British Journal of Dermatology (1992) 127, 566-570.
Mutant p53 oncogene expression in keratoacanthoma and squamous cell carcinoma T.J.STEPHENSON, J.ROYDS, P.B.SILCOCKS* AND S.S.BLEEHENf Departments of Pathology, *Public Health Medicine and -tDermatology, University of Sheffield Medical School, Sheffield, U.K. Accepted for publication 26 luly 1992
Summary
The tumour suppressor gene p53, located on the short arm of chromosome 17, encodes for a nuclear protein which regulates cell proliferation by inhibiting cells entering S-phase. p53 mutations are alleged to be the commonest genetic abnormality in human cancer. We studied mutant p53 oncoprotein expression, using PAbl801 monoclonal antibody immunohistochemistry, in 25 'ideal' keratoacanthomas and 26 well-, 19 moderately and 18 poorly differentiated squamous cell carcinomas of the skin. While there was a highly significant trend in the proportion of p53 oncoprotein-positive lesions from keratoacanthomas to poorly differentiated squamous cell carcinomas (/^ = 1 7 1 3 , d f = l , exact P=0-00003), p53 expression was inadequate for distinguishing keratoacanthoma from well-differentiated squamous cell carcinoma (;(-^ = 2 • 5 5, df = 1, exact P = 0 • 18; corresponding to a sensitivity of 0 84 and a specificity of only 0 36).
It has been proposed that p53 gene mutations are the commonest genetic abnormality in human malignancy.^ The gene, located on the short arm (p) of chromosome 17, encodes for a nuclear phosphoprotein involved in the inhibition of cell proliferation^ by preventing cells from entering the S-phase.^ Combined loss of one allele with mutation of the remaining allele, has been demonstrated at the p53 locus (17p 131), a finding typical for suppressor oncogenes.* Mutations and deletions ofthe p5 3 gene have been demonstrated in several forms of human malignancy,'"" including squamous cell carcinomas from internaF-^'^^ and epidermal sites.^* The latter study showed p53 mutations at dipyrimidine sites, and the presence of CC to TT double base alterations in epidermal, but not other, malignancies. These are findings typical of UV light-induced DNA damage, and have implications for the aetiology of squamous cell carcinoma ofthe skin. Present in exons of the p53 gene, the mutations lead to synthesis of an abnormal p53 protein.^'^' Normal p53 protein regulates cell proliferation and inhibits transformation, but can be converted by point mutations into a possibly dominant oncogene with transforming activity. ^^"^^ Owing to the short half-life of the wild-type protein, p53 protein levels in non-transformed cells under norCorrespondence: Dr T.I.Stephenson, Department of Pathology, University of Sheffield Medical School, Beech Hill Road, Sheffield SIO 2RX, U.K.
566
mal conditions are sufficiently low to prevent detection by immunohistochemistry." The elevated levels of p53 protein, demonstrable by immunohistochemistry'•^^•^' are thought to result from post-translational stabilization of the mutant protein by complexing with other proteins such as the wild-type product of the normal allele, viral proteins and heat-shock proteins.^^'^^''^'^ Immunohistochemistry for p53 protein is a valid screening method for mutations in the p53 gene, because previous combined immunohistochemical and molecular biological studies in various human tumours have shown strong correlation between p53 gene mutations and immunohistochemical staining for p53 protein. " • " • ^ 1 " We selected monoclonal antibody PAbl801 (Cambridge Research Biochemicals, U.K.), which recognizes an epitope between amino acids 32-3 7 close to the N-terminus of both wild-type and mutant p53 protein,^^ with which to perform immunohistochemistry. This highly specific monoclonal antibody reliably recognizes p53 protein; other antibodies (PAb240, PAb241) offer no advantage in recognizing mutant rather than wildtype p53 protein in human tissues.^^'^*"^' Furthermore, the effect of different fixation processes on immunohistochemical demonstration and localization of p53 protein by this antibody is known.^' We compared the expression of mutant p53 protein in keratoacanthomas, and squamous cell carcinomas with various degrees of differentiation, for three reasons:
MUTANT p53 ONCOGENE IN KA AND SCC
(i) because of continuing interest in pathological differences between keratoacanthoma and squamous cell carcinoma;^*-^'' (ii) because keratoacanthomas are an example of an invariably regressing neoplasm,^^ and analysis of p53 mutations in these lesions may add to the still-evolving story'^ of how far down the multistep pathway to neoplasia p53 mutation lies; (iii) if keratoacanthomas express mutant p53 protein, then either the p53 mutation is reversible or is not a determinant of obligate malignancy. This would be an interesting finding in the light ofthe recent demonstration^^ of H-ras oncogene in rabbit keratoacanthomas and its implication in the process of tumour regression.
Methods Selection of cases
Proof that a keratoacanthoma is correctly diagnosed as such is a perennial difficulty in this type of study.'^' As our 'gold standard' keratoacanthomas, we selected 25 in which the clinical presentation, naked eye and histological features were absolutely typical. ^° Eor comparison, 26 well-, 19 moderately, and 18 poorly differentiated squamous cell carcinoma excisions were selected by review of the departmental classified files.
567
was cut from each block for diagnostic review, and the next two sections were taken for the avidin-biotin complex (Vector Laboratories, U.K.) immunoperoxidase technique. The primary antibody, PAbl801, was applied at 1:400 dilution for 1 h at 20°C. Negative immunohistochemical controls were performed by omitting the primary antibody and substituting other monoclonal antibodies of the same immunoglobulin subtype. Normal skin and benign adnexal neoplasms were used as negative tissue controls. Positive tissue controls included prostatic adenocarcinomas which, as fresh tissue, had stained with the antibody PAb240, which is thought to be specific for mutant as opposed to wild-type p53 oncoprotein.^' The same two observers (T.J.S. and J.R.) viewed each case simultaneously, blind as to the histological diagnosis. Where any tumour staining was seen—either nuclear, or nuclear and cytoplasmic were accepted as genuine^'—the case was assigned to the positive category. Statistical analysis X^ tests on the proportion of p53 positive lesions were performed using the package STATXACT'-^ which yields exact P values even with small sample sizes.
1mm unohistochemistry
Results
Tissues were fixed for 12-36 h in 10% neutral-buffered formalin, and embedded in paraflin wax. One section
In all cases in which staining was detected, this was nuclear, with the exception of two cases of poorly
Figure 1. Tumour/stroma interface ofa keratoacarithoma at high power. Brownstained nuclei can be seen in the basal layers of the epithelium. (Immunoperoxidase for p53 protein with haematoxylin counterstain.)
568
T.J.STEPHENSON et aL
Table 2. Mutant p53 immunohistochemistry: the two most easily confused lesions
Table 1. Mutant p53 immunohistochemistry: all lesions
Squamous cell carcinoma: degree of differentiation Keratoacanthoma
p53 negative p53 positive Odds ratio
Keratoacanthoma
Well
Moderate
Poor
21 4 1
14 8 0 33
9 10 0-17
4 14 0-05
(trend). 1 degree of freedom = 17-13, P < 0 0 0 0 0 3 .
differentiated squamous cell carcinoma which also showed faint cytoplasmic staining. In the four keratoacanthomas showing staining, this was in nuclei of cells in the deep part, or the peripheral advancing margin of the tumour (Fig. 1). In the 32 squamous cell carcinomas with positive staining, this was in nuclei more widely distributed within the neoplasm (Fig. 2), Table 1 shows the profile of mutant p53 protein immunohistochemistry for all lesions. Four of 2 5 keratoacanthomas, 8 of 22 well-, 10 of 19 moderately, and 14 of 18 poorly differentiated squamous cell carcinomas showed mutant p53 protein expression. There was, therefore, a trend for keratoacanthomas to be negative, and for progressively less difi'erentiated squamous cell carcinomas to be positive (exact P=0-00003). However, if p53 protein immunohistochemistry were applied as a 'test' to differentiate between keratoacanthoma and
p53 negative p53 positive
21 4
Well-differentiated squamous cell carcinoma 14
Sensitivity of p53 immunostaining negative/positive in detecting keratoacanthoma = 0 84 Specificity of p53 immunostaining negative/positive in detecting keratoacanthoma = 0 3 6 X^ (1 degree of freedom) = 2-55; P = 0 1 8 (not significant).
well-differentiated squamous cell carcinoma—the main differential diagnosis^"—this would not be successful (;(2 = 1.55, d f = l , exact P = 0-18) (Table 2). This corresponds to a sensitivity of p53 protein expression negative/positive in detecting keratoacanthoma of 0-84 with a specificity of only 0-36.
Discussion Loss or alteration ofthe 54-kDa nuclear phosphoprotein encoded by the p53 gene,'* which is involved in suppression of cell proliferation, is thought to contribute to the deregulated proliferation of neoplastic cells.* p53 mutations, sometimes accompanied by chromosome 17
Figure 2. Centre of a poorly differentiated squamous cell carcinoma at high power. The brown-stained nuclei extended throughout the tumour. (Immunoperoxidase for p53 protein with haematoxylin counterstain.)
MUTANT p53 ONCOGENE IN KA AND SCC
deletions, were discovered in cancer of the colon.'' and subsequently described in breast cancer.*'^* astrocytoma,* lung cancer'^ and leukaemias.''^" These findings have led to the suggestion that p53 mutations are the commonest genetic abnormality in human cancer.' Although there are cases of immunohistochemical detection of p53 protein in non-neoplastic conditions of lymphoid tissues, indicating non-mutational stabilization ofthe p53 protein,'' in neoplasia the immunohistochemical detection of p53 protein has been shown to be synonymous with p53 gene mutations.'^-^'-^^ We have demonstrated p53 protein, and by implication p53 mutation, in a limited number (four of 25) of clinically and histopathologically typicaP" keratoacanthomas. Provided that these cases are not misdiagnosed squamous cell carcinomas—a misdiagnosis which is difficult to eliminate^'—then we have shown, by implication, p53 mutation in a 'neoplasm' which is invariably subject to spontaneous regression.^^ In addition, although the number of positive keratoacanthomas is small, we have shown a difference in distribution of the p53 protein-containing nuclei between the lesions. In keratoacanthomas, the stained nuclei were in the tumour immediately adjacent to the surrounding stroma (Eig. 1) whereas in squamous cell carcinomas nuclear staining was more widely distributed (Eig. 2). If p53 gene mutations occur in the spontaneously regressing keratoacanthoma. this is a surprising finding, in view ofthe proposed relatively late occurrence of p53 gene mutations during the multistep pathway to neoplasia.'^ Eurthermore, if p53 gene mutation occurs in keratoacanthomas, then this must in some way be either reversible, or not an invariable precursor of malignancy. In our small group of positive keratoacanthomas, the p53 protein-expressing nuclei were at the peripheral (possibly 'growing') edge of the tumour and not at the central area near the crater, where regressive features are common.^° In squamous carcinomas, which are normally relentlessly progressive and lack regressive features, p53 protein expression was uniform throughout the tumour. Other authors have reported a tendency for well-, but not poorly differentiated, squamous cell carcinomas to have selective p53 protein expression at the periphery. ^^ The reason for spontaneous regression in keratoacanthomas is unknown, although this may involve cellmediated immunity^''•^^ or recapitulation of the normal growth and involution phases of the hair follicle from which some consider it arises.^*"" Of additional interest to our finding of a selective distribution of p53 protein expression in keratoacanthomas is the demonstration.
569
in a rabbit keratoacanthoma model, of H-ras oncogene and its implication in tumour regression.-^^ While p53 mutations can include those typical of UVinduced DNA damage,'* p53 mutations occur in squamous carcinomas of internal'^'^ as well as epidermal'* sites. In internal sites, UV-damage is not likely to be a carcinogen. Our demonstration of mutant p5 3 protein in keratoacanthomas and squamous carcinomas carries no implication for the aetiology because from a proteinbased immunohistochemical study we cannot deduce the type of genetic mutations occurring. With regard to the popular topic of comparing attributes of well-differentiated squamous cell carcinoma and keratoacanthoma,^"-^^ we were not surprised that our study showed no clear-cut distinction (Table 2) between the two lesions. Our previous work,-^^-^'^'^* and that of others^* leads us to believe that keratoacanthoma and well-differentiated squamous cell carcinoma have more pathological similarities than differences. Their radically different behaviour awaits an explanation.
References 1 Harris AL. Mutant p53—the commonest genetic abnormality in human cancer? / Pathol 1990; 162: 5-6. 2 Finlay CA, Hinds PW. Levine AJ. The p53 proto-oncogene can act as a suppressor of transformation. Cell 1989; 57: 1083-93. 3 Martinez J, Georgotf I, Martinez J. Levine A. Cellular localization and cell cycle regulation by a temperature-sensitive p53 protein. Genes Dev 1991; 5: 151-9. 4 Nigro JM, Baker SJ, Preisinger AC et al. Mutations in the p53 gene occur in diverse human tumour types. Nature 1989; 342: 705-8. 5 Van der Berg FM. Tigges AJ, Schipper ME et al. Expression of nuclear oncogene p53 in colon tumours. / Pathol 1989; 157: 193-9. 6 Thompson AM. Steel CM, Chetty U et ai. p53 gene mRNA expression and chromosome 17p allele loss in breast cancer. Br J Cancer 1990; 61: 74-8. 7 Bressac B, Galvin KM. Liang TJ et al. Abnormal structure and expression of p53 gene in human hepatocellular carcinoma. Proc Natl Acad Sci USA 1990; 87: 1973-7. 8 Mulligan LM, Matlashewski GJ, Scrable HJ, Cavenee WK. Mechanisms ofp53 loss in human sarcomas. Proc Natl Acad Sci USA 1990; 87: 5863-7. 9 Mashal R, Shtalrid M, Talpaz M et al. Rearrangement and expression of p53 in the chronic phase and blast crisis of chronic myelogenous leukaemia. Blood 1990; 75: 180-9. 10 Sugimoto K, Toyoshima H, Sakai R et al. Mutations ofthe p53 gene in lymphoid leukaemia. Blood 1991; 77: 1153-6. 11 Villuendas R. Piris MA, Orradre ]L et al. p53 protein expression in lymphomas and reactive lymphoid tissue. / Pathol 1992; 166: 235-41. 12 Hollstein MC, Metcalf RA, Welsh JA et al. Frequent mutation ofthe p53 gene in human esophageal cancer. Proc Natl Acad Sci USA 1990; 87: 9958-61. 13 Iggo R. Gatter K, Bartek J et al. Increased expression of mutant forms of p53 oncogene in primary lung cancer. Lancet 1990; 335: 675-9.
570
T.J.STEPHENSON et aL
14 Brash DE, Rudolph JA, Simon JA et al, A role for sunlight in skin cancer; UV-induced p53 mutations in squamous cell carcinoma. Proc Natl Acad Sci USA 1991; 88: 10124-8. 15 Baker SJ. Fearon ER. Nigro JM et al. Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science 1989; 244: 217-21. 16 Finlay CA, Hinds PW, Tan TH et al Activating mutations for transformation by p53 produce a gene product that forms an hsc7O-p53 complex with an altered half-life. Mol Cell Biol 1988; 8: - 531-9. 17 Hinds PW. Finlay CA. Levine AJ. Mutation is required to activate the p53 gene for co-operation with the ras oncogene and transformation. / Virol 1989; 63: 739-46. 18 Lane DP, Benchimol S. p53; oncogene or anti-oncogene. Genes Dev 1990; 4: 1-8. 19 Gusterson BA, Anbazhagan R, Warren W et al. Expression of p53 in premalignant and malignant squamous epithelium. Oncogene 1991; 6: 1785-9. 20 Gannon JV. Greaves R, Iggo R, Lane SP. Activating mutations in p53 produce a common conformational effect. A monoclonal antibody speciflc for the mutant form. EMBO / 1990: 9: 1595602. 21 VarleyJM, Bramma WJ, Lane DP eta/. Loss of chromosome 17pl3 sequences and mutation of p53 in human breast carcinomas. Oncogene 1991; 6: 413-21. 22 Bartek J, Iggo R, Gannon J, Lane DP. Genetic and immunochemical analysis of mutant p5 3 in human breast cancer cell lines. Oncogene 1990; 5: 893-9. 23 Banks L, Matlashewski G, Crawford L. Isolation of human-p53speciflc monoclonal antibodies and their use in the studies of human p53 expression. Eur f Biochem 1986; 159: 529-34. 24 Cattoretti G, Rilke F. Andreola S et al. p53 expression in breast cancer. Int / Cancer 1988; 41: 178-83. 25 Purdie CA. O'Grady J. Piris J et al. p53 expression in colorectal tumors. Am ] Pathol 1991; 138: 807-13.
26 Stephenson TJ, Cotton DWK. Flow cytometric comparison of keratoacanthoma and squamous cell carcinoma. Br ] Dermatol 1988; 118: 582-3. 27 Stephenson TJ. Nichols CE, Cotton DWK. Can keratoacanthoma and squamous cell carcinomas be distinguished by flow cjftometry and immunohistochemistry.' / Pathol 1987; 152: 238A. 28 Corominas M, Leon J, Kamino H et al. Oncogene involvement in tumor regression; H-ras activation in the rabbit keratoacanthoma model. Oncogene 1991; 6: 645-51. 29 Iverson RE, Vistnes LM. Keratoacanthoma is frequently a dangerous diagnosis. Am J Surg 1973; 126: 359-65. 30 Kern WH, McCray MK. The histopathologic differentiation of keratoacanthoma and squamous cell carcinoma of the skin. f Cutan Pathol 1980; 2: 318-25. 31 Fox SB, Persad RA, Royds J etfl/.p53 and c-myc expression in stage Al prostatic adenocarcinoma: useful prognostic determinants? / Urol (in press). 32 StatXact User Manual Vol. 2, Cambridge, Mass.; Cytel Software Corporation. 1991. 33 Slater SD, McGrath JA, Hobbs C et al. Expression of mutant p53 gene in squamous carcinoma arising in patients with recessive epidermolysis bullosa. Histopathology 1992; 20: 237-41. 34 Stephenson TJ, Nichols CE, Cotton DWK. Keratoacanthomas contain a higher density of Langerhans cells than well differentiated squamous cell carcinomas. / Pathol 1987; 152: 238A. 35 Lawrence N, Reed RJ. Actinic keratoacanthoma. Speculations on the nature of the lesion and the role of cellular immunity in its evolution. Am J Dermatopathol 1990; 12: 517-33. 36 Ghadially FN. The role of the hair follicle in the origin and evolution of some cutaneous neoplasms of man and experimental animals. Cancer 1961; 14: 801-16. 3 7 Ghadially FN, Barton BW, Kerridge DF. The etiology of keratoacanthoma. Cflncer 1963; 16: 603-11.
Mutant p53 oncogene expression in keratoacanthoma and squamous cell carcinoma T.J.STEPHENSON, J.ROYDS, P.B.SILCOCKS* AND S.S.BLEEHENf Departments of Pathology, *Public Health Medicine and -tDermatology, University of Sheffield Medical School, Sheffield, U.K. Accepted for publication 26 luly 1992
Summary
The tumour suppressor gene p53, located on the short arm of chromosome 17, encodes for a nuclear protein which regulates cell proliferation by inhibiting cells entering S-phase. p53 mutations are alleged to be the commonest genetic abnormality in human cancer. We studied mutant p53 oncoprotein expression, using PAbl801 monoclonal antibody immunohistochemistry, in 25 'ideal' keratoacanthomas and 26 well-, 19 moderately and 18 poorly differentiated squamous cell carcinomas of the skin. While there was a highly significant trend in the proportion of p53 oncoprotein-positive lesions from keratoacanthomas to poorly differentiated squamous cell carcinomas (/^ = 1 7 1 3 , d f = l , exact P=0-00003), p53 expression was inadequate for distinguishing keratoacanthoma from well-differentiated squamous cell carcinoma (;(-^ = 2 • 5 5, df = 1, exact P = 0 • 18; corresponding to a sensitivity of 0 84 and a specificity of only 0 36).
It has been proposed that p53 gene mutations are the commonest genetic abnormality in human malignancy.^ The gene, located on the short arm (p) of chromosome 17, encodes for a nuclear phosphoprotein involved in the inhibition of cell proliferation^ by preventing cells from entering the S-phase.^ Combined loss of one allele with mutation of the remaining allele, has been demonstrated at the p53 locus (17p 131), a finding typical for suppressor oncogenes.* Mutations and deletions ofthe p5 3 gene have been demonstrated in several forms of human malignancy,'"" including squamous cell carcinomas from internaF-^'^^ and epidermal sites.^* The latter study showed p53 mutations at dipyrimidine sites, and the presence of CC to TT double base alterations in epidermal, but not other, malignancies. These are findings typical of UV light-induced DNA damage, and have implications for the aetiology of squamous cell carcinoma ofthe skin. Present in exons of the p53 gene, the mutations lead to synthesis of an abnormal p53 protein.^'^' Normal p53 protein regulates cell proliferation and inhibits transformation, but can be converted by point mutations into a possibly dominant oncogene with transforming activity. ^^"^^ Owing to the short half-life of the wild-type protein, p53 protein levels in non-transformed cells under norCorrespondence: Dr T.I.Stephenson, Department of Pathology, University of Sheffield Medical School, Beech Hill Road, Sheffield SIO 2RX, U.K.
566
mal conditions are sufficiently low to prevent detection by immunohistochemistry." The elevated levels of p53 protein, demonstrable by immunohistochemistry'•^^•^' are thought to result from post-translational stabilization of the mutant protein by complexing with other proteins such as the wild-type product of the normal allele, viral proteins and heat-shock proteins.^^'^^''^'^ Immunohistochemistry for p53 protein is a valid screening method for mutations in the p53 gene, because previous combined immunohistochemical and molecular biological studies in various human tumours have shown strong correlation between p53 gene mutations and immunohistochemical staining for p53 protein. " • " • ^ 1 " We selected monoclonal antibody PAbl801 (Cambridge Research Biochemicals, U.K.), which recognizes an epitope between amino acids 32-3 7 close to the N-terminus of both wild-type and mutant p53 protein,^^ with which to perform immunohistochemistry. This highly specific monoclonal antibody reliably recognizes p53 protein; other antibodies (PAb240, PAb241) offer no advantage in recognizing mutant rather than wildtype p53 protein in human tissues.^^'^*"^' Furthermore, the effect of different fixation processes on immunohistochemical demonstration and localization of p53 protein by this antibody is known.^' We compared the expression of mutant p53 protein in keratoacanthomas, and squamous cell carcinomas with various degrees of differentiation, for three reasons:
MUTANT p53 ONCOGENE IN KA AND SCC
(i) because of continuing interest in pathological differences between keratoacanthoma and squamous cell carcinoma;^*-^'' (ii) because keratoacanthomas are an example of an invariably regressing neoplasm,^^ and analysis of p53 mutations in these lesions may add to the still-evolving story'^ of how far down the multistep pathway to neoplasia p53 mutation lies; (iii) if keratoacanthomas express mutant p53 protein, then either the p53 mutation is reversible or is not a determinant of obligate malignancy. This would be an interesting finding in the light ofthe recent demonstration^^ of H-ras oncogene in rabbit keratoacanthomas and its implication in the process of tumour regression.
Methods Selection of cases
Proof that a keratoacanthoma is correctly diagnosed as such is a perennial difficulty in this type of study.'^' As our 'gold standard' keratoacanthomas, we selected 25 in which the clinical presentation, naked eye and histological features were absolutely typical. ^° Eor comparison, 26 well-, 19 moderately, and 18 poorly differentiated squamous cell carcinoma excisions were selected by review of the departmental classified files.
567
was cut from each block for diagnostic review, and the next two sections were taken for the avidin-biotin complex (Vector Laboratories, U.K.) immunoperoxidase technique. The primary antibody, PAbl801, was applied at 1:400 dilution for 1 h at 20°C. Negative immunohistochemical controls were performed by omitting the primary antibody and substituting other monoclonal antibodies of the same immunoglobulin subtype. Normal skin and benign adnexal neoplasms were used as negative tissue controls. Positive tissue controls included prostatic adenocarcinomas which, as fresh tissue, had stained with the antibody PAb240, which is thought to be specific for mutant as opposed to wild-type p53 oncoprotein.^' The same two observers (T.J.S. and J.R.) viewed each case simultaneously, blind as to the histological diagnosis. Where any tumour staining was seen—either nuclear, or nuclear and cytoplasmic were accepted as genuine^'—the case was assigned to the positive category. Statistical analysis X^ tests on the proportion of p53 positive lesions were performed using the package STATXACT'-^ which yields exact P values even with small sample sizes.
1mm unohistochemistry
Results
Tissues were fixed for 12-36 h in 10% neutral-buffered formalin, and embedded in paraflin wax. One section
In all cases in which staining was detected, this was nuclear, with the exception of two cases of poorly
Figure 1. Tumour/stroma interface ofa keratoacarithoma at high power. Brownstained nuclei can be seen in the basal layers of the epithelium. (Immunoperoxidase for p53 protein with haematoxylin counterstain.)
568
T.J.STEPHENSON et aL
Table 2. Mutant p53 immunohistochemistry: the two most easily confused lesions
Table 1. Mutant p53 immunohistochemistry: all lesions
Squamous cell carcinoma: degree of differentiation Keratoacanthoma
p53 negative p53 positive Odds ratio
Keratoacanthoma
Well
Moderate
Poor
21 4 1
14 8 0 33
9 10 0-17
4 14 0-05
(trend). 1 degree of freedom = 17-13, P < 0 0 0 0 0 3 .
differentiated squamous cell carcinoma which also showed faint cytoplasmic staining. In the four keratoacanthomas showing staining, this was in nuclei of cells in the deep part, or the peripheral advancing margin of the tumour (Fig. 1). In the 32 squamous cell carcinomas with positive staining, this was in nuclei more widely distributed within the neoplasm (Fig. 2), Table 1 shows the profile of mutant p53 protein immunohistochemistry for all lesions. Four of 2 5 keratoacanthomas, 8 of 22 well-, 10 of 19 moderately, and 14 of 18 poorly differentiated squamous cell carcinomas showed mutant p53 protein expression. There was, therefore, a trend for keratoacanthomas to be negative, and for progressively less difi'erentiated squamous cell carcinomas to be positive (exact P=0-00003). However, if p53 protein immunohistochemistry were applied as a 'test' to differentiate between keratoacanthoma and
p53 negative p53 positive
21 4
Well-differentiated squamous cell carcinoma 14
Sensitivity of p53 immunostaining negative/positive in detecting keratoacanthoma = 0 84 Specificity of p53 immunostaining negative/positive in detecting keratoacanthoma = 0 3 6 X^ (1 degree of freedom) = 2-55; P = 0 1 8 (not significant).
well-differentiated squamous cell carcinoma—the main differential diagnosis^"—this would not be successful (;(2 = 1.55, d f = l , exact P = 0-18) (Table 2). This corresponds to a sensitivity of p53 protein expression negative/positive in detecting keratoacanthoma of 0-84 with a specificity of only 0-36.
Discussion Loss or alteration ofthe 54-kDa nuclear phosphoprotein encoded by the p53 gene,'* which is involved in suppression of cell proliferation, is thought to contribute to the deregulated proliferation of neoplastic cells.* p53 mutations, sometimes accompanied by chromosome 17
Figure 2. Centre of a poorly differentiated squamous cell carcinoma at high power. The brown-stained nuclei extended throughout the tumour. (Immunoperoxidase for p53 protein with haematoxylin counterstain.)
MUTANT p53 ONCOGENE IN KA AND SCC
deletions, were discovered in cancer of the colon.'' and subsequently described in breast cancer.*'^* astrocytoma,* lung cancer'^ and leukaemias.''^" These findings have led to the suggestion that p53 mutations are the commonest genetic abnormality in human cancer.' Although there are cases of immunohistochemical detection of p53 protein in non-neoplastic conditions of lymphoid tissues, indicating non-mutational stabilization ofthe p53 protein,'' in neoplasia the immunohistochemical detection of p53 protein has been shown to be synonymous with p53 gene mutations.'^-^'-^^ We have demonstrated p53 protein, and by implication p53 mutation, in a limited number (four of 25) of clinically and histopathologically typicaP" keratoacanthomas. Provided that these cases are not misdiagnosed squamous cell carcinomas—a misdiagnosis which is difficult to eliminate^'—then we have shown, by implication, p53 mutation in a 'neoplasm' which is invariably subject to spontaneous regression.^^ In addition, although the number of positive keratoacanthomas is small, we have shown a difference in distribution of the p53 protein-containing nuclei between the lesions. In keratoacanthomas, the stained nuclei were in the tumour immediately adjacent to the surrounding stroma (Eig. 1) whereas in squamous cell carcinomas nuclear staining was more widely distributed (Eig. 2). If p53 gene mutations occur in the spontaneously regressing keratoacanthoma. this is a surprising finding, in view ofthe proposed relatively late occurrence of p53 gene mutations during the multistep pathway to neoplasia.'^ Eurthermore, if p53 gene mutation occurs in keratoacanthomas, then this must in some way be either reversible, or not an invariable precursor of malignancy. In our small group of positive keratoacanthomas, the p53 protein-expressing nuclei were at the peripheral (possibly 'growing') edge of the tumour and not at the central area near the crater, where regressive features are common.^° In squamous carcinomas, which are normally relentlessly progressive and lack regressive features, p53 protein expression was uniform throughout the tumour. Other authors have reported a tendency for well-, but not poorly differentiated, squamous cell carcinomas to have selective p53 protein expression at the periphery. ^^ The reason for spontaneous regression in keratoacanthomas is unknown, although this may involve cellmediated immunity^''•^^ or recapitulation of the normal growth and involution phases of the hair follicle from which some consider it arises.^*"" Of additional interest to our finding of a selective distribution of p53 protein expression in keratoacanthomas is the demonstration.
569
in a rabbit keratoacanthoma model, of H-ras oncogene and its implication in tumour regression.-^^ While p53 mutations can include those typical of UVinduced DNA damage,'* p53 mutations occur in squamous carcinomas of internal'^'^ as well as epidermal'* sites. In internal sites, UV-damage is not likely to be a carcinogen. Our demonstration of mutant p5 3 protein in keratoacanthomas and squamous carcinomas carries no implication for the aetiology because from a proteinbased immunohistochemical study we cannot deduce the type of genetic mutations occurring. With regard to the popular topic of comparing attributes of well-differentiated squamous cell carcinoma and keratoacanthoma,^"-^^ we were not surprised that our study showed no clear-cut distinction (Table 2) between the two lesions. Our previous work,-^^-^'^'^* and that of others^* leads us to believe that keratoacanthoma and well-differentiated squamous cell carcinoma have more pathological similarities than differences. Their radically different behaviour awaits an explanation.
References 1 Harris AL. Mutant p53—the commonest genetic abnormality in human cancer? / Pathol 1990; 162: 5-6. 2 Finlay CA, Hinds PW. Levine AJ. The p53 proto-oncogene can act as a suppressor of transformation. Cell 1989; 57: 1083-93. 3 Martinez J, Georgotf I, Martinez J. Levine A. Cellular localization and cell cycle regulation by a temperature-sensitive p53 protein. Genes Dev 1991; 5: 151-9. 4 Nigro JM, Baker SJ, Preisinger AC et al. Mutations in the p53 gene occur in diverse human tumour types. Nature 1989; 342: 705-8. 5 Van der Berg FM. Tigges AJ, Schipper ME et al. Expression of nuclear oncogene p53 in colon tumours. / Pathol 1989; 157: 193-9. 6 Thompson AM. Steel CM, Chetty U et ai. p53 gene mRNA expression and chromosome 17p allele loss in breast cancer. Br J Cancer 1990; 61: 74-8. 7 Bressac B, Galvin KM. Liang TJ et al. Abnormal structure and expression of p53 gene in human hepatocellular carcinoma. Proc Natl Acad Sci USA 1990; 87: 1973-7. 8 Mulligan LM, Matlashewski GJ, Scrable HJ, Cavenee WK. Mechanisms ofp53 loss in human sarcomas. Proc Natl Acad Sci USA 1990; 87: 5863-7. 9 Mashal R, Shtalrid M, Talpaz M et al. Rearrangement and expression of p53 in the chronic phase and blast crisis of chronic myelogenous leukaemia. Blood 1990; 75: 180-9. 10 Sugimoto K, Toyoshima H, Sakai R et al. Mutations ofthe p53 gene in lymphoid leukaemia. Blood 1991; 77: 1153-6. 11 Villuendas R. Piris MA, Orradre ]L et al. p53 protein expression in lymphomas and reactive lymphoid tissue. / Pathol 1992; 166: 235-41. 12 Hollstein MC, Metcalf RA, Welsh JA et al. Frequent mutation ofthe p53 gene in human esophageal cancer. Proc Natl Acad Sci USA 1990; 87: 9958-61. 13 Iggo R. Gatter K, Bartek J et al. Increased expression of mutant forms of p53 oncogene in primary lung cancer. Lancet 1990; 335: 675-9.
570
T.J.STEPHENSON et aL
14 Brash DE, Rudolph JA, Simon JA et al, A role for sunlight in skin cancer; UV-induced p53 mutations in squamous cell carcinoma. Proc Natl Acad Sci USA 1991; 88: 10124-8. 15 Baker SJ. Fearon ER. Nigro JM et al. Chromosome 17 deletions and p53 gene mutations in colorectal carcinomas. Science 1989; 244: 217-21. 16 Finlay CA, Hinds PW, Tan TH et al Activating mutations for transformation by p53 produce a gene product that forms an hsc7O-p53 complex with an altered half-life. Mol Cell Biol 1988; 8: - 531-9. 17 Hinds PW. Finlay CA. Levine AJ. Mutation is required to activate the p53 gene for co-operation with the ras oncogene and transformation. / Virol 1989; 63: 739-46. 18 Lane DP, Benchimol S. p53; oncogene or anti-oncogene. Genes Dev 1990; 4: 1-8. 19 Gusterson BA, Anbazhagan R, Warren W et al. Expression of p53 in premalignant and malignant squamous epithelium. Oncogene 1991; 6: 1785-9. 20 Gannon JV. Greaves R, Iggo R, Lane SP. Activating mutations in p53 produce a common conformational effect. A monoclonal antibody speciflc for the mutant form. EMBO / 1990: 9: 1595602. 21 VarleyJM, Bramma WJ, Lane DP eta/. Loss of chromosome 17pl3 sequences and mutation of p53 in human breast carcinomas. Oncogene 1991; 6: 413-21. 22 Bartek J, Iggo R, Gannon J, Lane DP. Genetic and immunochemical analysis of mutant p5 3 in human breast cancer cell lines. Oncogene 1990; 5: 893-9. 23 Banks L, Matlashewski G, Crawford L. Isolation of human-p53speciflc monoclonal antibodies and their use in the studies of human p53 expression. Eur f Biochem 1986; 159: 529-34. 24 Cattoretti G, Rilke F. Andreola S et al. p53 expression in breast cancer. Int / Cancer 1988; 41: 178-83. 25 Purdie CA. O'Grady J. Piris J et al. p53 expression in colorectal tumors. Am ] Pathol 1991; 138: 807-13.
26 Stephenson TJ, Cotton DWK. Flow cytometric comparison of keratoacanthoma and squamous cell carcinoma. Br ] Dermatol 1988; 118: 582-3. 27 Stephenson TJ. Nichols CE, Cotton DWK. Can keratoacanthoma and squamous cell carcinomas be distinguished by flow cjftometry and immunohistochemistry.' / Pathol 1987; 152: 238A. 28 Corominas M, Leon J, Kamino H et al. Oncogene involvement in tumor regression; H-ras activation in the rabbit keratoacanthoma model. Oncogene 1991; 6: 645-51. 29 Iverson RE, Vistnes LM. Keratoacanthoma is frequently a dangerous diagnosis. Am J Surg 1973; 126: 359-65. 30 Kern WH, McCray MK. The histopathologic differentiation of keratoacanthoma and squamous cell carcinoma of the skin. f Cutan Pathol 1980; 2: 318-25. 31 Fox SB, Persad RA, Royds J etfl/.p53 and c-myc expression in stage Al prostatic adenocarcinoma: useful prognostic determinants? / Urol (in press). 32 StatXact User Manual Vol. 2, Cambridge, Mass.; Cytel Software Corporation. 1991. 33 Slater SD, McGrath JA, Hobbs C et al. Expression of mutant p53 gene in squamous carcinoma arising in patients with recessive epidermolysis bullosa. Histopathology 1992; 20: 237-41. 34 Stephenson TJ, Nichols CE, Cotton DWK. Keratoacanthomas contain a higher density of Langerhans cells than well differentiated squamous cell carcinomas. / Pathol 1987; 152: 238A. 35 Lawrence N, Reed RJ. Actinic keratoacanthoma. Speculations on the nature of the lesion and the role of cellular immunity in its evolution. Am J Dermatopathol 1990; 12: 517-33. 36 Ghadially FN. The role of the hair follicle in the origin and evolution of some cutaneous neoplasms of man and experimental animals. Cancer 1961; 14: 801-16. 3 7 Ghadially FN, Barton BW, Kerridge DF. The etiology of keratoacanthoma. Cflncer 1963; 16: 603-11.
