Louisville Slogger

Sample: pancreatic cancer

Epidemiology: 2% of all cancer diagnosis in the U.S. 33,730 cases (2006): 17,150 men and 16,580 women (10th). Mortality: 32,300 men (4th) and women (5th) Higher incidence in AA population. Median age 72 years. Risk factors: Smoking (RR 2.0), High fat intake, Prior abdominal RT, Chronic Pancreatitis, Occupational exposure to DDT, 2-naphthylamine, benzidine (stone miners, cement workers, gardeners, textile workers). Tropical Pancreatitis. Hereditary: Breast/ovarian cancer (BRCA 2 ), Familial atypical multiple mole melanoma (FAMM), Peutz-Jeghers syndrome, Familial adenomatous polyposis, HNPCC (Lynch II), Hereditary pancreatitis (Sloan's disease), Ataxia telangiectasia, Li-Fraumeni, [Multiple Endocrine Neoplasia (AD) - pancreatic islet cell tumors (60% non-Beta cell tumors and most secrete gastrin ? Zollinger Ellison syndrome, parathyroid adenoma/hyperplasia, pituitary adenoma], Von-Hippel Lindau Syndrome (AD) pancreatic cysts, retinal angiomas, CNS hemangioblastomas in the cerebellum and spinal cord, renal cell carcinoma, pheochromocytoma, islet cell pancreatic tumor, epididymal cystadenoma. Molecular: oncogenes K-ras is most frequently activated in pancreatic adenocarcinoma (90%). p16 CDK inhibitory protein (tumor suppressor) on chromosome 9p (inactivated 90%), p53 (70%). Peripancreatic carcinoma survival: duodenum > ampulla > bile duct > pancreas.

Anatomy: Located at L1-2. Main duct (Wirsung) drains into second part of duodenum with common bile duct through ampulla of Vater. Accessory duct (Santorini) drains upper part of head and drains into duodenum above main duct or communicates with main duct. 2/3 of pancreatic tumors are at the head which is to the right of the SMV, body tumors are between the SMV and left border of the aorta and tail tumors are to the left of the aorta. LN - pancreaticoduodenal, porta hepatis, celiac, suprapancreatic. Body/Tail: lateral suprapancreatic, splenic, hilar.

Natural history: Patterns of spread: locally invasive, early mets to regional LNs and liver, peritoneum, lung. body/tail cancers more commonly present with peritoneal implants. 25% of adenomas on the papilla contain carcinoma.

Pathology: 80% are acinar cells in the pancreas and 20% are islet cells. Exocrine cancer: comprises 95% of all cases (only 2% benign). Adenocarcinoma: 80% of all cases. rare: squamous cell, giant cell carcinoma (5%), carcinosarcoma, cystadenocarcinoma (mucinous <2% ), papillary cystic carcinoma, sarcoma, lymphoma, small cell (early hematologic mets, managed like SCLC), pancreaticoblastoma (peds). Variable malignant potential: mucinous cystadenoma, intraductal papillary mucinous, cystic and solid papillary neoplasm (Hamoudi tumor, women in 3rd decade, present with distention & discomfort, good prognosis for resected tumors). Endocrine cancer: comprises 5% of all cases (most endocrine tumors are benign). Malignant subtypes: insulinoma, glucagonoma, gastrinoma, and islet cell carcinoma.

Presentation: Classic Triad - pain (85%), jaundice, weight loss. Glucose intolerance (10-15%) [60-80% diagnosed with DM in the 2 years prior to diagnosis]. Courvoisier's sign = palpable gallbladder (25% of patients), Trousseau's syndrome = migratory superficial phlebitis, Sister Mary Joseph's node = periumbilical mass. Peri-ampullary tumors - obstructive jaundice: pruritus, dark urine, clay colored stools, and gastric outlet obstruction. Body/tail: less symptomatic until later, tumors are larger. Physical: abdominal mass/ascites (20%), Corvoisier sign (non-tender palpable gallbladder), Virchow's node (left supraclavicular node).

Work-up: CBC, electrolytes, renal function, LFTs, Alk. Phos, Bili, amylase, lipase, CA 19-9 [if (-) FNA > 200-400 U/ml diagnostic] & to follow, Although not effective as a screening tool [elevated in pancreatitis, diabetes, obstruction of pancraticobiliary tree]. 10 to 15% of pancreatic tumors do not secrete CA 19-9, CEA. [EGD + EUS + FNA], ERCP, CXR, CT (abdomen) +/- MRA to assess for invasion or encasement of SMA or celiac axis. Tissue Dx (ERCP, laparoscopy, laparotomy); biopsy not mandatory if operable disease. Biliary obstruction can be relieved via ERCP, percutaneous or transhepatic cholangiography. MRCP effective to visualize biliary tree with no ductal contrast needed. If patients presents with jaundice, abdominal u/s can be 1st sensitivity 80%, specificity 95%. CT better sensitivity 90% and same specificity. For resectability should get a helical CT with dual phase IV contrast and oral contrast.

Prognostic factors: local extent / resectability, size of resected tumor (< 0.5 cm vs 0.5-2 cm vs > 2 cm), LN mets, distant mets, R0 resection is the most important prognostic factor as it is the only potential cure for this disease. R1 or R2 resection results in MS of 6-12 months which is similar to unresectable nonmetastatic disease. A decreasing CA 19-9 correlates with survival after surgery or chemotherapy.

Management: Resectable patients: Whipple resection is the gold standard for lesions of the pancreatic head. However, only between 10-20 % patients present as resectable disease (usually due to late onset of symptoms).What makes a patient unresectable?: a) stage III or IV disease, b) absence of vascular flow through the SMV-portal vein confluence (partial venous involvement of the SMV-PV is considered potentially resectable, c) direct involvement of IVC or aorta. At the tail, unresectable is distant mets, rib/vertebral invasion, SMA/Celiac encasement. Borderline resectable - adrenal, colon/mesocolon, kidney invasion, or biopsy proven peripancreatic lymph node (NCCN 2B recc). Of the 80% that are unresectable ½ present with advanced locoregional disease precluding resection (median survival 6-11 months) and ½ have metastatic disease (median survival 3-6 months). Potentially curative resection results in locoregional relapse in the pancreatic bed of 50-85%. Liver and intraabdominal metastasis is <15% without locoregional recurrence. Palliative surgery: Symptoms that may require intervention include obstruction (duodenal/biliary), hemorrhage, and pain. If a laparotomy is done and the patient is found to be unresectable, then a double bypass procedure (gastrojejunostomy, and either hapatico or choledocojejunostomy) is possible. Celiac plexus neurolysis to reduce abdominal and back pain also can be done. For resectable cases, retrospective data demonstrates about 3% perioperative mortality and OS₅ from 4-24% with R0 "curative" resection. Body/tail lesions - distal pancreatectomy

Adjuvant Chemoradiation
GITSG (Kalser, Arch Surg 120: 899-903, 1985) - 43 analyzable patients; closed early due to poor accrual. Randomized patients after curative resection (negative margins) to observation vs. chemoRT. Concurrent Chemo = 5-FU IV bolus 500 mg/m2/day x 3 days with each XRT course Maintenance Chemo: 4 weeks post-RT, given weekly for 2 years or until tumor progression RT = 40 Gy concurrent XBRT (split course: 20 Gy/10 → 2 week break → 20 Gy/10) APPA, <400cm2, Cover the entire pancreas and pancreatic bed, including celiac axis, pancreaticosplenic, peripancreatic, retroperitoneal Median survival time = 21 mos vs. 11 mos for surgery alone, p=0.03, 2 yr. survival (42% vs. 15%), LR: 30-50%, liver mets 50-66%
GITSG Confirmatory (Cancer 59: 2006-10, 1987) Additional 30 non-randomized patients; MS 18 months; OS₂ 43% OS₅ 17%.

EORTC 40891 (Klinkenbijl, Annals of Surgery 230(6): 776-784, 1999), n=218 Pancreatic head [n=114] (T1-2N0-1a) or periampullary Ca (T1-3N0-1a)
Chemo: 5-FU PVI 25 mg/kg/d d1-5 concurrent with RT (depending on toxicity during phase I), no maintenance chemo after RT. RT = 40 Gy concurrent XBRT (split course: 20 Gy/10 → 2 week break → 20 Gy/10). Median Survival 24.5 mos vs.19 mos, p=0.2, 2 yr OS 41% vs. 51%. 2 yr OS pancreatic head 26% vs. 34% p=0.099. 2 yr OS periampullary 63% vs. 67% p=0.74. 20% of patients in chemoRT arm did not receive treatment.

ESPAC-1 (Neoptolemos, Lancet 358:1576-85, 2001), n=530 (Update - Neoptolemos, N Engl J Med. 2004 350: 1200-10). 2x2 Randomization (chemo v. radiotherapy v. chemoradiotherapy v. observation), median f/u 9mos. Testing postop chemo (weekly 5-FU 425mg/m2 +folinic acid 20mg/m2 for 6 cycles) and chemoRT (40 Gy split course + 500mg 5-FU IV bolus during 1st 3 days of RT). No benefit for chemoRT (14 mos) vs obs (15.7mos), p=ns. Survival benefit for some pts having chemo (19.5 mos) vs observation 13.5 mos, p=0.003. Caveats: selection bias, no RT quality control, positive margins (20%), 1/3 of observation and chemo only patients received RT.

RTOG 9704 (Regine, J Clin Oncol 24:18S, 2006 (suppl) - abstract) - s/p GTR. n=538 (head 381) Arm 1. 5-FU 250mg/m2 daily- 3 weeks before EBRT and then for 12 weeks after vs. Arm 2. Gemzar 1000mg/m2 weekly for 3 weeks before & then 12 weeks after. 50.4 Gy/1.8Gy fx with 5-FU 250mg/m2/day. Median survival = 18.8 months OS₃ = 31% - gemcitabine arm versus 16.7 months and 21% P=.047, respectively, for the 5-FU arm [results are for head tumors]. No significant difference in survival with body/tail tumors only. Large proportion of patients in the non-Gemzar arm salvage Gemzar. More T3/T4 tumors in Gemzar arm. Grade 4 hematologic toxicity higher in the Gemzar arm (14% v 2%, p=s). Prodrug - phosphorylated.- S phase nucleoside analog incorporates into DNA radiosensitizer-(remains 24-48 hours after drug given).

Neoadjuvant ChemoRT: These studies are reported in the Gunderson textbook.
MDACC n=142. 91 patients with neoadjuvant after histologic identification of tumor and 51 laparotomy due to no pathology. 26% in neoadjuvant group were found to have progressive disease on restaging following chemoRT. Remaining 67 patients, 52 underwent successful whipple and 15 (22%) were found to have unresectable disease. RT = 50.4 Gy or 30Gy/10fx. No difference in OS. LRR 21% postop vs. 10% in preop group.

Fox Chase- Pingpank et. al. J. Gastrointest. Surg 2001. n=100. 53% preop ChemoRT & 47% postop ChemoRT. Preop - 7.5% >1 + margin. 44.7% postop group > 1+ margin. 50% of preop 0+ margin compared to 26% in the postop group. Decreased survival associated with + margins.

ECOG 1988: 53 patients treated with preop using Fox Chase regimen. 12 did not make it to surgery. 17 patients did not have a curative resection (11 hepatic/peritoneal mets, 6 locally unresectable) 24 patients undergoing resection MS 15.7 months.

Locally Advanced

GITSG (Moertel, et al. Cancer. 48:1705-1710, 1981). 194 pt with unresectable locally advanced pancreatic adenoCA without distant mets
randomization:(1) RT 60 Gy alone (n=25), RT 40 Gy + 5-FU (n=83), RT 60 Gy + 5-FU (n=86). Split course RT to 20Gy/2 Gy fx. with 2 week break to 40 Gy (AP-PA to entire pancreas). Then boost to 60 Gy (tumor only).

French FFCD-SFRO Phase III (ASCO 2006)- unresectable. ARM 1. 5-FU 300mg/m2/day continuous infusion/Cisplatin 20mg/m2/day + 60 Gy/2Gy fx. ? Gemzar 1000mg/m2 weekly for 3 out of four weeks until progression or limiting toxicity vs. Arm II. Gemzar alone 1000mg/m2 weekly for 7 out of 8 weeks ? Gemzar maintenance same as Arm 1. Results: Increased hematologic and GI toxicity in ARM 1. Gemzar better. Median survival 8.4 months vs. 14.2 months (p=.014). OS1 year 24% vs. 51%

CONKO-001 JAMA 2007. n=368. M0 patients. Unresectable. Arm 1 (186 pts). 6 cycles of Gemzar 3 out of 4 weeks or observation. >80% with R0 resection. DFS 13.4 vs. 6.9 months. MS 22.1 vs. 20.2 months (no diff). Almost all patients received Gemzar at relapse.

Head of pancreas tumor: Include pancreaticoduodenal, porta hepatis, celiac and suprapancreatic nodes. Include entire duodenal loop with margin as the lesion may invade medial wall of duodenum, placing entire circumference at risk. Often ? 50% R kidney in field due to duodenal inclusion, so have to exclude at least 2/3 of L kidney on APPA. Superior field border: T10/11 or mid T11 (for adequate margins on celiac vessels (T12/L1)). Inferior field border: to include PAN (L3/4) and/or to give 3 cm margin on GTV, whichever is lower. Right edge: cover preop pancreatic duodena by 2-3 cm. Left edge: margin 2-3 cm on preop tumor extent or 2cm from left edge of vertebral body, which ever most lateral. On lateral fields: ** must limit contribution via these fields to 18 - 20 Gy b/c kidney/liver in field**. Anterior field border 1.5 - 2 cm ant to CTV, at least 3.5-4 cm ant to vertebral body. Posterior field border split vertebral body in half (to allow good margin on PAN LN). Body or tail of pancreas tumor. LN at risk: pancreaticoduodenal, porta hepatis, lateral suprapancreatic and splenic hilum LN. Often ? 50% L kidney in field due to inclusion of above LN

IORT - Tumor bed + retroperitoneum. Borders - superior-distal extent of transected bile duct, lateral - medial edge of right kidney, medial - medial edge of pancreatic remnant, inferior - to L3. Dose: 10-20 Gy (higher doses cause duodenal perforation/bleeding). RTOG study (n=51)- major complication rate 12% with death in 2 patients. No improvement in MS. Local control improved @ Mayo & MGH but high degree of abdominal relapse.

Ongoing Trials: ESPAC III - Patients having adenocarcinoma of pancreas s/p curative resection. Arm 1. 5-FU/LV 425mg/m2 bolus/20 mg/m2), 5 days every 28 days for 6 cycles Arm 2. Gemcitabine 1000 mg/m2 weekly X 3 weeks then 1 week rest for six cycles, Observation.

ACOSOG: Multi-institutional Phase II: Postop 50.4Gy with 5-FU 175mg/m2 continuous infusion days 1-38, cisplatin 30mg/m2 IV weekly, Interferon-alpha (3 million units SQ every other day) ? Continuous infusion 5-FU for 2 cycles. Primary endpoints will be OS, toxicity, locoregional disease and distant metastases. Trial is based on Virginia Mason study of 33 resected patients demonstrating OS2 year of 84%. Toxicity much higher with Interferon-alpha
Closed trials: RTOG PA-0020. Phase II Arm1: Gemzar 75mg/m2 weekly & Taxol 40mg/m2 weekly with 50.4 Gy. Arm 2. Same as ARM 1 ? Farnesyl Transferase Inhibitor R115777 3-8 weeks after last RT dose (blocks K-Ras post-translational modification for binding to cell membrane)

RTOG 0411 Phase II Xeloda (BID on RT days) & Avastin (days 1, 15, 29) with 50.4 Gy.