Return to Alta: Patellar Height Increases Over Time After Tibial Tubercle Osteotomy With Distalization

Abstract

Background:

For patients who have patellar instability with patella alta, distalizing the patella via tibial tubercle osteotomy (TTO-d) is one method used to surgically restore stability. Although this procedure is generally associated with favorable outcomes, the effects on maintaining postoperative patellar height remain inadequately understood.

Purpose/Hypothesis:

This study aimed to assess changes in patellar height after TTO-d in skeletally mature patients. It was hypothesized that the patella would eventually return to a state of alta after TTO-d.

Study Design:

Case series; Level of evidence, 4.

Methods:

Skeletally mature patients who underwent a TTO-d procedure with a follow-up lateral knee radiograph at ≥3 months after surgery were included. Caton-Deschamp (CD) and Blackburne-Peele (BP) indices were calculated at various time intervals: preoperative, immediate postoperative, 3-month, 6-month, and ≥1-year follow-up on lateral knee radiographs as available. Paired Student t tests were used to compare patellar height indices at different time points.

Result:

A total of 35 knees in 28 patients were included in this study, 20 of which had imaging ≥1 year postoperatively. The mean age of the cohort was 22.3 ± 7.95 years (range, 15.5-43.0 years), mean body mass index was 28.0 ± 7.86, mean tibial tubercle distalization was 10.8 ± 4.83 mm at time of surgery, and mean follow-up using the latest available lateral knee radiographs occurred at 2.83 ± 2.96 years postoperatively. Mean preoperative CD and BP indices were 1.46 ± 0.193 and 1.24 ± 0.196, respectively, which decreased to 1.13 ± 0.189 (P < .001) and 0.855 ± 0.209 (P < .001) on immediate postoperative films. In those with imaging ≥1 year out from surgery, mean CD and BP indices increased to 1.34 ± 0.196 (P < .001) and 1.13 ± 0.131, respectively. Additionally, 88% (15/17) of patients with a minimum 1-year radiographic follow-up transitioned from normal CD and BP indices immediately postoperatively to patella alta at latest follow-up.

Conclusion:

Nearly all patients who underwent TTO-d had progressive increases in both CD and BP indices over time, suggesting a return to a state of patella alta over time. Further studies are needed to compare clinical outcomes of patients who have a confirmed return to alta versus those who do not in order to better understand the clinical significance of this observation.

Patellar instability can result from soft tissue abnormalities, limb malalignment, or bony deformities. ²³ Patella alta is a common osseous contributor, which is classically defined by elevations in various patellar height indices such as the Caton-Deschamps (CD) index, the Insall-Salvati ratio, and the Blackburne-Peel (BP) index.^3,4,8 Patella alta represents a form of patellofemoral malalignment, altering the mechanics of knee extension and increasing stress on the patellofemoral joint, and is linked to a higher risk for developing osteoarthritis in the patellofemoral joint.^19,22

First-line treatment for these patients is typically nonoperative management with bracing and physical therapy. ²³ Although this is the first-line treatment, instances of recurrent instability and dislocations after an initial episode in patients with patella alta can range from 37% to 86% depending on additional risk factors such as younger age and higher grade of trochlear dysplasia.^14,23 In patients with failed nonoperative management who experience recurrent symptoms of instability, surgical intervention may be warranted. Numerous surgical modalities for correcting this pathology have been described, and tibial tubercle osteotomy with distalization (TTO-d) is commonly used.^13,17 This procedure involves freeing up the tibial tubercle and displacing it distally to bring the patella into the trochlear groove to decrease subluxation and dislocation events and restore contact pressures at normal physiologic ranges of motion.^17,20 Improved pain and stability outcomes are associated with this surgical procedure in addition to a low overall complication rate when performed during proper indications.^1,10,18 Despite literature showing improved outcomes after TTO-d alone for recurrent instability, current controversies exist regarding TTO-d with or without medialization compared with medial patellofemoral ligament (MPFL) reconstruction alone even in patients with patella alta.^2,5,6

Although studies have shown a decrease in patellar height at time zero, reduction in instability events, and improved patient-reported outcomes after this procedure, little is known about how the patellar tendon may respond to TTO-d long term.^10,12 With other surgeries performed around the patellar tendon, there are reports of decreased patellar height demonstrating patellar tendon shortening after surgeries involving the extensor mechanism. ⁹ However, in our experience, we have noted a return to a state of patella alta on follow-up imaging in patients who underwent TTO-d. The purpose of this study was to determine whether the patellar height changes over time after TTO-d. Our hypothesis was that compared with the original patellar height correction obtained after TTO-d, the patellar height would increase in the postoperative follow-up and would not maintain the original correction.

Methods

Cohort Identification and Characterization

After institutional review board (IRB) approval was granted (IRB No. 71733), records of all skeletally mature patients who underwent a tibial tubercle osteotomy as identified by CPT code 27418 between January 1, 2011, and December 31, 2023, across 5 orthopaedic surgeons were pulled from an institutional electronic data warehouse. Patients’ charts were then reviewed by 2 fourth-year medical students (B.T.J. and C.D.R.H.) to ensure that the patients had (a) a confirmed primary distalizing tibial tubercle osteotomy (with or without medialization) and MPFL reconstruction for unstable patella alta and (b) lateral knee radiographs at least 3 months postoperatively. The study excluded (a) individuals without a follow-up lateral knee radiograph ≥3 months out from surgery (n = 8), (b) those undergoing a revision procedure (n = 5), and (c) those with traumatic complete patellar dislocations (n = 2) (Figure 1). Patients who underwent a hinge osteotomy were also excluded, and the amount of distalization at the time of surgery was recorded in the operative note by the attending surgeon as an approximation.

Figure 1.

Flow diagram demonstrating inclusion and exclusion criteria. TTO-d, tibial tubercle osteotomy with distalization.

Imaging Review

CD and BP indices were calculated by 2 trained fourth-year medical students (B.T.J. and C.D.R.H.) for preoperative, immediate postoperative, and 3-month, 6-month, and ≥1-year postoperative results as available (Figure 2). Reviewers were trained by the senior orthopaedic surgeon (S.K.A.) and performed measurements independent of one another at separate time points. Patella alta was defined as a CD index ≥1.30 and BP index ≥1.00, as previously established.^3,4 Given that rehabilitation protocols varied among the surgeons, immediate postoperative weightbearing images were defined as standing lateral images of the operative knee <6 weeks after surgery at a knee flexion angle of 20° to 30°.

Figure 2.

Measurements of Caton-Deschamp (CD) and Blackburne-Peel (BP) indices in a 17-year-old skeletally mature female (A) preoperatively (CD = 2.19; BP = 2.10), (B) immediately after tibial tubercle osteotomy with approximately 30 m of distalization (CD = 1.29; BP = 0.94), and (C) at 1 year out from surgery (CD = 1.81; BP = 1.38).

Statistical Analysis

All analyses were performed using SPSS Version 27 (IBM Corp). Intraclass correlation coefficients (ICCs) for both CD and BP indices were calculated via a consistency type 2-way mixed model. Paired Student t tests were conducted to compare the mean patellar height indices at various postoperative time points with the preoperative and immediate postoperative measurements. Additionally, patellar heights measured by CD and BP indices were compared in the immediate postoperative image versus the 3-month, 6-month, and ≥1-year images to determine the percentage of increase or decrease postoperatively at those specific time intervals. Bonferroni-corrected P values were used to account for multiple comparisons, with significance set at P < .05.

Results

A total of 35 knees (28 patients) were included in the final analysis (Table 1). The mean age of the cohort was 22.3 ± 7.95 years (range, 15.5-43.0 years), and the mean body mass index was 27.5 ± 7.63. The tibial tubercle was displaced distally by a mean of 10.8 ± 4.83 mm at the time of surgery per the surgeon’s operative note. In patients with available imaging ≥1 year after surgery (n = 20/35 knees), the mean follow-up was 2.83 ± 2.96 years. ICCs were 0.800 (range, 0.724-0.878) for CD index and 0.959 (range, 0.868-0.989) for BP index, indicating strong interrater agreement. The most common reasons for patients returning to clinic at 1 year were knee pain (40%), hardware discomfort (20%), recurrent subjective feelings of instability without frank dislocation (20%), and routine follow-up or contralateral knee pain (20%).

Table 1.

Patient Demographic Characteristics and Surgical Data ^a

Variables
Age, y	22.3 ± 7.95
Sex
Male	9 (32)
Female	19 (68)
Body mass index	27.5 ± 7.63
Amount distalized, mm	10.8 ± 4.83
Follow-up time in patients with imaging ≥1 y postoperative, y	2.83 ± 2.96
Reason for return to clinic ≥1 y (n = 20)
Knee pain	8 (40)
Hardware discomfort	4 (20)
Recurrent instability	4 (20)
Routine follow-up or contralateral knee pain	4 (20)

^aData are expressed as mean ± SD or n (%).

Mean CD index was 1.46 ± 0.193 before surgery, 1.13 ± 0.189 immediately after surgery (P < .001), and 1.34 ± 0.196 in the ≥1-year group (P = .17 compared with preoperative state; P≤ .001 compared with immediate postoperative state) (Table 2; Figure 3). Mean BP index was 1.24 ± 0.200 before surgery, 0.855 ± 0.209 immediately after surgery (P < .001), and 1.13 ± 0.216 in the ≥1-year group (P = .19 compared with preoperative state; P < .001 compared with immediate postoperative state). In a 15-knee subgroup comparison of measurements between patients who had imaging at both the 3-month and the ≥1-year intervals, 3-month CD and BP indices were 1.26 ± 0.175 and 1.03 ± 0.174, respectively, whereas the ≥1-year CD and BP indices were 1.33 ± 0.220 (P = .273) and 1.12 ± 0.240 (P = .235).

Table 2.

Change in CD and BP Indices Over Time ^a

	Caton-Deschamps (CD) Index			Blackburne-Peel (BP) Index
Time Period	Mean ± SD	P (pre) b	P (post) c	Mean ± SD	P (pre) b	P (post) c
Preoperative (n = 33)	1.46 ± 0.193	—	<.001	1.24 ± 0.200	—	<.001
Postoperative (n = 35)	1.13 ± 0.189	<.001	—	0.855 ± 0.209	<.001	—
3 mo postoperative (n = 27)	1.25 ± 0.170	<.001	<.001	1.03 ± 0.183	<.001	<.001
6 mo postoperative (n = 18)	1.33 ± 0.224	.14	<.001	1.10 ± 0.220	.01	<.001
≥1 y postoperative (n = 20)	1.34 ± 0.196	.17	<.001	1.13 ± 0.216	.19	<.001

^aP values compare preoperative, 3-month postoperative, 6-month postoperative, or ≥1-year postoperative state to the immediate postoperative state. Boldface indicates statistical significance.

^bPaired 2-tailed t test comparing CD or BP at the specified time point to the preoperative value.

^cPaired 2-tailed t test comparing CD or BP at the specified time point to the immediate postoperative value.

Figure 3.

Change in Caton-Deschamps (CD) and Blackburne-Peel (BP) indices over time. *P < .05 for both CD and BP. **P < .05 for CD only.

Of the 35 knees studied, all but 1 knee had an increase in patellar height at the 3-month and 6-month time intervals (Table 3). All knees at the 1-year mark had an increase in patellar height on CD and BP indices by a mean of 25% and 44%, respectively. At the ≥1-year imaging follow-up, 88% (15/17) of the knees that had a normalized patellar height on immediate postoperative films returned to a state of patella alta on both CD and BP indices. The 3 other knees (n = 20) with imaging ≥1 year out were still considered patella alta by CD and BP indices on the immediate postoperative film and were not included in this analysis.

Table 3.

Percentage Change in CD and BP Ratios Comparing Immediate Postoperative Radiographs to 3-Month, 6-Month, and ≥1-Year Postoperative Time Points ^a

	% Change in CD Ratio After TTO-d	% Change in BP Ratio After TTO-d
3 mo postoperative (n = 27)	12.7 ± 14.4	24.3 ± 24.5
6 mo postoperative (n = 18)	20.5 ± 22.6	26.1 ± 28.2
≥1 y postoperative (n = 20)	24.6 ± 23.0	44.4 ± 39.5

^aValues are expressed in percentages as mean ± SD. BP, Blackburne-Peel; CD, Caton-Deschamps; TTO-d, tibial tubercle osteotomy with distalization.

Discussion

Patients who underwent TTO-d had a progressive increase in both CD and BP indices over time. Although the limitations with the present study allow only speculation as to why this occurred, the finding highlights the need for rigorous biomechanical investigations to determine a potential mechanism for this observation. Future studies should compare clinical outcomes of patients who have a confirmed return to patella alta versus those who do not in order to better understand the clinical significance of this observation.

Leite et al ¹² investigated long-term clinical and radiographic outcomes after TTO-d in 31 knees (25 patients). The investigators showed that in the immediate postoperative period, mean CD index decreased from 1.37 to 1.02 across 31 knees, and at a later period with a minimum 1-year follow-up, the CD mean remained 1.02. Although we also found a decrease in the immediate postoperative period, the current study identified a consistent increase in patellar height over time from a mean preoperative CD index of 1.46 to a postoperative CD of 1.13, and eventually 1.34 at ≥1 year. These findings may be due to variation between patient populations or study protocols, given that surgical procedures were similar, but further investigations are needed to explain the discrepancy seen between our 2 studies.

Consistent with the present study’s findings, multiple investigations have demonstrated that TTO-d decreases patellar height as approximated by various radiographic measurements in the immediate postoperative period.^10,15 Uppstrom et al ²¹ showed that patellar height remained significantly below the preoperative height on magnetic resonance images at a mean follow-up of 9.6 months after TTO-d. Although we also found that patellar height remained below the preoperative height, our interval imaging showed a progressive increase after an initial postoperative image. Mayer et al ¹⁵ found that a TTO-d procedure performed in conjunction with patellar tendon tenodesis resulted in decreased patellar height without any substantially different patient-reported outcomes between those with and without tenodesis at a minimum of 5 years postoperatively. However, their study did not investigate how the patellar height changed over time with follow-up radiographs. The current study did not evaluate patients who underwent TTO-d with tenodesis, and further studies would be needed to determine whether a tenodesis prevents an increase in postoperative patellar height compared with TTO-d alone.

Although the surgical management of recurrent patellar instability in those with patella alta is typically associated with good outcomes, some controversy exists in that medialization alone may be sufficient.^10,12,24 Despite some biomechanical concerns that an MPFL reconstruction alone in patella alta may be prone to failure from an induced anisometry, ² Dennis et al ⁵ recently found that isolated MPFL reconstruction resulted in improved patient outcomes regardless of patellar height. Furthermore, a systematic review from Guevel et al, ⁶ investigating outcomes in patients undergoing TTO plus MPFL reconstruction versus MPFL reconstruction alone in those with an elevated tibial tubercle trochlear groove offset >20 mm, found no significant difference between the 2 groups. Further study is needed to determine whether the change in patellar height observed in the current study correlates with clinical outcomes and whether lateral patellar drift in MPFL reconstructions occurs over time.

Limitations

This study has some limitations. First, not all patients undergoing TTO-d were available for >1-year follow-up, and this limits the generalizability of our findings. Second, measuring patellar height is variable in nature and can be affected by multiple factors such degree of knee flexion, upright patient positioning, and tibial slope, leading to variable index values.^7,11,16 Third, this study did not include patient-reported outcomes, so the clinical significance of these findings remains unknown. Fourth, although imaging with weightbearing at a specified knee flexion angle was performed at the immediate postoperative scan, patient comfort may have prevented complete quadriceps contraction in some cases, which could artificially lower patellar height indices. This may be accounted for in future investigations by repeating lateral knee radiographs in various positions such as standing and decubitus positions with varying quadriceps contraction. Fifth, variations in surgical technique with regard to TTO-d and MPFL reconstruction may exist and affect apparent patellar height. Sixth, small sample size and retrospective design of the study may limit generalizability of the findings. Seven, significant selection bias exists among the patients included in the study, as not all patients who underwent TTO-d returned for follow-up imaging ≥3 months out.

Conclusion

All but 1 patient who underwent TTO-d had a progressive increase in both CD and BP indices over time compared with their immediate postoperative patellar height. Further research is needed to confirm this these finding and to determine the amount of increase in patellar height that may correspond with a decrease in patient-reported outcomes in order to better understand the clinical significance of this observation.