Abstract
Background: Hallmarks of a persistent crouched walking pattern exhibited by individuals with cerebral palsy usually include loss of an adequate plantar flexion/knee extension couple, hamstring and/or psoas tightness, or contracture in conjunction with quadriceps insufficiency. Traditional treatment addresses the muscle-tightness component, but not the contracture or the muscle insufficiency. This study was performed to evaluate the effectiveness of distal femoral extension osteotomy and/or patellar tendon advancement in the treatment of crouch gait in patients with cerebral palsy.
Methods: A retrospective, nonrandomized, repeated-measures design was used. Individuals with a diagnosis of cerebral palsy were included if they had had (1) a distal femoral extension osteotomy in combination with a distal patellar tendon advancement (thirty-three patients), (2) a distal femoral extension osteotomy without patellar tendon advancement (sixteen), or (3) a distal patellar tendon advancement only (twenty-four). All subjects were evaluated with preoperative and postoperative gait analysis. Gait, radiographic, strength, and functional measures were included in the analysis to assess changes in knee function.
Results: Seventy-three individuals met the criteria for inclusion. A single side was chosen for the analysis of each subject. Ninety percent of the subjects had additional, concurrent surgery. Improvements were noted in the index assessing the level of gait pathology and in functional variables across all groups, and pain was consistently decreased. All preoperative stress fractures healed. Strength levels were maintained across all groups. The Koshino index of patellar height improved from 1.4 to -2.3 in the group treated with patellar tendon advancement only and from 1.5 to -2.9 in the group treated with both osteotomy and tendon advancement. The range of knee flexion improved an average of 15° to 20°, and stance-phase knee flexion was restored to the typical range (9° to 10°) in the groups that had advancement of the patellar tendon as part of the procedure. Individuals who underwent a distal femoral osteotomy only were still in a crouch (a mean of 31° of knee flexion in midstance) at the final assessment.
Conclusions: Inclusion of patellar tendon advancement is necessary to achieve optimal results in the surgical management of a persistent crouch gait exhibited by adolescents and young adults with cerebral palsy. When this procedure is done alone or in combination with a distal femoral extension osteotomy (for the treatment of a knee flexion contracture), knee function in gait can be restored to values within typical limits, with gains in community function.
Level of Evidence: Therapeutic Level III. See Instructions to Authors for a complete description of levels of evidence.
Persistent crouch gait is a common problem exhibited by children, adolescents, and young adults with spastic cerebral palsy. A number of factors can contribute to this gait, including hamstring and/or psoas tightness due to spasticity or contracture1-3, lever-arm dysfunction4-6, weakness, and impaired balance4,7. Once a crouch gait has developed, body mechanics contribute to the progression of knee flexion, hip flexion, and ankle dorsiflexion8. The loss of an adequate plantar flexion/knee extension couple keeps the ground reaction force behind the knee joint and in front of the hip and ankle joints4. Fixed knee-flexion contractures often develop. Patella alta and elongation of the patellar tendon develop over time and become contributing factors. Knee pain, probably related to the increased quadriceps forces required to stabilize the flexed knee joint, increased patellofemoral compressive forces, and/or eventual patellar pole or tibial tubercle fracture, frequently develops as well9-11. Knee and back pain have been mentioned as two of the primary factors that limit or decrease the walking ability of these patients12,13.
Discerning the factors contributing to crouch gait in any individual case is complex. Advances in computer-aided analysis of typical and pathologic gait with use of modern motion-analysis laboratories, induced acceleration analysis6,14-17, musculoskeletal models for estimation of muscle-tendon lengths18-20, and the understanding of concepts such as lever-arm dysfunction4-6 all play key roles. When present, fixed knee-flexion contractures and quadriceps insufficiency contribute to crouch gait, and surgical procedures to address both should be considered21,22. Most traditional surgical procedures do not target either of these problems. Lotman reported that, without the treatment of patella alta, the rate of recurrence of knee flexion deformity was 32%21.
Because most individuals with a persistent crouch gait have had previous interventions, including muscle lengthening procedures, it is necessary to consider alternative methods of treatment. The use of distal femoral extension osteotomy and patellar tendon shortening/advancement-type procedures for treatment of the persistent crouch gait of individuals with neuromuscular disorders dates back to the early twentieth century23-25. Distal femoral extension osteotomy has been described for the treatment of fixed flexion deformities associated with poliomyelitis, arthritic conditions, and myelomeningocele23,26-30, but reports of its use in patients with cerebral palsy are limited31. Reported rates of complications, including angulation or hyperextension deformity, neurovascular compromise, and/or fracture, were high27,28,30. Recently, anterior distal femoral epiphysiodesis has been advocated (for skeletally immature patients) as a less invasive procedure32,33.
Patellar tendon advancement procedures, including distal transplantation of the patellar tendon with a block of bone in conjunction with hamstring lengthening, was first described by Chandler in 193324. A follow-up article described a modified technique to address the complication of genu recurvatum when the advancement was performed in an immature child25. The article also addressed issues of fixation and included the first descriptions of tension-band techniques. Use of this surgical procedure with modifications continued to be reported into the 1960s for the treatment of knee flexion deformity in children with cerebral palsy34-38, but there was very little mention of advancement of the patellar tendon in the literature after that time. Recently, patellar tendon shortening procedures have been described, with varied degrees of success reported22,39,40.
Despite the recognition of the importance of both quadriceps insufficiency and knee flexion contracture as causes of persistent crouch gait, to our knowledge the combined use of patellar tendon advancement and distal femoral extension osteotomy has not been the subject of previous research. Over the past decade, at our facility, patellar tendon advancement with or without distal femoral extension osteotomy (with use of modifications of techniques reported in the past) has become an integral part of the treatment of crouch gait exhibited by adolescents and young adults with cerebral palsy. The purpose of this study was to evaluate the effectiveness of these techniques.
A retrospective, nonrandomized, repeated-measures design was used. Institutional review board approval was granted. A search of the clinical database of the Center for Gait and Motion Analysis and the surgical schedules at Gillette Children's Specialty Healthcare, St. Paul, Minnesota, was performed to identify subjects seen between January 1994 and June 2007. The criteria for selection for the study were (1) a primary diagnosis of cerebral palsy; (2) the ability to walk about the house without the physical assistance of another person (level 4 or above on the walking scale of the Gillette Functional Assessment Questionnaire41); (3) treatment with a distal femoral extension osteotomy, patellar tendon advancement, or a combination of the two procedures at our institution between January 1994 and December 2005; and (4) performance of comprehensive preoperative and postoperative gait analysis at our Center for Gait and Motion Analysis, with the preoperative gait analysis data acquired no more than twenty-four months prior to the date of surgical intervention and the postoperative data acquired seven to thirty-six months after the surgical intervention. The level according to the Gross Motor Function Classification System for Cerebral Palsy (GMFCS)42 was not considered to be an inclusion or exclusion criterion.
Surgical Technique
A lateral approach to the distal part of the femur posterior to the vastus lateralis is used for the distal femoral extension osteotomy. The chisel for a 90° blade-plate (Synthes USA, West Chester, Pennsylvania) is inserted just proximal to a guidewire placed at a 90° angle to the femoral shaft and just proximal to the physis (or physeal scar) with the angle guide of the chisel parallel to the tibia. This placement helps to avoid varus or valgus displacement of the osteotomy. An anterior triangular wedge of bone that matches the degree of contracture is removed. Any bone protruding posteriorly from the distal fragment is also removed (Figs. 1-A, 1-B, and 1-C). Coronal and transverse plane deformities can be corrected simultaneously. The clinical finding of a fixed knee-flexion deformity is eliminated by this compensatory osteotomy.
Which of two methods of patellar tendon advancement is used depends on the skeletal maturity of the patient. If the physis is open, the patellar tendon is sharply divided from the tibial tubercle to avoid physeal injury and is advanced under a periosteal flap. If growth is complete, the tibial tubercle (with the attached patellar tendon) is transposed distally and secured with a compression screw. A tension-band of FiberTape (Arthrex, Naples, Florida) (or a 16-gauge Luque wire early in the series) placed transversely through the patella and the proximal part of the tibia is used to protect the repair (Figs. 1-B, 2-A, and 2-B).
Outcome Analysis
Subjects were categorized into three treatment groups on the basis of the type of surgical intervention performed: (1) distal femoral extension osteotomy only, (2) distal femoral extension osteotomy and patellar tendon advancement, and (3) patellar tendon advancement only. If bilateral surgery had been performed, a single limb was chosen at random to be included in the analysis. Six subjects who had undergone bilateral surgery with a different surgical procedure on each limb were excluded from the study.
Gait, radiographic, and functional measures were included in the analysis. Gait measures included three-dimensional lower-extremity kinematic data measured with a VICON 370 or 512 Motion Measurement System (Vicon Motion Systems, Lake Forest, California); energy expenditure measured with a MedGraphics CPX/D system (Medical Graphics, St. Paul, Minnesota); and strength43, range of motion44, and spasticity45 measured with standard clinical examinations. Radiographic data included measurement of patellar height on lateral knee radiographs. Functional measures included the Gillette Functional Assessment Questionnaire41 (FAQ) and the Pediatric Outcomes Data Collection Instrument46 (PODCI), which were completed by the parents.
Gait Analysis Variables
Seven discrete kinematic parameters were extracted from a single representative trial from each preoperative and postoperative gait analysis. The Gillette Gait Index (GGI), previously known as the Normalcy Index47, was used to assess global gait pathology. The measurement error of gait analysis at our institution has previously been described48. A mean error of approximately 3° was noted at the pelvis, hip, and knee in the sagittal plane.
Strength data were converted from the traditional five-grade scale to a percentage scale, as described by Kendall et al.43, to numerically represent all possible grades. Normal strength (grade 5) is transformed to 100%; grade-0 strength is maintained as 0 for the end range of the scale. Typical antigravity strength against partial resistance (grade 3+) is represented as 60%. Knee flexion contracture and the strength of the gluteus maximus, quadriceps, and plantar flexors were determined with clinical examination. Measurement errors for strength and goniometric measurements specific to our laboratory are not available but are considered to be consistent with those in the published literature43,49,50. Available information does suggest that the reliability of measurements improves with the experience of the testers and use of a standard protocol51. All of the therapists in our laboratory have seven to twenty years of gait laboratory experience.
The net non-dimensional normalized rate of oxygen consumption (gross rate minus resting rate) was calculated to assess energy expenditure52.
Radiographic Variables
Patellar height was measured on lateral knee radiographs with the method described by Koshino and Sugimoto53. The Koshino index has been shown to be a stable assessment for skeletally immature children53,54. A ratio is calculated with use of the epiphyseal line midpoints of the tibia and femur and the midpoint of the long axis of the patella (Figs. 1-A and 1-B). A single preoperative radiograph made closest, by date, to the date of the preoperative gait analysis and a single postoperative radiograph made closest, by date, to the date of the postoperative gait analysis were used. All measurements were performed by one of the authors (J.L.S.).
To ensure that author bias was not introduced in the data, a reliability study was conducted with use of a subset of eighty-eight radiographs included in the analysis. Three individuals who were not associated with the study (a research assistant, a nurse, and an engineer) were recruited to measure randomly ordered lateral knee radiographs with the Koshino method. Intraclass correlation coefficients (0.96 for the measurement of the patella-tibia distance, 0.98 for the femur-tibia distance, and 0.86 for the Koshino index) demonstrated excellent reliability among the independent raters, between the independent raters and the author (J.L.S.), and between the author and each independent rater.
Functional Variables
The Transfers and Basic Mobility, Sports and Physical Function, and Pain/Comfort subscales of the PODCI and the ten-level walking scale and eight of the twenty-two higher-level skills chosen to represent activities related to knee function from the FAQ were analyzed to assess community function.
Complications
Medical record data from the time of the surgical admission to the last available clinic note were reviewed for all subjects by one of the authors (J.L.S.). Any unwanted event was considered a complication. Complication types and rates were recorded separately for each procedure. Pain was considered a complication if it lasted six months or more after the surgery. Delayed union or nonunion was defined as a lack of bone-healing at six months or longer after the surgery. Postoperative deformities included malalignments resulting from incomplete correction, overcorrection, or angulation. Recurrent contracture was not considered a complication of the surgery. Short-term complications related to the initial inpatient stay or cast treatment were recorded but not considered in this analysis.
Statistical Methods
All statistical analyses were performed with use of pairwise exclusions for missing data. Statistical significance experimentwise was set at p < 0.05. A Bonferroni correction was used, resulting in a level of significance of p < 0.003 for the nineteen outcome parameters in the study. A paired t test or Wilcoxon signed-rank test (in the case of non-normal distributions) was used to assess the significance of changes from before to after the surgery.
Subjects
One hundred and sixty-three individuals with cerebral palsy who had undergone distal femoral extension osteotomy and/or patellar tendon advancement were identified in the hospital databases (Fig. 3). Seventy-three adolescents and young adults (thirty-one female and forty-two male) met the criteria for inclusion in the study: sixteen had undergone distal femoral extension osteotomy only; thirty-three, distal femoral extension osteotomy and patellar tendon advancement; and twenty-four, patellar tendon advancement only. Twenty-six had unilateral surgery and forty-seven, bilateral surgery. As stated previously, for those who underwent bilateral surgery, a single limb was chosen at random to be included in the analysis. Patient data, including age, sex, and GMFCS level, are presented in Table I. No differences were noted among the three groups with regard to age or the time between the surgery and the postoperative analyses.
As all subjects were required to have undergone both preoperative and postoperative gait analysis to be included in the study, complete kinematic and GGI values were available for 100% of the subjects. Energy-expenditure, radiographic, and functional questionnaire data (particularly PODCI data) were less complete across the patient groups, with an average of 81% of the paired data available for analysis (see Appendix).
Sixty-eight (93%) of the seventy-three subjects had had previous surgical intervention (see Appendix). Seven of the sixty-eight had had muscle tone reduction only, twenty-five had had a combination of muscle tone reduction and orthopaedic surgery, and thirty-six had had orthopaedic surgery only. Of the sixty-one subjects who had had orthopaedic surgery, 61% had had hamstring lengthening and 87% had had heel-cord lengthening.
Concurrent surgery at the time of the distal femoral extension osteotomy and/or patellar tendon advancement was common (Table II). Sixty-six (90%) of the seventy-three subjects had additional procedures performed simultaneously with the index operation. Seven had the distal femoral extension osteotomy or patellar tendon advancement performed as an isolated procedure or with implant removal only.
Gait
Differences between groups with regard to preoperative kinematic, clinical examination, radiographic, and functional variables were evaluated with use of one-way analysis of variance. No significant difference was found with respect to seventeen of the nineteen variables. Post hoc comparisons (with a Bonferroni correction) showed that knee flexion contracture and the minimum knee flexion in stance differed significantly between the group treated with the distal femoral extension osteotomy and patellar tendon advancement and the group treated with the patellar tendon advancement only.
The improvement in selected kinematic variables between the preoperative and postoperative assessments differed among the treatment groups. The group treated with both osteotomy and tendon advancement demonstrated the greatest improvement, and the osteotomy-only group demonstrated the least change. All groups exhibited an improvement in kinematic variables, but the group treated with both osteotomy and tendon advancement and the group treated with tendon advancement only demonstrated improvement in the total range of knee motion (an average of 15° to 20°), restoration of minimum stance-phase knee flexion to the typical range (9° to 10°), and resolution of knee flexion contracture whereas the osteotomy-only group did not. Greater variability in stance-phase knee motion among individuals was noted in the tendon-advancement-only group as compared with the group treated with both osteotomy and tendon advancement (Fig. 4). All groups demonstrated an increase in anterior pelvic tilt postoperatively, and no significant changes were noted at the hip in any of the groups. The greatest increase in anterior pelvic tilt occurred in the group treated with both osteotomy and tendon advancement (Fig. 5 and Appendix).
Radiographic Evaluation
The group treated with both osteotomy and tendon advancement and the group treated with tendon advancement only exhibited a significant change in the patellar position. The Koshino index indicated an overcorrection in both groups postoperatively (Fig. 6 and Table III). Preoperative stress fractures of the patella or the tibial tubercle healed in all patients postoperatively.
Functional Assessment
Functional variables exhibited varying degrees of improvement, with no loss of ability. The FAQ ten-level walking scale demonstrated the greatest improvement in the group treated with both osteotomy and tendon advancement, with an average increase of almost one full walking level (average, +0.8 level; range, -1 to +5 levels). Improvements in the osteotomy-only group were more modest (average, +0.7 level). Almost no change in the FAQ level was noted in the tendon-advancement-only group. None of these changes reached significance (Table III).
Eight advanced skills requiring knee range of motion and/or strength were assessed with the FAQ. The osteotomy-only group demonstrated less functional skill preoperatively (75% of the total [see Appendix]) than did either the group treated with both osteotomy and tendon advancement or the group treated with tendon advancement only (83% of the total in both groups) (see Appendix). The osteotomy-only and tendon-advancement-only groups exhibited a net addition of skills postoperatively. The group treated with both osteotomy and tendon advancement exhibited a net loss of skills (see Appendix).
Strength assessment revealed substantial weakness of the gluteus maximus, quadriceps, and plantar flexors in all groups preoperatively and postoperatively (Table III). The quadriceps muscles were the strongest, with an average preoperative strength ranging from grade 3 to grade 3+ in the osteotomy-only and tendon-advancement-only groups and grade 4- in the group treated with both osteotomy and tendon advancement. The plantar flexor muscles were the weakest, with an average preoperative strength of approximately grade 2 in all groups. All strength measures had returned to preoperative levels at the time of the postoperative assessment. The only significant change in strength was an improvement in quadriceps strength (from grade 3 to grade 4) in the tendon-advancement-only group.
The groups demonstrated differences in the preoperative extensor lag (active minus passive knee extension). The osteotomy-only group exhibited a significantly greater lag than the other two groups. Improvement was noted across groups postoperatively. The group treated with both osteotomy and tendon advancement and the group treated with tendon advancement only demonstrated a significant change between the preoperative and postoperative values. The improvement in the osteotomy-only group did not reach significance.
GMFCS Level
The GMFCS level had no main effect on the change in selected kinematic, clinical examination, and functional variables (knee flexion contracture, gait speed, FAQ walking level, and knee kinematics) between the preoperative and postoperative evaluations.
Pain
Preoperatively, thirty-four patients experienced pain that limited walking ability. The pain reported was not specifically knee pain. Stress fractures (of either the patella or the tibial tubercle) identified in twelve knees were associated with knee pain. Postoperatively, pain was perceived as a limitation to the walking ability of twenty-three patients. Patients in the tendon-advancement-only group exhibited the least improvement in this parameter. PODCI pain scores improved in the group treated with both osteotomy and patellar tendon advancement and the group treated with tendon advancement only, but these improvements did not reach significance (Table III).
Complications
The total complication rate associated with the distal femoral extension osteotomies was 19% (nine of forty-nine), and that associated with the patellar tendon advancement procedures was 18% (ten of fifty-seven) (Table IV). Seventy-eight percent of the patients in the study experienced no complications related to the operative procedure(s).
The type and frequency of complications differed between the two procedures. Stretch palsy and neuropathy and postoperative deformity (undercorrection or angulation) accounted for the largest percentage of the complications related to the distal femoral extension osteotomies. Knee pain (of more than six months' duration) was the most frequent complication related to the patellar tendon advancement procedures. Growth arrest at the tibial tubercle occurred in one patient.
In this series, the knee position improved across all groups regardless of the procedures performed (distal femoral extension osteotomy, patellar tendon advancement, or a combination of the two procedures). The GGI, reflecting overall gait pathology, improved in the osteotomy-only group and in the group treated with both osteotomy and tendon advancement. Improvements in function were noted as well, as indicated by healing of preoperative stress fractures and maintenance of, or often improvement in, walking function and basic mobility scores. However, the groups that had advancement of the patellar tendon as part of the procedure demonstrated resolution of crouch gait with a greater improvement in the total range of knee motion, restoration of minimum stance-phase knee flexion to the normal range, and resolution of knee flexion contracture. This was a clear difference from the outcomes in the osteotomy-only group.
The implication for surgical decision-making is clear: to restore upright posture and the plantar flexion/knee extension couple, it is necessary to correct both the static fixed knee-flexion contracture and the dynamic quadriceps (patellar tendon) insufficiency. Individuals in the osteotomy-only group demonstrated improvement in knee flexion but still walked in a crouch (an average of 31° of knee flexion) as seen clinically subsequent to the procedure. Additionally, at the time of the postoperative gait analysis, only six of the sixteen knees in the osteotomy-only group were free of contracture compared with twenty-nine of the thirty-three knees in the group treated with both osteotomy and tendon advancement. Consistent with this observation is the finding that all three knees in which a flexion contracture recurred within six months after the surgery were in individuals who had undergone a distal femoral extension osteotomy in isolation. The importance of addressing patella alta and restoring active knee extension to avoid recurrence of contracture has been recognized previously21,24,25,35,37, but the reported information on this issue has been limited22,31.
If the knee is better but the patient is worse, or if gait pathology is decreased but the patient continues to walk in a crouch, the functional goal of surgery was not achieved. Our results indicate that pain scores, strength, level of community walking, and activities requiring knee function were improved or maintained following the surgery across all three groups. The single exception is that the group treated with both osteotomy and tendon advancement demonstrated a net loss of activities requiring knee motion. The reason for this loss is unclear. It does not appear to be related to pain. Despite the weakness noted across the patient groups in this study, strength at the time of the postoperative gait analysis had returned to baseline or had improved in most muscles. Only the group treated with both osteotomy and tendon advancement demonstrated a clinically meaningful improvement in walking ability postoperatively41,55, suggesting that the level of walking function is an important distinguishing variable. The fact that all groups maintained baseline function is noteworthy because the natural history of the walking function of individuals with cerebral palsy is one of decline12,56 and a treatment that stabilizes or improves a current walking pattern or ability represents a positive outcome56,57.
Other procedures were performed simultaneously with the distal femoral extension osteotomy and patellar tendon advancement procedures. In only seven subjects was distal femoral extension osteotomy and/or patellar tendon advancement performed in isolation or at the time of plate removal without additional procedures. The other procedures done in our patients were consistent with the overall principles of lengthening contracted musculotendinous units that limit the range of motion and correcting lever-arm dysfunction to optimize muscle performance, and they could have accounted for some of the improvement in the walking patterns reported in this study. For the most part, the concomitant operations were similar and balanced across the groups. The only exception is that there were many more hamstring lengthening procedures in the osteotomy-only group; yet the correction of crouch was least effective in that group.
One of the adverse and unanticipated findings in the study was the increase in anterior pelvic tilt noted across all groups. The increase was significant, and was thought to be clinically important as well, in both the tendon-advancement-only group and the group treated with both osteotomy and tendon advancement. Maximum hip extension in stance phase did not change. Initially, we thought that perhaps hip flexion contractures had not been adequately identified or addressed, since persisting psoas or rectus femoris contracture can worsen pelvic tilt postoperatively. Comparison of the individuals who had had an intramuscular psoas lengthening and rectus femoris transfer with those who had not had such a procedure demonstrated no difference between the two groups. The level of extensor strength (i.e., the strength of the gluteus maximus, quadriceps, and plantar flexors) was unchanged between the preoperative and postoperative evaluations, with the exception of improved knee extensor strength in the tendon-advancement-only group, and loss of strength was not thought to be a factor contributing to the worsened pelvic position. In a retrospective analysis of a small subset of patients, it is difficult to fully evaluate the multiple factors that might cause worsening of an anterior pelvic position. This finding should alert clinicians to design rehabilitation protocols that include strengthening, balance, and postural alignment both preoperatively and postoperatively.
Most clinicians believe that impairments of strength, balance, and selective motor control are more severe for children with a GMFCS level of III or IV, which may lead to the expectation that the results in these individuals may differ from the results for less severely affected children. The use of upper-extremity assistive devices also has an impact on both the GGI47,58 and energy expenditure59. The individuals with a GMFCS level of III or IV in this study tended to have a greater pelvic tilt range of motion, poorer hip extension, a more abnormal loading response at the knee, higher GGI scores, and greater net energy expenditure preoperatively. Despite this, the improvements in knee kinematics seen in the surgical groups as a whole were also seen in the children with a GMFCS level of III or IV, and following the surgery those children had improvements in the GGI and energy expenditure that were similar to the improvements for their counterparts with a GMFCS level of I or II; this suggests that the GMFCS level did not have a main effect.
Analysis of outcomes of complex surgical procedures should include an analysis of complications. Consistent with the findings in previous reports, neurovascular compromise accounted for the largest percentage of the complications related to the distal femoral extension osteotomies in this series (three complications after forty-nine procedures)26-30. Although the complication occurred throughout the time period of the study, increased experience with the performance of the procedures as well as changes in operative technique and immediate postoperative care have helped to decrease the rate of this complication. The need for a "threshold" level of experience to decrease complication rates in association with other complex procedures has been previously reported60. Briefly, these changes in care included performing the osteotomy as distally as possible, limiting the degree of knee flexion deformity treated to 25° to 30°, trimming any posterior bone prominence, avoiding excessive retraction, and immobilizing the lower extremity with the knee in 20° to 30° of flexion for the first three days postoperatively with use of a soft Robert Jones dressing. Including a component of shortening within the osteotomy or removal of a cuneiform wedge can also minimize stretch injury to neurovascular structures. Zimmerman et al. reported that femoral shortening was the key factor in their reported success in avoiding nerve injury28. One concern with this approach is whether the quadriceps mechanism can be tightened enough to restore effective quadriceps function. The success of the regimens used for our patients is evidenced by the fact that, in addition to the three patients in whom a neurovascular complication developed, ten additional patients had only transient sensory and/or motor symptoms that resolved by the end of their inpatient stay—that is, the development of additional, or progression of, neurovascular symptoms was circumvented. A high index of suspicion for neurovascular compromise must be maintained. Frequent monitoring in the postoperative period is essential, and any sign or symptom of nerve stretch must be evaluated and immediately addressed.
In contrast to the observations in previous reports, hyperextension (overcorrection) deformities were not common in our study. Approximately one year following the surgery, six of the seventy-three individuals walked with 5° to 15° of recurvatum, as measured with motion analysis. Radiographs of two knees confirmed 10° of recurvatum, which persisted. Both cases were in individuals who had had both distal femoral extension osteotomy and patellar tendon advancement.
Persistent knee pain, not fixation failure, was the most frequent complication associated with the patellar tendon advancement procedures, a finding that is in contrast with those in earlier studies24,25,34-38. The pain occurred in four individuals and was noted in both the group with patellar tendon advancement only and the group with both distal femoral extension osteotomy and patellar tendon advancement. In two of the four patients, patellofemoral symptoms developed after the surgery, whereas two had knee pain associated with knee motion or prolonged sitting. The pain resolved, without specific treatment, in three of the four knees within two years after the surgery. The fourth patient underwent arthroscopic surgery for patellar débridement and a tibial tubercle elevation (a Maquet procedure61), with subsequent relief of symptoms.
The fixation failure in this series occurred in the immediate postoperative period. This failure demonstrates the importance of managing spasms and tightness of both the knee flexors and the knee extensors as part of the procedure. If the patient has not had a previous selective dorsal rhizotomy or rectus femoris transfer, then we recommend that the rectus femoris transfer be performed at the same time as the patellar tendon advancement to allow the patella to assume its new "typical" position without restraint. Correcting minor knee flexion contractures of 5° to 10° (by distal femoral extension osteotomy) is also advised, as they appear to pose a risk to patellar tendon fixation. Injection of botulinum toxin is commonly included to decrease postoperative spasm in both the quadriceps and the hamstrings.
Our study does not support the cautions raised by Tachdjian regarding the potential distal migration of the patella with the tibial tubercle causing patellofemoral incongruity, early arthritis, and patella baja62. The method that we currently use for patellar advancement (aligning the inferior pole of the patella at the joint line of the knee) is an overcorrection of the typical anatomy seen radiographically as described by Koshino and Sugimoto53. Overcorrection of the patellar position has proven to be effective without any apparent associated complications. While we are concerned about the possibility of patella baja developing as a result of subsequent growth and we take some precautions to prevent this, we have not observed this complication in our patients.
Early review of the outcomes of these procedures led to the evolution of treatment principles and the management of the complications over the eleven-year span of the study. For example, persistence of crouch in the osteotomy-only group led us to stop performing the distal femoral extension osteotomy in isolation. Similarly, early in the study, small rubbery knee-flexion contractures were accepted until it was noted that the contractures did not stretch out following the surgery and that small contractures contributed to fixation-related complications after the patellar tendon advancement. The tendon-advancement-only group spans both treatment philosophies, which probably led to more variable results in that group.
The indications for these procedures have also evolved on the basis of our experience. The typical patient is an adolescent or young adult walking with a crouch gait pattern defined by insufficient knee extension in midstance. Most have weak ankle plantar flexors and a persistent crouch despite prior treatment, which typically included failed hamstring lengthening. While pain and patellar stress fractures are common, they are not the primary indications for surgery. The absolute indication for distal femoral extension osteotomy is a knee flexion contracture (typically in the range of 10° to 30°) found on physical examination. The absolute indication for patellar tendon advancement is quadriceps insufficiency (typically =10°) as indicated by an extensor lag on physical examination or a patient in whom a distal femoral extension osteotomy is performed. The goal of the procedures is to achieve full knee extension and adequate quadriceps length to maintain extension during the stance phase of gait. The extension osteotomy of the distal part of the femur is performed to compensate for the knee flexion contracture by creating an extension deformity of the distal part of the femur. The "correction" of patella alta typically represents an overcorrection not only to restore the function of the quadriceps, but also to compensate for ankle plantar flexor insufficiency and to correct crouch gait.
This study was limited by its retrospective nature. Forty-seven individuals who had undergone preoperative gait analysis prior to the operative event did not return for a postoperative analysis. This loss to follow-up could raise concerns that the results of the study are biased. While postoperative referral for gait analysis after multilevel surgery is common practice, it is not always possible for a variety of reasons, such as insurance issues, inconvenience to the family, or a clinical evaluation suggesting satisfactory function and thus minimizing the importance of a follow-up study. This was the case for thirty of the forty-seven individuals. Review of their medical records suggests that their ambulatory function was stable and comparable with that of the individuals who did have postoperative gait analysis. An additional eleven patients had ongoing follow-up, including postoperative videotapes that demonstrated resolution of the crouch. However, six individuals were not able to walk sufficiently to undergo gait analysis testing postoperatively. Of these, four did not regain their preoperative level of walking and two had lost their ability to walk prior to the surgery.
This study was also limited by a short duration of follow-up (range, seven to thirty-six months), which isolated the effects of surgical intervention to a time frame that allowed the individuals to recover but minimized the covariate effects of growth and other variables. The average duration of follow-up was consistent across all groups (approximately fourteen months). At the time that this article was written, sixty-five of the seventy-three patients had been seen within the previous five years. More than half were still being actively followed and had been seen within the previous two years. Only one of the subjects included in this study had a repeat procedure after the study was closed to further data inclusion, and that individual had had a distal femoral extension osteotomy only as the original procedure. No long-term complications related to patellar tendon advancement (such as patella baja) have been noted in the group who had a patellar tendon advancement procedure in this study. While we believe that the procedures are safe and effective, longer-term follow-up will provide further insight into whether the results reported here are maintained.
In conclusion, we evaluated the management of persistent crouch gait with use of surgical techniques that address structural contracture of the knee (distal femoral extension osteotomy) and knee extensor insufficiency (patellar tendon advancement). The results suggest that inclusion of patellar tendon advancement is necessary to achieve optimal results. The combined procedure (for individuals with a knee flexion contracture) or the patellar tendon advancement procedure alone (for individuals without a knee flexion contracture) can restore knee function to within typical limits in a heterogeneous group of individuals with regard to the GMFCS level. Function is improved or maintained as evidenced by relief of pain, restoration of preoperative levels of strength, improvement in the community walking level, and improvement in skills that require more complex knee function. The study suggests that the procedures are safe and effective. Because deformity is less likely to recur in individuals in whom the growth rate has slowed or growth is complete, it appears reasonable to anticipate that these adolescents will be able to maintain walking ability into adulthood.
Tables presenting demographic data, previous surgical procedures, pelvic and hip motion values, and functional ability data are available with the electronic versions of this article, on our web site at (go to the article citation and click on "Supplementary Material") and on our quarterly CD/DVD (call our subscription department, at 781-449-9780, to order the CD or DVD). 
Note: The authors thank Patrick Cavanaugh, RN, BSN, Tina Given-Collins, RN, and Adam Rozumalski, MS, for their participation in the reliability study of the Koshino method of radiographic measurement.
Sutherland DH, Cooper L. The pathomechanics of progressive crouch gait in spastic diplegia. Orthop Clin North Am.1978;9:143-54.9143
1978
[PubMed]
Sutherland DH, Davids JR. Common gait abnormalities of the knee in cerebral palsy. Clin Orthop Relat Res.1993;288:139-47.288139
1993
Novacheck TF, Trost JP, Schwartz MH. Intramuscular psoas lengthening improves dynamic hip function in children with cerebral palsy. J Pediatr Orthop.2002;22:158-64.22158
2002
[CrossRef]
Gage JR, editor. The treatment of gait problems in cerebral palsy. 2nd ed. London: Mac Keith Press; 2004. Treatment principles for crouch gait; p 382-97.
2004
Schwartz M, Lakin G. The effect of tibial torsion on the dynamic function of the soleus during gait. Gait Posture.2003;17:113-8.17113
2003
[CrossRef]
Hicks J, Arnold A, Anderson F, Schwartz M, Delp S. The effect of excessive tibial torsion on the capacity of muscles to extend the hip and knee during single-limb stance. Gait Posture.2007;26:546-52.26546
2007
[CrossRef]
DeLuca PA, Ounpuu S, Davis RB, Walsh JH. Effect of hamstring and psoas lengthening on pelvic tilt in patients with spastic diplegic cerebral palsy. J Pediatr Orthop.1998;18:712-8.18712
1998
[CrossRef]
Hicks JL, Schwartz MH, Arnold AS, Delp SL. Crouched postures reduce the capacity of muscles to extend the hip and knee during the single-limb stance phase of gait. J Biomech.2008;41:960-7.41960
2008
[CrossRef]
Perry J, Antonelli D, Ford W. Analysis of knee joint forces during flexed knee stance. J Bone Joint Surg Am.1975;57:961-7.57961
1975
Rosenthal RK, Levine DB. Fragmentation of the distal pole of the patella in spastic cerebral palsy. J Bone Joint Surg Am.1977;59:934-9.59934
1977
Topoleski TA, Kurtz CA, Grogan DP. Radiographic abnormalities and clinical symptoms associated with patella alta in ambulatory children with cerebral palsy. J Pediatr Orthop.2000;20:636-9.20636
2000
[CrossRef]
Andersson C, Mattsson E. Adults with cerebral palsy: a survey describing problems, needs, and resources, with special emphasis on locomotion. Dev Med Child Neurol.2001;43:76-82.4376
2001
[CrossRef]
Murphy KP, Molnar GE, Lankasky K. Medical and functional status of adults with cerebral palsy. Dev Med Child Neurol.1995;37:1075-84.371075
1995
[CrossRef]
Anderson FC, Pandy MG. Individual muscle contributions to support in normal walking. Gait Posture.2003;17:159-69.17159
2003
[CrossRef]
Arnold AS, Anderson FC, Pandy MG, Delp SL. Muscular contributions to hip and knee extension during the single limb stance phase of normal gait: a framework for investigating the causes of crouch gait. J Biomech.2005;38:2181-9.382181
2005
[CrossRef]
Kimmel SA, Schwartz MH. A baseline of dynamic muscle function during gait. Gait Posture.2006;23:211-21.23211
2006
[CrossRef]
Schwartz MH, Kimmel SA. The effect of crouch gait on sagittal plane induced accelerations. Read at the Annual Meeting of the European Society of Movement Analysis for Adults and Children; 2005 Sept 22-24; Barcelona, Spain.
2005
Arnold AS, Blemker SS, Delp SL. Evaluation of a deformable musculoskeletal model for estimating muscle-tendon lengths during crouch gait. Ann Biomed Eng.2001;29:263-74.29263
2001
[CrossRef]
Arnold AS, Liu MQ, Schwartz MH, Ounpuu S, Delp SL. The role of estimating muscle-tendon lengths and velocities of the hamstrings in the evaluation and treatment of crouch gait. Gait Posture.2006;23:273-81.23273
2006
[CrossRef]
Delp SL, Arnold AS, Speers RA, Moore CA. Hamstrings and psoas lengths during normal and crouch gait: implications for muscle-tendon surgery. J Orthop Res.1996;14:144-51.14144
1996
[CrossRef]
Lotman DB. Knee flexion deformity and patella alta in spastic cerebral palsy. Dev Med Child Neurol.1976;18:315-9.18315
1976
[CrossRef]
Beals RK. Treatment of knee contracture in cerebral palsy by hamstring lengthening, posterior capsulotomy, and quadriceps mechanism shortening. Dev Med Child Neurol.2001;43:802-5.43802
2001
[CrossRef]
Osgood RB. A method of osteotomy of the lower end of the femur in cases of permanent flexion of the knee joint. Am J Orthop Surg.1913;s2-11:336-46.s2-11336
1913
Chandler FA. Re-establishment of normal leverage of the patella in knee flexion deformity in spastic paralysis. Surg Gynecol Obstet.1933;57:523-7.57523
1933
Chandler FA. Patellar advancement operation: a revised technique. J Int Coll Surg.1940;3:433-7.3433
1940
Bain AM. Treatment of flexion contracture of the knee following poliomyelitis. J Trop Med Hyg.1966;69:285-90.69285
1966
Leong JC, Alade CO, Fang D. Supracondylar femoral osteotomy for knee flexion contracture resulting from poliomyelitis. J Bone Joint Surg Br.1982;64:198-201.64198
1982
Zimmerman MH, Smith CF, Oppenheim WL. Supracondylar femoral extension osteotomies in the treatment of fixed flexion deformity of the knee. Clin Orthop Relat Res.1982;171:87-93.17187
1982
Grant AD, Small RD, Lehman WB. Correction of flexion deformity of the knee by supracondylar osteotomy. Bull Hosp Jt Dis Orthop Inst.1982;42:28-38.4228
1982
Asirvatham R, Mukherjee A, Agarwal S, Rooney RJ, Ellis RD, Watts HG. Supracondylar femoral extension osteotomy: its complications. J Pediatr Orthop.1993;13:642-5.13642
1993
Gage JR, Stout JL, Novacheck TF, Matsuo K. Distal femoral extension osteotomy for treatment of persistent crouch gait. Read at the Annual Meeting of the European Pediatric Orthopaedic Society; 2002 Apr 17-20; Istanbul, Turkey.
2002
Kramer A, Stevens PM. Anterior femoral stapling. J Pediatr Orthop.2001;21:804-7.21804
2001
[CrossRef]
Stevens PM. Guided growth for angular correction: a preliminary series using a tension band plate. J Pediatr Orthop.2007;27:253-9.27253
2007
[CrossRef]
Keats S, Kambin P. An evaluation of surgery for the correction of knee-flexion contracture in children with cerebral spastic paralysis. J Bone Joint Surg Am.1962;44:1146-54.441146
1962
Cleveland M, Bosworth DM. Surgical correction of flexion deformity of knees due to spastic paralysis. Surg Gynecol Obstet.1936;63:659-64.63659
1936
Eggers GW. Surgical division of the patellar retinacula to improve extension of the knee joint in cerebral spastic paralysis. J Bone Joint Surg Am.1950;32:80-6.3280
1950
Roberts WM, Adams JP. The patellar-advancement operation in cerebral palsy. J Bone Joint Surg Am.1953;35:958-66; passim.35958
1953
Baker LD. A rational approach to the surgical needs of the cerebral palsy patient. J Bone Joint Surg Am.1956;38:313-23.38313
1956
Ferraretto I, Machado PO, Filho EL, Selber P. Preliminary results of patellar tendon shortening as a salvage procedure for crouch gait in cerebral palsy [abstract]. In: Proceedings of the Annual Meeting of the Pediatric Orthopaedic Society of North America; 2000 May 1-4. Vancouver, BC, Canada. p 71.
2000
Kaastad TS. Painful knee and crouch gait treated by patellar advancement through shortening of the patellar tendon in ambulating patients with cerebral palsy. Read at the Annual Meeting of the European Pediatric Orthopaedic Society; 2002 Apr 17-20; Instanbul, Turkey.
2002
Novacheck TF, Stout JL, Tervo R. Reliability and validity of the Gillette Functional Assessment Questionnaire as an outcome measure in children with disabilities. J Pediatr Orthop.2000;20:75-81.2075
2000
[CrossRef]
Palisano R, Rosenbaum P, Walter S, Russell D, Wood E, Galuppi B. Development and reliability of a system to classify gross motor function in children with cerebral palsy. Dev Med Child Neurol.1997;39:214-23.39214
1997
[CrossRef]
Kendall HO, Kendall FP, Wadsworth GE. Muscles, testing and function. 2nd ed. Baltimore: Williams and Wilkins; 1971. p 2-15, 145-239.
1971
Norkin CC, White DJ. Measurement of joint motion: a guide to goniometry. 2nd ed. Philadelphia: Davis; 1985. p 117-77.
1985
Lee K-C, Carson L, Kinnin E, Patterson V. The Ashworth scale: a reliable and reproducible method of measuring spasticity. J Neurol Rehabil.1989;3:205-9.3205
1989
Daltroy LH, Liang MH, Fossel AH, Goldberg MJ. The POSNA pediatric musculoskeletal functional health questionnaire: report on reliability, validity, and sensitivity to change. Pediatric Outcomes Instrument Development Group. Pediatric Orthopaedic Society of North America. J Pediatr Orthop.1998;18:561-71.18561
1998
Schutte LM, Narayanan U, Stout JL, Selber P, Gage JR, Schwartz MH. An index for quantifying deviations from normal gait. Gait Posture.2000;11:25-31.1125
2000
[CrossRef]
Schwartz MH, Trost JP, Wervey RA. Measurement and management of errors in quantitative gait data. Gait Posture.2004;20:196-203.20196
2004
[CrossRef]
Stuberg WA, Fuchs RH, Miedaner JA. Reliability of goniometric measurements of children with cerebral palsy. Dev Med Child Neurol.1988;30:657-66.30657
1988
[CrossRef]
Enwemeka CS. Radiographic verification of knee goniometry. Scand J Rehabil Med.1986;18:47-9.1847
1986
Mayerson NH, Milano RA. Goniometric measurement reliability in physical medicine. Arch Phys Med Rehabil.1984;65:92-4.6592
1984
Schwartz MH, Koop SE, Bourke JL, Baker R. A nondimensional normalization scheme for oxygen utilization data. Gait Posture.2006;24:14-22.2414
2006
[CrossRef]
Koshino T, Sugimoto K. New measurement of patellar height in the knees of children using the epiphyseal line midpoint. J Pediatr Orthop.1989;9:216-8.9216
1989
Insall J, Salvati E. Patella position in the normal knee joint. Radiology.1971;101:101-4.101101
1971
Haley SM, Fragala-Pinkham MA. Interpreting change scores of tests and measures used in physical therapy. Phys Ther.2006;86:735-43.86735
2006
Bottos M, Feliciangeli A, Sciuto L, Gericke C, Vianello A. Functional status of adults with cerebral palsy and implications for treatment of children. Dev Med Child Neurol.2001;43:516-28.43516
2001
[CrossRef]
Gough M, Eve LC, Robinson RO, Shortland AP. Short-term outcome of multilevel surgical intervention in spastic diplegic cerebral palsy compared with the natural history. Dev Med Child Neurol.2004;46:91-7.4691
2004
[CrossRef]
Romei M, Galli M, Motta F, Schwartz M, Crivellini M. Use of the normalcy index for evaluation of gait pathology. Gait Posture.2004;19:85-90.1985
2004
[CrossRef]
Stout JL, Koop SE. Energy expenditure in cerebral palsy. In: Gage JR, editor. The treatment of gait problems in cerebral palsy. 2nd ed. London: Mac Keith Press; 2004. p 146-64.
2004
Dahl MT, Gulli B, Berg T. Complications of limb lengthening. A learning curve. Clin Orthop Relat Res.1994;301:10-8.30110
1994
Maquet P. Advancement of the tibial tuberosity. Clin Orthop Relat Res.1976;115:225-30.115225
1976
Tachdjian MO. Pediatric orthopedics. 2nd ed. Philadelphia: W.B. Saunders; 1990. p 1570.
1990