If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Concurrent pulmonary tuberculosis (TB) is common in patients with peripheral lymph node TB.
•
Symptoms, radiographic findings, and HIV are risk factors for concurrent disease.
•
The transmission risk of peripheral lymph node–associated pulmonary TB is low.
•
Multiple pathogenetic mechanisms explain the phenomenon of concurrent disease.
Abstract
Objectives
Relatively little is known about the prevalence, risk factors, and public health consequences of peripheral lymph node (PLN)–associated pulmonary tuberculosis (PTB).
Methods
We developed a 10-year (2010-2019) population-based cohort of PLNTB patients in Canada. We used systematically collected primary source data and expert reader chest radiograph interpretations in a multivariable logistic regression to determine associations between sputum culture positivity and demographic, clinical, and radiographic features. Public health risks were estimated among contacts of PLNTB patients.
Results
There were 306 patients with PLNTB, among whom 283 (92.5%) were 15-64 years of age, 159 (52.0%) were female, and 293 (95.8%) were foreign-born. Respiratory symptoms were present in 21.6%, and abnormal chest radiograph in 23.2%. Sputum culture positivity ranged from 12.9% in patients with no symptoms and normal lung parenchyma to 66.7% in patients with both. Respiratory symptoms, abnormal lung parenchyma, and HIV-coinfection (borderline) were independent predictors of sputum culture positivity (odds ratio [OR] 2.24 [95% confidence interval [CI] 1.15-4.39], P = 0.01, OR 4.78 [95% CI 2.41-9.48], P < 0.001, and OR 2.54 [95% CI 0.99-6.52], P = 0.05), respectively. Among contacts of sputum culture-positive PLNTB patients, one secondary case and 16 new infections were identified.
Conclusion
Isochronous PTB is common in PLNTB patients. Routine screening of PLNTB patients for PTB is strongly recommended.
In Canada, peripheral lymph node tuberculosis (PLNTB) is defined by the International Classification of Diseases (ICD) as clinical TB disease involving extra-thoracic and extra-abdominal lymph nodes. Its emergence as the most common site of extrapulmonary TB during the second half of the twentieth century was observed with a shift in country of birth of new immigrants to Canada from low TB incidence countries of Western Europe to high TB incidence countries of Asia and Africa [
]. PLNTB patients are not usually considered a public health threat; their incident disease only signals the likelihood of past exposure to an infectious pulmonary TB (PTB) patient. While contemporary series of extrapulmonary TB report minor proportions of patients with concurrent PTB [
], studies often had small sample sizes, were not limited to PLNTB, and were not systematic in screening for PTB. Primary infection with Mycobacterium tuberculosis (MTB) is usually via a lung portal, and lymphohematogenous spread is a natural sequela. Lymph node involvement is important to the adaptive immune response [
PLNTB-associated PTB may be subclinical, i.e., the patient denies respiratory symptoms, and their chest radiographs may evince no lung parenchymal abnormality [
]. A better understanding of PLN-associated PTB could lead to earlier diagnosis; the concomitant disease has public health implications and raises legitimate questions about pathogenesis. Herein, we examine in detail a 10-year cohort of PLNTB patients with three primary objectives, (i) determining the prevalence of PTB in PLNTB patients overall and by lymph node site, (ii) describing the chest radiographic features of PLNTB patients, with and without respiratory symptoms, and (iii) determining predictors of PTB in PLNTB patients. A final objective was to determine the public health consequences of PLN-associated PTB. The study was performed in the province of Alberta, Canada, which, during the mid-point of the study (2016), had a population of 4,252,900 and is one of four major immigrant-receiving provinces in Canada. These four provinces combined report >75% of all TB patients in Canada (British Columbia [14.1%], Alberta [12.6%], Ontario [37.6%], Quebec [12.1%]), and >80% of the patients in each are foreign-born [
]. The TB program in Alberta is uniquely centralized, with all diagnoses verified and patients treated by a small group of university-based physicians operating out of three dedicated public health TB clinics [
We used a retrospective cohort study design with data from the Canadian Tuberculosis Reporting System and ICD-9 codes (Public Health Agency of Canada). All patients diagnosed with TB in Alberta between 2010 and 2019 in the provincial registry and TB laboratory were identified and then grouped according to whether they did or did not have PLNTB using ICD-9 code 017.2. Patients with PLNTB were defined as those who were culture-positive for MTB complex species from their lymph nodes, those who had granulomatous inflammation in their lymph nodes with or without a positive culture elsewhere, or those who had a clinical diagnosis of PLNTB and were culture-positive for MTB complex species from elsewhere.
Clinical, demographic, and radiographic data were extracted from the hospital, public health, and laboratory records. Patients with PLNTB were compared to those without PLNTB by age, sex, population group (Canadian-born or foreign-born), disease type, culture status, anti-TB drug resistance (if culture-positive), and HIV status. The records for PLNTB patients were further interrogated to determine (i) whether a chest radiograph had been performed and (ii) whether airway secretions (spontaneously expectorated sputum, induced sputum, or bronchoscopy washing/bronchoalveolar lavage [BAL] specimens) had been submitted for acid-fast bacilli (AFB) smear and culture. Those patients without a chest radiograph performed and/or at least one induced sputum or bronchoscopy washing/BAL or a minimum of two spontaneously expectorated sputum samples were excluded from further analysis. The records of remaining patients were reviewed to determine whether respiratory symptoms (cough, hemoptysis, chest pain, or shortness of breath) were present at diagnosis.
Chest radiographs of PLNTB patients with or without respiratory symptoms were reread by two experienced chest radiologists. Discrepant interpretations were resolved by a third experienced chest radiologist. A data abstraction form was used to report and categorize the chest radiographic features following expert reader interpretation and for selected features following field reader interpretation at the time of diagnosis (see below and supplement 4) [
]. Computed tomographic (CT) scans of the thorax, if performed, were also reread by the same expert readers for the presence or absence of enlarged intrathoracic lymph nodes, and as below, in selected cases, for the presence or absence of lung parenchymal disease.
Chest radiographic features
(i)
Category or Pattern: radiographic findings were categorized as ‘typical’ for adult-type PTB (upper lung zone predominant disease, with or without cavitation, but no discernable adenopathy), ‘atypical’ for adult-type PTB (abnormalities inconsistent with the definition of typical), or normal. Lung zones were determined by visualizing a perpendicular line from the apex of the lung to the hemidiaphragm and dividing the lung in half; the upper lung zone included the superior segment of the lower lobe.
(ii)
Laterality: bilateral, unilateral, or normal.
(iii)
Cavitation: cavitary—with cavities defined as a gas-filled space within pulmonary consolidation, a mass, or a nodule—or non-cavitary disease.
(iv)
Endobronchial spread: acinar shadows defined as ill-defined nodules 4-8 mm in diameter).
Extent of parenchymal disease: normal, minimal, moderately advanced, far-advanced, or miliary according to criteria established by the US National Tuberculosis and Respiratory Disease Association (see supplement 5) [
Incidental note was also made of the presence of volume loss, pleural thickening/retraction, or effusion. If enlarged nodes alone were identified—i.e., the lung parenchyma appeared normal—then the category was reported as atypical, and the laterality and extent of disease were reported as normal. When the only abnormality was a small, calcified nodule, the radiograph was reported as normal. Normal radiographs included both posteroanterior and lateral views with few exceptions.
CT features
(i)
Adenopathy: hilar or mediastinal lymph node enlargement defined as a node >10 mm in short axis diameter.
Chest radiographic features were compared in PLNTB patients with and without respiratory symptoms. Demographic, clinical, and radiographic risk factors for sputum culture positivity were determined.
Public health consequences of PLN-associated PTB
The outcome of household contact tracing was compared in PLNTB patients with and without positive sputum cultures. In addition, given that, in Alberta, all initial isolates of MTB are routinely genotyped using restriction fragment-length polymorphism (RFLP) supplemented by spoligotyping in isolates with five or fewer copies of insertion sequence IS6110 (1990-2016) or, more recently, 24-loci mycobacterial interspersed repetitive units (MIRU) (January 2014 onwards). RFLP and MIRU data were used to obtain a crude estimate of the number of PLNTB patients who might (clustered isolate) or might not (unique isolate) have transmitted within the jurisdiction [
Proposal for standardization of optimized mycobacterial interspersed repetitive unit-variable-number tandem repeat typing of Mycobacterium tuberculosis.
]. As has been previously described, the presence of a matching DNA fingerprint (clustered isolate) was interpreted to mean possible transmission from the PLNTB patient [
]. If the TB patient was foreign-born, then matches that preceded their date of arrival were excluded. Molecular data were also used to (i) rule out laboratory cross-contamination [
Molecular strain typing of Mycobacterium tuberculosis to confirm cross-contamination in the mycobacteriology laboratory and modification of procedures to minimize occurrence of false-positive cultures.
] and (ii) screen for M. orygis—a zoonotic or zooanthroponotic member of the MTB complex—reported from extrapulmonary TB patients primarily in the Indian subcontinent [
Descriptive comparisons were performed using chi-square or Fisher's exact test for categorical variables and Student's t-test for continuous variables. Agreement between radiography readers was estimated using generalized kappa statistics. Univariable analyses calculating odds ratios (ORs) and 95% confidence intervals (95% CIs) for the association between prediction factors and sputum culture positivity were performed. The same variables were included in a multivariable regression model to determine independent predictors of sputum culture positivity. An interaction term between radiographic findings of lung parenchymal abnormality and enlarged intrathoracic nodes on the dependent variable in the model was also explored. The model was separately run using both expert reader and field reader interpretations of radiographs. All data were analyzed using Stata Statistical Software (StataCorp) version 14.0. Institutional approval was obtained from the Health Research Ethics Board at the University of Alberta, protocol ID: Pro00116880.
Results
All PLNTB patients
In the decade 2010 through 2019, there were 2043 TB notifications in the province of Alberta of whom 400 (19.6%) had PLNTB and 333 (16.3%) were classified as PLNTB according to Canada's hierarchical reporting system, which ranks respiratory, disseminated, and central nervous system disease above PLNTB (Figure 1). Cultures were positive from PLNs in 324/400 (81.0%) patients; the median (interquartile range [IQR]) time-to-culture positivity was 15 (11-20) days. M. bovis was recovered from four patients (1 each from India, Somalia, Ethiopia, and Morocco); no patient was culture-positive for M. orygis. Granulomatous inflammation was reported as “caseating” or “necrotizing” in 62/72 (86.1%) lymph node culture-negative patients; four PLNTB patients had a clinical diagnosis of PLN disease. Non-lymph node culture-positive sites included: 63 pulmonary (11 patients were PLN culture-negative and sputum culture-positive), five pleural (one patient was PLN culture-negative and pleura culture-positive), two blood, three bone and joint, two genitourinary, and two peritoneum. The leading countries of birth for foreign-born persons with PLNTB (n = 381) were the Philippines (37%), India (13.1%), Ethiopia (n = 11.5%), Somalia (n = 7%), and Vietnam (n = 3.4%). Among those with all other forms of TB (n = 1357), the leading countries of birth included the Philippines (30.3%), India (18.7%), China (6.3%), Ethiopia (5.0%), and Vietnam (4.8%). Compared to patients with other forms of TB, those with PLNTB disease were more likely to be younger, female, and foreign-born and less likely to have relapse/retreatment disease (Table 1). HIV co-infection was similarly uncommon in those with and without PLNTB. Of the total of 400 PLNTB patients, 306 (275 cervical, 31 axillary/inguinal), or 76.5%, had both a chest radiograph and sputum cultures. These patients comprise the study group going forward. The 94 patients who were excluded differed from the 306 patients who were included only by the proportion that were lymph node culture-positive (see supplement 1).
Figure 1Over the 10 study years (2010-2019) there were 400 patients with peripheral lymph node tuberculosis (ICD-9 code 017.2) diagnosed in the Province of Alberta. Among the 400 PLNTB patients there were 67 who also had a hierarchically more important site of disease, i.e., in addition to ICD-9 code 017.2 they had another hierarchically more important ICD-9 code, and it was with this latter code that was used to classify their disease in Canada's national TB Reporting System. The diagnostic criteria and culture status of each PLNTB patient is reported. One patient, whose PLNTB diagnosis was based on clinical criteria, was culture-negative from any site; they were included in the group of 400 (but not in the main study group of 306) based on imaging, a fine needle aspirate that was negative for an alternative diagnosis and a positive therapeutic response. Among the 400 PLNTB patients, 94 were excluded from the main analysis, not having had sputum submitted or a radiograph performed (90 cervical; four axillary/inguinal). Among the 306 patients included in the main analysis, 275 had cervical lymph node TB and 31 had axillary/inguinal lymph node TB. “Cervical” included all lymph node stations in the head and neck.
324 (81.0%) patients were culture-positive from a peripheral lymph node, with or without a positive culture from another site; 12 patients were culture-negative at their peripheral lymph node site but culture-positive elsewhere
First-line drug susceptibility testing was not possible on one lymph node isolate. Four PLNTB patients had multidrug-resistant TB. Ten PLNTB patients were culture-positive from both a lymph node and a non-lymph node site and were drug-resistant; in each case the drug susceptibility test results on the lymph node and other site isolates were the same.
a 324 (81.0%) patients were culture-positive from a peripheral lymph node, with or without a positive culture from another site; 12 patients were culture-negative at their peripheral lymph node site but culture-positive elsewhere
b First-line drug susceptibility testing was not possible on one lymph node isolate. Four PLNTB patients had multidrug-resistant TB. Ten PLNTB patients were culture-positive from both a lymph node and a non-lymph node site and were drug-resistant; in each case the drug susceptibility test results on the lymph node and other site isolates were the same.
Among the 306 PLNTB patients in the main study group, chest radiographs were performed, and sputum was collected at a median (IQR) of 6 days (2-18) and 3 days (1-10), respectively, from the start date of treatment. A single induced sputum or bronchoscopy wash/BAL had been collected in 34 patients; multiple spontaneous or induced sputum specimens (median [IQR], 3 [3-3]) had been collected in 272 patients. Sputum was culture-positive in 63 (20.5%) PLNTB patients (see below). The mean (± SD) and median (IQR) time-to-culture positivity of sputum were 20.6 ± 8.3 and 19.0 (15.0 to 23.3) days, respectively. At the time of diagnosis, field readers had reported chest radiographs as normal in 210 (68.6%), and clinicians had reported respiratory symptoms as absent in 240 (78.4%) PLNTB patients.
Kappa statistics showed perfect expert reader agreement for cavitation and substantial agreement (0.61-0.80) for all other features, see supplement 2. Expert reader chest radiographic interpretations in PLNTB patients with and without respiratory symptoms are reported in Table 2. Radiographs were reported as normal in 235 (76.8%) patients. Cavitation and acinar shadows were extremely rare (0.7% and 0.3%, respectively), and in the total of 71 PLNTB patients among whom a lung parenchymal abnormality was present, it was moderately advanced, far-advanced, or miliary in only 10 (14.1%). Patients with respiratory symptoms were more likely to have an atypical pattern, which, in some, was due to the presence of enlarged intrathoracic lymph nodes. Pleural thickening/retraction/effusion and/or volume loss were uncommon and reported only twice (volume loss) in the absence of lung parenchymal disease.
Table 2Chest radiographic features, according to respiratory symptom status, in peripheral lymph node tuberculosis patients whose sputum culture status was assessed (n = 306).
When reporting on “laterality” and “extent of parenchymal disease”, normal referred to the state of the lung parenchyma; 19 patients with normal lung parenchyma had enlarged intrathoracic lymph node.
When reporting on “laterality” and “extent of parenchymal disease”, normal referred to the state of the lung parenchyma; 19 patients with normal lung parenchyma had enlarged intrathoracic lymph node.
235 (76.8)
48 (72.7)
187 (77.9)
Minimal
61 (19.9)
13 (19.7)
48 (20.0)
Mod-advanced
7 (2.3)
4 (6.1)
3 (1.3)
Far-advanced
0 (0.0)
0 (0.0)
0 (0.0)
Miliary
3 (1.0)
1 (1.5)
2 (0.8)
a See text for definition of features; the field, rather than the expert reader interpretation, was used in one patient whose radiograph had been purged
b When reporting on “laterality” and “extent of parenchymal disease”, normal referred to the state of the lung parenchyma; 19 patients with normal lung parenchyma had enlarged intrathoracic lymph node.
Among the 306 PLNTB patients from whom sputum was collected and a chest radiograph performed, 83 (27.1%) had also undergone a CT thorax, 79 (95.2%) of which were enhanced. Most CT scans were performed during the workup of a patient and before a definitive diagnosis of TB. The kappa statistics for inter-reader variability for the presence of enlarged intrathoracic nodes on CT thorax was 0.829 (see supplement 2). Enlarged intrathoracic nodes were present in 51/83 (61.4%) patients; those with enlarged intrathoracic nodes were more likely than those without to be sputum culture-positive, 19/51 (37.3%) vs 5/32 (15.6%), P = 0.03. The chest radiograph was not sufficiently sensitive for detecting enlarged intrathoracic nodes; only 18/51 (35.3%) patients with enlarged nodes on CT thorax had enlarged nodes on chest radiograph. A CT scan of the thorax (1 positron emission tomography with concurrent computerized tomography [PET/CT]) had been performed in eight patients with normal lung parenchyma on chest radiograph, but yet who were sputum culture-positive; in four of the eight, the lungs appeared normal on CT scan; see Table 3 and Figure 2.
Table 3Lung parenchyma on computed tomography thorax in sputum culture-positive peripheral lymph node tuberculosis patients whose lung parenchyma appeared normal on chest radiograph.
Figure 2A posterior-anterior chest radiographic image (a) and an axial CT image (b), performed 3 days before and 18 days after, respectively, the start date of treatment in a 21-year-old female emigrant from the Congo, who grew Mycobacterium tuberculosis from a cervical lymph node and from sputum. The chest radiographic image demonstrates two non-calcified nodules in the left upper lung zone; the CT image demonstrates multiple non-calcified nodules in the apical-posterior segment of the left upper lobe, some in a tree-in-bud formation. Intrathoracic nodes were not enlarged. An AP chest radiographic image (c) and a selected AP projection from an FDG PET/CT study (d), performed 1 day and 6 days after, respectively, the start date of treatment in an 89-year-old male emigrant from Tanzania who grew M. tuberculosis from a cervical lymph node and from sputum. The chest radiographic image is unremarkable other than demonstrating coarsened intrapulmonary markings and a small, calcified nodule in the left mid-lung zone; the FDG PET/CT image demonstrates minimal bilateral hilar node FDG uptake, normal physiologic myocardial uptake, abnormal uptake in left supraclavicular neck nodes and no lung parenchymal uptake. Intrathoracic nodes were not enlarged.
The proportions of PLNTB patients whose airway secretions were culture-positive are depicted in Table 4 according to the site of PLN disease, the presence or absence of respiratory symptoms, and the presence or absence of abnormal lung parenchyma on chest radiograph. Among cervical lymph node TB patients, 60/275 (21.8%) were sputum culture-positive; 21/62 (33.9%) if they had respiratory symptoms, 39/213 (18.3%) if they had no respiratory symptoms. Among cervical lymph node TB patients who were symptomatic, 12/18 (66.7%) were culture-positive if their lung parenchyma was abnormal on chest radiograph; 9/44 (20.5%) if their lung parenchyma was normal on chest radiograph. Among cervical lymph node TB patients with no reported respiratory symptoms, 17/42 (40.5%) of those with and 22/171 (12.9%) of those without a lung parenchymal abnormality on chest radiograph were sputum culture-positive. Among patients with axillary/inguinal lymph node TB, only 3/31 (9.7%) were sputum culture-positive; 3/26 (11.5%) of those with axillary disease and none with inguinal disease. Only 4/63 (6.3%) PLNTB patients who were sputum culture-positive were smear-positive; only 2/63 (3.2%) were on the contact list of a recently diagnosed patient with pulmonary TB. Most (41/63 or 65.1%) sputum culture-positive PLNTB patients had subclinical pulmonary disease.
Table 4Sputum culture positivity in PLN TB patients according to PLN site, presence or absence of respiratory symptoms, and presence or absence of a lung parenchymal abnormality on chest radiograph.
Factors predicting sputum culture positivity in PLNTB patients
In univariable regression analysis, and using the expert reader radiograph interpretations, the presence of respiratory symptoms, abnormal lung parenchymal on chest radiograph, enlarged intrathoracic nodes on chest radiograph, and a positive HIV test were all associated with sputum culture positivity. In multivariable regression analysis, respiratory symptoms and the presence of a lung parenchymal abnormality on chest radiograph were independent predictors of sputum culture positivity; the model was underpowered for HIV positivity and while being statistically borderline it is a practically relevant consideration. There was no impact of interaction between parenchymal abnormality and enlarged intrathoracic nodes on sputum culture positivity in the multivariable model (see Table 5). The results were unchanged when the field reader radiograph interpretations were used (see supplement 3).
Table 5Predictors of sputum culture positivity in patients with PLNTB (using expert reader radiograph interpretations).
There was no impact of interaction between abnormal lung parenchyma and enlarged intrathoracic nodes on chest radiograph on sputum culture positivity in the multivariable model, OR 1.02 (0.018 – 5.74), P = 0.981
There was no impact of interaction between abnormal lung parenchyma and enlarged intrathoracic nodes on chest radiograph on sputum culture positivity in the multivariable model, OR 1.02 (0.018 – 5.74), P = 0.981
a There was no impact of interaction between abnormal lung parenchyma and enlarged intrathoracic nodes on chest radiograph on sputum culture positivity in the multivariable model, OR 1.02 (0.018 – 5.74), P = 0.981
Household contacts of sputum culture-positive PLNTB patients (when identified and assessed) were slightly more likely to have a new positive tuberculin skin test (TST)/interferon-gamma release assay (IGRA) or a TST/IGRA conversion than household contacts of sputum culture-negative PLNTB patients (when identified and assessed) though the difference was not statistically significant, see Table 6. Only one household contact of a symptomatic, sputum culture-positive PLNTB patient with abnormal lung parenchyma on chest radiograph was identified as a secondary case (matching MIRUs). Assuming the new positive TST/IGRA and TST/IGRA converter rate in household contacts of sputum culture-negative PLNTB patients to be the background rate, then an excess of 16 new positive TST/IGRA or TST/IGRA converters and one secondary case was estimated to have occurred from sputum culture-positive patients. Among sputum culture-positive and sputum culture-negative PLNTB patients whose MTB isolates were DNA fingerprinted, the proportion clustered with another isolate was similar, 16/60 (27%) vs 41/185 (22%), P = 0.47; data not shown.
Table 6Household contact tracing in peripheral lymph node tuberculosis patients with and without positive sputum cultures.
PLNTB was common in this 10-year population-based cohort of notified TB patients in Canada; 16.3% of all notified patients were classified as PLNTB, and another 3.3% had PLNTB in addition to a hierarchically more important site of disease. Cervical lymph nodes were most involved (91.3%). Patients with PLNTB were more likely than those without PLNTB to be younger, female, and foreign-born. They were less likely to have relapse/retreatment disease. Leading countries of birth of foreign-born PLNTB patients were the Philippines, India, Ethiopia, Somalia, and Vietnam. Respiratory symptoms were absent, and the chest radiograph was normal in most PLNTB patients. Moderately advanced or greater lung parenchymal disease on chest radiograph was rare. Yet, sputum cultures were not uncommonly positive, especially in patients with cervical lymph node disease. Independent predictors of a positive sputum culture were respiratory symptoms, abnormal lung parenchyma on chest radiograph, and HIV seropositivity (assessment of the latter was underpowered – not many patients were HIV positive; statistical significance was borderline but clinical significance is considered high). Transmission events attributable to sputum culture-positive PLNTB patients were few; an estimated 16 new positive TST/IGRAs or TST/IGRA conversions and one secondary case.
Allowing for differences in context, study design, and definitions our findings are largely in line with the existing peer-reviewed literature. In a review by Fontanilla et al in 2011[5], pulmonary involvement (in some cases inferred from an abnormal chest radiograph, but not proven by culture) was present in 0-28% of PLNTB patients from non-TB-endemic countries and 0-42% of PLNTB patients from TB-endemic countries. In Santa Clara County, California, Polesky et al[8] reported a group of 106 PLNTB patients, 19 (18%) of whom reported cough and 40 (38%) of whom reported an abnormal chest radiograph. Of those with an abnormal chest radiograph sputum was collected in 34 and were positive on culture in 14 (41%). Among patients with a normal chest radiograph only one was sputum culture-positive but the number whose sputum culture status was assessed was not reported [
]. In King County, Washington, Parimon et al[7] reported positive sputum cultures in 2/22 (9.1%) PLNTB patients, 2/18 (11.1%) patients with a normal chest radiograph, and 0/4 (0.0%) patients with an abnormal chest radiograph. In Riyadh, Saudi Arabia, El-Hazmi et al[4] reported an abnormal chest radiograph in 19/76 (25.0%) PLNTB patients. Sputum was collected in 18 patients and cultures were positive in 3/10 (30%) patients with a normal, and 1/8 (12.5%) patients with an abnormal, chest radiograph.
Two studies are reported from TB-endemic countries. In Malawi, sputum was systematically collected in 272 PLNTB patients and found to be smear-positive in nine (3.3%) [
]. In a prospective study of extrapulmonary TB patients in India, half of whom had PLNTB, Shivakumar et al[9] reported an abnormal chest radiograph consistent with TB in 28/90 (34%) patients with cough and 33/185 (20%) patients without cough; positive sputum cultures were present in 15/90 (18%) patients with cough and 21/185 (12%) patients without cough. Of 70 patients with a normal chest radiograph and no cough, 14 (20%) had a positive sputum microbiology test result (AFB smear, nucleic acid amplification test, or culture). In the previously mentioned reports from both non-TB-endemic and TB-endemic countries, the proportion of PLNTB patients that were HIV co-infected, when indicated, was low. Recovery of MTB in sputum in the absence of a chest radiographic abnormality has been reported in advanced HIV-positive patients who have an epidemiologic risk for TB infection, independent of co-existent extrapulmonary TB [
MTB complex, with rare exceptions, is introduced into the host and reaches PLNs through the oropharynx, the lungs, or the gastrointestinal tract. In the past, perhaps because of frequent exposure of children to advanced pulmonary disease (larger infecting doses) or greater prevalence of M. bovis (understood to infect lymph nodes first and lungs later) [
]. A more intriguing question is just how MTB arises in the lungs of PLNTB patients and is recoverable in small numbers (smears were usually negative, and time-to-culture positivity was usually long) – in the absence of respiratory symptoms or radiologically demonstrable lung parenchymal disease. In Figure 3, we propose five possible mechanisms for PLN-associated PTB: (i) contemporaneous reactivation of remote infection in the lung and lymph nodes; (ii) passage of MTB from lung to cervical lymphatics via apical adhesions [
]; (iv) lymphohematogenous seeding of the lung from active lymph node disease, or (v) retrograde migration of bacteria from lymph nodes to lung after compromise of the normal anatomic and physiologic basis of antegrade lymph flow [
). Panel II: Two pathogenetic mechanisms associated with abnormal lung parenchyma on chest radiograph; simultaneous reactivation in previously seeded lung (a) and lymph node (a); reactivation in previously seeded lung (a) with secondary seeding - through apical adhesions - and reactivation in cervical lymph node (b). These mechanisms are estimated to account for 28/63 (44.4%) PLN-associated PTB patients. Panel III: A third pathogenetic mechanism occurs when primary infection of the lung (a) is followed by secondary seeding and early reactivation in cervical lymph nodes (b) - the lung parenchyma may be normal on chest radiograph. Based on contact history this mechanism is estimated to account for 2/63 (3.2%) PLN-associated PTB patients. Panel IV: A fourth and fifth, hypothetical, pathogenic mechanisms may explain concomitant PTB in PLNTB patients having what may be normal lung parenchyma on chest radiograph; in one Mycobacterium tuberculosis reaches the lung (b) via lympho-hematogenous spread, in the other via retrograde lymphatic spread, from a lymph node (a). These mechanisms are estimated to account for 33/63 (52.4%) PLN-associated PTB patients.
That transmission events from sputum culture-positive PLNTB patients were few (approximately one new positive TST/IGRA or TST/IGRA conversion for every four patients and only a single secondary case) was not surprising given that most sputum culture-positive patients were smear-negative and had subclinical disease [
]. Undiagnosed PTB in extrapulmonary TB patients, in general, may account for the not-uncommon diagnosis of active TB in contacts of extrapulmonary TB patients [
Strengths of our study include the size of the cohort and the quality and completeness of the epidemiologic, radiologic, and mycobacteriology data, a perquisite of the program's organization. The prospective interpretation of the radiographs by multiple highly qualified readers and the performance of the logistic regression using both field and expert reader interpretations were other strengths. The primary weakness is the retrospective study design. Although not all PLNTB patients had a chest radiograph and sputum, those who did and those who did not differed minimally. The number and type of airway secretion specimens submitted on each patient were inconsistent; however, all patients who had not undergone sputum induction or a bronchoscopy washing/BAL submitted at least two spontaneously expectorated sputum specimens and the incremental yield of a third specimen in diagnosing PTB is small (5-8%) [
In summary, isochronous PTB in PLNTB patients is common, especially in those with respiratory symptoms, abnormal lung parenchyma on chest radiograph, or HIV co-infection. No single pathogenetic mechanism is likely to explain this phenomenon. Its public health consequences, although limited, may have heretofore been overlooked. Routine screening of PLNTB patients for PTB, with a symptom inquiry, chest radiograph, and sputum, is strongly recommended.
Declaration of competing interest
The authors have no competing interests to declare.
Acknowledgments
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. The authors are very grateful to the staff of the Edmonton Tuberculosis Clinic, the Calgary Tuberculosis Clinic, the Provincial Tuberculosis Clinic, Alberta Health Services, Alberta; the staff of Alberta Precision Laboratories, Alberta Health Services, Alberta; and the staff of the Radiology and Diagnostic Imaging Department, University of Alberta Hospital, and Edmonton-based artist, Jill Stanton for their invaluable contributions to this study.
Author contributions
RL, CH, AD, and AL participated in conception. RL, CH, AD, AL, and CP contributed to the methodology and formal analysis. All authors contributed to investigation and validation. CH administered, and RL supervised the project and, along with MLE, JB, CW, GA, and GT, provided resources and data curation. RL and CH wrote the original draft, and all authors reviewed and edited subsequent versions. All authors approved the final version.
Data sharing
Data will be available once results of all planned primary and secondary outcomes have been published, upon written request and provision of a detailed statistical analysis plan to the authors.
Proposal for standardization of optimized mycobacterial interspersed repetitive unit-variable-number tandem repeat typing of Mycobacterium tuberculosis.
Molecular strain typing of Mycobacterium tuberculosis to confirm cross-contamination in the mycobacteriology laboratory and modification of procedures to minimize occurrence of false-positive cultures.
00026-7&id=gr1.jpg){kind=link}
00026-7&id=gr2.jpg){kind=link}
00026-7&id=gr3.jpg){kind=link}