ability to increase its tendon excursion. Tendon load is the tension in the tendon during flexion. Tendon load increases as resistance to tendon gliding increases. Load data was used in two ways: both to calculate endpoint load efficiency and to calculate work. The tendon load at endpoint was used to calculate the endpoint load efficiency entries in Table 1. However, endpoint load efficiency provides a limited view of tendon forces; therefore, we drew few quantitative conclusions directly from our load-at-endpoint efficiency data. Instead, we focused

331

on the work of flexion data because it measures similar changes in tendon load throughout flexion and hence is the preferred barometer of forces which resist tendon gliding. Work of flexion (work -- force x distance) was calculated as the area under the instantaneous loadversus-excursion curve collected by the computerintegrated tensiometer. The calculation of work takes into account the frictional forces that resist tendon gliding. Because work is the product of two components, excursion and load, one must consider why each may change owing to a pathologic digital sheath anatomy. Tendon load during flexion would increase if resistance to tendon gliding was increased. This would increase total work performed and hence would decrease work efficiency when compared to the control (see equation). Similarly, tendon excursion would increase if the moment arm at the joint increases, such as during tendon bowstringing. This will also increase total work performed and would decrease work efficiency when compared to the control. It is clear that these two effects are intimately related and difficult to distinguish. However, the work parameter allows one to measure their combined effect and verify that it is greater than either the excursion or load effect alone.

Results Pulleys Excised/Flexor Digitorum Profundus Intact Excursion Efficiency. Data from single pulley cuts (Table I) show that excising one of the annular pulleys caused consistent decreases in excursion efficiency: A1, 96.9%; A2, 92.4%; A3, 96.7%. Combined loss of two pulleys caused slightly greater decreases in excursion efficiency: A1-A2, 88%; A2-A3, 90%; A2-PA, 93.3%. Loss of two annular pulleys plus the PA resulted in similar decreases in excursion efficiency. Loss of all three annular pulleys, A1-A2-A3, notably lost the most excursion efficiency (87.4%). These losses indicate a longer tendon pull was needed--namely, that caused by tendon bowstringing during flexion. Loss of A2 consistently resulted in significant loss of excursion efficiency and hence played the most important role in maintaining excursion efficiency by decreasing bowstringing. Endpoint Load Efficiency. Combined loss of A2-A3 pulleys produced the only statistically significant decrease in endpoint load efficiency (90.3%). This means that resistance to FDS tendon gliding approaching our full-fist endpoint significantly increased without the A2 and A3 pulleys. Although the variability in the other data was too large to per-

332

Hamman et al./A2, A3 Pulleys, FDP Needed for FDS Function

Table 1. Efficiencies From Cutting Protocols Section 2: FDP(-) and Pulley Effects

Section 1: Pulley Excision Effects Pulleys Excised At A2 A3 PA A1, A2 A 1, A3 A2, A3 A2, PA A1, PA A3, PA PA, A1, A2 PA, A 1, A3 PA, A2, A3 A1, A2, A3 None

n

Work Effect

Load Effect

Excursion Effect

Work Effect

Load Effect

Excursion Effect

6 5 5 7 5 5 5 5 4 4 7 5 5 5 11

0.973 0.849* 0.875* 0.949 0.801" 1 0.825* 0.945 0.97 0.92 0.917 0.925 0.864 0.861" NA

0.989 0.943 0.914 0.998 0.88 1.02 0.903* 0.993 1 0.962 1.014 0.952 0.92 0.973 NA

0.969* 0.924* 0.967* 0.992 0.88* 0.971 0.9* 0.933* 0.981 0.969 0.905* 0.95" 0.902* 0.874* NA

0.844"t 0.833 0.908 0.901 0.818 1.043 0.885"t 0.921 0.986 0.874 0.936 0.873 0.905 0.947 0.924*

0.885"t 0.9 19 0.9 28 0.93 0.876 1.054 0.921 0.969 0.999 0.933 1.061 0.917 0.961 0.995 0.955

0.92"t 0.905"t 0.937"t 0.943"t 0.858"t 0.988"t 0.877"t 0.908 *t 0.944"t 0.936"t 0.893 * 0.914*t 0.896* 0.861"t 0.95 2*

These fingers are a random sample of the 30. * Values are statistically significant from the intact control at p < .05. ~ FDP-out data are significantly different from the corresponding FDP-in data by paired t-test, p < .05. FDP, flexor digitorum profundus; FDP-in, FDP intact; PA, palmar aponeurotic pulley.

mit statistical significance, trends in the endpoint load efficiency data point to drag effects that are statistically significant when the whole excursion is considered (see Work Efficiency below). In general, pulley loss increases the angle a tendon takes when entering the next pulley; as that angle increases, so too can the resistance to tendon gliding increase. 3 Work Efficiency. Excision of all three annular pulleys showed significant loss in work efficiency (86.1%). However, one may break this result down into components by examining the data from single and double pulley cuts. Data from single pulley cuts shows that loss of A2 or A3 each causes significantly decreased work efficiency, while loss of the A1 or PA did not: A2, 84.9%, and A3, 87.5%. Combined excision of the A1-A2 caused significant loss of work efficiency (80.1%). But even more dramatic was the loss of the A2-A3 combination, the only point where all three parameters, work (82.5%), load (90.3%), and excursion efficiency (90%) were significantly decreased. This result shows that A2 and A3 are the most important combination for preserving optimum FDS function. A1 seems to be important only in combination with A2. These results suggest that loss of A2 increases the angle at which FDS enters A1, that loss of A3 increases the FDS angle at A2, and that combined loss of A2 and A3 further increases the FDS angle at A1. This explanation allows for the dual effects of increased moment arm due to bowstringing across

the PIP and MP joints and the increased friction due to increased tendon angle at pulleys.

Pulleys Intact/Flexor Digitorum Profundus Excised Removing the FDP from a finger with an otherwise intact digital sheath caused statistically significant losses of both work and excursion efficiency when compared to the FDP-in control (Table 1). This result suggests that removal of the FDP causes slack in the digital sheath and increases the amount of sliding friction once the finger attempts full flexion and active pinching force. The ability to account for this effect and suggest a mechanism is enhanced by the data from the pulley-excised, FDP-excised states. Pulleys Excised/FDP Excised In order to distinguish the FDP excision effect from the pulley excision effect, we compared the FDP-out data for each pulley excision state to the corresponding FDP-in data using paired t-tests (p < .05). Statistically significant differences are noted by daggers in Table 1. The following results focus on how loss of the FDP alters FDS biomechanics. Excursion Efficiency. As Table 1 demonstrates, in almost every case, FDP removal caused a statistically significant additional loss in FDS excursion efficiency when compared to each finger's FDP-in state. These results suggest that regardless of which

The Journal of Hand Surgery / Vol. 22A No. 2 March 1997

pulley was cut, FDP removal caused consistent increases in the excursion needed to perform a given task. For example, excursion efficiency for A 1(-) fingers decreased from 96.9% when FDP was in place to 92% when FDP was removed. This occurred because the average FDS moment arm across joints increased slightly. Such an additional increase in moment arm could have been caused by the slackened pulleys' assuming a tentlike shape during fingertip pinching (Fig. 2). Endpoint Load Efficiency. Only FDP-out fingers lacking the A1 pulley showed significant decrease in endpoint load efficiency (88.5%). This effect suggests that in an A I ( - ) finger, additional loss of the FDP caused some additional resistance to tendon gliding during active pinch. Work Efficiency. Two results were statistically significant: FDP removal caused a decrease in work efficiency in AI(-) fingers (84.4%--significantly decreased from the insignificant 97.3% when the FDP was intact) and caused an increase in work efficiency in A2-A3(-) fingers (88.5%--increased from 82.5% when the FDP was intact). The first result clarifies the result from the endpoint load efficiency data; namely, that without the FDP in an AI(-) finger, the FDS experiences additional drag throughout the pinching motion. It is possible that the

FDP

-

in

333

pulley tenting presumed to be responsible for the consistent loss of excursion efficiency may also play a role in increasing resistance to tendon gliding. It is plausible that the longest pulley, the A2, would buckle in the center when loaded during pinch. Such buckling could increase resistance by allowing the ends of A2 to create an even sharper tentlike shape (Fig. 2). Variability in the rest of the data was too large to allow other quantitative conclusions, though a decreasing trend in endpoint work and load efficiency is clearly evident, which suggests that A2 is probably not unique in this buckling effect. The second significant result shows that when a finger lacks A2 and A3, FDS sliding through A1 became easier when the FDP did not fill up the rest of the pulley. Although this is true biomechanically, a finger lacking the A2 and A3 is already so compromised in excursion efficiency that such a decrease in sliding friction and hence increase in work efficiency cannot be seen as an improvement in the system. This result was a plausible biomechanical artifact which has little to no clinical relevance, but demonstrates the sensitivity of the experiment. While not statistically significant, a similar trend of increased work efficiency was observed in the related PA-A2-A3(-) and A1-A2-A3(-) states. These trends show that sliding friction in the digital sheath is a complex vari-

FDP - out

Figure 2. This illustration demonstrates (1) the buckling that occurs at the A2 pulley when the flexor digitorum profundus (FDP) is removed, which causes an increase in the work required to flex the finger (lateral view), and (2) the increased moment arm experienced by the flexor digitorum superficialis (FDS) when the FDP is removed from the sheath, which is referred to as tenting (best seen at bottom, in cross-section). FDP-in, FDS intact; FDP-out, FDP excised.

334

Hamman et al. /A2, A3 Pulleys, FDP Needed for FDS Function

able affected by the angles the tendons take when sliding through pulleys, by the resistance between the tendons themselves, and by the resistance between the tendons and the pulleys. Additional experiments are needed to further elucidate this system.

Discussion This study characterizes the basic biomechanics of the FDS when the parameters of tendon excursion, endpoint tendon load, and work of flexion are used. We investigated the role of the digital sheath anatomy by pulling the FDS with an intact sheath first, and then at each point of a pulley excision protocol in which we used the combinations of one, two, and three pulleys excised. In order to appreciate the delicate balance between the flexor sheath and tendons, we pulled the FDS before and after the FDP was removed at each point of the pulley excision protocol. The equations derived by Landsmeer z0 and the basic biomechanical descriptions of Idler 1 explain how losses of efficiency are related to sheath anatomy. For example, excursion efficiency is lost when the moment arm at one or more joints has been increased. Landsmeer states that (1) the distance from the pulley edge to the joint axis and (2) the perpendicular distance from the pulley edge to the longitudinal phalangeal axis both influence tendon excursion via the moment arm. Our excursion efficiency data confirm Landsmeer's findings. Pulley excision increases the distance (1) from the pulley to the joint axis, thus permitting tendon bowstringing during flexion; FDP removal increases the distance (2) between pulley and phalangeal axis by creating space in the sheath and allowing the pulleys to assume tentlike shapes during flexion when pulled by the normal forces of a sliding FDS (Fig. 2). Pulley tenting was previously noted by Idler, 1 who pointed out how diminutive flexor tendon grafts can affect the tendon moment arm at a pulley. Slack in the digital sheath caused by FDP removal affected both excursion and work efficiency, as shown at the bottom of Table 1. Without the FDP to fill up the rest of the sheath, the FDS probably slid against some internal buckling of the sheath and pulleys (Fig. 2). This effect was most dramatic when the AI was excised in a FDP-out finger, leaving the long A2 pulley alone to restrain the FDS at the MP joint (Table 1). The data shown in Table 1 confirm the importance of the digital pulleys for maintaining flexor tendon

function. We found that A2 and A3 were the most important pulleys for the FDS because their individual losses caused significant decreases in work and excursion efficiency and their combined loss caused a significant decrease in all three parameters. A1 was found to be important only when sectioned in conjunction with A2. However, the A1-A2 combination is indeed very important, as shown by the large decreases in work efficiency (80.1%) and excursion efficiency (88%) caused by their excision. Loss of the A1 by itself did not have a large effect so long as the FDP was present. However, removing the FDP from A I ( - ) fingers caused deficits on the same order as the combined loss of the A2 and A3. This suggests that, in an FDS-only finger, the A1 pulley should be preserved if the sheath is otherwise intact. Each entry in Table 1 is the mean of a number of independent measurements; thus, statistical significance was calculated within a pulley type. As mentioned before, this kind of significance points to the clustering of the data and, hence, to the consistency of the effect across different hands; this permits small but reliable deviations to be considered statistically significant, as was the case in the excursion efficiency of the AI(-)-FDP-in finger (96.9%). Savage 16 used this same approach when highlighting pulley excision effects. Statistical significance does not necessarily imply clinical significance. However, effects such as those found in the A1-A2(-)-FDP-in, A2-A3(-)-FDP-in, and the AI(-)-FDP-out cases are likely to be clinically significant because the data were consistent and the absolute values of the efficiency losses were large. The biomechanical parameters in this study focused on the digital sheath. No attempt was made to model the function of the FDS muscle itself. In living muscle, force generated per unit excursion rises from 0 to a peak and then declines as maximum flexion approaches. 2~ Because we used a constant rate of excursion with a high torque motor, the absolute values of the forces generated throughout these excursions were undoubtedly different from those occurring in vivo. It is impossible to predict, for each cadaver, the length-force curve that would have been generated in vivo. However, each tendon was pulled at the same power by the same high-torque 4 cm/min excursion speed. Standardizing input power let us focus on how changes in sheath anatomy diminished how well this standard power was transmitted to the fingertip. It is clear that excising pulleys or removing the FDP affect how efficiently input force via tendon excursion is translated into active full-fist pinching

The Journal of Hand Surgery / Vol. 22A No. 2 March 1997

force. Efficiency numbers compare the relative differences measured within a given finger and, hence, allow each finger to be its own control. This eliminated the influence of hand-to-hand variability in tendon moment ann, phalanx length, and tissue compliance, all of which affect the raw data for tendon excursion, load, and work. Excursions applied to these fingers never exceeded the physiologic range, as proposed by Brand et al. 22 Excursion efficiency was most affected by tendon bowstringing, the most important result of pulley excisions. In addition to sapping excursion efficiency, tendon bowstringing also affected the load and work efficiency by increasing the resistance to tendon gliding in two distinct ways. First, a bowstringing tendon develops sharp angles as it enters or leaves the remaining pulleys. Sliding friction increases as this angle becomes more acute. 3 Second, an inevitable result of tendon bowstringing is that the tendon is pulled through the fat and fascia under the skin. One group of researchers 18 has shown that palmar skin adds drag to the system during bowstringing. Although the cadaveric skin cannot simulate chronic changes in the living hand, its measurable affect on work data makes it worth including in a model of flexor tendon function because similar effects may be found in vivo. 18 Two clinical correlations can be suggested from this work. The first is the importance of pulley tightening in a sublimus finger to obtain optimum function. The second correlation comes during tendon grafting and/or reconstruction: either a graft should be chosen to fill the sheath or the pulleys need to be reconstructed more tightly so that the FDP minus effect of FDS bowstringing and pulley tenting is not allowed. This research may be helpful to surgeons considering reconstruction of the flexor tendons and sheath. In patients who require power grip or who need excellent finger flexion during wrist flexion, as may be the case for many musicians and mechanics, repair of the FDS could become a priority. Whenever the digital sheath must be excised so that the injured area can be reached, biomechanical data predicting the potential losses associated with a resected sheath might prove useful. The possibility that postoperative adhesions could further inhibit tendon gliding is an incentive to restore the normal sheath geometry whenever possible. The FDP-out data further contribute to the prevailing belief that both flexor tendons should be repaired whenever possible.

335

References 1. Idler RS. Anatomy and biomechanics of the digital flexor tendons. Hand Clin 1985;1:3-11. 2. Smith RJ. Balance and kinetics of the fingers under normal and pathological conditions. Clin Orthop 1974;104:92-111. 3. Brand PW. Clinical mechanics of the hand. St. Louis: CV Mosby, 1993:61-66. 4. Long C, Conrad PW, Hall EA, Furler SL. Intrinsic-extrinsic muscle control for the hand in power grip and precision handling. J Bone Joint Surg 1970;52A:853-867. 5. Napier JR. The prehensile movements of the human hands. J Bone Joint Surg 1956;38B:902-913. 6. Barton NJ. Experimental study of optimal location of flexor tendon pulleys. Plast Reconst Surg 1969;43:125-129. 7. Brand PW, Cranor KC, Ellis JC. Tendon and pulleys at the metacarpophalangeal joint of the finger. J Bone Joint Surg 1975;57A:779-784. 8. Doyle JR, Blythe W. The finger flexor tendon sheath and pulleys: Anatomy and reconstruction. In: AAOS symposium on tendon surgery in the hand. St. Louis: CV Mosby, 1975:81-87. 9. Idler RS, Strickland JW. The effects of pulley resection on the biomechanics of the PIP joint. J Hand Surg 1984;9:595. 10. Lin G-T, Amadio PC, An K-N, Cooney WP, Chao EYS. Biomechanical analysis of finger flexor pulley reconstruction. J Hand Surg 1989;14B:278-282. 11. Manske PR, Lesker PA. Palmar aponeurosis pulley. J Hand Surg 1983;8A:259-263. 12. Widstrom CJ, Johnson G, Doyle JR, Manske PR, Inhofe P. A mechanical study of six digital pulley reconstruction techniques: Part I. Mechanical effectiveness. J Hand Surg 1989; 14A:821-825. 13. Goldstein SA, Greene TL, Ward WS, Matthews LS. A biomechanical evaluation of the function of the digital pulleys. Orthopaedic Transactions 1985;8:342. 14. Hunter JM, Cook JE The pulley system: rationale for reconstruction. In: Strickland JW and Steichen JB, ed. Difficult problems in hand surgery. 3rd ed. St. Louis: Mosby, 1982. 15. Hume EL, Hutchinson DT, Jaeger SA, Hunter JM. Biomechanics of pulley reconstruction. J Hand Surg 1991;16A: 722-730. 16. Savage R. The mechanical effect of partial resection of the digital fibrous flexor sheath. J Hand Surg 1990; 15B:435-442. 17. Lane JM, Black J, Bora FW Jr. Gliding function following flexor-tendon injury. J Bone Joint Surg 1976;58A:985-990. 18. Peterson WW, Manske PR, Bollinger BA, Lesker PA, McCarthy JA. Effect of pulley excision on flexor tendon biomechanics. J Orthop Res 1986;4:96-102. 19. Greenwald D, Shumway S, Allen C, Mass D. Dynamic analysis of profundus tendon function. J Hand Surg 1994; 19A:626-635. 20. Landsmeer JME Anatomical and functional investigation of the articulations of the human fingers. Acta Anat 1955; 25(suppl. 24):59-84. 21. Elftman H. Biomechanics of muscle. J Bone Joint Surg 1966;48A:363-377. 22. Brand PW, Beach RB, Thompson DE. Relative tension and potential excursion of muscles in the forearm and hand. J Hand Surg 1980;6A:209-214.