RSTMH Early Career Grant Programme case study 2: Project statement

1. Project Background

Current diagnostic algorithms for pulmonary tuberculosis (PTB) rely primarily on sputum microscopy, chest x-ray (CXR), and, more recently, nucleic acid amplification tests (NAAT). However, microscopy has relatively poor sensitivity, and many clinicians in resource-limited areas where PTB is common may not have ready access to CXR or NAAT. Lung ultrasound can detect abnormalities of the pleura and parenchyma, and has been increasingly used for diagnosis of pneumonia with excellent performance characteristics compared to conventional radiography. For instance, a systematic review and meta-analysis demonstrated lung ultrasound had a pooled sensitivity and specificity of 94% and 97% for the diagnosis of pneumonia in adults (Chavez M, et al. Lung Ultrasound for the diagnosis of pneumonia in adults: a systematic review and meta-analysis. Respir Res. 2014;15(1):50). Despite the extensive literature base for lung ultrasound in pneumonia, literature on its use to aid in the diagnosis of PTB is extremely limited. Two recent case series suggested several abnormal lung ultrasound findings in patients with pulmonary tuberculosis, but these studies were small (combined n=70), and included no control arm. This study therefore will seek to describe abnormal sonographic lung findings in patients with confirmed PTB, and compare these findings to those in control groups without tuberculosis.

Accurate and timely diagnosis of pulmonary tuberculosis is critical to TB control programs, and improved point-of-care tests and diagnostic algorithms have been identified as research priorities by WHO. If abnormal lung ultrasound can be shown to correlate with a diagnosis of pulmonary tuberculosis, then incorporation of ultrasound into diagnostic algorithms could help expedite accurate diagnosis and treatment initiation, especially in resource constrained settings without access to chest x-ray and in smear negative patients. I am currently conducting a small, pilot prospective study of lung ultrasound in smear positive pulmonary tuberculosis patients and changes in ultrasound findings during treatment. Building on these results, the present study will compare lung ultrasound findings between pulmonary tuberculosis patients and patients without pulmonary tuberculosis. To our knowledge, this type of comparison has not yet been published in the literature.

2. Aims and Objective:

The aim of this study is to describe abnormal sonographic lung findings in patients with confirmed PTB, and compare these findings to those in patients without tuberculosis. This work will build on an ongoing prospective study of lung ultrasound in smear positive PTB patients, and evaluate the potential for lung ultrasound to be incorporated into a diagnostic algorithm that could reduce time to diagnosis and treatment initiation. The study objectives are:
Objective 1: To assess the correlation of abnormal lung ultrasound findings with microbiologically confirmed PTB.
Objective 2: To explore a comparison of the predictive value of lung ultrasound findings for microbiologically confirmed PTB with the predictive value of chest x-ray.
Objective 3: To assess the correlation of lung ultrasound findings with chest x-ray findings in patients with microbiologically confirmed PTB.

3. Study Design:

Overview
This multi-center prospective observational study will recruit adult patients with smear-positive pulmonary tuberculosis who have started anti-tuberculosis treatment within the previous one week, and three control groups without tuberculosis -- one group with pneumonia, one group with asthma or COPD exacerbation, and one group of asymptomatic volunteers. Each participant will have a lung ultrasound scan performed at enrollment. Each participant at enrollment will also complete a questionnaire and, if not already obtained as part of routine medical care, a chest x-ray.

Patient Population
The study will enroll participants from four distinct groups:
Group 1: Patients with confirmed pulmonary tuberculosis within 1 week of starting anti-tuberculosis treatment
Group 2: Patients diagnosed with lower respiratory tract infection within the past 3 days
Group 3: Patients diagnosed with acute asthma or COPD exacerbation within the past 3 days
Group 4: Healthy volunteers

Inclusion Criteria
For all groups: Age greater than or equal to 18 years; Agreement to participate through comprehension and signing informed consent.

The following are additional inclusion criteria for each group:
Group 1: Diagnosis of pulmonary tuberculosis confirmed by positive smear, Xpert MTB/RIF, MODS liquid culture or Ogawa medium solid culture; Within one week of initiation of treatment for pulmonary TB; Sputum for AFB smear and drug susceptibility testing obtained as part of routine medical care.
For Group 2: Diagnosis of LRTI by hospital pulmonologist in past 3 days by combination of clinical signs and suggestive chest x-ray.
For Group 3: Diagnosis of asthma or COPD exacerbation by hospital pulmonologist by clinical signs within the previous 3 days.
For Group 4: No respiratory symptoms – specifically cough, shortness of breath or chest pain – within the previous 2 months.

Exclusion Criteria
History of TB treatment prior to the current episode. These patients will be excluded due to high frequency of baseline pleural pathology.
Pregnancy
Unable to cooperate with lung ultrasound protocol
Unwilling or unable to provide fully informed consent

Study Setting
The study will be conducted in Lima, Peru at Hospital de Huaycan and Hospital Nacional Dos De Mayo.

Study Procedures
After discussion of the study and agreement to participate confirmed by the signing of the informed consent form, patients in all four study groups meeting the inclusion criteria will have the following study procedures at the baseline visit:
(1) Questionnaire administration
(2) Limited chart review - sputum AFB and drug sensitivity results, HIV status, CD4 count and ART status if applicable
(3) Comprehensive lung ultrasound, performed by clinician-sonographer blinded to participant's diagnosis and other clinical information
(4) Chest x-ray, if not already obtained as part of routine medical care

In addition, all patients in Groups 1-3 will have a medical chart review 8 weeks after enrollment to ensure accurate study diagnosis.

Data Analysis
The primary objective will be addressed by first measuring the sensitivity and specificity of each individual abnormal lung ultrasound finding for microbiologically confirmed PTB. We will then combine these findings to develop a receiver operating characteristic (ROC) curve and explore development of a predictive model. For the secondary objectives, sensitivity and specificity of chest x-ray for microbiologically confirmed PTB will be calculated across all 4 study groups and then, using all ultrasound abnormalities to predict PTB, we will compare the area under the ROC curve (AUC) for ultrasound and x-ray. Finally, we will assess the difference between ultrasound and chest x-ray for identifying consolidation, miliary pattern, cavitation and pleural effusion using the McNemar test.

Sample Size
The study is powered to Objective 1. Based on preliminary data regarding pulmonary ultrasound findings in PTB and pneumonia, we predict that a model incorporating all lung ultrasound abnormalities will have a sensitivity of 90% and a specificity of 70%. Given a sample of 30 confirmed PTB cases, and 90 participants without PTB (30 from each of 3 groups), the 95% confidence interval around a predicted sensitivity of 90% will be 73-98%, and around a predicted specificity of 70% will be 59-79% when considering all 3 control groups together. Therefore a total sample of 120 patients will allow an adequate assessment of Objective 1.

4. Approach used to maximize the impact of research outputs:

The impact of research outputs will be maximized by presenting the results at scientific conferences, and publishing the results in a peer-reviewed journal. We will also present the findings to local stakeholders, including XXX Hospital and YYY Group.

5. Expected Outcomes:

If specific lung ultrasound findings are repeatedly demonstrated in PTB, lung ultrasound has the potential to be incorporated into point-of-care diagnostic algorithms and shorten time to diagnosis and treatment initiation, especially in resource constrained settings without access to chest x-ray and in smear negative patients. A manuscript based on the study data will be submitted to a peer-reviewed journal for publication. An abstract based on the study data will be submitted to several conferences by way of application for oral or presentations. The findings from this study will provide guidance for the methodology of a future study to evaluate the accuracy of lung ultrasound for the diagnosis of PTB in a cohort of patients with suspected PTB.

6. My role in the project:

I will be the Co - Primary Investigator, along with my mentor and supervisor Dr. ABC.