Sternal mycobacterial infections

Abstract

Sternal mycobacterial infections are rare. Due to the rarity, its clinical characteristics, diagnoses, and regular management strategies are still scanty. A total of 76 articles on this topic were obtained by a comprehensive literature collection. The clinical features, diagnosis, management strategies and prognosis were carefully analyzed. There were totally 159 patients including 152 (95%) cases of tuberculosis (TB) and seven (5%) cases of non-TB sternal infections. Sternal mycobacterial infections can be categorized into three types: Primary, secondary, and postoperative, according to the pathogenesis; and categorized into isolated, peristernal, and multifocal, according to the extent of the lesions. Microbiological investigation is more sensitive than medical imaging and Mantoux tuberculin skin test in the diagnosis of sternal infections. Most patients show good responses to the standard four-drug regimen and a surgical intervention was necessary in 28.3% patients. The prognoses of the patients are good with a very low mortality. A delayed diagnosis of sternal mycobacterial infections may bring about recurrent sternal infections and sustained incurability. An early diagnosis and prompt antibiotic regimens may significantly improve the patients' outcomes.

Sternal tuberculosis (TB) infection is rare.[1,2] Sternal infection with rapidly growing mycobacteria is even rarer, but outbreaks can occur in high epidemiologic areas.[3] For sternal mycobacterial infections, diagnosis may be delayed, predisposing to recurrent sternal infections and sustained incurability.[4,5,6,7,8,9] Besides, the differential diagnosis should be established with other conditions including chronic pyogenic osteomyelitis, tumors, and other conditions of the like.[1] Due to the rarity, its clinical characteristics, diagnoses, and regular management strategies are still scanty. The present study aims to give a comprehensive analysis of sternal mycobacterial infections.

Methods

Medline and Google search engine were searched for publications reporting on sternal mycobacterial infection of the year range 2000-2013. The search terms included “sternum”, “tuberculosis”, “non-tuberculosis”, and “mycobacterium”. All the articles, titles, and subject headings were screened carefully for potential relevance.

Data were extracted mainly from the text. Variables included study population, demographics, clinical manifestations of sternal infection, predisposing risk factors, previous heart operation, interval between cardiac surgery and sternal infection, sites of infections, diagnostic imaging, pathogen investigations, and anti-mycobacteria as well as surgical management strategies, follow-up length, and main outcomes.

The effect time was defined as duration from the start of an anti-mycobacterial regimen to the significant improvement of the clinical manifestations, such as weight gain, swelling subside, and discharge stop, etc.

Numerical data were expressed as mean ± standard deviation (SD) and compared with the independent samples t-test; while count data were expressed as percentages and compared with Fisher's exact test. P < 0.05 was considered statistically significant.

Results

Patients' demographics

As a result, 76 articles[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76] were obtained, including 56 (73.7%) case reports,[1,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57] seven (9.2%) case series,[2,58,59,60,61,62,63] eight (10.5%) original articles,[64,65,66,67,68,69,70,71] two (2.6%) quizzes,[72,73] one (1.3%) congress abstract,[74] one (1.3%) how-to-do-it,[75] and one (1.3%) imaging.[76] A total of 159 patients with sternal mycobacteria were involved. There were 98 (64.9%) males and 53 (35.1%) females, while eight patients' gender was not stated. The patients aged 41.3 ± 21.9 (range, 0.75-84; median, 42) years (n = 96). A significant age difference was noted between male and female patients (45.0 ± 21.1 years vs 34.9 ± 22.5 years, P = 0.038).

A total of 107 (67.3%) cases were primary,[2,5,9,10,13,14,16,17,18,19,21,22,23,28,30,31,32,33,34,38,39,40,42,43,44,48,50,52,53,56,59,60,61,62,64,65,66,67,69,70,72,73,74,75,76] 33 (20.8%) were secondary,[6,11,12,15,20,24,25,26,29,35,37,46,47,51,55,62,64,66,67] and 19 (11.9%) were postoperative.[1,3,4,7,8,27,36,41,45,49,54,58,62,63,68,71]

Regional epidemiology

Regional epidemiology analysis revealed that the reports of sternal mycobacterial infections were mostly reported from India and USA, and the highest incidences of the patients were in India, South Africa, and USA [Table 1].

Table 1.

Regional epidemiology of sternal mycobacterial infections

Predisposing risk factors

Sternal mycobacterial infections were categorized into three types: Primary (107, 67.3%), secondary (33, 20.8%), and postoperative (19, 11.9%).

The predisposing risk factors responsible for primary sternal TB were described in 29 patients including infective spondylitis (14, 48.3%),[67] human immunodeficiency virus infection (six, 20.7%),[67] diabetes mellitus (five, 17.2%),[2,28,39] travelling (two, 6.9%),[53,56] trauma (ankle injury; one, 3.4%),[17] and immunocompromised (underling disorder not indicated) plus malnutrition (one, 3.4%).[10]

The underlying causes for secondary sternal TB were a previous history of lung TB in 12 (36.4%)[15,25,55,62,66,67,73] (one was reactivated lung TB[73] and one was associated with lymph node TB[55]), TB contact in 10 (30.3%),[20,24,46,47,66] extrapulmonary TB in six (18.2%) including TB lymphadenitis in three[6,51,55] (cervical,[55] supraclavicular,[6] and unstated[51] in one each) and vertebral TB,[67] multifocal skeletal TB,[67] and abdominal and spinal TB[37] in one each; and Bacille de Calmette et Guérin vaccination in five (15.2%) patients.[11,12,26,29,35]

The underlying operations for postoperative TB were coronary artery bypass grafting in eight (42.1%);[1,8,27,36,41,45,54,58] aortic valve replacement in three (15.8%);[27,58,63] and atrial septal defect repair,[62] mitral valve replacement,[7] left atrial myxoma excision with reexploration,[3] and redo-Bentall operation for root aneurysm[4] in one (5.3%) each; and cardiac (cardiothoracic) operation (the procedure details were not given) in four (21.1%) patients.[49,68,71] The latency from the cardiothoracic operation to the onset of sternal TB was 24.9 ± 58.2 (range, 0.5-228; median, 3) months (n = 15)[1,3,4,7,8,27,36,41,45,54,58,62,63] by excluding one patient whose latency was recorded as “rapid”.[17] There was no statistical significance between the latency after coronary artery bypass grafting and that after noncoronary artery bypass grafting operation, although the latter was longer (15.3 ± 19.8 vs 35.9 ± 84.8 months, P = 0.514).

Clinical manifestations

Isolated sternal TB was observed in 96 (60.4%) patients,[1,2,3,4,5,7,9,12,18,19,20,21,22,23,26,29,30,33,34,35,38,39,40,41,42,47,48,49,50,53,60,61,63,64,65,66,67,69,70,72,74] sternal TB with peristernal tissue invasions (muscles, cartilages, and joints) in 32 (20.1%) patients [Table 2], and sternal TB with concurrent TB of other organs in 31 (19.5%) patients [Table 3]. The infection sites of the sternum were recorded in 109 (68.6%) patients with manubrium being the most commonly involved portion [Table 4].

Table 2.

Peristernal tissue invasion

Table 3.

Concurrent tuberculosis of other organs

Table 4.

The infection sites of the sternum of 109 patients

Clinical manifestations could not be tracked in two original articles including three patients.[68,71] In the remaining 156 patients, sternal swelling and local pain were the most common symptoms [Table 5].

Table 5.

Clinical manifestations of 156 patients with sternal mycobacteria infection

Lymphadenopathy was noted in 17 (10.7%) patients,[14,16,17,20,27,28,33,34,35,38,46,52,55,65,66] of which mediastinal lymphadenopathy was the most common [Table 6].

Table 6.

Lymphadenopathy associated with sternal mycobacteria infection

Diagnosis

Osteolytic lesions were present in 10/39 (25.6%) patients on chest X-ray films and in 18/77 (23.4%) patients on chest computed tomography (χ² = 0.1, P = 0.788) [Figures 1 and 2].

Figure 1.

Signs of sternal tuberculosis on chest X-ray films

Figure 2.

Signs of sternal tuberculosis on chest computed tomography

The alternative diagnostic techniques included magnetic resonance imaging in 17 (all 17 were positive),[2,9,24,61,67] bone scan in five (four were positive[5,8,12,30,44] and one was negative[26]), ultrasound in two (both were positive[37,74]), and positron emission tomography-computed tomography in one patient (it was positive[29]). A Mantoux tuberculin skin test (purified protein derivative, 5 tuberculin units) was reported in 22 patients. There were 18 (81.8%) positive[2,6,18,19,21,24,26,33,34,38,39,52,57,65,72] with a diameter of 19.7 ± 6.7 (range, 10-32; median, 20) mm (n = 11), and four (18.2%) were negative results.[30,31,45,55]

Two patients had an additional diagnosis unrelated to sternal infection: One was pseudoaneurysm of the innominate artery[17] and the other was gynecomastia.[22]

Microbiology

There were 152 (95%) cases of TB and seven (5%) cases of non-TB sternal infections.[3,7,41,49,51,58,63] The non-TB mycobacteria did not show any predilection to any strains [Figure 3].

Figure 3.

Distribution of pathogens of sternal nontuberculosis mycobacterial infections

The pathogens were evidenced by at least one of the four investigations including Ziehl–Neelsen stain, histopathology, culture, and molecular biology. Statistical analysis revealed that the microbiological investigations were superior to medical imaging, and Mantoux tuberculin skin test; and the histopathology for pathogen determination was superior to other microbiological investigational methods in term of sensitivity [Table 7].

Table 7.

Mycobacterial investigations

Management and prognosis

The anti-TB regimens were not mentioned in 25 (15.7%) patients and unspecified in 35 (22.0%),[12,15,17,20,27,36,51,53,56,60,62,64,66,70,71,72,73,75,76] and the first-line anti-TB regimen was unspecified in one (0.6%) patient.[46]

In the remaining 98 patients, a standard four-drug regimen was prescribed in 55 (56.1%) patients[1,4,5,6,13,14,16,19,22,23,24,27,33,34,35,38,39,40,42,43,44,47,48,50,52,55,57,59,61,64,68,74] (28 patients with a successive two-drug[14,19,24,25,35,42,44,48,64,74] and two patients with a successive three-drug regimen[60]), a three-drug regimen in seven (7.1%) patients,[8,9,10,11,18,28,29] a two-drug regimen in two (2.0%) patients,[26,63] second-line drug in 13 (13.3%) patients,[21,32,41,69] successive first- and second-line drugs in 14 (14.3%) patients,[67] adjusted regimen in one (1.0%),[54] and alternative antibiotic regimen in six (6.1%) patients[3,7,46,49,58] [Figure 4].

Figure 4.

Antibiotic regimens for sternal mycobacterial infections

The effect time of the anti-TB regimen was described in 23 (14.5%) patients. It was expressed as “immediate” or “rapid” in 3 patients.[4,27,50] In the remaining 20 patients, the effect time was 6.9 ± 3.8 (range, 2-12; median, 8) weeks.[6,20,21,24,25,30,34,39,41,42,43,44,48,52,54,59,64,74,75] The duration of treatment with the anti-TB regimen was described in 53 (33.3%) patients. It was expressed as “long-term” in one patient.[4] In the remaining 52 patients, the duration of treatment was 12.3 ± 5.9 (range, 0.5-30; median, 12) months. Most patients responded well to the treatment, but poor responses were reported in at least 13 (8.2%) cases of this patient setting due to multidrug resistance.[40,67,69] One patient required corset for vertebral lesions.[76] Surgical management was necessary in 45 (28.3%) patients, but one of them declined.[32] A three-stage operation was necessary in two patients[4,27] and a two-stage in four[3,8,27,40] patients. Reconstruction of the chest wall was the most common surgical procedure applied [Table 8].

Table 8.

Surgical managements of sternal mycobacterial infections

Patients were under a follow-up (15.4 ± 12.8; (range, 1-60; median, 12) months)[1,3,7,8,9,11,13,15,16,20,21,22,23,25,26,27,28,29,30,31,36,40,41,42,43,44,47,48,52,54,58,59,62,64,66,67,69,73,75] Prognosis was described in 124 patients: Event-free survival in 117 (94.4%), one to two recurrences in two (1.6%),[4,9] failure to cure in four (3.2%),[5,6,7,8] and death in one patient (0.8%).[63] By comparison, the event-free survival was significantly higher in the conservative than in the surgical patients [Table 9].

Table 9.

Prognosis of sternal mycobacterial infections

Discussion

Recent investigations of global epidemiology of TB revealed that the majority of TB patients are in Asia (59%) and Africa (26%), with the largest estimated incidence in 2011 in India and China.[77] The epidemiology of sternal mycobacterial infections described in the present study may partly correspond to that of the TB cases in terms of the regional distribution in particular in India. Due to the limitation of literature collection to English publications, the sternal mycobacterial infections in China where the authors publish articles in Mandarin were neglected.

The mechanisms for primary sternal TB with a basis of TB of other organs are uncertain.[54] Subclinical contamination of TB[78] and immunocompromised condition[10] can be the possible pathogenesis. Living or travelling to a TB endemic area, and positive TB contact are predisposing risk factors of sternal TB infection.[36] Left internal mammary artery harvest in coronary artery bypass grafting and the presence of diabetes mellitus may increase the risk of postoperative sternal infections.[68] Isolated sternal TB is more common than concurrent TB of the sternum and other organs,[54] as it was illustrated in this study that concurrent TB infections accounted for 19.5% of the cases. Moreover, postoperative sternal mycobacterial infection might have a latency that varied from weeks to years.[36]

Imaging can be helpful to characterize sternal TB infection; however, the positivity rate of chest radiograph seems to be lower than that of chest computed tomography for the diagnosis of sternal infections. The characteristic osteolytic changes of the sternal lesions were only shown in 25% cases by chest radiograph as compared to 55.9% by chest computed tomography as identified in the present study. Furthermore, a Mantoux tuberculin skin test can be positive in 81.8% of the tested patients.

An alternative reliable method of diagnosis of sternal mycobacterial infections is microbiology investigation, either by Ziehl–Neelsen stain, histopathology, culture, or molecular biology. Smear for Ziehl–Neelsen stain was positive for acid-fast bacilli, or a positive culture for Mycobacterium tuberculosis and/or a positive smear for acid-fast bacilli.[67] Histopathology of fine-needle aspiration or of resected infected tissues may show clusters of epithelioid cells and giant cell of Langerhan's cells. There has been agreement between polymerase chain reaction (PCR) and histopathological studies in determining the pathogens responsible for sternal TB infections.[54] In general, the microbiological investigation is superior to medical imaging, Mantoux tuberculin skin test, and other investigation methods in terms of sensitivity and positive rate.

The target of sternal TB treatment relies on good responses to the anti-TB medication with standard four-drug regimen for 6-9 months.[53] Nevertheless, in the present study, two patients (14.2%) had multidrug-resistant TB but responded to second-line anti-TB therapy with streptomycin, ethionamide, ofloxacin, pyrazinamide, and ethambutol for 6 months; followed by ethionamide, ofloxacin, pyrazinamide, and ethambutol for 14 months more.[67] The surgical operation can be drainage of abscess, debridement, removal of wires, curettage, resection, and/or chest wall reconstruction.[49] In addition, hyperbaric oxygen therapy can be an adjunct therapy of the sternal infection.[71] In addition, a combined multidrug regimen and surgical debridement can be a precondition of successful management of sternal mycobacterial infections.[54] However, sternal TB infections do not always require a surgical treatment,[27] which is always indicated for the refractory cases.[1] Lahiri et al.,[79] reported that 2.2% of cases warranted a surgical procedure, of which 0.9% was for TB of the ribs and sternum.

Patients with anti-TB regimen often take effect in shorter period of time, such as weight gain in 3 months, swelling subside in 8 months, and discharge stop in 10.5 months, etc.[64] Complete cold abscess resolution has been observed within 4 months of treatment.[67] The clinical recovery can precede radiological healing for a mean of 3.2 months.[67]

Conclusions

Sternal mycobacterial infections are rare. Sternal TB infections show a much higher incidence than that of sternal non-TB infections. Microbiological investigation is more sensitive than medical imaging and Mantoux tuberculin skin test in the diagnosis of sternal infections. Most patients show good responses to multidrug regimens and a surgical intervention is necessary in 28.3% patients. The prognoses of the patients are good with a very low mortality.