The purpose of this phase III study is to evaluate the safety and efficacy of the combination of eflornithine and sulindac compared to single agent sulindac or eflornithine in reducing the number of polyps in patients with familial adenomatous polyposis (FAP).

Inclusion Criteria:

• Diagnosis of phenotypic Familial Adenomatous Polyposis (FAP) of the colorectum based on meeting the criteria in one of two groups: Group 1-Greater than 100 adenomatous colorectal polyps prior to age 40. Group 2-Greater than 10 adenomatous polyps and age <40 or greater than 25 polyps and age >40; combined with a dominant family history or genotype: More than 100 polyps in a first-degree relative; More than 25 polyps in 2 relatives in 2 generations, including a first-degree family member; Genetic diagnosis in a relative; Genetic diagnosis by in vitro synthesized truncated protein or similar assay.

• No colorectal surgery or prior colon surgery for polyposis at least 1 year prior (total abdominal colectomy with ileal-rectal anastomosis, or total proctocolectomy with ilea pouch-anal reconstruction.

• Baseline endoscopy

  1. If no prior colorectal surgery, at least 3 polyps in a cluster each ≥ 2 mm in diameter; or
  2. If rectum is in situ and to be assessed, baseline rectal segment endoscopy documenting 3 or more rectal polyps each at least 2 mm in diameter in a defined cluster and/or at least 6 polyps, ≥ 2 mm in diameter, in the distal 10 cm of rectum
  3. If ileal pouch neo-rectum is in place, 3 or more pouch polyps in a cluster ≥ 2 mm in diameter, or at least 6 polyps, ≥ 2 mm in diameter, in the distal 10 cm of pouch.
  4. Clinical/pathological grading of duodenal polyps will utilize the Spigelman Classification.

• Hematopoietic: no significant hematologic dysfunction; WBC ≥3,000/mm3; platelet count ≥100,000/mm3; hemoglobin ≥10g/dL; no known or prior clinical coagulopathy.

• Hepatic: bilirubin ≤ 1.5 times ULN; AST and ALT ≤ 1.5 times ULN; Alkaline phosphatase ≤ 1.5 times ULN.

• Renal: No significant renal dysfunction; creatinine ≤ 1.5 times ULN.

• Hearing: no clinically significant hearing loss that affects everyday life.

• Not pregnant or nursing.

• Negative serum pregnancy test if female of child-bearing potential.

• Absence of gross blood in stool.

• Fertile patients must use effective contraception.

• Stool occult blood either negative or minimal (1+).

• No prior hypersensitivity to cyclooxygenase-2 inhibitors, sulfonamides, NSAIDs, or salicylates; no NSAID associated symptoms of gastritis.

• No discrete gastric or duodenal ulcer greater than 5 mm within the past year except Helicobacter pylori-related peptic ulcer disease treated successfully with antibiotics (as documented by an endoscopy.

• No invasive malignancy within the past 5 years except stage I or II colon or rectal cancer or resected nonmelanomatous skin cancer.

• No other significant medical or psychiatric problems that would preclude study participation.

• No chronic adrenocorticosteroids.

• No prior pelvic irradiation.

• At least 3 months since prior investigational agents.

• Patients may not be receiving or plan to receive corticosteroids.

• Concomitant NSAID use outside this study may not exceed 4 days per month.

• Use of 81 mg daily aspirin or 650 mg aspirin not more than once a week.

• No concurrent warfarin, fluconazole, or lithium.

• Must be willing and able to sign informed consent.

Exclusion Criteria:

• High Risk for cardiovascular disease including clinical diabetes mellitus (Type I or II) requiring glycemic medications; Prior personal history of cardiovascular disease or, two or more of the following - hypertension or use of anti-hypertensive medications, hyperlipidemia or use of lipid-lowering medications or current smoker.
• Hearing loss that affects everyday life and or for which a hearing aid is required.