Advertisement

International Journal of Colorectal Disease

, Volume 30, Issue 12, pp 1617–1626 | Cite as

Genome-wide copy number changes and CD133 expression characterized distinct subset of colon polyps: differentiation between incidental polyps and cancer-associated polyps

  • Chih-Yung Yang
  • Ju-Yu Tseng
  • Chian-Feng Chen
  • Teh-Ying Chou
  • Hong-Wei Gao
  • Chia-Ling Hua
  • Chi-Hung Lin
  • Jen-Kou Lin
  • Jeng-Kai JiangEmail author
Original Article

Abstract

Purpose

Colorectal polyps are generally believed to be the precursors of colorectal cancers (CRC); however, the proportion and speed of progression differed widely in different subsets of polyps. Using microarray-based comparative genomic hybridization (aCGH) platform and CD133 immunostaining, we characterized colon polyps according to their association with CRC that developed in the same individual.

Patients and methods

aCGH was performed to unveil genomic changes in 18 cancer-synchronous polyps (CSP), and 9 cancer-preceding polyps (CPP), together with their corresponding cancers and 16 cases of incidental polyps (IP), were examined for comparison. aCGH profiles were analyzed to determine the clonal relationship (CR) between the paired adenoma and carcinoma. CD133 expressions in each subset of polyps were quantified by immunohistochemistry (IHC) staining.

Results

Progressive genomic changes were observed from IP, CSP/CPP to CRC; they encompass an entire chromosomal region in IP and sub-chromosomal region in CSP/CPP and CRC. CR analyses demonstrated that 50 % of CSP and 67 % of CPP were clonally related to the concurrent or later developed carcinomas, respectively. The CD133 expression levels were significantly higher in CSP/CPP than those in IP (P < 0.0001) and even higher in CSP/CPP that were clonally related to their corresponding carcinomas than CSP/CPP that were unrelated (P < 0.05).

Conclusions

There were more genomic changes in CSP/CPP than IP; more than half of the CSP/CPP were clonally related to the corresponding carcinomas. Genomic changes at sub-chromosomal regions and/or high CD133 expression were associated with CSP/CPP and highlighted their carcinogenic potential.

Keywords

Colorectal polyp Colorectal cancer CD133 Array CGH 

Introduction

Colorectal cancer (CRC) is the most frequently diagnosed cancer and the third leading cause of cancer death in Taiwan [1]. The emergence of CRC has been believed to follow the well-known “adenoma–carcinoma sequence,” a term that describes the stepwise progression from normal to dysplastic epithelium to carcinoma associated with accumulation of multiple clonally selected genetic changes [2, 3, 4]. This concept becomes the basis for preventing CRC by endoscopic removal of precursor lesions. However, more than 95 % of the adenomas remain benign for decades, while, on the other hand, the time for the adenoma–carcinoma progression varies widely. Hence, until now, there have been no reliable criteria that can predict adenoma progression or recurrence, and the follow-up protocol following polypectomy remains unclear. It is therefore of clinical importance to identify biomarkers from the adenoma that can predict the polyp-bearing patients’ adenoma and/or carcinoma risk [4, 5, 6, 7].

According to the conventional unifocal cell lineage view of CRC carcinogenesis, each transformed cell gives rise to only a single cell lineage that advances to adenoma and to carcinoma [8]. Each cell lineage is independent from the descendants of other transformed cells [8, 9, 10, 11, 12]. In this scenario, one would predict that the adenoma coexisting with CRC (i.e., cancer-synchronous polyp or CSP) should be clonally unrelated to the concurrent cancer. Similarly, an adenoma previously removed (i.e., cancer-preceding polyp or CPP) and the CRC developed later in the same patient should come from distinct cell lineages. Here, we used comparative genomic hybridization technique that employed microarray as a template (array-based CGH or microarray-based comparative genomic hybridization (aCGH) [13, 14, 15, 16, 17]) to unveil and quantify genome-wide copy number imbalances [18, 19, 20, 21, 22]. By comparing the aCGH profiles between CSP/CPP and their corresponding cancers, we were able to determine their clonal relationships.

CD133 (also known as Prominin-1 or AC133), a five-transmembrane protein, is a marker for colon cancer stem cell or caner initiating cell. Several authors demonstrated that CD133 expression in tumor tissue adversely affects CRC patients’ survival outcome [23, 24]. Nonetheless, reports on CD133 expression in colorectal adenoma were scanty and the clinical significance was unclear. Here, we compared CD133 expression between different subsets of colon polyps under the assumption that CD133 immunostaining would be different in each subset of colon polyps and would correlate with clonal relationship between the CSP/CPP and their corresponding cancers.

Materials and methods

Clinical specimens and DNA extraction

From May 2007 through May 2008, a prospective study was conducted at Taipei Veterans General Hospital and Tri-Service General Hospital Taipei, Taiwan, following the ethical regulations approved by the Institutional Review Board. Eighteen patients were admitted with the diagnosis of concurrent CSP and CRC. Sixteen cases of IP were enrolled; they were volunteers who received physical checkup when colorectal polyps were found incidentally. They did not have a family history of colorectal cancer, nor of colorectal polyps, and during the follow-up period, no CRC was found. For CPP, a retrospective survey of clinical database was conducted to look for CRC patients who had undergone polypectomy at least 1 year prior to CRC resection. The paired CPP/CRC tissues were retrieved from the archive and their genomic DNA extracted using modified phenol-chloroform method [25] or QIAamp Mini Kit according to the manufacturer’s protocols (Qiagen). All tissue samples were diagnosed by experienced pathologists (Supplement Fig. S1).

aCGH and data analysis

Genome copy number imbalances were revealed by aCGH method using the Agilent Human Genome Microarray Kit 244K (Agilent Technologies, Santa Clara, CA, USA) as previously described [26], except for case no. 18 (using 185K). DNA extracted from the tumor adjacent normal tissue or from the pooled peripheral blood of 10 healthy donors was used as reference DNA for aCGH in the cases of CSP/CRC or CPP/CRC and IP, respectively.

We defined gene copy number increase at each microarray probe by setting log2 T/R or P/R ≥ 0.3 (T = tumor, P = polyp, R = reference), and gene copy number decrease by log2 T/R or P/R ≤ −0.3. An aCGH (gain or loss) aberration event was determined by Aberration Detection Method 2 (ADM-2, provided by Agilent) [26]. We set the ADM-2 threshold at 5, so a genome region was defined as an event of aCGH gain or loss only when a stretch of DNA showed gene copy number increases or decreases for more than five consecutive microarray probes and marked as a dot above or below the aCGH profile, respectively (Fig. 2). The aCGH results were validated by quantitative PCR, which showed over 80 % consistency (Supplementary Fig. S2). The microarray raw data had been deposited in European Bioinformatics Institute (EBI) ArrayExpress database (ArrayExpress accession: E-MEXP-3149).

Clonal relationship analysis

The clonal relationship (CR) between the paired adenoma–carcinoma was quantified according to the methods described previously [27, 28]. An aCGH gain or loss event shared by a paired adenoma and carcinoma was defined by not only having the same location but also the same extent of lesion (see Fig. 2b). We first calculated the probability (X) for all shared aCGH abnormality events to occur purely by chances. If I or R represents the aCGH gains and losses found in a polyp and its corresponding carcinoma, respectively, then IR = c 1, c 2, c 3, . . ., where c k is representing the shared aCGH lesions by the adenoma–carcinoma pair. For any given aCGH event to occur by chance, the probability is P(c i ) (i = 1, 2, 3,…, k), which equals to the number of samples carrying c i divided by all samples examined. The overall probability of having all shared aCGH events to occur by chance is therefore P(c 1) × P(c 2) × P(c 3) × . . . × P(c k ) = X (0 < X < 1). We defined CR value as 1 − X. Larger CR values (closer to 1) indicate positive clonal relationship between the paired adenoma–carcinoma. The cutoff CR value was set at 0.8 (dashed lines, Fig. 3a, b).

CD133 immunostaining and quantification

CD133 immunohistochemistry (IHC) was performed on formalin-fixed paraffin-embedded tissue sections (3-μm thick) mounted on silane-coated microscope slides (Dako) and treated with CD133/AC133 antibody (MACS, Miltenyi Biotec) at 1:50 dilution. Five medium-power fields per section were first viewed by the experimenter and then confirmed by a naïve observer who scored the overall intensity of IHC staining as 1~3 (I, Fig. 4c). If >50 % of the cells in a gland (asterisks, Fig. 4a, b) were stained positively for CD133, the gland is considered CD133-positive. For each tissue section, at least 100 glands were examined; among them, the percentage of CD133-positive glands was calculated as D. The value of D × I for each sample, ranging from 0 to 3, represents the level of its CD133 content.

Statistical analysis

The correlation between the clonal relationships of the paired adenoma–carcinoma and the patients’ clinical manifestations (Table 1) was analyzed by independent t test using Statistical Package for Social Science (SPSS®, SPSS, Inc., Chicago, IL).
Table 1

Descriptions of the polyps and tumors used in this study

Case

Gender

Age

Pathological diagnosis

CRC

Remark

Grading

Staging (TNM)

Cancer synchronous polyps (CSP)

Distance between polyp and concurrent tumor (cm)

Clonal relationship

1

F

76

Tubular adenoma

Well

0 (is, 0, 0)

2

+

2

M

73

Tubular adenoma

Moderately

I (2, 0, 0)

10

3

M

68

Tubular adenoma

Moderately

IIA (3, 0, 0)

4.7

+

4

M

75

Tubular adenoma with moderate dysplasia

Moderately

IIA (3, 0, 0)

3.5

+

5

M

80

Tubular adenoma

Moderately

IIA (3, 0, 0)

2.58, >15

+

6

F

73

Tubular adenoma

Moderately

IIA (3, 0, 0)

>15

+

7

M

84

Tubular adenoma

Moderately

IIA (3, 0, 0)

4

8

F

50

Tubular adenoma

Moderately

IIA (3, 0, 0)

ND

9

M

68

ND

Moderately

IIA (3, 0, 0)

7.08

10

M

79

ND

Moderately

IIA (3, 0, 0)

>15

11

M

70

ND

Moderately

IIA (3, 0, 0)

5

12

M

72

ND

Moderately

IIIB (3, 1, 0)

2.14

13

M

86

ND

Moderately

IIIB (3, 1, 0)

5

+

14

F

59

Tubular adenoma with focal moderate dysplasia

Moderately

IIIC (3, 2, 0)

ND

+

15

M

51

ND

Moderately

IIIC (3, 2, 0)

2.67

16

M

70

Tubular adenoma with focal moderate dysplasia

Poorly

IV (2, 1, 1)

5.26

+

17

M

73

Tubular adenoma with moderate dysplasia, adenoma T1

Moderately

IV (3, 1, 1)

3.75

18

F

72

ND

Moderately

ND (3, x, 0)

4

+

Cancer preceding polyps (CPP)

Time between polypectomy and CRC surgery (month)

Distance between polyp and subsequent tumor (cm)

Clonal relationship

19

M

81

Tubular adenoma

Moderately

I (2, 0, 0)

65

7

20

F

63

Villous adenoma

Moderately

IIIA (3, 1, 0)

23

74

+

21

M

55

Tubular adenoma

Moderately

IIIC (3, 2, 0)

14

44

22

M

81

Tubular adenoma

Well

IIIC (4, 2, 0)

91

41

23

M

73

Villous adenoma with focal malignant

Moderately

IV (4, 1, 1)

74

2

+

24

M

71

Tubular adenoma

Moderately

IV (3, 0, 1)

40

ND

+

25

M

80

Tubular adenoma

Well

IIIB (3, 1, 0)

12

10

+

26

M

71

Tubulovillous adenoma

Moderately

IV (3, 2, 1)

47

95

+

27

M

83

Tubular adenoma

Well

I (1, 0, 0)

38

45

+

Incidental polyps

CGH abnormalitiesa

1

F

46

Tubular adenoma

  

Gains at 6, 9, 13, 20

2

F

55

Tubular adenoma

  

Nil

3

F

70

Villous adenoma with focal severe dysplasia

  

Gains at 7, 8, 12, 19, 21, X

4

M

48

Tubular adenoma

  

Nil

5

M

80

Tubulovillous adenoma with moderate dysplasia

  

Gains at 13, 20

6

F

50

Tubular adenoma

  

Gains at 7, 20

7

M

61

Tubular adenoma

  

Nil

8

F

67

Tubulovillous adenoma

  

Nil

9

F

71

Tubular adenoma

  

Nil

10

M

75

Tubular adenoma

  

Nil

11

F

64

Tubulovillous adenoma with moderate dysplasia

  

Nil

12

F

57

Tubular adenoma

  

Nil

13

M

51

Tubular adenoma

  

Gains at 7

14

F

68

Tubular adenoma

  

Nil

15

M

58

Tubular adenoma with focal severe dysplasia

  

Gains at 6, 13, 20

16

F

58

Tubulovillous adenoma with moderate dysplasia

  

Nil

aIn most cases, the gains involve the entire chromosome indicated

Results

Analysis of the pooled aCGH results

Table 1 demonstrated clinicopathological data of the patients. A total of 18 pairs of CSP/CRC, 9 pairs of CPP/CRC, and 16 cases of IP were enrolled in this study. Distances between the coexisting CSP and CRC ranged between 2 and 15 cm. Time intervals between CPP polypectomy and CRC resection ranged from 12 to 91 months. The pooled aCGH results from our experiments and from a public database consisting of 797 sporadic CRC cases [29] are summarized in Fig. 1. Grossly, a trend toward more genetic changes was observed from IP, cancer-associated polyp (CSP/CPP) to polyp-paired CRC, while those of the latter were similar to those of the sporadic CRC.
Fig. 1

Pooled aCGH results of different subsets of colon polyps (incidental polyp, IP; cancer-synchronous polyp, CSP; cancer-preceding polyp, CPP) and colorectal cancers. Results as indicated show percent aCGH lesions (Y-axis) along the chromosomes (X-axis). aCGH profile of sporadic CRC was from Progenetix database

Clonal relationship between the adenoma–carcinoma pair

The clonal relationship between each pair of CSP/CPP and its corresponding CRC was determined by comparing both their whole-genome aCGH profiles (Fig. 2a) and high-resolution aCGH maps (Fig. 2b), which revealed gain or loss at individual microarray probes. In case no. 3, for example, the overall aCGH profile of the polyp (3P) was almost identical to that of the tumor (3T). When zooming into the small ~3-kbp amplicon located at chromosome 3p (red arrow and shade), we found 36 gene probes. Twenty-eight of these probes exhibited gains (red dots) in the polyp; among these gains, 96 % (27 out of 28) were also found in the tumor. The paired adenoma and carcinoma of case no. 3 is considered clonally related to each other. Two CSP (5Pa and 5Pb) were obtained in case no. 5. From their aCGH profiles, we noticed that many (closed arrowheads), but not all, of the lesions were shared by the adenomas and carcinoma. When zooming into the small ~7-kbp amplicon located at chromosome 1p (green arrow and shade), we found 141 gene probes; 105 and 110 of these probes exhibited gains in 5Pa and 5Pb, respectively. Among these gains, >80 % were also present in 5T. These results support the notion that the paired adenoma–carcinoma of case no. 5 is clonally related to each other.
Fig. 2

Comparisons of aCGH profiles reveal clonal relationships between the paired CSP/CPP and CRC. a Overall aCGH profiles showing ratio plots along the chromosome set and b high-resolution aCGH map showing ratio data at individual microarray probes (dots) are demonstrated for the paired polyp (P) and tumor (T) of cases no. 3, no. 5 (that contains two polyps, 5Pa and 5Pb, and one cancer, 5T), and no. 12. Upward deviations of the ratio trace or dots from the baseline indicate increase in gene copy numbers, and downward deviations indicate copy number decrease. Along the chromosome set shown in (a), a dot/line is marked above or below the ratio trace at the genome regions where gene gains or losses, respectively, stretch for more than five probes. Arrows in (a) indicate the small amplicons further analyzed in (b), where red, green, or black dots indicate microarray probes that exhibit gene gains, losses, or balances

On the contrary, in case no. 12, we found many small amplicons/deletions (open arrowheads) in the polyp, but none of them were shared by the carcinoma. When zooming into the polyp’s small ~7-kbp amplicon located at chromosome 1p (blue arrow and shade), we found 89 of the 125 gene probes exhibited gains in 12P, but only <30 % of them were present in 12T. Note that the gain probes of 12T were dispersed, without forming any “longer than five-probe” stretch to be defined as an amplicon. We concluded from these results that the paired adenoma–carcinoma of case no. 12 is clonally unrelated to each other.

An algorithm (see “Materials and methods”) was conducted to quantify the value of CR for each pair (Fig. 3a for CSP and Fig. 3b for CPP). This method weights individual shared aCGH events differently. For a shared aCGH event that seldom occurred in CRC, its power as a clonality determinant was weighed higher than if the shared aCGH event was frequently found in CRC. The calculated CR values range from 0, when there is no shared aCGH event, to 1, when the aCGH profiles of the pair are identical. By setting the threshold at 0.8, we found that 50 % (9 out of 18) of CSP and 67 % (6 out of 9) of CPP are clonally related to their corresponding CRC. Moreover, we were surprised to note that seven out of nine CPP patients had their cancers diagnosed at advanced stage, despite their routine follow-up after receiving polypectomy (Table 1).
Fig. 3

Clonal relationship (CR) value for each and every pairs of CSP/concurrent tumor (a) and CPP/later developed tumor (b) are calculated (see “Materials and methods”); the threshold is set at 0.8. Filled symbols indicate clonally related adenoma–carcinoma pairs, whereas open symbols indicate clonally unrelated cases. CR between 5Pa and 5Pb is also shown

High CD133 expression in the adenoma is a carcinogenic risk factor

A thorough histopathological examination on CSP/CPP did not reveal any pathological feature that can differentiate them from IP, nor can it determine whether or not they were clonally related to the corresponding CRC (Supplement Fig. S1). Since the clonally related adenoma and carcinoma might originate from the same initiating cell, we decided to perform a series of immunostaining to determine if CD133 could function as a predictive biomarker for adenoma patients’ subsequent carcinogenic risk.

As shown in Fig. 4a, most IP expressed low level of CD133 and had few CD133-positive cells surrounding the gland (asterisks). On the other hand, the paired CSP/CPP and carcinoma contained highly abundant CD133 (Fig. 4b). The presence of CD133 at the apical membrane facing the gland lumen was evident in the carcinoma (arrowheads, inset), while very little apical CD133 was observed in CSP/CPP. Immunostaining of CD133 was quantified by calculating D × I value (see “Materials and methods”), where D represents the percent CD133-positive glands and I stands for staining intensity. Note the levels of CD133 immunostaining were significantly higher in CSP/CPP than IP (P < 0.0001, Fig. 4d). An even higher CD133 immunostaining was noted in CPP that were clonally related to carcinomas than CPP that were unrelated (P < 0.05, Fig. 4e). Tumor adjacent normal tissues had very little CD133 immunostaining.
Fig. 4

CD133 immunostaining in incidental polyp (a), cancer-associated polyps (CSP/CPP), and corresponding cancers (b). Asterisks mark the glands. The level of CD133 IHC is scored quantified by calculating D × I, where D stands for the percentage of CD133-positive glands, and I stands for the staining intensity of IHC that is scored 1~3 as exemplified (c). Each symbol in (d, e) represents one patient; mean for each category is shown (bar). Note that CD133 staining for CSP/CPP is comparable to CRC but significantly higher (P < 0.0001 by Mann–Whitney test) than IP (d). Tumor adjacent tissues show very low CD133. CD133 is higher in CPPs that are clonally related to the later developed cancers than the ones that are clonally unrelated (e, P < 0.05). Scale bar is 50 μm

Discussion

Although only a small proportion of adenomas progress to malignancy, the treatment strategy and follow-up protocol differ little among different subsets of polyps and depend largely on clinicopathological characteristics of the polyps. In this study, unlike previous authors, we stratified the polyps according to their association with CRC developed in the same individual. Using aCGH as a platform, we found that there were progressive genomic changes from incidental polyp to cancer-associated polyps (CSP/CPP). Clonal relationship analysis and CD133 expression further stratified CSP/CPP and their carcinogenic risks.

aCGH is a powerful tool for the screening of chromosomal variations on a whole-genome scale with high resolution. Using this method, copy number changes at the level of 5–10 kbp of DNA sequences could be detected. Nonetheless, the results necessitate validation. Our aCGH data were validated by qRT-PCR which showed a >80 % consistency. Furthermore, the detected aCGH defects of CRC that paired with CSP/CPP were very similar to the sporadic CRC from public database, both showing sub-chromosomal gains in chromosomes 7, 8q, 13, and 20 and losses in chromosomes 1p, 4, 8p, 14, 15, 17p, and 18, as were also reported by other authors [30]. On the other hand, adenomas (IP/CSP/CPP) contain much less aCGH abnormalities than CRC [7, 31, 5, 25]. Around one third of the IP have aCGH gains and, interestingly, such gains often encompass the entire chromosomes, such as in chromosomes 7, 13, and 20. In contrast, gains or losses in CSP/CPP, like those of CRC [7, 25, 31, 32, 33], occur mostly at sub-chromosomal regions. We then intended to find a genomic marker that can differentiate among IP, CSP/CPP, yet no significant characteristic that can differentiate each subset was found.

Systemic cancer progression can be accounted for by at least two models [34]. The commonly accepted linear progression model states that the accumulation of genetic changes in selected cell clones may first lead to primary tumor formation and subsequently to regional spread and distant metastasis. However, recent evidence in both animal model [35] and clinical patients [35, 36, 37] suggests an alternative parallel progression model whereby dissemination of some tumor cells may occur at an early stage of cancer before clinical metastasis is evident. Further genetic and epigenetic alterations at these distant microscopic lesions then cause some of them to “grow” into metastatic tumors. Moreover, there are evidences supporting the field cancerization. The proliferation of a common precursor or cancer initiation cells that disseminate over a wide distance contributes to the field effect in CRC, head and neck cancer, and urothelial cancer [38, 39, 40, 41, 42]. However, it is unclear when or by what mechanism the early dissemination of primary tumor occurs. In this study, based on clonal relationship analysis, we proposed the parallel progression view alluded to cancer spreading may actually occur earlier at the cancer initiation stage. The evidence came from the adenoma and carcinoma tissues obtained from the same patients. Following the conventional unifocal carcinogenesis pathway, the progression from a transformed cell to an adenoma and to a carcinoma is independent from each other and different cell lineages progress at different rates. Removal of a polyp essentially eliminates the chance of having a cancer developed from that specific cell lineage. Thus, CPP should be clonally unrelated to the later developed CRC, and CSP should be clonally unrelated to the concurrent cancer. However, we found that more than half of CSP/CPP were clonally related to the tumors. This observation can be explained by the multifocal cell lineage model, whereby a transformed cell can proliferate and spread (microscopically) at very early stage, forming multifocal cell lineages that can undergo adenoma–carcinoma progressions independently and in parallel. Only a small proportion of the polyps later develop carcinomas and, among these cancers, we found that the invasion and metastasis appear to advance faster if the cancers were clonally related to the preceding or synchronous adenoma than if they were clonally unrelated, i.e., carcinogenesis via the multifocal pathway exhibits a more malignant outcome than the unifocal pathway.

Conventional histopathological examinations on tissue sections are not able to differentiate CPP/CSP from IP. However, we noticed that genomic losses or gains at sub-chromosomal regions or high CD133 expression are associated with cancer-associated polyps (CSP/CPP), especially those undergo multifocal progressions. Detection of CD133 expression in polyp has been readily applicable in clinical examinations. CD133-positive cells are thought to represent the putative colorectal cancer-initiating cells and play crucial roles in cancer metastasis and conferring resistance to chemo- or radiotherapy [43, 44, 45, 46, 47, 48]. In CRC, CD133 expression is an independent prognostic marker for metastasis and low survival [24, 49, 50, 51]. Given that the major challenge facing clinicians is how to determine the appropriate surveillance intervals following the finding of a benign polyp, the high CD133 expression in the polyp is a carcinogenic risk factor which will be a potential implication for decision on medical checkup intervals. The multifocal carcinogenic pathway in the high CD133 content colorectal cancer-initiating cell can therefore be a possible explanation for the patients’ poor clinical outcome.

One drawback of the present study is that we were not able to collect more CPP/CRC pairs. This was the result of a stringent follow-up protocol after polypectomy, making a tumor formation lateral on extremely rare. On the other hand, some of the cancer patients reported a previous polypectomy at other institute certain years ago, making tissue retrieval impossible. Nonetheless, from this limited group of CPP, statistical differences were observed in CR analysis and CD133 immunostaining. One would challenge the definition of IP which might turn into CPP if the follow-up is long enough and a cancer developed. Nonetheless, during the follow-up period (mean = 64.69 ± 18 months, range = 52 to 84 months), no CRC was found in all the IP patients. Also, as we know that only a small proportion of polyps progressed to carcinoma, the chances of carcinoma formation would be minimal among the IP patients who already had their polyp removed.

In this study, using aCGH platform and CD133 immunostaining, we were able to characterize different subsets of colon polyps. We found that there were more genomic changes in CSP/CPP than IP; more than half of the CSP/CPP were clonally related to the corresponding carcinomas. Moreover, genomic changes at sub-chromosomal regions and/or high CD133 content were associated with CSP/CPP and highlighted their carcinogenic potential.

Notes

Acknowledgments

We thank Dr. Li-Li Li and Felicia Hsin-Yu Jiang for language revision of the manuscript. This work was supported by research grants MOST 104-2319-B-010-001, NSC97-3112-B-075-002, NSC98-3112-B-010-023-B4, and NSC 97-2320-B-010-024-MY3 from National Science Council and Aim for the Top University Plan from the Ministry of Education, Taiwan.

Conflict of interest

The authors declare that they have no competing interests.

Supplementary material

384_2015_2319_MOESM1_ESM.docx (435 kb)
Fig. S1 (DOCX 435 kb)
384_2015_2319_MOESM2_ESM.docx (79 kb)
Fig. S2 (DOCX 78 kb)

References

  1. 1.
    Chen CJ, You SL, Lin LH, Hsu WL, Yang YW (2002) Cancer epidemiology and control in Taiwan: a brief review. Jpn J Clin Oncol 32(Suppl):S66–S81CrossRefPubMedGoogle Scholar
  2. 2.
    Kinzler KW, Vogelstein B (1996) Lessons from hereditary colorectal cancer. Cell 87(2):159–170CrossRefPubMedGoogle Scholar
  3. 3.
    Fearon ER, Vogelstein B (1990) A genetic model for colorectal tumorigenesis. Cell 61(5):759–767CrossRefPubMedGoogle Scholar
  4. 4.
    Vogelstein B, Fearon ER, Hamilton SR, Kern SE, Preisinger AC, Leppert M, Nakamura Y, White R, Smits AM, Bos JL (1988) Genetic alterations during colorectal-tumor development. N Engl J Med 319(9):525–532CrossRefPubMedGoogle Scholar
  5. 5.
    Hoglund M, Gisselsson D, Hansen GB, Sall T, Mitelman F, Nilbert M (2002) Dissecting karyotypic patterns in colorectal tumors: two distinct but overlapping pathways in the adenoma-carcinoma transition. Cancer Res 62(20):5939–5946PubMedGoogle Scholar
  6. 6.
    Molatore S, Ranzani GN (2004) Genetics of colorectal polyps. Tech Coloproctol 8(Suppl 2):s240–s242CrossRefPubMedGoogle Scholar
  7. 7.
    Postma C, Hermsen MA, Coffa J, Baak JP, Mueller JD, Mueller E, Bethke B, Schouten JP, Stolte M, Meijer GA (2005) Chromosomal instability in flat adenomas and carcinomas of the colon. J Pathol 205(4):514–521CrossRefPubMedGoogle Scholar
  8. 8.
    Sedivy R, Wolf B, Kalipciyan M, Steger GG, Karner-Hanusch J, Mader RM (2000) Genetic analysis of multiple synchronous lesions of the colon adenoma-carcinoma sequence. Br J Cancer 82(7):1276–1282. doi: 10.1054/bjoc.1999.1091 PubMedCentralCrossRefPubMedGoogle Scholar
  9. 9.
    Harris AL, Neal DE (1992) Bladder cancer—field versus clonal origin. N Engl J Med 326(11):759–761CrossRefPubMedGoogle Scholar
  10. 10.
    Slaughter DP, Southwick HW, Smejkal W (1953) Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer 6(5):963–968CrossRefPubMedGoogle Scholar
  11. 11.
    Wistuba II, Behrens C, Milchgrub S, Bryant D, Hung J, Minna JD, Gazdar AF (1999) Sequential molecular abnormalities are involved in the multistage development of squamous cell lung carcinoma. Oncogene 18(3):643–650. doi: 10.1038/sj.onc.1202349 CrossRefPubMedGoogle Scholar
  12. 12.
    Jang SJ, Chiba I, Hirai A, Hong WK, Mao L (2001) Multiple oral squamous epithelial lesions: are they genetically related? Oncogene 20(18):2235–2242. doi: 10.1038/sj.onc.1204311 CrossRefPubMedGoogle Scholar
  13. 13.
    Albertson DG, Ylstra B, Segraves R, Collins C, Dairkee SH, Kowbel D, Kuo WL, Gray JW, Pinkel D (2000) Quantitative mapping of amplicon structure by array CGH identifies CYP24 as a candidate oncogene. Nat Genet 25(2):144–146. doi: 10.1038/75985 CrossRefPubMedGoogle Scholar
  14. 14.
    Garcia MJ, Pole JC, Chin SF, Teschendorff A, Naderi A, Ozdag H, Vias M, Kranjac T, Subkhankulova T, Paish C, Ellis I, Brenton JD, Edwards PA, Caldas C (2005) A 1 Mb minimal amplicon at 8p11-12 in breast cancer identifies new candidate oncogenes. Oncogene 24(33):5235–5245. doi: 10.1038/sj.onc.1208741 CrossRefPubMedGoogle Scholar
  15. 15.
    Sonoda I, Imoto I, Inoue J, Shibata T, Shimada Y, Chin K, Imamura M, Amagasa T, Gray JW, Hirohashi S, Inazawa J (2004) Frequent silencing of low density lipoprotein receptor-related protein 1B (LRP1B) expression by genetic and epigenetic mechanisms in esophageal squamous cell carcinoma. Cancer Res 64(11):3741–3747. doi: 10.1158/0008-5472.CAN-04-0172 CrossRefPubMedGoogle Scholar
  16. 16.
    Massion PP, Kuo WL, Stokoe D, Olshen AB, Treseler PA, Chin K, Chen C, Polikoff D, Jain AN, Pinkel D, Albertson DG, Jablons DM, Gray JW (2002) Genomic copy number analysis of non-small cell lung cancer using array comparative genomic hybridization: implications of the phosphatidylinositol 3-kinase pathway. Cancer Res 62(13):3636–3640PubMedGoogle Scholar
  17. 17.
    Kuuselo R, Savinainen K, Azorsa DO, Basu GD, Karhu R, Tuzmen S, Mousses S, Kallioniemi A (2007) Intersex-like (IXL) is a cell survival regulator in pancreatic cancer with 19q13 amplification. Cancer Res 67(5):1943–1949. doi: 10.1158/0008-5472.CAN-06-3387 CrossRefPubMedGoogle Scholar
  18. 18.
    Marchio A, Meddeb M, Pineau P, Danglot G, Tiollais P, Bernheim A, Dejean A (1997) Recurrent chromosomal abnormalities in hepatocellular carcinoma detected by comparative genomic hybridization. Genes Chromosomes Cancer 18(1):59–65. doi: 10.1002/(SICI)1098-2264(199701)18:1<59::AID-GCC7>3.0.CO;2–0 CrossRefPubMedGoogle Scholar
  19. 19.
    Kusano N, Shiraishi K, Kubo K, Oga A, Okita K, Sasaki K (1999) Genetic aberrations detected by comparative genomic hybridization in hepatocellular carcinomas: their relationship to clinicopathological features. Hepatology 29(6):1858–1862. doi: 10.1002/hep.510290636 CrossRefPubMedGoogle Scholar
  20. 20.
    Shiraishi K, Kusano N, Okita S, Oga A, Okita K, Sasaki K (1999) Genetic aberrations detected by comparative genomic hybridization in biliary tract cancers. Oncology 57(1):42–49CrossRefPubMedGoogle Scholar
  21. 21.
    Hui AB, Lo KW, Leung SF, Teo P, Fung MK, To KF, Wong N, Choi PH, Lee JC, Huang DP (1999) Detection of recurrent chromosomal gains and losses in primary nasopharyngeal carcinoma by comparative genomic hybridisation. Int J Cancer 82(4):498–503. doi: 10.1002/(SICI)1097-0215(19990812)82:4<498::AID-IJC5>3.0.CO;2-S CrossRefPubMedGoogle Scholar
  22. 22.
    Wong N, Lai P, Lee SW, Fan S, Pang E, Liew CT, Sheng Z, Lau JW, Johnson PJ (1999) Assessment of genetic changes in hepatocellular carcinoma by comparative genomic hybridization analysis: relationship to disease stage, tumor size, and cirrhosis. Am J Pathol 154(1):37–43PubMedCentralCrossRefPubMedGoogle Scholar
  23. 23.
    Horst D, Kriegl L, Engel J, Jung A, Kirchner T (2009) CD133 and nuclear beta-catenin: the marker combination to detect high risk cases of low stage colorectal cancer. Eur J Cancer 45(11):2034–2040. doi: 10.1016/j.ejca.2009.04.004 CrossRefPubMedGoogle Scholar
  24. 24.
    Horst D, Kriegl L, Engel J, Kirchner T, Jung A (2008) CD133 expression is an independent prognostic marker for low survival in colorectal cancer. Br J Cancer 99(8):1285–1289. doi: 10.1038/sj.bjc.6604664 PubMedCentralCrossRefPubMedGoogle Scholar
  25. 25.
    Hermsen M, Postma C, Baak J, Weiss M, Rapallo A, Sciutto A, Roemen G, Arends JW, Williams R, Giaretti W, De Goeij A, Meijer G (2002) Colorectal adenoma to carcinoma progression follows multiple pathways of chromosomal instability. Gastroenterology 123(4):1109–1119CrossRefPubMedGoogle Scholar
  26. 26.
    Fan YS, Jayakar P, Zhu H, Barbouth D, Sacharow S, Morales A, Carver V, Benke P, Mundy P, Elsas LJ (2007) Detection of pathogenic gene copy number variations in patients with mental retardation by genomewide oligonucleotide array comparative genomic hybridization. Hum Mutat 28(11):1124–1132. doi: 10.1002/humu.20581 CrossRefPubMedGoogle Scholar
  27. 27.
    Chen YJ, Yeh SH, Chen JT, Wu CC, Hsu MT, Tsai SF, Chen PJ, Lin CH (2000) Chromosomal changes and clonality relationship between primary and recurrent hepatocellular carcinoma. Gastroenterology 119(2):431–440CrossRefPubMedGoogle Scholar
  28. 28.
    Kuukasjarvi T, Karhu R, Tanner M, Kahkonen M, Schaffer A, Nupponen N, Pennanen S, Kallioniemi A, Kallioniemi OP, Isola J (1997) Genetic heterogeneity and clonal evolution underlying development of asynchronous metastasis in human breast cancer. Cancer Res 57(8):1597–1604PubMedGoogle Scholar
  29. 29.
    Baudis M, Cleary ML (2001) Progenetix.net: an online repository for molecular cytogenetic aberration data. Bioinformatics 17(12):1228–1229CrossRefPubMedGoogle Scholar
  30. 30.
    Gaiser T, Camps J, Meinhardt S, Wangsa D, Nguyen QT, Varma S, Dittfeld C, Kunz-Schughart LA, Kemmerling R, Becker MR, Heselmeyer-Haddad K, Ried T (2011) Genome and transcriptome profiles of CD133-positive colorectal cancer cells. Am J Pathol 178(4):1478–1488. doi: 10.1016/j.ajpath.2010.12.036 PubMedCentralCrossRefPubMedGoogle Scholar
  31. 31.
    He QJ, Zeng WF, Sham JS, Xie D, Yang XW, Lin HL, Zhan WH, Lin F, Zeng SD, Nie D, Ma LF, Li CJ, Lu S, Guan XY (2003) Recurrent genetic alterations in 26 colorectal carcinomas and 21 adenomas from Chinese patients. Cancer Genet Cytogenet 144(2):112–118CrossRefPubMedGoogle Scholar
  32. 32.
    Camps J, Armengol G, del Rey J, Lozano JJ, Vauhkonen H, Prat E, Egozcue J, Sumoy L, Knuutila S, Miro R (2006) Genome-wide differences between microsatellite stable and unstable colorectal tumors. Carcinogenesis 27(3):419–428. doi: 10.1093/carcin/bgi244 CrossRefPubMedGoogle Scholar
  33. 33.
    Grady WM (2004) Genomic instability and colon cancer. Cancer Metastasis Rev 23(1–2):11–27CrossRefPubMedGoogle Scholar
  34. 34.
    Klein CA (2009) Parallel progression of primary tumours and metastases. Nat Rev Cancer 9(4):302–312. doi: 10.1038/nrc2627 CrossRefPubMedGoogle Scholar
  35. 35.
    Husemann Y, Geigl JB, Schubert F, Musiani P, Meyer M, Burghart E, Forni G, Eils R, Fehm T, Riethmuller G, Klein CA (2008) Systemic spread is an early step in breast cancer. Cancer Cell 13(1):58–68. doi: 10.1016/j.ccr.2007.12.003 CrossRefPubMedGoogle Scholar
  36. 36.
    Stoecklein NH, Hosch SB, Bezler M, Stern F, Hartmann CH, Vay C, Siegmund A, Scheunemann P, Schurr P, Knoefel WT, Verde PE, Reichelt U, Erbersdobler A, Grau R, Ullrich A, Izbicki JR, Klein CA (2008) Direct genetic analysis of single disseminated cancer cells for prediction of outcome and therapy selection in esophageal cancer. Cancer Cell 13(5):441–453. doi: 10.1016/j.ccr.2008.04.005 CrossRefPubMedGoogle Scholar
  37. 37.
    Weckermann D, Polzer B, Ragg T, Blana A, Schlimok G, Arnholdt H, Bertz S, Harzmann R, Klein CA (2009) Perioperative activation of disseminated tumor cells in bone marrow of patients with prostate cancer. J Clin Oncol 27(10):1549–1556. doi: 10.1200/JCO.2008.17.0563 CrossRefPubMedGoogle Scholar
  38. 38.
    Dakubo GD, Jakupciak JP, Birch-Machin MA, Parr RL (2007) Clinical implications and utility of field cancerization. Cancer Cell Int 7:2. doi: 10.1186/1475-2867-7-2 PubMedCentralCrossRefPubMedGoogle Scholar
  39. 39.
    Tabor MP, Brakenhoff RH, van Houten VM, Kummer JA, Snel MH, Snijders PJ, Snow GB, Leemans CR, Braakhuis BJ (2001) Persistence of genetically altered fields in head and neck cancer patients: biological and clinical implications. Clin Cancer Res Off J Am Assoc Cancer Res 7(6):1523–1532Google Scholar
  40. 40.
    Tabor MP, Brakenhoff RH, Ruijter-Schippers HJ, Van Der Wal JE, Snow GB, Leemans CR, Braakhuis BJ (2002) Multiple head and neck tumors frequently originate from a single preneoplastic lesion. Am J Pathol 161(3):1051–1060. doi: 10.1016/S0002-9440(10)64266-6 PubMedCentralCrossRefPubMedGoogle Scholar
  41. 41.
    Aivado M, Gynes M, Gorelov V, Schmidt WU, Roher HD, Goretzki PE (2000) “Field cancerization”—an additional phenomenon in development of colon tumors? K-ras codon 12 mutations in normal colonic mucosa of patients with colorectal neoplasms. Chirurg 71(10):1230–1234, discussion 1234–1235CrossRefPubMedGoogle Scholar
  42. 42.
    Leedham SJ, Graham TA, Oukrif D, McDonald SA, Rodriguez-Justo M, Harrison RF, Shepherd NA, Novelli MR, Jankowski JA, Wright NA (2009) Clonality, founder mutations, and field cancerization in human ulcerative colitis-associated neoplasia. Gastroenterology 136(2):542–550. doi: 10.1053/j.gastro.2008.10.086, e546CrossRefPubMedGoogle Scholar
  43. 43.
    Pallini R, Ricci-Vitiani L, Montano N, Mollinari C, Biffoni M, Cenci T, Pierconti F, Martini M, De Maria R, Larocca LM (2010) Expression of the stem cell marker CD133 in recurrent glioblastoma and its value for prognosis. Cancer. doi: 10.1002/cncr.25581 PubMedGoogle Scholar
  44. 44.
    Kojima M, Ishii G, Atsumi N, Nishizawa Y, Saito N, Ochiai A (2010) CD133 expression in rectal cancer after preoperative chemoradiotherapy. Cancer Sci 101(4):906–912. doi: 10.1111/j.1349-7006.2009.01478.x CrossRefPubMedGoogle Scholar
  45. 45.
    Elsaba TM, Martinez-Pomares L, Robins AR, Crook S, Seth R, Jackson D, McCart A, Silver AR, Tomlinson IP, Ilyas M (2010) The stem cell marker CD133 associates with enhanced colony formation and cell motility in colorectal cancer. PLoS One 5(5), e10714. doi: 10.1371/journal.pone.0010714 PubMedCentralCrossRefPubMedGoogle Scholar
  46. 46.
    Ishii H, Iwatsuki M, Ieta K, Ohta D, Haraguchi N, Mimori K, Mori M (2008) Cancer stem cells and chemoradiation resistance. Cancer Sci 99(10):1871–1877. doi: 10.1111/j.1349-7006.2008.00914.x CrossRefPubMedGoogle Scholar
  47. 47.
    Liu JM, Mao BY, Hong S, Liu YH, Wang XJ (2008) The postoperative brain tumour stem cell (BTSC) niche and cancer recurrence. Adv Ther 25(5):389–398. doi: 10.1007/s12325-008-0050-x CrossRefPubMedGoogle Scholar
  48. 48.
    Klonisch T, Wiechec E, Hombach-Klonisch S, Ande SR, Wesselborg S, Schulze-Osthoff K, Los M (2008) Cancer stem cell markers in common cancers—therapeutic implications. Trends Mol Med 14(10):450–460. doi: 10.1016/j.molmed.2008.08.003 CrossRefPubMedGoogle Scholar
  49. 49.
    Horst D, Scheel SK, Liebmann S, Neumann J, Maatz S, Kirchner T, Jung A (2009) The cancer stem cell marker CD133 has high prognostic impact but unknown functional relevance for the metastasis of human colon cancer. J Pathol 219(4):427–434. doi: 10.1002/path.2597 CrossRefPubMedGoogle Scholar
  50. 50.
    Takahashi S, Kamiyama T, Tomaru U, Ishizu A, Shida T, Osaka M, Sato Y, Saji Y, Ozaki M, Todo S (2010) Frequency and pattern of expression of the stem cell marker CD133 have strong prognostic effect on the surgical outcome of colorectal cancer patients. Oncol Rep 24(5):1201–1212CrossRefPubMedGoogle Scholar
  51. 51.
    Artells R, Moreno I, Diaz T, Martinez F, Gel B, Navarro A, Ibeas R, Moreno J, Monzo M (2010) Tumour CD133 mRNA expression and clinical outcome in surgically resected colorectal cancer patients. Eur J Cancer 46(3):642–649. doi: 10.1016/j.ejca.2009.11.003 CrossRefPubMedGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2015

Authors and Affiliations

  • Chih-Yung Yang
    • 1
  • Ju-Yu Tseng
    • 2
  • Chian-Feng Chen
    • 3
  • Teh-Ying Chou
    • 4
  • Hong-Wei Gao
    • 5
  • Chia-Ling Hua
    • 3
  • Chi-Hung Lin
    • 1
    • 2
    • 3
    • 6
  • Jen-Kou Lin
    • 6
    • 7
  • Jeng-Kai Jiang
    • 6
    • 7
    Email author
  1. 1.Department of Education and ResearchTaipei City HospitalTaipeiTaiwan
  2. 2.Institute of Microbiology and ImmunologyNational Yang-Ming UniversityTaipeiTaiwan
  3. 3.VYM Genome Research CenterNational Yang-Ming UniversityTaipeiTaiwan
  4. 4.Department of PathologyTaipei Veterans General HospitalTaipeiTaiwan
  5. 5.Department of PathologyTri-Service General HospitalTaipeiTaiwan
  6. 6.School of MedicineNational Yang-Ming UniversityTaipeiTaiwan
  7. 7.Division of Colon & Rectal Surgery, Department of SurgeryTaipei Veterans General HospitalTaipeiTaiwan

Personalised recommendations