juvenile polyposis coli - journal of medical geneticsclinical features and pathology of our cases...

7.med. Genet. (I966). 3, 5.

Juvenile Polyposis ColiA. M. 0. VEALE, I. McCOLL, H. J. R. BUSSEY, and B. C. MORSON

From the Research Department, St. Mark's Hospital, London

The term 'polyposis coli' applies to a conditionof multiple polyps of the large intestine and isused synonymously with 'familial polyposis' and'multiple adenomatosis'. The latter name empha-sizes the fact that the individual polyps of thisdisease in no way differ from adenomas observedas solitary lesions.The object of this paper is to describe the histo-

pathology, clinical features, and family histories ofii patients with polyposis in whom the polypswere not adenomas but had the structure of juvenilepolyps. The latter are often confused with adeno-mas despite clear differences in pathology andbehaviour.

Juvenile polyps, like adenomas, are usuallypresent as solitary lesions, but just as adenomatouspolyps may be present in large numbers (adeno-matous polyposis coli), so the large intestinalmucosa may be covered with multiple polyps ofthe juvenile variety, henceforward called 'juvenilepolyposis coli'.

Large intestinal polyposis has been described inchildren by Kerr (I948), Gordon, Hallenbeck,Dockerty, Kennedy, and Jackman(I957), and Hines,Hanley, Ray, and Rush (I959), but these authors,wrongly in our opinion, referred to the polyps as'adenomas'. Castro (1955), MalIam and Thomson(I959), and Knox, Miller, Begg, and Zintel (I960)also reported cases of colonic polyposis in children,but did not describe any investigation of the familyhistories of their patients. A summary of theclinical features and pathology of our cases hasbeen published elsewhere (McColl, Bussey, Veale,and Morson, I964).

Diagnosis of Juvenile PolypThe typical adenomatous polyp (Fig. i) has a

lobulated surface appearance with intercommuni-cating cracks or crevices separating the lobules.In contrast the juvenile polyp (Fig. 2) iS charac-

Received May 13, I965.

teristically smooth in its surface configuration.Occasionally juvenile polyps have a papillary orlobulated surface pattern which makes any precisedistinction between the two types of tumourunreliable on simple inspection alone. However,the cut surface of juvenile polyps usually showsmultiple cysts filled with mucin (Fig. 2), hence thename 'retention polyp' sometimes applied to thesetumours. Cystic changes in true adenomas areexceptional. The stalks of juvenile polyps appearno different from those of adenomas, except thatin surgical specimens containing multiple juvenilepolyps there may be some stalks without tumours.

FIG. 3i. Adenomatous polyp: the tumour is composed of lobulesseparated from one another by intercommunicating crevices.

5

on April 11, 2021 by guest. P

rotected by copyright.http://jm

g.bmj.com

/J M

ed Genet: first published as 10.1136/jm

g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

Veale, McColl, Bussey, and Morson

FIG. 2. Juvenile polyp: the surface of the tumour is characteristicallysmooth in its surface outline. The cut surface shows numerous cystsfilled with mucin.

In such cases it may be presumed that the tumourhas fallen off the stalk, a process called 'autoampu-tation', which is characteristic of juvenile polyps(Lane, I865). A possible explanation for thisbehaviour is the absence of any participation bythe muscularis mucosa in the structure of juvenilepolyps.The size of juvenile polyps varies from small

sessile nodules a few millimetres in diameter topedunculated tumours about 3 cm. across. Theyare often very haemorrhagic in appearance, whichcould be due to infarction from twisting of theirpedicles.

Microscopically, both juvenile and adenomatouspolyps have an epithelial and a connective tissueelement. The muscularis mucosae seems to playno part in the growth of juvenile polyps, but inadenomatous polyps smooth muscle fibres aresometimes found in the stroma. In an adenomathe proliferating epithelial tubules are packedclosely together, being separated by supportingtissue which is morphologically similar to normallamina propria (Fig. 3 and 4). The epitheliumshows a diminution in the number of goblet cells,crowding ofthe nuclei with much hyperchromatism,and an increased number of mitotic figures, allfeatures that are characteristic of a neoplasm. Incontrast the epithelial lining of the tubules in

FIG. 3. Adenomatous polyp: the proliferating tubules are packedclosely together and there is a well-marked lobulated surfaceconfiguration. (H. and E. x 5.)

6



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

juvenile Polyposis Coli

FIG. 4. Adenomatous polyp: the closely packed epithelial tubulesshow diminution in the number of goblet cells, crowding of nuclei,hyperchromatism, and an increased number of mitoses. The tubulesare separated by lamina propria which is qualitatively and quanti-tatively normal. (H. and E. x ioo.)

juvenile polyps (Fig. 5 and 6) consists of cells thatshow little sign of increased nuclear activity suchas hyperchromatism and mitosis, although theremay be some diminution in the number of gobletcells. The tubules readily become dilated byretained mucus leading to flattening of the epithe-lial lining.The most striking feature of juvenile polyps is

the greatly increased amount of supporting connec-tive tissue compared with adenomas. The epithelialtubules are widely separated by a fine areolarnetwork of reticulin fibres and connective tissuecells containing a sprinkling of lymphocytes andplasma cells.True bone has been reported in the stroma of

juvenile polyps (Todd, I963; Marks and Atkinson,I964), probably the result of metaplasia, andlymphoid tissue with central germinal follicles isregularly seen.The surface of juvenile polyps may be covered

by a single layer of normal-appearing columnarepithelium. This is readily breached, probably bytrauma from the movement of faeces, leading tosecondary infection of the tumour. In such cases,there is a greatly increased number of lymphocytes

and plasma cells in the connective tissue and evenmicro-abscess formation. Haemorrhage into thesubstance of these tumours also causes changes thatreadily obscure the correct diagnosis. In particularinfarction from twisting of the pedicle may convertthe juvenile polyp into a shadow of its former self.

In those juvenile polyps that have a pronouncedpapillary surface appearance the microscopicalstructure shows some variation from that usuallyseen. The excess of supporting connective tissueis not so great, though the tubules are still muchmore widely spaced than in normal mucosa or inadenomatous polyps. In such cases, however, itcan be seen that the epithelial cells are normaland show little evidence of hyperplasia or neoplasia.The diagnosis of juvenile polyp rests not only onthe excess of connective tissue, but also on theabsence of those epithelial changes that are charac-teristic of benign neoplastic proliferation.

Materials and MethodsThe pathology of I45 patients with polyposis coli in

the records of the research department at St. Mark'sHospital was reviewed. In 6 patients the polyps hadthe microscopical structure of the juvenile type. A

7



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/


4. )]. lE. .*..... .. s - ,4

t 's't..t>,,tA~~4

*... .* -At -g... F. .e .t.

FIG. S. Juvenile polyp: the tumour has a smooth outline and itssubstance contains numerous cystic spaces. (H. and E. x 4.)

seventh patient, for whom there is no histological con-firmation (Case III.3, Family 2), was accepted as suf-fering from juvenile polyposis on clinical and follow-up evidence. These 7 patients belonged to 4 differentfamilies.

In addition 4 other cases of juvenile polyposis wereobtained from the Royal Marsden Hospital (I) andGuy's Hospital (3). These occurred in 4 families, andthus there was a total of ii patients from 8 families forstudy.

v : * t wtv.r:A.

FIG. 6. Juvenile polyp: the epithelial tubules are widely separatedby the supporting connective tissue which has a 'mesenchymal'appearance unlike normal lamina propria. There is cystic dilatationwith flattening of the epithelial lining but no sign of epithelialhyperplasia. (Mucicarmine. x 68.)



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

Juvenile Polyposis Coli

Case HistoriesCase i (female, aged I8 months). She passed bright

red blood with her stools at the age of i8 months.Examination revealed three rectal polyps which couldeasily be made to prolapse through the anus. Thetumours were plucked off with the fingers. Two morepolyps were removed through the sigmoidoscope andbarium enema examination indicated others in thedescending colon. Intermittent rectal bleeding con-tinued, and two more polyps were removed sigmoido-scopically at the age of 4. A year later laparotomy showed2 polyps in the terminal ileum, one each in the caecumand ascending colon, 2 in the transverse colon, and 4 inthe descending colon. At the age of 7 years I2 morerectal polypi were removed, since when there has beenno further bleeding for the past three years.

Histological examination of all the polyps removed,including those from the small intestine, showed themto be juvenile polyps.

Case 2 (female, aged I3 years). She was admitted tohospital when she was 13 years old for appendicectomywhen the caecum was found to be subhepatic. Abdominalpain continued, and laparotomy two weeks later showedoedema of the caecum thought to be due to intussuscep-tion; a right hemicolectomy was performed. Theresected specimen contained numerous polyps, particu-larly in the caecum, all of which had the structure ofjuvenile polyps. After discharge from hospital she beganto pass blood and mucus per rectum with prolapse ofpolyps on defaecation. Tenesmus became a prominentsymptom and the patient was admitted to hospitalagain when 20 years old. She was thin, pale, andweighed only 86 lb. (39 kg.). Sigmoidoscopy showedscores of polyps ranging in size from o 5 to 3 cm. indiameter. They were particularly numerous in therectum but less frequent in the lower sigmoid colon.Double contrast barium enema demonstrated polypsfrom the mid part of the descending colon to the pelviccolon. Laboratory investigations revealed low levels ofserum potassium, reduced total plasma protein, andporphyria.

Fifteen polyps were removed through two colo-tomies in the sigmoid and ascending parts of the colon.The rectum was so thickly covered by polyps that nonormal mucosa could be seen. The largest polyps werefound in the lower rectum and several of these pro-lapsed out of the anus on long thin strap-like stalks.Histologically they were all juvenile polyps.

Case 3 (female, aged 4 years). She was the product ofa full-term birth, weighing i i lb (5-2 kg.). Her headwas said to be disproportionately large and her leftfoot had a sixth toe. At i8 months she began to pass alittle blood and on one occasion a small polyp. Thebleeding increased, with frequent prolapse of polyps,diarrhoea, and abdominal pain during defaecation.On examination the most striking feature was her

large head. Hypotonia was present in all limbs withweakness of the pelvic and shoulder girdles. The knee

and ankle-jerks were absent. She had an exaggeratedlumbar lordosis with a protuberant abdomen. Therewas marked clubbing ofthe fingers and toes. The rectumwas full of polyps. A hypochromic anaemia of 6 g./I00 ml. was corrected by blood transfusion, but herdiarrhoea became more severe. A diagnosis of hyper-telorism, amyotonia congenita, and adenomatous poly-posis was made.

Ileostomy and total proctocolectomy were carriedout. A total of 8o polyps were scattered throughoutthe excised colon. Histologically they were all juvenilepolyps. The patient made a good recovery from theoperation but died two years later, the cause of deathbeing given as inanition, polyposis, and hydrocephalus.

Case 4 (male, aged 2 years). He was bom with anumbilical faecal fistula and a Meckel's diverticulum,and was kept under observation for a mild communicat-ing hydrocephalus. Rectal bleeding was first noticedat the age of 2 years and sigmoidoscopy showed a polypat 15 cm. At laparotomy 3 polyps were removed fromthe caecum and 2 from the descending colon, but otherswere palpable in other parts of the colon. Malrotationof the gut was present, the transverse colon beingposterior to the jejunum. The left testis was in theabdomen and the right in the inguinal canal.The following year there was a recurrence of the

rectal symptoms accompanied by prolapse of polyps.At the age of 4 several polyps were removed through thesigmoidoscope. Over subsequent years his healthremained good apart from an iron deficiency anaemiaof 8 g./Ioo ml. However, his symptoms became moresevere and sigmoidoscopy and radiological examinationrevealed further polyps. It was decided to resect thecolon as he was thought to have adenomatous polyposiscoli. Colectomy with ileo-rectal anastomosis was carriedout, and a single polyp was excised from the ileum.The operation specimen showed a few polypoid

excrescences in the ileum. Pedunculated polyps rangingin size up to 2 cm. in diameter were scattered throughoutthe colon. The cut surface of the larger tumours wascystic. Sections of the nodules in the terminal ileumshowed them to be hypertrophic Peyer's patches. Thesingle ileal polyp as well as the colon polyps were all ofthe juvenile type.The patient is now aged io and is in good health. A

year ago several polyps were seen on sigmoidoscopy andremoved.

Case 5 (female, aged I3 years: III.2 in Pedigree i(Fig. 7)). She first passed blood per rectum at the ageof I3. Polyps prolapsing through the anus were thoughtto be haemorrhoids. Sigmoidoscopy at the age of i8showed many polyps in the rectum. The two largestwere excised and the remainder destroyed by diathermy.During the ensuing 30 years another I0 polyps wereremoved, the last being at the age of 52. They all had thehistology of the juvenile variety.

Case 6 (female, aged 8 years: III.3 in Pedigree i(Fig. 7), sister of Case 5). At the age of 8 years she

9



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/


FAMILY

2 3 4

7 8 9III 4

2 3 4 5 to

IV L a6 7 8 11 12 13

FAMILY 3 l!4

FAMILY 2

1 2 3 4 5 6 7 10 11 12 13

2 3 4 5

00 Unaffected or unexaminedI e Multiple intestinal adenomatosis

Q Carcinoma of large bowel1 Multiple juvenile polyposis* Sex unknown or, when figured, a mixed sibship

FAMILY 4

11 4v@ 4v&Wet1 2 3 4 5 6 7

III & & fI

1 2 3 4 5 6 7 8 9 10

IV 13$I4 5

2 3

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

14 1/ 18 14 13 (y31 32 33 34 35 36 37

FIG. 7. Pedigrees of Families x, 2, 3, and 4.

began passing mucus in her stools, followed later bybleeding diarrhoea and prolapse of polyps. The patientwas thought to have adenomatosis coli. At I9 years many

polyps were seen on sigmoidoscopy and removed. Thesewere all found to be juvenile polyps. Barium enemasat the age of 38 and 50 years were normal.

Case 7 (female, aged 40 years: IL.i in Pedigree i

(Fig. 7), mother of Cases 5 and 6). This patient never

complained of any symptoms but was examined at theage of 40 because her daughter (III.2) had polyposis.Two rectal polyps were found and removed. Duringthe following 2 years 7 more were excised, since whenshe has been clinically and radiologically normal.Histologically all the polyps were of the juvenile type.

Case 8 (female, aged 20 years: III.4 in Pedigree 2(Fig. 7)). She first presented at the age of 20 with a

history of prolapsing polyps and passage of mucus for'some years'. Sigmoidoscopy showed multiple polypsranging in size from 5 to 25 mm. A diagnosis of adeno-matous polyposis was made and a colectomy with ileo-rectal anastomosis was performed. During the subse-quent 37 years small rectal polyps have appeared atirregular intervals and have been removed. All thetumours have had the typical structure of juvenilepolyps.

Case 9 (male, aged I2 years: III.3 in Pedigree 2

(Fig. 7), brother of Case 8). Seen at St. Mark's Hospitalin I9IO with multiple rectal polypi, some of which hadprolapsed. No histology is available, but the descriptionof the clinical picture certainly suggests juvenile poly-posis. Scores of polyps were removed together with a

3-in. (7-6 cm.) length of rectal mucosa and otherswere destroyed by diathermy. Further polyps wereremoved at intervals during the following 8 years, butthe patient had no rectal trouble thereafter. He diedfrom a heart attack at the age of 63.

Case I0 (female, aged I0 years: III.4 in Pedigree 3(Fig. 7)). She first had rectal bleeding when 4 yearsold. Barium enema examination and sigmoidosopy at theage of io showed many polyps which bled readily on

contact. Colectomy and ileo-rectal anastomosis were

performed. During the operation a lymphangioma was

excised from the mesentery of the ileum.The operation specimen showed a total of I30 tumours

varying in size from tiny pinpoint dome-shaped mucosalelevations up to tumours 2 cm. in diameter on long stalks.All the polyps were deeply haemorrhagic and friable,some had a smooth surface while others were lobulated.In the caecum, where the tumours were most numerous,there were many branching stalks without polyps, as ifthe latter had sloughed off. The polyps examinedmicroscopically were all of the juvenile type with muchsecondary infection and haemorrhage.

IO



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

Jtuvenile Polyposis Coli

Case iI (male, aged 8 years: IV.34 in Pedigree 4(Fig. 7)). This boy had rectal bleeding, diarrhoea,and prolapsing polyps at the age of 5. When 8 yearsold he was admitted to hospital with severe abdominalpain, and laparotomy revealed polyps in the caecum.Polyps were also seen on sigmoidoscopy, and an x-rayexamination a year later confirmed their presencethroughout the colon. At the age of 9, colectomy withileo-rectal anastomosis was performed.

FIG. 8. Juvenile polyposis coli, Case II. The mucous membraneis covered with numerous pedunculated polyps, most of whichhave a smooth surface configuration.

A total of 235 polyps was scattered over the mucosa ofthe excised specimen, but mainly in the caecum andascending colon (Fig. 8). The tumours had naked-eyeappearances and microscopical characters similar tothose described in Case IO. Since the operation a fewsmall rectal polyps have been removed. The patientdied suddenly at the age of 17, probably from acuteintestinal obstruction.

Family HistoryThe ii cases described above come from 8 families,

3 from one family, and 2 from another, the remaining 6being solitary examples. In 4 of these 6 families noother case of bowel disease was discovered in spite ofinquiries involving a total of I40 family members. Thehistories of the remaining 4 families are as follows.

Family i (Fig. 7). It has already been mentionedthat Cases 5 and 6 (III.2 and III.3) are sisters and thatCase 7 (II. i) is their mother. The maternal grandfather(I.I) died from cancer of the rectum aged 57, but noinformation is available about either of his wives. Hisdaughter (II.3) by his second marriage died from carci-noma of the colon at 32 years of age. No bowel troublewas found in any other member of the family.

Family 2 (Fig. 7). In this family both brother andsister (III.3 and III.4) had juvenile polyposis (Cases 8and 9). Their father (II.7) was treated at St. Mark'sHospital for an inoperable cancer of the rectum.

Family 3 (Fig. 7). The propositus (III.4, CaseIO) has a sister (III.3) who was found to have 3 rectalpolyps at the age of ii years, one brother (III.6) whohad a small polyp at 5 years, and another (III.7) who had2 polyps at the age of 4. Unfortunately, no histology isavailable of any of these tumours.The father of the propositus (II.i) is known to have

died aged 46 from cancer of the rectum and adeno-matosis coli, as did also his brother (II.3) at the age of 33.

Family 4 (Fig. 7). In this family, the mother ofthe propositus (III.I4) and 4 of her I3 sibs died fromintestinal cancer. Her father (II. i) and grandfather(I.i) and two cousins (III.i8 and III.I9) are also said tohave died from the same cause. Polyposis was presentin the mother and her two cousins, but its nature isunknown as no histological material is available forstudy. The general pattern, however, is that of adeno-matosis.The histories of the 8 families may be summarized

as follows. In 4 there was a single case of juvenile poly-posis but no evidence of bowel disease in other membersof these families. In the other 4 families intestinalcancer was also present, being associated with multipleadenomatosis in 2 families, and with more than one caseof juvenile polyposis in the other 2 families.

DiscussionGenetics. There are two families in which

more than one person is affected with juvenilepolyposis (Families i and 2); in addition there are3 families (i, 2, and 4) in which a number of rela-tives are known to have died from cancer of thecolon or rectum. In Family 3 the father of thepropositus suffered from adenomatous polyposisand died from cancer of the rectum. All 4 familiescontain other relatives known to have died from

II



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

V2Veale, McColl, Bussey, and Morson

large bowel cancer but in whom the diagnosis ofadenomatous polyposis was not made or cannotbe confirmed with certainty. It is noteworthy thatat one time each of these 4 families was thought to

be suffering from adenomatous polyposis but was

included in the juvenile polyposis series when thehistology of the polyps was reviewed. Of especialsignificance is Family 3 in which the initial ascer-

tainment was through II.i who was affected withclassical adenomatous polyposis and cancer of therectum. His daughter III.4 began to have bowelsymptoms at the age of 4 years and was latertreated by colectomy and ileo-rectal anastomosis.It was only after some years that the significance ofthe histology of the polyps in III.4 was appreciated.The simplest explanation for the appearance of

multiple juvenile polypi would be that this is a

manifestation of the polyposis gene in children. Anecessary corollary of this hypothesis would bethat as these children grow older they would cease

to make juvenile polyps and that adenomas wouldbegin to appear in increasing numbers. Such a

change from juvenile polyposis to adenomatouspolyposis has never been observed in any of theyoung patients with polyposis who have been treatedat St. Mark's Hospital.A further explanation is that unknown environ-

mental factors modify the action of the polyposisgene and produce juvenile polyps instead ofadenomas. It is not clear how genetic or environ-mental factors could account for the high incidenceof gross congenital abnormalities in the cases ofjuvenile polyposis reported here.

If there is a separate gene for juvenile polyposisit appears to have been passed from parent tooffspring in a number of cases, but the difficultiesof explaining the family history of bowel cancer

are considerable. Histologically juvenile polypspresent none of the features of a pre-cancerouscondition. Furthermore the appearance in one

family of proved adenomatous polyposis and thestrong family history of bowel cancer in three otherfamilies must be regarded as more than a coinci-dence.

Veale (I962) examined 76 families with adenomatouspolyposis collected at St. Mark's Hospital over a periodof 40 years. Correlations between relatives of variousage parameters connected with polyposis (age at appear-ance of polypi, age at onset of symptoms, age at deathfrom cancer) all showed consistent values near + 0-5 forcorrelations between sibs (brothers and sisters); andvalues of zero between affected parents and offspring.In the absence of significant environmental influencesthe values of both correlations are expected to be + o 5.Environment was not thought to be important in adeno-

matous polyposis as the average age for onset of symp-toms was 29 years, and for death from cancer 41 years.At these ages differences in environment between sibsare likely to be just as great as those between parentsand children. Consequently some other explanation ofthe absence of a parent-child correlation was required.Later Veale (i965) increased the number of familiesstudied to I02 with confirmation of the previous results.A genetic model involving three allelic genes was pro-posed. The genes were:

'P' the adenomatous polyposis gene; 'p' the 'polyp'gene; '+' the normal gene.

Persons with polyposis could thus have two possiblegenotypes, Pp and P+ (a third genotype PP is possiblebut will be so rare that it can safely be neglected). Itwas proposed that persons of genotype Pp suffered fromsevere adenomatous polyposis with early age of onsetand early death from cancer. Persons of genotype P+were thought to have a later age of onset and a laterdeath from cancer. Evidence was produced to showthat persons with polyposis probably consisted of twooverlapping distributions. The genes 'p' and '+' aretermed modifying alleles, as the action of the polyposisgene is modified according to which allele accompaniesit. Since an affected child receives the polyposis genefrom his affected parent, the modifying allele must comefrom the normal parent who will have one of three pos-sible genotypes: pp, p +, or + +. There is thus noreason why the genotype of the affected childrenshould necessarily be the same as that of their affectedparent. The parent-child correlation is thus zero. Itcan also be shown that under these circumstances thesib-sib correlation will be + o05 for all frequencies ofthe 'p' and '+' genes in the general population. Anestimate of the possible frequency of the 'p' gene inthe population is possible if it is supposed that personsof genotype 'pp' are those people in whom isolatedadenomas of the colon and/or rectum appear. Anelaboration of this genetic model allows adenomatouspolyposis families, i.e. the few families in which isolatedadenomas appear to be inherited, and families with astrong history of bowel cancer in the absence of poly-posis, to be included in a single genetic theory which iscapable of further confirmation.

By an extension of this model it is possible toinclude the families in which juvenile polyposishas appeared. It is now necessary to postulate afurther allele which so modifies the action of thepolyposis gene that juvenile polypi are producedinstead of adenomas. The following polyposisgenotypes are now possible:

Pp : adenomatous polyposis with early onset.P+ : adenomatous polyposis with late onset.Pj : juvenile polyposis.

Among the population without polyposis there will

12



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

Juvenile Polyposis Coli

TABLE IPOSSIBLE CLINICAL EFFECTS FOR DIFFERENT

GENOTYPES

Genotype Possible Clinical Effect (Phenotype)

+ + NormalPi } Probably both normalpi ?? Isolated adenomas and/or juvenile polyps??P Isolated adenomas

Isolated juvenile polyps

be 6 possible genotypes. In Table I we suggespossible clinical effects for the different possibilities.The term 'isolated' is here taken to mean theoccurrence of a single polyp or just a few (e.g. lessthan io) so that the question of multiple polyposiswould not ordinarily arise. Unfortunately wecannot estimate the frequency of persons of thedifferent genotypes without some knowledge ofthe relative frequencies of the genes 'p', 'j', and'+'. The frequency of persons with isolatedadenomas of the colon and rectum increases withthe age of the population studied. In persons overthe age of 40 the incidence is commonly reckonedto be in the region of io% so that we may assume

that this io% is represented by the persons of the'pp' genotype. Unfortunately, no estimate of theproportion of children with isolated juvenile polypiis available, and an accurate estimate may not even

be possible. It is reasonably certain, however, thatthe proportion of children affected is considerablyless than io% and is probably even less than i%.If we assume that the frequency of persons carryingthe 'j' gene is low, we could expect that in a smallproportion of adenomatous polyposis families,with most of the affected persons showing poly-posis and cancer, an occasional person woulddevelop juvenile polyposis even though one parentappeared to have adenomatous polyposis (e.g.Family 3). If the normal parent is homozygous forthe 'j' gene, then all children in the sibship whoreceive the polyposis gene will develop juvenilepolyposis (e.g. Family 2). Much more commonly,the normal parent will be heterozygous for the 'j'gene (i.e. either pj or j +), so that in such sibshipshalf the children receiving the polyposis gene willdevelop juvenile polyposis while the other halfwill develop adenomatous polyposis. This situationhas not yet been observed, so that confirmation, orotherwise, of this genetic model must await thestudy of families in which juvenile polyposis andadenomatous polyposis are present in sibs.The occurrence of 4 cases of juvenile polyposis

without any family history of polyposis or cancer

of the bowel seems to require some explanation.Isolated cases of adenomatous polyposis are wellknown, and it has been suggested that such casesare non-familial. The writers do not agree withthis hypothesis. Veale (I962) has shown thatmutation of a hitherto normal gene to the polyposisgene is required in order to maintain the frequencyof the condition at its observed level in the popula-tion. Isolated cases (mutations) are seen to occurwith just that order of frequency that independentcalculations lead us to expect. There is thus nonecessity to postulate the existence of a non-familial form of adenomatous polyposis. Such casesare more correctly regarded to have arisen as aresult of mutation, and these persons can beexpected to pass the condition on to half of theiroffspring. This has been observed to happen witha number of cases which were at first regarded asisolated and 'non-familial'.

Since mutation to the 'P' form of the gene isthought to occur, this will from time to time takeplace in persons carrying one or two 'j' genes, sothat the new genotype will be 'Pj' and juvenilepolyposis will develop. The present series ofjuvenile polyposis cases was obtained by reviewingthe histology of adenomatous polyposis familiesand from a search for children affected with 'poly-posis'. In no sense can this investigation beregarded as a population survey, so that no con-clusion can be drawn with respect to the relativeincidence of cases with and without family history.

Sex Ratio in Juvenile Polyposis. Threemales and eight females were affected. This is nota statistically significant deviation from the hypo-thesis that the sexes are equally affected. Additionalcases will have to be collected before any conclusioncan be drawn.

Natural History. Very little is known of thecourse of juvenile polyposis, but II.i (Family I) isat present alive and well aged 75. No person withadenomatous polyposis who has attained this agewithout developing carcinoma is known to the writ-ers, so that confirmation is obtained of the histo-logical evidence that the risk of malignancy is low.The age of onset of symptoms in patients with

juvenile polyposis has already been mentioned, buta striking difference is found if the patients aredivided into two groups on the basis of the presenceor absence of a family history of large bowel patho-logy. In those patients with no family history, theage at onset of symptoms is 4-5 years, whereas inthose with a family history the age is 9-5 years.This latter figure excludes the one patient who has

13



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

4Veale, McColl, Bussey, and Morson

never had any symptoms. A similar difference inthe age at onset of symptoms was noted for adeno-matous polyposis when isolated cases of polyposiswere compared with the propositi of families inwhich other cases were found (Veale, I962). Inthe isolated cases symptoms appeared at an averageage of 23 years and at 29 years in the others. Thegenetic interpretation of these observations is thatin those families where affected parents are sur-viving to pass their abnormal polyposis gene on totheir children, they survive by virtue of the fact thattheir total genetic background (genome) favourstheir continued existence. Natural selection thustends to perpetuate those genomes upon whichabnormal genes have the least effect. Whenmutation occurs the genome without previous ex-perience of the abnormal gene will be less welladapted to its effects and an average earlier andmore severe effect can be expected. Consequently,the earlier age of onset in those cases without afamily history can be interpreted as evidence infavour of mutation.

Congenital Anomalies. No satisfactoryexplana-tion of the various congenital anomalies found in3 of these patients is at present suggested. It iscurious that the patients with congenital anomaliesare the ones without a family history of polyps orintestinal cancer.

Pathology. Whatever the nature of juvenilepolyps the microscopical appearances are quitedistinctive compared with adenoma (Roth andHelwig, I963). The latter is a neoplasm of theintestinal epithelium, whereas the juvenile polyp iscertainly non-neoplastic. It has been postulatedelsewhere (Morson, I962a, b) that the juvenilepolyp is a hamartoma or malformation of the con-nective tissue element of the intestinal mucosawithout the participation ofthe muscularis mucosae.This could explain the generally accepted viewthat juvenile polyps have no pre-cancerous poten-tial (Knox et al., I960). However, a number ofauthorities, notably Roth and Helwig (I963),believe they have an inflammatory origin.

There is some evidence in support of the conceptthat these tumours are hamartomas. First, thepolyps are mostly found in babies and youngchildren with a decreasing incidence as age advances.Secondly, associated congenital abnormalities werepresent in 4 of the i i patients reported here(Table II). Lastly, the histopathology is usuallydistinctive and unlike the neoplastic types oftumour, such as adenoma and villous papilloma orindeed the inflammatory polyps of ulcerative colitis

TABLE IIASSOCIATED CONGENITAL ANOMALIES AND

HAMARTOMAS

Case No. Associated Anomalies

2 Malrotation of gut;porphyria

3 Hypertelorism;amyotonia congenita;extra toe

4 Umbilical faecal fistula and Meckel's diverticulum;hydrocephalus;malrotation of gut;undescended testes

IO Mesenteric lymphangioma

and other diffuse mucosal inflammations of thelarge intestine.

Willis (1958) defined hamartoma as a primarilynon-neoplastic malformation characterized by anabnormal mixture of tissue indigenous to the partwith an excess of one or more of these. In juvenilepolyps it is the connective tissue element of themucosa or lamina propria which appears to undergoexcessive growth. Willis also allows the use of theterm for an inborn tissue anomaly which manifestsitself by excessive growth during postnatal matura-tion. In some of our cases reported here as well asin others seen recently juvenile polyps have con-tinued to appear during adult life after the initialremoval. This tendency to recurrence is notevidence against the concept that juvenile polypsare hamartomatous. Research on epithelial cellrenewal in recent years (Leblond and Stevens,1948; Creamer, Shorter, and Bamforth, I96I;Gastroenterology, I962) has shown beyond questionthat the intestinal epithelium throughout postnatallife is a highly dynamic structure which replacesits cell population every few days. Nor shouldwe presume that the supporting structure of themucosa or lamina propria is wholly inert, thoughit is likely that its turnover rate is slower than theepithelial element. The pathogenesis of juvenilepolyps might well be explained by an inborn(probably genetic) tissue anomaly which expressesitself in a morphological pattern that does not pre-dispose to malignant change (i.e. it is non-neo-plastic), in contrast to adenomatous polyposis coliin which the gene expresses itself in the form of abenign neoplastic state of the intestinal epitheliumprone to malignant degeneration.

Clinical Features. The passage of bright redblood per rectum occurred in 9 out of ii patientsand was the commonest presenting symptom(Table III). This was often accompanied by

I4



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

juvenile Potyposis Coli

TABLE III

Symptoms No. of PresentingPatients Symptoms

Rectal bleeding 9 6Prolapse of polyps 8 3Mucus 5 IDiarrhoea 4Abdominal pain 2

mucus and diarrhoea. Two patients underwentemergency laparotomy for right-sided abdominalpain. These symptoms are not uncommon inadenomatous polyposis, however, and may heraldthe onset of malignant change. Prolapse of polypsout of the rectum is a common feature of juvenilepolyposis but rare in the adenomatous variety.One patient had no symptoms. The average age ofonset of symptoms was 6 years, and 8 of the i ipatients were female. In contrast the averageage of onset of symptoms in adenomatous poly-posis is approximately 30 years and the sexes areequally affected (Veale, I962).The appearances of these juvenile polyps was

characteristic. They were bright red, smooth,glistening spheres which bled readily on contact.The size varied from 3 to 30 mm. in diameterand most were pedunculated. They could be madeto prolapse out of the rectum and could be pluckedoffthe rectal wall with ease and with only a minimumof bleeding. Spontaneous passage of polyps alsooccurred. In contrast adenomatous polyps areusually dark red and lobulated, do not bleed easilyon contact, cannot be readily plucked off, and arenot usually passed spontaneously.

Sigmoidoscopy and barium enema radiographsindicated the extent of the disease, but the numberof polyps present was usually underestimated. Alli i patients had rectal polyps, and in 4 they werelimited to the rectosigmoid and did not exceed40 (see Table IV). In the other 7 patients the polypswere found throughout the colon, mainly gathered

TABLE IVDISTRIBUTION OF JUVENILE POLYPS

Site No. of No. ofCases Polyps

Recto-sigmoid only 4 I0/40Mainly rectum and 3 I50/250

right colonThroughout

rectum and Mainly rectum, also 3 3o/8ocolon ileum in 2 cases

Distribution c I > 50unknown

in the rectosigmoid and in the right side of thecolon. Their numbers varied from 30 to 300.Two patients had a few polyps in the terminalileum. Adenomatous polyposis coli does not involvethe ileum. Adenomatous polyps and juvenilepolyps have been reported in the appendix (Shnitkaand Sherbaniuk, I957), though this part of thelarge gut was not involved in any of our patientswith juvenile polyposis as far as this could beascertained. I. M. P. Dawson (I964, personal com-munication) carried out a post-mortem examina-tion on a girl of i5 months in whom there was astate of juvenile polyposis throughout the entiregastro-intestinal tract. The cause of death wasanaemia due to intestinal haemorrhage.

Congenital anomalies or malformations occurredin 4 patients (see Table II). This has not beendescribed in adenomatous polyposis. Two of thefour with congenital anomalies also had low plasmaproteins and serum potassium. Both were anaemic.Two other patients not included in this series alsohad low plasma proteins. One girl of i8 yearsdeveloped generalized oedema and another aged 8presented with such gross plasma protein depletionthat a prophylactic colectomy had to be performed.In a recent patient admitted to St. Mark's Hospitalthe presenting symptoms were profound anaemiaand prolapse of polyps out of the rectum. Thispatient, again not included in this series of fullyinvestigated cases, also had a patent interventricularseptum. Biochemical abnormalities, in the experi-ence of this hospital, are not a feature of the present-ing symptoms of adenomatous polyposis. Thefeatures of juvenile and adenomatous polyposis arecompared in Table V.

TABLE VCOMPARISON OF JUVENILE AND ADENOMATOUS

POLYPOSIS

Juvenile Adenomatous

Average age of 6 30onset ofsymptoms

Symptoms Bleeding; Diarrhoeaprolapse;auto-amputation

Congenital Yes Noanomalies

Premalignant No YesHistology Hamartoma Neoplasm

Treatment. The diagnosis of juvenile poly-posis should be confirmed on histological examina-tion of several polyps before considering treatment.As far as cancer prevention is concerned radical

I5



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/


surgery appears to be unnecessary, as there is yetno evidence that juvenile polyps are pre-cancerous.

Five patients had radical surgery as it was wronglyassumed at the time that their polyposis was adeno-matous and therefore premalignant. Althoughjuvenile polyposis is not thought to have any malig-

nant potential, the loss of blood and protein may besevere enough to warrant a colectomy.

SummaryThe histopathology, clinical features, and family

histories of ii patients with juvenile polyposis ofthe large intestine are described.

Juvenile polyposis coli is clinically and patho-logically distinct from adenomatous polyposis coli.It usually presents in children with rectal bleedingand prolapse of polyps. Associated congenitalabnormalities were found in 4 of the ii patientsstudied.

Juvenile polyposis should be classified among thehamartomatous syndromes. There is no evidence,as yet, that it is a pre-cancerous condition.

In a proportion of cases, there is a family historyof juvenile polyposis and/or cancer of the largebowel. Various genetic explanations are discussed,and it is concluded that juvenile polyposis may resultfrom modification of the action of the gene foradenomatous polyposis by an allele received fromthe normal parent. Isolated cases of juvenile poly-posis may be the result of fresh mutations.

We are much indebted to Mr. M. H. Harmer, Mr.Rex Lawrie, and the consultant staff of St. Mark'sHospital for permission to study their patients. Thephotographs were taken by Mr. Norman Mackie. Mr.

Lloyd Soodeen has given valuable technical assistance.The expenses of this investigation have been providedout of a block grant to the Research Department ofSt. Mark's Hospital from the British Empire CancerCampaign.

RbFERENCESCastro, A. F. (I955). Adenomatous polyps of the rectum and colon

in children. Clin. Proc. Child. Hosp. ((Wash.), II, 29.Creamer, B., Shorter, R. G., and Bamforth, J. (I96I). The turnoverand shedding of epitheL;al cells. Gut, 2, 110.

Gastroenterology (I962). Cell renewal in the gastrointestinal tractof man. 43,472.

Gordon, D. L., Hallenbeck, G. A., Dockerty, M. B., Kennedy,R. L. J., and Jackman, R. J. (1957). Polyps of the colon inchildren. Arch. Surg., 75, 90.

Hines, M. O., Hanley, P. H., Ray, J. E., and Rush, B. M. (I959).Polyps of the colon and rectum in children: Experience with 27cases. Dis. Colon Rect., 2, I6I.

Kerr, J. G. (1948). Polyposis of the colon in children. Amer. J7.Surg., 76, 667.

Knox, W. G., Miller, R. E., Begg, C.F., and Zintel, H. A. (I96o).Juvenile polyps of the colon. Surgery, 48, 201.

Lane, J. R. (I865). Clinical observations on diseases of the rectum.Lancet, 2, 87.

Leblond, C. P., and Stevens, C. E. (1948). The constant renewal ofthe intestinal epithelium in the albino rat. Anat. Rec., 100, 357.

Mallam, H. S., and Thomson, S. A. (I959). Polyps of the rectumand colon in children. Canad. J. Surg., 3, 17.

Marks, M. M., and Atkinson, K. G. (I964). Heterotopic bone injuvenile rectal polyp. Dis. Colon Rect., 7, 345.

McColl, I., Bussey, H. J. R., Veale, A. M. O., and Morson, B. C.(I964). Juvenile polyposis coli. Proc. roy. Soc. Med., 57, 896.

Morson, B. C. (1962a). Some peculiarities in the histology ofintestinal polyps. Dis. Colon Rect., 5, 337.

(x962b). Precancerous lesions of the colon and rectum. J.

Amer. med. Ass., 179, 3I6.Roth, S. I., and Helwig, E. B. (I963). Juvenile polyps of the colonand rectum. Cancer, I6, 468.

Shnitka, T. K., and Sherbaniuk, R. W. (19s7). Adenomatouspolyps of the appendix in children. Gastroenterology, 32, 462.

Todd, I. 0. (1963). Juvenile polyps. Proc. roy. Soc. Med., 56, 969.Veale, A. M. 0. (1962). Familial intestinal polyposis. Ph.D.

Thesis, University of London.(I965). Intestinal Polyposis Eugenics Laboratory Memoirs

Series, 40. Cambridge University Press, LondonWillis, R. A. (I958). The Borderland of Embryology and Pathology.

Butterworth, London.

i6



g.bmj.com

/J M


g.3.1.5 on 1 March 1966. D

ownloaded from

http://jmg.bmj.com/

juvenile polyposis coli - journal of medical geneticsclinical features and pathology of our cases...

Documents