Pancreas SBRTMatthew Spraker, MD, PhDFaculty Mentor: Smith Apisarnthanarax, MDUniversity of Washington/Seattle Cancer Care AllianceSeattle, WAJanuary 19, 2017

Case Presentation 70 year old gentleman presents withrefractory abdominal pain, bloating, and 16 lbweight loss over 1 monthJanuary 19, 2017

Patient History Past Medical/Surgical History– Liver cysts Family History– Father with CAD, Mother DM, no malignancy Social History– Retired English teacher– Occasional EtOH use ( 2 drinks per week)– Non-smoker, no drug useJanuary 19, 2017

Physical ExaminationVitals: Ht 72 cm, Wt 157 lbs, BP 142/88, HR 65,Temp 97.8, SpO2 98% on RAGen: Well-appearing male in NADLungs: Clear with symmetric air movementHeart: RRR, S1S2 normal, no murmursAbd: Soft, non-distended, left-mid tenderSkin: No jaundiceLymph: No cervical, supraclavicular, axillary, oringuinal LAD noted

Laboratory Studies CBC– Within normal limits Chemistries– Within normal limits– Cr 0.95 Liver panel– AST 15, ALT 26, AP 115– Albumin 4.1 Ca 19-9 is 250

CT Scan – Pancreatic phase (ax)Hypoenhancingpancreaticbody massPresumedsplenic a.involvementJanuary 19, 2017

CT Scan – Pancreatic phase (cor)Hypoenhancingpancreatic body massJanuary 19, 2017

Diagnostic work up CT scan with pancreas protocol– 15 mm mass in the mid-body of the pancreas withpancreatic ductal dilation and superior mesentericvein encasement, and apparent involvement ofthe splenic artery posterior to the pancreas. Thetumor abuts the celiac artery. Multiple hepaticcysts are seen diffusely. No hepatic metastasesidentified.– “Borderline resectable” ( 180 contact with CA) EUS – pancreatic body mass FNA showsAdenocarcinoma CT chest – negative

Epidemiology Female Male, peak incidence: 65-80 yo Despite #12 cancer incidence, #4 deaths– 53,070 cases, 41,780 deaths est. in 2016 Risk factors: cigarette smoking, diets high inanimal fat, prior abdominal radiation (i.e.testicular CA), chronic DM, chronicpancreatitis, obesity 10% familial : BRCA2, Peutz-Jeghers, HNPCC,Hereditary pancreatitis, MEN I, VHL, ATJanuary 19, 2017

Diagnostic workup Per NCCN, imaging should include a dedicated pancreaticCT or MRI– CT angiography with thin slices ( 1mm) using dual-phasecontrast (pancreatic and portal phase) is preferred Biopsy via EUS versus CT-guidance– Reduced risk for peritoneal seeding Diagnostic laparoscopy considered in selected patients CA 19-9 at baseline not diagnostic, but may be helpful forresponse assessmentJanuary 19, 2017

Resectability statusResectability statusArterial (celiac, SMA,common hepatic)Venous (SMV, PV)ResectableNo contactNo contact or 180ocontact without veinirregularityBorderline resectable Contact with SMV/PV 180o withoutirregularity Contact with IVCUnresectable Contact with CA/SMA Unresectable due to 180otumor involvement or Contact with first jejunalocclusionbranch of SMA Involvement of firstjejunal branch of SMVCH contact w/o celiacSMA 180o contactContact with CA 180oPresence of variantanatomySee radiology reporting template in NCCN guidelines (adapted from Al-Hawaryet al., Radiology, 2014)January 19, 2017

Treatment paradigm Pancreas SBRT is one of several potentialtreatment paradigms for unresectable orborderline resectable pancreatic cancer– Chemotherapy alone– Definitive chemoradiotherapy– Chemotherapy chemoradiotherapy– Chemotherapy SBRTJanuary 19, 2017

Neoadjuvant therapy Patient underwent 4 cycles of gemcitabineand nab-paclitaxel Response by CT– “Small mass within the mid-body of the pancreasappears less conspicuous. There is decreasedcompromise of the portal venous componentsevidenced by decreased dilation of the splenicvein as well as decreased dilation of thepancreatic duct.” Response by CA 19-9 of 250 27January 19, 2017

Why Pancreas SBRT Local control is important in advanced pancreaticcancer– Pain, bleeding, obstructive jaundice Chemoradiotherapy is arduous––––50.4-59.4 Gy over 5-6 weeksConcurrent 5FU or capecitabineAcute GI toxicities in up to 30%LC improved over chemotherapy alone per LAP-07 SBRT can offer equal or better LC, less acute toxicity,and shorter treatment courseJanuary 19, 2017

Single fraction SBRT First, dose-escalation study of pancreas SBRTby Stanford (Koong et al., IJROBP 2004)– Escalated 15, 20, then 25 Gy x1 fraction– 7 patients treated at 25 Gy with no GI grade 3 orgreater acute toxicity. Single and multi-fraction regimens were latertried, with finding that multi-fractionregimens had less SBRT-related toxicityJanuary 19, 2017

Outcomes for pancreatic SBRTNumberpatientsDoseNumberfractionsLocalcontrol (%)Mediansurvival (mo)Koong et al., 200516251948.25Hoyer et al., 200522453575.7Mahadevan et al., 20103624-3637814.3Schellenberg et al., 2011122519411.8Goyal et al., 20122022-301-38114.4Lin et al., 20152035-4558020Moningi et al., 20158825-3356114.4-18.4Shaib et al., 20161336-453NR11StudyAdapted from Kim et al., J Gastro Onc, 2016All studies on unresectable patients, except Moningi, which included borderline resectable as wellJanuary 19, 2017

Phase II Multicenter Trial Herman, et al., Cancer 2014– Phase II Multi-institutional Study– Johns Hopkins, MSKCC, Stanford Locally advanced, N 49, Median F/U 13.9mos– 3 wks Gemcitabine SBRT Gemcitabine– SBRT 33Gy/5fx, centrally reviewed– 83% local progression free survival, median OS13.9 mos– 2% acute gr 2, 11% late grade 2 toxicityJanuary 19, 2017

Cost effectiveness of pancreas SBRTStanford cost-effectiveness study (Murphy et al., 2012) Compared cost effectiveness of 4 models: Gem aloneGem conventional RTGem IMRTGem SBRT SBRT gem added 0.20 QALY at inc. cost of 13,700compared with gem alone Unlike IMRT, SBRT gem increases the clinicaleffectiveness beyond gem alone at a potentiallyacceptable cost.January 19, 2017

SBRT Simulation Implanted Fiducials– Gold markers implanted endoscopically– Minimum 2, but 3-5 recommended to account forpotential migration or loss– Plan for simulation 5-14 days following implant to allowseeds to “settle” Motion management via expiratory breath hold orabdominal compression Patient immobilization– NPO– Supine in vac-lock, arms up on wing board– IV contrast & small bowel contrastJanuary 19, 2017

Biliary Stent versus Fiducial Markers van der Horst et al. IJROBP 2014– Eleven PA patients with stents and fiducials– Daily CBCT registered to planning CT by boneyanatomy, stent, or fiducials Stent better then bony anatomy in 67% offractions Found that stent-tumor relation was not rigid,with deviations up to 8.4mm Conclusion: fiducials stent bony anatomyJanuary 19, 2017

Treatment Planning Planned with VMAT, minimize tx time Prescription– 25-33Gy in 5 fractions– Consider SIB 35-50Gy to areas of vesselinvolvement– Range of fractionations from daily to BIW/TIWJanuary 19, 2017

Dose Kidneys (combined)V30 0.5 ccV18 5 ccV12.5 10 ccStomachMean dose 10 GySpinal CordV30 0.035 ccV25 0.5 ccSpinal Cord 5 mmV30 0.5 ccV25 5 ccLiverMean dose 15 Gy*These are institutional constraints for 5 fraction treatment created as anamalgam of published constraints (Moffitt, JHU) and those used in aninstitutional trial.January 19, 2017

This patient’s plan 30Gy in 5 fractions with SIB to vessels (40Gy/5 fx) Multiple scans acquired during CT simulation:– If ABC: contrast scan (for volume delineation) andnon-contrast scan (for dosimetry planning)– If free breathing: helical contrast scan (for volumedelineation) and 4DCT (time series for ITVcreation, average for planning)January 19, 2017

Institutional Method for PTV If breath hold: compared contrast-enhancedand non-contrast helical scans, construct PTVbased on set up error between the two scans If free breathing: measure maximum range offiducial markers in each of three planes, usetravel distance/direction to construct PTV– Note this might result in an asymmetric PTV(typically SI size radial size)January 19, 2017

January 19, 2017

January 19, 2017

January 19, 2017

January 19, 2017

January 19, 2017