superDimension Navigation System to help diagnostic sampling of peripheral lung lesions

Deng et al. (2018)

Study size, design and location

Systematic review with limited meta-analysis (search dates January 2000 to November 2017).

n=1,648 ENB cases from 31 selected studies.

China.

Intervention and comparator(s)

ENB biopsy of lesion in lung parenchyma and mediastinal area.

CT-guided PTNB biopsy of lesion in lung parenchyma mapped on CT images.

EBUS-TBNA biopsy of lesion in subcarinal and bilateral hilar area.

Mediastinoscopy biopsy of the lesion or lymph node in the vicinity of the trachea, the subcarinal and the bronchi area.

CTC biopsies of tumour cells shed from solid tumour lesion into peripheral blood.

Key outcomes

No significant correlation was found between ENB detection rate and number of cases, average age of patients, sex, nodule size, lobar location of nodule, mean distance from pleura to nodule and operation time.

ENB complication rate: pneumothorax=5.2% (86/1,648), significantly negatively correlated with nodule size (p=0.018).

CT-guided PTNB complication rate: pneumothorax=23.0% (1,111/4,822).

Strengths and limitations

Large study size with some limited meta-analysis.

Outcome measures such as detection rate and diagnostic yield are not adequately defined.

There may be double counting of cases in the meta-analysis, because the selection of the 1,648 cases is not reported in a way that can be replicated.

At least 1 of the 31 ENB citations is a non-systematic review which simply quotes an earlier primary study; however, both the review and the primary study have been tabulated as separate studies of 932 cases by the authors.

Zhang et al. (2015)

Study size, design and location

Systematic review with meta-analysis (search dates 2000 to 2015).

n=1,161 lung nodules in 1,106 patients from 19 studies in the systematic review.

n=892 patients from 15 studies in the meta-analysis.

China.

Intervention and comparator(s)

superDimension Navigation System.

superDimension Navigation System combined with fluoroscopy.

superDimension Navigation System combined with radial EBUS.

superDimension Navigation System combined with ROSE.

Key outcomes

ENB alone; pooled values (95% CI):

Sensitivity=0.82 (0.79 to 0.85)

Specificity=1.00 (0.98 to 1.00)

PLR=18.67 (9.04 to 38.55)

NLR=0.22 (0.15 to 0.32)

DOR=97.36 (43.75 to 216.69)

Summary ROC AUC=0.9842 (SE=0.0113)

Complication rate: 40 pneumothorax in 681 procedures=5.9%.

ENB plus ROSE (6 studies):

Sensitivity range=85 to 92%

Specificity range=96.5 to 100%.

Strengths and limitations

Well-reported large study with clinically relevant diagnostic outcomes.

Limited comparative data, as none of the included studies compared ENB to surgery as a gold standard. Some minor inconsistencies in reported results in tables in the paper, versus narrative.

Gex et al. (2014)

Study size, design and location

Systematic review with meta-analysis (search date March 2012).

n=1,033 lung nodules or masses in 971 patients from 15 studies.

Switzerland.

Intervention and comparator(s)

ENB (14/15 studies used the superDimension Navigation System).

ENB combined with fluoroscopy.

ENB combined with radial EBUS.

ENB combined with ROSE.

Key outcomes

Rate of successful navigation=97.4% (95% CI 95.4 to 98.5).

ENB alone; pooled values (95% CI):

Sensitivity for malignancy=71.1% (64.6 to 76.8)

Accuracy for malignancy=78.6% (72.8 to 83.4)

Diagnostic yield=64.9% (59.2 to 70.3)

Diagnostic accuracy=73.9% (68.0 to 79.2)

NPV=52.1% (43.5 to 60.6).

Complication rate: 32 pneumothorax in 1,033 procedures=3.1% (95% CI 2.1 to 4.3).

Combined procedures:

ENB combined with fluoroscopy: diagnostic yield=56.3%

ENB combined with ROSE: sensitivity=80.2%.

Strengths and limitations

Well-reported large study with clinically relevant diagnostic outcomes. Issues of variable definitions of diagnostic yield across the constituent primary studies are addressed by study selection criteria.

Limited comparative data, because none of the included studies compared ENB with surgery as a gold standard.

Wang Memoli et al. (2012)

Study size, design and location

Systematic review with meta-analysis (search date up to October 2010).

n=3,052 lesions in 3,004 patients from 39 studies.

US.

Intervention and comparator(s)

ENB.

Virtual bronchoscopy.

Radial EBUS.

Ultrathin bronchoscope.

Guide sheath.

Key outcomes

Diagnostic yield (inverse variance weighted, with 95% CI values):

ENB=67.0% (62.6 to 71.4)

Virtual bronchoscopy=72.0% (65.7 to 78.4)

Radial EBUS=71.1% (66.5 to 75.7)

Ultrathin bronchoscope=70.0% (65.0 to 75.1)

Guide sheath=73.2% (64.4 to 81.9)

Overall=70.0% (67.1 to 72.9).

Complication rate: pneumothorax=1.5% (range 0.0 to 7.5).

Strengths and limitations

Well-reported large study with clinically relevant diagnostic outcomes.

Heterogeneity of diagnostic yield results across the constituent primary studies is acknowledged and discussed, with subgroup analysis between the different technology types reported.

Eberhardt et al. (2007)

Study size, design and location

Dual centre RCT between January 2003 and August 2006.

n=120 patients.

Germany and US.

Intervention and comparator(s)

ENB (superDimension/Bronchus system)

EBUS.

ENB with EBUS.

Key outcomes

ENB:

Overall diagnostic yield=59%

Pneumothorax rate=5%.

EBUS:

Overall diagnostic yield=69%

Pneumothorax rate=5%.

ENB/EBUS combined procedure:

Overall diagnostic yield=88%

Pneumothorax rate=8%.

Strengths and limitations

Well-reported RCT with clear definitions of all baseline characteristics and outcome measures. 118/120 patients had confirmed histological findings (from either the bronchoscopy, or, if this was non-diagnostic, from a follow-up, gold standard, surgical biopsy) and were included in the analyses, giving robust data for diagnostic yield.

The biopsy technique using bronchoscopy was standardised to forceps therefore avoiding confounders from alternatives such as needle, brush or washing biopsies.

This is a relatively small RCT, with 40 patients randomised to each arm.

Bhatt et al. (2018)

Study size, design and location

Retrospective comparative cohort between 2013 and 2015.

n=150 biopsies in 146 patients with ENB-guidance.

n=150 biopsies in 149 patients with CT-guidance.

US.

Intervention and comparator(s)

ENB-guided biopsy using the superDimension system.

CT-guided biopsy.

Key outcomes

ENB-guided biopsy:

Overall diagnostic yield=66.0% (99/150)

Any pneumothorax=4.0% (6/150)

Pneumothorax needing chest tube / admission=2.7% (4/150)

Any haemorrhage=3.3% (5/150)

Symptomatic haemorrhage=2.0% (3/150).

CT-guided biopsy:

Overall diagnostic yield=86.0% (129/150)

Any pneumothorax=28.7% (43/150)

Pneumothorax requiring chest tube/admission=1.3% (2/150)

Any haemorrhage=16.7% (25/150)

Symptomatic haemorrhage=1.3% (2/150).

Strengths and limitations

Clinically relevant, patient level data presented, comparing ENB with standard care.

The authors highlight a lack of statistical power as a possible limitation.

Folch et al. (2019)

Study size, design and location

Prospective, single-armed, multicentre global, pragmatic cohort study from April 2015 to August 2016 (12-month results from the NAVIGATE study).

n=976/1,215 patients with 12-month results from 29 sites (of 37 sites in total).

US.

Intervention and comparator(s)

ENB using the superDimension navigation system, version 6.0 or higher.

n=1,157/1,215 with ENB-guided lung lesion biopsy, 258/1,215 ENB with fiducial placement, 23/1,215 ENB with plural dye marking and/or 30/1,215 ENB with lymph node biopsy.

Biopsy tools include aspirating needles, biopsy forceps, cytology brushes, core biopsy, and bronchoalveolar lavage.

Key outcomes

Pneumothorax needing hospitalisation or intervention (CTCAE grade 2 or greater)=2.9% (35/1,215)

Any grade pneumothorax=4.3%

Navigation success rate=94.4% (1,092/1,157)

Diagnostic yield ([TP+TN]/all attempted biopsies)=72.9% (768/1,053)

Excluding 104/1,157 deferred cases, for malignancy:

Sensitivity=68.8% (484/704)

Specificity=100% (284/284)

PPV=100% (484/484)

NPV=56.3% (284/504).

Strengths and limitations

Well-reported, large study with clear definitions of baseline characteristics, outcome measures and data flow diagrams.

Pragmatic, 'real-world' study gives heterogeneous patient pathways, including some ENB combined procedures.

Abbreviations: AUC, area under the curve; CI, confidence interval; CT, computed tomography; CTC, circulating tumour cell; CTCAE, Common Terminology Criteria for Adverse Events; DOR, diagnostic odds ratio; EBUS, endobronchial ultrasonography; ENB, electromagnetic navigation bronchoscopy; NLR, negative likelihood ratio; NPV, negative predictive value; PLR, positive likelihood ratio; PPV, positive predictive value; PTNB, percutaneous trans-thoracic needle biopsy; RCT, randomised controlled trial; ROC, receiver operating characteristic; ROSE, rapid on-site cytological evaluation; SE, standard error; TBNA, trans-bronchial needle aspiration; TN, true-negative for malignancy; TP, true-positive for malignancy.