Advice
Clinical and technical evidence
Clinical and technical evidence
A literature search was carried out for this briefing in accordance with NICE's interim process and methods statement. This briefing includes the most relevant or best available published evidence relating to the clinical effectiveness of the technology.
Further information about how the evidence for this briefing was selected is available on request by contacting mibs@nice.org.uk.
Published evidence
Five studies are summarised in this briefing that are considered the most relevant to the technology. They include 2 systematic reviews, 1 randomised controlled trial and 2 cross-sectional studies. One systematic review analysed data of 2,418 people from 5 studies to evaluate the safety and acceptability of Cytosponge (Januszewicz et al. 2019). The other review analysed the efficacy and safety of Cytosponge for diagnosing oesophageal diseases (Iqbal et al. 2018). The randomised controlled trial is a multicentre UK study (n=109 in GP surgeries) and compared the risk of detecting Barrett's oesophagus for people having Cytosponge with those having standard care.
The clinical evidence with its strengths and limitations are summarised in the overall assessment of the evidence.
Overall assessment of the evidence
The review by Januszewicz et al. (2019) is low quality because it was limited to data largely from Barrett's Esophagus Studies (BEST). Also, there was no quality assessment of the individual studies. Iqbal et al. (2018) is a moderate quality review with pre-defined eligibility criteria and quality assessment of the included studies. But, methods for data analysis were not described in the review.
The randomised controlled trial had over 13,000 people who were randomised, but only a small proportion of people had the Cytosponge procedure (n=1,750; 25.6%). Two cross-sectional studies were abstracts. Information provided in the abstracts was limited, and the quality of these studies was not assessed.
Fitzgerald et al. (2020)
Intervention and comparator
A total of 6,834 people in the intervention group had a Cytosponge Trefoil Factor 3 (TFF3) test and 6,388 people in the usual care group had standard care. In the usual care group, people had prescriptions for acid-suppressant medicine and their GP might have provided lifestyle advice or referral for an endoscopy, depending on symptom severity.
Key outcomes
The analysis showed that 140 people were diagnosed with Barrett's oesophagus in the intervention group (127 people who had the Cytosponge-TFF3 procedure, and 13 people who did not), compared with 13 people diagnosed in the usual care group. The estimated adjusted relative risk of detecting Barrett's oesophagus was 10.6 (95% confidence interval [CI] 6.0 to 18.8) comparing the intervention group with the usual care group in a 12‑month follow up.
A total of 1,750 people chose to have the Cytosponge procedure. Of 1,654 people in the intervention group who swallowed Cytosponge, 221 (13%) with a positive TFF3 result had a subsequent confirmatory endoscopy. Of these, 127 people (57%) were diagnosed with Barrett's oesophagus and 4 people (2%) were diagnosed with stage I oesophago-gastric cancer. The Cytosponge TFF3 procedure had a positive predictive value of 59% (131 of 221 confirmatory endoscopies in people with a positive Cytosponge TFF3 result) for Barrett's oesophagus, dysplasia or oesophago-gastric cancer.
Less than 1% (n=9) of people were diagnosed with dysplastic Barrett's oesophagus (n=4) or stage I oesophago-gastric cancer (n=5) in the intervention group. No one was diagnosed with dysplastic Barrett's oesophagus or stage I oesophago-gastric cancer in the usual care group (who had the endoscopy procedure).
In 1 person, the Cytosponge detached from the thread, needing endoscopic removal. The most common side effect was a sore throat in 63 (4%) of 1,654 people.
Strengths and limitations
Over 13,000 people were included in the study analysis. Randomisation happened at a GP clinic level and at an individual patient level. People in the Cytosponge group were offered a choice of Cytosponge or usual care. People who chose to have the Cytosponge procedure might have had more problematic symptoms than those who did not choose to have the procedure. More women than men had the Cytosponge procedure.
Leeds et al. (2020, an abstract)
Study size, design and location
A cross-sectional study of 113 people at risk of oesophageal adenocarcinoma. The study location was not reported in the abstract.
Key outcomes
A total of 76 (67.2%) people were offered the Cytosponge procedure and 46 (60.5%) chose to have the procedure. Most people (n=45; 97.8%) completed the procedure. People who chose to have the procedure were significantly more likely to have a diagnosis of gastro-oesophageal reflux disease compared with those who chose to not have the test (69.6% compared with 36.7%; p=0.009). Also, they were more likely to have never smoked (p=0.09), have a higher body mass index (p=0.054) and have a morning appointment (p=0.12). The most common reasons for not having the test were 'not interested' (46.7%) and 'not enough time' (23.3%).
Two people (4.4%) had a positive test; 1 had a 4-cm Barrett's oesophagus segment, and the other had diffuse gastric intestinal metaplasia.
Januszewicz et al. (2019)
Study size, design and location
A systematic review assessed safety and acceptability of Cytosponge. A total of 5 prospective trials were included in the review.
Intervention and comparator
Cytosponge was used to collect oesophageal cells in the intervention group.
Key outcomes
Data were analysed from 2,418 people from 5 studies between May 2008 and August 2017. Of them, 2,284 people could complete the Cytosponge test. The overall acceptability of Cytosponge was satisfactory, with a median score of 6.0 (interquartile range 5.0 to 8.0). This was significantly higher when compared with endoscopy without sedation, with a median score of 5.0 (interquartile range 3.0 to 7.0; p<0.001). But, it was lower than endoscopy with sedation (medical score of a visual analogue scale 8.0, interquartile range 5.0 to 9.0; p<0.001).
There were 84 people (3.5%) who could not swallow the Cytosponge. The proportion of people who were unable to swallow the device was over 2 times higher in people with Barrett's oesophagus than those with gastro-oesophageal reflux disease. Of the 2,672 Cytosponge tests done, there were 12 serious adverse events reported, and 2 could be directly associated with Cytosponge. These included 1 detachment of the sponge and 1 pharyngeal bleeding after Cytosponge withdrawal. The others were related to endoscopic therapy done immediately after the Cytosponge test.
Shaheen et al. (2019, an abstract)
Key outcomes
Of 191 people in the study, 99.5% could swallow the Cytosponge. There was no significant difference in the acceptability of Cytosponge compared with the endoscopy procedures. Most people (93%) would be willing to repeat Cytosponge if the doctor indicated that it was medically necessary. There were 65% of people who preferred Cytosponge over endoscopy.
The most common adverse events included oropharyngeal pain (n=4; 2%) and throat irritation (n=2; 1%). The sponge detached in 2 people, both were retrieved during a subsequent endoscopy.
Iqbal et al. (2018)
Study size, design and location
A systematic review assessed safety and efficacy of Cytosponge in the diagnosis of oesophageal pathology. A total of 13 studies were included in the review (8 UK studies).
Intervention and comparator
Cytosponge was used to collect oesophageal cells in the intervention group. The efficacy was compared with upper gastrointestinal endoscopy with endoscopic biopsy, the gold standard test.
Key outcomes
A total of 6 studies were identified, which observed the efficacy of the Cytosponge in screening Barrett's oesophagus for people having upper endoscopy (esophagogastroduodenoscopy). A pooled sensitivity and specificity of Barrett's oesophagus detection using sponge devices were 81% and 91%, respectively.
Recent and ongoing studies
Introducing a non-endoscopic diagnostic test into the clinical pathway to identify high-risk patients with Barrett's oesophagus. ISRCTN registry identifier: ISRCTN91655550. Status: recruiting. No interim results published. Indication: oesophageal cancer. Devices: Cytosponge. Last update on 30 June 2020.