While the importance of the patella in the normal function of the knee is currently well recognized , patellectomy was a common and accepted treatment for patellofemoral pathology, most commonly for patellar fracture or severe osteoarthritis, until the last few decades . Patients with prior patellectomy can nonetheless develop tibiofemoral arthrosis, for which the mainstay of treatment is TKA. However, comparison studies on these patients suggest inferior function after TKA relative to patients with intact patellae [10, 13, 14]. Quadriceps and hamstring torque and strength , rate of complication, and knee outcome score [13, 14] are all reportedly worse in patients with TKA who have had previous patellectomy.
Although patellectomized patients have outcomes after TKA that are less likely to be satisfactory and have a higher rate of complications, improvements in pain and function can be achieved [10, 13, 15]. However, the studies on the outcomes of TKA after previous patellectomy [2, 5, 10, 11, 13-16, 18] have all been small (the largest studies numbering 22 patients) with variations in the types of prosthesis and a mixture of primary and revision cases. These studies also vary in their use of outcome measures, making comparison between results difficult. Thus, while the literature suggests worse function, it remains unclear to what degree these patients improve.
We therefore asked whether (1) TKA provides improvements in clinical outcome scores in patellectomized knees and (2) clinical outcomes of TKA in patellectomized knees are comparable to those in knees with intact patellae.
Patients and Methods
We retrospectively reviewed a prospectively collected database of TKAs performed at our institution, identifying 60 primary TKAs performed on previously patellectomized knees in 57 patients between 1984 and 2009. The indications for prior patellectomy were patellofemoral dysfunction (34), trauma (15), and infection (two for tuberculosis, one for hardware infection, one for osteomyelitis). Two patients had insufficient data collected preoperatively on two knees (one of the two patients remained included for the opposite patellectomized knee, for which there was sufficient preoperative data). Six patients were lost to followup. The remaining 52 knees in 50 patients were followed for a minimum of 24 months (mean, 69 months; range, 24-204 months) (Table 1). There were eight deaths; as all eight had been followed for at least 24 months, the data from these patients were included in the analysis. All TKAs were performed for pain secondary to tibiofemoral arthrosis. Forty-eight of 52 knees received a posterior-stabilized (PS) implant, with the remaining four receiving cruciate-retaining (CR) implants. The implants used were four AMK® (DePuy Orthopaedics Inc, Warsaw, IN, USA), two Congruency® (DePuy), two Coordinate® (DePuy), six Genesis® I (Smith & Nephew Inc, Memphis, TN, USA), 17 Genesis® II (Smith & Nephew), seven Genesis® II Oxinium (Smith & Nephew), one Miller/Galante® (Zimmer, Inc, Warsaw, IN, USA), two Miller-Galante II® (Zimmer), one Natural-Knee® I (Zimmer), one Natural-Knee® II (Zimmer), two PCA® (Howmedica Inc, Rutherford, NJ, USA), two SAL® (Sulzer AG, Winterthur, Switzerland), and five Sigma® (DePuy). No patients were recalled specifically for this study; all data were obtained from medical records.
Using the same database, we created a set of 52 control knees in 52 patients with intact patellae matched for age, sex, implant, and surgical year (Table 1).
Patients were evaluated both preoperatively and at latest clinic visit, with followup postoperatively at 6 weeks, 3 months, 6 months, 1 year, 2 years, and then annually or biannually thereafter. Radiographs were taken and assessed at each visit. Knees were examined on all parameters used in the Knee Society Clinical Rating System . Measurement of ROM and extensor lag was performed either with a goniometer or by clinical judgment at the clinician’s discretion. Clinical outcome scores used were the Knee Society score (KSS) , SF-12 score , and WOMAC score . Knee scores were calculated according to the KSS, with a score for pain with a maximum of 50 and a total score including function and knee scores with a maximum of 200. For TKAs performed after 1995, SF-12 and WOMAC questionnaires were filled out by patients both preoperatively and at latest clinic visit and the scores calculated. SF-12 and WOMAC scores were available for 31 of 52 knees in the study group and 29 of 52 knees in the control group.
We performed statistical analyses to identify differences in mean preoperative and postoperative total outcome scores within groups using the paired Student’s t-test and in mean total outcome scores between groups using Student’s t-test. A Bonferroni correction was used such that the overall alpha equaled 0.05. All statistical analyses were performed using Excel® (Microsoft Corp, Redmond, WA, USA).
We observed increases in the mean WOMAC score and the mean KSS in both the control and patellectomized groups after TKA (Table 2). The mean overall WOMAC score in the control group improved (p < 0.001) from 41.8 (range, 7.5-72.4) preoperatively to 69.1 (range, 17.0-100.0) postoperatively, with an increase seen in all three subcategories of pain, stiffness, and function (Table 3). It similarly improved (p < 0.001) in the patellectomized group from 35.8 (range, 5.2-62.2) preoperatively to 61.3 (range, 17.5-96.2) postoperatively, also with increases in all three subcategories. The mean total KSS in the control group improved (p < 0.001) from 80.4 (range, 4.0-143.0) preoperatively to 161.4 (range, 69.0-200.0) postoperatively, with an increase in the pain, function, and knee scores. It similarly improved (p < 0.001) in the patellectomized group from 76.9 (range, 5-134) preoperatively to 136.8 (range, 7-199) postoperatively, with an increase in all subcategory scores. The mean SF-12 score was not different in either the control group (p = 0.024) or the patellectomized group (p = 0.032) after TKA.
The mean postoperative overall WOMAC scores were similar (p = 0.203) between the control and patellectomized groups. However, the mean total postoperative KSS was lower (p = 0.001) in the patellectomized group than in the control group. With respect to the mean SF-12 score, there was no difference (p = 0.194) between the control and patellectomized groups.
In the control group, seven of the 52 knees (13.5%) had subsequent revision, one for patellofemoral arthritis (1.9%), one for infected hardware (1.9%), and five for instability or polyethylene wear requiring polyethylene exchange or upsizing (9.6%). In comparison, the patellectomized group had 10 knees go on to require revision (19.2%), one for pain (1.9%), two for infected hardware (3.8%), one for sepsis (1.9%), two for instability requiring polyethylene upsizing (3.8%), and four for polyethylene wear requiring polyethylene exchange (7.7%). Of the two revisions for infected prosthesis, one knee ultimately went on to an arthrodesis due to recurrent infection and pain. In addition to the implant revisions, one patient required an extensor mechanism reconstruction for extensor lag, while two patients had postoperative manipulation under anesthesia for knee stiffness, with successful treatment of all three.
Patellectomy results in loss of mechanical leverage to the knee extensor mechanism but does not prevent development of tibiofemoral arthrosis. Historically, studies on TKA in patellectomized knees have been small and have shown mixed results [2, 5, 10, 11, 13-16, 18]. We therefore asked whether (1) TKA provides improvements in clinical outcome scores in patellectomized knees and (2) clinical outcomes of TKA in patellectomized knees are comparable to those in knees with intact patellae.
Certain limitations of our study should be noted. First, our study was retrospective and nonrandomized. However, the two groups seemed well-matched according to demographic data (Table 1). Second, although this is the largest reported series of postpatellectomy TKA, it remains small compared to series of intact-patella TKA. Third, multiple implants were used, with varying lengths of followup. The groups were too small to perform a multivariable analysis to determine whether any given implant was associated with lower scores. Fourth, our patellectomized group contained bilateral TKAs in two patients while our control group had none. While the functional scores will to some degree reflect the function of both knees, bilateral arthroplasties might lead to worse function. However, two patients would not affect the means to any major degree. Finally, multiple comparisons also risked the possibility of differences being found by chance, although this was mitigated by use of a Bonferroni correction.
In our study, patellectomized patients who underwent TKA showed improvement as measured by the WOMAC score and the KSS. Comparison of our results against other postpatellectomy TKA studies is difficult, given the lack of common outcome measures (Table 4). The mean ± SD KSS knee score in our patellectomized group was 78.2 ± 19.5 (range, 17-100), which is similar to the scores from two other studies that also used the KSS. Kang et al.  reported a mean knee score of 85.6 ± 7.2 in their postpatellectomy primary TKA group, while Paletta and Laskin  reported a score of 89 ± 6 in their postpatellectomy PS TKA group. No studies used the WOMAC or SF-12. Using categorical scores of “good” or “excellent” based on a KSS knee score grading system  for comparison purposes, 38 of 52 patellectomized knees (73%) in our study had “good” or “excellent” results, compared to 44 of 52 controls (85%). While the patellectomized group had fewer such results compared to controls, these numbers are consistent with the higher end of reported categorical results. Joshi et al.  reported “good” or “excellent” results in 14 of 19 knees (74%) and Cameron et al.  in 11 of 16 knees (69%). Martin et al. , however, reported such results in only 13 of 22 knees (59%), Larson et al.  in only seven of 14 knees (50%), and Lennox et al.  in only five of 11 knees (45%).
In comparison to patients with intact patellae, our patellectomized group had lower scores post-TKA measured with the knee-specific and physician-derived KSS. Only four other studies have included an intact-patella comparison group [10, 13, 14, 16]. Paletta and Laskin  also looked at the KSS and found the mean knee score in patellectomized knees with CR implants (67 ± 10) to be lower than that in knees with intact patellae using either CR (88 ± 5) or PS (90 ± 4) implants, although this difference was eliminated in patellectomized knees with PS implants (89 ± 6). Lennox et al. , Larson et al. , and Joshi et al.  all noted fewer “good” or “excellent” categorical outcomes in the patellectomized groups compared to controls post-TKA. From a patient perspective, however, our study may not indicate worse outcomes post-TKA in patellectomized patients compared to those with intact patellae. The postoperative scores for the WOMAC and SF-12, both patient-derived outcome measures, were similar between the two groups. This may reflect lower perceived starting baseline function in these patients, even in the absence of tibiofemoral arthrosis; Peeples and Margo , in their followup of 34 patients postpatellectomy, found more than ½ reported some limitation in activity.
The change in alignment of the patellar and quadriceps tendon with patellectomy and consequent loss of reinforcement against AP translation of the flexed knee has been cited as a possible cause for worse results in patellectomized patients . Some authors have thus recommended the use of more constrained implants such as PS or hinged prostheses [2, 5, 16]. Paletta and Laskin  reported predictably good results with a PS prosthesis in patellectomized patients, which were comparable to results with intact patellae, while Cameron et al. reported superior results in a series with PS implants  compared to a previous series with cruciate-sacrificing implants . Most knees in our series had PS implants, with only four CR prostheses, and this may have contributed to our better clinical results. Interestingly, three of the CR TKAs (two from the same patient) had total KSSs of greater than 160 while the remaining knee had a score of 7. A less constrained implant does not by itself predict worse outcomes. However, further conclusions cannot be drawn, given the small number of patients with this prosthesis in the study.
Other studies have stressed the importance of patient selection and that the patellectomized patient with severe tibiofemoral arthritic changes is a better candidate for TKA than the same patient with pain in the presence of only minimal or moderate arthritic changes [14, 15]. The indication for surgery in our study was pain secondary to tibiofemoral arthrosis, as opposed to pain alone. The resultant patient selection may have contributed to the better categorical outcomes in our series.
In terms of complications, the number of patients that went on to revision surgery was not markedly higher in the patellectomized group, with 10 knees going on to revision compared to seven controls. The relatively high revision rate in both groups may be secondary to several factors, including the use of multiple implants and the wide range in length of followup (upwards of 21 years). At 15 to 20 years of followup, especially with older-model implants and noncrosslinked polyethylene, polyethylene wear may be substantial and revision for polyethylene exchange would not be unexpected. That five of seven control and four of 10 patellectomized revisions were for polyethylene wear may reflect this.
Another three patellectomized knees required further procedures in the form of extensor mechanism repair for extensor lag in one knee (1.9%) and manipulation under anesthesia for stiffness in two (3.8%). Prior surgery is a known risk factor for postoperative stiffness in TKA [4, 19] and patellectomized patients have at least one and often additional other prior knee procedures. As such, it would not be unexpected to see an increased prevalence of stiffness in the patellectomy group. Compared to larger series of primary TKA with intact patellae, there appears to be a higher incidence of postoperative stiffness in our patellectomized group. Various criteria have been used for knee stiffness, with a wide range in reported prevalence post-TKA [9, 12]. Kim et al.  defined knee stiffness as flexion contracture of greater than 15° or full flexion of less than 75° and reported an incidence of 1.3% in a series of 1000 primary TKAs. In contrast, Ipach et al.  defined it as full flexion of less than 90° and reported an incidence of 4.54%. Using these two different definitions, post-TKA stiffness rates in our study group were higher at 3.8% and 9.6%, respectively, versus control rates of 1.9% and 3.8%, respectively.
While physician-based outcome measures of TKA in patellectomized patients are lower compared to those in patients with intact patellas, the chance of achieving “good” or “excellent” scores is likely higher than historically reported. Additionally, patient perception of the pain relief and functional improvements is comparable to that of patients with TKA with an intact patella. Surgical indication and choice of implant type should factor into consideration for TKA in these patients. For the patellectomized patient with tibiofemoral arthrosis, TKA is an option that can provide marked improvements in pain and function.
1. Asif, S. and Choon, DS. Midterm results of cemented Press Fit Condylar Sigma total knee arthroplasty system. J Orthop Surg (Hong Kong).
2005; 13: 280-284.
2. Bayne, O. and Cameron, HU.TKA following patellectomy. Clin Orthop Relat Res.
1984; 186: 112-114.
3. Bellamy, N., Buchanan, WW., Goldsmith, CH., Campbell, J. and Stitt, LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. J Rheumatol.
1988; 15: 1833-1840.
4. Bong, MR. and DiCesare, PE. Stiffness after TKA. J Am Acad Orthop Surg.
2004; 12: 164-171.
5. Cameron, HU., Hu, C. and Vyamont, D. Posterior stabilized knee prosthesis for total knee replacement in patients with prior patellectomy. Can J Surg.
1996; 39: 469-473.
6. Gandek, B., Ware, JE., Aaronson, NK., Apolone, G., Bjorner, JB., Brazier, JE., Bullinger, M., Kaasa, S., Leplege, A., Prieto, L. and Sullivan, M. Cross-validation of item selection and scoring for the SF-12 Health Survey in nine countries: results from the IQOLA project. J Clin Epidemiol.
1998; 51: 1171-1178. 10.1016/S0895-4356(98)00109-7
7. Grelsamer, RP. and Weinstein, CH. Applied biomechanics of the patella. Clin Orthop Relat Res.
2001; 389: 9-14. 10.1097/00003086-200108000-00003
8. Insall, JN., Dorr, LD., Scott, RD. and Scott, WN. Rationale of the Knee Society Clinical Rating System. Clin Orthop Relat Res.
1989; 248: 13-14.
9. Ipach, I., Schafer, R., Lahrmann, J. and Kluba, T. Stiffness after knee arthrotomy: evaluation of prevalence and results after manipulation under anesthesia. Orthop Traumatol Surg Res.
2011; 97: 292-296. 10.1016/j.otsr.2011.01.006
10. Joshi, AB., Lee, CM., Markovic, L., Murphy, JC. and Hardinge, K.TKA after patellectomy. J Bone Joint Surg Br.
1994; 76: 926-929.
11. Kang, JD., Papas, SN., Rubash, HE. and McClain, EJ. Total knee arthroplasty in patellectomized patients. J Arthroplasty.
1993; 8: 489-501. 10.1016/S0883-5403(06)80214-7
12. Kim, J., Nelson, CL. and Lotke, PA. Stiffness after TKA: prevalence of the complication and outcomes of revision. J Bone Joint Surg Am.
2004; 86: 1479-1484. 10.1302/0301-620X.86B7.15255
13. Larson, KR., Cracchiolo, A., Dorey, FJ. and Finerman, GA.TKA in patients after patellectomy. Clin Orthop Relat Res.
1991; 264: 243-254.
14. Lennox, DW., Hungerford, DS. and Krackow, KA.TKA following patellectomy. Clin Orthop Relat Res.
1987; 223: 220-224.
15. Martin, SD., Haas, SB. and Insall, JN. Primary TKA after patellectomy. J Bone Joint Surg Am.
1995; 77: 1323-1330.
16. Paletta, GA, Jr, and Laskin, RS.TKA after a previous patellectomy. J Bone Joint Surg Am.
1995; 77: 1708-1712.
17. Peeples, RE. and Margo, MK. Function after patellectomy. Clin Orthop Relat Res.
1978; 132: 180-186.
18. Railton, GT., Levack, B. and Freeman, MA. Unconstrained knee arthroplasty after patellectomy. J Arthroplasty.
1990; 5: 255-257. 10.1016/S0883-5403(08)80080-0
19. Scranton, PE. Management of knee pain and stiffness after TKA. J Arthroplasty.
2001; 16: 428-435. 10.1054/arth.2001.22250
© 2013 Lippincott Williams & Wilkins LWW
20. Sledge, CB. and Ewald, FC.TKA: experience at the Robert Breck Brigham Hospital. Clin Orthop Relat Res.
1979; 145: 78-84.