G Model EURR-7146; No. of Pages 8

ARTICLE IN PRESS European Journal of Radiology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad

Role of digital tomosynthesis and dual energy subtraction digital radiography in detection of parenchymal lesions in active pulmonary tuberculosis Madhurima Sharma a,1 , Manavjit Singh Sandhu a,2 , Ujjwal Gorsi a,∗ , Dheeraj Gupta b , Niranjan Khandelwal a,3 a b

Department of Radiodiagnosis and Imaging, PGIMER, Chandigarh 160012, India Department of Pulmonary Medicine, PGIMER, Chandigarh 160012, India

a r t i c l e

i n f o

Article history: Received 17 February 2015 Received in revised form 13 April 2015 Accepted 26 May 2015 Keywords: Tuberculosis Digital tomosynthesis Dual-energy subtraction digital radiography CT

a b s t r a c t Objective: To assess the role of digital tomosynthesis (DTS) and dual energy subtraction digital radiography (DES-DR) in detection of parenchymal lesions in active pulmonary tuberculosis (TB) and to compare them with digital radiography (DR). Materials and methods: This prospective study was approved by our institutional review committee. DTS and DES-DR were performed in 62 patients with active pulmonary TB within one week of multidetector computed tomography (MDCT) study. Findings of active pulmonary TB, that is consolidation, cavitation and nodules were noted on digital radiography (DR), DTS and DES-DR in all patients. Sensitivity, specificity, positive and negative predictive values of all 3 modalities was calculated with MDCT as reference standard. In addition presence of centrilobular nodules was also noted on DTS. Results: Our study comprised of 62 patients (33 males, 29 females with age range 18–82 years). Sensitivity and specificity of DTS for detection of nodules and cavitation was better than DR and DES-DR. Sensitivity and specificity of DTS for detection of consolidation was comparable to DR and DES-DR. DES-DR performed better than DR in detection of nodules and cavitation. DTS was also able to detect centrilobular nodules with sensitivity and specificity of 57.4% and 86.5% respectively. Conclusion: DTS and DES-DR perform better than DR in detection of nodules, consolidation and cavitation in pulmonary TB. DTS gives better results than DES-DR, particularly in detection of cavitation and has moderate sensitivity for detection of centrilobular nodules. Thus DTS can be used for evaluation of patients of suspected pulmonary TB, thereby giving a more confident diagnosis of active disease and also in follow up. © 2015 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Pulmonary TB is a common worldwide infection and is a major cause of mortality and morbidity especially in developing coun-

Abbreviations: AEC, automatic exposure control; AFB, acid fast bacilli; ATT, antitubercular therapy; BAL, bronchoalveolar lavage; CR, computed radiography; CT, computed tomography; DES-DR, dual energy subtraction digital radiography; DR, digital radiography; DTS, digital tomosynthesis; MDCT, multidetector computed tomography; FNAC, fine needle aspiration cytology; TB, tuberculosis. ∗ Corresponding author. Tel.: +91 9914200324. E-mail addresses: [email protected] (M. Sharma), [email protected] (M.S. Sandhu), [email protected] (U. Gorsi), [email protected] (D. Gupta), [email protected] (N. Khandelwal). 1 Tel.: +91 9592232819. 2 Tel.: +91 9914209384. 3 Tel.: +91 9914209381.

tries. According to World Health Organisation global tuberculosis report, globally estimated new cases of tuberculosis were 8.7 million in 2011 (125 cases per 100,000 population) with 59% of the estimated cases from Asia. India had largest number of incident cases in 2011 (2.0–2.5 million cases) and accounted for 26% of all global cases [1]. Prompt diagnosis of TB is highly desirable for adequate treatment of patients and control of disease spread. However clinical features are often non specific, thereby posing difficulty in making a definitive clinical diagnosis. The gold standard for diagnosis is demonstration of mycobacteria in respiratory secretions. However, only few patients with active TB show acid fast bacilli in their sputum. Thus a radiological diagnosis of the disease can facilitate earlier treatment for patients thereby ensuring better disease control [2]. Though CT is not confirmatory for pulmonary TB, patients with high suspicion of TB on imaging can be selected, and other diseases can be excluded [3].

http://dx.doi.org/10.1016/j.ejrad.2015.05.031 0720-048X/© 2015 Elsevier Ireland Ltd. All rights reserved.

Please cite this article in press as: M. Sharma, et al., Role of digital tomosynthesis and dual energy subtraction digital radiography in detection of parenchymal lesions in active pulmonary tuberculosis, Eur J Radiol (2015), http://dx.doi.org/10.1016/j.ejrad.2015.05.031

G Model

ARTICLE IN PRESS

EURR-7146; No. of Pages 8

M. Sharma et al. / European Journal of Radiology xxx (2015) xxx–xxx

2 Table 1 Baseline characteristics of study population. Demographics: Age Male:female History: 1. Cough 2. Fever 3. Loss of appetite and weight 4. Loss of appetite 5. Haemoptysis

tivation TB [2]. So it cannot be always relied upon for an accurate diagnosis. In cases with normal or inconclusive radiographs, multidetector computed tomography (MDCT) is the investigation of choice to assess for the extent and distribution of the disease. MDCT is accurate in 91% of patients for diagnosing TB [10]. However considering the young age of many patients and the socioeconomic burden of tuberculosis, especially in developing countries, both radiation dose and cost of MDCT remain the limiting factors. Dual-energy subtraction digital radiography (DES-DR) and digital tomosynthesis (DTS) are recent advancements in digital radiography (DR) having certain advantages over conventional radiography, with a lower cost and radiation dose than computed tomography (CT). These modalities can be useful in better demonstration of pulmonary parenchymal abnormalities. The purpose of our study was to prospectively assess the diagnostic performance of DES-DR and DTS in detection of parenchymal lesions in patients of active pulmonary TB, keeping MDCT as reference standard.

18–82 years 33:29 41 (66%) 32 (52%) 14 (22.6%) 2 (3.2%) 5 (8.1%)

Diagnosis: 1. Sputum smear positivity 2. Sputum culture positivity 3. BAL fluid positivity 4. Cytological diagnosis 5. Presumptive diagnosis

20 (32.3%) 17 (27.4%) 7 (11.3%) 3 (4.8%) 15 (24.2%)

Findings on CT: 1. Nodules 2. Consolidation 3. Cavitation 4. Centrilobular nodules

61 (98.4%) 44 (71%) 31 (50%) 40 (64.5%)

Pulmonary mycobacterial disease can give rise to various non specific radiologic abnormalities depending upon certain host factors like age, immunity and past exposure to TB [4]. Primary TB commonly manifests as mediastinal and hilar lymphadenopathy, homogenous air space consolidation and pleural effusion (usually unilateral) [4]. Post primary TB most commonly manifests as patchy or focal heterogeneous consolidation (usually in the apical and posterior segments of the upper lobes and the superior segments of the lower lobes), ill defined nodules and cavities (seen in 20–45% of patients) [4–6]. It is important to differentiate between active and inactive pulmonary TB rather than categorizing it as primary or post primary disease, as activity of the disease will decide further course of management. Consolidation (usually patchy and multifocal), cavitation and nodules (air space as well as branching centrilobular nodules) are common radiological abnormalities in active pulmonary TB. Scattered ill defined air space nodules (usually 5–10 mm in size) and branching centrilobular nodules showing tree-in-bud appearance indicate endobronchial spread of the disease and are most helpful for establishing diagnosis of active disease [7]. According to Hatipoglu et al. [8] presence of branching centrilobular nodules with tree in bud appearance is most useful radiological feature distinguishing active form inactive TB. Chest radiography has a major role in tubercular diagnosis, screening and follow up [2] and is often the first radiological investigation. It is easily available and imparts very low radiation dose, but has certain limitations due to overlap of anatomic structures, variable X-ray transmission, scattered radiation and variability in perception of abnormalities [9]. The diagnosis of TB on radiography is accurate in 49% of all cases – 34% in primary TB and 59% in reac-

2. Materials and methods This prospective study was carried out in a tertiary care hospital after approval from the institutional review committee. 70 patients were enrolled in the study with a clinical suspicion of active pulmonary TB. All the patients had undergone MDCT in the last one week. Informed written consent was obtained from all the patients. Clinical presentation and relevant investigations of all the patients were also noted. The diagnosis of active pulmonary TB was confirmed by a positive sputum/bronchoalveolar lavage (BAL) culture for acid fast bacilli (AFB), demonstration of AFB on sputum/BAL smears or histopathological/cytological diagnosis. In few patients, diagnosis of pulmonary TB could not be confirmed on sputum, fluid or tissue sampling. In view of strong clinical and radiological suspicion of pulmonary TB, and failure to respond to antibiotics, these patients were started on anti tubercular therapy (ATT). MDCT was done in 64/128 slice scanner depending upon the availability. Parameters for 64-detector CT system (Lightspeed VCT of GE Healthcare) were: detector collimation 64 × 0.625 mm; helical pitch 1.375; rotation time 0.6 s; tube voltage 120 kVp; AEC controlled tube current; 10 mm section thickness; and reconstruction at 2.5 mm using high spatial frequency reconstruction algorithm. Parameters for 128-detector CT system (Somatom definition flash by Siemens) were: detector collimation 128 × 0.6 mm; helical pitch 1.2; rotation time 0.5 s; tube voltage 120 kVp; Automatic exposure control (AEC) controlled tube current; 10 mm section thickness; and reconstruction at 2 mm with high spatial frequency reconstruction algorithm. MDCT images were analysed by a radiologist having 10 years of experience in chest radiology.

Table 2 Comparison of performance of DR, DES-DR and DTS: Table demonstrating sensitivity, specificity, positive and negative predictive values of DR, DES-DR and DTS in detection of lung nodules, consolidation and cavitation. Observer 1

Observer 2

Lesion

Modality

Sensitivity

Specificity

PPV

NPV

p value

Sensitivity

Specificity

PPV

NPV

p value

Nodules

DR DES-DR DTS

82% (52/61) 90.2% (55/61) 98.4% (60/61)

100% (1/1) 100% (1/1) 100% (1/1)

100% 100% 100%

10% 14.3% 50%

.040 .005 .000

73.8% (45/61) 85.2% (52/61) 100% (61/61)

100% (1/1) 100% (1/1) 100% (1/1)

100% 100% 100%

5.9% 10% 100%

.101 .022 .000

Consolidation

DR DES-DR DTS

75% (33/44) 88.6% (39/44) 88.6% (39/44)

72.2% (13/18) 72.2% (13/18) 66.7% (12/18)

86.8% 88.6% 86.7%

54.2% 72.2% 70.6%

.001 .000 .000

77.3% (34/44) 88.6% (39/44) 88.6% (39/44)

77.8% (14/18) 72.2% (13/18) 66.7% (12/18)

89.5% 88.6% 86.7%

58.3% 72.2% 70.6%

.000 .000 .000

Cavitation

DR DES-DR DTS

67.7% (21/31) 90.3% (28/31) 100% (31/31)

83.9% (26/31) 80.6% (25/31) 90.3% (28/31)

80.8% 82.4% 91.2%

72.7% 89.3% 100%

.000 .000 .000

64.5% (20/31) 80.6% (25/31) 100% (31/31)

83.9% (26/31) 83.9% (26/31) 83.9% (26/31)

83.3% 83.3% 86.1%

81.3% 81.3% 100%

.000 .000 .000

Please cite this article in press as: M. Sharma, et al., Role of digital tomosynthesis and dual energy subtraction digital radiography in detection of parenchymal lesions in active pulmonary tuberculosis, Eur J Radiol (2015), http://dx.doi.org/10.1016/j.ejrad.2015.05.031

G Model

ARTICLE IN PRESS

EURR-7146; No. of Pages 8

M. Sharma et al. / European Journal of Radiology xxx (2015) xxx–xxx

3

Table 3 Performance of DR, DES-DR and DTS in detection of cavities (on per cavity basis). Observer 1

Number of cavities Mean SD Maximum Minimum

Observer 2

CT

DR

DES-DR

DTS

DR

DES-DR

DTS

141 2.27 3.667 15 0

58 0.94 1.638 8 0

104 1.68 2.468 10 0

126 2.03 2.92 13 0

57 0.92 1.623 8 0

90 1.45 2.324 9 0

116 1.87 2.525 11 0

DES-DR was done in Definium 8000 DR system of GE Healthcare with flat panel detector (41 cm × 41 cm with amorphous silicon photodiode transister array and caesium iodide scintillator). Erect Postero-Anterior (PA) chest image was acquired in deep inspiration using the standard dual shot method (120 kVp and 60 kVp) with automatic exposure control (AEC) determined tube load. Routine DR image without subtraction and subtracted soft tissue and bone images were generated. DTS was done immediately after DES-DR in the same system using the volumeRAD option. Erect PA scout view was taken with AEC settings. DTS images were acquired with a tube sweep angle of approximately ±15◦ and a stationary detector. The breathhold time was 10 s. The raw data was processed to give 53 images with 5 mm slice interval. The images were analysed independently by two radiologists having experience of 20 years and 7 years respectively. Radiologists were blinded to the patient’s clinical profile and to MDCT findings. Analysis was done using the commercial DICOM image viewer (RadiAnt – 64 bit – 1.8.8) and manipulation of windowing, zoom/pan parameters was allowed. The DR, DES-DR and DTS images were analysed separately in 3 different sessions with an interval of 15 days between each session. Pulmonary nodules, consolidation, cavitation and number of cavities in each patient, were noted. In addition presence of centrilobular nodules with tree in bud appearance was also noted. The findings were interpreted according to the glossary of terms from the Fleischner Society [11] as follows: Consolidation: An essentially homogenous opacity in the lung characterized by little or no loss of volume, with or without air bronchogram (on radiograph) and associated with effacement of blood vessels (on CT). Cavity: A gas filled space, seen as a lucency (on radiograph) or low attenuation area (on CT) within an area of pulmonary consolidation, a mass or a nodule. It may or may not contain a fluid level. Nodule: A sharply defined, discrete, near circular opacity less than 30 mm in diameter on radiograph. On CT it is seen as a round or irregular opacity, well or poorly defined and measuring upto 3 cm in diameter. Presence of tree in bud appearance was also noted which is defined as centrilobular nodules with branching pattern on CT. It has not been described in radiography. However Kim et al. [12] described tree in bud appearance on digital tomosynthesis images. Since there are no well established criteria to define parenchymal findings in DTS, we followed criteria used in radiography as described above. Keeping MDCT as the reference standard, sensitivity, specificity and predictive values of DR, DES-DR and DTS in detecting pulmonary nodules, consolidation and cavitation in pulmonary tuberculosis were calculated separately by the two observers. Number of cavities detected on DR, DES-DR and DTS were compared with each other using Wilcoxon signed rank test. Kappa test of agreement was used to assess the inter-observer agreement for all observations. All calculations were performed using SPSS® version 17 (Statistical Packages for the Social Sciences). All statistical tests were two-sided and performed at a significance level of ˛ = .05.

3. Results A total of 70 patients were enrolled in the study. In 8 patients, diagnosis of pulmonary TB was not confirmed and were subsequently excluded from the study. Prevalence of pulmonary TB in our study population is 88.6%. Such a high prevalence can be attributed to selection of only those patients which were referred to us for MDCT, with clinical suspicion of pulmonary TB. Baseline characteristics of study population have been summarised in Table 1.The age range of patients was from 18 to 82 with a mean of 43.56 years and standard deviation of 17.98 years. Of the total 62 patients 33 were male and 29 were female. All patients had duration of symptoms of more than 1 month except for 5 patients who presented with onset of symptoms within last one month with duration ranging from 1 to 3 weeks. 52 patients had more than one presenting symptom. Cough was the most common symptom present in 41(66%) patients. Low grade fever was next most common symptom present in 32 (52%) patients. Fever was the only presenting symptom in 3 patients. 20 patients had sputum smear positive for AFB. 17 patients had sputum culture positive for mycobacteria. Seven patients had BAL fluid positive for AFB. In 3 patients cytological diagnosis of TB was confirmed. Two patients underwent CT guided fine needle aspiration cytology (FNAC) from mass like consolidation in lung. One patient underwent CT guided FNAC from mediastinal lymph node. Thus diagnosis of TB was made based upon positive results from sputum, fluid or tissue sampling in 47 (75.8%) patients. In remaining 15 patients (24.2%) both sputum culture and sputum smear were negative. Other confirmatory tests like polymerase chain reaction (PCR) for mycobacterial DNA and T spot tests could not be performed in these patients due to non affordability. In these patients a possible diagnosis of pulmonary TB was given based upon clinical and radiological findings. After no clinical response to course of antibiotics, these patients were started on ATT and subsequently improved. CT images of all patients were analysed for presence of nodules, consolidation, cavitation and centrilobular nodules. All the patients had more than one finding on CT. 20 patients (32.3%) had all 4 findings on CT. Sensitivity, specificity and predictive values for detection of each abnormality were calculated for the two observers (Table 2). p value was

Role of digital tomosynthesis and dual energy subtraction digital radiography in detection of parenchymal lesions in active pulmonary tuberculosis.

To assess the role of digital tomosynthesis (DTS) and dual energy subtraction digital radiography (DES-DR) in detection of parenchymal lesions in acti...
2MB Sizes 2 Downloads 6 Views