Tuberculosis (TB) is a major cause of morbidity and one of the leading causes of death worldwide (1). TB/HIV coinfection and drug resistant TB pose challenges to the TB elimination efforts. Effective treatment of TB, particularly multidrug-resistant TB (MDR-TB) and extensively drug resistant TB (XDR-TB) is long, expensive and toxic. The long treatment duration [6–22 months (2)], and severe adverse effect of anti MDR-TB medications often lead to poor patient adherence to the treatment, putting not only these patients, but also the people in their close proximity, at risk (3, 4). Discontinuous treatment and partial treatment are two leading causes for drug resistance (5).

Coinfection of TB and HIV exacerbates the symptoms of both diseases in a synergetic manner, as both pathologies impair the immune system’s ability to defend against these pathogens (6). Coinfection with HIV significantly increases mortality among TB patients (7). For example, a study conducted in Eastern Europe reported a median survival time of 5.9 years following diagnosis with MDR-TB or XDR-TB, which decreases to 1.9 years for patients also diagnosed with HIV (8). Additional risk factors for death among TB patients include alcoholism, male gender and old age, among others (7, 9).

Mycobacterium tuberculosis (Mtb), the major causative agent of TB, can exist inside the human host either in an active, replicating state, or in a dormant state. Dormancy is a programed response which enables Mtb to survive inside macrophage phagosomes under conditions of starvation, hypoxia, low pH, and oxidative and nitrosative stress. During dormancy, bacteria cease replication, downregulate central metabolism, and switch to anaerobic metabolism (10). The induction of dormancy is regulated by the dormancy survival DosR regulon (11, 12). This dormant state allows latent infection that can persist for many years or even a lifetime without clinical symptoms. However, if a critical number of bacteria manage to evade the host immune system and resume an active, replicating state, the patient develops an active, symptomatic TB disease (13, 14).

The chronic nature of infection with Mtb allows for within-host evolution of the infecting bacteria over time. This process involves the accumulation of mutations, which are believed to contribute to the pathogen’s adaptation to the environmental conditions within the host. Indirect evidence supporting the selective advantage of these mutations is the observation that they often occur in the same genes across different patients, suggesting homoplastic evolution. A major mycobacterial function influenced by these mutations is drug resistance, frequently accompanied by compensatory mutations that mitigate the fitness cost of drug resistance mutations. Genes linked to virulence, lipid and cell wall metabolism and glycerol-3-phosphate metabolism have also been demonstrated to accumulate mutations during host infection (15, 16).

Compared to other bacteria, the genome of Mtb is highly conserved. Modern Mtb strains do not typically acquire new genomic features through horizontal gene transfer (17, 18). However, Mtb strains often lose fragments of their genomes. Baena et al. (19) found 4,111 genomic deletions longer than 1,000 bp in 522 Mtb isolates of L4 lineage. One of those deletions, that is 6,479 bp long and results in the loss of 10 genes, was observed in 249 genomes from four different countries. Other large deletions were also detected in multiple strains. Most of these strains were obtained from sputum, which is produced exclusively during active respiratory tuberculosis. Assuming each unique deletion originated from a distinct event, it is therefore evident that some large genomic deletions do not impede pathogeny and transmission.

Israel has maintained ongoing surveillance of all TB cases since 1997. Patients with suspected pulmonary TB in Israel are requested to provide sputum or other biological specimens for culture. All isolates are sent by primary laboratories to the National Mycobacterium Reference Laboratory (NMRL) in Tel Aviv, Israel. NMRL identifies the strain, performs drug susceptibility tests, and maintains a bank of strains as part of routine work. Since 2019, all new TB strains are fully sequenced (WGS) for epidemiological purposes and for drug susceptibility prediction. Between 2010 and 2019, NMRL received 35 TB samples from a patient diagnosed with both TB and HIV. These samples were found to belong to a single clone of MDR-TB characterized by a unique genomic deletion spanning 18 genes. In spite of the patient’s long progressive AIDS and active TB, incomplete treatment, and multiple risk factors, this patient survived for almost 9 years. Here we describe this exceptional case, and discuss the distinct features of the infecting pathogen that may have contributed to this survival.

In this study, we present a case of exceptionally long survival in a man over 60 years of age who concurrently had progressive AIDS, active tuberculosis caused by MDR-TB and alcoholism. This remarkable extended survival is particularly notable, given the patient’s failure to complete his anti-TB treatment regimens, and his refusal of anti-HIV treatment as reflected by his very low CD4 cells counts and high viremic state. A previous study on survival rates in patients with drug resistance TB found a median survival of 5.9 years in MDR-TB and XDR-TB patients and 1.9 years in patients co-infected with HIV (8). The advanced age, male gender, and alcoholism of our patient makes this case even more exceptional (8). The severe consequences and poor prognosis for patients co-infected with HIV/MDR-TB, particularly those that do not take ART is well documented in the literature (26–29). Furthermore, prior HIV infection, like in our case, significantly increase the chance of acquiring MDR-TB (1.42 times higher) (30).

Over the years, since his first TB diagnosis, Mr. X started treatment several times, but never completed it. He was treated with 13 different antibiotics in total. His treatment ultimately failed to cure his TB, presumably due to the combination of his co-morbidities, the pathogen’s multiple drug resistance and his low compliance (Figure 1; Supplementary Table S1). Initially the standard combination of RIF, INH, EMB, and PZA was administered, but when DST results from the NMRL indicated that the strain was resistant to RIF, INH, and PZA, these drugs were replaced with second line drugs. His treatment later included EMB again, despite most of his Mtb isolates exhibiting intermediate resistance to this anti-mycobacterial agent, and two isolate being fully resistant.

The isolation and genome sequencing of the Mtb isolates from this patient revealed an unusual genomic deletion, termed Del-X. This deletion spans from bases 2,274,648 to 2,290,386 in the H37Rv reference genome (RefSeq id NC_000962.3), causing the loss of 18 genes (Rv2028c to Rv2045c in the standard Mtb loci numbering). What effect could Del-X have on the bacterium’s physiology, and the nature of the disease? Many of the genes inside Del-X are involved in dormancy (Figure 2; Supplementary Table S3). The hspX gene (Rv2031c, also known as acr) encodes alpha-Crystallin-like heat shock protein (Hsp16.3), which plays a pivotal role in Mtb dormancy response (23, 31, 32). DosR, a transcription factor that activates the transcription of Mtb dormancy regulon, was shown to induce the expression of hspX and acg (33). In turn, Rv2028 and Rv2034 positively regulate the expression of the dosR gene (32, 34). Many of the genes within Del-X are induced by the conditions typical of macrophages phagosomes, such as hypoxia, starvation, low levels of nitric oxide, and acidic pH (32, 35–41). hspX actively contributes to slowing the growth of Mtb immediately following infection and is essential for its resistance to low oxygen stress (31, 32). Therefore, it is reasonable to speculate that the strain infecting Mr. X was unable to enter a dormant state. Studies have shown that a knockout strain of Mtb lacking hspX exhibits reduced growth in macrophages (23), and that depletion of HspX protein deteriorates Mtb tolerance to anaerobic conditions (32). Rv2037c too was shown to enhance in vivo and ex vivo intracellular survival (39). Thus, we suggest that this genomic deletion not only prevented the pathogen from entering a dormant state, but might have also abolished its ability to survive inside macrophages’ phagosomes, imposing an escape from the phagosomes into cells’ cytosol (42).

Several of the genes in Del-X region play an important role in the host immune response to Mtb. hspX, Rv2029c, Rv2034, Rv2041c, and possibly Rv2028c encode strong immunogenic antigens (43, 44) (Figure 2; Supplementary Table S3). HspX in particular plays a major role in activation of dendritic cells and induces the production of TNF-α in lymphocytes (44, 45). acg is important for development of the typical tuberculosis granulomas (46). In a study on MDR-TB patients, Rv2037c was shown to induce the production of antibodies and of pro-inflammatory cytokines (39). Rv2041c increases the production of cytokines, which are important for macrophage maturation (M1 polarization) (47). Therefore, Del-X is predicted to weaken the immune response against the pathogen, compared to the response against wild type Mtb. The patient’s immune system was severely compromised by his HIV infection, as evidenced by the low CD4 counts (about 80 per μL), low CD4/CD8 ratio (10%), and high viral load (near 350,000 and 900,000 per mL) measured in 2017 and in 2018, respectively. The combined effect of impaired immune system and lack of Mtb determinants important for immunogenic recognition and activation, likely diminished the host ability to neutralize the infection, resulting in a constantly active TB.

On the other hand, Del-X may have severely affected the pathogen’s proliferation inside the host. Two of the deleted genes, Rv2028 and Rv2040c, have been previously identified in a comprehensive screen for genes that are important for in-host survival (48). acg has also been shown to be an essential virulence factor, as its deletion attenuated the bacterial growth in acute and persistent mice infections (46). The Rv2038c-Rv2041c operon encodes a sugar importer of the ABC transporter family, which is thought to play an important role in virulence or in-host survival (49), and its absence may limit the bacterial energy source. Furthermore, Del-X may impair lipid metabolism, due to the deletion of Rv2034 and Rv2037c (34, 39), and nucleotide metabolism, due to the deletion of Rv2030c (32). These deficiencies could further attenuate the growth of this strain. We hypothesize that Del-X genomic deletion resulted in an attenuated Mtb strain, unable to enter dormancy, yet with poor proliferation rate. The reduced virulence of this strain may have been a major factor contributing to the long survival of the patient. However, in contrast to this hypothesis, it is worth noting that some evidences have shown that deletion of hspX increases bacterial growth in vivo (31). Notably, the Mtb mutant used in the paper of Hu et al. (31), was deleted only in hspX gene, on the other hand, the del-X mutants in our study, contain an unusually large genomic deletion that covers 18 genes involved in multiple cellular functions, as described. Therefore, these strains cannot be compared. An obvious effect of Del-X was PZA resistance, due to the deletion of pncA. Treatment with PZA likely selected for the strain with this deletion, allowing it to take over the population, despite its defects. Indeed, although deletions in this genomic region are rare, almost all known Mtb genomes with a deletion of 5 kb or more in this region also exhibit a deletion of pncA (Figure 4; Supplementary File S5). In line with this, the work of Godfroid et al. (50), focusing on analysis of multidrug resistant MTB outbreak isolates, discovered that genomic insertions and deletions are significantly enriched in genes conferring antibiotic resistance.

The discovery of Del-X underscores the importance of detecting large genomic deletions in routine analysis of Mtb whole genome sequencing. Standard pipelines based on mapping to reference genomes and variant call may miss such deletions. Indeed, following this case, our laboratory has incorporated a step of identifying deletions based on uncovered genomic regions to its routine analysis.

We found a few Mtb genomes with genomic deletions in Del-X, from different parts of the world, mainly belonging to the Beijing or LAM lineages. As mentioned, the isolates from Mr. X also belong to the Beijing lineage that is known to be hyper-mutagenic, virulent, that rapidly acquires mutations for resistance to anti-tuberculosis drugs (51–54).

Interestingly, two of these genomes were isolated from a parent and their child, and are genetically clonal. Hence, one of these family members probably infected the other (25). This finding suggests that the Del-X genomic deletion does not prevent transmission from one patient to another and may have already been present in the strain when Mr. X was infected. However, none of this strain’s known closely related isolates had a deletion in the Del-X region. Thus, we cannot exclude the possibility that the deletion arose within this patient (Figure 3; Supplementary Table S4).

The isolates obtained from Mr. X at different time points during his nearly 9 years of infection belong to the same clone, yet there are several variations differing between them. These variations may originate from co-existing sub populations randomly sampled, or reflect changes that accumulated over time during the infection. Our phylogenetic analysis also attempts to separate these cases and include only mutations likely to have emerged during the infection. The later possibility is generally the case in Mtb infections (17). Many of the genes mutated during the infection were previously identified as “hot spots” for in vivo evolution during TB infection (Figure 5) (15, 16). For example, the gyrA A90V mutation discussed above, which explains the emergence of FQ resistance in isolates obtained between August 2016 and November 2017, and was probably selected for by the treatment with MFX and LFX (55). Furthermore, the fact that the patient did not receive continuous treatment is known to be a major risk factor for the appearance of resistance (5). After the exposure to these drugs was removed, an isolate taken in February 2019 lacked this mutation, probably representing a sub-population that evaded FQs and became dominant once they were no longer used. Another interesting variation is the insertion of C in the homopolymeric tract of 7 cytosine in the glpK gene, relative to the TB reference H37Rv genome. This mutation is common to all isolates except for the one from August 2016 and the last isolate from February 2019. The glpK gene encodes a glycerol kinase involved in glycerol metabolism. This mutation is known to serve as a mechanism for drug tolerance and growth slow-down, which occurs and reverses at high frequency (56). Notably, among the 26 genes mutated in-host there were six PE-PGRS genes, the functions of which are largely unknown (Figure 5).

The medical records of Mr. X indicate that he started TB treatment at least six different times, probably reflecting worsening of the symptoms (Figure 1). In all of these episodes, the patient stopped the treatment before completing it, at times by simply disappearing. This demonstrates the challenges faced by the health authorities in treating TB, which often involve social and personal difficulties. The case of Mr. X underscores the importance of patient adherence to appropriate medical treatment and highlights the significant challenges in achieving this goal due to personal and social issues. It is probable that with a combined therapy, including ART and suitable anti-TB drugs, this patient could have successfully overcome the infection with the del-X mutant. Treatment of low compliant and unmotivated patients is often unsuccessful and requires a tailored plan based on establishing a long-term trust relationship. Directly observed therapy (DOT) is an important part of such a plan (57). Mr. X’s case raises complex ethical, legal, and personal dilemmas. Israeli law grants district doctors the authority to forcibly hospitalize individuals suffering from diseases that jeopardize public health, provided they have obtained judicial approval. The issue of forced medical treatment is even more intricate, and, on rare occasions, the law allows for the treatment of such patients. However, it is important to note that the successful treatment of drug-resistant tuberculosis hinges on the patient’s cooperation, which, unfortunately, was unattainable in this particular case.

Taken together, we hypothesize that the unusual prolonged TB illness of Mr. X was a result of a unique intricate interplay between the host and the pathogen. The potentially low immunogenic signature of the pathogen together with the severe immune deficiency of the host prevented the elimination of the pathogen. The Del-X genomic deletion prevented latency, resulting in constant active TB. At the same time, Del-X probably resulted in an attenuated pathogen, allowing the long-term survival of the patient.

The importance of this report stems from the unusual long survival of Mr. X, considering the circumstances, as well as the uniqueness of the genomic deletion described here. This case raises important questions about the influence of Mtb genetic factors on disease progression and the potential implications for prognosis and treatment. It is interesting to find out whether in additional cases of patients who survive active tuberculosis for many years without treatment, mutated Mtb with genomic mutations in important genes are involved. Further investigations may elucidate the precise mechanisms by which the Del-X deletion affects the pathogenicity and virulence of Mtb, as well as the host immune response to the infection. In addition, this special isolate, can be used as a model bacterium for a better understanding of host-pathogen fundamental interactions, such as Mtb escape from the phagosome to the cytosol, dormancy and more.

The raw genomic sequences from this work have been deposited to NCBI and are available in BioProject PRJNA834625.

This study was approved by the National Helsinki Committee for Medical Experiments on Humans, of the Israeli Ministry of Health (# MOH-081-2021). The studies were conducted in accordance with the local legislation and institutional requirements. The human samples used in this study were acquired from a by-product of routine care or industry. Written informed consent for participation was not required from the participant or the participant’s legal guardians/next of kin in accordance with the national legislation and institutional requirements. Written informed consent was not obtained from the individual for the publication of any potentially identifiable images or data included in this article because the patient this article is about died several years before publication. Therefore, a written informed consent was obtained from the participant’s next of kin, for the publication of this case report, and any potentially identifiable images or data included in this article.

MR: Conceptualization, Data curation, Formal analysis, Writing – original draft, Writing – review & editing. AM: Data curation, Formal analysis, Visualization, Writing – original draft, Writing – review & editing. YL: Formal analysis, Writing – original draft, Writing – review & editing. GV: Data curation, Formal analysis, Writing – review & editing. YG: Data curation, Formal analysis, Writing – review & editing. HM: Writing – review & editing. EF: Formal analysis, Writing – review & editing. IK: Formal analysis, Writing – review & editing. OM: Formal analysis, Writing – review & editing. DZ: Formal analysis, Writing – review & editing. MS: Investigation, Writing – review & editing. LT: Investigation, Writing – review & editing. HK: Writing – review & editing. ZD: Writing – review & editing. ER: Project administration, Supervision, Writing – review & editing. IN: Conceptualization, Data curation, Formal analysis, Methodology, Writing – original draft, Writing – review & editing.