hyperplastic polyposis syndrome: phenotypic presentations and the role of mbd4 and myh

Ha

ELDJ*QB

B(agceecBcptoslawhHtoimitubsudn

Haan(a

GASTROENTEROLOGY 2006;131:30–39

yperplastic Polyposis Syndrome: Phenotypic Presentationsnd the Role of MBD4 and MYH

LIZABETH CHOW,* LARA LIPTON,* ELLY LYNCH,* REBECCA D’SOUZA,* CLELIA ARAGONA,*INDY HODGKIN,* GREGOR BROWN,* INGRID WINSHIP,* MELISSA BARKER,‡

ANIEL BUCHANAN,‡ SHANNON COWIE,§ STEVE NASIOULAS,§ DESIREE DU SART,§

OANNE YOUNG,‡ BARBARA LEGGETT,¶ JEREMY JASS,� and FINLAY MACRAE*Familial Cancer Clinic, Royal Melbourne Hospital, Melbourne, Victoria; ‡The Molecular Epidemiology Laboratory, QIMR, Herston,ueensland; §Murdoch Children’s Research Institute, Royal Children’s Hospital, Victoria; ¶Conjoint Gastroenterology Laboratory, Royal
risbane Hospital, Brisbane, Queensland, Australia; and �Department of Pathology, McGill University, Montreal, Canada
rsrdH

mlttig

emcmpa

athdi

asayi

ackground & Aims: Hyperplastic polyposis syndromeHPS) is defined phenotypically with multiple, largend/or proximal hyperplastic polyps. There is no knownerm-line predisposition. We aimed to characterize thelinicopathologic features of 38 patients with HPS andxplore the role of germ-line mutations in the basexcision repair genes MBD4 and MYH. Methods: Utilizinglinical databases of The Royal Melbourne Hospitalowel Cancer Surveillance Service and the Familial Can-er Clinic, 38 patients with HPS were recruited. Theatients were analyzed for age at first diagnosis, fea-ures of hyperplastic polyposis, family histories of polyp-sis and colorectal cancer (CRC), coexisting adenomas,errated adenomas, incidence of CRC, and microsatel-ite instability in the tumours. Mutation analysis of MBD4nd MYH were performed. Results: Serrated adenomasere common (26%), and 19 (50%) of the 38 patientsad a first-degree relative with CRC. Family history ofPS was uncommon, with only 2 cases found. Ten pa-

ients developed CRC, and 3 required surgery for polyp-sis. No pathogenic mutations in MBD4 were detected

n the 27 patients tested, but 6 single nucleotide poly-orphisms of uncertain functional significance were

dentified. Pathogenic biallelic MYH mutations were de-ected in 1 patient. Conclusions: Mutations in MBD4 arenlikely to be implicated in HPS; MYH mutations shoulde studied, especially when adenomas occur in theame patient. The clinical, histopathologic, and molec-lar findings of this study should contribute to our un-erstanding of HPS and its relationship to the serratedeoplasia pathway.

yperplastic polyposis syndrome (HPS) is a raresyndrome best characterized by multiple, large,

nd/or proximal hyperplastic polyps (HPs) and, occasion-lly, smaller numbers of serrated adenomas (SAs), ade-omas, and admixed hyperplastic/adenomatous polypsADMPs). Recent studies suggest a link between HPS
nd colorectal cancer (CRC) with variable levels of mic-
osatellite instability (MSI), possibly arising through aerrated neoplasia pathway. Reported case series of HPSemain few, with the largest series published so farescribing 28 patients, 20 of whom had more than 20Ps.1

We have studied a cohort of 38 patients with eitherultiple (more than 20), large (�1 cm), or proximally

ocated HPs. The patients’ clinical and pathologic fea-ures are described, with particular attention to inheri-ance and association with CRC. Molecular genetic stud-es have been performed to explore the role of 2 candidateenes in this syndrome: MBD4 and MYH.Genomic integrity is challenged by endogenous and

xogenous DNA damaging agents, and a variety of repairechanisms exist to counteract these mutagenic and

arcinogenic effects. MBD4 plays an important role inaintenance of genomic integrity by impacting on 3

rocesses: base excision repair, DNA mismatch repair,nd cell-cycle response to DNA damage.

MYH-associated polyposis (MAP) is a recently char-cterized autosomal recessive syndrome comprising mul-iple colorectal adenomas, HPs, and CRC. It is the firstuman cancer predisposition disorder to be linked toefects in the base excision repair (BER) pathway. Its rolen HPS has not been previously explored.

Abbreviations used in this paper: ADMP, admixed hyperplastic/denomatous polyp; AFAP, attenuated familial adenomatous polypo-is; BER, base excision repair; CRC, colorectal cancer; FAP, familialdenomatous polyposis; HP, hyperplastic polyp; HPS, hyperplastic pol-posis syndrome; MAP, MYH-associated polyposis; MSI, microsatellitenstability; MSS, microsatellite stable; SA, serrated adenoma.© 2006 by the American Gastroenterological Association Institute

0016-5085/06/$32.00
doi:10.1053/j.gastro.2006.03.046

HCHImftsm5

amscHRAi

caFpunprnagwpa

8

(fepupBt

D1wY

tp

yyHr

T

P

M

M

T

a

July 2006 HYPERPLASTIC POLYPOSIS SYNDROME 31

Materials and Methods

Patient Selection

Utilizing clinical databases of The Royal Melbourneospital Bowel Cancer Surveillance Service and the Familialancer Clinic, patients satisfying the diagnostic criteria ofPS according to the World Health Organization (WHO)

nternational Classification were recruited2 (Table 1). The Fa-ilial Cancer Clinic (FCC) database incorporates patients re-

erred to the FCC because of suspicion of genetic predisposi-ion to polyposis and/or CRC. The clinic and hospital has atatewide reputation for managing polyposis, acting as a com-on but not exclusive referral resource for a state population ofmillion people.The WHO diagnostic criteria of HPS is somewhat arbitrary,

nd future knowledge of molecular and genetic mechanismsay lead to improved classification and definition of this

yndrome. Although some of our patients fall outside theseriteria, all have developed multiple, large, and/or right-sidedPs. Approval for the study was obtained from the Humanesearch Ethics Committee of the Royal Melbourne Hospital.ll participants provided written informed consent for partic-

pation in the study.

Patient Characteristics

The pedigrees were verified, and sequential colonos-opy and histopathology findings were ascertained. Gender,ge, and age at first diagnosis of HPS were documented.amily history of polyposis and CRC were ascertained. Allatients have had colonoscopic surveillance to follow the nat-ral progression of their disease. Information regarding theumber, size, distribution, gross morphology, and histology ofolyps were derived from the colonoscopy and histopathologyeports. Polyps were classified as HPs, SAs, ADMPs, or ade-omas. The total numbers of each polyp type were estimateds rigorously as possible during colonoscopy or from the sur-ical specimen if a colectomy was performed. Details of CRCere reported. Testing for MSI in a subset of cancers andolyps was performed. Thirty-eight germ-line DNA samplesnd cell lines were collected.

Mutation Detection

MBD4 gene testing. The MBD4 gene (gene ID:

able 1. Diagnostic Criteria of HPS According to the WHOInternational Classification

(1) At least 5 histologically diagnosed hyperplastic polyps proximalto the sigmoid colon, of which 2 are greater than 10 mm indiameter, or

(2) Any number of hyperplastic polyps occurring proximal to thesigmoid colon in an individual who has a first-degree relative withhyperplastic polyposis, or

(3) greater than 30 hyperplastic polyps but distributed throughoutthe colona

More than 20 hyperplastic polyps used in other publications.7

930) is on chromosome 3q21-q22. The messenger RNA

mRNA) (NM003925) sequence contains 8 exons and codesor a 580 amino acid protein. Genomic DNA (gDNA) wasxtracted from whole blood using a standard ethanol-saltrecipitation method. We amplified the 8 exons of this genesing polymerase chain reaction (PCR) and sequenced PCRroducts using the ABI 3100 Genetic Analyzer (Appliediosystems, Foster, CA) (Primers and conditions on applica-

ion to authors).MYH gene testing. For MYH mutation testing,

NA was amplified in 2 PCR fragments containing exons3–14 and 6–9, respectively. The common G382D variantas detected by restricting the PCR product with BglII. The165C variant was detected by direct sequencing.

Results

Table 2 summarizes the phenotypic characteris-ics and MBD4 and MYH testing results of our cohort ofatients.

Phenotypic Characteristics

The median and mean ages of our cohort were 52ears and 57 years, respectively, with a range of 32 to 79ears. The median and mean ages at first diagnosis ofPS were 44 years and 49 years, respectively, with a

ange between 27 and 78 years. There were 21 female

able 2. Summary of Findings

henotypic characteristics of 38 patientswith HPS

Median age (range) 52 (32–79)Female/male 21/17Median (range) age at first diagnosis 44 (27–78)Number of patients with proximal HPs 37 (97%)Number of patients with HPs �10 mm 15 (39.5%)Number of patients with more than 5

coexisting adenomas12 (32%)

Number of patients with at least 1serrated adenoma

11 (29%)

Number of patients with at least 1 ADMP 6 (16%)Number of patients with CRC 10 (26%)Median (range) age of diagnosis of CRC 45 (31–71)Number of families with family history of

hyperplastic polyposis2

Number of patients with first-degreerelative with CRC

19 (50%)

Number of patients receiving major colonicresection for CRC

10

Number of patients receiving major colonicresection for polyposis

3

MSI testing results for 5 polyps and 4cancers

MSS

BD4 gene testingCompleted in 27/38 patients 6 single nucleotide

polymorphismsYH gene testingCompleted in 38/38 patients 1 biallelic mutation

carrier Y165C
homozygote

ac

f

4Sp

T

11111111

11

22

2

2

22

222

2

3333

3

3333

Bsa

32 CHOW ET AL GASTROENTEROLOGY Vol. 131, No. 1

nd 17 male patients. The mean number of surveillanceolonoscopies was 5 with a range of 1 to 15.

Thirty-two of 38 patients satisfied the WHO criteria

able 3. Phenotypic Characteristics of Patients With HPS

Patient Sex Total HPs (Bx)R/S HPs

(Bx)HPs prox toSigmoid (Bx)

1 M (34) (10) (24)2a F 65 (41) 25 (17) 40 (24)3a F (3) (1) (2)4a M (9) (3) (6)5a M Multiple (3) 4 (1) Multiple (2)6a M �30 (24) 20 (20) 10 (4)7a F (Multiple) (Multiple) (1)8a F Multiple (31) Multiple (30) (1)9a F Multiple (20) 5–10 (5) Multiple (15)0a F �50 (Multiple) (Multiple) (Multiple)1a F (14) (2) (12)2a M Multiple (80) Multiple (40) Multiple (40)3a M Multiple (23) Multiple (20) (3)4a F (20) (16) (4)5a F Multiple (12) Multiple (4) (8)6a F Multiple (17) Multiple (13) (4)7a M 50–100 (multiple) Multiple (3) (Multiple)

8a F Multiple (27) Multiple (1) Multiple (�6)9a F Multiple (30) �20 (18) Multiple (12)

0a M Multiple (54) Multiple (45) Multiple (9)1a HPS IIfamily

F (40) (20) (20)

2 HPS IIfamily

M 50–100 (30) Multiple (24) Multiple (6)

3a HPS IIfamily

M (21) (19) (2)

4a M Multiple (8) Multiple (8) 3 (Not retrieved5a F Multiple (43) Multiple Multiple

6a M Multiple �100 Multiple Multiple7a M Multiple (54) Multiple (46) Multiple (8)8a HPS Ifamily

F Multiple (33) Multiple (9) Multiple (24)

9a HPS Ifamily

F Multiple (20) Multiple (10) Multiple (10)

0 F 100 12 (no Bx) 90 (35)1 F 20 (17) 10 (7) 10 (10)2 M Multiple �30 (8) �20 (6) Multiple (2)3 M 15–20 (15) 15 (12) (3)

4 M Multiple (32) Multiple (23) Multiple (9)

5 F 15 (4) 15 (4) 06 M (10) (5) (5)7 F �40 (10) �30 (4) (6)8 F 17 (7) 15 (5) (2)

x, biopsy; R/S, rectosigmoid; AC, ascending colon; TC, transverseerrated adenoma; ADMP, admixed hyperplastic adenomatous polypPatients who completed MBD4 testing.

or HPS (Table 1). The remaining 6 patients (patients 3, (

, 11, 35, 36, 38) had large, proximal HPs, or multipleAs (Table 3). Thirty seven patients (97%) had HPsroximal to the rectosigmoid. Twenty-seven patients

s �10 mmax size mm)

Site of HP�10 mm

AdenomasBx SA/ADMP

CRCsite

2 (14) Sigmoid 1 00 0 02 (15) HF 0 01 (12) TC 1 0 DC0 2 1 SA DC1 (10) Caecum 0 00 0 1 SA: TC, 30 mm Sigmoid0 3 00 0 0

15 (20) Pancolonic 2 00 11 1 ADMP TC

ltiple (20) Pancolonic 12 0 TC2 (10) AC, TC 2 00 1 00 0 03 (10) Sigmoid 2 00 3 1 SA: rectum,

15 mmDC

0 40 02 (12) DC, R/S 1 1 SA: TC, 7 mm

1 ADMP: rectum,15 mm

0 5 00 5 0

0 1 0

0 1 0

2 (10) Sigmoid 1 0certain (15) Uncertain Multiple

(5)3 ADMPs, 3 rectal

SAs0 Many 01 (12) Sigmoid 5 1 ADMP 2 mm TC0 8 2 ADMPs:

caecum, AC1 SA R/S

1 (10) TC 3 0 AC

0 3 1 SA rectum SF1 (10) TC 0 00 2 00 1 1 SA 10 mm

sigmoid0 15 5 SAs in

colectomyspecimen

1 ADMP R/S

Rectum

0 0 1 SA: 10 mm TC AC0 22 0 Caecum1 (10) R/S 1 0

10 1 SA 8 mm R/S

; DC, descending colon; SF, splenic flexure; HF, hepatic flexure; SA,

HP(M

Mu

)Un

colon.

71%) had more than 5 HPs proximal to the rectosig-

mptt(Tp1(

dcw7�nctpm

PyaPtcda(

hhFc1

cftccHrtan2aoohg

5wFtan

HPS


oid. Most HPs were �5 mm in diameter; however, 15atients (39.5%) had HPs larger than 10 mm, and 9 ofhese had large HPs proximal to the sigmoid. HPs werehe most prevalent lesions (100%); however, 32 patients84%) had concomitant ADMPs, SAs, or adenomas.welve patients (32%) had more than 5 adenomas, andatient 18 had 40 adenomas. At least 1 SA was found in1 patients (29%), and ADMPs were found in 6 patients16%).

CRC was diagnosed in 10 patients, 4 proximal and 6istal. All 10 patients had CRC diagnosed at their initialolonoscopy. Median and mean ages at diagnosis of CRCere 45 and 49 years, respectively, with a range of 31 to1 years. Three of these 10 patients had coexistent HPs10 mm in size. Three patients had more than 5 ade-

omas. Six patients had coexisting SAs. MSI testing wasompleted in 5 polyps and 4 cancers (polyps from pa-ients 2, 10, 18, 26, 27 were tested, and cancers fromatients 4, 5, 7, 12 were tested). All tumors tested wereicrosatellite stable (MSS).All 10 patients with CRC had colonic resections.

atient 7 had a sigmoid colectomy at the age of 31ears for sigmoid cancer and a right hemicolectomy for30-mm SA in the transverse colon 10 years later.

atient 12 had an extended left hemicolectomy for aransverse colon cancer with subsequent right hemi-olectomy for a tubulovillous adenoma with severeysplasia. Three patients (7.9%) had total colectomynd ileorectal anastomosis for management of HPS

Figure 1.

patients 10, 25, and 26). Twenty-four patients (63%) w

ad family history of CRC. Nineteen patients (50%)ad one or more first-degree relatives with CRC.amily history of HPS was very uncommon in ourohort of patients, with only 2 families found (Figuresand 2).In family 1, patients 28 and 29 had a family history of

onsanguinity (Figure 1). Patient 29 was a 73-year-oldemale who developed cancer of the ascending colon athe age of 67 years. She also had 40 HPs throughout theolon. Three adenomas were biopsied from the ascendingolon. Patient 28 was a 71-year-old female who had 70Ps throughout the colon, as well as 1 SA in the

ectosigmoid, 1 ADMP in the rectum, and 1 ADMP inhe caecum. Eight adenomas were removed from thescending and transverse colons. A 2-cm duodenal ade-oma requiring polypectomy was also found. Patient9’s son has mixed polyposis with 4 adenomas in caecumnd descending colon and 1 HP in sigmoid. The presencef the duodenal adenoma in patient 29 led to a suspicionf attenuated familial adenomatous polyposis (AFAP);owever, extensive adenomatous polyposis coli (APC)ene testing including deletion studies was negative.In family 2, patient 21 was a 78-year-old lady who hadadenomas plus 40 biopsy-proven HPs, 20 of which

ere located proximal to the rectosigmoid (Figure 2).ive of her siblings were affected: 3 with CRCs while inheir 60s, 1 with a severely dysplastic tubular adenoma,nd 1 with multiple rectosigmoid polyps and an ade-oma at the splenic flexure. Patient 21 had 4 sons, 2 of

family 1.

hom expressed the HPS phenotype. Patient 22 was a

5wPs

tMpighScncI

alwtbmm

OhtclaamfljMy

Mipb

HPS

T

IE

E

I3

3


1-year-old male who had an estimated 50–100 HPs,ith 30 biopsied. A single adenoma was also found.atient 23 had 21 HPs, 2 of which were proximal to theigmoid.

MBD4 Testing

Twenty-seven of 38 patients had MBD4 mutationesting completed (Table 3). No pathogenic mutations inBD4 were detected. Six single nucleotide polymor-

hisms (SNP) of uncertain functional significance weredentified (Table 4). Four of 27 patients were heterozy-ous for IVS2�27 T¡C. Four of 27 patients wereeterozygous for exon 3 nucleotide (nt) 993G¡A. ThisNP causes an amino acid substitution, Ala to Thr, atodon 273. Five of 27 patients were heterozygous atucleotide 1589 C¡T in exon 6 with no amino acidhange. Six of 27 patients were heterozygous forVS6�14C¡T. Nine of 27 patients were heterozygous,

Figure 2.

able 4. MBD4 Testing Results

Single nucleotidepolymorphism Patient sample Genotype

VS2�27 T�TC (rs140692) 7, 9, 26, 29 TCxon 3 nt 993G�GA,A273T (rs10342)

7, 9, 26, 29 GA

x 6 nt 1589 C�CT(rs140696)

7, 9, 17, 26, 29 CT

VS6�14 C�CT (rs140697) 7, 9, 17, 18, 26, 29 CT=UTR 2302 A�AG(rs2307285)

3, 10, 11, 12, 13, 15, 1721, 23—HPS family II

AG

2, 5, 628, 29—HPS family I

GG

= UTR 2366 G�GC 7, 9, 26, 29 GC

r(rs3138366)

nd 5 of 27 patients were homozygous for 3=-untrans-ated region (UTR) nt 2302 A¡G. Four of 27 patientsere heterozygous for 3=-UTR nt 2366 G¡C. None of

hese SNPs were considered to be of clinical significanceecause they would not cause frame shifts or truncatingutations and were unlikely to cause significant confor-ational changes in the protein.

MYH Testing

All 38 patients had complete testing for MYH.ne patient (patient 18) was found to be a Y165Comozygote. This was a 54-year-old female with morehan 150 polyps. Multiple HPs were noted at 10olonoscopies. Twenty-seven were biopsy-proven HPs, ateast 6 of which were proximally located. She also had 40denomas scattered throughout the colon. Her sister diedt the age of 36 years from CRC in the setting ofultiple adenomas. A brother had 1 HP at the hepatic

exure at age 46 years. Of her 3 daughters, 1 had 5 rectaluvenile polyps at the age of 26 years. These polyps were

SS. Another daughter had rhabdomyosarcoma at age 7ears (Figure 3).

Discussion

HPS

HPS was first described in the late 1970s.3–5 Aedline search revealed to date 31 papers published

dentifying patients with HPS (Table 5). Among theublished data, there was no consensus regarding num-ers of polyps needed for the diagnosis of HPS. The cases

family 2.

anged from 1 HP to those with hundreds of HPs.

Eom

rrpesSosfmfistqdwm5(nCsctCol

(ttbo

uwi(ohlsHg

oisealCtfpai

bialle


ighty-four percent of our cohort of 38 patients devel-ped simultaneous HPs, ADMPs and SAs, and adeno-as, similar to other published series (Table 5).6–8

Considerable evidence supports the concept of a ser-ated neoplasia pathway of colorectal tumorigenesis. Ser-ated polyps demonstrate serrated or “sawtoothed” ap-earance on histology because of infolding of the cryptpithelium. The classification is complex and controver-ial, covering a spectrum of lesions ranging from HPs toAs and ADMPs. SAs have the sawtoothed architecturef HPs, but the cells show histologic features of dyspla-ia. Some experts believe that SAs and ADMPs developrom HPs and are precursors to cancer with a variableicrosatellite stability profile.9–12 Studies are yet to de-

ne precisely the burden of CRC attributable to theerrated neoplasia pathway. Dysplastic lesions arisinghrough this pathway may have different rates and fre-uency of progression to malignancy. Williams et alescribed 7 patients, all with more than 50 HPs whoere followed for 3 to 13 years, with none developingalignancy.13 Subsequent case series report a more than

0% incidence of CRC in patients with HPS6,7,14–16

Table 5), although the true risk in unselected cases isot known. Ten (26%) of our cohort of 38 patients hadRC, but, because our series is not a population-based

eries of HPS, with a bias therefore toward ascertainingases presenting with symptomatic CRC, we cannot con-ribute to data on the risk of CRC in HPS. Patients withRC complicating HPS have been found to have featuresf early age of onset, multiplicity, and proximal colon

Figure 3. Pedigree of the

ocation and are either MSI-high (MSI-H) or MSI-low i

MSI-L)/MSI stable (MSS).6,7,9,12,17 In our patients, MSIesting in 5 polyps and 4 cancers revealed that all tumorsested were MSS. Absence of MSI in our series is notableut not inconsistent with the variability described inther series (Table 5).

An inherited genetic basis of HPS is likely, given thenusual clinical features of HPS. Familial aggregationith apparent dominant inheritance of HPS is exceed-

ngly rare, with only 5 case reports published to date7,8,17

excluding those in the current series). These case reportsf familial HPS suggest possible genetic predisposition;owever, information regarding genetic changes remainsimited, given the relative rarity of the syndrome. In oureries, one family with apparent dominant inheritance ofPS and one with a recessive pattern (including consan-

uinity) were identified.Although it may be impossible for the gastroenter-

logist to remove every minute polyp, especially thosen the distal colon and rectum, special attentionhould be paid to those with certain clinical andndoscopic features that indicate “high-risk” HPs suchs multiple (�20), large (�10 mm), and proximalocation and those with a family history of CRC.linicians should count the number of polyps each

ime a patient has a colonoscopy and submit all polypsor histologic examination if possible. It is also im-ortant that the endoscopist traces back the numbernd histopathology of all previous polyps to providenformed management. Pathologists should be famil-

lic MYH mutation carrier.

ar with the characteristics of SAs, including sessile

sd

ftpiafttt

lchhc

fioast

T


errated polyps and ADMPs, which seem to be highlyiscriminative for HPS.The current endoscopic surveillance recommendation

or the individual with HPS is 1 to 3 yearly examina-ions, depending on the number, size, and histology ofolyps.18 Proximal polyps should be resected completelyf possible because adenomatous tissue may be presentnd some of these polyps may be SAs.19 Surveillanceollowing polypectomy is then warranted, and, althoughhe precise frequency is unknown, it has been suggestedhat “high-risk” HPs and SAs should be approached in

able 5. Published Series of HPS

Reference Cases of HPS HPs Adenom

37 1 30 No5 9 Multiple No3 1 30 No4 1 Multiple Yes

13 7 50–156 Yes38 1 Hundreds No39 1 Multiple No40 1 1 No41 1 15 No42 1 20 Yes43 1 0 Yes44 1 28 No45 1 60 Yes46 1 18 Yes47 2 19–61 No48 2 12–18 Yes49 1 16 Yes50 1 Multiple Yes51 1 Multiple Yes52 6 �50–100 No17 5 1 to multiple Yes

53 3 Multiple to 68 No54 1 200 Yes55 1 �100 Not repo14 6 50–100 Yes7 13 Multiple to �100 Yes

9 4 6–50 Yes

12 1 75 No6 12 �15 to �100 Yes

16 1 Multiple Yes1 28 �20 Yes

18 15 15–207 Yes15 13 �20 Yes8 14 19 to �100 Yes

he same way as adenomas.20 Australian NHMRC guide- s

ines recognize the risk, recommending biannualolonoscopy. Colectomy may be justified in those with aigh number of HPs, concomitant SAs, or multipleigh-risk adenomatous lesions (more than 1 cm, villousomponent, high-grade dysplasia).8

Screening colonoscopies should be recommended forrst-degree relatives of affected individuals, independentf the presence of cancer in the proband. The appropriatege for commencement of screening relatives and thecreening interval remain matters for debate. Our prac-ice follows the Australian NHMRC recommendation to

SAs or MPs CRC MSI status (polyp or cancer)

Yes NoYes YesNo NoNo YesNo NoYes NoNo NoNo YesNo NoYes YesYes YesYes YesYes YesYes YesNo YesNo YesNo YesNo YesNo NoYes YesYes Yes MSI H: 2 cancers, 1 20 mm HP all

in 1 individualNo YesYes YesNo YesNo YesYes Yes MSI H: 4/129 HPs

MSI L: 2/129 HPs, 1/6 SAs, 1/3MPs

Yes Yes MSI H: 6/11 cancers, 4/13dysplastic foci

MSI L: 2/11 cancers, 3/13dysplastic foci, 1/13 HPs

MSS: 12/13 HPs, 6/13 dysplasticfoci, 3/11 cancers

Yes Yes MSS in cancer, SA, HPsYes Yes MSI H: 1/47 HPs, 2/8 cancers

MSI L: 2/47 HPs, 2/8 SAsMSS: 44/47 HPs, 1/1 MP, 6/8

SAs, 2/2 adenomas, 6/8cancers

No YesYes YesYes YesYes YesYes Yes

as

rted

creen first-degree relatives of affected subjects with

5Hyf

mlncsttmCMCCdtfltitcblCMaJMf

pehfty

�tacpabus

vmodssyiafblnMe

ch4Wem

HhftTahMmda

cHhcpimandtse

oe


-year colonoscopy, increasing if adenomas or multiplePs are detected. Starting age should be 40 years or 10

ears earlier than the earliest age at diagnosis in theamily.

Molecular Basis of HPS

MBD4 gene. MBD4 (human DNA repair proteinethyl-CpG binding endonuclease 1) is 1 of 5 mamma-

ian methyl binding domain family proteins.21 Sponta-eous hydrolytic deamination of 5 methylcytosine andytosine to thymine and uracil, respectively, occur con-tantly. If not corrected, this will lead to G:C to A:Transition mutations in the next round of DNA replica-ion, contributing to tumorigenesis. Nearly 50% of so-atic mutations of the tumor suppressor gene TP53 inRC are G:C to A:T transitions within CpG sites.22

BD4 functions as a caretaker of genomic fidelity atpG sites by surveying methylated and unmethylatedpG sites in the genome and excising spontaneouslyeaminated cytosine (ie, uracil) or methylcytosine (ie,hymine).23–25 Mutations causing a loss of normal MBD4unction with or without concurrent APC mutation mayead to an increase in genome-wide G:C to A:T transi-ions and thus promote tumourigenesis via genomicnstability.26,27 The MBD4 gene itself is mutated in 26%o 43% of human colorectal, endometrial, and pancreaticarcinomas that exhibit MSI.28,29 There is evidence foriallelic inactivation of MBD4 in human cancers withoss of heterozygosity at the MBD4 locus detected inRC specimens.29 One such family with a germ-lineBD4 has been described, with 1 family member having

n MSI-H cancer with extensive DNA methylation (JassR, personal communication). We considered thatBD4 is a strong candidate gene for testing in HPS

amilies.MBD4 testing in our cohort of patients yielded several

olymorphisms of uncertain functional significance. Ourvidence does not support a role for MBD4 in HPS;owever, pathogenic mutations in the gene cannot beully discounted without further experience relating tohe variants we identified (for example, segregation anal-sis) and studies to exclude deletions.

MYH gene. Approximately 30% of patients with15 adenomas that do not carry pathogenic APC mu-

ations are biallelic MYH mutation carriers.30 MYH-ssociated polyposis (MAP) is difficult to differentiatelinically from FAP or attenuated familial adenomatousolyposis (AFAP). MAP tends to present later with meannd median ages in the mid-50s years. Adenoma num-ers range from 5 to hundreds. These patients do notsually develop the thousands of adenomas seen in clas-
ical FAP. Small, mildly dysplastic tubular and tubulo- s
illous adenomas predominate, and HPs are also com-on.31,32 The majority of affected patients develop CRC,

ften at the time of presentation. They do not appear toiffer in stage, grade, or histologic appearance fromporadic CRCs. Extracolonic features are not well de-cribed, but a small proportion develops duodenal pol-posis. Northern European populations have a predom-nance of Y165C and G382D mutations (82% of mutantlleles).30,33–35 Other pathogenic variants have beenound in other ethnic groups. No unaffected carriers ofiallelic MYH mutations have been reported in pub-ished studies. However, population-based studies haveot been done to date. The increased risk, if any, of singleYH mutation carriers is difficult to estimate. A recent

stimate placed the risk at 1.68 (95% CI: 1.07–2.95).36

In our patients with HPS, 1 biallelic MYH mutationarrier was identified. This patient had a recessive familyistory of adenomas and CRC but also was found to have0 adenomas and more than 20 HPs, satisfying theHO HPS diagnostic criteria. This association is inter-

sting, and further studies focusing on the role of MYHutations in HPS may be informative.In summary, among this series of 38 patients withPS, SAs are common (26%). Half of the 38 patients

ave a first-degree relative with CRC. Only 2 cases ofamilial aggregation with HPS were described. Ten pa-ients developed CRC at a young mean age of 49 years.hree patients required total colectomy and ileo-rectalnastomosis for polyposis, and all 10 patients with CRCsad lesser colonic resections. No pathogenic mutations inBD4 were detected in 27 patients tested, but 6 poly-orphisms occur commonly and require further study to

etermine significance. Pathogenic biallelic MYH vari-nts were detected in 1 patient.

There is some evidence that serrated polyps form aontinuum from hyperplastic aberrant crypt foci throughPs, ADMPs and SAs, culminating not only in MSI-

igh CRC but possibly also a subset of MSI-low/MSSancers. Tumors arising from this proposed serrated neo-lasia pathway may have marked molecular heterogene-ty. In our study of patients with HPS, many exhibitixed polyposis, and there is a lack of evidence for

ssociation with MSI-high CRC. MBD4 mutations doot contribute to the pathogenesis of HPS; however,eletion studies were not performed, and some uncer-ainties exist regarding the pathogenicity of the unclas-ified variants. MYH may have a role to play in HPS,ither as an initiator or as a genetic modifier.

The clinical, histopathologic, and molecular findingsf this study contribute to our understanding of HPS andxplore the link to CRC that may evolve through the
errated neoplasia pathway. The rarity of HPS calls for

lgamc

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

2

2

2

3

3

3

3


arge studies with pooled series to determine the demo-raphic, clinical, genetic, morphologic, and molecularspects of serrated polyps including HPs and will deter-ine the appropriate management strategies with spe-

ific surveillance program and/or chemoprevention.

References1. Renaut AJ, Douglas PR, Newstead GL. Hyperplastic polyposis of

the colon and rectum. Colorectal Dis 2002;4:213–215.2. Burt RW, Jass JR. Hyperplastic polyposis. In: Hamilton SR, Aal-

tonen LA, eds. World Health Organisation classification of tu-mours. Pathology & genetics. Tumours of the digestive system.Lyon: IARC Press, 2000.

3. Cooke SA. Polyposis coli. The clinical spectrum in adults. S AfrMed J 1978;53:454–457.

4. Cooper HSPA, Marks G. Adenomatous and carcinomatouschanges within hyperplastic colonic epithelium. Dis Colon Rec-tum 1979;22:152–156.

5. Spjut H, Estrada R. The significance of epithelial polyps of thelarge bowel. Pathol Ann 1977;12:147–170.

6. Leggett BA, Devereaux B, Biden K, Searle J, Young J, Jass J.Hyperplastic polyposis: association with colorectal cancer. Am JSurg Pathol 2001:177–184.

7. Rashid A, Houlihan PS, Booker S, Petersen GM, Giardello FM,Hamilton SR. Phenotypic and molecular characteristics of hyper-plastic polyposis. Gastroenterology 2000;119:323–332.

8. Lage P, Cravo M, Sousa R, Chaves P, Salazar M, Fonseca R, ClaroI, Suspiro A, Rodrigues P, Raposo H, Fidalgo P, Nobre-Leitao C.Management of Portuguese patients with hyperplastic polyposisand screening of at-risk first degree relatives: a contribution forfuture guidelines based on a clinical study. Am J Gastroenterol2004;99:1779–1784.

9. Jass JR, Iino H, Ruszkiewicz A, Painter D, Solomon MJ, Koorey DJ,Cohn D, Furlong KL, Walsh MD, Palazzo J, Edmonston TB, FishelR, Young J, Leggett BA. Neoplastic progression occurs throughmutator pathways in hyperplastic polyposis of the colorectum.Gut 2000;47:43–49.

0. Jass JR, Whitehall VL, Young J, Leggett BA. Emerging concepts incolorectal neoplasia. Gastroenterology 2002;123:862–876.

1. Iino H, Jass JR, Simms LA, Young J, Leggett B, Ajioka Y, Wa-tanabe H. DNA microsatellite instability in hyperplastic polyps,serrated adenomas, and mixed polyps: a milder mutator pathwayfor colorectal cancer? J Clin Pathol 1999;52:5–9.

2. Hawkins NJ, Gorman P, Tomlinson IP, Bullpitt P, Ward RL. Colo-rectal carcinomas arising in the hyperplastic polyposis syndromeprogress through the chromosomal instability pathway. Am JPathol 2000;157:385–392.

3. Williams GT, Arthur JF, Bussey HJ, Morson BC. Metaplastic pol-yps and polyposis of the colorectum. Histopathology 1980;4:155–170.

4. Place RJ, Simmang CL. Hyperplastic-adenomatous polyposis syn-drome. J Am Coll Surg 1999;188:503–507.

5. Hyman NH, Anderson P, Blasyk H. Hyperplastic polyposis and therisk of colorectal cancer. Dis Colon Rectum 2004;47:2101–2104.

6. Koide N, Saito Y, Fujii T, Kondo H, Saito D, Shimoda T. A case ofhyperplastic polyposis of the colon with adenocarcinomas inhyperplastic polyps after long-term follow-up. Endoscopy 2002;34:499–502.

7. Jeevaratnam P, Cottier DS, Browett PJ, Van De Water NS, PokosV, Jass JR. Familial giant hyperplastic polyposis predisposing tocolorectal cancer: a new hereditary bowel cancer syndrome.J Pathol 1996;179:20–25.

8. Ferrandez A, Samowitz W, DiSario JA, Burt RW. Phenotypic char-
acteristics and risk of cancer development in hyperplastic polyp-
osis: case series and literature review. Am J Gastroenterol2004;99:2012–2018.

9. Rex DK, Ulbright TM. Step section histology of proximal colonpolyps that appear hyperplastic by endoscopy. Am J Gastroen-terol 2002;97:1530–1534.

0. Huang CS, O’Brien MJ, Yang S, Farraye FA. Hyperplastic polyps,serrated adenomas, and the serrated polyp neoplasia pathway.Am J Gastroenterol 2004;99:2242–2255.

1. Hendrich B, Abbott C, McQueen H, Chambers D, Cross S, Bird A.Genomic structure and chromosomal mapping of the murine andhuman MBD1, MBD2, MBD3, and MBD4 genes. Mamm Genome1999;10:906–912.

2. Kleihues P, Schauble B, zur Hausen A, Esteve J, Ohgaki H.Tumours associated with p53 germline mutations: a synopsis of91 families. Am J Pathol 1997;150:1–13.

3. Hendrich B, Hardeland U, Ng HH, Jiricny J, Bird A. The thymineglycosylase MBD4 can bind to the product of deamination atmethylated CpG sites. Nature 1999;401:301–304.

4. Petronzelli F, Riccio A, Markham GD, Seeholzer SH, Stoerker J,Genuardi M, Yeung AT, Matsumoto Y, Bellacosa A. Biphasickinetics of the human SNA repair protein MED1 (MBD4), a mis-match specific DNA N-Glycosylase. J Biol Chem 2000;275:32422–32429.

5. Petronzelli F, Riccio A, Markham GD, Seeholzer SH, Genuardi M,Karbowski M, Yeung AT, Matsumoto Y, Bellacosa A. Investigationof the substrate spectrum of the human mismatch-specific DNAN-glycosylase MED1(MBD4): fundamental role of the catalyticdomain. J Cell Physiol 2000;185:473–480.

6. Millar CB, Guy J, Sansom OJ, Selfridge J, MacDougall E, HendrichB, Keightley PD, Bishop SM, Clarke AR, Bird A. Enhanced CpGmutability and tumorigenesis in MBD4 deficient mice. Science2002;297:403–405.

7. Wong E, Yang K, Kuraguchi M, Werling U, Avdievich E, Fan K,Fazzari M, Jin B, Brown AM, Lipkin M, Edelmann W. MBD4inactivation increases C¡T transition mutations and promotesgastrointestinal tumour formation. Proc Natl Acad Sci U S A2002;99:14937–14942.

8. Bader SWM, Hendrich B, Bird A, Bird C, Hooper M, Wyllie A.Somatic frameshift mutations in the MBD4 gene of sporadiccolon cancers with mismatch repair deficiency. Oncogene 1999;18:8044–8047.

9. Riccio AAL, Godwin AK, Loukola A, Percesepe A, Salovaara R,Masciullo V, Genuardi M, Paravatou-Petsotas M, Bassi DE, Rug-geri BA, Klein-Szanto AJP, Testa JR, Neri G, Bellacosa A. The DNArepair gene MBD4 (MED1) is mutated in human carcinomas withmicrosatellite instability. Nat Genet 1999;23:266–268.

0. Sieber OM, Lipton L, Crabtree M, Heinimann K, Fidalgo P, PhillipsRK, Bisgaard ML, Orntoft TF, Aaltonen LA, Hodgson SV, ThomasHJ, Tomlinson IP. Multiple colorectal adenomas, classic adeno-matous polyposis, and germ-line mutations in MYH. N Engl J Med2003;348:791–799.

1. Sampson JR, Dolwani S, Jones S, Eccles D, Ellis A, Evans DG,Frayling I, Jordan S, Maher ER, Mak T, Maynard J, Pigatto F, ShawJ, Cheadle JP. Autosomal recessive colorectal adenomatous pol-yposis due to inherited mutations of MYH. Lancet 2003;362:39–41.

2. Lipton L, Halford SE, Johnson V, Novelli MR, Jones A, CummingsC, Barclay E, Sieber O, Sadat A, Bisgaard ML, Hodgson SV,Aaltonen LA, Thomas HJ, Tomlinson IP. Carcinogenesis in MYH-associated polyposis follows a distinct genetic pathway. CancerRes 2003;63:7595–7599.

3. Gismondi V, Meta M, Bonelli L, Radice P, Sala P, Bertario L, VielA, Fornasarig M, Arrigoni A, Gentile M, Ponz De Leon M, AnselmiL, Mareni C, Bruzzi P, Varesco L. Prevalence of the Y165C,G382D and 1395delGGA germline mutations of the MYH gene inItalian patients with adenomatous polyposis coli and colorectal
adenomas. Int J Cancer 2004;109:680–684.

3

3

3

3

3

3

4

4

4

4

4

4

4

4

4

4

5

5

5

5

5

5

Ge

Po


4. Enholm S, Hienonen T, Suomalainen A, Lipton L, Tomlinson I,Karja V, Eskelinen M, Mecklin JP, Karhu A, Jarvinen HJ, AaltonenLA. Proportion and phenotype of MYH-associated colorectal neo-plasia in a population-based series of Finnish colorectal cancerpatients. Am J Pathol 2003;163:827–832.

5. Jones S, Emmerson P, Maynard J, Best JM, Jordan S, WilliamsGT, Sampson JR, Cheadle JP. Biallelic germline mutations inMYH predispose to multiple colorectal adenoma and somaticG:C¡T:A mutations. Hum Mol Genet 2002;11:2961–2967.

6. Tenesa ABR, Wiltshire A, Prendergast J, Porteous M, Campbell H,Dunlop MG, Farrington SM. Inherited base excision repair defectspredispose to colorectal cancer. Fam Cancer 2005;4:33.

7. Goldman H, Ming S, Hickock DF. Nature and significance ofhyperplastic polyps of the human colon. Arch Pathol 1970;89:349–354.

8. Sumner HW, Wasserman NF, McClain CJ. Giant hyperplasticpolyposis of the colon. Dig Dis Sci 1981;26:85–89.

9. Cohen SM, Brown L, Janower ML, McCready FJ. Multiple meta-plastic (hyperplastic) polyposis of the colon. Gastrointest Radiol1981;6:333–335.

0. Franzin G, Novelli P. Adenocarcinoma occurring in a hyperplastic(metaplastic) polyp of the colon. Endoscopy 1982;14:28–30.

1. Tulman AB, Bradford S, Lee E, Brady PG. Giant hyperplasticpolyps associated with vasculitis of the colon. J Fla Med Assoc1982;69:380–383.

2. Yamada N FS, Ogawa T, Nishikage S, Adachi M, Muto T. Multiplemetaplastic polyps of the colon associated with adenomas com-plicating focal carcinoma, report of a case. Stomach Intestine1983;18:529–530.

3. Urbanski SJ, Kossakowska AE, Marcon N, Bruce WR. Mixedhyperplastic adenomatous polyps—an underdiagnosed entity.Report of a case of adenocarcinoma arising within a mixedhyperplastic adenomatous polyp. Am J Surg Pathol 1984;8:551–556.

4. Bengoechea O, Martinez-Penuela JM, Larrinaga B, Valerdi J, BordaFJ. Hyperplastic polyposis of the colorectum and adenocarcinoma ina 24-year-old man. Am J Surg Pathol 1987;11:323–327.

5. McCann BG. A case of metaplastic polyposis of the colon asso-ciated with focal adenomatous change and metachronous ade-
nocarcinomas. Histopathology 1988;13:700–702. H
6. Teoh HH, Delahunt B, Isbister WH. Dysplastic and malignantareas in hyperplastic polyps of the large intestine. Pathology1989;21:138–142.

7. Kusunoki M, Fujita S, Sakanoue Y, Shoji Y, Yanagi H, YamamuraT, Utsunomiya J. Disappearance of hyperplastic polyposis afterresection of rectal cancer. Report of two cases. Dis Colon Rec-tum 1991;34:829–832.

8. Shepherd NA. Inverted hyperplastic polyposis of the colon. J ClinPathol 1993;46:56–60.

9. Warner AS, Glick ME, Fogt F. Multiple large hyperplastic polyps ofthe colon coincident with adenocarcinoma. Am J Gastroenterol1994;89:123–125.

0. Lieverse RJ, Kibbelaar RE, Griffioen G, Lamers CB. Colonic ade-nocarcinoma in a patient with multiple hyperplastic polyps. NethJ Med 1995;46:185–188.

1. Beusnel C, Le Berre N, Pagenault M, Luherne Y, Bargain A,Ramee MP, Gosselin M, Bretagne JF. Giant hyperplastic polypo-sis with adenomatous tissue. Gastroenterol Clin Biol 1996;20:294–297.

2. Torlakovic E, Snover DC. Serrated adenomatous polyposis inhumans. Gastroenterology 1996;110:748–755.

3. Jorgensen H, Mogensen AM, Svendsen LB. Hyperplastic polypo-sis of the large bowel. Three cases and a review of the literature.Scand J Gastroenterol 1996;31:825–830.

4. Orii S, Nakamura S, Sugai T, Habano W, Akasaka I, Nakasima F,Kazama H, Hasimoto Y, Takahasi H, Sugawara M, Sato S. Hy-perplastic (metaplastic) polyposis of the colorectum associatedwith adenomas and an adenocarcinoma. J Clin Gastroenterol1997;25:369–372.

5. Keljo DJ, Weinberg AG, Winick N, Tomlinson GA. Rectal cancer inan 11-year-old girl with hyperplastic polyposis. J Pediatr Gastro-enterol Nutr 1999;28:327–332.

Received November 30, 2005. Accepted March 23, 2006.Address requests for reprints to: Elizabeth Chow, MD, Department of

astroenterology, Flinders Medical Centre, Adelaide, South Australia.-mail: [email protected]; fax: (61) 08 8267 3562.Supported by The Hicks Foundation Scholarship; Edith Viola Reid

ostgraduate Medical Scholarship, Faculty of Medicine, the Universityf Melbourne; and The Victor Hurley Grant, the Royal Melbourne
ospital (to E.C.).

hyperplastic polyposis syndrome: phenotypic presentations and the role of mbd4 and myh

Documents