Radiation Oncology Synopsis

Pancreatic Adenocarcinoma

Epidemiology and Demographics

There are bout 30,000 cases of pancreatic cancer per year in the United States. The disease is highly fatal and is the 10th most common cancer diagnosis but the 4th most common cause of cancer death in the U.S. The incidence is similar among men and women, but there is a predilection toward African descent. The peak age of incidence of pancreatic cancers is in the 6th - 7th decades.

Environment, Natural History and Genetics

The most common environmental risk factors for pancreatic cancers include:

Approximately 5% of pancreatic cancers are familial. Genetic mutations of p16 and BRCA2 are the most common familial genetic abnormalities. The K-ras oncogene is present in about 85% of all pancreatic cancers.

Chronic pancreatitis is not associated with the risk of pancreatic adenocarcinoma. Commonly pancreatic adenocarcinoma arises in the head of the pancreas, next in the body and least commonly in the tail.

  1. 75% Head of Pancreas
  2. 15% Body of Pancreas
  3. 10% Tail of Pancreas

About half of all pancreatic cancers have distant metastases at diagnosis. About 25% have regional nodal metastases at diagnosis. The most common sites of pancreatic cancer metastases include liver, peritoneal surface and lungs.

Periampullary tumors arise frm the ampulla of Vater, the distal common bile duct r the adjacent duodenum.

Pancreas Anatomy

There is no proven role for screening pancreatic cancers There are studies evaluating EUS roles but no proven benefit.

Nearly all pancreatic cancers are exocrine (95%). There are four common pathologic subtypes of adenocarcinoma of the pancreas:

  1. Ductal adenocarcinoma (40%)
  2. Mucinous cystadenocarcinoma
  3. Acinar cell carcinoma
  4. Adenosquamous carcinoma

The primary presenting signs of pancreatic adenocarcinoma are pancreatic or bilary duct obstruction, jaundice and abdominal pain.

Workup and Staging

The differential diagnosis of a pancreatic mass includes

The general workup is H&P, CBC, CMP, CA 19-9, tri-phasic thin sliced CT of the abdomen, chest imaging ± ERCP or EUS with FNA for diagnosis and possibly, stent placement. Patients with Lewis Antigen AB negative cannot secrete CA 19-9. This phenotype is present in about 5 - 10% of the population.

Tissue diagnosis is obtained from ERCP, endoscopic ultrasound guided fine needle aspiration, CT guided FNA or pancreatic resection. Histologic diagnosis is not necessary before resection of a pancreatic mass. ERCP and endoscopic resection (2%) is associated with a lower risk of seeding the peritoneum than CT guided FNA (26%).

CA 19-9

CA 19-9 has been reported as a prognostic indicator. The RTOG 9704 study enrolled 53 patients (14%) with CA 19-9 > 90 U/ml. Only 2 of these patients survived up to 3 years.

Staging and Classification

Pancreatic cancer is a surgical disease with roles for adjuvant and definitive chemotherapy and radiation therapy. The NCCN 2010 classified pancreatic cancers as :

The NCCN criteria for resectable pancreatic adenocarcinoma are:

The NCCN defines borderline resectability as

Pancreatic vessels -- blue SMV

For pancreatic tail lesions, borderline resectable lesions refer to lesions that invade the adrenal gland, colon, mesocolon or kidney. Pancreatic head lesions are frequently more resectable because they become symptomatic earlier in the disease.

Over 80% of pancreatic cancers deemed resectable by CT are resectable at surgery.

AJCC Staging

T1 limited to the pancreas ≤ 2 cm
T2 limited to pancreas > 2 cm
T3 extends beyond pancreas but without celiac or SMA involvement
T4 celiac axis or SMA involvement

N Stage
N0 No nodal metastases
N1 Nodal metastases

AJCC Stage Grouping
T N0 N1 M1
T1 IA IIB IV
T2 IB IIB IV
T3 IIA IIB IV
T4 III III IV

Positive washings at them time of laparoscopy are staged Stage IV (M1).

The AJCC has established a separate staging system for Ampulla of Vater, distal CBD and duodenal carcinomas.

Am pulla of Vater and Bile Duct Staging
T Ampulla of Vater Distal Common Bile Duct Perihilar Bile Ducts
T1 Limited to Amuplla of Vater or sphincter of Oddi Confined to Bile Duct histologically Confined to bile duct with extension up to muscle layer or fibrous tissue
T2a Invades duodenal wall Invades beyond wall of bile duct Invades wall to surrounding adipose tissue
T2b (T2 Only) (T2 Only) invades adjacent hepatic parenchma
T3 Invades pancreas Invades gall bladder, pancreas, duodenum or other adjancet organs without celiac axis or SMA involvement invades unilateral branches of teh portal vein or hepatic artery
T4 Invades peripancreatic soft tissues or other adjacent organs or structures other than pancreas Involves celiac axis, or SMA invades main portal vein or its branches bilaterally;
or the common hepatic artery
or the second order bilary radicals bilateraly

Treatment and Prognosis

Surgery is the primary treatment. Surgery used in head of the pancreas lesions (75% of all cases) involves a classic Whipple procedure. The Whipple procedure is a pancreaticoduodenectomy. Alternatively, a pyloris preserving pancreaticoduodenectomy may be performed.

Whipple Procedure

The Whipple procedure involves resecting the pancreas with three anastomoses:

  1. pancreaticojejunostomy
  2. choledochojejunostomy (hepaticojejunostomy)
  3. gastrojejunostomy

Post operative favorable prognostic indicators include:

Surgery is only beneficial if there is a complete resection (R0). Retrospective evidence suggests that survival is similar with patients who have R1 (microscopic residual) and R2 (gross residual) disease and those who have definitive chemo-radiation therapy. Planned resection should be performed in patients where complete resections (R0) resections are likely. Debulking surgery does not improve outcome over definitive chemo-radiation therapy. Autopsy findings in in 78 pancreatic adenocarcinoma patients demonstrate local recurrence in 78% was a component of failure and hepatic recurrence in 61.5%.

There has been no demonstrated survival benefit demonstrated in performing and extended lymphadenectomy. Resectable patients should not undergo extended retroperitoneal lymphadenectomy. Riall in 2005 published a randomized controlled trial demonstrating no statistically significant differences in survival at 25% and 31% 5 year survival.

Where possible, surgery remains the first line treatment of pancreatic adenocarcinoma. The Japanese Pancreatic Adenocarcinoma Study Group compared surgery alone against definitive chemo-radiation therapy in a randomized controlled trial. This trial, published in 2008 was stopped early due to clear benefit of surgery. Median survival was 12 months in the surgery arm and 9 months in the chemo-radiation arm. Radiation therapy was delivered to 50.4 Gy with ci5FU. Not only was the median survival much worse in the chemo-radiation arm, but the 5 year survival in the chemo-radiation arm was dismal. Surgical 5 year survival rates were not good, but at 10% were better than 0%.

Post resection adjuvant treatment options after pancreaticoduodenectomy include

Post Operative Adjuvant Therapy

Standard and total dose/fractionation after surgical resection uses a dose to regional lymphatics and tumor bed of 45 Gy with a boost to the tumor bed of between 50.4 Gy and 54 Gy (boost 5.4 Gy - 9 Gy). All doses are given at 1.8 Gy/fraction. The standard radiotherapy fields are tumor bed and at-risk regional lymph nodes. Add 1 -2 cm margins for motion and setup error. The boost fields is the tumor bed plus margin.

Studies Favoring Post-operative Chemotherapy and Radiation Therapy

GITSG 91-73 Post-op observation compared to Post-Op radiation with Split Course Radiation (Kalser 1985 Arch. Surg) GITSG 9173 Trial first reported a benefit to adjuvant post-operative concurrent chemotherapy with radiation therapy in 1985. The trial information is:

Most radiation oncologists do not believe that split course radiation is as effective as continuous course treatment.

Mayo Clinic (Corsini 2008) Retrospective review of resectable T1-T3 N0 patients who received adjuvant chemotherapy/radiotherapy or observation.

This study used a more conventional continuous course radiation therapy treatment at 1.8 Gy to 50.4 Gy. 98% of all patients received continuous infusion 5FU with radiation. This cohort was compared with those who were observed with no further treatment. The study showed that chemo-radiation therapy improved median survival, and two and five year overall survival:

Johns Hopkins 2008 Retrospective review of Post-operative Adjuvant Chemo-radiation v. Observation (Herman 2008)

This study looked at 616 patients who were treated with pancreaticoduodenectomy who recieved RT/5FU or observation. Chemotherapy was 5FU based. The study demonstrated improved median survival at 21 months in the treated cohort compared with 14 months in the observed cohort.

SEER Review of 3008 patients receiving re-operative or post-operative radiation therapy or no adjuvant treatment. (Hazard 2007)

This study compared cohorts who either did or did not get radiation therapy treatments. It reported improved survival in patients who had direct extension beyond the pancreas or in node positive patients with adjuvant radiation therapy. .

While not a part of this study, a re-analysis of the SEER data looking at pre-operative v. post-operative radiation therapy in showed improved overall survival in patients with pre-operative radiation therapy.

Johns Hopkins-Mayo Clinic Collaborative Study (Hsu 2009)

This study was another retrospective study of post-operative chemotherapy+radiation compared with observation alone. This study showed a median survival, 2 and 5 year overall survival improvement with chemo/radiation therapy over observation alone.The details:

EORTC 40891 Similar to GITSG 91-73 except no post radiation 5FU (Klinkenbijl, 1999 Ann. Surg.) This European study supported the benefit of post-operative radiation therapy combined with 5FU. The details:

Studies Favoring Chemotherapy without Radiation Therapy

ESPAC-1 Four Arm Trial of Surgery ± CRT or ± Chemotherapy (Neoptolemos, 2001, Lancet) The ESPAC-1 Study enrolled patients with grossly resected pancreatic cancer. It was a 4 arm trial that compared ± post-operative chemotherapy/radiation therapy, and ± chemotherapy alone.

ESPAC-1 Trial (Y J Chua et al, JCO 2005; 23:4532)

The study was intended to be a 2x2 factoral randomization comparing Chemotherapy+radiation therapy to chemotherapy to CRT followed by chemotherapy to observation. The study introduced two further separate randomizations: chemoradiotherapy or observation and chemotherapy or observation to allow the inclusion of patients when either clinician or patients were unwilling to accept the 2x2 randomization. The original study design is indicated with box and the newer arms are indicated with an A.

Here are the key features of the ESPAC 1 Trial:

Criticisms of this Study:

ESPAC Meta-analysis 2009 (Neoptolemas) examined 822 patients who received either adjuvant 5FU/folinic acid or observation after surgical resection. Adjuvant 5FU/FA improved median survival to 23.2 months from 16.8 months.

CONKO-001 Randomized study of Gemcitabine compared with observation alone (Oettle 2007).. This study was a randomized study of Gemcitabine for six cycles compared with observation alone in post-operative pancreatic cancer patients. It demonstrated, adjuvant gemcitabine improved disease free survival to 13.4 months from 6.9 months. No survival benefit was identified. The study excluded patients with a pre-operative CA 19.9 ≥ 2.5 x normal level

Studies of Different Chemotherapy/Radiation Regiments

RTOG 9704/SWOG/ECOG Randomized study of patients with GTR of pancreatic cancer treated with weekly gemcitabine or ci-5FU for 3 weeks before and 12 after concurrent chemo/RT (Ragine, 2008, JAMA) This study enrolled 451 patients with R0/R1 resection of pancreatic cancer. They were randomized to 3 weeks of either gemcitabine or ci-5FU → 5FU+RT (50.4 Gy @ 1.8 Gy/fraction) → 12 weeks of the original chemotherapy.

Neoadjuvant Treatment

There are no completed phase III trials for neoadjuvant treatment.

Stessin reanalyzed the SEER database in 2008 and found 70 patients with pre-operative RT, 1478 patients with post-operative RT and 2337 patients treated with surgery alone. Median survival was 23 months in pre-op RT cohort, 17 months in the post-op RT cohort and 12 months in the surgery alone cohort.

Evans reported in 2008 a Phase II trial of 86 patients randomized to Chemo-RT with radiation given at 30 Gy at 3 Gy/fraction with concurrent chemotherapy of weekly gemcitabine for seven weeks followed by surgery. Radiation Therapy fields included the pancreaticoduodenal, portahepatic, superior mesenteric and celica axis lymph nodes. All patients were re-staged after chemo-radiation therapy. 85% went on to surgery. Median survival was 22.7 months, improved from 7 months for surgery alone and OS5 was 27%. For patients who received surgery median survival was 34% compared with 7 months for unresectable patients.

Unresectable Pancreatic Adenocarcinoma

There are a number of studies examining the treatment of unresectable pancreatic adenocarcinomas. The sentinel study was the GITSG 9273 study (Moertel) published in 1981. This study demonstrated a definitive role in chemo-radiation therapy as definitive treatment in unresectable cases. The FFCD/SFRO French study used non-standard chemo-radiation therapy and reported that CRT was more toxic and had worse survival outcomes. The non-standard regimen was very poorly tolerated.

GITSG 9273 Compared split course radiation therapy (40 Gy) with bolus 5FU to split course RT (60 Gy) with bolus 5FU to RT alone to 60 Gy. This study compared radiation alone, radiation with chemotherapy and dose escalated radiation therapy with chemotherapy. Both concommitant chemotherapy arms prolonged median survival compared with radiation alone.

GERCOR Phase II/III Impact on chemo-radiotherapy with chemotherapy in locally advanced pancreatic cancer (Hugeut, 2007, JCO) This study examined 181 patients with locally advanced pancreatic cancer (unresectable) treated initially with 5FU or gemcitabine who had no evidence of progression after 3 months and then received additional chemotherapy or chemo-radiotherapy (physician's choice). Chemo-radiotherapy improved median progression free survival from 7.4 months to 10.8 months and overall survival from 11.7 months to 15 months.

FFCD/SFRO (French) Study 2 Arm trial of CRT (RT+5FU+CDDP followed by maintenance gemcitabine) compared to gemcitabine alone Cahuffert 2008, Ann Oncology.. This French study compared chemo-radiotherapy against gemcitabine chemotherapy alone. Induction chemo-radiotherapy was more toxic and had worse survival outcomes.

ECOG 4201 Compared Gemcitabine to gemcitabine + radiation (ASCO 2008). This study closed early due to slow accrual. It compared RT + gemcitabine to gemcitabine alone. 71 patients were accrued, Chemo-radiotherapy arm did better in median survival as well as overall survival at 2 years.

Treatment Recommendations

At present there is controversy and various studies have come to differing conclusions and re-analysis of those same studies have come to different conclusions from the original author conclusions.

For resectable disease, R0 resection is indicated, followed by chemotherapy. The use of chemo-radiotherapy is controversial with the GITSG study recommending CRT, the ESPAC-1 (flawed by failure to standardize treatment, failure to complete or initiate planned treatments) failed to show an advantage. EORTC 20891 did, on reanalysis show an advantage to CRT.

Presently, for borderline resectable disease, there is no standard neoadjuvant treatment for pancreatic adenocarcinoma. There is an ongoing University of Michigan trial of neoadjuvant gemcitabine/Radiation, a local protocol at Dartmouth for induction docetaxel/gemcitabine → gemcitabine/RT and an MDACC protocol of gemcitabine+cisplatin → gemcitabine/RT for resectable disease. Radiation therapy is given to 45 - 50.4 Gy in 1.8 - 2 Gy fractions or 30 Gy in 3 Gy fractions. The current treatment paradigms are staging laparoscopy, stent placemnt if jaundice, and neoadjvant chemo/radiation → resection.

Presently for locally advanced pancreatic cancer in the US, the Standard definitive treatment recommendations are for unresectable pancreatic cancer are: CRT continuous infusion 5FU with Radiation therapy to 50 - 60 Gy in 1.8 - 2 Gy fractions or 30 Gy in 3 Gy fractions.

Metastatic Disease, Metastatic disease is treated with gemcitabine alone, gemcitabine with erlotinib or on a clinical trial.

NCIC Study of gemcitabine ± erlotinib (Moore 2007 JCO) demonstrated improved median survival of 5.9 months from 6.2 months and OS1 to 23% from 17%.There may be a role for bevacizumab as well (AViTA Trial).

AViTA Trial compared the addition of bevacizumab (Avastin) to gemcitabine/erlotinib (2008 ASCO). There was better PFS at 4.6 months from 3.6 months but no overall survival difference at 7.1 months and 6 months (not statistically significant) with the addition of avastin.

Radiation Fields and Techniques

Pancreatic cancer was traditionally treated with fairly large fields bearing in mind certain landmarks:

The classic adjuvant radiation fields for pancreatic head disease cover the tumor bed, the pancreaticoduodenal nodes, local suprapancreatic nodes (but not the entire pancreas), celiac nodes, porta hepatis nodes, and the SMA/SMV nodes. For neoadjuvant and unresectable tumors the present trend is toward smaller fields. In the absence of CT/imaging guided planning, the following classical borders are used:

Stereotactic Body Radiotherapy has been the topic of several emerging studies.SBRT in unresectable pancreatic cancer may have promise for local control, but there has been significant duodenal toxicity. In addtion, RTOG 9704 demonstrated the general patterns of failure are distant metastases, in 71+% of cases, followed by a 23-25% local control failure.

Dose escalation trials using IMRT have been performed (Spalding 2007), increasing does from 52-66 Gy and achieving a boost dose of up to 85 Gy. This is technically possible as are the SBRT studies with cyberknife type technology, but again the distant disease is the primary mechanism of failure. There can be no durable control of distant disease without local control, but at present according to RTOG 9702 distant spread is the main failure rate in resected disease. Thus, improved local control, especially with added toxicity may not be an improvment.

Toxicity and Normal Tissue Tolerances

Liver < 30 Gy to ≤ 50%, Kidneys: ≥ 2/3 of one kidney ≤ 18 Gy; Spinal Cord ≤ 45 Gy; stomach and small bowel, < 50 Gy to small volumes.