Pancreatic cancer

Epidemiology
32,000 cases per year, about the same number of deaths
20% of tumors are resectable, 20% of resected patients are long term survivors

Risk factors
Age (increased after 45 years)
Males (RR 1.3)
Smoking
Black men
Chronic pancreatitis
Family history
Diabetes

Anatomy
Retroperitoneal organ
Head lies in the curve of the duodenum
Tail approaches the spleen
Transverse colon and greater omentum lie in front
More than 2/3 of tumors are found in the head of the pancreas
Nodes: peripancreatic, porta hepatis, celiac, para-aortic, splenic (tail lesions)
Vascular structures determine resectability: SMV invasion may be resectable with experienced surgeon: lesions invading the SMA or celiac A. are generally considered unresectable

Pathology
90% of tumors are adenocarcinoma
Other histologies: neuroendocrine, cystadenocarcinoma, islet cell and other rare endocrine tumors
Tail and body lesions have a worse prognosis; peri-ampullary tumors are better
K-ras is activated in 90% of pancreatic cancers
Other abnormalities include p53, p16, DPC4, BRCA2 mutations

Clinical
Most common presentation is jaundice
Back pain suggests a more advanced lesion
Prognostic factors: resectability*; resection margin

Therapy
Surgery is the definitive treatment
Some controversy over what is the optimum adjuvant treatment

Key adjuvant trials
GITSG (PMID 4015380)
Randomization to 40Gy + 5FU (split course RT) vs. observation
CRT was associated with:
-improved overall survival, median 11 vs 20 months, 5-year 5 vs 19% and improved PFS

EORTC (PMID 17968163)
Randomization to observation vs. 40Gy + 5FU (split course RT)
Ampullary carcinomas also included (almost half the patients)
Other criticisms: no maintenance chemo, no margin assessment, 20% non-compliance rate
Trend to improved median (19 vs 24 months) and 2-yr OS (41 vs 51%)

ESPAC (original PMID 11716884; 5-year update PMID 15028824)
4-arm trial (CRT, CRT + maintenance chemo, chemo only, observation)
Not randomized with significant rate of cross-over between arms
Observation arm associated with the best survival
Trial has been extensively criticized

RTOG 9704 (PMID 21499862)
Randomization to 5-FU vs. gemcitabine + 50.4Gy RT
Five-year update showed trend to OS benefit in head of pancreas patients with gemcitabine therapy but overall survival was not improved in the entire cohort.

CONKO-001 (PMID 17227278)
Randomization was to adjuvant gemcitabine vs. observation (postop CA19-9 had to be less than 90)
Trend to improved OS in gemcitabine group (20 vs 23 months)
DFS was significantly improved with chemo (7 vs 14 months)
High rates of crossover to salvage gemcitabine in the observation arm

Hopkins/Mayo collaboration (PMID 2840672)
Retrospective analysis of 1,092 cases of resected pancreatic cancer
53% received adjuvant CRT
Patients treated with CRT were younger, more likely to have high grade disease and positive margins
Matched pair analysis showed that CRT was associated with improved median, 2-year, and 5-year overall survival compared with CRT (21.9 vs 14.3 months, 45.4 vs. 31.4%, and 25.4 vs. 12.2%, respectively)

Locally advanced/unresectable, non-metastatic pancreatic cancer
CRT has been shown to provide benefit in several trials:

Mayo Clinic: 40Gy EBRT +/- 5FU (PMID 4186452)
Median OS: 6 vs 10 months (better with chemo)

GITSG: 60Gy split course EBRT vs. 40Gy + 5FU vs. 60Gy + 5FU (PMID 7284971)
RT alone arm closed early (1-yr OS 11%); both CRT arms had similar survival (36/38%)

GITSG: 40Gy + adriamycin vs. 60Gy + 5FU (PMID 2864997)
Overall survival was similar but toxicity rates higher in adriamycin group

GITSG: SMF vs. 54Gy + concurrent 5FU and adjuvant SMF (PMID 2898536)
1-year OS improved in the RT group (41 vs 19%)

ECOG: 5-FU only vs. 5-FU + 40Gy
Median OS about 8 months in both groups
Only negative trial for CRT in locally advanced disease
Study also included patients with residual disease after resection

FFCD/CFRO phase III study (PMID 18467316):
119 patients randomized to 60Gy with concurrent 5FU/cisplatin --> maintenance gemcitabine vs. single agent gemcitabine chemotherapy alone
Median OS 8.6 vs 13 months in CRT--> chemo vs. chemo alone
1-year OS 32 vs 53% respectively
Chemoradiation arm associated with higher rates of grade 3/4 toxicity