Letter to the Editor: Evidence for a two-step species-level pulmonary nocardiosis diagnostic approach

Abstract

Pulmonary nocardiosis, a rare and diagnostically challenging infection, usually presents with heterogeneous radiographic findings, compounded by the low sensitivity of traditional confirmatory cultures. In their most recent work, Wang et al analyzed 102 patients with pulmonary nocardiosis to address these concerns, investigating species-characteristic imaging patterns, pathological associations, and the role of metagenomic next-generation sequencing (mNGS) in the diagnostic approach. High-resolution computed tomography (CT) in adult patients with pulmonary infections caused by Nocardia wallacei was demonstrated to have a sensitive (85.71%) and specific (83.34%) presentation of bronchopneumonia in relation to the five Nocardia species in the sample with CT data. The authors also compared traditional cultures to mNGS, finding that traditional cultures and mNGS were concordantly positive in only 3.3% of cases. This letter supports a combined radiologic and molecular diagnostic approach, enabling earlier and more accurate species identification in pulmonary nocardiosis, thereby informing treatment decisions, and enhancing epidemiologic understanding.

Core Tip: The current understanding of pulmonary nocardiosis as presenting with heterogeneous computed tomography (CT) findings is changing, as in their recent work, Wang et al recently suggested a potential role for radiology in species-level diagnosis. By integrating high-resolution CT and molecular-based diagnostics, the authors provide evidence for a two-step diagnostic workflow for species identification in pulmonary nocardiosis, using imaging to rapidly build species-level suspicion followed by confirmation with metagenomic next-generation sequencing.

TO THE EDITOR

We read with great interest the high-quality article by Wang et al[1] recently published in the World Journal of Radiology. In retrospectively analyzing 102 patients with pulmonary nocardiosis, the authors further our collective understanding of the radiographic features, common co-infections, and associations of pulmonary nocardiosis, all while demonstrating the importance of metagenomic next-generation sequencing (mNGS) diagnostics in this disease. Furthermore, by systematically comparing imaging manifestations across different Nocardia species, this study found a species-specific and sensitive computed tomography (CT) chest presentation for Nocardia wallacei (N. wallacei), as bronchopneumonia, spurring the need for further investigation into this and other species-characteristic imagery presentations. In this letter, we aim to contextualize the authors’ findings within the evolving landscape of radiology and microbiological diagnosis of pulmonary nocardiosis, with emphasis on the implications of identifying species-specific imaging features, the limitations of conventional culture-based confirmation, and the potential impact of mNGS on both diagnostic practice and epidemiologic understanding of Nocardia infections.

Radiographic patterns in pulmonary nocardiosis

Previous literature described pulmonary nocardiosis as radiographically heterogeneous, commonly presenting with consolidation, nodules, cavitation, and pleural effusion, without reliable species-level distinctions with imaging alone[2,3]. However, Wang et al[1] in their recent work, have identified a unique CT presentation of a relatively rare species of Nocardia, N. wallacei, as bronchopneumonia without cavitation. Specifically, the authors found that of the seven patients in their sample with N. wallacei, six presented with bronchopneumonia and none with cavitation. To our knowledge, this finding represents the first such multi-patient report of species-level distinction of Nocardia species on CT alone. It is important to recognize that bronchopneumonia is also a rare but possible finding present in pulmonary nocardiosis caused by Nocardia species other than N. wallacei. Despite this, when using the authors’ data, the finding of bronchopneumonia is both sensitive (85.71%) and specific (83.34%) at discriminating N. wallacei from other species in patients with pulmonary nocardiosis within the sample of five species with CT data when considering cases caused by species other than N. wallacei as “negatives”. These findings are promising but are limited by the commonality of coinfections in pulmonary nocardiosis, previously reported at 21%, which can mask or complicate radiographic findings[4]. This challenge is heightened by the fact that pulmonary nocardiosis is more common in immunocompromised hosts, in which cytomegalovirus, Aspergillus spp., and Pneumocystis jirovecii, among others, are common co-infections[5,6]. Furthermore, several other causes of pulmonary disease besides Nocardia species present similarly to this proposed pattern. Pulmonary infections with Pseudomonas aeruginosa, Staphylococcus aureus, Haemophilus influenzae, Actinomyces israelii, Aspergillus spp., Mycobacterium tuberculosis, and the Mycobacterium avium complex pathogens are all known to possibly present with similar CT findings of bronchopneumonia[7,8]. Even when limited to Nocardia spp. caused pulmonary disease, 35 species are known to cause pulmonary nocardiosis, and no prior studies have characterized or compared the prevalence of different CT findings across all species[9]. These observations underscore the reality that CT alone is limited in absolute diagnostic specificity when used in isolation, as each patient will present at a varying stage of disease and with varying immune status, which influences the radiographic presentation[2,10,11]. The work by Wang et al[1], comparing the features of nine species thus represents a strong first step toward understanding the radiographic presentation of specific Nocardia species, with gaps that require additional confirmatory testing in the diagnosis of pulmonary nocardiosis and potential causative species.

Clinical implications of Nocardia species differentiation and the role of molecular diagnostics

The ability to differentiate species can influence treatment and prognosis prediction. For instance, the largest meta-analysis to date on nocardiosis found that infections with specific species can have significantly different mortality rates: 26.6% and 23.4% for Nocardia asteroides and Nocardia farcinica, respectively, compared to 1.6% for N. wallacei[12]. Another study also found that species influences drug resistance, with 2.3% of N. farcinica being resistant to the first-line treatment of TMP-SMX[13]. These findings highlight the importance of the authors’ work, as they suggest that radiology can play a role in identifying the causative species in pulmonary nocardiosis at the point of care before molecular-based diagnostic results are returned. Current molecular-based diagnostics, such as mNGS, have an average turnaround time of 48 hours, excluding delivery, underscoring the need for rapid imaging-based diagnosis, especially in remote hospitals and clinics[14]. Although slower than radiographic evaluation, molecular-based diagnostics still play a crucial role. Confirmatory testing is required for diagnosis in pulmonary nocardiosis, regardless of clinical and radiographic presentation, and traditional cultures are much slower. One report found that the turnaround time for traditional cultures was 7.5 ± 1.92 days, and another found that the speed of diagnosis greatly influenced patient outcomes[15,16]. Adding to this, one of the more surprising findings in the article by Wang et al[1] is that traditional cultures had very low sensitivity, detecting only 3.3% of those diagnosed using mNGS in this paper. Similar reports have previously found sensitivity of traditional cultures ranging from 36.84% to 66.67% compared to mNGS in pulmonary nocardiosis. Still, none, to our knowledge, have reported rates quite so low[4,15,17]. This finding by Wang et al[1] strongly supports the use of molecular diagnostics, such as mNGS, for confirming pulmonary nocardiosis rather than cultures. An added benefit of mNGS is that, by sequencing all RNA and DNA within a sample in a non-targeted manner, it is ideal for identifying the specific causative Nocardia species, as well as other species or pathogens present[4]. That level of detail, in combination with improved species-level discrimination via CT, as the work by Wang et al[1] has set forth, can provide a two-step diagnostic workflow for identifying the causative species in pulmonary nocardiosis. Furthermore, mNGS paired with CT imaging would address the limitations of CT for species identification posed by coinfections, which complicate radiographic findings.

Epidemiologic insights from molecular diagnostics

The latest research by Wang et al[1] demonstrates another use case for mNGS in providing insights into the epidemiology of pulmonary nocardiosis. By enabling highly specific and sensitive species identification, mNGS can characterize the relative prevalence of Nocardia species. The authors report Nocardia gelsenkin (N. gelsenkin) as the most common Nocardia species in their sample of pulmonary infections. However, upon searching the literature, the N. gelsenkin species has not previously been described. Elsewhere in the paper, they mention Nocardia gelsenkirchenensis (N. gelsenkirchenensis), which also does not belong to a valid species of Nocardia and can be presumed to refer to the same species (N. gelsenkin) discussed previously. Based on the available information, we believe the name N. gelsenkirchenensis may refer to Nocardia cyriacigeorgici (N. cyriacigeorgici), which was first isolated in the German town of Gelsenkirchen and described by Yassin et al[18]. After changing the names as described, Wang et al[1] would have found N. cyriacigeorgici (22%) to be the most common, followed by Nocardia abscessus (14%) and N. farcinica (12%), the three most common in China, where the study was conducted[13]. While similar epidemiological data have previously been collected in this manner, the authors build on this idea, using mNGS to characterize the prevalence of common coinfections in patients with pulmonary nocardiosis and bronchiectasis. The authors found Aspergillus to be the most common within their sample, which, in this context, has previously been described as a rare co-infection with Nocardia cyriacigeorgica[19]. The authors’ findings are in line with the shared risk factor between pulmonary nocardiosis and Aspergillus, namely bronchiectasis, and provide further evidence for an association previously reported only sparingly[19-21].

Conclusion

The most recent work done by Wang et al[1] meaningfully advances the diagnostic understanding of pulmonary nocardiosis by demonstrating the potential for species-specific CT patterns and common associations via mNGS. The authors’ findings suggest a potential complementary two-step diagnostic framework in which imaging, with fast turnaround, raises species-level suspicion, and molecular diagnostics, such as mNGS, provide later definitive confirmation. With further research into this area with larger sample sizes, these approaches may enable earlier, more accurate diagnosis, refine epidemiologic knowledge, and ultimately improve clinical outcomes in pulmonary nocardiosis.