The timing of total hip replacement (THR) in patients with active tuberculosis (TB) of the hip is controversial, because of the potential risk of reactivation of infection. There is little information about the outcome of THR in these patients. We conducted a systematic review of published studies that evaluated the outcome of THR in patients with active TB of the hip. A review of multiple databases referenced articles published between 1950 and 2012. A total of six articles were identified, comprising 65 patients. TB was confirmed histologically in all patients. The mean follow-up was 53.2 months (24 to 108). Antituberculosis treatment continued post-operatively for between six and 15 months, after debridement and THR. One non-compliant patient had reactivation of infection. At the final follow-up the mean Harris hip score was 91.7 (56 to 98). We conclude that THR in patients with active TB of the hip is a safe procedure, providing symptomatic relief and functional improvement if undertaken in association with extensive debridement and appropriate antituberculosis treatment. Cite this article: Bone Joint J 2013;95-B:578–82

Aims

The surgical treatment of tuberculosis (TB) of the spine consists of debridement and reconstruction of the anterior column. Loss of correction is the most significant challenge. Our aim was to report the outcome of single-stage posterior surgery using bone allografts in the management of this condition.

We report 60 patients with tuberculosis of the hip treated by Charnley low friction arthroplasty and followed for eight to 13 years. Eight of them had active tuberculosis of the hip at the time of operation, and all were covered by relatively short courses of antituberculous drugs. Our study suggests that arthroplasty can be recommended for these patients provided that adequate chemotherapy is given both before and after operation

1. Three hundred and twenty cases of tuberculosis of the hip joint have been analysed and the late results assessed three or more years after discharge from hospital. 2. The primary bone focus involved the acetabulum alone in 39·3 per cent; the acetabulum and femoral head in 34·1 per cent; the head of femur alone in 19·2 per cent, and the femoral neck alone in 7·4 per cent. In 101 cases widespread destruction of the joint had taken place by the time the patient first came under observation. 3. Premature epiphysial fusion round the knee joint of the affected side occurred in 23 per cent of all patients under the age of fifteen years. 4. When hip disease was complicated by multiple foci of active tuberculosis or by secondarily infected abscesses and sinuses, the prognosis was seriously worsened. 5. Significant late deformity occurred in 38·3 per cent of patients discharged with "sound" fibrous ankylosis, and in 60·5 per cent of those with an unstable fibrous ankylosis. 6. Of 187 patients observed for more than three years after discharge from hospital 174 returned to full activity, seven were partly incapacitated and six were totally incapacitated. 7. Analysis of the late results suggests that the prognosis is best when an adequate period of conservative treatment is followed by some form of arthrodesis operation

After tuberculosis of the spine, hip and knee, tuberculosis of the foot and ankle is the most common occurring area of skeletal tuberculosis seen in our unit. We retrospectively reviewed 14 patients (14 feet and/or ankles) seen over the 16-year period 1982 to 1997. The average age of the children was 5.2 years (range 1.5 to 11 years). The duration of symptoms ranged from 1 week to 1 year. The most common presentation was swelling and pain of the involved joint, but three patients each presented with a chronic discharging sinus after being drained elsewhere as an acute abscess. Radiographs revealed osteo-penia with or without lytic areas, joint margin destruction or joint space narrowing. The average sedimentation rate (ESR) was 52.3 mm/hour (range 9 to 120). The Mantoux test was positive in 13 out of the 14 patients. Chest x-rays demonstrated latent or active tuberculosis in 50% of patients. Open biopsy was performed in all patients. Hypertrophic synovium was found in all cases except one, where atrophic tuberculosis with joint space narrowing was present. A positive diagnosis of tuberculosis was made in all cases, either by demonstrating caseating granulomatous tissue on histology, or by growing a positive culture for mycobacterium tuberculosis or both. Histology was positive in 86%, acid-fast bacteria were seen in 28.5% and a positive culture was obtained in 82% of the patients. At an average follow up of 7.4 years (range 1 to 17 years) all patients were assessed both clinically and radiologically. Patients with lytic lesions and destruction of joint margins reconstituted well radiologically, had a good clinical outcome with a good range of movement of the affected joint, however the one patient with atrophic tuberculosis remained with a narrowed joint space, stiffness and a poor clinical result compared to the rest

1. Nine hundred and fourteen cases of tuberculosis of the spine are analysed and the late results ascertained three or more years after discharge from hospital. 2. The relative frequency with which the various segments of the spine are involved has been found. Cervical disease was present in 3·5 per cent of cases, thoracic in 43·l per cent, lumbar in 32·9 per cent, thoraco-lumbar in 16· 7 per cent and lumbo-sacral in 3·8 per cent. 3. The mortality rate was 16·7 per cent. In patients with multiple lesions 25·5 per cent died, compared with 12·3 per cent in the group without complications. When chronic secondarily infected abscesses and sinuses were present the mortality rate was 19·1 per cent, and of patients with paraplegia 24·8 per cent died. 4. In the late results the working capacity of 390 patients was ascertained. It was full in 86 per cent, partial in 5·8 per cent and nil in 8·2 per cent. 5. An attempt has been made to determine the site of the primary bone focus from the radiograph. Early "epiphysial" changes were present in 33 per cent ; the central focus beginning in the spongy tissue of a vertebral body was present in 11·6 per cent; subperiosteal lesions were present in 2·1 per cent; and infection of the neural arch was present in only 0·5 per cent. In 52·8 per cent, however, widespread destruction had taken place when the patient first came under observation. 6. The ill effect of complications upon the prognosis is stressed—especially in the cases of multiple foci of active tuberculosis, secondarily infected abscesses and sinuses, and paraplegia. Paraplegia occurred in 31·2 per cent of the cases of thoracic disease. 7. An attempt has been made to determine the frequency with which tuberculosis of the spine heals by spontaneous bony fusion of the affected vertebral bodies. It was found in 27·3 per cent of the cases in the present series. 8. Reasons for further treatment after the initial discharge of the patient are examined

The case histories of these Black Notley patients show that no serious harmful effects have been caused by pregnancies going to full term. Most of the patients had normal confinements; Caesarian sections were few and were usually done for obstetric reasons, not for fear of reactivation of the disease. This experience bears out a remark made to me by Marcel Galland. When asked if Caesarian sections were required for women who became pregnant after treatment at Berck for skeletal tuberculosis, he replied: "Jarnais! Tout passe!". Two patients in the series had artificial abortions, and one aborted spontaneously. Two of these three patients did not do as well as other patients whose pregnancies went to full term. Quite another question is whether the onset of skeletal tuberculosis sooii after pregnancy may be attributed to the pregnancy. In a group of women of child-bearing age, it would be difficult to distinguish between post hoc and propter hoc in considering this question. One patient of this series, H. Q. (Case 25), was admitted eight months pregnant with active tuberculosis of the knee; but pain in the joint had started fifteen months before admission, and the onset of disease no doubt preceded the pregnancy. Some of the patients admitted to Black Notley were mothers of young children hut there is no note of any pregnancy having preceded the onset of tuberculosis by a short interval. The follow-up reports of the married patients compare favourably with those of the unmarried. Further, it may be argued that pregnancy increases resistance to tuberculosis. Such a belief was current from the time of Hippocrates until the middle of the nineteenth century, and clinical evidence can be quoted in favour of such a view. In America, Lyman (1943) followed up 1,818 women who had been treated for tuberculosis in a sanatorium; many had married despite medical advice to the contrary. Lyman found that the late results of treatment were four times as good in the married patients as in the single women. Lyman says, "When we consider that the married group established this record in spite of the fact that 192 out of 315 gave histories of pregnancies (averaging 2·25 children each), it is clear that some factor not yet accounted for has exerted a profound influence in their cases.". I have speculated elsewhere as to what this factor may be (Wilkinson 1949). But it seems clear that increased sterol circulation in the body is associated with increased reticulo-endothelial activity (Fraser 1935). Thus reticulo-endothelial proliferation has been observed in the lesions of lupus following the administration of calciferol (Dowling, Gauvain and Macrae 1948). An increase of blood cholesterol is found during pregnancy. The sex hormones are sterols and might be expected to be raised in the marital state; it is perhaps significant that marital contacts form a group relatively immune to tuberculosis. There would appear to be good grounds for reassuring the married woman who has suffered from skeletal tuberculosis regarding the prospect of normal pregnancy. Many letters received from patients in a follow-up of this sort are poignant documents. There is no doubt that the majority of these women desire children as ardently as normal women, and that a safe confinement following skeletal tuberculosis is an excellent form of rehabilitation

We report a case of peri-prosthetic tuberculous infection nine years after total hip arthroplasty in a patient with no history of tuberculosis before the procedure. Further investigation revealed active pulmonary tuberculosis which was thought to have spread haematogeneously to the arthroplasty. The infection did not respond to standard antituberculous drugs. Removal of the prosthesis and insertion of an antibiotic spacer were required

In this retrospective observational cohort study, we describe 17 patients out of 1775 treated for various fractures who developed mycobacterium tuberculosis (MTB) infection after surgery. The cohort comprised 15 men and two women with a mean age of 40 years (24 to 70). A total of ten fractures were open and seven were closed. Of these, seven patients underwent intramedullary nailing of a fracture of the long bone, seven had fractures fixed with plates, two with Kirschner-wires and screws, and one had a hemiarthroplasty of the hip with an Austin Moore prosthesis. All patients were followed-up for two years. In all patients, the infection resolved, and in 14 the fractures united. Nonunion was seen in two patients one of whom underwent two-stage total hip arthroplasty (THA) and the other patient was treated using excision arthoplasty. Another patient was treated using two-stage THA. With only sporadic case reports in the literature, MTB infection is rarely clinically suspected, even in underdeveloped and developing countries, where pulmonary and other forms of TB are endemic. In developed countries there is also an increased incidence among immunocompromised patients. In this paper we discuss the pathogenesis and incidence of MTB infection after surgical management of fractures and suggest protocols for early diagnosis and management.

Cite this article: Bone Joint J 2015;97-B:1279–83.

A prospective series of 32 cases with tuberculosis of the hand and wrist is presented. The mean age of the patients was 23.9 years (3 to 65), 12 had bony disease and 20 primarily soft-tissue involvement. The metacarpal of the little finger was the most commonly involved bone. Pain and swelling were the usual presenting features and discharging sinuses were seen in three cases.

All patients were given anti-tubercular chemotherapy with four drugs. Operative treatment in the form of open or arthroscopic debridement, or incision and drainage of abscesses, was performed in those cases where no response was seen after eight weeks of ATT. Hand function was evaluated by the modified score of Green and O’Brien. The mean was 58.3 (25 to 80) before treatment and 90.5 (80 to 95) at the end. The mean follow-up was for 22.4 months (6 to 43). Conservative treatment was successful in 24 patients (75%). Eight who did not respond to chemotherapy within eight weeks required surgery.

Although tuberculosis of hand has a varied presentation, the majority of lesions respond to conservative treatment.

We describe the results of cemented total hip replacement in 23 patients (23 hips) with active tuberculous arthritis of the hip with a mean follow-up of 4.7 years (4 to 7). In two patients the diagnosis was proved by pre-operative biopsy, whereas all others were diagnosed on a clinicoradiological basis with confirmation obtained by histopathological examination and polymerase chain reaction of tissue samples taken at the time of surgery. All patients received chemotherapy for at least three months before surgery and treatment was continued for a total of 18 months. Post-operative dislocation occurred in one patient and was managed successfully by closed reduction. No reactivation of the infection or loosening of the implant was recorded and function of the hip improved in all patients.

Total hip replacement in the presence of active tuberculous arthritis of the hip is a safe procedure when pre-operative chemotherapy is commenced and continued for an extended period after operation.

We studied 15 patients with healed tuberculosis of the spine and a resultant kyphosis. We selected only those with no neurological deficit and performed a wedge resection of the vertebra using a transpedicular approach. The wedge was removed from the apex of the deformity. For those with a neurological deficit, we chose the conventional anterior debridement and decompression with 360° circumferential fusion. At a mean follow-up of 26.8 months (8 to 46) the outcome was good with an increase in the mean Oswestry Disability Index from 56.26 (48 to 62) pre-operatively to 11.2 (6 to 16) at the latest follow-up.

There are few reports of the treatment of lumbar tuberculous spondylitis using the posterior approach. Between January 1999 and February 2004, 16 patients underwent posterior lumbar interbody fusion with autogenous iliac-bone grafting and pedicle screw instrumentation. Their mean age at surgery was 51 years (28 to 66). The mean follow-up period was 33 months (24 to 48). The clinical outcome was assessed using the Frankel neurological classification and the Kirkaldy-Willis criteria.

On the Frankel classification, one patient improved by two grades (C to E), seven by one grade, and eight showed no change. The Kirkaldy-Willis functional outcome was classified as excellent in eight patients, good in five, fair in two and poor in one. Bony union was achieved within one year in 15 patients. The mean pre-operative lordotic angle was 27.8° (9° to 45°) which improved by the final follow-up to 35.8° (28° to 48°). Post-operative complications occurred in four patients, transient root injury in two, a superficial wound infection in one and a deep wound infection in one, in whom the implant was removed.

Our results show that a posterior lumbar interbody fusion with autogenous iliac-bone grafting and pedicle screw instrumentation for tuberculous spondylitis through the posterior approach can give satisfactory results.

Injury to the spinal cord and kyphosis are the two most feared complications of tuberculosis of the spine. Since tuberculosis affects principally the vertebral bodies, anterior decompression is usually recommended. Concomitant posterior instrumentation is indicated to neutralise gross instability from panvertebral disease, to protect the anterior bone graft, to prevent graft-related complications after anterior decompression in long-segment disease and to correct a kyphosis. Two-stage surgery is usually performed in these cases. We present 38 consecutive patients with tuberculosis of the spine for whom anterior decompression, posterior instrumentation, with or without correction of the kyphus, and anterior and posterior fusion was performed in a single stage through an anterolateral extrapleural approach. Their mean age was 20.4 years (2.0 to 57.0).

The indications for surgery were panvertebral disease, neurological deficit and severe kyphosis. The patients were operated on in the left lateral position using a ‘T’-shaped incision sited at the apex of kyphosis or lesion. Three ribs were removed in 34 patients and two in four and anterior decompression of the spinal cord was carried out. The posterior vertebral column was shortened to correct the kyphus, if necessary, and was stabilised by a Hartshill rectangle and sublaminar wires. Anterior and posterior bone grafting was performed.

The mean number of vertebral bodies affected was 3.24 (2.0 to 9.0). The mean pre-operative kyphosis in patients operated on for correction of the kyphus was 49.08° (30° to 72°) and there was a mean correction of 25° (6° to 42°). All except one patient with a neural deficit recovered complete motor and sensory function. The mean intra-operative blood loss was 1175 ml (800 to 2600), and the mean duration of surgery 3.5 hours (2.7 to 5.0). Wound healing was uneventful in 33 of 38 patients. The mean follow-up was 33 months (11 to 74). None of the patients required intensive care.

The extrapleural anterolateral approach provides simultaneous exposure of the anterior and posterior aspects of the spine, thereby allowing decompression of the spinal cord, posterior stabilisation and anterior and posterior bone grafting. This approach has much less morbidity than the two-stage approaches which have been previously described.

We studied 51 patients with osteo-articular tuberculosis who were divided into two groups. Group I comprised 31 newly-diagnosed patients who were given first-line antituberculous treatment consisting of isoniazid, rifampicin, ethambutol and pyrazinamide. Group II (non-responders) consisted of 20 patients with a history of clinical non-responsiveness to supervised uninterrupted antituberculous treatment for a minimum of three months or a recurrence of a previous lesion which on clinical observation had healed. No patient in either group was HIV-positive. Group II were treated with an immunomodulation regime of intradermal BCG, oral levamisole and intramuscular diphtheria and tetanus vaccines as an adjunct for eight weeks in addition to antituberculous treatment. We gave antituberculous treatment for a total of 12 to 18 months in both groups and they were followed up for a mean of 30.2 months (24 to 49). A series of 20 healthy blood donors served as a control group.

Twenty-nine (93.6%) of the 31 patients in group I and 14 of the 20 (70%) in group II had a clinicoradiological healing response to treatment by five months.

The CD4 cell count in both groups was depressed at the time of enrolment, with a greater degree of depression in the group-II patients (686 cells/mm³ (sd 261) and 545 cells/mm³ (sd 137), respectively; p < 0.05). After treatment for three months both groups showed significant elevation of the CD4 cell count, reaching a level comparable with the control group. However, the mean CD4 cell count of group II (945 cells/mm³ (sd 343)) still remained lower than that of group I (1071 cells/mm³ (sd 290)), but the difference was not significant. Our study has shown encouraging results after immunomodulation and antituberculous treatment in non-responsive patients. The pattern of change in the CD4 cell count in response to treatment may be a reliable clinical indicator.