Morphologic Characterization of Hamartomatous Gastrointestinal Polyps in Cowden Syndrome, Peutz-Jeghers Syndrome and Juvenile Polyposis Syndrome Ruthy Shaco-Levy MD, Kory W. Jasperson MS, CGC, Katie Martin MD, N. Jewel Samadder MD, Randall W. Burt MD, Jian Ying PhD, Mary P. Bronner MD PII: DOI: Reference:
S0046-8177(15)00429-3 doi: 10.1016/j.humpath.2015.10.002 YHUPA 3732
To appear in:
Human Pathology
Received date: Revised date: Accepted date:
16 May 2015 5 October 2015 8 October 2015
Please cite this article as: Shaco-Levy Ruthy, Jasperson Kory W., Martin Katie, Samadder N. Jewel, Burt Randall W., Ying Jian, Bronner Mary P., Morphologic Characterization of Hamartomatous Gastrointestinal Polyps in Cowden Syndrome, PeutzJeghers Syndrome and Juvenile Polyposis Syndrome, Human Pathology (2015), doi: 10.1016/j.humpath.2015.10.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Morphologic Characterization of Hamartomatous Gastrointestinal Polyps in Cowden Syndrome, Peutz-Jeghers Syndrome and Juvenile Polyposis Syndrome Ruthy Shaco-Levy, MD, 3,4 Kory W. Jasperson, MS, CGC, 4 Katie Martin, MD, 3,5,6 N. Jewel
T
1,2,3
RI P
Samadder, MD, 3,5,6 Randall W. Burt, MD, 3,5,7 Jian Ying, PhD, 2,3 Mary P. Bronner, MD 1
Department of Pathology, Soroka Medical Center, Ben-Gurion University of the Negev, Beer
SC
Sheva, 84101, Israel; 2Department of Pathology & ARUP Laboratories, 3Huntsman Cancer
MA NU
Institute, 4Genetic Counseling, 5Department of Internal Medicine, 6Division of Gastroenterology, 7Division of Epidemiology, University of Utah, Salt Lake City, UT, 84112, USA.
Keywords: Cowden syndrome, Peutz-Jeghers Syndrome, Juvenile Polyposis Syndrome ,
ED
hamartoma, polyp.
PT
Running title: Hamartomatous Gastrointestinal Polyposis
CE
Conflict of Interest: None
Funding: We acknowledge support of funds and use of the Biorepository and Molecular
AC
Pathology, in addition to the Genetic Counseling shared resources, supported by NCI grant 5P30CA042014-22, awarded to the Huntsman Cancer Institute and to the Gastrointestinal Cancer Program at Huntsman Cancer Institute. Abstract: The morphological features of the gastrointestinal polyps in hamartomatous polyposis syndromes are poorly defined. Our aim was to better characterize the gastrointestinal hamartomas in these syndromes. A blinded review was performed regarding many histological features for every polyp. The study included 15 Cowden syndrome (CS), 13 Peutz-Jegher's (PJS), 12 juvenile polyposis (JuvPS) patients, and 32 cases of sporadic 1
ACCEPTED MANUSCRIPT hamartomatous polyps. A total of 375 polyps were examined. CS polyps were characteristically colonic, sessile, small, without surface erosion, showing mildly inflamed
T
fibrotic lamina propria with smooth muscle proliferation and lymphoid follicles. They
RI P
showed the least degree of cystic glands and had no thick mucin. Uncommon but specific features were ganglion cells and nerve fibers within the lamina propria and mucosal fat. PJS
SC
polyps were typically of small or large bowel origin, often exophytic, seldom eroded, with inflamed edematous and fibrotic lamina propria and dilated cystic glands filled with often
MA NU
thick mucin. All PJS polyps showed smooth muscle proliferation, frequently widespread. The polyps of JuvPS were typically colonic, large, exophytic, eroded, with strikingly edematous, fibrotic markedly inflamed lamina propria, cystic glands filled with frequently thick mucin, and the least degree of smooth muscle proliferation. Non-syndromic hamartomatous polyps
ED
were similar to JuvPS polyps; however, they were more often colonic, smaller, showed more widespread smooth muscle proliferation, and were less likely to contain thick mucin. In
PT
conclusion, we were able to define the characteristic hamartomatous polyp for each
CE
hamartomatous polyposis syndrome. Awareness to these features may aid in the diagnosis
AC
of these rare syndromes. Introduction:
The hamartomatous polyposis syndromes comprise a diverse group of genetic,
clinical and pathologic entities. The most common and important, due to their increased cancer rates, are Cowden's syndrome (CS), Peutz-Jegher's syndrome (PJS), and juvenile polyposis syndrome (JuvPS) [1-5]. CS, an autosomal dominant condition affecting 1 in 200,000 people, is characterized by multiple hamartomatous lesions [1-7]. Gastrointestinal (GI) polyposis is a common manifestation, can occur throughout the GI tract, and is reported to have a markedly varied histology, with predominantly hamartomatous but also adenomatous, inflammatory, 2
ACCEPTED MANUSCRIPT hyperplastic, lipomatous and ganglioneuromatous polyps [5-7]. Consensus criteria have been established to assist in the diagnosis of CS, which frequently presents with various
T
signs and symptoms [2, 8]. CS patients are at particularly high risk of developing cancer of
RI P
the breast, thyroid, ovary, endometrium, uterine cervix, and urinary bladder [1-5, 7]. CS probably also confers a risk of colorectal cancer [6]. Germline mutations in the PTEN
SC
(phosphatase and tensin homolog) tumor suppressor gene located on chromosome 10q are found in approximately 80% of patients with CS [5, 7].
MA NU
PJS is an autosomal dominant inherited hamartomatous polyposis syndrome with a prevalence of approximately 1 in 200,000. PJS is characterized by melanotic mucocutaneous hyperpigmentation, which often fades with age, and GI hamartomatous polyps, mostly in
ED
the small bowel but also in the colon and stomach [7, 9-11]. These hamartomatous polyps can cause abdominal pain and intussusception, sometimes leading to bowel obstruction and
PT
severe GI bleeding [10]. PJS is now recognized as a cancer predisposition syndrome, since these patients are at very high relative risk for colorectal cancer and a variety of extracolonic
CE
malignancies. The malignancies recognized include breast, pancreas, thyroid, stomach, small
AC
intestine, ovary, endometrium, uterine cervix, testis, multiple myeloma and skin [11-13]. Patients with PJS have a 93% cumulative lifetime risk for cancer [14, 15], including an almost 70% risk of GI cancer [2, 3]. Germline mutations in the tumor suppressor gene STK11/LKB1 gene on chromosome 19p are responsible for the PJS [16]. These mutations are found in up to 80% of affected individuals; up to 25% of documented cases are sporadic [7, 15]. JuvPS, the most common of the hamartomatous polyposis syndromes, affects 1 in 100,000 and occurs as an autosomal-dominant inherited disorder in approximately 30% of the patients; the remaining cases represent de-novo mutation [15, 17]. JuvPS is characterized by the presence of multiple hamartomatous polyps affecting the colon and rectum. Unlike sporadic colorectal juvenile polyps, which are relatively common, occurring in
3
ACCEPTED MANUSCRIPT up to 2% of children younger than 10 years of age, the polyps of JuvPS are more numerous and may affect the proximal GI tract polyps [3, 5, and 16]. Patients with JuvPS are at
T
increased risk of colorectal, pancreatic and upper GI cancer, with an overall risk of GI
RI P
malignancies at 55% [7, 15, 18, and 19]. Both sporadic and inherited forms share similar genetics: germline mutations of SMAD4 (Mothers against Decapentaplegic Homolog4, also
SC
known as MADH4 and DPC4), located on the chromosome 18q, are detected in approximately 15% of patients with JuvPS. The BMPR1A gene (Bone Morphogenetic Protein
MA NU
Receptor-Type 1A), located on chromosome 10q, is mutated in about 25% of JuvPS patients [15, 20, 21]. ENG germline mutations have been found in JuvPS presenting in early childhood [18]. All 3 genetic changes cause disruption of the TGFb (transforming growth factor beta)
ED
signal transduction pathway [22].
Pathologists frequently fail to raise the suspicion of a hamartomatous polyposis
PT
syndrome based on the gastrointestinal polyps' morphology. CS polyps are sometimes interpreted as hyperplastic polyps, juvenile polyps or merely hamartomatous polyps [7, 23].
CE
Patients with CS can present with multiple juvenile colonic polyps and therefore be
AC
misdiagnosed as having JuvPS [3]. Also, although PJS polyps are thought to demonstrate characteristic histological features, with "classic" tree branchlike pattern, this arborizing smooth muscle configuration is more pronounced in the small intestine than in the colon, where these polyps may be misdiagnosed as mucosal prolapse polyp [24]. Furthermore, as there are no known histological differences between sporadic and syndromic juvenile polyps, it is currently not possible to raise the suspicion of JuvPS based on a solitary or even a couple of colorectal juvenile polyps [7]. The genetic bases have been determined in large part for all of these syndromes. However, careful systematic genotype-phenotype analysis specifically of gastrointestinal polyp morphology has not been done. The prevalence of various morphologic features of
4
ACCEPTED MANUSCRIPT these polyps in known genetic backgrounds remains poorly defined. This needs to be rectified as polyp morphology is one of the major factors contributing to correct diagnosis,
T
clinical management and research study design in these syndromes. Appropriate
RI P
identification of individuals and families affected with these syndromes is crucial as these syndromes are associated with high rates of malignancy and a thorough cancer screening is
SC
necessary in order to enhance the opportunity for early detection [2].
The aim of this study was to systematically review the histological features of JuvPS,
MA NU
PJS, CS, and non-syndromic hamartomatous polyps within the University of Utah polyposis registry and surgical pathology archives, blinded to known genetic and clinical syndromic backgrounds, to assess the value of histologic features for diagnostic utility.
ED
Materials and Methods:
PT
Approval for the conduction of this study was obtained by the University of Utah's Institutional Review Board. The pathology database at the University of Utah Medical Center
CE
was searched from January 2000 to December 2011 using the terms "hamartomatous polyp(s)", "juvenile polyp(s)", "peutz-Jegher's syndrome", "Cowden Syndrome" and
AC
"intestinal ganglioneuroma". This search yielded 476 cases that were then scanned by the Clinical Genetics Department, in order to identify those who have confirmed CS, PJS, or JuvPS, either by a genetic mutation or by clinical criteria. Among the non-syndromic cases, 45 that matched the polyposis cases by age were selected to serve as a control group. After comprehensive review of their medical charts, 32 were confirmed as definite sporadic cases. The archival histologic slides from the 4 groups (CS, PJS, JuvPS, and non-syndromic patients) were obtained and intermixed with each other. A systematic review was performed by 2 pathologists with interest in gastrointestinal pathology using a doubleheaded microscope, and a consensus regarding each of the following histological features
5
ACCEPTED MANUSCRIPT was rendered for every polyp. The gross features recorded were total number of polyps, their location, polyp size, and configuration (exophytic/sessile). The microscopic features
T
examined included surface erosion, expanded and inflamed lamina propria, active
RI P
inflammation, expanded edematous lamina propria, expanded fibrotic lamina propria, smooth muscle fibers in the lamina propria, glandular distortion, dilated cystic glands, mucin
SC
in cystic glands, serrated architecture, thick intra-luminal mucin (defined as dense deeply eosinophilic mucin), "onion-skin" arrangement of fibroblasts around glands, lobular clusters
MA NU
of glands, stromal ganglion cells, nerve fiber proliferation in the lamina propria, lymphoid follicles, mucosal stromal fat, mild (low-grade) dysplasia, and severe (high-grade) dysplasia. For each feature the diagnosis given was either present or absent, except smooth muscle fibers in the lamina propria that was graded as follows: zero for absence of smooth muscle
ED
fibers in the lamina propria, 1 when scattered delicate fibers were present, and 2 when conspicuous broad strands of smooth muscle crossed the lamina propria. The review was
PT
blinded to knowledge of the clinical syndrome and genetics.
CE
The catalogued histological features were then analyzed for prevalence in each
AC
group. The prevalence was then compared between the 4 groups in an attempt to find features that differ significantly between the groups, in order to better define diagnostic polyp morphology relative to known genetic and clinical backgrounds. Genetic analysis of PTEN, STK11, and SMAD4: Genetic testing of PTEN, STK11, and SMAD4 was performed using standard clinical techniques, which typically consisted of sequencing of all exons and adjacent introns, in addition to large rearrangement testing. Or in the case of a known mutation in the family, site specific testing for the familial mutation may have been performed. Statistical analysis:
6
ACCEPTED MANUSCRIPT The proportion of polyps with interested features was calculated for each of the 4 groups. The difference of the proportion among the 4 groups was assessed by permutation
T
of group at individual level. Specifically, observed chi-square statistic was calculated from
RI P
the original data. The symptom group of each subject was permuted and chi-square statistic was calculated for the permutated data. 1000 permutations performed for 1000 times and p
SC
value was calculated the proportion of the permuted statistic that not less than observed statistic. This permutation method was applied to all comparison of proportion on interested
the lamina propria and location. Results:
MA NU
feathers. Fisher exact test was used to assess the association between smooth muscle in
ED
The group of patients with clinically confirmed hamartomatous polyposis syndromes included 15 with CS from 14 different families (9 confirmed mutations; 6 clinical diagnoses),
PT
13 with PJS from 11 different families (8 confirmed mutations; 5 clinical diagnoses), and 12
CE
with JuvPS from 10 different families (6 confirmed mutations; 6 clinical diagnoses). Thirty two cases of sporadic hamartomatous polyps served as a control group. A total of 375 polyps
AC
were examined. Table 1 summarizes the various polyps seen in each category. Hamartomatous and juvenile polyps comprised the vast majority of polyp type in each group. A polyp was classified as juvenile polyp if it had the characteristic features: cystically dilated glands filled with mucus, separated by an inflamed and edematous stroma. Hamartomatous polyps without these features were classified as hamartoma. Table 2 compares the various gross and histological features between the 4 groups. Colonic predominance with >75% colonic polyps and <10% small bowel polyps was seen in CS, JuvPS and the sporadic hamartomatous polyp groups; on the other hand, approximately 40% of the PJS polyps derived from the colon and small bowel each (p<0.001). It is possible that sporadic hamartomatous polyps in the stomach were diagnosed as hyperplastic or 7
ACCEPTED MANUSCRIPT inflammatory polyps, accountable for the colonic predominance in this group; however, we cannot confirm or reject this assumption. Additionally, we assume that the colonic
T
predominance of sporadic polyps is not attributed to selection bias (as they do not usually
RI P
undergo upper endoscopy), because one would expect that if sporadic polyps were located in the stomach or small bowel, they would be symptomatic. Our search for "hamartomatous
SC
polyps" wasn't limited to location, and detected almost only colonic polyps. The largest polyps were those of JuvPS, followed by non-syndromic hamartomas, PJS, and CS
MA NU
polyps (p<0.001).
Surface erosion (Fig. 1A) was seen predominantly in juvenile polyps, hence it was much more common in JuvPS and non-syndromic polyps than in CS and PJS polyps
ED
(p<0.001).
PT
In many histological aspects CS polyps were found to differ from the other 3 groups. On average CS polyps were the smallest (p<0.001) and most of them were sessile. Exophytic
CE
polyps with a fibrovascular stalk were more prevalent in the JuvPS, PJS, and non-syndromic hamartoma groups compared to the CS group (p=0.01). The lamina propria of CS polyps was
AC
significantly less edematous (p=0.004) with lower degree of both chronic and active inflammation (p<0.001) (Figs. 1B-1D). Although glandular distortion was a common feature, seen in ≥76% of the polyps in all 4 groups, it was less frequent in CS polyps (p=0.01). Moreover, dilated cystic glands filled with mucin (Fig. 1E) were evident in ≥90% of the PJS, JuvPS, and non-syndromic hamartomatous polyps, compared to 36% of CS polyps (p<0.001). In addition, thick luminal mucin (Fig. 1F), that was a frequent characteristic of PJS and JuvPS polyps, seen in 44% and 47%, respectively, was completely absent in CS polyps (p<0.001). Furthermore, some features distinguished CS polyps from the others: ganglion cells, nerve fibers and adipose tissue within the lamina propria (Figs. 2A-2C) were seen almost exclusively in CS polyps (p=0.02). The adipose tissue consisted of clusters of adipocytes in 8
ACCEPTED MANUSCRIPT the lamina propria (not in the sub-mucosa), separating glands. Additionally, lymphoid follicles (Fig. 2C) were more prevalent in CS polyps, compared to the other 3 groups
T
(p=0.007). Glands clustering surrounded by "onion-skinning" fibroblasts were rarely
RI P
observed features that although almost restricted to CS polyps, did not reach statistical significance.
SC
Although some degree of smooth muscle proliferation in the lamina propria was evident in each group (Figs. 3A-3D), it was considerably more characteristic of PJS polyps
MA NU
(p<0.001). The lowest degree of smooth muscle fibers within the lamina propria was observed in JuvPS polyps. The degree of smooth muscle proliferation was found to be correlated with polyp size in all groups; larger polyps showed higher level of smooth muscle
ED
proliferation (p=0.01). Additionally, smooth muscle proliferation was higher in small bowel PJS polyps compared to colonic PJS polyps (p<0.001).
PT
PJS and JuvPS polyps were found to be similar in many aspects; they were frequently
CE
exophytic with edematous and fibrotic lamina propria showing marked acute and chronic inflammation. Dilated cystic glands, often filled with thick mucin, were a usual component in
AC
both polyp types. They differed from each other in the following characteristics: location (p=0.05), size (p<0.001), surface erosion (p<0.001), and smooth muscle proliferation in the lamina propria (p=0.001). JuvPS polyps originated predominantly from the colon, while the distribution between colon and small bowel was almost equal for PJS polyps. JuvPS polyps were considerably larger and displayed surface erosion much more frequently. Smooth muscle proliferation in the lamina propria was more conspicuous in PJS polyps. Sporadic and syndromic juvenile polyps shared many features. They were the larger polyps in this series (although JuvPS polyps were the largest), they were frequently exophytic with surface erosion, and had inflamed lamina propria. They practically always exhibited glandular distortion with dilated cystic glands. They differed from each other by 9
ACCEPTED MANUSCRIPT location: almost all sporadic juvenile polyps were colonic and none were of gastric origin, while the syndromic polyps were more equally distributed between the colon, small bowel
T
and stomach, although they also showed colonic predominance. Additionally, edema of the
RI P
lamina propria was more common in the syndromic polyps (p=0.004), as well as thick luminal mucin (p=0.02). The sporadic polyps showed more conspicuous smooth muscle
SC
proliferation within the lamina propria (p=0.03).
Low-grade dysplasia and adenoma were seen in the following groups in a decreasing
low-grade dysplasia (p=0.045).
MA NU
order of frequency: CS, JuvPS, and PJS. The difference was statistically significant only for
No statistically significant differences were observed while comparing each
ED
histological feature between patients with and without detectable mutation in each group.
PT
Taking into account the common histological features, a typical polyp profile can be described for each group (Table 3). CS polyps (Fig. 4A) are usually colonic, relatively small
CE
(mean 0.35 cm), rarely exophytic, and are almost never eroded. They are characterized by expanded and fibrotic lamina propria with some degree of chronic inflammation, smooth
AC
muscle proliferation and presence of lymphoid follicles. Uncommon but specific features are ganglion cells and nerve fiber proliferation within the lamina propria as well as mucosal fat. They show the least degree of glandular distortion and no thick mucin. PJS polyps (Fig. 4B) derive usually from the small or large bowel, are often exophytic, measure on average 0.54 cm, and are seldom eroded. They typically have expanded edematous and fibrotic lamina propria with substantial acute and chronic inflammation. They are characterized by dilated distorted glands filled with mucin, which is often thick. Practically all PJS polyps, either from the stomach, small or large bowel, show smooth muscle proliferation in the lamina propria, not uncommonly of grade 2. Of 103 PJPs, only 1 had no smooth muscle proliferation. Since smooth muscle proliferation is seen also in the other groups, one should look for other 10
ACCEPTED MANUSCRIPT features, for example, dilated glands containing thick mucin, in order to make a diagnosis of PJP.
T
The polyps of JuvPS (Fig. 4C) are usually colonic, large (mean 1.06 cm), and often exophytic.
RI P
They are characterized by strikingly edematous and fibrotic markedly inflamed lamina propria, often with surface erosion and formation of granulation tissue. They practically
SC
always show glandular distortion with dilated cystic glands filled with mucin, which is frequently thick. This group shows the least degree of smooth muscle proliferation in the
MA NU
lamina propria; almost half of these polyps show no smooth muscle proliferation at all. Nonsyndromic hamartomatous polyps (Fig. 4D) are similar to JuvPS polyps; however, they are more often colonic, are smaller, show more frequent and widespread smooth muscle
ED
proliferation, and are less likely to contain thick mucin.
PT
Discussion:
The hamartomatous polyposis syndromes represent a small but significant number
CE
of the inherited gastrointestinal cancer predisposition syndromes [1-5, 18]. CS and PJS also carry a substantial risk for developing malignancies outside the gastrointestinal tract [1-5,
AC
12-14, 18]. Therefore, although these syndromes are rare, their proper identification is very important for appropriate disease management and prevention, since the malignant potential in these syndromes is quite high, and the proper surveillance for each syndrome is different [18]. Various polyp types are reported to occur in CS, including hamartoma, juvenile polyp, lipoma, inflammatory/hyperplastic polyp, ganglioneuroma, lymphoid hyperplasia, and adenoma [2]. Whether such diverse polyp pathology is common in patients with CS is, however, questioned by some experts in this field [5]. According to our results, patients with CS showed some extent of varied polyp histology: the far most common
11
ACCEPTED MANUSCRIPT gastrointestinal polyp was hamartoma, followed by adenoma, inflammatory/hyperplastic polyp and fundic gland polyp. In contrast to previous reports, which claimed that patients
T
with CS can present with multiple juvenile colonic polyps and consequently be misdiagnosed
RI P
as having JuvPS [2, 3], only 1 juvenile polyp was seen among the CS patients in our series. This may be attributed to different terminology (the terms hamartoma and juvenile polyp
SC
are sometimes mixed). However, the CS polyps in our series differed from JuvPS polyps in several aspects: they were less likely to be exophytic, usually lacked the expanded
MA NU
edematous and markedly inflamed lamina propria characteristic of juvenile polyps, and had more conspicuous smooth muscle proliferation in the lamina propria. Additionally, dilated cystic glands filled with mucin, especially thick, as well as surface erosion, were much more common in the JuvPS polyps than in CS polyps. Thus, we do not think that the usual
ED
appearance of CS polyp should mislead towards the diagnosis of JuvPS.
PT
Small CS polyps may be mislabeled as hyperplastic polyps due to their expanded lamina propria [22, 23]. However, in CS polyps the expanded lamina propria is more dense
CE
and fibrous than edematous.
AC
CS polyps differed from the other 3 groups in several aspects: their lamina propria was more fibrous than inflamed and edematous; they had less conspicuous glandular distortion and cysts and lacked thick mucin. In addition, they exhibited few distinguishing features. ganglioneuromatous polyps were evident only in CS patients and not in the other groups (p=0.02). These polyps were not common, comprising 10% of the CS polyps, but they were seen in 7 (46.7%) CS patients. Our study supports the results of previous studies, which described stromal ganglion cells and nerve fibers in CS polyps [2]. This feature was neither found in a large series of JuvPS, nor in PJS studies [18], and it was not evident in the JuvPS or PJS polyps in our series as well. Another distinctive feature was adipose tissue proliferation in the lamina propria, seen in 14% of CS polyps and practically in none of the other polyps
12
ACCEPTED MANUSCRIPT (p=0.02). This feature is mentioned occasionally in the literature as a possible histological component of CS hamartomatous polyps [25-27].
T
The characteristic PJS polyp described in the literature is composed of architecturally
RI P
complex glands lined by a cytologically bland epithelium [2]. The glands are separated from each other by extensive smooth muscle proliferation forming a distinctive tree-like
SC
configuration. This unique pattern is considered characteristic of small bowel polyps, and less prominent or absent from gastric and colonic polyps [2]. Practically all PJS polyps in our
MA NU
study showed some degree of smooth muscle proliferation; however, broad bands of smooth muscle dissecting the lamina propria were seen more frequently in small bowel compared to large bowel or gastric polyps (p<0.001). Yet, grade 2 smooth muscle
ED
proliferation was evident in 53% and 39% of the colonic (Fig. 3B) and gastric PJS polyps, respectively. Thus, although grade 2 smooth muscle proliferation is more common in small
PT
bowel PJS polyps, seen in 88% of them, it is seen quite frequently in colonic PJS polyps as well. Noteworthy, even though smooth muscle proliferation, particularly of grade 2, was
CE
significantly more common in PJS polyps than in CS, JuvPS and non-syndromic
AC
hamartomatous polyps (p<0.001), it was not restricted to it. Smooth muscle fibers within the lamina propria were seen in the majority of polyps from all groups (Table 2 and Figs. 3C, 3D). Hence, finding smooth muscle in the lamina propria of a gastrointestinal hamartomatous polyp does not indicate that it is a PJS polyp. Rather, the specific classification of a hamartomatous polyp should be based on additional features. Also, our observation of smooth muscle fibers in all hamartomatous polyp types does not agree with the view that there is no smooth muscle infiltration in JuvPS polyps [5]. Smooth muscle proliferation in the lamina propria, especially in the stomach, is a feature shared with other polyps: JuvPS polyps as well as hyperplastic polyps [15]. Presence of strikingly edematous and inflamed lamina propria with surface erosion and granulation
13
ACCEPTED MANUSCRIPT tissue formation is suggestive of JuvPS or hyperplastic polyp. Thick mucin, on the other hand, is more consistent with PJS or JuvPS polyp.
T
Mucin cysts are a well-known finding seen in intestinal PJS polyps [11]. Our results
RI P
support this notion, as dilated cystic glands filled with mucin were seen in 90% of the PJS polyps in our series, and 44% had thick mucin. The PJS polyps in our study shared this
SC
feature with JuvPS polyps, which had dilated cystic glands in 97% of cases, and thick luminal mucin in 47%. Additional features common to PJS and JuvPS in our series were edematous
MA NU
fibrotic and markedly inflamed lamina propria with chronic active inflammation. Hence, juvenile polyps may be mistaken as PJS polyps, and vise versa. On the other hand, JuvPS polyps were more often colonic, typically larger, much more likely to exhibit surface erosion,
ED
and less likely to show widespread smooth muscle proliferation. Previous studies stated that the polyps associated with JuvPS differ from the non-
PT
syndromic juvenile polyps: their surface epithelium is rarely eroded, in contrast to sporadic
CE
juvenile polyps, which generally show erosion of the external surface. Additionally, their lamina propria is less conspicuously expanded compared to the solitary juvenile polyps, and
AC
some of them show focal dysplasia, a feature that is very unusual in sporadic juvenile polyps [1, 19, and 23]. These studies also found that solitary juvenile polyps occasionally show a serrated architecture. Most of the above statements were not confirmed by our results. In fact, we found many shared features by sporadic and syndromic juvenile polyps. Typically, they were both large and frequently exophytic polyps with surface erosion, inflamed lamina propria, and dilated cystic glands. Moreover, in contrast to the previous studies, in our series expansion and edema of the lamina propria were more prominent in the syndromic than in the sporadic juvenile polyps (p=0.004), as well as thick luminal mucin (p=0.02). Smooth muscle proliferation in the lamina propria was more conspicuous in the sporadic polyps (p=0.03). In our series JuvPS polyps differed from sporadic hamartomatous polyps by
14
ACCEPTED MANUSCRIPT location: although colonic predominance was seen in both, almost all sporadic polyps were colonic and none were of gastric origin, while the syndromic ones were more equally
T
distributed between the colon, small bowel and stomach. Glandular serration was a very
RI P
rare occurrence in our study, including in the juvenile polyps. Our results, with no evidence of dysplasia in sporadic hamartomatous polyps, support the statement that dysplasia is a
SC
very rare occurrence in these polyps. Furthermore, low-grade dysplasia was found to be significantly more common in JuvPS than in sporadic hamartomatous polyps (p=0.045),
MA NU
supporting the results of previous studies [19, 23].
No statistically significant differences were observed while comparing the histological features in each group between patients with and without detectable mutation.
ED
Although it is possible that the histological features of the polyps correlate with the phenotypic syndromes and not with their genotype, it may also be due to the small sample
PT
size.
CE
In summary, in this study we aimed to find distinctive features that would enable the pathologist raising the possibility of specific hamartomatous polyposis syndrome based
AC
on microscopic examination of hamartomatous polyps. Despite overlapping features, we managed to define histological features characteristic of CS, PJS, JuvPS, and sporadic hamartomatous polyps. Diagnosing the specific hamartomatous polyposis syndrome is important, as each is associated with specific risks of malignancy and requires different surveillance. Meticulous examination of gastrointestinal hamartomatous polyps with attention to specific histological features may facilitate the diagnosis of these rare syndromes.
15
ACCEPTED MANUSCRIPT Acknowledgments: We acknowledge the use of the Genetic Counseling Shared Resource supported by
RI P
T
P30 CA042014 awarded to Huntsman Cancer Institute.
SC
References:
MA NU
1. Calva D, Howe JR. Hamartomatous polyposis syndromes. Surg Clin North Am 2008;88:779817.
2. Gammon A, Jasperson KW, Kohlmann W, Burt RW. Hamartomatous polyposis syndromes. Best Pract Res Clin Gastroenterol 2009;23:219-231.
ED
3. Jasperson KW, Tuohy TM, Neklason DW, Burt RW. Hereditary and familial colon cancer.
PT
Gastroenterology 2010;138:2044-2058.
CE
4. Gryfe R. Inherited Colorectal Cancer Syndromes. Clin Colon Rectal Surg 2009;22:198-208. 5. Zbuk KM, Eng C. Hamartomatous polyposis syndromes. Nat Clin Pract Gastroenterol
AC
Hepatol 2007;4:492-502.
6. Stanich PP, Owens VL, Sweetser S, et al. Colonic polyposis and neoplasia in Cowden Syndrome. Mayo Clin Proc 2011;86:489-492. 7. Schreibman IR, Baker M, Amos C, McGarrity T. The hamartomatous polyposis syndromes: a clinical and molecular review. Am J Gastroenterol 2005;100:476-490. 8. Pilarski R, Eng C. Will the real Cowden Syndrome please stand up (again)? Expanding mutational and clinical spectra of the PTEN hamartoma tumour syndrome. J Med Genet 2004;41:323-326.
16
ACCEPTED MANUSCRIPT 9. Jeghers H, Mc KV, Katz KH. Generalized intestinal polyposis and melanin spots of the oral mucosa, lips and digits; a syndrome of diagnostic significance. N Engl J Med 1949;241:1031-
T
1036.
SC
Peutz-Jeghers syndrome. J Med Genet 2004;41:327-333.
RI P
10. Amos CI, Keitheri-Cheteri MB, Sobripour M, et al. Genotype-phenotype correlations in
11. McGarrity TJ, Kulin HE, Zaino RJ. Peutz-Jeghers Syndrome. Am J Gastroenterol
MA NU
2000;95:596-604.
12. Giardiello FM, Brensinger JD, Tersmette AC, et al. Very high risk of cancer in familial Peutz-Jeghers syndrome. Gastroenterology 2000;119:1447-1453.
ED
13. Lim W, Olscheang S, Keller JJ, et al. Relative frequency and morphology of cancers in STK11 mutation carriers. Gastroenterology 2004;126:1788-1794.
PT
14. Giardiello FM, Welsh SB, Hamilton SR, et al. Increased risk of cancer in the Peutz-Jeghers
CE
syndrome. N Engl J Med 1987;316:1511-1514. 15. Lam-Himlin D, Park JY, Cornish TC, Shi C, Montgomery E. Morphologic characterization of
AC
syndromic gastric polyps. Am J Surg Pathol 2010;34:1656-1662. 16. Mehenni H, Resta N, Guanti G, et al. Molecular and clinical characteristics in 46 families affected with Peutz-Jeghers syndrome. Dig Dis Sci 2007;52:1924-1933. 17. Burt RW, Bishop DT, Lynch HT, et al. Risk and surveillance of individuals with heritable factors for colorectal cancer. WHO Collaborating Centre for the Prevention of Colorectal Cancer. Bull World Health Organ 1990;68:655-665. 18. Howe JR, Mitros FA, Summers RW. The risk of gastrointestinal carcinoma in familial juvenile polyposis. Ann Surg Oncol 1998;5:751-756.
17
ACCEPTED MANUSCRIPT 19. Jass JR, Williams CB, Bussey HJ, Morson BC. Juvenile polyposis – a precancerous condition. Histopathology 1988;13:619-630.
T
20. Howe JR, Roth S, Ringold JC, et al. Mutations in the SMAD4/DPC4 gene in juvenile
RI P
polyposis. Science 1998;280:1086-1088.
SC
21. Howe JR, Sayed MG, Ahmed AF, et al. The prevalence of MADH4 and BMPR1A mutations in juvenile polyposis and absence of BMPR2, BMPR1B, and ACVR1 mutations. J Med
MA NU
Genetics 2004;41:484-491.
22. Sweet K, Willis J, Zhou XP, et al. Molecular classification of patients with unexplained hamartomatous and hyperplastic polyposis. JAMA 2005;294:2465-2473.
Practice 2008;204:431-447.
ED
23. Jass JR. Colorectal polyposes: from phenotype to diagnosis. Pathology Research and
PT
24. Burkart AL, Sheridan T, Lewin M, Fenton H, Ali NJ, Montgomery E. Do sporadic Peutz-
CE
Jeghers polyps exist? Experience of a large teaching hospital. Am J Surg Pathol 2007;31:1209-1214.
AC
25. Allen BA, Terdiman JP. Hereditary polyposis syndromes and hereditary non-polyposis colorectal cancer. Best Pract Res Clin Gastroenterol 2003;17:237-258. 26. Chan OTM, Haghighi P. Hamartomatous polyps of the colon. Ganglioneuromatous, stromal and lipomatous. Arch Pathol Lab Med 2006;130:1561-1566. 27. Merg A, Howe JR. Genetic conditions associated with intestinal juvenile polyps. Am J Med Genet 2004;129C:44-55.
18
ACCEPTED MANUSCRIPT Legends to figures: Figure 1: Distinguishing morphological features (H&E).
RI P
T
Figure 1A: Surface erosion (arrow), often accompanied by underlying granulation tissue, was common in JuvPS polyps and sporadic hamartomatous polyps and very rare in CS and PJS
SC
polyps ( X200).
Figure 1B: Expanded edematous lamina propria was more common in PJS and JuvPS polyps
MA NU
than in CS and sporadic hamartomatous polyps (X200).
Figure 1C: Chronically inflamed lamina propria was seen in almost all PJS, JuvPS and sporadic hamartomatous polyps; it was considerably less common in CS polyps (X400).
ED
Figure 1D: Active inflammation was a common feature of PJS, JuvPS and sporadic
PT
hamartomatous polyps, and was significantly less common in CS polyps (X200). Figure 1E: Dilated cystic glands filled with mucin were evident in almost all PJS, JuvPS, and
CE
sporadic hamartomatous polyps; they were remarkably less common in CS polyps (X40).
AC
Figure 1F: Dense deeply eosinophilic thick luminal mucin seen within glands and secreted to the mucosal surface was significantly more frequent in PJS and JuvPS polyps than in sporadic hamartomatous polyps. It was not seen at all in CS polyps (X40).
Figure 2: Uncommon but specific features seen only in CS polyps (H&E). Figure 2A: Ganglion cells scattered within the lamina propria were a distinctive feature seen in 9% of CS polyps and none of the other polyps (X400). Figure 2B: Nerve fiber proliferation within the lamina propria was evident in 10% of CS polyps and practically none of the other polyps (X200). 19
ACCEPTED MANUSCRIPT Figure 2C: Mucosal fat was observed in 14% of CS polyps and practically none of the other polyps; lymphoid follicles (arrow) were significantly more common in CS polyps compared to
RI P
T
the other 3 groups (X100).
SC
Figure 3: Smooth muscle fibers within the lamina propria were significantly more common and widespread in PJS polyps, but seen also in the other polyp types (H&E).
MA NU
Figure 3A: Small bowel PJS polyp with broad bands of smooth muscle fibers dissecting the lamina propria (X20).
Figure 3B: Colonic PJS polyp with extensive smooth muscle proliferation within the lamina
ED
propria; although smooth muscle proliferation was considerably more widespread in small than large bowel PJS polyps, it was quite extensive in many colonic PJS polyps (X20).
PT
Figure 3C: Colonic CS polyp with smooth muscle fibers in the lamina propria (arrow) (X100).
AC
(X100).
CE
Figure 3D: Colonic JuvPS polyp with smooth muscle fibers in the lamina propria (arrow)
Figure 4: Classic polyps illustrating some of the characteristic morphological features for each group (H&E). Figure 4A: Colonic CS polyp with fibrotic mildly inflamed lamina propria, lymphoid aggregate (arrow), and no significant glandular distortion. Smooth muscle fibers were evident on higher magnification (X100). Figure 4B: Characteristic small bowel PJS polyp with broad bands of smooth muscle fibers forming a tree-like configuration; expanded edematous-fibrotic and inflamed lamina propria
20
ACCEPTED MANUSCRIPT and dilated glands are also evident. Thick mucin is another common feature not depicted in this picture (X20).
T
Figure 4C: Colonic JuvPS polyp with dilated cystic glands filled with mucin, focally thick
RI P
(deeply eosinophilic and dense mucin seen in the middle and left bottom fields), separated by markedly inflamed lamina propria; surface erosion is another common feature not
SC
depicted in this picture (X40).
MA NU
Figure 4D: Sporadic colonic hamartomatous polyp appearing similar to JuvPS polyp. These polyps usually exhibited more smooth muscle proliferation and were less likely to contain
AC
CE
PT
ED
thick mucin compared to JuvPS polyps (X20).
21
AC
CE
PT
ED
MA NU
SC
RI P
T
ACCEPTED MANUSCRIPT
Figure 1
22
AC
CE
PT
ED
MA NU
SC
RI P
T
ACCEPTED MANUSCRIPT
Figure 2
23
ED
MA NU
SC
RI P
T
ACCEPTED MANUSCRIPT
AC
CE
PT
Figure 3
24
ED
MA NU
SC
RI P
T
ACCEPTED MANUSCRIPT
AC
CE
PT
Figure 4
25
ACCEPTED MANUSCRIPT Table 1 – Polyp histology in each category
Peutz Jegher's Syndrome (N=13) 94 0 3
Juvenile polyposis (N=12) 12 69 0
Sporadic polyps (N=32)
T
Cowden Syndrome (N=15) 112 1 8
AC
CE
PT
ED
MA NU
SC
RI P
Hamartoma 7 Juvenile Polyp 30 Hyperplastic 0 polyp Adenoma 14 1 6 0 Fundic gland 8 3 1 0 polyp Miscellaneous 4 2 0 0 Total 147 103 88 37 Miscellaneous includes foveolar hyperplasia (PJS),submucosal leiomyoma, benign spindle cell lesion, and prominent lymphoid follicle (CS).
26
ACCEPTED MANUSCRIPT Table 2 – comparison of gross and histological features of gastrointestinal polyps from Cowden syndrome, Peutz-Jegher's syndrome, juvenile polyposis syndrome and nonsyndromic hamartomatous polyps
Small bowel Stomach
PJS
p value
147
37
88
103
NA
85% 3% 12%
95% 5% 0%
75% 7% 18% 1.06 (0.84, 1.29)
37% 41% 22% 0.54 (0.31, 0.77)
<0.001
Polyp size (cm): mean 0.35 (95% CI) (0.16,
<0.001
12%
43%
47%
44%
0.01
Surface erosion
3%
68%
68%
4%
<0.001
Expanded edematous LP
24%
22%
55%
54%
0.004
Expanded fibrotic LP
65%
92%
83%
70%
0.20
Inflamed LP
41%
92%
92%
88%
<0.001
Chronic inflammation
41%
92%
92%
88%
<0.001
Active inflammation
18%
81%
85%
70%
<0.001
Gland Distortion
76%
100%
99%
97%
0.01
Dilated cystic glands filled with mucin Thick luminal mucin
36%
92%
97%
90%
<0.001
0%
19%
47%
44%
<0.001
Stromal ganglion cells
10%
0%
0%
0%
0.02
Nerve fiber proliferation
10%
3%
0%
0%
0.02
Stromal (lamina propria) fat Lymphoid follicles
14%
0%
0%
1%
0.02
50%
19%
30%
21%
0.007
16% 66% 18% 8%
5% 73% 22% 0%
47% 45% 8% 0%
1% 35% 64% 1%
<0.001
2%
0%
1%
0%
0.60
2%
0%
1%
3%
0.81
Low grade dysplasia
10%
0%
14%
1%
0.045
Adenoma
10%
0%
7%
1%
0.16
AC
CE
PT
Exophytic with stalk
ED
MA NU
0.54)
0.66 (0.38, 0.93)
T
JuvPS
RI P
Location Colon
Control
SC
Total polyp count
CS
Smooth muscle fibers in lamina propria 0 1 2 Onion skinning of fibroblasts around glands Lobular clusters of glands Serrated architecture
CS – Cowden Syndrome; JuvPS – Juvenile polyposis syndrome; PJS – Peutz-Jeghers syndrome; LP – lamina propria.
27
0.12
ACCEPTED MANUSCRIPT Table 3 – Characteristic features of specific hamartomatous polyp types Peutz-Jegher's syndrome
Juvenile polyposis syndrome
Location
Usually colon
Small or large bowel
Usually colon
Mean size
Small (mean 0.4 cm)
Small (mean 0.5 cm)
Large (mean 1.1 cm)
Architecture Erosion Lamina propria
Sessile Almost never Expanded, fibrotic
Inflammation
Mild chronic
Exophytic Almost never Expanded, edematous and fibrotic Marked chronic active
Cystic glands Thick mucin Smooth muscle proliferation Lymphoid follicles Ganglion cells Nerve fibers Mucosal fat
Least occurrence No Mild
Common Common Highest level
Exophytic Common Expanded, strikingly edematous and fibrotic Marked chronic active with granulation tissue Almost always Common Common
No
Occasionally
Uncommon
No No No
No No No
No Very rare No
28
RI P
SC
MA NU
AC
CE
PT
Rare Rare Rare
ED
Common
T
Cowden syndrome
Sporadic hamartomatous polyp Almost always colon Medium (mean 0.7 cm) Exophytic Common Expanded, fibrotic
Marked chronic active with granulation tissue Almost always Uncommon Common