Cellvizio is a confocal laser endomicroscopy (CLE) system with a fibre-optic probe for real-time imaging of tissues. It is designed for use as an adjunct to the standard endoscopic ultrasound-guided fine-needle aspiration (EUS‑FNA) procedure, to characterise pancreatic cysts and provide additional information to help guide therapeutic decisions. The evidence summarised in this briefing comes from 2 feasibility and 3 pilot studies with a total of 138 adult patients. The diagnostic accuracy for Cellvizio was reported to be between 71% and 87% in 3 studies compared with histopathology, EUS‑FNA or a committee consensus. In 2 studies, images were successfully obtained in all patients. In another study, images were successfully obtained in most (17 of 18) patients. The device was used to identify and validate new diagnostic criteria for pancreatic cyst types in 4 studies. The numbers of safety incidents in 2 Cellvizio studies were higher than reported in a previous EUS‑FNA‑only study. The main capital component of the Cellvizio system costs £79,000 with installation, commissioning and initial training costs of £2,145. Each fibre-optic miniprobe (AQ‑Flex 19) can be used up to 10 times and costs £4,000. All costs are excluding VAT.

Product summary and likely place in therapy

  • Cellvizio is a CLE system with a fibre-optic miniprobe. The system allows imaging of tissues by generating 'optical biopsies', that is, real-time microscopic images of tissue, taken during an endoscopic procedure. These images can be used as well as or instead of information obtained by standard biopsies.

  • The Cellvizio CLE system (see model details below) is designed for use as an adjunct to the standard EUS‑FNA procedure for the characterisation of pancreatic cysts.

Effectiveness and safety

  • Five studies are included in this briefing. Of these, 2 were feasibility studies, and 3 were pilot studies. The studies involved a total of 138 adult patients.

  • In 1 prospective, single-arm feasibility and safety study, images were obtained in 17 of 18 people enrolled. Two people (11%) developed pancreatitis.

  • One prospective, single-arm feasibility study (n=30) found that Cellvizio had a diagnostic accuracy of 87% for detecting potentially malignant pancreatic cysts compared with histopathology or a committee consensus.

  • One prospective, single-arm pilot study (n=57) suggested that epithelial villous structures visualised using Cellvizio were strongly associated with potentially malignant cysts. Cellvizio identified these cysts with a diagnostic accuracy of 71% compared with histopathology or a committee consensus.

  • One prospective, single-arm pilot study in 31 patients demonstrated that Cellvizio could identify benign cysts with an accuracy of 87% compared with histopathology, EUS‑FNA or a committee consensus.

  • A retrospective analysis of a prospective single-arm pilot study (n=33) identified and validated new diagnostic criteria for benign or potentially malignant cysts using Cellvizio.

Technical and patient factors

  • The focus of this briefing is the Cellvizio 100 Series F400‑v2 Laser Scanning Unit (LSU) operating at a wavelength of 488 nm in combination with the AQ‑Flex 19 Confocal Miniprobe. Other Cellvizio models used in research are not considered.

  • The AQ‑Flex 19 is loaded into a 19‑gauge EUS needle and attached by a locking device. This is then passed through the accessory channel of a standard endoscope.

  • Fluorescein dye is used to enhance image contrast.

  • The system is intended for use in secondary or tertiary care units that have the relevant equipment as well as appropriate expertise and experience of performing EUS‑FNA. The intended users are endoscopists trained in the EUS‑FNA procedure.

  • The Cellvizio system can also be used for other clinical applications, each requiring the use of a different type of miniprobe. Other miniprobes and clinical applications are outside the scope of this briefing.

Cost and resource use

  • The main capital components of the Cellvizio system cost £79,000.

  • Installation, in‑servicing and initial training cost £2,145. The optional digital imaging and communications in medicine (DICOM) connectivity module costs £2,730.

  • Each AQ‑Flex 19 costs £4,000. Each miniprobe can be used for up to 10 procedures in multiple patients.

  • All costs stated are excluding VAT.

  • No published evidence on cost consequences and resource use was available.