A 20-year retrospective study of osteoarticular tuberculosis in a pediatric third level referral center

Tuberculosis is a disease that has been in existence for centuries, with report of skeletal lesions in early Neolithic population [1]. The causative agent is the bacillus Mycobacterium tuberculosis, a strict aerobic microorganism, alcohol-acid resistant first described on 24th March, 1882 by Robert Koch.

TB has been a persistent global health problem with annual incidence rate of approximately 10 million cases, and one of the 10 leading causes of death in the world. In the last 5 years, it has been the leading cause of death by specific infectious agent, even above human immunodeficiency virus (HIV) [2].

The latest global report indicated that a total of 10.4 million peoples are infected and that out of these, 6% corresponds to children under 18 years of age. In the pediatric population, the most frequent site affected is the lung in 80%. Of the extrapulmonary presentations, 67% corresponds to lymph node TB, 13% to meningeal, 6% to pleural, 5% to miliary, and 4% to musculoskeletal manifestations [3]. Of the total cases of TB infection, osteoarticular involvement represents 4% to 5% with reports in the pediatric population of up to 7%. The diagnosis of extrapulmonary tuberculosis in the pediatric age is a challenge, since it presents with insidious onset without constitutional signs and symptoms in up to 72% m [4]. The objective of the present study is to describe the clinical, diagnostic, radiological and therapeutic aspects of OATB in patients in a tertiary pediatric hospital.

This is a retrospective, descriptive study of the cases of Tuberculosis diagnosed in the Infectious Diseases Department of National Institute of Pediatrics, Mexico City from 1999 to 2018. The hospital has a unique program of Tuberculosis in Pediatrics and all the patients with TB diagnosis are directly managed by personnel who are specialists in the disease. We collected demographic data, BCG vaccination status, clinical presentation, imaging studies, performance of purified protein derivative (PPD), diagnostic time and method, medical-surgical treatment and follow up. Those patients who do not comply with previous criteria were excluded. HIV testing with ELISA was performed to detect immunodeficiencies. The operative procedure varied according to the case and comprises of arthrotomy and debridement, cryotherapy, application of bone substitutes (glass BONE™) and decompression with column stabilization. All methods were carried out in accordance with relevant guidelines and regulations under Ethics approval and consent to participate. As this was a retrospective study, all the information was obtained from the electronic records of the patients. However, the study was submitted to and approved by the Institutional Ethics and Academic Committee of National Institute of Pediatrics. Written informed consent was obtained from a parent or guardian for participants under 16 years old. The study was carried out with strict adherence to the ethical principles enunciated in the Declaration of Helsinki and to the CONSORT guidelines, http://​www.​consort-statement.​org/​consort-2010.

The diagnosis of OATB in pediatrics is a challenge due to its insidious clinical presentation. In this study, 43 cases of OATB handled under the TB unique program of the Infectious Diseases Department, National Institute of Pediatrics, Mexico City during a 20-year period were analyzed. Contrary to the national epidemiology where the pulmonary infection is the most common modality of tuberculosis, in our hospital, 85% of the cases were extrapulmonary presentations because lung TB are mostly managed in the second level of attention [5]. The presentation of OATB represented 29% of the general population, which is above the percentage in regions such as Europe, but coincides with what is reported in developing countries, particularly in Asia [6‐8]. Patients ≤ 6 years old were the most prevalent due to the high risk of progression.

The most important clinical symptom was pain in all case followed by fever. Contact with TB diagnosis was identified in 23% of the patients; however, this finding did not modify the delay in the diagnostic time with an average of 8 months (in one case, the diagnostic delay was 2 years), unlike the report of Buonsenso et al. [9], where the patients with a history of positive contact had more timely diagnosis. However, the median time of pre-diagnostic symptoms in our study of 8 months coincides with the UK cohort reported by Kenyon PC et al., with a median duration of symptoms before diagnosis of 7 months and the Taiwan cohort reported by Chin-Yun et al., with a median time of 2 months [10, 11].

In the present study, TB location was not related to the age or gender of the patients. The localization of OATB was defined as conventional only in 24%,in comparison to the 42% reported in adult population by Jutte PC et al. [12]. Although this presentation might be associated to a delay diagnosis in children, more data is necessary to get a solid conclusion.

Bacteriological confirmation was poor with 85% and 71% negative culture and ZN staining respectively. The sensitivity of PPD in the present study was 46%, which was below the range of 50% to 90% in some studies of OATB [13, 14]. PCR test presented a sensitivity of 55% which is agreement with what was reported by WHO expert group [15]. The diagnostic method par excellence was histopathological study with a sensitivity of 97%. Of the 15% positive cultures, the presence of three M. bovis BCG Tokyo 172 strains calls for attention and represents the basis of a report on the adverse effect of the vaccine. This strain contains 30 million CFU/ml as compared to Pasteur 1173P3 sub-strain and Danesa 1331 both with 2 to 8 million CFU/ml; hence representing the highest concentration of this unit/ml. However, the strain presents less report of reactogenicity [16], and the presence of osteitis by BCG has a frequency of 1 case per million doses applied [17]. BCG osteitis cases have been described in countries with routine immunization programs. This finding highlights that M. bovis should be considered as an etiological agent in these countries. Further analysis in other countries is required to confirm this data.

All the patients in this study presented osteolytic lesions followed by abscess in 79% of the cases. However, the typical triad known as Phemister triad was only present in 9%, highlighting the importance of being on alert in the presence of unique osteolytic lesions. In patients with diagnosis of Pott’s disease, the use of simple radiographs may cause a delay of up to 19 months in the diagnosis; therefore, more advanced studies such as computer axial cosmography (CAT) magnetic resonance (RM) are extremely necessary [18]. In our population, the most common site of affectation was from T5 to T9 with the presence of osteolytic lesions, xiphosis and spinal cord compression in all the cases. Chest X-rays were also performed in search of concomitant pulmonary tuberculosis, finding alterations in 16% of the cases. However, it was confirmed only in two patients with sputum positive culture for M. tuberculosis which corresponds to 4% of the population and which was below the range of 6.9% to 29% reported in the literature [19, 20].

For the medical treatment, first-line anti-Tb drugs were used in all cases and antibiotics such as macrolides and quinolones were only added in three patients due to poor clinical evolution and isolation of M bovis BCG which is resistant to pyrazinamide [20]. The average total duration of treatment was 13 months which is in line with the recommendation of WHO [21]. During the treatment, only two adverse effects were reported. Evolution was favorable in all the cases and mortality rate was 0% and coincides with what was reported in other studies where the frequency of 0% to 16% was found [22].

We described a significantly low treatment abandonment of 2% of one patient after the intensive phase. According to a systematic review carried out by Kruk et al. [23], a lineal increase of 7% is estimated in the abandonment of treatment after the fourth week.

Surgical treatment in patients with vertebral involvement was necessary in 25% to 50% of the cases and in patients with diagnostic delay; this treatment may be as high as 98% [24, 25]. In the present study, all patients with Pott’s disease required surgical management with drainage of abscess, debridement and stabilization due to diagnostic delay which in these cases was 9 months on average. In the patients with extraaxial involvement, 60% underwent surgical procedures such as arthrodesis, curettage, cryotherapy and placement of bone substitution. A joint approach with the Immunology Department in search of primary immunodeficiency was undertaken without finding positive results; although, some reports described a frequency of concomitant presentation of HIV with osteoarticular TB of 10 to 27%, mainly in adult population [26].