This retrospective study was approved by the Institutional Review Board of the Kinki-Chuo Chest Medical Center (approval code 689) and has been performed in accordance with the ethical standards in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. We required that all patients provide written informed consent before information was collected.

Case-patient 1 was an 88-year-old HIV-negative man admitted to our hospital in 2018. He had complications from stable interstitial pneumonia and an increased productive cough and fever. Chest radiograph and transverse computed tomography showed left-side consolidation.

Two months later, sputum cultures were positive for M. shigaense, identified by partial DNA sequences of the 16S rRNA, hsp65, and rpoB genes. Thereafter, multidrug treatment was orally administered for 12 months. After »1 month of treatment, his sputum was culture negative and negative for acid-fast bacilli (AFB). The patient showed clinical improvement and decreased symptoms. He has remained culture negative for >1 year (Figure 1, panel A).

Case-patient 2 was a 78-year-old HIV-negative woman referred for an evaluation of previously diagnosed chronic BE since 2016. She had an NB-type radiologic pattern. AFB smear and culture test results were negative on 3 consecutive sputum samples.

This patient was not initially given antimicrobial drugs and was evaluated by expectorated sputum examinations at follow-up. After 2 years, M. shigaense was isolated from subsequent sputum samples on 3 occasions. Because there were no respiratory symptoms, the patient was not initially given antimicrobial drugs after diagnosis. Spontaneous culture conversion was found after 3 consecutive negative sputum cultures, and the negative status was maintained during the follow-up period of >1 year (Figure 1, panel B).

We performed drug susceptibility testing considering MAC breakpoints and using broth microdilution according to Clinical and Laboratory Standards Institute guidelines (7). Testing showed that isolates of both patients were susceptible to clarithromycin, amikacin, moxifloxacin, and linezolid.

Whole-genome sequencing was performed on an initial isolate from each case-patient (strain KC8 from case-patient 1 and strain KC354 from case-patient 2). DNA sequence libraries were prepared by using the QIAseq FX DNA Library Kit (QIAGEN, https://www.qiagen.com) using 50 ng M. tuberculosis genomic DNA, followed by paired-end sequencing using Illumina MiSeq Reagent Kit version 3 (600 cycles) (Illumina, https://www.illumina.com).

We conducted average nucleotide identity basic local sequence alignment tool analysis by using JSpecies version 1.2.1 (8). Average nucleotide identity values for strains KC8 and KC354 clearly indicated that these 2 patients were infected by M. shigaense (Table 1). Identification of single-nucleotide polymorphisms (SNPs) in these isolates, compared with those in M. shigaense JCM 32072^T, was conducted by using the BactSNP pipeline (9). KC354 had 19 SNPs, whereas KC8 had 6,826 SNPs (Table 2).

A core genes phylogenetic tree for genome sequences of 19 M. simiae complex and M. avium 104 was reconstructed by using Roary version 3.11.2 (10) (Figure 2). M. shigaense strains were most closely related to M. rhizamassiliense (11); KC8 was less related to M. shigaense JCM 32072^T than to M. shigaense SCY. Raw sequence reads of M. shigaense KC8 and M. shigaense KC354 were deposited in the DNA Data Bank of Japan Sequence Read Archive (DNA Sequence Read Archive, https://www.ddbj.nig.ac.jp/dra/index.html) under study accession no. DRA009490.

NB-type infection with M. shigaense is considered sufficiently indolent that careful longitudinal appraisal without therapy is safe and poses little risk for rapid progression (12,13). The previously reported case-patient with NB-type pulmonary M. shigaense disease did not have a history of immunosuppressive therapy or clinical symptoms but was treated successfully (6). Clinical data, including the presence or absence of underlying BE as a concurrent condition, was insufficient (6). Our NB-type patient with chronic BE was considered not to have a clinically serious condition and required no treatment. In contrast, the patient with progressive FC-type infection was given chemotherapy that resulted in improvement observed by computed tomography (Figure 1). We showed that clinical features of M. shigaense disease resemble those of MAC disease, but radiographic differences indicated that MAC disease was more serious.

We also observed genomic diversity of M. shigaense. Our comparative genomic analysis showed that strain KC354 obtained from the NB-type patient was closely related to M. shigaense JCM 32071^T. In contrast, strain KC8 obtained from the FC-type patient showed a large number of SNPs when compared with the type strain. This result might explain the increased virulence of strain KC8.

More than 50% of stable pulmonary MAC disease patients have spontaneous sputum conversion without treatment (3,14). Bacterial genotypic comparison between patients with spontaneous sputum conversion and those with serial sputum-positive cultures might identify patients who are likely to profit from antimicrobial drug therapy. Our next-generation sequencing findings indicated that differences in the genetic background of the pathogens might aid physicians with clinical decisions regarding therapy initiation. We believe that accumulation of genomic data for clinical strains should be helpful for future comparative studies and will probably lead to diagnosis of more cases. Further studies of this relatively new pathogen with larger sample sets are needed to identify clear, reliable, and clinical markers that predict the virulence of M. shigaense.

Current treatment for pulmonary mycobacterial disease recommends ethambutol, rifampin, and macrolides (1,13). Little is known regarding response of antimicrobial agents and clinical outcome for this rare species. Therefore, a better understanding of drug susceptibility of the pathogen is necessary to provide suitable treatment. Most patients reported with M. shigaense disease were given multiple agents: 4 patients showed improvement after receiving a clarithromycin-based treatment for 4–12 months (no data were available for linezolid) (6,15). In comparison to previous cases of M. shigaense disease (5), for which the pathogen was susceptible to clarithromycin and moxifloxacin, our isolates showed susceptibility to amikacin. We believe that a synergistic response of antimicrobial drugs against M. shigaense requires further evaluation.

In summary, we found that the NB-type M. shigaense pulmonary strain was closely related to the cutaneous reference strain, but the more pathogenic FC-type strain differed considerably. Our results for this rare species open possibilities for further investigation into this neglected NTM disease and provide indications for the need for therapy.

Additional information on 2 new cases of pulmonary infection by Mycobacterium shigaense, Japan.