Advice

Evidence review

Evidence review

Clinical and technical evidence

Regulatory bodies

A search of the Medicines and Healthcare Products Regulatory Agency website revealed no manufacturer Field Safety Notices or Medical Device Alerts for this device.

Four reports of adverse events were identified from a search of the US Food and Drug Administration (FDA) Manufacturer and User Device Facility Experience (MAUDE) database. One event was reported twice, leaving 3 unique events involving an ERBE cryoprobe. It is unclear which cryoprobe or system model was used in each case. Two of the adverse events were listed as device malfunctions (missing probe tip and loud noise near the connection to the ERBE unit) and did not result in patient harm. The remaining adverse event was a patient death. However, this was deemed to be a consequence of the patient's underlying disease state and unrelated to the device.

Clinical evidence

Of the 118 relevant papers identified, 90 were excluded because they were duplicates, case studies, reviews, animal studies, about peripheral lung lesions or did not mention the device used. There were 28 appropriate studies that used an ERBE flexible cryoprobe. This briefing includes only diagnostic randomised controlled trials for cryobiopsy and all cryorecanalisation studies (see evidence selection). Studies on cryonecrosis were excluded from further assessment on the basis of small sample sizes, date of publication, or because they were reviews.

Diagnostic randomised controlled trials for biopsy

Hetzel et al. (2012) compared the diagnostic yield and safety when using the ERBE flexible cryoprobe for cryobiopsy with conventional sampling methods (forceps) for potentially cancerous endobronchial lesions. The ERBE model used was not stated. Diagnostic yield is the ratio of diagnostic findings and the total number of procedures done using the technique. It is defined as the likelihood that a test or procedure will provide the information needed to establish a diagnosis. The study randomised 593 patients into the cryoprobe (n=296) or forceps (n=297) groups. There was no statistically significant difference in baseline characteristics between the study groups. The diagnostic yield was 95.0% for cryobiopsy and 85.1% for standard forceps (p<0.001). No bleeding occurred in 19.9% of patients in the cryoprobe group and 30.6% of patients in the forceps group (p=0.009). Mild bleeding occurred in 61.8% of cryoprobe group patients and 51.5% of forceps group patients (no p‑value reported). There was no significant difference in the number of bleeding complications needing intervention between the 2 groups (p=0.90). The proportions of non‑diagnostic results in patients with non‑small cell lung cancer were 11.8% for forceps biopsy and 5.5% for cryobiopsy (p=0.025); non‑diagnostic results for patients with small cell lung cancer were also lower for cryobiopsy (3.6% versus 16.1% for forceps biopsy, p=0.024).

Jabari et al. (2012) compared the diagnostic yield with and safety of endobronchial biopsies using the ERBE flexible cryoprobe (900 mm in length, 2.3 mm diameter) with forceps biopsy. Patients (n=60) with an established endobronchial lesion were included in the study. All patients had 3 biopsies, a forceps biopsy and 2 cryotherapy biopsies (1 for 3 seconds and 1 for 5 seconds), in a random order. The median biopsy size was 0.5 cm for forceps biopsy, 0.8 cm for 3‑second cryobiopsy and 1.6 cm for 5‑second cryobiopsy. The diagnostic yield was 66.7% for forceps biopsy, 80% for 3‑second cryobiopsy and 76.6% for 5‑second cryobiopsy; differences in diagnostic yield were not statistically significant (p>0.05). However, a combination of both cryobiopsy methods resulted in an increase of diagnostic yield to 90.0% (p=0.02). Bleeding complications occurred in 8 cases; there were no differences in the incidence or bleeding type among the 3 sampling methods.

Schumann et al. (2010a) compared the diagnostic yield with, and the feasibility and safety of, endobronchial biopsies using the ERBE flexible cryoprobe (780 mm in length, 2.3 mm diameter) with forceps biopsy. patients (n=296) with an endoscopically visible endobronchial lesion were included in the study. The first 55 patients were randomised consecutively for the order of the biopsy procedures (forceps biopsy followed by cryobiopsy or vice versa). The remaining patients were not randomised and only had cryobiopsy. The diagnostic yield was 89.1% and 65.5% for cryobiopsy and forceps biopsy respectively (p<0.05). In addition, tissue samples taken by cryobiopsy were significantly larger and freer from artefacts than tissue obtained from forceps biopsy (p<0.001).

Cryorecanalisation

Hetzel et al. (2004) did a prospective study to examine the safety and effectiveness of cryorecanalisation with an ERBE flexible cryoprobe (780 mm in length, 2.3 mm diameter) in patients with respiratory tract stenosis caused by lung tumours. The study enrolled 60 patients with high‑grade stenosis in the central respiratory tract (caused by tumours) or in a segment of a bronchus with post‑obstructive pneumonia (as a consequence of lung tumours). Recanalisation was defined as successful if there was no residual stenosis detectable by endoscopy and partially successful if residual stenosis was evident by endoscopy but passable with the bronchoscope. Recanalisation was successful in 61% of the patients and partially successful in 22%. Bleeding that stopped spontaneously within minutes occurred in 54 patients, but 6 patients had more intense bleeding (100–300 ml of blood loss). However, the bleeding was controlled and stopped with intervention in all 6 patients.

Schumann et al. (2010b) investigated the efficacy and safety of the ERBE flexible cryoprobe (length and diameter not stated) for immediate tumour ablation in patients with endobronchial or tracheal tumour obstruction. In a retrospective study of 225 patients, the authors examined the success rate and bleeding complications associated with cryorecanalisation. The intervention was defined as successful if the endobronchial tumour mass could be removed so that either drainage of secretions or reduced airway stenosis led to an improvement in the patient's condition and symptoms. The procedure was successful in 91.1% of patients. Bleeding complications occurred in 27 patients (12.0 %). Of these, 9 had mild bleeding (ice‑cold sodium chloride or adrenaline solution was needed to treat the bleed) and 18 had moderate bleeding (argon plasma coagulation or a bronchus blocker was needed to treat the bleed).

Recent and ongoing studies

Three ongoing or in‑development trials using ERBE cryoprobes were identified in the preparation of this briefing.

  • NCT01475084 – the purpose of the study is to compare cryobiopsies using an ERBE flexible cryoprobe (2.4 mm diameter, 900 mm length) and ERBOKRYO CA with forceps biopsies during semi‑rigid thoracoscopy. The Slovenia‑based study began in November 2011 and its status is unknown.

  • NCT02075762 – the objective of this study is to compare the sample size, architectural preservation and diagnostic yield of bronchoscopic cryoprobe lung biopsy with that of standard (using forceps) and video‑assisted thoracoscopic surgery lung biopsy for the diagnosis of interstitial lung disease. This USA‑based study began in August 2013 and is expected to complete in August 2015. It is a prospective cohort study in which 20 subjects will be enrolled.

Costs and resource consequences

ERBE flexible cryoprobes can be used in place of diagnostic or therapeutic devices for lung biopsy, recanalisation or selective tissue necrosis, including: forceps, electrocautery, neodymium‑doped yttrium aluminium garnet laser, argon plasma coagulation or photodynamic therapy. A flexible bronchoscope is needed alongside the technology as well as general or local anaesthesia, sedation and intubation as appropriate. Medical professionals must be trained to use the device according to the instructions for use. The ERBECRYO 2 and the ERBOKRYO CA systems are currently used in the NHS.

No published evidence on the resource consequences of using ERBE flexible cryoprobes was identified in the systematic review of evidence.

Strengths and limitations of the evidence

Three procedures were done in the Jabari et al. (2012) study (forceps biopsy, cryobiopsy in 3 seconds and cryobiopsy in 5 seconds). Although the tissue samples obtained by cryobiopsy in 5 seconds were larger than those obtained by cryobiopsy in 3 seconds, the diagnostic yield of cryobiopsy in 3 seconds was higher. This might be because multiple samples were taken from the same target area and the mechanical damage to the small area may have influenced the results.

In all 3 biopsy randomised controlled trials, diagnostic yield was a primary outcome measure. The studies reported the proportion of patients who were diagnosed as a proportion of the total number of patients assessed. However, none of the studies presented diagnostic accuracy, so it is not possible to discern the proportion of patients who were correctly diagnosed.

Hetzel et al. (2012) was the only included study that did a power calculation for sample size. It is unclear whether the remaining studies had an adequate sample size to draw reliable conclusions.

The study by Hetzel et al. (2012) also eliminated performance bias by blinding the evaluating pathologist to the intervention technique. In the studies by Schumann et al. (2010a) and Jabari et al. (2012), patients had biopsies via both forceps and flexible cryoprobe procedures. Both studies randomised the order of sampling within each patient to eliminate order‑effect bias.

Hetzel et al. (2004) and Schumann et al. (2010b) both examined cryorecanalisation using the cryoprobe over different time periods within the same German healthcare institution. Neither were controlled studies, and both reported the safety and effectiveness of the technique without a reference standard. Schumann et al. (2010b) showed a higher rate of successful recanalisation than Hetzel et al (2004). Schumann et al. (2010b) highlighted that Hetzel et al. (2004) was a feasibility study, and that the increased success rate in their later publication may have been because of the learning gained in the cryorecanalisation procedure. Although neither study provided a sample size calculation, the study by Schumann et al. (2010b) examined a larger sample size of 225 patients, which reduced the likelihood of recanalisation successes occurring by chance when compared with Hetzel et al. (2004), which studied only 60 patients. However, the Hetzel et al. (2004) study was done prospectively, whereas the Schumann et al. (2010b) study was retrospective; sources of error due to confounding and bias are more common in retrospective studies than in prospective studies.

All of the studies use cryoprobes for ERBOKRYO CA. The difference between ERBECRYO 2 and ERBOKRYO CA lies in the updated unit display and design of ERBECRYO 2 rather than the clinical functionality of the device. Therefore, results for the ERBOKRYO CA may be generalisable to the ERBECRYO 2.

None of the studies included in this briefing took place in the UK, so the generalisability of the results to the NHS is unclear. Similarly, although the device is not contraindicated for children, no relevant evidence was found on the safety and effectiveness of ERBECRYO 2 or ERBOKRYO CA in patients under 18 years old.

Finally, Jabari et al. (2012) was the only study without an author related to, or funding from, the ERBE company. The manufacturer directly contributed to the studies by Hetzel et al. (2004) and Hetzel et al. (2012). At least 1 author from Schumann et al. (2010a) and Schumann et al. (2010b) disclosed that they received lecture fees from ERBE. Manufacturer involvement in the project may potentially introduce bias in the reporting of outcomes.