56th Annual Ob-Gyn UpdateOvarian Cancer:

Finally We Can Start To Talk About

Prevention!

Walter Kinney M.D.Division of Gynecologic OncologyThe Permanente Medical Group

Clinical Professor of Ob/Gyn, U.C. DavisSacramento, California

Dr. Bethan Powell

Gynecologic OncologistKP San Francisco

Slide designs and images!

Uncompensated Colleagues

Dr. Helen Dinkelspiel Gyn Oncology Fellow

Columbia University First edition of this talk

Why Not Screening? Prevention

Requires a “detectable preclinical phase” of long duration (like colon and cervix). We have now identified a preclinical phase for some of the serous cancers (STIC) but presently there is no screening test that can detect this.

Mortality ReductionDiagnosis at earlier stage. Requires long stage

duration. Stage duration for high grade serous (HGS) ovarian cancer is measured in weeks or months

Why Not Screening? PLCO Trial – General Population

•Large, federally funded, prospective randomized controlled clinical trial of annual CA-125 and ultrasound: 78,216 women, age 55-74, median followup 13 years•Outcome – Screening arm experienced:

– No change in rate of occurrence of ovarian/peritoneal/fallopian tubal cancer (hereafter “ovarian cancer” or “HGS cancer”)

– No change in stage distribution– No survival advantage– 1080 surgeries including 163 “serious complications”

Buys SS, JAMA 2011 305:2295-2303

Why Not Screening? UKFOCSS – High Risk Women

•Annual CA-125, and TVUS for abnormals•3663 women with >=10% lifetime risk•Women diagnosed with cancer who were screened in the preceding year before diagnosis: 26% Stage IIIC+ versus 86.7% Stage IIIC+ and OS 72 versus 48 months BUT•60% of Stage I were Lynch syndrome; BRCA associated cancers were 7/8 Stage IIIC+ AND•Only 2/13 incident screen detected cancers were early stage•Conclusion – screening interval changed to 4 months and CA-125 interpretation changed

Long KC, JCO 2013 31:8-10 and Rosenthal AN, JCO 2013 31:49-57

Why Not Screening? Symptoms and p53 on Pap

•Value of symptom triggered diagnostic evaluation for ovarian cancer

– 5012 women age 40+; 241 positive symptom screen (4.8%)

– 6 surgeries for ovarian masses, 2 cancers diagnosed with 6 months, one stage 1A, symptom negative and one Stage III+, symptom positive

•Cells from most endometrial cancers and some HGS pelvic cancers in liquid Pap specimens

– Pap screening for p53?– Untested for screening or more importantly for

precursor detection

Anderson MR Ob Gyn 2015;123:73-9, and Kunde, Sci Transl Med 2013; 5:167

Important Insights

• Screening is ineffective for High Grade Serous cancer in the pelvis. There are no effective tests for early stage disease and the stage duration is too short

• Ovarian cancer is several (probably many) diseases in two general categories: HGS (70%) and everything else

• New data suggests 70% of HGS carcinomas may originate in the fallopian tube

• A precursor state is identifiable (STIC)• Explosion in the world of genetics: panel testing

now available and about 25% of ovarian cancer is related to an inherited gene.

Classic thinking….

Epithelial (EOC)(90%)

Germ Cell(3%-5%)

Sex Cord-Stromal(2%-3%)

Secondary(Metastatic)

(5%)

Figure modified from Gartner, L.P. & Hiatt, J.L. eds. In Color Atlas of Histology.3rd ed. (2000) Lippincott Williams & Wilkins: Philadelphia, PA.

  Serous

Endometrioid

Mucinous

Clear cell

Transitional

Undifferentiated

Grade 1  3.5%

 10%

     

Grade 2

     4%    

Grade 3  70%

     10%  

2% NOS

EOC Grouping reflects epidemiology, germline genetics, somatic genetics, animal models, clinical presentation and response to therapy

Ovarian cancer: different diseases

Gilks, Mod Path 2008Kurman and Shih, July 2011

Origin of ovarian cancer: Paradigm shift

TYPE 1• Clear cell,low grade

endometrioid• Low-grade serous• Mucinous• Molecular Markers: BRAF,

KRAS, PTEN, MEK, MAPK, Beta-catenin

• Early changes and precancer surrounds cancer

• Chemotherapy resistent

TYPE II• High grade serous,

endometrioid, carcinosarcoma.

• Molecular markers: P53, CCNE2, Ki-67, MIB-!, PAX2, PAX8, PIK3CA, WT-1

• Fallopian tube markers• Most late stage, rapidly

growing.• Chemotherapy sensitive

Timeline: Origin of serous cancer in the FT

1993BRCA1

1997 FTCA assoc with BRCA

2001Tubal dysplasiaIn RRSO

2004 Progression ofp53 signature to STIC

200770% of PSC involve tube

2010P53 mutations match, molecular genetics match

Today:Clinical Implications of FTorigin

Bowtell et al, Nat Rev Can 2011 Crum, Clin Med 2007 Kurman et al, Human Path 2011

Stepwise Progression: Precursor Lesions

Levanon JCO 2008, Lee et al

P53 Signature

P53 Signature – Possible Precursor Lesion

STIC and cancer in proximity

Pathology Review of 150 RRSO Cases Performed at UCSF

Case Ovary FT Side Size Met Stage FU

1 Serous Neg Uni 0.9 mm None IA PP, 7y

2 Neg Serous, TIC, I Uni 8.2 mm None IA-0 NED,

7y

3 Neg Serous, TIC, F Uni 2.0 mm +

Cyto IC NED, 4y

4CIS,

surface

TIC, F Uni 1.0 mm None 0 NED, 3y

5 Serous Neg Bi 11, 3 mm LN IIIC AWT, 2y

6 Neg TIC, F Uni 1 mm None 0 NED, 1.5y

7 Neg TIC, F Uni 2 mm + Cyto IC NED,

1y

8 Neg TIC, I Uni 2 mm None 0 R

9 Serous Serous, TIC, F

Ovary BiFT Uni

Ovary 13, 5.5 mmFT 6 mm

LN+

CytoIIIC R

STIC and invasive Ca at RRSO

• STIC found in 3% of 2035 RRSOs – Median age 50– 70% BRCA1

• Invasive Ca in 2.7% of 3030 RRSOs– 71% of tubal origin– Median age 51– 79% BRCA1

Powell, Gyn Onc 2014

Evidence compelling that some if not most HGS cancers originate in the fallopian tube

• Consistent with epidemiology• Precursor lesion in tube, not in ovary• Mullerian molecular markers: PAX8, not

calretinin• Molecular markers match, P53 mutation in

>80%• STIC or tube involved in 70% of PSC in sporadic

and BRCA related cancer.• Mouse model

Genetics: Who Gets Referred?

Genetics: Germline mutations in ovarian cancer

• Approximately 24% of ovarian cancers have a germ line mutation, 17% BRCA and 7% nonBRCA (18+ other genes)

• Lynch is a small contributor to the total: 8/1890 cases (0.4%) sequences by UW or GOG

• 31% -44% have no family history and 40% are >age 60 at diagnosis

• Approximately 12% of women alive in 2013, with ovarian cancer diagnosed in the last 5 years have been tested for BRCA1 and BRCA2

Walsh, PCOS 2012, myriad annual report 2013

Not your mother’s BRCA: Gene Panels

• BRCA 1 ,BRCA 2• MLH1, MSH2,• MSH6, PMS2, EPCAM• STK11• TP53• CHEK2, BRIP1, RAD50,

RAD51C, RAD51D,• PALP2, ATM, • TP53, BARD1, NBN, MRE11A

Genetics: Germline mutations in ovarian cancer

• Approximately 24% of ovarian cancers have a germ line mutation, 17% BRCA and 7% nonBRCA (18+ other genes)

• Lynch is a small contributor to the total: 8/1890 cases (0.4%) sequences by UW or GOG

• 31% -44% have no family history and 40% are >age 60 at diagnosis

• Approximately 12% of women alive in 2013, with ovarian cancer diagnosed in the last 5 years have been tested for BRCA1 and BRCA2

Walsh, PCOS 2012, myriad annual report 2013

Genetics: Who gets referred?March 2014 SGO Statement on Genetic testing

All women with ov/ft/pp cancer should considered for genetic counseling and testing in the absence of family history.

All women with endometrial cancer should undergo a clinical or molecular evaluation for lynch syndrome with molecular tumor evaluation the preferred method.

Genetics: Germline mutations in ovarian cancer

• Approximately 24% of ovarian cancers have a germ line mutation, 17% BRCA and 7% nonBRCA (18+ other genes)

• Lynch is a small contributor to the total: 8/1890 cases (0.4%) sequences by UW or GOG

• 31% -44% have no family history and 40% are >age 60 at diagnosis

• Approximately 12% of women alive in 2013, with ovarian cancer diagnosed in the last 5 years have been tested for BRCA1 and BRCA2

Walsh, PCOS 2012, myriad annual report 2013

Compliance with referral to genetic counseling

• 59% with breast cancer were referred while only 21% with ovarian cancer were referred (p<0.0001).

• Opportunity missed in ovarian cancer due to death.

• Attrition every additional step to testing.

• Family history and ethnicity were difficult to interpret with very few documented Jewish women in cohort.

Powell Int J Gyn Cancer, 2011

All guidelines recommend: Risk Reducing salpingo-oophorectomy for BRCA mutation

carriers

90 % effective in reducing ovarian cancer50% reduction in breast cancer if performed before age 50Increase life expectancy 6.6-11.7 years for combined BSO, mastectomy.

.

Microsectioning the tubes and ovaries

Medeiros et al, Am J Surg Path, 2006

Sectioning of RRSO Specimens

BRCA Risk of Ovarian Cancer

AGE BRCA1 BRCA2

40 2.2% <1%

50 8.7% 1.9%

60 22% 7.4%

Chen et al, JCO 2007

Oophorectomy reduces Breast Cancer Risk in BRCA mutation carriers

Risk reduction if performed by age 40• BRCA 1: 56% (OR= 0.44;95% CI 0.29,0.66)• BRCA 2: 46% (OR=0.57;95% CI 0.28,1.15),

Domchek JAMA, 2010Eisen et al, J Clin Oncol, 2005

Compliance with NCCN guidelines in a community based health care system

• 305 women with BRCA mutations in KPNC• Mean age at BRCA test: 47!

• 74% elected RRSO• 17% of these women <40yrs• Median time 6 mos from test date to surgery date

Garcia, C et al Gyn Onc 2013

Oral contraceptives reduce Ovarian Cancer Risk in BRCA mutation carriers

Cancer risk reduction:• Ovary: OR= 0.58; 95% CI 0.46-0.73)• Breast: OR=1.21; 95% CI 0.93-1.58),

Moorman PG, J Clin Oncol 2013 31:4188

Adherence to Cancer Screening Among Women With BRCA1 or 2 Who Retained

Tubes and Ovaries

Garcia, Powell, et al, WAGO, 2013

NCCN guidelines for BRCA mutation carriers

• Remove the tubes and ovaries when between the age of 35-40 or when completed childbearing.

• Screening with CA 125 and ultrasound q 6 months.

Hot flashes

Cognitive impairment

sexualityosteoporosis

Cardiac mortality

Early Menopause

Salpingectomy in mutation carriers

Pros• Some cancer risk reduction• Avoid premature

menapause• Maintain option for IVF

pregnancy• Option for those unwilling to haveBSO

Cons• Two stages to surgery• Delay of removing the

ovaries• May not be as effective• No Breast cancer risk

reduction

Cost Effectiveness Model

Average Discounted Costs (Can $)

Avg Life Expectancy Gain

Avg QALY Expectancy Gain

Incremental C-E Ratio (cost per QALY)

BRCA1

BSO @ 40 $25,987 21.15 17.56 ---

Tubes @ 40 $38,208 20.74 18.17 $20,050

Tubes @ 40, Ov @ 50

$41,577 20.83 18.26 $37,805

BRCA2

BSO @40 $16,932 22.62 18.87 ---

Tubes @ 40 $33,150 22.08 19.51 $25,658

Tubes @ 40, Ov @ 50

$37,686 22.14 19.56 $89,680

Kwon et al, Obstet Gynecol 2013Delayed oophorectomy with the greatest QALY

Why leave the tube in women at average risk?

In US 30% women undergo hysterectomy, 50% have ovaries and fallopian tube left in situ

~20% of women who develop ovarian cancer have had a prior hysterectomy

Up to 20% of ovarian cancer patients have had a tubal ligation

Most women with inherited risk are unidentified.

Walsh et al, PNAS 2011

Davis et al, J La Soc 2003

Removal of the tube at hysterectomy

• No detrimental effect on: ovarian function

hormonal levels blood supply to the ovary• If tube is left in situ, incidence of hydrosalpinx: 28%,

requiring surgery 7.8%

Dar et al, 2000 Sezik et al, 2007 Ghezzi et al, 2009,

Morse et al, 2006 Morelli 2013, Findley 2013

Salpingectomy Instead of tubal ligation or at time of any pelvic surgery

Pros• Decreased tubal pathology, • Decreased ectopic rate• Improved sterilization

efficacy• Potential cancer risk

reduction

Cons• Requires surgery• May be difficult to access

tube• Potential for bleeding • May require additional

equipment/incision/cost• Uncommon cancer

Salpingectomy versus Tubal Ligation

• Effect of tubal sterilization on risk of serous epithelial or primary peritoneal cancer

• Nested case control, 194 cases and 388 controls• Controlled for prior hyst or SO, oral contraceptive use,

endometriosis, infetility, gravidity and parity• HGS cancer risk

– Any tubal sterilization OR 0.59 (0.29-1.17) p = 0.13– Salpingectomy OR 0.36 (0.13-1.02) p = 0.054

Lessard-Andersen, Gyn Onc 2014; 135:423-7

Technique

BRCA 1 or 2 mutation• Inspect entire abdomen• Peritoneal cytology• Remove adjacent ovarian

capsule (fimbria ovarica)• Remove all the tube • Place in an endoscopic bag • Pathology: microsectioning entire tubes

No known genetic risk• Preserve the utero-ovarian

ligament• Remove or cauterize any

attachments to the ovary • If the entire fallopian tube

cannot be removed, consider removing the tubal fimbrial end

• Pathology: examine the fimbrae in 2-3 cassettes.

Ovarian Cancer Prevention• Improve compliance with referral guidelines to

genetic counseling for women at hereditary risk

• RRSO for mutation carriers: big change in overall survival

• OCs or preferably salpingectomy for a bad family history and negative testing or mutation carriers who are not disposed to consider RRSO

• Opportunistic salpingectomy