Due to lack of specific presentation most patients with TB of the knee are in severe disability with joint destruction and loss of articular congruity when they are diagnosed as TB. While several studies have indicated that TKA is effective to treat the patients with advanced TB of the knee joint a consensus of views regarding surgical timing, prosthesis selection, and peri-operative antitubercular therapy has not been reached [11–14, 16, 18]. The results of the present study revealed that 2 to 4 weeks of preoperative antitubercular therapy was enough for the patient with advanced active TB of the knee. We recommend one-stage TKA for the patients with advanced active TB of the knee unless there is a discharging sinus tract or abscess that cannot be debrided during the surgery for TKA. In line with previous studies we recommend a long period of postoperative antitubercular therapy which should be at least 12 months after surgery to reduce the rate of local recurrence of TB [11–14, 16, 18].
TKA can provide the patients with quiescent TB of the knee with an excellent function, but joint replacement for active tuberculous infection was opposed by most authors because they believed the risks were too high . Other authors recommend a long interval between the treatment of the active infection and arthroplasty [11, 12, 14, 18]. They believe that the prolonged preoperative antitubercular therapy could reduce the reactivation risk of TB infection. However, the optimal timing for TKA in patients with TB of the knee joint is still controversy. Su et al  reported sixteen cases of tuberculosis of the knee treated with TKA. Eight cases were treated with antitubercular therapy for 2 (3 cases) to 12 (5 cases) months before and 1 year after arthroplasty. Another 8 cases were not diagnosed primarily and therefore received only postoperative antitubercular therapy. Five cases had a recurrence of TB. Four of these 5 cases did not receive preoperative antitubercular treatment. The last case of recurrent infection occurred in a patient who had received longstanding corticosteroid therapy. They suggested that effective preoperative and postoperative antitubercular chemotherapy were mandatory for the prevention of tuberculous reactivation after TKA. Oztürkmen et al.  reported twelve patients with recent onset tuberculous arthritis of the knee. Two-stage TKA was performed in all patients with 7 primary prostheses and 5 revisions. The time interval between the first and the second stage was not more than 6 months. All patients were given postoperative antitubercular treatment for more than 1 year, and within an average follow-up period of 6.1 ± 1.8 years, no reactivation of tuberculous infection was found. Kim et al.  performed TKA in 19 patients (22 knees) with tuberculous arthritis of the knee. The interval between the subsidence of the signs of tuberculous infection and arthroplasty ranged from 3 months to 5 years. They found 3 recurrent cases during the follow-up. Habaxi et al  analyzed 10 patients with active tuberculosis of the knee. One-stage TKA was performed in all cases. The patients were given preoperative antitubercular treatment for 2–4 weeks and more than 1 year postoperatively. They had one recurrent case. Our study is in line with Habaxi et al. In our series, preoperative antitubercular therapy was administered for 2–4 weeks in six patients and no recurrence was found during follow-up. Two other cases received antitubercular treatment for more than 3 months because popliteal abscess and discharging sinus tract had to be eliminated before TKA. No recurrence of TB was found in these two cases. From the results of our study and reviews of previous studies we find that the recurrence of TB after TKA is not related to the interval between initiation of preoperative antitubercular treatment and arthroplasty. The recurrence of TB after TKA may occur even when TB is quiescent for several years [13, 19] while patients without preoperative antitubercular treatment may be free from recurrence [11, 16, 21]. We recommend short-term preoperative antitubercular treatment so as to shorten the course of treatment of TB. In our opinion 2–4 weeks of preoperative antitubercular treatment is enough before TKA if there are no special conditions such as sinus tract or extra-articular abscess.
The reason that we recommend short-term preoperative antitubercular treatment lies in that M. tuberculosis has specific behavioral characteristics differ from pyogenic bacteria . M. tuberculosis reproduces slowly and has little tendency to adhere to implants. In vitro studies also demonstrated that M. tuberculosis has little or no biofilm formation making it susceptible to antitubercular therapy . So antitubercular drugs can eliminate M. tuberculosis efficiently even with implants existed. Even if TB is reactivated, the disease can be controlled by antitubercular drugs alone or in combination with debridement without removal of the prosthesis in most cases [11, 13, 19, 24].
Complete debridement of the infected tissues is the mainstay of the treatment to lower the potential risk of reactivation of TB [11–14]. However, it’s difficult to cure tuberculous infection using debridement in combination with antitubercular drugs. The risk of recurrence is high according to previous reports [13, 14]. Moreover, the patients have to suffer from disability of the knee before arthroplasty because the joint surface would be more ragged after debridement. Additionally it’s difficult to perform thorough debridement in that surgeons have to keep bone tissues as more as possible for fear of massive bone defect. The situation is different in TKA because the infected tissues are eliminated to the most extent due to extensive intraoperative exposure and bone osteotomy. Intraoperative pulsed irrigation may further reduce M. tuberculosis located in trabecular space.
Some authors recommended two-stage surgery in patients with active tuberculous arthritis [12, 18, 25]. However, two-stage TKA may cause so much loss of bone that revision prostheses have to be needed in some cases which may impair the function of the knee and the life of the prosthesis [12, 25]. In our series, two-stage TKA was performed in two cases due to popliteal abscess or discharging sinus tract. In fact, the surgical procedures were more difficult than usual due to the stiffness caused by less movement of the knee after the first-stage surgery. At the last follow-up, although no recurrence of TB was found the function of the knee with two-stage TKA was poorer than that with one-stage surgery (Table 1). In our opinion, two-stage TKA for active tuberculosis of the knee is not recommended unless there is discharging sinus tract or extra-articular abscess unable to remove during TKA.
As for postoperative antitubercular therapy our opinion is in line with others [11–14, 16, 18]. We recommend a long-term postoperative antitubercular therapy for at least 12 months. Once recurrence of TB is detected a longer antitubercular course is recommended until TB is cured. Debridement with or without removal of implants may be needed sometimes.
Several authors considered the level of ESR as an indicator of the efficacy of antitubercular drugs [11, 18]. We agree with this point. In our study TKA was not performed until the ESR decreased more than 20% or below 50 mm/h. Normal ESR levels are not prerequisite for TKA because the signs of active infection (persistent pain, effusion in the joint, and positive radiographic features) may exist despite the fact that laboratory tests show no evidence of active infection . The present study reveals that TKA for patients with active TB of the knee is a safe procedure even though the ESR level is higher than normal.