The application of metagenomic next generation sequencing in diagnosing tuberculous otitis media: a case report and review of the literature

Abstract

Background

Tuberculous otitis media is a chronic Mycobacterium tuberculosis infection of the middle ear tissues. Diseases with varied and insidious clinical features can make diagnosis difficult and delay treatment.

Case presentation

Here, we document a case of tuberculous otitis media in a 46-year-old ethnic Han woman that manifested as nonspecific chronic otitis media. A mastoidectomy and tympanoplasty were performed for the initial diagnosis of cholesteatoma. The histopathology of the tissue specimen revealed granuloma formation with necrosis. Staining for acid-fast bacilli and the polymerase chain reaction method for Mycobacterium tuberculosis yielded negative results. However, the chest computed tomography scan demonstrated a pulmonary miliary nodule. Next, metagenomic next-generation sequencing was applied and the Mycobacterium tuberculosis was identified. The patient recovered after receiving antituberculous treatment.

Conclusion

This report highlights the application of novel diagnostic tools such as metagenomic next-generation sequencing as a supplementary method for the diagnosis of tuberculous otitis media in highly suspected patients.

Introduction

Tuberculous otitis media is a rare type of chronic suppurative infection of the middle ear and mastoid, accounting for 0.04–0.9% of chronic suppurative otitis media cases [1]. The historically classical presentations of tuberculous otitis media (TOM) include painless otorrhea, multiple tympanic membrane (TM) perforations, and facial palsy (FP) [2]. However, these typical symptoms are often absent nowadays. The clinical signs are typically diverse and resemble other types of chronic otitis media cases [3, 4]. Complications occur as disease progresses including FP, labyrinthitis, osteomyelitis, abscess formation, and infection to the central nervous system [5]. Early diagnosis and prompt treatment are therefore imperative.

Microbiology or histopathologic characteristics for tuberculosis are commonly used to diagnose TOM, although these conventional approaches frequently produce false-negative results [6,7,8]. The diagnosis of TOM is sometimes delayed due to nonspecific clinical characteristics that induce a low index of suspicion, as well as false-negative test results [9]. Clinical metagenomic next-generation sequencing (mNGS), an unbiased culture-independent and hypothesis-free sequencing technology, can detect etiologic pathogens quickly and accurately and has been widely used in the diagnosis of infectious disorders, particularly when conventional testing is negative [10]. Here, we present a case of TOM in which traditional laboratory findings were negative, and the diagnosis was later confirmed using mNGS.

Case presentation

Here, we present a 46-year-old Han woman came in with a 5-month history of unilateral, left-sided tinnitus and hearing loss. She had no family history of ear disease. She was diagnosed with primary tinnitus and treated with Ginkgo biloba extract. Nonetheless, her problems persisted. Two months later, she had refractory painless otorrhea and was otherwise well. She then came to our clinic for aid. Otoscopy revealed a 50% central perforation with pus and granulation tissue covering the mesotympanum. Audiometry showed a mixed hearing loss at 0.25, 0.5, 1, and 2 kHz with air-bone gaps of 35, 25, 40, and 35 dB, respectively. During this visit, the diagnosis was amended to probable cholesteatoma. A high-resolution CT scan of the temporal bone revealed complete opacification of the left mastoid air cells and middle ear cavity, with erosion of the tympanic scutum. The bony labyrinth was intact (Fig. 1). The patient underwent mastoidectomy and tympanoplasty. During surgery, an abundance of granulated tissue was seen in the mastoid and middle ear cavities. Tegmen mastoideum was eroded. Tissue samples were sent for histology, which demonstrated granuloma formation with necrosis. Acid-fast bacilli (AFB) staining and polymerase chain reaction (PCR) amplification showed no evidence of Mycobacterium tuberculosis (M.Tb). However, her chest computed tomography (CT) scan indicated a pulmonary miliary nodule. As a suspicion of TOM remained given clinical findings, mNGS was done, which revealed M.Tb in formalin-fixed tissue specimens (Fig. 2). Antituberculous medication was taken for 6 months including isoniazid, rifampicin, ethambutol, and pyrazinamide. Following therapy, the patient’s ear was dry, and her TM was intact. The follow-up time was 19 months.

Fig. 1

Images of the patient. Axial (A) and coronal (B) temporal bone computed tomography images. Complete opacification was detected in the left mastoid air cells and middle ear cavity, along with erosion of the tympanic scutum. The bony labyrinth was intact. The white arrows indicate the granulomatous inflammation in A and erosion of the tympanic scutum in B

Fig. 2

The metagenomic next-generation sequencing results. Mycobacterium tuberculosis was detected in formalin-fixed tissue specimens. Mycobacterium tuberculosis accounted for 15.65% of all microbial reads

Discussion

TOM is a rare form of extrapulmonary tuberculosis, accounting for 0.1% of all tuberculosis cases. It is usually caused by an infection of the lungs, larynx, pharynx, or nose [11]. A chronic suppurative ear infection associated with known or suspected tuberculosis raises clinical suspicion of TOM. The pathogenesis of TOM typically involves aspiration through Eustachian tube, hematogenous dissemination from distant tuberculous foci, and direct implantation through the external auditory canal and tympanic membrane perforation [11, 12].

Clinical characteristics

The characteristic features of TOM were described in 1953 as painless otorrhea, multiple tympanic perforations, ipsilateral FP, severe hearing loss, and pale granulation tissue [13]. Otorrhea is the most common presentation. It may be painless or coupled with otalgia due to potential bacterial superinfection [3, 14]. Although otorrhea can be varied, it is always refractory to standard treatments in all cases.

The appearance of TM varies in degree, which can be single or multiple perforations or even intact. Recent studies have reported a frequency of multiple perforations ranging from 2% to 10% [2, 6, 7, 15, 16]. It was presumed that multiple perforations were only an early pathologic feature that would eventually fuse to form a large single perforation at a later stage. Most individuals who did not have perforations at the time of assessment would probably develop perforations if left untreated [17]. Aside from the multiple perforations of TM, several otoscopic findings (pale debris, exuberant granulation tissue, or whitish exudates not readily aspirated) may also indicate tuberculous etiology [9, 18].

According to the literature, the incidence of FP was about 26.9% in TOM cases [4], which was greater than that in nontuberculous otitis media. FP seems to arise during the acute phase of the disease [7]. Thus, the onset of FP after a short period of symptoms should raise suspicion of TOM, especially in the absence of cholesteatoma [11, 19].

Profound hearing loss is classically known to be a characteristic feature of TOM, which can be conductive, sensorineural, or mixed. Conductive hearing loss begins early in the illness, with the onset of TM perforation. When labyrinthitis happens, mixed or sensorineural hearing loss is noted. Moderately severe hearing loss is more frequent in TOM than in nonspecific chronic otitis media and generally more marked than suggested by otoscopy [6, 7, 11].

The most frequent CT findings in TOM were soft tissue attenuation throughout the middle ear cavity with no bone erosion and the preservation of mastoid air cells without sclerosis [6, 20]. Cho et al. reported that cortical erosion of the mastoid bone in patients with the typical otoscopic imaging of chronic otitis media was a significant finding connected to TOM [7]. Cochlear fistula formation and widespread bone destruction of the skull base may indicate advanced TOM [20]. Although specific CT results for TOM may not exist, they are useful in determining the extent of involvement of various regions of the temporal bone.

Diagnosis: microbiological testing, histopathology, and mNGS

TOM cannot be established through clinical manifestations due to the scant incidence, and diverse and non-specific features mimicking other types of chronic otitis media. It is generally diagnosed via microbiology or conclusive histology. Most studies utilized more than one diagnostic tool. Common microbiological tests include AFB staining, ear exudate or tissue culture, and purified protein derivative (PPD) assays. However, AFB test or culture is readily disturbed and not as reliable as other bacteriological examinations. Tuberculous lesions of the ear usually present with low bacillus counts as other extrapulmonary tuberculosis [14]; continuous use of ear drops, particularly those containing neomycin, can further reduce bacterial concentration. [18]; in addition, secondary infection by other microorganisms can interfere with AFB staining or culture, hence AFB tests often give a poor yield, and delay the microbiological confirmation [8, 14, 19]. A positive PPD exam simply discloses a mycobacterial infection at some point in the past, therefore it is only an indicator, not diagnostic of tuberculosis, and is not the preferred method of diagnosing [5].

Histopathological evaluation is an ideal diagnostic tool for TOM. Typical tuberculous lesions include well-formed granulation with epithelioid cells, giant Langerhans cells, and caseous necrosis [5, 6, 11]. The generally acknowledged histopathologic criteria can help to establish a diagnosis. The diagnostic efficacy of histopathology was reported to range from about 80–90% [3, 4, 7], compared with 50% reported by Vaamonde et al. [6].

The PCR approach has been shown to improve the diagnostic efficiency and accuracy of TOM. It can rapidly detect M. Tb DNA in a variety of clinical specimens when AFB tests or histologic examinations are negative [7, 21]. PCR provides higher sensitivity for detecting M.Tb than acid-fast culture, with reported positive rates of up to 88% [7]. Although the utility of PCR for TOM diagnosis has not been proved, patients who exhibit suggestive clinical indications, distinctive CT features, and a positive PCR can be considered to have TOM.

The negative results of necessary exams to prove tuberculosis infection made it difficult to diagnose our patient. To identify pathogenic bacteria, we used mNGS technology on formalin-fixed tissue specimens. mNGS is an impartial and comprehensive approach for detecting and classifying microorganisms. It can detect multiple organisms including bacteria, viruses, fungi, atypical pathogens, and parasites in a clinical specimen at once [10]. Antibiotics rarely alter its diagnostic efficiency. Studies have revealed that mNGS offered high sensitivity and specificity for diagnosing tuberculous diseases. An analysis performed by Li et al. reported that the combined sensitivity and specificity of mNGS for diagnosing M.Tb were 0.69 (0.58–79) and 1.00 (0.99–1.00), respectively. Cerebrospinal fluid had a lower sensitivity of 0.58 (0.39–0.75) [22]. As a result, the mNGS might be employed as a first-line diagnostic tool. Recently, mNGS has outperformed established approaches in the diagnosis of complex ear infections by detecting uncommon bacteria and fungi [23,24,25]. One advantage of this technology is the ability to detect culture-negative bacterial and fungal infections. To our knowledge, no study has reported the application of mNGS for diagnosing TOM. Our case reinforces the fact that mNGS can be a supplementary method for identifying M.Tb in suspected cases with negative results from conventional examinations.

Xpert Mycobacterium Tuberculosis/Rifampicin (MTB/RIF) is a widely used molecular diagnostic tool for tuberculosis, providing rapid results by detecting M. Tb DNA and rifampicin resistance directly from clinical samples. Its primary advantage is its rapid turnaround time (typically within hours), making it suitable for point-of-care settings. Xpert MTB/RIF has shown high sensitivity for pulmonary tuberculosis, but its effectiveness in extrapulmonary infections is more limited compared with mNGS [26,27,28]. Furthermore, Xpert MTB/RIF’s reliance on a specific target gene for detection limits its ability to detect other pathogens or coinfections that may be present in TOM cases. Considering this situation, Xpert MTB/RIF was not used in this patient.

While mNGS offers several advantages, it is not without challenges. The feasibility of widespread use of mNGS for TOM diagnosis remains a concern due to its high cost and the need for specialized equipment and expertise. While mNGS holds great promise, its use may be limited to specialized centers or research settings in the short term. The high cost of sequencing, coupled with the need for bioinformatics support to interpret results, means that it may not yet be practical for routine clinical application, particularly in resource-limited settings. However, it may be justified for diagnostically challenging cases as long as economic conditions allow.

Conclusion

The clinical signs and symptoms of TOM are diverse and nonspecific. Diagnosis remains a significant challenge for clinicians. A combination of clinical vigilance and appropriate diagnostic procedures will aid in the early identification of TOM. Traditional methods of detecting tuberculosis frequently produce false-negative results. When traditional investigations are negative in patients with a high suspicion of TOM, a mNGS test is recommended.