Embed Size (px)
Citation preview
4/8/13
1
Jeffrey Fox, MD, MPH UCSF Primary Care Medicine: Update 2013
4/8/13
2
Smoking Obesity Diet NSAIDs Symptoms usually mean it’s too late
Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research; 2007.
4/8/13
3
American Cancer Society, 2011 estimates
4/8/13
4
Colorectal cancer: We’re winning!
Just the facts – colorectal cancer
• 2010 NCI es*mates for US: – 142, 570 new CRC diagnoses – 51,370 CRC deaths
• 3rd leading cause of cancer in men and women
• 2nd leading cause of cancer death in men and women
• Over 90% 5-‐year survival when caught early • Under 40% are caught early
Horner MJ, SEER Cancer Sta*s*cs Review, NCI, 2010
Race/Ethnicity Male Female
All Races 57.3 per 100,000 men 42.8 per 100,000 women
White 56.9 per 100,000 men 42.1 per 100,000 women
Black 69.3 per 100,000 men 53.5 per 100,000 women
Asian/Pacific Islander 46.9 per 100,000 men 34.6 per 100,000 women
American Indian/Alaska Native
43.1 per 100,000 men 41.2 per 100,000 women
Hispanic 46.3 per 100,000 men 32.2 per 100,000 women
Horner MJ, SEER Cancer Sta*s*cs Review, NCI, 2009
Overall CRC incidence: 50 per 100,000 per year
4/8/13
5
CRC incidence and mortality declining Influence of treatment? Influence of diet and lifestyle? Influence of screening?
Overall 5-‐year survival about 65% (SEER data) 30-‐35% in China and Eastern Europe
Colorectal cancer mortality has declined by 36.5% since 1985
Colorectal cancer mortality has declined by 36.5% since 1985
What happened here?
4/8/13
6
.
The impact of a celebrity promo*onal campaign on the use of colon cancer screening: the Ka*e Couric effect. Cram P, Fendrick AM, Inadomi J, Cowen ME, Carpenter D, Vijan S. Arch Intern Med. 2003 Jul 14;163(13):1601-‐5
The rate of colonoscopy screening went up 20-‐40% aaer her na*onally televised colonoscopy
4/8/13
7
Edwards BK, Cancer, 2009
Stool occult blood testing First method proven in randomized trials to decrease colorectal cancer incidence and mortality
33% decrease in mortality with annual use 20% decrease in colorectal cancer incidence with annual use
Detects cancers at earlier stages than controls Limited in test characteristics and adherence
Mandel JS, et al. NEJM ,1993 & 2000
Flexible sigmoidoscopy Randomized trials show mortality decreased by 26-‐31% 10+ years after sigmoidoscopy
30% of proximal advanced adenomas missed
Assumes L colon pathology predicts R colon pathology ▪ In women miss up to 2/3 of lesions this way
Lieberman DA, et al. NEJM, 2000
Atkin WS, Lancet 2010 Schoen RE, et al. NEJM 2012
Schoenfeld P, et al. NEJM, 2005
4/8/13
8
Colonoscopy More sensitive than other tests (>90%) One stop shopping National Polyp study ▪ 50% reduction in CRC mortality relative to expected rate
Higher complication rate than other tests Most costly of all the screening tests
Zauber AG, et al, NEJM 2012
Zauber AG et al, NEJM, 2012
Modality Odds ratio
Distal colon cancer Mortality
Odds ratio Proximal colon cancer
Mortality
Modality Odds ratio
Distal colon cancer Mortality
Odds ratio Proximal colon cancer
Mortality
0.41 (0.25-‐0.69) 0.96 (0.61-‐1.50)
0.33 (0.28-‐0.39) 0.99 (0.86-‐1.14)
4/8/13
9
Modality Odds ratio
Distal colon cancer Mortality
Odds ratio Proximal colon cancer
Mortality
Flexible sigmoidoscopy1
0.41 (0.25-‐0.69) 0.96 (0.61-‐1.50)
Colonoscopy2 0.33 (0.28-‐0.39) 0.99 (0.86-‐1.14)
1Selby JV, et al. NEJM, 1992 2Baxter NN, et al. Ann Intern Med, 2009
Endoscopy really good
4/8/13
10
Endoscopy really good
Endoscopy may fall short
• Even colonoscopy appears to be limited in the proximal colon relative to the distal colon
• Why the difference? – Technical difficulty of inspection in R colon – Inferior bowel prep – Less skilled practitioners – Flat polyps – Gender differences – Separate biology
adenoma carcinoma normal mucosa adenoma carcinoma normal mucosa
R colon > L colon Flat/depressed > polypoid
4/8/13
11
Modality Interval ACS-‐MSTF USPSTF
Hemoccult II 1 year No Yes
High-‐sensitivity Hemoccult or fecal immunochemical test
1 year Yes Yes
Flexible sigmoidoscopy 5 years Yes Yes
CT colonography 5 years Yes Insufficient evidence
Colonoscopy 10 years Yes Yes
Stool DNA testing 5 years Yes Insufficient evidence
ACS-MSTF: American Cancer Society-US Multi-Society Task Force, 2008 USPSTF: US Preventive Services Task Force, 2008
Identify who is at risk Effective treatment for pre-‐cancer Optimal screening study Optimal screening intervals Improving colonoscopy Overcoming barriers
Average risk 5.2% lifetime risk
Taylor et al, Gastroenterol, 2010
Family members affected RR (95% CI)
First degree relative, any age 1.91 (1.82-‐2.00)
First degree relative >50 2.02 (1.93-‐2.11)
First degree relative <50 3.31 (2.79-‐3.89)
2 first degree relatives, any age 3.01 (2.66-‐3.38)
Second degree relative only, no 1st degree 1.05 (0.99-‐1.11)
“Dose response”: your risk associated with number, age, and “degree” of your affected rela*ves
Only 15% of patients with colorectal cancer have an affected 1st or 2nd degree relative
Do we need to be more inclusive? HNPCC: endometrial, gastric, pancreatic, small bowel, biliary, ovarian, urothelial, brain, skin
Family history of polyps? ▪ Advanced adenomas in 1st degree relative RR=2 ▪ Non-‐advanced adenoma in 1st degree relative NO increased risk
Cojet, Gastroenterol, 2007
4/8/13
12
Risk factor Effect on CRC risk
Smoking Increased
Alcohol >2 drinks per day (men) or >1 drink per day (women)
Increased
Diet high in red meat or processed foods Increased
Diet high in fruits and vegetables Decreased
Active lifestyle Decreased
Sedentary lifestyle Increased
Risk factor Effect on CRC risk
Personal history of polyps or CRC Increased
Family colon cancer syndrome Increased
Ulcerative colitis or Crohn’s disease Increased
Type 2 diabetes mellitus Increased
Obesity Increased
Prior pelvic irradiation Increased
• Polypectomy –National Polyp Study • Chemoprevention
– NSAIDs/Asprin – Vitamins
• Folic Acid • Calcium • Vitamin D • Vitamin B6
– HRT – RCT and meta-‐analysis – decreased CRC but incr breast CA, thrombotic/embolic events
Zauber AG et al, NEJM, 2012
Chan AT, JAMA, 2005
Average risk screening: the list of options approach (USPSTF) Highly sensitive fecal testing annually Flexible sigmoidoscopy every 5 years Colonoscopy every 10 years Start age 50, stop age 75
4/8/13
13
Average risk screening: the list of options approach (USPSTF) Highly sensitive fecal testing annually Flexible sigmoidoscopy every 5 years Colonoscopy every 10 years Start age 50, stop age 75
Endorsed by ACP, AAFP, and CDC Sharaf RN, et al. Am J Gastroenterol, 2013
Strong family history: start age 40 or 10 yrs prior to youngest 1st degree relative, then q 5 yrs 1st degree relative under 60 with CRC or advanced adenoma
Multiple 1st degree relatives of any age CRC/AA Familial colorectal cancer syndromes (various) Chronic inflammatory bowel disease: q 2-‐3 yr starting 10 yrs after disease onset
Blacks: start age 45
Risk appears to exceed benefit at some point Life expectancy should be >5 years because screening doesn’t improve mortality until then
Screening ages >80 yields less than 1/6 the life-‐years gained compared to screening ages 50-‐54
We overscreen sick 80+ and underscreen healthy 60-‐75
Walter LC, et al. Ann Intern Med, 2009
Lin OS, et al. JAMA, 2006
4/8/13
14
Most should stop at 75 (USPSTF) Prior adenomas/cancer should stop at 80 If never screened prior to 75, consider single time screening 75-‐80
No screening/surveillance beyond 85
Only 60.8% of >50 screened as of 2006 Demystifying
Thank you Katie! Herd effect
Increase convenience Offer a buffet of options
Inreach/outreach
• 55,300 randomized to an outreach to either: – FIT q 2 years – colonoscopy one time
• Adherence low • FIT 33% • Colonoscopy 20%
• Number of colon cancers found equivalent between groups (baseline exam only)
• Higher in colonoscopy group were: – benign polyps – complications (24 vs 10)
Quintero et al NEJM 2012 Khalid de Bakker C et al. Endoscopy 2011
4/8/13
15
Gastro, 2011 Nov;141(5):1551-‐5
FIT Kit Mailed
PCP Pre-letter Mailed
Robo-call Reminder
One week prior to kit
3 weeks after kit
Reminder Postcard
6 weeks after kit
Secure Msg
Regional Local
Increased regional initiatives based on previous pilots and local innovation
MA Calls
Region-wide 2nd kit mailing to non-responders
Distribute Kit At Office visit Or Flu Clinic
4/8/13
16
SEER summary stage
Barrett’s Esophagus: Much ado about
what exactly?
4/8/13
17
Barrett’s esophagus (BE): A change in the esophageal epithelium of any length that can be recognized at endoscopy and is confirmed to have intestinal metaplasia by biopsy
Pre-‐malignant lesion for esophageal adenocarcinoma (EAC)
Most EAC accompanied by BE
Primary identifiable risk factors for Barrett’s esophagus and esophageal adenocarcinoma male Caucasian chronic symptomatic GERD age obesity smoking
4/8/13
18
GERD a risk for esophageal adenoCA
GERD sx at least once/wk Controls
Esophageal adenocarcinoma
OR (Adjusted)
No 685 (84%) 76 (40%) ref
Yes 135 (16%) 113 (60%) 7.7
Lagergren et.al., N Eng J Med, 1999
Incidence of EAC has increased 10-‐fold over last 30 years
Incidence of EAC between was increasing 4-‐10% per year through
Incidence of Barrett’s esophagus appears to be rising at a comparable rate
Van Soest, et.al. Gut 2005
Overall mortality from EAC increasing (reflecting increasing incidence since it is rarely found in curable stage)
In those diagnosed with cancer, survival to 1 and 5 yrs improved over last 25 years, but is still poor 15-‐20% overall 5-‐year survival 37% 5-‐year survival if initial staging shows localized disease
Esophagectomy only “cure” ▪ Large morbidities associated
SEER data, 2006
4/8/13
19
Initial estimates of EAC arising in BE as high as annual risk of 1/48 (2.1%)
Strong correlation between the size of the study and estimated cancer risk smaller studies predict higher risk than larger, population-‐based studies
PUBLICATION BIAS likely overestimated CA risk1
Barrett’s Esophagus Study Trial (“BEST”)2 1376 patients followed avg 4.12 years 1/200 annual risk EAC (0.5% per pt-‐yr)
1Shaheen, et al. Gastroenterol 2000 2Sikkema M, et.al. Clin Gastroenterol Hepatol 2010
Newest estimates of progression from BE to EAC are MUCH LOWER than previous
Study locale
# of patients Years of f/u
Annual progression to EAC
Denmark1 11,028 5.2 (median) 0.12%
Ireland2 8,522 7.0 (mean) 0.13%
1Hvid-‐Jensen F et al. N Eng J Med 2011 2Bhat S et al. JNCI 2011
4/8/13
20
GERD is the most common risk factor, but about 40% of patients found to have EAC don’t complain of prior GERD symptoms Hard to select subset to screen for BE
Those diagnosed with BE tend to die of non-‐esophageal causes much more frequently than of EAC and identical adjusted life expectancy to those unaffected by BE
Lagergren et.al., N Eng J Med, 1999
Anderson, et.al. Gut, 2003 Solaymani-‐Dodaran, et.al. Am J Gastroenterol 2005 Sikkema M, et.al. Clin Gastroenterol Hepatol 2010
Medical therapy (proton pump inhibitors) and surgical therapy 75-‐90% effective at treating symptomatic GERD and esophagitis
GERD contributes to risk of dysplasia and EAC arising from BE
Other factors are important medical GERD therapy & antireflux surgery not proven to revert BE to normal mucosa or to prevent EAC
Smith et al, Am J Surg Pathol, 1984 Sharma et.al., Am J Gastroenterol 1997 Spechler et.al., JAMA, 2001 Parilla et.al., Ann Surg, 2003
Ablation Photodynamic therapy, radiofrequency ablation Concern for subepithelial recurrence Sham-‐controlled randomized trial for dysplastic BE showed decreased risk of progression to higher grade of dysplasia and adenoCA1
Biomarkers Identifying genetic abnormalities in tissue Help to risk stratify
1Shaheen NJ et.al. NEJM 2009
4/8/13
21
• Multiple retrospective studies suggest NSAIDs and ASA use associated with reduced risk of GI malignancy, including esophageal CA
• Prospective study of 570 BARRETT’S patients (median f/u 4.5 years):
– Use of NSAIDs: RR 0.47 – Use of statins: RR 0.46 – Use of NSAIDs AND statins: RR 0.22
Risk of esophageal CA
Kastelein F et al. Gastroenterol 2011
Endoscopic surveillance is currently the only widely accepted means for attempting to decrease the risk of advanced EAC in patients with non-‐dysplastic BE
Goals detect dysplasia (risk stratification) detect early cancer (improve outcomes)
No dysplasia Every 3-5 yrs
Low-grade dysplasia Every 1 yr
High-grade dysplasia
Focal: Every 3 months
Multifocal: Ablation
Mucosal Irregularity: EMR
Wang KK, et al, Am J Gastroenterol, 2008
Effect of surveillance
TNM Stage at diagnosis Corley et.al., Gastroenterol, 2002
4/8/13
22
Retrospective data show that performance of BE surveillance is associated with earlier diagnosis and improved survival
No randomized or prospective trials showing surveillance prolongs or improves lives
Surveillance in patients with BE q5yrs is cost-‐effective when CA rate in BE is > 1.9%/yr Actual incidence appears to be much lower
Screening patients with GERD for BE and cancer is not cost-‐effective unless we ignore those with non-‐dysplastic BE Not likely, difficult to justify “ignoring” pre-‐malignant condition
Inadomi et.al., Ann Int Med, 2003
Lagergren et.al., Ann Int Med 1999; JAMA 2000
4/8/13
23
CONSIDER white, obese men with chronic GERD (e.g. >5yrs), especially those who smoke, for a one-‐ time EGD to screen for Barrett’s These are at the highest risk for Barrett’s and EAC
Screening of general asymptomatic population NOT recommended
Capsule endoscopy: the next frontier?
All patients with well-‐documented Barrett’s esophagus Consider forgoing surveillance where treatment of early-‐detected lesion unlikely to be tolerated (eg advanced age, comorb)
All patients with Barrett’s should be on a daily proton pump inhibitor whether or not they have symptoms
Hepatocellular Carcinoma: No Longer a
Death Sentence?
4/8/13
24
• In US, incidence rising rapidly – 1.4/100,000 (1976-80) 2.4/100,000 (1991-95)
• 82% of 530,000 cases in world caused by viral hepatitis – 316,000 with hepatitis B – 118,000 with hepatitis C
• Worldwide ‒ – 5th most common cancer, 3rd most common cause of cancer death
– Varies geographically, mostly due to hepatitis B
El-Serag, NEJM, 2000"
Aggressive tumor Median survival following diagnosis is
approximately 6 to 20 months 5-year survival 15%
4/8/13
25
Cirrhosis (all cause) Chronic HBV
5-‐15 fold increased risk (don’t need cirrhosis) Chronic HCV (almost exclusively cirrhotics) Exposure to aflatoxin Less accepted independent risk factors
DM, obesity, smoking, OCP, alcohol, iron overload
4/8/13
26
Vaccinate for HBV Treat HBV and HCV Identify and treat other underlying liver diseases that lead to cirrhosis Etoh abuse Hemochromatosis
Statins? Meta-‐analysis: yes for observational, no for RCT1
1Singh S, et.al. Gastroenterol 2013
Definable population Early detection is necessary for treatment with surgical resection, ablation or transplant
Prognosis for advanced disease is poor Cannot wait for symptoms
Schafer, Lancet, 1999
Methods for screening are limited High cost of testing High frequency needed to detect early Usually detected in advanced stages
16-‐year population-‐based prospective cohort study of 1487 HBsAg-‐positive Alaska natives
AFP levels measured 6-‐monthly elevated AFP followed by ultrasound examination
Elevated AFP in 61 men, 39 non-‐pregnant women (6.7% of total study population)
HCC diagnosed in 32 tumors <6 cm in 23 patients compared to historical control population 5-‐ and 10-‐year survival
rates significantly improved with screening Screening (semiannual AFP measurements) of HBsAg-‐
positive Alaskan natives effective in detecting HCC at resectable stage
McMahon et al, Hepatology 2000; 32: 842
4/8/13
27
18,816 HBsAg+ patients in China Randomized to screening (US+AFP) or no screen 37% reduced risk of HCC related mortality ▪ Not intention-‐to-‐treat analysis ▪ Randomized by cluster (ie factory, school) but analyzed by individual
▪ Overall mortality not reported
Other RCTs comparing screening to NONE unlikely to be performed
Zhang BH et al. J Cancer Res Clin Oncol, 2004
There are no good RCTs comparing surveillance to no surveillance
Cohort studies and cost-‐effectiveness modeling support screening
Consensus recommendations are to screen, but evidence is weak and therefore not as widely adhered to as other forms of cancer screening
Currently covered by insurance anyway
Management of Hepatocellular Carcinoma, AASLD Practice Guideline; Bruix J and Sherman M. Hepatology. 2011 Mar;53(3):1208-‐36.
Newest recommendation recommends NOT using AFP for screening, imaging only
EASL-‐EORTC Clinical Practice Guidelines: Management of Hepatocellular Carcinoma; Llovet JM, et al. Journal of Hepatology. 2012 56:908-‐943.
Hepatitis B carriers (incidence >0.2 %/year) Asian males >40 yo (0.4-‐0.6 %/yr) Asian females >50 yo (0.2 %/yr) Africans >20 yo (unknown, likely>0.2%/yr) Hepatitis B cirrhosis (3-‐5%/yr) Family history of HCC (usually Asian, African)
Non-‐hepatitis B cirrhotics (incidence >1.5 %/yr) Hepatitis C cirrhosis (3-‐5%/yr) Alcoholic cirrhosis Primary biliary cirrhosis
Bruix, Sherman, AASLD guidelines
4/8/13
28
6 month screening interval improved survival relative to 12 month interval1
3 month screening interval did NOT improve detection of small, potentially curable HCC over 6 month2
Bottom line: every 6 months
1San* V et al. J Hepatol 2010 2Trinchet JC et al. Hepatology 2011
Does not correlate well with extent of HCC Elevated AFP is seen in chronic liver disease
without HCC 23% of HCV patients had AFP >10 ng/mL
Test characteristics depend on study population, cutoff level, and gold standard Specificity 41-‐65% Sensitivity 80-‐94%
However, a rise in AFP should raise suspicion Magnitude of AFP elevation in patients with HCC
predicts worse outcome
Gupta, Ann Int Med, 2003
Bottom line on AFP Stand alone screen for HCC: NO Screen for HCC in addition to US: MAYBE
(but still often used)
Diagnosis of liver nodule/mass: YES Check prior elevation for rise: YES
Ultrasound Hard to distinguish small lesions Sensitivity 40-‐78%, specificity 90% Frequently requires confirmatory CT
CT Hard to distinguish tumor from cirrhotic nodules Sensitivity 68%, specificity 81% Higher sensitivity helical CT with arterial phase contrast
MRI Higher sensitivity and specificity in cirrhosis in differentiating benign regenerative nodules
4/8/13
29
Ultrasound Hard to distinguish small lesions Sensitivity 40-‐78%, specificity 90% Frequently requires confirmatory CT
CT Hard to distinguish tumor from cirrhotic nodules Sensitivity 68%, specificity 81% Higher sensitivity helical CT with arterial phase contrast
MRI Higher sensitivity and specificity in cirrhosis in differentiating benign regenerative nodules
Bruix, Sherman. AASLD guidelines. Hepatology 53 (3)1020-‐1022, 2 MAR 2011
Bruix, Sherman. AASLD guidelines. Hepatology 53 (3)1020-‐1022, 2 MAR 2011
Patient comes to you with abdominal pain How do I make sure that my patient does not have cancer? CT scan? US? MRI? Endoscopy/EUS?
4/8/13
30
1 in 1000 risk of iatrogentic cancer from 10 mSv exposure (average abdominal CT)
Need to weigh risk of patient having cancer with potential downsides of the imaging: Iatrogenic cancer from ionizing radiation ▪ less important with older patients
Contrast induced nephropathy ▪ more important with older patients
Incidentalomas
Stop smoking and lose weight Colorectal cancer – get screened Barrett’s: less concern than ever for progression; screening NOT cost-‐effective
HCC – curable if caught early, screening essential management for cirrhotics and certain Hep B patients
Imaging – more judicious use of ionizing radiation
4/8/13
31
Colorectal cancer screening with odour material
by canine scent detection. Sonoda H, Kohnoe S, Yamazato T, Satoh Y, Morizono G, Shikata K, Morita M,
Watanabe A, Morita M, Kakeji Y, Inoue F, Maehara Y. Gut. 2011 Jan 31.
97% Sensitivity
4/8/13
32
