SpyGlass DS peroral cholangioscopy — direct optical vision guiding lithotripsy of difficult bile duct stones and targeted stricture biopsy
Roughly 85–90% of common bile duct (CBD) stones are cleared with conventional ERCP extraction using balloon or basket sweeps, with mechanical lithotripsy as a backup for larger stones. The remaining 10–15% are "difficult stones" — large, impacted, barrel-shaped, intrahepatic, or situated above a stricture — that resist standard tools. A parallel problem exists for biliary strictures: standard blind brush cytology and forceps biopsy carry a sensitivity of only 40–60% for malignancy, leaving many patients in diagnostic limbo. Peroral cholangioscopy with the SpyGlass DS system was developed to solve both problems by putting a camera and working channel directly inside the duct.
A stone becomes "difficult" when its size, shape, or position exceeds the mechanical capability of standard retrieval tools. Recognized risk factors include: stone diameter greater than 15mm (and especially above 25–28mm), a barrel or square cross-sectional shape that resists basket capture, impaction against the ductal wall, intrahepatic location out of reach of a straight basket, a stricture distal to the stone that prevents passage, and altered surgical anatomy (Billroth II gastrectomy, Roux-en-Y bypass) that limits scope positioning and elevator control.
When standard balloon/basket sweeps fail, the next line is mechanical lithotripsy — a basket that captures the stone and crushes it against a metal sheath by hand-crank force. This resolves many remaining cases, but a meaningful subset still fails, or the basket itself becomes impacted around the stone ("basket entrapment"), historically a surgical emergency. Extracorporeal shockwave lithotripsy (ESWL) and peroral cholangioscopy-guided lithotripsy (EHL or laser, delivered under direct vision) are the two endoscopic salvage options recommended before resorting to surgical bile duct exploration.
Distinguishing malignant strictures (cholangiocarcinoma, pancreatic adenocarcinoma invading the duct, metastatic nodal compression) from benign ones (primary sclerosing cholangitis, IgG4-related sclerosing cholangitis, post-surgical or post-inflammatory fibrosis, chronic pancreatitis) has enormous management consequences — one pathway leads to major hepatobiliary resection, the other to surveillance or medical therapy.
Standard ERCP tissue sampling — cytology brushing across the stricture plus intraductal forceps biopsy — has long been recognized as insensitive, typically in the 40–60% range for malignancy. The reasons are structural: cholangiocarcinoma is frequently densely desmoplastic and scirrhous, shedding relatively few malignant cells for the brush to collect, and the sampling is performed blindly over fluoroscopy without the operator ever actually seeing the lesion. Non-diagnostic results often force repeat ERCPs, delaying diagnosis and, in resectable cases, delaying surgery. Direct cholangioscopy addresses this by letting the endoscopist see the stricture and choose exactly where to sample.
In published series, adding cholangioscopy to the diagnostic workup of an indeterminate stricture changes clinical management in an estimated 60–75% of cases — either by confirming malignancy without further non-diagnostic ERCPs, or by providing sufficiently reassuring visual and histologic findings to avoid an unnecessary major resection.
The SpyGlass Discover / SpyGlass DS platform (Boston Scientific) replaced the original 2007-era fiberoptic SpyGlass system, which required a second operator simply to hold and torque the scope while the primary endoscopist manipulated the duodenoscope. The DS generation integrates a digital sensor at the catheter tip itself, eliminating the fragile fiber-optic image bundle, enabling true single-operator control, and delivering a substantially sharper, wider field of view directly to the endoscopy monitor alongside the fluoroscopic image.
The legacy SpyGlass system (introduced 2007) transmitted its image through a coherent fiberoptic bundle threaded down the catheter — a design that degraded resolution with every reprocessing cycle as individual fibers fractured, and that required a dedicated second operator to hold the SpyScope stable while the primary endoscopist worked the duodenoscope and accessories with both hands.
SpyGlass DS (2015) and the current DS II moved the digital imaging sensor to the catheter tip itself, sending an electronic signal up the shaft instead of an optical fiber bundle. This single change removed the resolution degradation problem, widened the field of view to roughly 70°, and — critically — allowed one endoscopist to control both the duodenoscope's elevator and the cholangioscope's steering wheel with the same two hands, since the catheter no longer needed a second person's grip to stay stable in the duct. The catheter is single-use and disposable, removing the cross-contamination risk that dogged the reusable fiberoptic scopes.
After standard biliary cannulation and sphincterotomy, the SpyGlass catheter is passed down the duodenoscope's 4.2mm therapeutic working channel and advanced into the bile duct under fluoroscopic guidance, exactly as an accessory would be. Once its tip emerges past the duodenoscope elevator, the operator switches to a dedicated steering handle controlling four-way tip deflection, allowing active negotiation of the cystic duct takeoff, the hilar bifurcation, and acute intrahepatic angles that a straight catheter cannot reach.
A dedicated irrigation channel — separate from the duodenoscope's own accessory/contrast channel — continuously flushes saline to clear blood, bile, and debris from the optical field, since even a thin film of bile is enough to blind a 2–3mm sensor. The endoscopist now works from two simultaneous displays: the fluoroscopic silhouette of the duct system for global orientation, and the direct-vision monitor for the close-up optical view — a combined technique that neither imaging modality alone can replicate.
Once the SpyGlass tip is positioned above the target, the endoscopist obtains something fluoroscopy can never provide: an actual picture of the stone's surface and the ductal mucosa surrounding it. Stone color and texture hint at composition and hardness, informing how much lithotripsy energy will likely be required, while ductal wall appearance — smooth versus irregular, pale versus friable, normal versus tortuous vasculature — is the single most useful real-time clue to whether a stricture is benign or malignant.
Stone appearance under direct light correlates loosely but usefully with composition and hardness. Black pigment stones — composed of calcium bilirubinate polymer, associated with hemolytic conditions and cirrhosis — tend to be hard and brittle. Brown pigment stones, which form around bacterial deconjugation of bilirubin during bile stasis or infection, are typically softer and more friable. Cholesterol stones, usually of gallbladder origin and secondarily migrated into the duct, are often yellow and comparatively soft.
Seeing the stone directly lets the operator anticipate roughly how many lithotripsy pulses will be needed and, more importantly, choose a probe trajectory that strikes the stone perpendicular to its face for maximal energy transfer while keeping the ductal wall entirely out of the direct energy path — something that is impossible to guarantee under fluoroscopy alone, where the stone and the duct wall are superimposed shadows.
A well-described set of cholangioscopic visual criteria helps separate malignant from benign strictures in real time: dilated, tortuous, irregular-caliber "tumor vessels"; an irregular, nodular, or ulcerated mucosal surface; and infiltrative, friable tissue that bleeds on contact all favor malignancy. In contrast, a smooth, symmetric, pale stricture with a regular, unremarkable vascular pattern favors a benign process such as primary sclerosing cholangitis, post-surgical fibrosis, or stone-related inflammatory narrowing.
Pooled data from meta-analyses of cholangioscopic visual impression alone report roughly 90% sensitivity and 87% specificity for malignancy — a substantial jump over blind brush cytology's 40–60% sensitivity. Direct visualization also frequently reveals additional stones that a contrast cholangiogram missed entirely: reported rates of stones seen on cholangioscopy but not on the prior fluoroscopic cholangiogram range from roughly 13% to 30%, because small stones or stones layered behind larger ones can be radiographically silent.
Visual impression is a powerful triage tool, but it is not a substitute for tissue diagnosis — false positives and false negatives both occur. Its real clinical value is in directing where the SpyBite forceps should take their biopsies, which is what ultimately raises the combined diagnostic yield to roughly 95%.
With the stone in clear view, a thin probe is passed through the SpyGlass working channel to deliver either electrohydraulic (EHL) or holmium:YAG laser energy directly onto the stone surface. Both modalities work by generating a rapid, localized mechanical shockwave in the saline-filled duct; both require the probe tip to be positioned in near-contact with the stone under continuous direct vision, because the same energy that fractures a stone will just as readily injure or perforate the ductal wall if misdirected.
An EHL probe is a bipolar electrode, typically 1.9–4.5F in diameter, that is advanced through the working channel until its tip is nearly touching the stone, fully submerged in the irrigating saline. A high-voltage discharge (on the order of 100V) across the spark gap instantaneously vaporizes a small volume of the surrounding fluid, creating a plasma bubble that expands and collapses within microseconds. That rapid expansion/collapse cycle generates a mechanical shockwave that propagates through the aqueous medium and fractures the stone by tensile and shear stress at its surface, rather than by heat.
Because the shockwave is generated in the fluid itself, EHL is critically dependent on two conditions: the field must be fully irrigated (an air-filled duct blunts the shockwave and risks arcing against the endoscope), and the probe tip must be aimed directly at the stone, in close or actual contact, under continuous visual confirmation. Fired blindly, or aimed even slightly off-target, the same shockwave that fragments a stone can perforate the thin ductal wall — which is precisely why EHL delivered through a cholangioscope, under direct vision, is considered dramatically safer than the earlier blind fluoroscopic-only EHL techniques.
The alternative modality is a solid-state holmium:YAG laser, delivering pulsed infrared energy at 2100nm through a flexible quartz fiber passed down the same 1.2mm working channel. This wavelength sits almost exactly at a peak of water's absorption spectrum, so the energy is absorbed within roughly 0.4mm of the fiber tip — largely by the interstitial water within and immediately around the stone rather than by deeper tissue.
That absorbed energy rapidly vaporizes water at the fiber tip, forming an expanding vapor bubble whose collapse generates a photoacoustic shockwave that mechanically cracks the stone, while direct surface photothermal ablation contributes an additional, more gradual erosive effect. Typical clinical settings run roughly 0.5–1.0 Joules per pulse at 5–10Hz (about 4–10 Watts average power). Comparative series report broadly similar stone clearance rates between EHL and laser lithotripsy, with laser's shorter, more predictable penetration depth sometimes cited as a theoretical safety advantage for ductal wall sparing — though both require the same fundamental discipline of direct-vision, near-contact firing.
Pooled results across multiple large multicenter series — including prospective registries evaluating digital single-operator cholangioscopy for stones that had already failed conventional balloon/basket and mechanical lithotripsy — report complete ductal clearance in a single session in roughly 71–77% of patients, with cumulative clearance climbing to approximately 90–97% within one to two sessions. Success is largely independent of stone size, including stones exceeding 25mm, provided enough lithotripsy sessions are performed.
Adverse event rates run roughly 5–9%, dominated by mild-to-moderate post-ERCP pancreatitis and cholangitis (cited around 1–7% in various series), with direct-vision-guided energy delivery associated with a markedly lower perforation risk (well under 1%) compared with the historical era of blind fluoroscopic EHL or basket-impaction emergencies, some of which previously required emergency surgical intervention.
Cholangioscopy-guided lithotripsy achieves roughly 90% or higher stone clearance in patients for whom standard endoscopic extraction had already failed — converting a population that would otherwise face open or laparoscopic bile duct exploration, with its own 15–20% complication rate in older, comorbid patients, into a population managed successfully through the endoscope.
The final phase converts lithotripsy or visualization into a completed procedure. Stone fragments, generally reduced to pieces under 3–4mm, are swept out of the duct with a retrieval balloon or basket and confirmed clear on a completion cholangiogram. For strictures, SpyBite mini-forceps take biopsies from exactly the tissue the endoscopist just watched behaving suspiciously — a targeted approach that substantially outperforms blind ERCP sampling.
Once lithotripsy has reduced the stone to fragments generally under 3–4mm, a standard extraction balloon is inflated just above the fragments and withdrawn toward the papilla, or a Dormia-type basket is used to actively capture larger remaining pieces, sweeping the duct systematically from the most proximal extent down to the papilla. A completion cholangiogram — a fluoroscopic contrast injection performed at the end of the case — confirms the absence of residual filling defects.
Even after a technically successful lithotripsy and sweep, a small residual fragment rate of roughly 5–10% is typical in reported series, usually addressed either with a second endoscopic session or, when fragments are small enough, by leaving a biliary stent to promote natural passage and prevent obstruction in the interim.
The SpyBite Max mini-forceps (Boston Scientific) pass through the same 1.2mm working channel and are steered, under direct vision, to the exact tissue that looked most suspicious moments earlier — nodular, ulcerated, or friable areas overlying irregular tumor vessels — rather than sampling blindly at the fluoroscopic center of a stricture as conventional ERCP forceps must. Typically four to six biopsy passes are obtained to maximize tissue yield and account for sampling variability across a heterogeneous lesion.
Because each bite captures only a superficial 1–2mm of tissue, deeply infiltrative or densely desmoplastic malignant strictures can still occasionally under-sample. Despite this limitation, multiple published series report cholangioscopy-guided biopsy sensitivity in the 85–90% range for malignancy — roughly double the 40–60% sensitivity of conventional blind brush cytology and forceps biopsy.
Taken together, direct visualization plus targeted SpyBite biopsy pushes combined diagnostic sensitivity for malignant strictures toward roughly 95%, while cholangioscopy-guided lithotripsy resolves stones that had already defeated standard endoscopic therapy in roughly 90% or more of cases within one or two sessions. The net clinical effect is a meaningful reduction in non-diagnostic repeat ERCPs, in unnecessary or delayed surgical referrals, and in open or laparoscopic bile duct exploration for stones — all while keeping the intervention entirely endoscopic and largely outpatient or short-stay.
Across stone and stricture indications alike, the unifying principle of peroral cholangioscopy is the same: replace an inference made from a two-dimensional contrast shadow with a direct look at the tissue itself, then act — fragment, sweep, or biopsy — with the precision that only seeing the target actually provides.