Overview

Conventionally fractionated chemoradiation for pancreatic cancer. The target is small but sits against the duodenum and stomach and moves with breathing, so most of the work is in the simulation — fasting, contrast timing, motion management and a careful image fusion — and in keeping the bowel below its limits.

Start here: I’m planning my first GI case — a step-by-step learning roadmap.

Schedules

Conventionally fractionated chemoradiation is used for unresectable (locally advanced) and borderline resectable pancreatic cancer, usually after a period of chemotherapy. One protocol gives 50.4–54 Gy in 28–30 fractions for unresectable disease and 45–50.4 Gy in 25–28 fractions for borderline resectable disease, in 1.8 Gy fractions. The role of radiotherapy is still debated, and practice varies between centers.

Shorter and dose-escalated approaches — five-fraction SBRT, and 15- and 25-fraction regimens — are increasingly used, often with MRI guidance or adaptive planning. The NRG consensus atlas covers their target volumes. This page covers conventional fractionation; a pancreas SBRT page is planned.

Simulation & Setup

The pancreas sits among bowel, stomach and vessels that move with breathing and change with every meal. Most of the setup is about making that anatomy the same each day.

Position
An arms-up setup in a custom bag keeps the beam paths clear and the position reproducible.

What is it?

Supine with the arms above the head, in a vacuum bag or on a wingboard.

Why does it matter?

Arms up keeps them out of the lateral and oblique beam paths through the upper abdomen, and a custom bag helps the patient return to the same position.

Fasting and a consistent drink
A consistent fasting and drinking routine keeps the stomach and duodenum where they were planned.

What is it?

No food or drink for about two hours before simulation and each treatment, then a small, consistent amount of water (or dilute oral contrast at simulation) just beforehand.

Why does it matter?

Stomach filling moves the stomach and duodenum — the organs that usually limit the plan. The same routine every day keeps them close to where they were planned. A small drink also helps show the duodenum on the planning CT.

IV contrast timing
Timed IV contrast makes visible a tumor that is often barely seen on a non-contrast CT.

What is it?

IV contrast for CT-based planning, timed to show the tumor. The NRG atlas suggests considering a triple-phase scan — arterial, pancreatic and portal venous — or, if only one contrast scan is possible, a single scan at about 40–50 seconds.

Why does it matter?

Pancreatic tumors are often barely visible on a non-contrast CT; they show best as a low-density area during the pancreatic phase. Contrast also shows the vessels the target is built around.

Motion management
The pancreas moves with breathing, so motion is assessed in every patient and controlled or tracked when it is large.

What is it?

Breath-hold CT — at deep inspiration or deep expiration — or a 4DCT when breath hold isn't feasible. Fiducial markers are often placed, and a biliary stent can serve as a surrogate.

Why does it matter?

The pancreas moves with the diaphragm. The NRG atlas recommends assessing motion in every patient and managing it when the target moves more than 5 mm. Fiducials give something visible to match to on daily imaging; stents move more from day to day, which the atlas suggests allowing for with larger margins.

Image fusion — diagnostic CT, MRI and PET
Diagnostic multiphase CT, MRI and PET used alongside the planning CT.

What is it?

Image fusion aligns an outside scan with the planning CT so the target can be drawn using information the CT doesn't show. For pancreatic cancer the outside scans are the diagnostic multiphase (pancreas-protocol) CT, MRI and PET, registered rigidly to the planning CT.

Why does it matter?

Fusion can be difficult here: the diagnostic scans are taken in a different breathing state, with different stomach and bowel filling. Check the registration locally around the tumor rather than on the spine, and use the fused images to guide the contours rather than define them — especially when the planning scan is a breath-hold CT without multiphase contrast.

Daily imaging
Daily cone-beam CT is matched to the target rather than the spine, and shows where the duodenum and stomach are that day.

What is it?

Cone-beam CT, recommended daily, matched to bone and then to fiducials, a stent or soft tissue.

Why does it matter?

The spine is a poor surrogate for a target that moves with breathing and filling. The match that matters is the one on the target, and the daily scan also shows where the duodenum and stomach are that day.

Target Delineation

What each volume represents. Expansions come only from your protocol.

GTV — Primary and involved nodes

Definition

The primary tumor and any involved nodes, drawn on the contrast planning CT with help from the fused diagnostic scans. In conventional-fractionation protocols the whole pancreas and the uninvolved draining nodes are not included.

Why it matters

Treating uninvolved nodes or pancreas has not been shown to help and adds toxicity, so the target stays close to the visible disease.

Contouring resources

Organ contouring guides →

ITV — Motion

Definition

With 4DCT, the tumor across the breathing cycle, drawn on each phase or on a maximum-intensity projection and checked in cine mode. Concept page →

Why it matters

Accounts for breathing motion when the patient can't hold their breath. With breath hold the remaining motion is smaller; whether an ITV is still used depends on the protocol.

Contouring resources

Organ contouring guides →

CTV — Microscopic spread

Definition

A small expansion of the GTV or ITV in conventional protocols. Dose-escalated approaches also include tissue around the nearby arteries, as described in the NRG atlas.

Why it matters

Pancreatic cancer spreads along nerves around the celiac axis and superior mesenteric artery, which is why vessel-based volumes appear in newer guidelines.

Contouring resources

Organ contouring guides →

PTV — Planning target volume

Definition

The CTV expanded for setup uncertainty and any motion not already included.

Why it matters

The PTV usually touches the duodenum or stomach. The protocol's priorities decide whether coverage or the bowel limit gives way where they overlap.

Contouring resources

Organ contouring guides →

Organs at Risk & Dose Constraints

Click an organ for its anatomy, why it matters and contouring resources. Superscripts link to the study behind each value.

Organ50.4 Gy / 28 fx45–54 Gy / 25–30 fx
DuodenumAvoid hotspots. The D15% limit is often hard to meet and is weighed against coverage.
  • V55 Gy <15 cc1
  • D15% <45 Gy2
Small BowelAvoid hotspots. Each limit applies only to the contouring method it names.
  • Loops V15 Gy <120 cc3
  • Bowel cavity V45 Gy <195 cc3
StomachAvoid hotspots.
  • Whole organ D100% <45 Gy3
KidneysIf one kidney's mean exceeds 18 Gy, keep the other kidney's V6 Gy under 30%.
  • Mean <15–18 Gy (both kidneys)3
  • V20 Gy <32%3
LiverAssumes normal liver function; a mean under 20 Gy should be achievable.
  • Mean <28 Gy2
  • V30 Gy <30%2
Spinal Cord
  • Max ≤45 Gy2
Heart / Pericardium
  • V40 Gy <30%4,5
  • The QUANTEC stomach value is a whole-organ limit for ulceration (risk <7%); the source protocol itself asks only to avoid hotspots in the stomach.
  • The source also lists kidney V12 Gy <55%, V23 Gy <30% and V28 Gy <20% (QUANTEC).
References (5)
  1. Verma 2014 — Dosimetric predictors of duodenal toxicity after intensity modulated radiation therapy for treatment of the para-aortic nodes in gynecologic cancer. Int J Radiat Oncol Biol Phys. 2014;88(2):357–362 Source
  2. eviQ 3581 — Pancreas definitive/neoadjuvant EBRT chemoradiation (eviQ ID 3581). eviQ Radiation Oncology protocol 3581 Source
  3. QUANTEC summary — Use of normal tissue complication probability models in the clinic. Int J Radiat Oncol Biol Phys. 2010;76(3 Suppl):S10–S19 Source
  4. CROSS — Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med. 2012;366(22):2074–2084 Source
  5. Leong 2005 — 3D conformal radiotherapy for gastric cancer — results of a comparative planning study. Radiother Oncol. 2005;74(3):301–306 Source

Values quoted from eviQ 3581 (last reviewed 14 June 2024; review due 30 June 2028) and QUANTEC summary, checked by DoseLab on 2026-09-29. Verify against your protocol, institutional policy and physician directive before clinical use.

Search all GI constraints in the Constraint Finder →

Protocols & Literature

Contouring atlasSource Verified

International NRG consensus atlas for target volumes in dose-escalated pancreatic cancer radiotherapy. Besides the contouring steps, it gives practical simulation advice: IV contrast for CT-based plans with a triple-phase protocol considered (arterial, pancreatic and portal venous at 25, 40 and 70 seconds), a small oral volume on the table, motion management for more than 5 mm of motion, fiducials, and fusion with diagnostic MRI or PET to aid contouring. It also tabulates suggested organ metrics for its dose-escalated schedules. Citation checked against Europe PMC.

Clinical resourceSource Verified
NRG Oncology Protocol Development Template for Radiation Therapy — Gastrointestinal Cancer: Pancreatic Cancer
Yock A, Parker W · NRG Oncology Medical Physics Subcommittee / Radiation Oncology Committee · 2021

The template NRG Oncology gives investigators for the radiation therapy section of a pancreatic cancer trial. It asks patients not to eat or drink for at least two hours before simulation and treatment and then drink a consistent volume, allows breath hold or fiducial-based gating or tracking, and recommends rigid registration of MRI and PET with the planning CT.

Consensus guidelineSource Verified
Use of normal tissue complication probability models in the clinic
Marks LB, Yorke ED, Jackson A, et al. · QUANTEC · 2010

The QUANTEC overview paper, whose summary table gathers the approximate dose/volume/outcome data from each organ chapter. Its footnotes stress that the data are estimates from the literature, should be applied with caution, and largely do not reflect modern IMRT. DoseLab read the esophagus row (mean dose <34 Gy for grade ≥3 acute esophagitis, 5–20%, 3D-CRT) from the open-access manuscript, and the lung, heart, liver, kidney, stomach and small bowel rows used on the GI pages from the journal's full-text table.

AAPM reportSource Verified

AAPM task group report on respiratory motion in thoracic, abdominal and pelvic treatment. It groups management approaches into motion-encompassing methods, respiratory gating, breath-hold techniques, forced shallow breathing (such as abdominal compression) and respiration-synchronized (tracking) techniques. Read in the AAPM report PDF (Report No. 91, July 2006): using ICRU Report 62 nomenclature, the PTV adds margins to the CTV for respiratory (intrafraction) motion, interfraction motion and set-up error; without a respiratory management device the entire range of motion should be considered in the internal margin; and for lung tumors a maximum intensity projection (MIP) can give the motion-encompassing volume, provided there is no mediastinal involvement.

AAPM reportUnder Review

AAPM Task Group 132 report on image registration and fusion in radiotherapy. It reviews rigid and deformable registration, points out that every registration carries an uncertainty that should be understood and communicated, and gives recommendations for commissioning and quality assurance of registration software and for checking individual registrations. Citation checked against Europe PMC.

Clinical resourceSource Verified

Endorsed eviQ protocol for conventionally fractionated chemoradiation of unresectable or borderline resectable pancreatic cancer. It recommends breath-hold CT (or 4DCT when breath hold isn't feasible), IV contrast, optional fiducial markers, and PET, CT or MRI to aid tumor delineation — noting that adequate image fusion may be difficult. The target is the gross tumor and involved nodes only. Its organ table covers the bowel, duodenum, stomach, heart, kidneys, liver and spinal cord.

All GI literature →