The patient was allergic to TMP-SMX and was treated for 5 weeks with amikacin and ceftriaxone (the reasoning behind the choice is unknown), and discharged home on amoxicillin-clavulanate 625mg TDS.

Three months later, his outpatient notes state that the sinus on chest wall now closed but patient has a central neck mass. Goitre? Cyst? Patient noticed increasing size over last 6 days. Patient referred to the surgeons. Ultrasound revealed a complex cystiEDc mass in the left lobe of thyroid most likely infective. Over the next week the anterior neck swelling enlarged with associated dysphagia and dyspnea.

On examination, he was wasted, afebrile, with a tender, enlarged goitre and generalized shotty lymphadenopathy. Investigations showed:

• WCC 10.6 x 109/L, Hb 8.4g/dl, Platelets 298 x 109/L
• CRP 202 mg/L
• TSH < 0.01, t3- 3.1, t4- 43.0
• Blood cultures – negative
• CT scan - thyroid abscess with oesophageal compression but no tracheal compromise.

On the advice of the endocrinologists, prednisone and Luggol’s iodine was started. An ultrasound-guided aspiration of thyroid mass aspirated 18ml of pus from the left lobe followed by incision and drainage. Definitive surgery was not performed due to the risk of thyroid storm. Steroids were stopped after 10 days.

Four days after the I&D, the patient complained of diarrhoea and stool was positive for C. difficile toxin. Treated with metronidazole.

Pus cultured from thyroid aspirate grew N. brasiliensis

Modified Kinyoun stain of thyroid biopsy

The patient developed a right supraclavicular swelling which was incised & drained.

On the basis of his new resistance profile, he stopped amoxicillin-clavulanate and started linezolid 600mg BD.

A CT brain confirmed a poorly enhancing multi-cystic right cerebellar hemisphere mass with surrounding oedema, early hydrocephalus and attenuation of 4th ventricle. Large necrotic submental nodes and multiple necrotic mediastinal and anterior neck nodes were noted on lower cuts.

Discussion

Laboratory Diagnosis of Nocardia Species

Diagnosis of nocardiosis can be challenging as the organism is relatively slow-growing and often overgrown by other micro-organisms. The first clue to the presence of a Nocardia species is the Gram stain, which typically demonstrates filamentous beaded Gram positive bacilli. A modified Kinyoun stain that then demonstrates acid-fast bacilli is further supportive evidence and attempts should be made to isolate this organism through use of selective media.

Upon isolation of what appears to be a Nocardia species (typical chalky-white, cotton-wool appearance, musty smelling colony) a definitive genus and species identification can only be made using molecular diagnostic modalities. Biochemical tests e.g. hydrolysis reactions are generally indiscriminatory and unreliable.

Commercial identification systems have similar limitations to traditional phenotypic methods and are unreliable. Despite its widespread use and applicability in bacterial identification MALDI-TOF has not demonstrated sufficient accuracy in the identification of aerobic actinomycetes, including Nocardia spp. This is possibly related to the lack of a suitably representative database.

Serological tests play no role in the diagnosis of nocardiosis. Molecular methods including 16S rRNA and other target gene (e.g. secA, HSP-65) sequencing are currently the gold standard for species ID, although there are limitations in interpretation of sequence data.

Antimicrobial Susceptibility Testing of Nocardia

• The CLSI recommends the broth microdilution (BMD) method for antimicrobial testing of aerobic actinomycetes.
• Alternative methods include the E-test™ and BACTEC radiometric methods.
• A study in South Africa did comparative antimicrobial susceptibility testing (AST) of E-test™ and BMD.
• All 39 Nocardia isolates demonstrated susceptibility to TMP-SXT, amikacin & linezolid. Only these 3 antibiotics demonstrated 100% categorical agreement when comparing the 2 AST methods.
• Other antibiotics fell short of the international standard of >90% concordance and < 3% major errors />no very major errors.
• AST of Nocardia spp is challenging and difficult to standardize due to issues with inoculum preparation. There is also poor correlation between susceptibility patterns and the species, thus AST is essential. However, it should be done by a reference laboratory experienced in the field of nocardia AST. Based on local and global susceptibility data in the absence of confirmatory AST, TMP-SXT should form the backbone of any treatment regime. Amikacin and linezolid also provide reliable empiric cover.

Norcadia Thyroiditis

Thyroiditis due to nocardia is a rare presentation and there are only four cases reported in the literature:

• 1993- N. asteroides in patient with SLE and on treatment with prednisone and cyclophosphamide.
• 1999- N. asteroides in a liver-kidney transplant patient on tacrolimus, azathioprine and prednisone
• 2007- Norcadia spp in an elderly diabetic patient with disseminated nocardiosis.
• 2011- N. asteroides in an elderly COPD patient who received a short course of prednisone. Also had brain lesions on MRI.

It is important to note that these cases present N. asteroides as the causative species. This is probably inaccurate and these are considered cases of misidentification. N. asteroides was the name assigned to the reference strain ATCC 19247, but this was prior to molecular testing and given the inert biochemical nature of nocardia isolates many different species appear biochemically the same. Clinical strains of nocardia have not been found to match the original type strain (according to sequence-type). The taxonomic nomenclature of nocardia remains in a state of flux but the species name N. asteroides should no longer be used to designate the most commonly identified cause of nocardiosis. In all likelihood this species name will be confined to the microbiological history books. Interestingly isolates defined as N. asteroides are typically found to be identical to N. cyriacigeorgica.

References:

1. Brown-Elliott et al. Clin Microbiol Rev 2006; 19(2):259-282
2. Lewin et al. Journal of Infect 1993; 26:339-40
3. Carriere et al. J Clin Microbiol 1999; 37(7):2323-2325
4. Indumathi, Indian J of Med Microbiol 2007; 25(3):294-296
5. Su et al. J. Micr, Imm & Infect, 2011; (44):238-240
6. Lowman & Aithma; J Clin Microbiol 2010; 48(12):4534-4540
7. Conville et al. J Clin Microbiol 2012; 50(4):1270-1280
8. Brown-Elliott et al. J Clin Microbiol 2012; 50(3):670-672