Professional Documents
Culture Documents
BACKGROUND. p53 is the most commonly mutated gene in cancer, including soft
tissue sarcoma (STS). The authors characterized p53 alterations (protein accumulation and gene mutation) in STS to evaluate possible associations with patient
outcomes.
have not been previously described. Four p53 exon 4 mutations were identified, a
mens expressed p53 when the authors used the clinical IHC assay of their institu-
p53 region previously unknown to be mutation prone. Eighteen of the 31 specition. Interassay concordance of 48% was observed; only 6 of 10 sequencingidentified p53 mutated specimens exhibited nuclear p53 protein expression by
bored sequencing-identified wild-type p53. Decreased survival was observed in
STS patients bearing sequencing-identified mutated p53 versus wild-type p53, as
was a correlation between IHC-determined nuclear p53 protein expression and
decreased survival.
DOI 10.1002/cncr.22680
Published online 11 April 2007 in Wiley InterScience (www.interscience.wiley.com).
2324
CANCER
2325
TABLE 1
Primer pairs and PCR Conditions for p53 Sequencing
Exon
50 Primer
30 Primer
23
4
56
7
89
10
11
ATCCCCACTTTTCCTCTTGC
CGTTCTGGTAAGGACAAGGG
TCTTTGCTGCCGTCTTCC
TGCTTGCCACAGGTCTCC
GGACAGGTAGGACCTGATTTCC
TTACTTCTCCCCCTCCTCTGTTG
TCATCTCTCCTCCCTGCTTCTGTC
GAAAAGAGCAGTCAGAGGAC
ACAAAAGAAATGCAGGGGG
AGGGCCACTGACAACCAC
GTCAGAGGCAAGCAGAGGC
AAACAGTCAAGAAGAAAACGGC
GCTTTCCAACCTAGGAAGGCAG
TGCTTCTGACGCACACCTATTG
360
412
517
235
441
193
224
55
55
55
55
55
55
55
were analyzed using Chromas-Pro software (Technelysium; Tewantin, Queensland, Australia) and verified
against the sequence deposited in GenBank (USA
National Center for Biotechnology Information
[NCBI]).
Immunohistochemistry
Immunohistochemistry was performed as previously
described.10 Briefly, 5 lm formalin-fixed, paraffinimbedded, tissue sections were dewaxed and rehydrated through step-wise immersion in xylene and
standard ethanol dilutions to water. Sections were
microwaved in a 10 mM sodium citrate solution with
a pH of 6.0 for antigen retrieval. Endogenous peroxidase activity was quenched with 1% hydrogen peroxide in Tris-buffered saline (TBS)/calcium solution at
room temperature for 30 minutes before blocking
with 5% horse serum in 1% bovine serum albumin
(BSA), or Dako protein block, in phosphate-buffered
saline (PBS) overnight at 48C. Anti-p53: DO1 monoclonal (Santa Cruz Biotechnology Inc., Calif) was
applied in 1% BSA-PBS according to manufacturers
instructions and detected by using Dako Envision kit
(DakoCytomation, Carpinteria, Calif) with diamino
benzidine (DAB) development before counterstaining
with hematoxylin and performing step-wise dehydration. Simultaneously, the specimens were subjected
to p53 immunohistochemistry in the Pathology Clinical Laboratory of the University of Texas M. D.
Anderson Cancer Center according to protocol by
using the clone DO7 anti-p53 antibody (DakoCytomation, Carpinteria, Calif).
The DO1 monoclonal antibody maps to amino
acid residues 2025 in the N-terminus, whereas the
DO7 monoclonal antibody maps to an epitope
between amino acids 19 and 26, and, thus, these
2 antibodies have overlapping specificity. Antigen
retrieval is important for reliable results with DO7.
p53 immunohistochemistry expression was evaluated by 2 independent reviewers. The samples were
2326
CANCER
STATISTICAL ANALYSIS
Statistical analysis was performed by using the Fisher
exact test; significance was set as P < .05. Cumulative
incidence was estimated by using the method of
Gooley et al26 with death as competing risk. The estimates of cumulative incidences by various risk factors were calculated by using a publicly available SAS
(SAS Institute; Cary, NC) macro created by Erik Bergstralh of the Mayo Clinic (Rochester, Minn). The correlation between p53 expression (as determined by
immunohistochemistry) and p53 mutational status
(determined by sequencing) was analyzed for sensitivity (probability of positive p53 immunohistochemistry staining when a p53 mutation was identified
by sequencing), specificity (probability of negative
p53 immunohistochemistry staining when p53 was
found to be wild-type by sequencing), and concordance (probability of exact agreement between the
2 methods).
RESULTS
Thirty-one stage III soft tissue sarcoma specimens
were included in the study (Table 2). Specimens
were collected from 20 men and 11 women with a
median age of 56 years. Median follow-up time was
78 months, during which 23 (74%) patients died of
their disease because of metastasis, unresectable
local recurrence, or both.
TABLE 2
Patient and Tumor Characteristics
Characteristics
Sex
Males
Females
Median age, y (range)
Men
Women
Histological subtype
Synovial sarcoma(SS)
Liposarcoma(LS)
Malignant fibrous histiocytoma (MFH)
Rhabdomyosarcoma(RMS)
Leiomyosarcoma (LMS)
Fibrosarcoma (FS)
Tumor site
Upper extremity
Lower extremity
Head & neck
Visceral
Retroperitoneal
Pelvic
Superficial trunk
Tumor status
Primary (P)
Recurrent (R)
No.
20
11
56 (1876)
56 (2485)
7
3
11
4
5
1
7
10
1
3
4
5
1
15
16
tional sequencing of corresponding autologous normal tissues for these 10 identified p53-mutated soft
tissue sarcoma samples verified that these were true
mutations rather than polymorphisms.
Although the incidence of identified p53 mutations was concordant with previous reports, only 3 of
the mutations were in the classic exon 5 to 8 p53
core-binding domain. Instead, the mutations were
spread throughout the p53 gene, and it is of special
interest that 40% of the mutations were found in
residues 6396, corresponding to exon 4 of the p53
gene. Although p53 is, perhaps, the most extensively
investigated tumor-suppressor gene, the structure
and function of p53 exon 4 still remains obscure. To
the best of our knowledge, this is the first report to
identify p53 exon 4 mutations in soft tissue sarcoma.
Seven (70%) of these mutations were missense synonymous mutations, whereas 3 mutations were
either insertions or frameshift mutations resulting in
early truncation proteins. A search of literature,
established p53 databases,22 and the Universal Mutation Database23 confirmed that the point mutation at
codon 250 we observed in 2 specimens has been previously observed in soft tissue sarcoma.29 The point
mutation in codon 96, although not previously
reported in soft tissue sarcoma, has been previously
observed in bladder cancer.30 In contrast, the other 7
p53 mutations are apparently novel and have not
2327
TABLE 3
Mutation Type and Location in STS Samples. IHC Status Corresponding to Each Mutation Is Shown in Comparison
Pat
ID
Tumor histological
subtype
Tumor
status
Exon
Codon
Nucleotide
Change
Amino acid
substitution
IHC
561
121
344
390
474
583
537
500
244
508
SS
SS
SS
SS
SS
FS
LS
MFH
MFH
RMS
P
P
P
R
P
P
R
R
P
R
2
7
7
4
4
4
11
4
6
11
8
250
250
96
63
75
377
77
213
393
144
868
868
407
307
344
1249
350
759
1297
C in at 144*
CCC >ACCy
CCC>ACCy
TCT>TGT{
GCT>CCT*
CCT>CAT*
T insertion at nt.1249*
CCA>CAA*
CGA>CGG(Insertion)*
GAC>AAC*
Negative
Negative
Weak
Negative
Moderate
Weak
Weak
Negative
Strong
Moderate
Pat ID indicates patient identification code; IHC, immunohistochemistry; SS, synovial sarcoma; FS, fibrosarcoma; LS, liposarcoma; MFH, malignant fibrous histiocytoma; RMS, rhabdomyosarcoma; P, primary;
R, recurrent.
* Novel mutations not reported in p53 mutation databases.
y
Reference 30.
{ Reference 31.
FIGURE 1. (A) Chromatographs demonstrating p53 mutations were identified by direct sequencing in 3 selected soft tissue sarcoma samples. (B) p53 mutation spectra in soft tissue sarcoma. The entire transcribed p53 gene is depicted (exons 211). Thin arrows above represent the mutations found in our study
compared with common hotspot p53 mutations represented by arrows below. The functional protein domains of transactivation, DNA binding, and tetramerization are depicted in blue, red, and green, respectively. Codon boundary numbers for the exons are directly below the figure.
2328
CANCER
TABLE 4
Tumor Characteristics and Patient Outcome Stratification According
to p53 Mutational Status
p53 status by sequencing
Histology
SS (n7)
MFH (n11)
RMS (n4)
LS (n3)
LMS (n5)
FS (n1)
Site
Extremity (n17)
Nonextremity (n14)
Tumor status
Primary (n15)
Recurrent (n16)
Survival status
Dead
Alive
Median survival, mo
wt (n = 21)
Mut (n = 10)
2
9
3
2
4
0
5
2
1
1
1
1
11
10
6
4
9
12
6
4
15
6
52
8
2
33
wt indicates wild type; Mut, mutated; SS, synovial sarcoma; MFH, malignant fibrous histiocytoma;
RMS, rhabdomyosarcoma; LS, liposarcoma; LMS, leiomyosarcoma; FS, fibrosarcoma; mo, months.
2329
FIGURE 2. p53 immunohistochemistry staining score. Stained sections representing the 4 distinct staining patterns identified, (A) negative, (B) weak, (C) moderate, (D) strong. (E) A biphasic synovial sarcoma demonstrates high expression of p53 in the epithelioid components of the tumor (magnification, 3200).
2330
CANCER
TABLE 5
Tumor Characteristics and Patient Outcome Stratification According
to p53 IHC Staining Score
IHC p53 status
Histology
SS (n7)
MFH (n11)
RMS (n4)
LS (n3)
LMS (n5)
FS (n1)
Site
Extremity (n17)
Nonextremity (n14)
Tumor status
Primary (n15)
Recurrent (n16)
Survival status
Dead
Alive
Median survival, mo
Negative
(n = 13)
Weak
(n = 7)
Moderate
(n = 6)
Strong
(n = 5)
3
7
1
1
1
0
1
0
2
1
2
1
3
0
1
1
1
0
0
4
0
0
1
0
7
6
4
3
3
3
3
2
7
6
2
5
3
3
3
2
5
8
84
7
0
38
6
0
26
5
0
22*
IHC indicates immunohistochemistry; SS, synovial sarcoma; MFH, malignant fibrous histiocytoma;
RMS, rhabdomyosarcoma; LS, liposarcoma; LMS, leiomyosarcoma; FS, fibrosarcoma; mo, months.
* P < .05
FIGURE 3. Increased p53 protein expression in soft tissue sarcoma inversely correlates with time to local recurrence (A) and time to metastasis (B)
as demonstrated by cumulative incidence curves.
TABLE 6
IHC Staining Score Correlation With p53 Mutational Status According
to Direct Sequencing
IHC*
Negative
(n = 13)
Weak
(n = 7)
Moderate
(n = 6)
Strong
(n = 5)
Sequencing
wtp53 (n 21)
Mutp53 (n 10)
9
4
4
3
4
2
4
1
DISCUSSION
A search of PubMed (USA National Library of Medicine searchable journal database) using the keyword
p53 revealed more than 40,000 publications referring to this gene and its protein product that have
been published since its first description just 3 decades ago, rendering p53 the most studied molecule
in the context of cancer, yet it is still not fully understood. More than 10,000 p53 mutations have been
IHC indicates immunohistochemistry; wtp53, wild type p53; Mutp53, mutated p53; n, number.
* IHC sensitivity 60%; IHC specificity 42%; IHC to sequencing concordance rate 48% (when
any degree of staining is considered positive).
2331
2332
CANCER
signifies increased environmental stress signals leading to stabilization of wtp53. However, in that situation, we would probably expect to observe increased
tumor apoptosis and a relatively less aggressive clinical behavior, which was not the case. A second
option is that down-stream alterations in the p53
pathway could lead to overexpression of p53 without
mutation. In such cases, sequencing will not be of
much identifying help, whereas immunohistochemistry may serve as a better indicator of a functionally
abnormal pathway leading to a more aggressively
dysregulated tumor.
Even though all these hypotheses are possible,
there is another explanation for the discrepancy
between sequencing and p53 immunohistochemistry
that may be unique to soft tissue sarcoma; soft tissue
sarcoma can commonly attain large size (as was the
case in all the cases evaluated in the present study)
and are known to be highly heterogeneous; it is possible that results of the 2 assays represent 2 different
fields of the tumor. Thus, the part of the tumor used
for DNA collection harbors a p53 mutation while the
segment used for immunohistochemistry does not,
and vice versa. Although it is impossible to assay an
entire soft tissue sarcoma specimen, analyzing p53
alterations by using both sequencing and immunohistochemistry may reduce the chance for error. In
addition, different sections of the tumor should be
studied.
Whereas our data cannot support replacing p53
sequencing by immunohistochemistry for the identification of p53 mutations in soft tissue sarcoma, it
does suggest the potential benefit of using p53
immunohistochemistry as a prognostic marker in
stage III soft tissue sarcoma. p53 immunopositivity
has been previously shown to correlate with soft tissue sarcoma outcome.7,8 Current soft tissue sarcoma
prognostic staging systems are based on crude clinical and tumor variables such as size, grade, and the
presence or absence of overt regional nodal or distant visceral metastasis. There is a growing awareness that more sensitive prognostication and
therapeutic decision-making algorithms will need to
incorporate relevant molecular determinants. In light
of the findings reported above, it may be particularly
useful to prospectively investigate the role of p53
expression as potentially integral to the soft tissue
sarcoma staging system.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
REFERENCES
1.
19.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
2333