RETRACTED ARTICLE: Patelloplasty in total knee arthroplasty with circumpatellar denervation versus without denervation – a randomized prospective study

Introduction Anterior knee pain is one of the major problems in total knee arthroplasty (TKA) and is often etiologically associated with a patellofemoral parts etiology. There is no consensus as to etiology or treatment. Denervation of the patella by electrocautery and patelloplasty along with removal of osteophytes have been used for treatment of anterior knee pain in TKA. The purpose of our study was to compare, in terms of the anterior knee pain and clinical outcomes of patelloplasty in total knee arthroplasty (TKA), patellar denervation by electrocautery and non-patellar-denervation treatment in a 2 year follow-up. Materials and methods This study was conducted in a total of 108 patients, who underwent TKA at our institution between June 2015 and December 2016. Patients age 55 to 80 years, who are suffering from osteoarthritis, rheumatoid arthritis of knee were included in this study. Patients were randomly allocated into patelloplasty with denervation group and non-denervation group. The denervation of the patella was done in electrocautery group using a monopolar coagulation diathermy set to 50 W. (Valleylab Inc., Boulder, CO). Postoperatively, patients were assessed at regular intervals of 3, 6, 9, 12, 24 months. To assess patient outcomes, we used questionnaires to determine the Knee Society score (KSS - knee and function scores), a specific patellofemoral pain questionnaire (Kujala score) range of motion (ROM) and a visual analogue scale (VAS) to assess anterior knee pain. Results The data obtained were analyzed using SPSS version 17.0. Continuous variables were expressed as mean ± SD. Of the 108 patients, 9 patients were lost to follow-up. Among the remaining 99 patients, 50 were included in denervation group and 49 in non-denervation group. In our study, there was no statistically significant difference in Mean KUJALA score preoperatively (p > 0.05). Postoperatively, the mean KUJALA score was significantly higher in denervation group at 3, 6, 9, 12, 24 months of follow-up when compared to TKR with no denervation (p < 0.05). There was no statistically significant difference in Mean VAS score preoperatively (p > 0.05). However, 6, 12 and 24 months after the operation, the mean VAS score was significantly lower in denervation group. There was no statistically significant difference in Mean KSS score preoperatively and postoperatively (p > 0.05). The mean ROM was significantly higher in denervation group than in the group of TKR with no denervation (p < 0.05). Conclusion In our study, less postoperative anterior knee pain, increased range of motion, significantly lower VAS scores were seen in the denervation group compared with non-denervation group. Circumferential denervation of patella during primary TKA along with patellar resurfacing is a safe procedure that improves patient satisfaction, decreases anterior knee pain and improves range of flexion in the postoperative period and at postoperative follow-ups.


Introduction
Anterior knee pain is still a major problem in total knee arthroplasty (TKA) reported in 4-49% of patients. Often due to patellofemoral etiology, there is no clear consensus as to etiology or treatment [1] of anterior knee pain. This anterior patellofemoral knee pain is one of the major cause of dissatisfaction and morbidity in post-TKR patients.
Unfortunately, many aspects of anterior knee pain after TKA have yet to be fully understood. Patellar resurfacing for anterior knee pain reduction in TKA remains controversial, because it has been associated with fracture, subluxation and dislocation of the patella, aseptic loosening, and patella necrosis [2].
In general, patelloplasty with removal of osteophytes along with denervation of the patella by electrocautery was used for treatment of anterior knee pain [1]. The purpose of our study was to compare the anterior knee pain and clinical outcomes of patelloplasty in TKA with patellar denervation by electrocautery and without patellar denervation for a follow-up period of 2 years.

Materials and methods
This study was conducted among a total of 108 patients, who underwent TKA at our institution between June 2015 and December 2016. All the surgeries were performed by a single surgeon. Patients between age 55 to 80 years, who were suffering from osteoarthritis, rheumatoid arthritis of knee were included in this study. Patients with previous patella surgery/fracture, previous high tibial osteotomy, revision TKA were excluded from the study. Of the 108 patients, 55 were included in patelloplasty with electrocautery group and other 53 in nonelectrocautery group. The allocation was done by using simple randomization method. Inside the operation theatre, patients were allocated to denervation or nondenervation group based on a chit. Neither the operating surgeons nor the assistants knew to which group the patient was allocated prior to the procedure.
All the patients who are planned admitted for Total knee arthroplasty underwent routine blood investigations for surgical profile. Physician fitness & consent are taken prior to surgery. X-rays of involved knees in AP and lateral views were taken. Patellofemoral and tibiofemoral arthritis was graded. Radiological and clinical assessment was done in terms of deformities, crepitus, and anterior knee pain.

Surgical technique
Standard surgical technique involving anterior midline incision ( Fig. 1) and medial parapatellar approach ( Fig. 2) with retropatellar fat pad excision ( Fig. 3) was used for all the patients. Cruciate sacrificing PFC SIGMA implants were used in all cases. Patellar tracking was then checked with 'no thumb test' after seating of both trial and definitive implants.
After implantation of femoral and tibial components, the patellar surface was resurfaced using an electric saw. Both facets of the patella were reshaped to anatomically mimic the normal shape of the patella. All the marginal osteophytes were removed. Peripheral denervation around patella was performed with electrocautery for a depth of 2 to 3 mm in electrocautery group (Fig. 4). The denervation of the patella was done using a monopolar coagulation diathermy set to 50 W. (Valleylab Inc., Boulder, CO). In non-electrocautery group only patelloplasty was performed. Later, wound was closed in layers over suction drain.

Patient evaluation
Postoperative X-rays were taken in AP and lateral views.
To assess patient outcomes, we used questionnaires to determine the Knee Society score (KSS -knee and function scores), a specific patellofemoral pain questionnaire (Kujala score), range of motion (ROM) and a visual analogue scale (VAS) to assess anterior knee pain. The total KSS consists of two components, with first one being the Knee Society Function Score and the second the Knee Society Pain Score.
Written informed consent was obtained from all patients, and approval to use their medical records and to re-evaluate each patient was taken from the Local Research Ethics Committee.

Radiological assessment
Standard weight-bearing anteroposterior, lateral and skyline views were taken preoperatively, immediately postoperatively and at the follow-up visits. Pre operatively, Insall -salvati ratio was assessed. Those with patella alta/baja are excluded from study based on insall ratio.

Method of statistical analysis
SPSS statistical software version 17.0 (SPSS Inc., Chicago, IL, USA) was used to process the data. We used the following parametric tests: (1) paired t test for intragroup comparison of two variables i.e. Pre-and postoperative values and (2) non-paired t test for intergroup comparison of two variables. Pearson correlation coefficient was used to measure correlation. Continuous variables were be expressed as mean ± SD. A P value < 0.05 was considered significant.

Results
Of the 108 patients who underwent TKA, 9 were lost to follow-up. Among the 99 who were studied, 50 were in electrocautery group and 49 in non-electrocautery group. Patients were followed up 3, 6, 12 and 24 months after operation. At all preoperative and postoperative visits, all clinical scores were determined with respect to function, the range of motion VAS etc. The data obtained were analyzed using SPSS version 17.0. Continuous variables were expressed as mean ± SD. Appropriate statistical tests were used to determine outcomes of patelloplasty in TKA, with and without coagulation diathermy.
Patients in both study groups were comparable in terms of age, gender, mean age being 63.6 years. There was no significant difference in the distribution of patients in terms of age (p > 0.05). Mean age in patients with denervation was 62.7 ± 7.5 and the mean age in patients without denervation was 63.6 ± 8.11 (Table 1). There was no statistically significant difference in mean age distribution between groups (p > 0.05) ( Table 2). The gender distribution between the groups is also not significant statistically (p > 0.05). Male to female ratio was 0.26:1.
In our study, there was no statistically significant difference in mean KUJALA score preoperatively (p > 0.05) ( Table 3, Fig. 5). Postoperatively, the mean KUJALA score was significantly higher in denervation group at 3, 6,9,12 months follow-up when compared to TKR with no denervation (p < 0.05). There was no statistically significant difference in mean VAS score preoperatively (p > 0.05). However, postoperatively at 6, 12 and 24 months, the mean VAS score was significantly lower in denervation group (Table 4, Fig. 6). There was no statistically significant difference in mean KSS score preoperatively and postoperatively (p > 0.05) (Fig. 7). The mean ROM was significantly higher in denervation group than in non-denervation (p < 0.05) (Fig. 8). The patient satisfaction rate was also higher in denervation group (Table 5) compared with non-denervation group.

Discussion
Anterior knee pain is reported to postoperatively occur in up to one-half of all patients receiving TKR. The ultimate goal of TKR is to relieve pain and to improve the knee function. The presence of AKP after TKR is negatively correlated with patient satisfaction and quality of life.
The pathophysiology of anterior knee pain in osteoarthritis is often uncertain and is frequently multifactorial. Surface incongruities in patella and cartilage is one of the main reason for AKP [3,4].
The major source of this pain is both peri-patellar soft tissue and retropatellar fat pad [5]. Substance-p nociceptive fibres in the peripatellar soft tissue were found to be one of the major cause of this pain [6,7]. Electrocautery disables these pain receptors and achieve desensitization or denervation of the anterior knee region.
Disagreement exists in regard of circumpatellar denervation, with some studies finding it beneficial and some studies not [8][9][10][11]. We hypothesized that patellar denervation with electrocautery along with patellar   In a study, van Jonbergen et al. [9,10] found that 56% of Dutch orthopedic surgeons performing TKA used circumpatellar electrocautery to prevent anterior knee pain whereas 32% did not.
We designed a randomized control study with a null hypothesis that there is no difference in the outcome between the two groups.
Our results showed that postoperative Knee Society score, ROM, Kujala score and VAS were significantly better in the denervation group. These findings indicated that patellar denervation by electrocautery would decrease anterior knee pain and achieve good clinical outcomes after TKA.
Rand and Gaffey [1] described that electrocautery has potentially harmful effects on the articular cartilage and must be performed with care intraoperatively to avoid cartilage trauma.
We have applied electrocautery only to the peripheral rim of the patella with a depth of 2-3 mm to prevent surface exposure.
Vega and Golano et al. [12] summarised that circumpatellar electrocautery would denervate the patellar rim and effectively desensitize the anterior knee area.
Our findings were consistent with studies conducted by Pulavarthi RS et al. [13] in 2013 and Gupta et al. [8], in terms of age, gender distribution.
In our study there was no statistically significant difference in mean KUJALA score preoperatively (p > 0.05). However, posoperatively the patients were followed for 3 months (p <0.001), 6 months (p < 0.001), 12 months (p <0.001) and 24 months (p < 0.001) and it was observed that the mean KUJALA score was significantly higher in denervation group than in those without denervation (p < 0.05). Ramnadh et al. [13] also observed significant improvement in knee pain (in terms of Kujala score) (p = 0.02) at 3 months (p = 0.14) in denervation group but at 6 months (p = 0.13), 9 months (p = 0.17), 12 months (p = 0.21) and 2 years (p value = 0.20) the improvement was not significant compared to both groups.
There was no statistically significant difference in mean VAS score preoperatively (p > 0.05). No significant difference in mean VAS score was observed at 3 months postoperatively (p > 0.05). However, at 6, 12 and 24 months, the mean VAS score was significantly lower in denervation group than when compared to TKR with no denervation (p < 0.05). These results are in contrast with the findings by Ramnadh Pulavarthi et al., where VAS scores were not significant at 12 months (p = 0.1) and 24 months (p = 0.35) in denervation group compared to non-denervation group.
Another study by Saoud [5] involving a total 40 patients who underwent TKA showed that denervation group had a significantly more favorable pain scores.
Sun et al. retrospectively studied clinical outcomes of patelloplasty and non-patelloplasty treatment in TKA for a mean follow-up time of 55 months. The patelloplasty group had significantly higher Feller patellar scores, Lonner patellar scores and better patient satisfaction [14] but there was no significant difference found in KSS scores. Whereas, in our study, patients in both groups, patelloplasty with and without denervation, had higher KSS than those observed by Sun et al. We can attribute this to a longer follow-up time, as it seems that the KSS scores improve with time.
Ramnadh et al. [13] reported significant difference in preoperative ROM (p = 0.48) and postoperative ROM (p<0.01) between denervation group and non-denervation group. Alay et al. [15] also reported that there was significant difference in postoperative ROM (p = 0.015) between denervation and non-denervation groups. In this study, there was no statistically significant difference found in mean ROM score preoperatively (p > 0.05). Postoperatively, the mean ROM was significantly higher in denervation group than in non-denervation group (p < 0.05).
The pain relief achieved in the denervation group, at early stages of postoperative period helped achieve a better ROM 12 and 24 months after operation. This resulted in higher patient satisfaction in the denervation group. Fig. 6 The mean Visual Analogue Score was significantly lower in denervation group than in the non-denervation group. The mean VAS score in denervation group was 2.52 (6 months) and 1.36 (12 months). Whereas in denervation group, it was 3.1 and 2.5 at 6 and 12 months respectively Single surgeon involvement, using a standardized technique, robust inclusion/exclusion criteria and a minimum 24 months of follow-up time are major strengths of our study. The limitations of the included a small sample size and short-term follow-up. Larger, prospective randomized controlled trials and longer follow-up studies are needed to better evaluate the effect and long-term results of patellar denervation. Also better scoring systems have to developed to evaluate the knee pain after TKA.

Conclusion
The concept of patellar denervation is intriguing, but the which technique is proper remains unclear.
In this study, less postoperative anterior knee pain, increased ROM, significantly lower VAS scores in denervation group were observed as compared with non-denervation group. However, the study had the limitations of shorter period of follow-up, small sample size. In conclusion, circumferential denervation of patella during primary TKA with patellar resurfacing is a safe procedure which improves patient satisfaction, decreases anterior knee pain and improves range of flexion in the postoperative period and at subsequent follow-up visits.  Fig. 8 Mean ROM between two groups. In denervation group the mean ROM was 96, 99, 105, 106 at 3,6,12, 24 months respectively whereas in non-denervation group mean ROM was 93, 96, 101, 101 respectively. The difference was statistically significant