~53 Protein Expression in Ulcerative
Colitis-Associated Colorectal Dysplasia
and Carcinoma
N0A.M HARPAZ, MD, PHD, ALLAN La PECK, MD, JING YIN, DDS,
ISABEL FIEL, MD, MARIA HONTANOSAS, MD,
TOMMY R. TONG, MD, JACQUELINE N. LAURIN, MD,
JOHN M, ABRAHAM, PHD, BRUCE D. GREENWALD, MD,
AND STEPHEN J. MELTZER, MD
The frequency and timing of p53 inactivationin ulcerative colitis
(UC)-associated tumorigenesiswere investigatedusing immunohis-
tochemistry((IHC)to detect p53 protein overexpressionin 56 carcino-
mas and 40 dysplastic epithelia derived from 58 patients with UC
undergoing colectomy for neoplasia. p53 DNA in 25 of the carcino-
mas also was evaluatedby single-strandconformation polymorphism
analysis(SSCP) to detect point mutationsin exons 5-8 and by loss of
heterozygosityanalysii to detect allelic deletions. Point mutations
were detected in 20 of the 25 carcinomas (80.0%) undergoing both
IHC and DNA analysis.One carcinoma contained an allelic deletion
but no mutationsof the correspondingallele withinthe region tested.
p53 overexpressionoccurred in 16 (76.2%) of the 21 carcinomaswith
point mutationsand/or allelic deletions but not in any of those with
wild type DNA. Of the 56 carcinomasevaluatedby IHC, p.53 overex-
pression occurred in 34 carcinomas(60.7%). The proportion of posi-
tive tumors was independent of stage, anatomiclocation, differentia-
tion, and histologicalsubtype. Overexpressionwas observed in nine
of 20 dysplastic masses devoid of and situated remote from carci-
noma (45.0%) and correlated positivelywith increasinggrade of dys-
plasia (P < .025). In contrast, overexpressionoccurred in 16 of 20
dysplasticepitheliasituatedadjacentto carcinoma(80.0%) and corre-
lated with overexpressionby the corresponding carcinomasbut not
with the grade of dysplasiapresent (P = .013). It is concluded that
p53 overexpression can be detected by IHC in most, although not
all, UC-associatedcarcinomaswith p53 mutationsand/or allelic dele-
tions. Based on this method, p53 overexpression occurs frequently
in UC-associated carcinomas regardless of stage and pathological
characteristics,in noncancerous dysplastic masses with high grade
dysplasia,and in dysplasiasof all grades situatedadjacentto carcino-
mas. These findings implicate p53 inactivationin the progression
from dysplasiato carcinoma in UC and suggest that its occurrence
in dysplasticepitheliummay be an independent marker of malignant
potential. HUM PATHOL 25:1069-1074. Copyright 0 1994 by W.B.
Saunders Company
I,ong-stmding ulcerative colitis (UC) is a risk factor
for the Iclevelopment of colorectal adenocarcinoma.’
From the Department of Pathology. The Mount Sinai Medical
School of the City University of New York. New York. NY; and the
Department of Medicine, GdStrOenterOlOp Division, University of
Maryland School of Medicine and Veterans Affairs Hospital of Balti-
mow. Baltimorr. MD. Accepted for publication April 4, 1994.
Supported in part by a National Institutes of Health, Bethesda,
.MD, grant no. DK47’717-01 and a grant from the Crohn’s and Colitis
Foundation of America, NY. NY.
Kq w&s: ulcerative colitis, colorectal dysplasia. ~5.7 immunohis-
tochemistry. ~53 mutation.
Address correspondence and reprint requests to Noam HarpaT,
MD, PhD, Department of Pathology, Box 1194, The Mount Sinai
Medical Crntrr, One Gustave L. Levy Place, New York, NY 10029.
C:opyIight Q 1994 by W.B. Saunders Company
0046-~~177i1)4/‘2510-0013$5.00/O
1069
Most, if not all, colitis-associated carcinomas are derived
from dysplastic colorectal epithelium. Thus dysplasia is
used clmically as an end point of endoscopic surveil-
lance programs aimed at identifying patients with UC
at imminent risk of carcinoma.”
The molecular pathogenesis of UC-associated dys-
plasia and its progression to carcinoma are not yet com-
pletely understood, but important insights have
emerged from the study of sporadic colorectal carci-
noma. The progression from adenomas to carcinomas
involves the accrual of successive mutations in specific
genes,” including the ‘rasproto-oncogenes”.’ and several
tumor suppressor genes, including adenomatous polyp-
osis coli (APC) ,6 mutated in colon cancer (MCC),’ de-
leted in colon cancer (DCC) ,8 and ~53.’ p53 mutations
in sporadic colon cancer are found mainly in severely
dysplastic adenomas and invasive carcinomas and thus
seem to be late events heralding the development of
invasive cancer?
How closely the molecular events in UC-associ-
ated tumorigenesis parallel those in sporadic colo-
rectal cancer is unclear. Point mutations of rus proto-
oncogenes are rarer in dysplastic lesions and cancers
of UC than in sporadic adenomas and cancers.4,fl,‘“-‘2
Allelic deletions and/or point mutations of the APC,
MCC, DCC, and p53 tumor suppressor genes occur
frequently in both sporadic and colitis-associated can-
cers’“.‘“; moreover, an unexplained tendency for al-
lelic losses to occur more frequently in cancers of
the left colon and rectum than in right-sided cancers
seems to be common to cancers of both types.“‘.‘”
Less is known regarding genetic alterations in UC-
associated dysplasia. Allelic deletions and point muta-
tions of the p53 gene in UC have been implicated in
relatively early stages in the dysplasia to carcinoma
sequence 14,” in contrast with their late timing in spo-
radic colorectal carcinoma.
Mutationally inactivated forms of the p53 protein
product are frequently hyperstable. Detection of p53
overexpression by immunohistochemistry (IHC) may
afford a convenient marker of ~53 inactivation in rou-
tinely processed histological specimensI Using this
technique 55% to 61% of sporadic colorectal carci-
nomas have shown p53 overexpression.“,” Recently,
Taylor et al”” observed p53 overexpression in 11 of 21
UC-associated carcinomas, in a similar proportion of
sporadic colonic carcinomas, and in six of 20 specimens
with dysplasia.
 HUMAN PATHOLOGY Volume 25, No. 10 (October
In the present study we compared the prevalence
of altered ~53 in a series of UC-associated carcinomas
as detected by IHC, single-strand conformation poly-
morphism analysis (SSCP), and loss of heterozygosity
(LOH). We then used IHC to investigate the prevalence
and timing of p53 inactivation in the progression of
dysplasia to carcinoma.
MATERIALS AND METHODS
Tumor samples were obtained from the colectomy speci-
mens of 58 patients with UC who had undergone surgery for
colorectal adenocarcinoma, high grade dysplasia, or a dyspla-
sia-associated lesion or mass. Normal control tissue was ob-
tained from uninvolved lymph nodes, connective tissue, or
ileal mucosa. Tissues were fixed in 10% neutral-buffered for-
malin and were embedded in paraffin.
Loss of heterozygosity’s and SSCP analysisI (exons 5-8)
were performed on epithelial tissues microdissected from 10.
pm par&n sections as previously described.‘” Loss of hetero-
zygosity was detected using polymerase chain reaction-based
methods.” Once a mutation was found in any exon by SSCP,
analysis of the remaining exons was not performed. Because
of limited amounts of tissue, complete analyses of all exons
were not performed in all patients; however, only lesions in
which all four exons were analyzed and found to be negative
are reported as SSCP negative. Of the 56 adenocarcinomas
reported here, the SSCP and LOH results of 13 have been
reported previously.15
Immunohistochemistry was performed on serial 4pm
paraffin sections of the same tissue blocks used for SSCP and
LOH. The primary antibody, murine monoclonal antibody
BP5312-1 (Biogenex Laboratories, San Ramon, CA), was used
at the dilution supplied (1:80). It recognizes a fixation-resis-
tant epitope on the N-terminal domain of both mutant and
wild type p53 protein. Parallel controls were performed with
nonimmune mouse serum in phosphate-buffered saline. The
sections were prepared for IHC by fixation to glass slides at
37”C, deparafKnization in Microclear (Micron Diagnostics,
Fairfax, VA) or fresh xylene, and pretreatment with 3% hydro-
gen peroxide in methanol for 10 minutes at room tempera-
ture. A known immunopositive control section of adenocarci-
noma was included. Incubation with the primary antibody
and control serum was performed for 30 minutes at room
temperature and followed by sequential incubations at room
temperature with Super Sensitive Multilink reagents (Bio-
genex Laboratories) supplied in kit form, including second-
ary antibody (20 minutes), horseradish peroxidase-conju-
gated streptavidin (20 minutes), and 3,3’-diaminobenzidine
with enhancer (10 to 15 minutes). Counterstaining was per-
formed with methyl green or hematoxylin. To assure repro-
ducibility at least two serial sections of each tumor were evalu-
ated by IHC in separate runs.
Staining intensity was graded semiquantitatively as 0
(negative) to 3+ (intensely positive) and the percentage of
positive cells was assessed subjectively. A positive result was
recorded when more than 10% of epithelial cells were stained.
Grading of dysplasias was performed in an unblinded
fashion by one pathologist (N.H.) according to the criteria
of the standardized classification system for dysplasia in in-
flammatory bowel disease.? Dysplastic masses were defined as
sessile, exophytic neoplasms composed of benign but dysplas-
tic epithelium of at least 1.0 cm diameter, occurring within
colorectal mucosa affected by UC, and separated from carci-
noma by at least 3 cm. Although the recommended two-tier
classification of low and high grade dysplasia was used, a cate-
gory of “mixed dysplasia’* was established to accommodate
1070
1994)
masses containing substantial foci of both low and high grade
dysplasia. Classification of dysplastic masses based on their
dimensions was precluded by the indistinct borders and irreg-
ular contours that frequently characterize such lesions. Dys-
plasia was considered to be adjacent to carcinoma when pres-
ent in the same histological sections as, but not directly
mdying. a carcinoma. Staging of carcinomas was based on
the Astler-Caller modification of the Dukes system. Mutinous
carcinomas were defined as tumors in which mucin pools
occupied 50% or more of the tumor area as determined from
the available histological sections.
Differences between populations based on positive or
negative IHC results were analyzed by the x’ test or, for data
with low expected values, Fisher’s two-sided exact test. Differ-
ences between ordered populations were analyzed by the $
test for trend or, for variables with three or four ranklngs. b!
Bartholomew’s test.”
RESULTS
Immunohistochemistry was performed on 76 colo-
rectal tumors from 58 UC colectomy specimens, com-
posed of 55 primary adenocarcinomas (20 of which
were accompanied by adjacent dysplastic mucosa), one
metastatic adenocarcinoma, and 20 benign dysplastic
masses. The staining was nuclear in all cases (Fig 1)
except for one carcinoma that stained cytoplasmically.
Twenty-five of the carcinomas were selected at ran-
dom and were evaluated for p53 mutations by molecu-
lar assays. As shown in Table 1, nonsilent p53 point
mutations were detected in 20 carcinomas (80.0%) b)
SSCP and allelic deletions were detected in I2 carcino-
mas (48.0%) by LOH; seven tumors were from uninfor-
mative patients and could not be evaluated by LOH.
One carcinoma was LOH positive despite being SSCP
negative, suggesting the presence of a mutation outside
the regions analyzed by SSCP. Thus, the combined re-
sults of these assays yielded 21 of 25 carcinomas (84.0%)
containing mutations and/or deletions. Of these 2 1 tu-
mors, 16 were IHC positive (76.2%), corresponding to
15 of the 20 SSCP-positive tumors and to the single
LOH-positive/SSCP-negative tumor. No IHC-positive
tumors were identified among the four carcinomas that
were negative by the molecular assays.
Of the 56 carcinomas evaluated by IHC, 34
(60.7%) were IHC positive. As shown in Table 2. the
prevalence of IHC positivity was independent of the
tumors’ pathological characteristics, including Dukes’
stage, anatomic location, differentiation, and mutinous
or signet ring cell histological subtype.
Of the 20 dysplastic epithelia situated adjacent to
carcinomas, eight lesions were exophytic and 12 were
flat. The immunohistochemical staining of the dysplas-
tic epithelia was more intense than that of the corre-
sponding carcinomas in 10 cases (50.0%) and of similar
intensity in the other IO (50.0%) (Fig 1). There was a
tendency for the staining intensity to diminish gradually
from the crypt bases to the surface epithelium, but at
least a few positively stained nuclei could be identified
in the superficial epithelial layer.
The IHC results for these dysplastic epithelia were
concordant with the IHC results for the corresponding
adjoining carcinomas, regardless of the grade of dyspla-
sia present (Table 3). Thus. the dysplastic epithelia aclja-
 ~53 EXPRESSION IN UC NEOPLASIA (Harpaz et al)

FIGURE 1. p53 immunostaining of UC-associated neoplastic lesions. (A) Invasive, well-differentiated adenocarcinoma. (Original
magnification x160.) (B) Signet ring cell adenocarcinoma. (Original magnification x160.) (C) Low grade dysplasia. (Original
magnification x100.) (D) Villiform high grade dysplasia. (Original magnification x80.) (All of the figures show the streptavidin-
biotin peroxidase method with hematoxylin counterstain.)

1071
 HUMAN PATHOLOGY Volume 25, No. 10 (October 1994)

TABLE 1. Carcinomas Evaluated by IHC, SSCP, and TABLE 2. lmmunohistochemistry Results and
LOH Analyses Pathological Characteristics of
Colorectal Carcinomas
DNA
NO. IHC LOH SSCP Ex0n IHC Positive
Pathological Characteristic N (%)
HlTl-2 + + + 5
HZT + + + 7 Stage
H3Tl-2 + + + 7 A 2 0 (0)
H7T + + + 6 Bl 19 11 (57.9)
H8T2 _ - _ B2 5 3 (60.0)
HlOTl + + + 5 Cl 2 2 (100)
HlOT2 + + + .5 C? 23 15 (65.2)
HllTl - - _ D 4 2 (50.0)
104T - - + 7 Unknown 1 1 (100)
115T + + + 8 Location
12lT + ho + 6 Right colon 26 18 (69.2)
126T _ ho sil 7 Left colon 30 16 (53.3)
653T + + + 6 Differentiation
656T _ + + 5 Well 21 13 (61.9)
665/8T + + + 8 Well/moderate 3 1 (33.3)
6891 _ + + 7 Moderate 15 10 (66.7)
691T + ho + 8 Moderate/poor 3 2 (66.7)
727T + + _ Poor 14 8 (57.1)
733T _ ho + 8 Histological subtype
775T + ho + 8 Mutinous 14 7 (50)
81 IT + _ + 8 Signet ring 9 7 (77.8)
815T - ho + 8 Total 56 34 (60.7)
817T + ho + 8
862T + _ + 8 NOTE. P > .2 for all differences among populations.
880-3 _ - _ Abbreviation: IHC, immunohistochemistry.

NOTE. AI1 tumors contained missense mutations. Mutations in

13 cases were reported in a previous study.”
Abbreviations: IHC, immunohistochemistry, SSCP. single-strand suggested that p53 inactivation may participate in the
conformation polymorphism analysis: LOH, loss of heterozygosity: progression from dysplasia to carcinoma in UC. To fur-
ho, homozygous (uninformative for LOH); sil. silent mutation. ther investigate the frequency and timing of p53 inacti-
vation in UC, we used IHC to detect the overexpression
of p53 protein in histological sections of carcinomas
cent to 15 of the 16 IHC-positive carcinomas (93.8%) and dysplastic epithelia.
were IHC positive, including nine of 10 cases of low Immunohistochemical overexpression is an indi-
grade dysplasia (90.0%), whereas those adjacent to rect manifestation of p53 inactivation. Thus, we com-
three of the four IHC-negative carcinomas (75.0%) pared our IHC results with those of two molecular DNA
were IHC-positive (P = .013). assays-SSSCP, which detects point mutations, and
Of the 20 isolated dysplastic masses evaluated, nine LOH, which detects allelic deletions-performed on
(45.0%) were IHC positive (Table 4). In contrast with epithelia microdissected from the same tissue blocks.
dysplastic mucosae situated adjacent to cancer a sig- Based on 25 carcinomas examined it was determined
nificant positive correlation was noted between IHC that positive immunostaining correlated with the pres-
positivity and high grade dysplasia (P < .025). The dis- ence of p53 point mutations and/or allelic deletions in
parity between dysplasias remote from and adjacent to 16 of 21 tumors (76.2%)) whereas no staining occurred
carcinoma was greatest among low grade dysplasias, the in any of the neoplasms in which they were absent.
prevalence of IHC positivity in each group being 20.0% The failure of IHC to detect some tumors containing
and 75.0%, respectively (P = .03). However, among mutations by SSCP has been noted previously,” and
high grade dysplasias, equally high proportions were
IHC positive, regardless of any association with carci-
noma. TABLE 3. lmmunohistochemistry of Carcinomas and
With the exception of occasional foci of faint (I+) Adjacent Dysplastic Mucosa
nuclear staining within the basal portions of benign Dysplasia
colonic crypts, we did not observe any IHC positivity
in mucosae that were unequivocally free of dysplasia. IHC Positive IHC Negative
However, there were a few instances of intense staining
of mucosa that were considered indefinite for dysplasia. Low grade 9 1.0~ grade I
IHC Positive 15 Mixed 2 1 Mixed 0
High grade 4 High grade 0
Carcinoma
DISCUSSION Low grade 0 Low grade 2
IHC Negative 1 Mixed 0 3 Mixed 0
High grade 1 High grade 1
Recent results demonstrating mutations’” and al-
lelic deletions”‘.‘” of the p53 tumor suppressor gene in NOTE. P = .013.
UC-associated dysplastic epithelia and carcinomas have Abbreviation: IHC, irnmunohistochwnistq.

1072
 p53 EXPRESSIONIN UC NEOPLASIA (Harpazet al)
TABLE 4. lmmunohistochemistry of Dysplastic
Mucosa Adjacent To and Remote From Carcinoma
Remote From
Carcinoma
Adjacent to Carcinoma
IH(: Positive/Total (%) IHC Positive/Total
Dvsplasia N = 20 N = 20
1,ow grade 9,‘12 (75)* 2/10
Mixed 2,‘2 (100) 2/4
High grade .!,I6 (83) <i/6
Total 16,/20 (80)tt 9/20
Abbreviation: IHC. immunohistochemistry.
* Significant difference between low grade dysplasias adjacent to
and remote from carcinoma; P = .03.
t Significant trend toward IHC+ with increasing grade of dyspla-
sia: P < ,025 (Bartholomew’s test”).
tt Significant difference between total dysplasias adjacent to and
remote from carcinoma, P = 0.05.
indicates that some mutations do not result in detect-
able overexpression of the p53 protein. It should be
noted that although we did not observe p53 expression
in the absence of mutation and/or deletion in our
cases, therse have been several reports of immunohisto-
chemical positivity in noncolorectal tissues without p53
mutations.“‘-‘” Nevertheless, the high prevalence of al-
tered p53 in our cases based on the DNA assays (84.0%)
supports t:he notion that p53 inactivation plays an im-
portant role in UC-associated tumorigenesis.
Immumostaining has the capacity to reveal small
populations of cells that overexpress ~53. Although not
as exquisiltely sensitive, SSCP easily detects mutations
present in less than 10% of total DNA.27 Thus because
we used 10% positive cells as our IHC cutoff, the evalua-
tions by these two methods were comparable in sensitiv-
ity. On the other hand, LOH assays require tumor cell
purities in excess of 50%“,‘” and thus are not as sensitive
as IHC or SSCP. Nevertheless, in this relatively small
series good agreement was observed.
Based on 56 UC-associated carcinomas examined
bv IHC, 1~53 overexpression was observed frequently
(34 cases) and independently of such pathological char-
acteristics as Dukes’ stage, anatomic location, differenti-
ation, and histological subtype. Although it had been
reported that p53 immunostaining is infrequent among
sporadic mutinous colorectal carcinomas,‘R no such
negative correlation was evident in our series.
In addition to cancers p53 protein also was fre-
quently overexpressed in both groups of dysplastic epi-
thelia eva.luated, those adjacent to and those remote
from carcinomas. Thus, p53 inactivation occurs during
the intraepithelial phase of tumor development that
precedes invasive malignancy. Nevertheless, the preva-
lence of overexpression in the adjacent group was
higher than the remote group (80.0% ZI45.0%, respec-
tively). In addition, overexpression in the adjacent
group correlated closely with overexpression in the cor-
responding carcinomas but not with the grade of dys-
plasia present. In other words, dysplastic epithelia situ-
ated adjacent to IHC-positive carcinomas tended to be
IHC negative regardless of their grade. In contrast,
overexpression in the remote group showed a positive
correlation with grade. This dichotomy was greatest
among low grade dyplasias, few of which were positive
in the remote group (20.0%) but most of which were
positive in the adjacent group (‘75.0%).
The implications of these observations are twofold.
(770) First, the concordance in p53 expression between carci-
nomas and the dysplastic epithelia adjoining them is
consistent with the existence of clonal relationships be-
(20)
tween UC-associated carcinomas and the dysplastic sur-
(50)
(83)t roundings in which they arise. Direct evidence for such
(45) relationships has recently emerged from flow cytomet-
ric studies of aneuploid populations within dysplastic
and cancerous mucosae.‘” Second, the disproportion-
ately high prevalence of p53 overexpression in dyspla-
sias that had progressed to carcinoma and the indepen-
dence of p53 overexpression of the grade of dysplasia
suggest that p53 inactivation in dysplastic epithelia in
UC may be an independent marker of malignant poten-
tial.
A conceptual basis for this hypothesis is afforded
by recognition that one of the key functions of the
p53 protein product is to prevent the emergence of
transformed clones from populations of replicating
cells.2Y When functional, the p53 protein arrests the
entry of cells sustaining DNA damage into S phase,
thereby providing an opportunity for DNA repair to
proceed through normal repair mechanisms, whereas
it triggers apoptosis in cells in which DNA damage is
irreversible. Loss of these protective mechanisms
through mutation or other means might thereby predis-
pose abnormal epithelia to malignancy, especially un-
der conditions of persistent epithelial injury and turn-
over, such as those prevailing in UC.“”
Assessment of the value of p53 determination as
an adjunct to conventional histopathological evaluation
in the management of high-risk UC patients awaits fur-
ther investigation, including longitudinal studies of pa-
tients enrolled in endoscopic surveillance programs.
Such studies might ascertain, for example, whether p53
status influences the prognostic significance of biopsy
specimens with low grade dysplasia, a possibility implied
by our findings.
Despite the fact that some p53 point mutations
escape detection by IHC, this technique has advantages
over molecular DNA assays that make it more suitable
for clinical use, including technical accessibility, inex-
pensiveness, accommodation of small tissue samples,
and capability of detecting alterations in small clones.
The efficacy of IHC may perhaps be further enhanced
by the development of other antibodies or antibody
combinations directed against different p53 epitopes.“’
Nevertheless, the potential for p53 overexpression to
occur despite the absence of p53 mutations, as recently
reported in certain noncolorectal tissues. mandates cau-
tion in interpreting results based solely on this tech-
nique.
REFERENCES
1. Lashner BA. Silverstein MD, Hanaurr SB: Hazard rates for
dysplasia and cancer in ulcerative colitis. Results from a surveillance
program. Dig Dis Sci 34:153&1541. 1989
2. Riddell RH, Goldman H, Ransohoff DF. et al: Dysplasia in
1073
 HUMAN PATHOLOGY
inflammatory bowel disease: Standardized classification with provi-
sional clinical applications. HUM PATHOL 14:931-966, 1983
3. Vogelstein B, Fearon ER, Hamilton SR, et al: Genetic alter-
ations during colorectal-tumor development. N Engl J Med 319:525-
532, 1988
4. Bos JL, Fearon ER, Hamilton SR, et al: Prevalence of ras gene
mutations in human colorectal cancers. Nature 327:293297, 1987
5. Forrester K, Almoguera C, Han K, et al: Detection of high
incidence of K-ras oncogenes during human colon tumorigenesis.
Nature 327:298-303, 1987
6. Nishisho I, Nakamura Y. Miyoshi Y, et al: Mutations of chro-
mosome 5q21 genes in FAP and colorectal cancer patients. Science
253:665-669, 1991
7. Kinzler KW, Nilbert MC, Vogelstein B, et al: Identification of
a gene located at chromosome 5q21 that is mutated in colorectal
cancers. Science 251:13661370, 1991
8. Fearon ER, Cho KR, Nigro JM, et al: identification of a chro-
mosome 18q gene that is altered in colorectal cancers. Science 247:49-
56, 1990
9. Baker SJ, Fearon ER, Nigro JM, et al: Chromosome 17 dele-
tions and p53 gene mutations in colorectal carcinomas. Science
244:217-221, 1989
10. Meltzer SJ, Mane SM, Wood PK, et al: Activation of c-Ki-ras
in human gastrointestinal dysplasias determined by direct sequencing
of polymerase chain reaction products. Cancer Res 50:3627-3630,
1990
11. Burmer GC, Levine DS, Kulander BC, et al: c-Ki-ras muta-
tions in chronic ulcerative colitis and sporadic colon carcinoma. Gas-
troenterology 99:416-4420, 1990
12. Chen J, Compton C, Cheng E, et al: c-Ki-ras mutations in
dysplastic fields and cancers in ulcerative colitis. Gastroenterology
102:19831987, 1992
13. Greenwald BD, Harpaz N, Xn J, et al: Loss of heterozygosity
affecting the ~53, Rb, and mcc/apc tumor suppressor gene loci in
dysplastic and cancerous ulcerative colitis. Cancer Res 52:741-745,
1992
14. Burmer GC, Rabinovitch PS, Haggitt RC. et al: Neoplastic
progression in ulcerative colitis: Histology, DNA content, and loss of
a p53 allele. Gastroenterology 103:1602-1610, 1992
15. Yin J, Harpaz N, Tong Y, et al: p53 Point mutations in dysplas-
tic and cancerous ulcerative colitis lesions. Gastroenterology
104:16331639, 1993
16. Delattre 0, Olschwang S, Law DJ, et al: Multiple genetic
alterations in distal and proximal colorectal cancer. Lancet 2:353-
356, 1989
Volume 25, No. 10 (October 1994)
17. BQrtek J. Bartkov& J, Vojtesek B, et al: Aberrant expression
of the p53 oncoprotein is a common feature of a wide spectrum of
human malignancies. Oncogene 6:1699-1703, 1991
18. van den Berg FM, Tigges AJ, Schipper MEI, et al: Expression
of the nuclear oncogene p53 in colon tumours. J Path01 157:193199,
1989
19. Yamaguchi A, Kurosaka Y, Fushida S, et al: Expression of
p53 protein in colorectal cancer and its relationship to short-term
prognosis. Cancer 70:2778-2784, 1992
20. Taylor HW, Boyle M, Smith SC, et al: Expression of p53 in
colorectal cancer and dysplasia complicating ulcerative colitis. Br J
Surg 80:442-444, 1993
21. Fleiss JL: Comparing proportions from independent sam-
ples. Statistical Methods for Rates and Proportions (ed 2). New York,
NY, John Wiley, 1981. pp 147-149
22. Burns JE, Baird MC, Clark LJ, et al: Gene mutations and
increased levels of p53 protein in human squamous cell carcinomas
and their cell lines. Br J Cancer 67:12741284, 1993
23. Rubio MP, von Dieimling A, Yandell DW, et al: Accumulation
of wild type p53 protein in human astrocytomas. Cancer Res 53:3465-
3467, 1993
24. Helander SD, Peters MS, Pittelkow MR: Expression of p53
protein in benign and malignant epidermal pathologic conditions. J
Am Acad Dermatol 29:741-748, 1993
25. Scarpa A,Capelli P, Mukai K, et al: Pancreatic adenocarcino-
mas frequently show p53 gene mutations. Am J Pathol142:15341543,
1993
26. Thompson AM, Anderson TJ, Condie A, et al: p53 Allele
losses, mutations and expression in breast cancer and their relation-
ship to clinico-pathological parameters. IntJ Cancer 50:528-532,1992
27. Suzuki Y, Orita M, Shiraishi M, et al: Detection of ras gene
mutations in human lung cancers by single-strand conformation poly-
morphism analysis of polymerase chain reaction products. Oncogene
5:1037-1043, 1990
28. Hanski C, Bornhoeft G, Shimoda T, et al: Expression of p53
protein in invasive colorectal carcinomas of different histologic types.
Cancer 70:2772-2777, 1992
29. Lane DP: ~53, Guardian of the genome. Nature 358:15-16.
1992
30. Serafini EP, Kirk AP, Chambers TJ: Rate and pattern of epi-
thelial cell proliferation in ulcerative colitis. Gut 22:648-652, 1981
31. Bartek J, Bartkova J, Lukas J, et al: lmmunohistochemical
analysis of the p53 oncoprotein on paraffin sections using a series of
novel monoclonal antibodies. J Pathol 169:27-34, 1993
1074