p53 Mediated tumor cell response to chemotherapeutic dna damage: A preliminary study in matched pairs of breast cancer biopsies

Original Contributions p53 Mediated Tumor Cell Response to Chemotherapeutic DNA Damage: A Preliminary Study in M a t c h e d Pairs of Breast Cancer Biopsies UTE M. MOLL, MD, ANNE G. OSTERMEYER, BA, JEAN-CHARLES AHOMADEGBE, PHD, MARIE-CHRISTINEMATHIEU, MD, AND GUY RIOU, MD W'dd type p53 plays a crucial role in maintaining genomic stability in both normal and tumor cells in vitro. When DNA damage occurs, p53 acts as a cell cycle checkpoint and induces a cellular response that aims at restoring genomic integrity, p53 may either allow the repair of damaged DNA by inducing a transient G1 arrest or may eliminate the damaged cells by triggering apoptosis. Mutant p53 fails to mediate any of these effects. From this, a p53 status-dependent response to therapy might be expected when tumors are treated with DNA-damaging genotoxic agents: Although wild type p53-harboring tumors have an intact checkpoint that might allow them to restore genomic integrity back to a pre-exposure level, mutant p53 tumors have a corrupted checkpoint that could lead to an accelerated loss of genomlc stability. Until now, no studies have been described that examine such a p53-mediated effect in vivo. The authors tested this response model in vivo comparing 32 matched biopsy pairs from patients with breast cancer before and after rigorously standardized polychemotherapy. Four of the five drugs specifically induce a wild type p53-mediated checkpoint response. Tumor tissue from matched pairs of untreated and treated biopsies of the same patient were analyzed for treatment-associated changes of p53 protein expression by immunocytochemistry and, in a few available specimens, of p53 genotype changes by pulymerase chain reaction-based DNA analysis. Treatment-associated changes of the p53 immunophenotype, which

the authors speculate to reflect clonal selection, occurred in 39% (12 of 31) of the specimens. One specimen was not informative. Most tumors undergoing clonal selection originally harbored mutant p53 (nine of 12), and only three of 12 tumors were wild type. This study shows that exposure to genotoxic agents is commonly associated with a change in p53 immunophenotype. Although the limited material in this cohort prevented direct analysis of genetic instability, these results suggest that tumors with altered p53 may be genomically less stable and, therefore, may be more likely to undergo treatment-induced clonal changes than wild type tumors. This study also shows that the rigorous matched sample approach, although difficult to obtain, is an important tool that allows the in vivo assessment of the tumor response to genotoxic therapy in a controlled fashion. HUM PATHOL 26:1293--1301. Copyright © 1995 by W.B. Saunders Company Key words: p53, DNA damage, chemotherapy, breast cancer, tumor cell. Abbreviations: CAD gene, trifunctional enzyme carbamoyl-P synthetase, aspartate transcarbamylase, dihydroorotase; PALA, N-phosphonacetyl-L-aspartate; inhibitor; SSCP single strand conformation polymorphism; N/C, nuclear/cytoplasmic; ALL, acute lymphoblastic leukemia; IBC, inflammatory breast carcinoma.

P53, DNA DAMAGE AND GENOMIC STABILITY

sponse to anticancer therapy would be of great benefit and might enable us to predict in which patient these agents will exhibit their desired tumor-specific cytoytoxicity, and conversely, in which patients they will only exert pressure to select more aggressive, resistant tumor cell populations. Decreased cellular uptake of chemotherapeutic agents (amplified multidrug resistance 1 gene MDR-1) can be one reason for resistance. However, it has become clear that important events that determine the ultimate fate o f a t u m o r cell occur after the drug interacts with its cellular target (DNA duplex or DNA metabolism) rather than from the genotoxic action o f the agents themselves. 2 They induce a cellular response that requires an active genetic program with three possible outcomes: (1) cell death through apoptosis or mitotic failure, (2) complete reconstitution with survival, or (3) accelerated genomic rearrangements and mutations with selection of more aggressive and resistant tumor cell clones. The p53 tumor suppressor plays a crucial role in the execution of this cellular checkpoint response after DNA damage in both normal and t u m o r cells. T h e exposure of cells to DNA damaging agents, such as ioniz-

A major clinical problem for successful radiation and c h e m o t h e r a p y is the unresponsiveness of some cancers or the appearance of highly aggressive, resistant tumors on relapse of an initially responsive neoplasm. A m o r e complete understanding of the cellular reFrom the Department of Pathology, State University of New York at Stony Brook, Stony Brook, N ~ and the Laboratoire de Pharmacologie Clinique et Mol6culaire and Service de Histopathologie C, Institut Gustave Roussy, Villejuif, France. Accepted for publication July 12, 1995. Supported in part by grant No. R29 CA60664-01 (to UMM) from the National Cancer Institute at the National Institute of Health, a Junior Faculty Research Award No. JFRA-477 (to UMM) from the American Cancer Society, the Catacasinos Cancer Research Award (to UMM), the Lilly Wehrli Ammann Breast Cancer Fund (to UMM), and a grant from Ligue Nationale contre le Cancer, Paris (to GR). Address correspondence and reprint requests to Ute M. Moll, MD, Department of Pathology, State University of New York at Stony Brook, NY 11794-8691. Copyright © 1995 by W.B. Saunders Company 0046-8177/95/2612-000355.00/0

1293

HUMAN PATHOLOGY

Volume 26, No. 12 (December 1995)

ing radiation, ultraviolet (UV), or c h e m o t h e r a p e u t i c agents specifically induces high levels of wild type p53 protein. This leads to either cell cycle arrest in G1, which may allow time for DNA repair ~-5 or cell death by apoptosis. 2'6-9 (1) After exposure to T-irradiation or actinomycin D, h u m a n myeloblastic leukemia cells and n o r m a l b o n e m a r r o w p r o g e n i t o r cells, b o t h h a r b o r i n g e n d o g e n o u s wild type p53, go into transient G1 arrest, c o n c o m i t a n t with an increase in p53 protein levels, s In contrast, hematopoietic cells that lack wild-type p53 protein or overexpress m u t a n t p53 are incapable of G1 arrest after irradiation (while their G2 arrest is unaffected), s (2) W h e n wild type p53 cDNA is transfected into cells that lack p53, G1 arrest after T-irradiation is partially restored. Conversely, w h e n m u t a n t p53 cDNA is transfected into t u m o r cells with e n d o g e n o u s wild type p53, it causes a loss of irradiation-induced G1 arrest (because o f a d o m i n a n t negative inhibitory interaction between m u t a n t and wild type p53).4 T h e introduction of wild type p53 into a myeloid leukemia cell line that normally lacks p53 results in rapid apoptosis. 6 Also, wild type p53 initiates apoptosis in thymocytes specifically when the inducing agents cause DNA strand breaks. 7,s T h e basis of the decision whether to react with G1 arrest or apoptosis is not yet clear. Some evidence suggests the c o n c o m i t a n t presence of an unscheduled, proliferative signal f r o m an o n c o g e n e m i g h t drive cells into selfdestruction rather than repair. 2'1° Alternatively, apoptosis rather than G1 arrest m i g h t be chosen in tumors that derive f r o m cells whose n o r m a l terminal differentiation involves apoptosis, such as colonic epithelium. ° In contrast, cells in which the p53 signaling pathway is inactivated because o f mutation, gene rearrangements, or protein Sequestration (ie, by MDM2) do n o t induce p53 a n d do n o t arrest in G1, but go on to replicate d a m a g e d DNA templates, which results in m o r e r a n d o m mutations a n d a n e u p l o i d y ) 1 Both wild type p53-mediated responses are p a r t of the role o f p53 in restoring genomic integrity. W h e n operating normally, wild type p53 reduces the probability of aquiring new or additional oncogenic mutations by suppressing genetic scrambling events, lm2 Importantly, the loss of wild type p53 is a c c o m p a n i e d by g e n o m i c instability. Experimentally, this is m e a s u r e d as cells acquiring the potential for g e n e amplification. n'l~ Cells are subjected to a strong selection pressure for g e n e amplification with the purine synthesis inhibitor PALA (N-phosphonacetyl-L-aspartate). Only cells that are capable of amplifying their CAD g e n e (trifunctional enzyme carbamoyl-P synthetase, aspartate transcarbamylase, dihydroorotase) will be able to circ u m v e n t the block and, therefore, can proliferate. Primary diploid fibroblasts derived f r o m p53 knock-out mice a n d their control littermates that carried either none, one, or two disrupted p53 alleles were tested) 1'12 T h e results clearly show that the loss of b o t h c o p i e s of wild type p53 is sufficient to allow gene amplification to occur. In contrast, ceils that retained at least o n e or b o t h wild type p53 alleles are incapable of CAD amplification. 13 ' 14 /Conversely, restoration of wild type p53 in ceils that contain only m u t a n t p53 alleles causes growth arrest a n d suppresses gene amplification. '4 These data

fit beautifully with the in vivo observation that primary h u m a n tumors have such an extraordinary frequency of b o t h p53 abnormalities a n d aneuploidy. T a k e n together, this is compelling in vitro evidence o f the role of wild type p53 as the guardian of the g e n o m e . I f true in vivo. it would also force a reexamination of the biological t u m o r response to anficancer therapy with respect to the t u m o r ' s original p53 status. Based on the strong in vitro evidence, the p53 status is likely to be an i m p o r t a n t predictor of o u t c o m e after genotoxic treatment. This r e p o r t is the first in vivo study of p53-associated g e n o m i c stability of h u m a n cancers in response to chemotherapy. These data indicate that m u t a n t breast cancers tend to be genomically less stable than wild type tumors.

MATERIALS AND METHODS Tissue The 32 matched pairs of breast cancer tissue came from patients who presented in 1989 to the Institut Gustave Roussy at Villejuif, France. Their ages ranged from 35 to 63 years. Patients were selected for the following criteria: less than 70 years of age, clinical size of the tumor greater than 3 cm but absence of distant metastasis, no inflammatory type, and no previous cancer treatment. All tumors were infiltrating ductal carcinomas. Before treatment, two needle biopsies were obtained. One was embedded in paraffin, and used for pathological diagnosis and immunocytochemistry. The other one was used for DNA and RNA extraction. Each patient then received monthly cycles of doxorubicin (50 m g / m 2, day 1), vincristin (1 m g / m 2, day 1), cyciophosj0hamide (200 m g / m 2, days 2 to 4), methotrexate (10 m g / m z days 2 to 4) and 5-fluoronracil (300 m g / m 2 days 2 to 4) for 3 consecutive months. All tumors decreased significantly in size; unfortunately, the extent was not recorded in the patient's chart. Four weeks after the last cycle of chemotherapy, patients underwent surgical resection of the residual tumor consisting of either modified mastectomy (>3 cm) or lumpectomy (<3 cm) depending on the size of the tumor. All patients had an axillary lymph node dissection. Two to 3 weeks after surgery, patients received three additional cycles every 3 weeks consisting of 5-fluorouracil (500 mg/m2), adriamycin (50 mg/m2), and cyclophosphamide (500 m g / m 2) followed by radiotherapy according to the type of surgery. Currently, 4 to,5 years of clinical followup are available on all patients except case nos. 21 and 32. Immunocytochemistry The tumor tissues differed in their sizes: needle biopsies obtained before chemotherapy averaged 1 × 20 mm, whereas the much larger resection specimens measured about 15 × 20 mm. 4 to 5 #m sections were cut, dewaxed, and processed essentially as described. 15 The sections were microwave processed (2 × 5 minutes/1.3 kW) in 1% zinc sulfate (Pab 1801, DO-I, and CM-1) or 10 mmol/L citric acid buffer, pH 6 (DO7), quenched in .01% H20~ and blocked with 10% normal serum. Several antibodies specific for human wild type and mutant p53 were used. (None of the existing antibodies can distinguish between both forms in tissue.) Monoclonal antibodies DO-l, Pab 1801 (400 to 500 ng/mL) (Santa Cruz Biotechnologies, Santa Cruz, CA) and DO-7 (1:100) (Vector Laboratories, Burlingame, CA) and the polyclonal serum CM-1 (1:500 to 1,000) (gift of Dr David Lane) were used. Pal) 1801, DO-l, and DO-7 recognize N-terminal epitopes between

1294

P53 MEDIATED RESPONSETO ANTICANCER THERAPY(Moll et al)

TABLE 1. p53 Status of M a t c h e d Pairs of Breast Carcinomas Before and After Chemotherapy Pretreatment Case No.

p53 ICC

Posttreatment

Comment

Gase No.

p53 ICC

Comment

Mutation Pretreatment/ Posttreatment (SSCP or Sequ)

Survival

Wildtype p 5 3 tumors

21a

--

A l l tumor cells

21bJ"

N+

No M u t / n d

3/89-3/90 Rein

N+/++

Regions of N+ cancer cells admixed with p53 -neg. regions; 80% C+ with DO-7 A l l tumor cells

22a

--

All tumor cells

22b

C++

A l l tumor ceils; 10% N+ +

No M u t / n d

7/89-2/91 Brain met 4/91 Death 3/89-3/93 Rein

23a

--

A//tumor cells

23b

with DO-7 and CM-1 Altered p 5 3 tumors

24a

N+ +

All tumor cells

24b

O

25a 26a

N+ N++

A l l tumor cells All tumor cells

25b 26b

O N+/++

27a 28a

N++ N+

A l l tumor ceils A l l tumor cells

27b 28b

-

29a 30a

N++ N+/++

All tumor cells A l l tumor cells

29b 30bt

31a 32a

N+/++ N+

A l l tumor cells A l l tumor cells;

desmoplastic

31b 32b

-

C++

C+ C+

No tumor tissue present

4/89-5/93

No tumor tissue present 30% of tumor cells; sclerosing adenosis Only single tumor cells N+ + Only single tumor cells N+

Contralateral recurrence; 10/93 Alive 4/89-9/93 Rem 11/89-12/92 Bone met 10/93 Alive 6/89-8/93 Rein 3/89-6/93 Rein

A l l tumor cells All tumor cells; single tumor

cells N + / + + ; 60% N + + and single cells C+ with DO-7 All tumor cells All cells, only in situ CA and hyperplastic ducts present; single CA cells are N+

Exon 5, codon 161 Ala ~ Asp pretreatment only

4/89-7/93 Rein 11/89-10/93 Rem

5/89-3/90 Death 4/89-3/90 Rein

Abbreviations: ICC, immunocytochemistry; N, nuclear; C, cytoplasmic; N/C, diffuse; +, + +, + + +, positive; - , negative; ~ , no residual tumor left; CA, carcinoma; rem, remission; met, metastasis; sequ, direct sequencing; no rout, no mutation; nd, not determined. * Tumor nos. 1-20 showed no change in their p53 status. Tumor nos. 21-32 with change of p53 patterns. Tumor nos. 1-8 show undetectable p53 protein and are considered wild type. Accordingly, tumor 5a and 6a showed no mutation in exons 4-10. Tumor nos. 9-20 showed accumulation of p53 protein and are considered altered for p53 expression. Among them, nine tumors showed nuclear overexpression, one tumor showed cytoplasmic overexpression, and two tumors showed diffuse overexpression. Tumor nos. 17a (N++) and 17b ( N + / + + ) carries an exon 7, codon 245 Gly ~ Val (GCC ~ GAC) mutation (See Fig 2); tumor no. 18a (N/C+) and 18b ( N / C + + ) carries an exon 9, silent codon 313 AGC ~ AGT mutation; tumor nos. 20a ( N + / + + ) and 20b ( N + + + ) carries an intron 6 mutation by SSCP and a heterozygous codon 124 Cys ~ Trp (TGC "-* TGG) mutation. J- Discrepant staining results between DO-7 and the other p53 three antibodies. a m i n o acids 32-79 (PAb 1801) 1%26 ( D O - l ) a n d 37-45 (DO1 a n d DO-7). CM-1 is p r o d u c e d in r a b b i t against a bacterially e x p r e s s e d h u m a n wild type p53. After i n c u b a t i o n with p r i m a r y a n t i b o d y in 2% b o v i n e s e r u m a l b u m i n / p h o s p h a t e - b u f f e r e d saline (BSA/PBS), sections were w a s h e d a n d i n c u b a t e d in b i o t i n y l a t e d s e c o n d a r y IgG for 45 m i n u t e s at r o o m t e m p e r a ture. B o t h s t r e p t a v i d i n - c o u p l e d h o r s e r a d i s h - p e r o x i d a s e / D A B (Zymed, San Francisco, CA) a n d alkaline p h o s p h a t a s e / F a s t R e d (Biogenex, San R a m o n , CA) systems were u s e d for detection. I n all runs, color d e v e l o p m e n t was strictly s t a n d a r d i z e d for t i m e a n d s u b s t r a t e c o n c e n t r a t i o n to allow a m e a n i n g f u l a s s e s s m e n t o f s t a i n i n g intensity. B o t h p r e t r e a t m e n t a n d p o s t t r e a t m e n t samples o f all speci m e n s were s t a i n e d with DO-1 a n d CM-1. I n a d d i t i o n , all exc e p t n i n e s p e c i m e n s were s t a i n e d with DO-7. F u r t h e r m o r e , all p o s t t r e a t m e n t samples were also s t a i n e d with PAb 1801; however, o n p a r a f f i n sections PAb 1801, as h a d b e e n o b s e r v e d by others, was t h e weakest a n d only c o n f i r m e d results obt a i n e d with t h e o t h e r a n t i b o d i e s , Generally, D O - l , DO-7, CM1, a n d PAb 1801 s h o w e d e x c e l l e n t c o n c o r d a n c e . F o u r cases d i d show d i s c r e p a n t results with DO-7 (see T a b l e 1, m a r k e d

by t ) . N o r m a l m o u s e IgG at 500 n g / m L served as negative control. A d j a c e n t sections were s t a i n e d with hematoxylin-eosin to evaluate t u m o r m o r p h o l o g y , p53 i m m u n o s t a i n i n g was assessed by t h e a u t h o r s . T h e selection modalities for p53 overe x p r e s s i o n were t h e following: o v e r e x p r e s s i o n was e i t h e r nuclear (N), cytoplasmic (C), o r diffuse ( N / C ) ; t h e s t a i n i n g intensity was semiquantitatively assessed, u s i n g a 3 g r a d e syst e m ( + to + + + ) by c o m p a r i s o n with o u r negative c o n t r o l tissue ( n o r m a l b r e a s t tissue) a n d positive c o n t r o l tissue (a b r e a s t c a n c e r with k n o w n n u c l e a r o v e r e x p r e s s i o n b e c a u s e o f a p53 missense m u t a t i o n ) , w h i c h were i n c l u d e d in e a c h run.

Single-Strand Conformation Polymorphism (SSCP)/cDNA Sequencing D N A a n d total RNA were e x t r a c t e d f r o m t h e same p i e c e o f f r o z e n tissue by t h e g u a n i d i n e - i s o t h i o c y a n a t e m e t h o d . 16F o r SSCP analysis, e x o n s 5 to 9 were amplifyed individually u s i n g p r i m e r s within f l a n k i n g i n t r o n s a n d gels analyzed as described. 17 F o r s e q u e n c i n g , 2 # g served as t e m p l a t e for cDNA synthesis with M o l o n e y m u r i n e l e u k e m i a virus reverse tran-

1295

HUMAN PATHOLOGY

Volume 26, No. 12 (December 1995)

scriptase (Gibco BRL, Gaithersburg, MD) and random primers. Codons 118 to 353 of p53 were amplified with primer pair F1F2 as the authors previously described. 1~ Amplicons (400 ng) were subjected to direct sequencing using the fluorescent DyeDeoxySequencing Kit (ABI, Applied Biosystems, Foster City, CA) and run on an Automatic DNA sequencer (ABI Model 370A). In SSCP positive cases, the altered exon was sequenced, whereas in SSCP negative cases, the entire amplicon was sequenced using several internal primers. Statistical analysis

Fisher's exact test for small samples was used to calculate significance of changes in p53 expression.

RESULTS The authors analyzed the treatment-associated changes in p53 immunophenotype as a measure of genomic stability after challenge by chemotherapy in breast cancer. Changes in tumor cell composition may reflect either altered p53 expression in existing cell clones or clonal instability including both the appearance of treatment-induced new clones and the positive selection of particularly adapted pre-existing clones within the heterogeneous tumor mass. This clonal process could lead to a significant redistribution in tumor composition and can be determined by changes in p53 expression patterns. Using a matched sample approach, tumors were followed in vivo through a controlled DNA damage event. The authors analyzed 32 tissue pairs at diagnosis and after 3 months of standardized polychemotherapy, p53 protein expression was determined by immunocytochemistry using four different p53-specific antibodies. Nuclear overexpression was used as a marker for p53 mutations, 15'18 a lack of staining indicated normal p53, and cytoplasmic overexpression before treatment suggested nuclear exclusion of wild type p53.15'19 DNA/RNA was only available from four pairs of tumor samples and four additional pretreatment cases, and their mutational status was determined by SSCP analysis and direct sequencing. Four of the five drugs used in our study (doxornbicin, cyclophosphamide, methotrexate, and 5-fluorouracil) specifically induce rapid nuclear accumulation of wild type p53 protein within a few hours after treatment, suggesting that they trigger the checkpoint response in cells with functional p53. 2° This response is transient, reaching a maximum at 16 to 24 hours and declining to normal (undetectable) levels thereafter. Eleven of 31 pretreatment specimens with none or only occasional tumor cell staining were grouped into p53 wild type tumors. One specimen (tumor no. 9)

showed diffuse staining before and after treatment. Because no RNA was available, this pattern could not be correlated with a specific p53 alteration. Therefore, tumor no. 9 was not included for further analysis, which reduced the total number of informative specimens from 32 to 31. Concordant with the staining results, SSCP analysis on tumor nos. 5 and 6, both with only rare N+ cells interspersed among large numbers of p53-negative tumor ceils, did not show mutations because of limited sensitivity of detection. Twenty of 31 pretreatment specimens showed nuclear (19 specimens) or cytoplasmic (one specimen) overexpression and were put into the "altered p53" group. One specimen (tumor no. 18) originally showed a peculiar diffuse (N/C) staining pattern. After chemotherapy, diffuse (with Pab 1801, DO-1 and CM-1) as well as purely cytoplasmic accumulation of p53 (with DO-7) was present. Sequence analysis showed a silent C ~ T transition of the third nucleotide in codon 313 but no other mutations in both samples (data not shown). No constitutional DNA was available to resolve whether this was a mutation or a polymorphism. Twenty cases (nos. 1 to 20) showed no change in both p53 expression (see Fig 1A for case no. 11) and gene mutation (see Fig 2 for SSCP/DNA sequences). In contrast, treatment-associated change in p53 expression occurred in 12 of 31 (39%) of the tumors (Table 1; Fig 1B-D). Among the 12 tumors undergoing clonal change, the majority (nine specimens) originally harbored mutant p53 compared with only three original wild type tumors (Table 1). This corresponds to a clonal rearrangement of 45% (9 of 20) among the altered tumors and 27% (3 of 11) among the wild type tumors, a difference that is not significant (P = .28). The 4-year follow-up of the available 29 patients shows that among the original "wild-type" group, 30% (3 of 10 patients) died or are alive with metastatic disease, whereas this occurred in 42% (8 of 19) of the original "altered" group, a difference that is not significant, possibly because of the small sample size. The changes in expression pattern were complex, suggesting both global genomic rearrangements as well as mutations of p53 or p53-associated genes (Table 1). Two negative (p53 wild type) cases converted to regional or global nuclear overexpressors (tumor nos. 21 and 22). Tumor 21b is difficult to interpret because it shows N+ tumor cells only in some regions of the section. Also, the DO-7 staining is discordant with the other antibodies, showing cytoplasmic overexpression in 80% of the cells. Furthermore, the authors cannot exclude a sampling bias in the pretreatment sample. If regional nuclear positivity had been present before

FIGURE i . (A) Breast cancer with nuclear p53 overexpression (pretreatment); no change after therapy (posttreatment) (case no. 1l a and b). All tumor cells stain. Pretrecrlment, CM-1, 1:1000: posttreatment DO-1,500 ng/mL (B) Breast cancer with strong nuclear p53 overexpression conveffs into a p53 negative tumor after therapy. Note unstained tumor cells invading fat in the posttreatment photomicrograph (case no. 29a and b). PrelTeatment, CM-1,1:1000, posttreatment DO-l, 400 ng/mL. (C) Breast cancer with moderate nuclear overexpression of p53 converts into cytoplasmic overexpression after therapy (case no. 31a and b). Pretreatment and posttreatment, DO-l, 400 ng/mL. (D) Breast cancer with nuclear p53 overexpression of all cells shows no detectable residual tumor after chemotherapy (case no. 24a and b). Note the negative epithelium in two normal ducts to the right. Pretreatment, DO-l, 400 ng/ml4 posttreatment CM-1, 1:500. (Horseradish peroxidase DAB No counterstqin; Original magnification x400).

1296

P53 MEDIATED REPONSE TO ANTICANCER THERAPY (Moll et al)

1297

HUMAN PATHOLOGY

( ~ 17 a

b

18 a

b

Volume 26, No. 12 (December 1995)

20 a

b

C FIGURE 2. (A) Example of an SSCPanalysis of the p53 gene on genomic DNA before (a) and after (b) chemotherapy. The migration pattern of exon 7 is shown for case nos. 17, 18, and 20. Note the presence of DNA fragments with abnormal mobility in case no. 17.* (Lane C) DNA from the human cancer cell line HT3 with a codon 245 mutation was used as positive control for shifted migration. (B) Direct sequencing of cDNA from tumor 17 with an antisense primer. A codon 245 GCC --* GAC transversion (Gly --* Val mutation, see I~) is present in both pretreatment and posttreatment tumors. This mutation had caused the SSCPshift.

therapy, it could have been missed by the relatively small needle biopsy. When the authors eliminate case no. 21 from the study for these reasons and recalculate significance, the new Pvalue is .17, which further supports the trend. Four cases with global nuclear overexpression converted to tumors with rare or no nuclear staining (case nos. 26 to 29; eg, Fig 1B), suggesting that drug-induced genomic rearrangement events occurred that (among others) eliminated the mutant p53 genes altogether. This was suggested in tumor 28, which harbored a codon 161 Ala --, Asp mutation in the pretreatm e n t tumor that could no longer be detected in the posttreatment tumor (data not shown). The loss of the mutant signal is unlikely to be merely caused by a simple sampling error given the large Size of all the posttreatm e n t samples. The wild type signal that the authors detected after treatment could derive from contaminating normal tissue or from a hemizygous wild type tumor that has lost its mutant allele. Three cases with nuclear (tumor nos. 30 to 32) and one case without detectable (tumor no. 23) p53 expression converted to cytoplasmic overexpression of all tumor cells after treatm e n t (eg, Fig 1C), suggesting the selection of an abnormal p53 phenotype probably because of a mutation in the p53 regulatory pathway. 15'19 Interestingly, two mutant p53 cases (tumor nos. 24 and 25) only showed fibrosis but no remaining cancer tissue after chemotherapy, indicating a dramatic initial cell killing of mutant p53 cells (Fig 1D). One patient (case no. 25) is still in remission after 4 years. However, another patient (case no. 24) suffered a contralateral malignancy 4 years later, which probably represents a second primary breast carcinoma rather than a metastasis. The degree of concordance in staining results between CM-1, DO-l, DO-7, and PAb 1801 was generally excellent. The authors noticed a gradient in sensitivity for detection of p53 overexpression between the four antibodies, with DO-1 being the most sensitive and PAB 1801 the least sensitive. DO-7 showed nonconcordance in four cases (case no. 9 pretreatment and posttreatment; case nos. 18, 21, and 30 posttreatment only) (Table 1). In all

of them, DO-7 detected overexpression in the "opposite" compartment. The nature of the p53 species recognized by DO-7 in this circumstance and its tendency to occur after treatment remains to be clarified. Many larger tissue sections also contained uninvolved breast tissue. Normal epithelium of ductal and lobular origin was usually negative for p53 staining. Occasionally, uninvolved ducts showed diffuse p53 staining of slight to moderate intensity (minor nos. 15a, 20b, 21b, and 26b). However, it is not clear if this represents elevated levels of normal p53 throughout the cell or a nonspecific cross-reactivity, possibly because of drug treatment in some cases.

DISCUSSION I n v a s i v e Breast C a n c e r a n d G e n o t o x i c Therapy

A major advance in the treatment of" localized breast cancer came with early adjuvant systemic therapy, which prolongs the disease-free interval and overall survival.2123 Polychemotherapy (cyclophosphamide, methotrexate and fluorouracil with or without additional drugs) is clearly beneficial for women with node positive tumors, in particular before menopause. 24 Preoperative chemotherapy for tumors larger than 3 cm, as in the 32 patients in this study, has not yet become a mainstream therapeutic modality in the United States but is common in France. 25 This protocol has a strong rationale: to shrink the tumor, to decrease the possibility of facilitated dissemination of metastatic cells during sur~cal manipulation, and to start systemic treatment immediately when micrometastasis are at their smallest. In a preliminary randomized trial, overall survival was better when the same chemotherapeutic regimen was given preoperatively rather than postoperatively. 25 It also offered us a unique opportunity to study the gen o m e guarding function of wild type p53 in vivo after being challenged by anticancer therapy. p53 is commonly inactivated in breast cancer, and patients can be classified into p53 altered and p53 wild

298

P53 MEDIATED RESPONSE TO ANTICANCER THERAPY (Moll et al)

type groups. Compared with activation of oncogenes (erb B-2/neu, erb B-3, myc, ras, hst, or int-2) and loss of other tumor suppressor genes (Rb and Brca-1), which are all implicated in the oncogenesis of breast cancer, p53 mutations are the most frequent genetic change in this disease. 26 The r e p o r t e d prevalence of mutant p53 in primary breast cancer varies from 13% to 58%, but most studies find a r o u n d 40 %.ls'27-3s Most mutations are missense mutations within the central specific DNAbinding region of the molecule. 39 The aberrant p53 protein stabilizes in the nucleus to such high levels that the normally undetectable protein becomes detectable by immunocytochemistry and serves as a marker for mutations. Recently, the authors described evidence for a mutation-independent mechanism of p53 inactivation in several wild type p53 h u m a n tumors including inflammatory breast carcinoma (IBC).15'19 One third of IBC show nuclear exclusion with cytoplasmic sequestration of wild type p53.15'37 Because nuclear localization of p53 is absolutely essential for its function,4°-43p53 could be inactivated in these tumors. In untreated tumors, p53 mutations are associated with t u m o r size, high nuclear grade, and absence of h o r m o n e receptors. 34'35'37'~8 Many recent studies on c o m m o n types of invasive breast c a r c i n o m a 34-36'~8'44'45 a s well as on inflammatory type breast carcinoma 37 show that nuclear p53 overexpression is a p o o r prognostic factor for patients with and without lymph node involvement, and correlates with low 5 and 10 year survival. Although these studies establish p53 mutations as a p o o r prognostic marker, they pay no attention to the type of treatment that patients received after diagnosis. It is important to note that most of these patients underwent radiation or adjuvant chemotherapy, given the treatment modalities that have been applied in the last 10 to 15 years. The observed p o o r e r clinical outcome could be a reflection of an intrinsically higher aggressiveness of mutant p53 tumors (possibly related to the association between mutant p53 and increased cell proliferation of breast c a n c e r cells 44'4"~versus wild type p53 tumors i n d e p e n d e n t o f therapeutic intervention. This is supported by the fact that certain human-derived p53 mutants behave like an oncogene in vivo and, therefore, represent a true "gain of function."46 In addition, these mutants, unlike the wild type protein, can transactivate the p r o m o t e r of the multidrug resistance gene MDR-1, an adenosine t r i p h o s p h a t e - c o n s u m i n g effiux p u m p for cytotoxic drugs induced in chemotherapeutically treated cancer cells. 47 This could explain why such a large spectrum of mutant p53 genes are retained and overexpressed in virtually all h u m a n cancers that have u n d e r g o n e inactivating p53 mutations. T h e third and possibly most important reason for a p o o r e r clinical outcome in breast cancer patients with nonfunctional mutant p53 could be the fact that mutant p53, with the associated incompetence for DNA repair, poses a high selection pressure among surviving tumor cells for more unstable and clinically more aggressive clones after therapeutic DNA damage. This assumes that the drug-induced killing is less than 100% efficient, which is usually the case. In mice harboring fibrosarcomas, p53 deficiency or mutation is associated

with treatment resistance and relapse of the tumors. 48 This hypothesis raises the disturbing specter that this group of patients might be h a r m e d rather than helped by the standard genotoxic therapy. Chemotherapeutic drugs from five different classes are strong inducers of the wild type p53-mediated checkpoint response. 3'2°'49 These include cisplatinum (metal complex that causes DNA strand breaks), mitomycin C and cyclophosphamide (DNA alkylating agents), actinomycin D and doxorubicin (DNA intercalating drugs that inhibit topoisomerase II), and 5-fluorouracil and methotrexate (nucleotide analogs that inhibit DNA synthesis).3,20 Four of these drugs (doxorubicin, cyclophosphamide, methotrexate, and 5 fluorouracil) have been used in this study. In contrast, cytotoxic drugs with other modes of action (that are not DNA damaging), such as vinca alkaloids or taxoids (inhibit microtubule formation) and arabinofuranosyl-cytosine (blocks S-phase), azacytidine or bromodeoxyuridine (base analogs), fail to induce p53. ~° The cellular p53 response is transient, reaching a maximum at 16 to 24 hours and declining to normal levels thereafter. Therefore, the prolonged detection of overexpressed p53 protein 4 weeks after the last drug exposure most likely reflected stable phenotypes with altered expression rather than a transient drug effect. 2° In an attempt to elucidate if p53 plays an active role in a tumor's biological response to anticancer therapy, the authors asked if the response is associated with a change in p53 expression of the target cells. Comparing a patient's tissue before and after standardized chemotherapy, the authors analyzed tumor tissue pairs for their stability of the p53 phenotype and, to a limited extent, the genotype of the tumor cell population. The authors speculate that stable alterations in a tumor's p53 status after chemotherapy most likely reflect a clonal selection process that occurred among the tumor cell population and therefore serves as a marker for genomic instability rather than a drug-related epiphen o m e n o n (ie, alteration of protein expression in unselected cells). However, the authors cannot formally exclude the possibility that changes in p53 protein levels in the posttreatment samples are caused by some sustained drug effect that affects the p53 turnover. Physiologically, normal p53 protein is so tightly regulated that it remains undetectable by immunocytochemistry, whereas p e r m a n e n t p53 overexpression is always associated with a stable genetic trait. Therefore, it seems reasonable to assume that the p53 overexpression that we observe in the posttreatment specimens is permanent, because the tissue was removed 4 weeks after the last drug exposure. In contrast, DNA damaging agents ab ways induce a transient overexpression that lasts up to 48 hours. 2-5'2° Because this retrospective study used small, formalin-fixed tissue samples, a direct measurement of genomic instability by fluorescent activated cell storing (FACS) ploidy analysis was precluded. All alternative assays of genomic instability are limited to in vitro analysis, and use either cytogenetic analysis of karyotypes or measures the amplification of the CAD gene with PALA. T M O u r preliminary results on 32 matched breast cancer samples show that 39% of the

1299

HUMAN PATHOLOGY

Volume 26, No. 12 (December 1995)

tumors exhibit p53-associated clonal instability in response to c h e m o t h e r a p y in vivo. Most of the tumors undergoing selection originally h a r b o r e d mutant p53 (nine of 12), whereas only three of 12 tumors were wild type. Although not significant, possibly because of small case numbers, a trend is observed that tumors with altered p53 are genomically less stable and, therefore more likely to u n d e r g o treatment induced clonal tumor cell selection than wild type tumors. Because of the inability of mutant p53 to elicit a checkpoint response, tumor cells u n d e r g o r a n d o m genetic rearrangements. Consequently, the authors observe a spectrum of clonal rearrangements within the tumor. Kinetic studies of h u m a n t u m o r cells have shown an average doubling time of 60 hours for solid tumors. 5° Thus, a 3-month period is equivalent to about 36 cell doublings. Therefore, selection processes can lead to a dramatic redistribution of clonal composition within a tumor mass. Redistribution of the clonal composition of tumors might not only be a function of m u t a n t p53-associated genomic instability but might also occur u n d e r circumstances in which wild type p53 induces tumor apoptosis rather than G1 arrest and repair. It has been speculated that wild type p53 cells choose apoptosis rather than repair if the damage is too substantial, but we currently lack an understanding of this dichotomy. One determinant might be cell differentiation. Indeed, the p53 apoptotic pathway might be restricted to certain cell types with specific differentiation, such as myeloid and thyrnocyte progenitor cells, 6~ or, in the case of epithelium, colonic epithelium whose normal terminal differentiation involves apoptosis. 9 It is unknown why the same type and dosage of DNA damage causes G1 arrest in fibroblasts but apoptosis in leukemia cells. 2'4 In addition, evidence exists that the concomitant presence of an unscheduled, proliferative signal from an oncogene might drive cells into self-destruction rather than repair. 2,t° Also, no association has b e e n observed with the class of DNA-damaging agent. Importantly, it is unclear if wild type p53-mediated apoptosis occurs in breast cancer. Currently available studies on breast cell lines with functional wild type p53 report either resistance to apoptosis 5~ or only limited apoptotic response after DNA damaging drugs. 5t'52 Wild type p53-mediated G1 arrest and apoptosis are an immediate response to DNA damage and occur rapidly within 72 hours after the event 2-5 By nature of its design, this clinical in vivo study does not measure these rapid events but rather measures clonal stability of t u m o r cells within a 3-month window, which reflects the overall consequence o f the immediate response. Likewise, this study does not give insight whether these p53 status-associated checkpoints result in increased or decreased chemosensitivity of breast cancer cells toward chemotherapy. This would largely d e p e n d whether in a given breast cancer wild type p53 causes arrest or apoptosis, or, conversely, m u t a n t p53 allows increasing aneuploidy and selection of resistant clones. Decreased radioresistance and chemoresistance has been shown experimentally in breast and other neoplasms. 2'48'53'~4 Conversely, three brief reports based on p53 immunostaining of pretreatment breast cancer biopsies show

either increased chemosensitivity55 or no significant correlation between the initial p53 protein status and clinical response. 56'57 In nonbreast malignancies, it is known from clinical experience that Wilms's tumors (J. Pelletier, personal communication, 1993), testicular germ cell cancers, 5s and acute lymphoblastic leukemia (ALL) of childhood 59 are sensitive to chemotherapy and radiation. They represent a class of tumors with a low incidence of p53 mutations at diagnosis. Conversely, failure of therapy correlates with the presence of p53 mutations in anaplastic Wilms's t u m o r (D. Housman, personal communication, 1994) and in relapsed A L L . 60,61

Taken together, these data support the notion that the p53 status of a tumor might be an important determinant of the biological response to anticancer therapy. Given the importance and controversy on this issue, more well-controlled clinical in vivo studies are urgently n e e d e d to determine if and how the p53 status has to enter anticancer treatment decisions. The authors propose that a rigorous matched sample approach is a powerful strategy for such studies.

Acknowledgment. The authors thank Drs M. Spielmann, J.P. Travagli, and L. Barreau-Plouhaer from Institut Gustave Roussy Departments of Medicine and Surgery for clinical data. They also thank Dr Thomas Rocek for advice in the statistical analysis and R. E1-Maghrabi and G. D'Angelo for technical assistance in sequencing. REFERENCES 1. Chin KV, Pastan I, Gottesman MM: Function and regulation of the human mulfidrug resistance gene. Adv Cancer Res 60:157-180, 1993 2. Lowe SW, Ruley HE, Jacks T, et al: P55-dependent apoptosis modulates the cytotoxicity of anticancer agents. Cell 74:957-967, 1993 3. Kastan MB, Onyekwere O, Sidransky D, et al: Participation of p53 protein in the cellular response to DNA damage. Cancer Res 51:6304-6311, 1991 4. Kuerbitz SJ, Plunkett BS, Walsh WV, et al: Wild-type p53 is a cell cycle checkpoint determinant following irradiation. Proc Natl Acad Sci U S A 89:7491-7495, 1992 5. Kastan MB, Zhan Q, E1-Deir WS, et al: A mammalian cell cycle checkpoint pathway utilizing p53 and gadd 45 is defective in Ataxia Telangiectasia. Cell 71:587-597, 1992 6. Yonish-Rouach E, Resnitzky D, Lotem J, et al: Wild-type p53 induces apoptosis of myeloid leukemic cells that is inhibited by interleukin-6. Nature 352:345-347, 1991 7. Lowe SW, Schmitt EW, Smith SW, et al: P53 is required for radiation induced apoptosis in mouse thymocytes. Nature 362:847849, 1993 8. Clarke AR, Purdie CA, Harrison DJ, et al: Thyrnocyte apoptosis induced by p53-dependent and independent pathways. Nature 362:849-852, 1993 9. Shaw P, Bovey R, Tardy S, et al: Induction of apoptosis by wild-type p53 in a human colon tumor derived cell line. Proc Natl Acad Sci U S A 89:4495-4499, 1992 10. Wang Y, Szekely L, Okan I, et al: Wild-type p53-triggered apoptosis is inhibited by bcl-2 in a v-myc induced T-cell lymphoma line. Oncogene 8:342%3431, 1993 11. Lane DP: P53, the guardian of the genome. Nature 358:1516, 1992 12. Papathanasion M, Fornace AJ Jr: DNA damage inducible genes, in Ozols RF (ed) : Drug Resistance (ed 2). Boston, MA, Kluwer, 1991, pp 13-36 13. Livingstone LR, White A, SprouseJ, et al: Altered cell cycle

1300

P53 MEDIATED RESPONSE TO ANTICANCER THERAPY (Moll et al) arrest and gene amplification potential accompany loss of wild-type p53. Cell 70:923-935, 1992 14. ]fin Y, Tainsky MA, Bischoff FZ, et al: Wild-type p53 restores cell cycle control and inhibits gene amplification in cells with mutant p53 alleles. Cell 70:937-948, 1992 15. Moll UM, Riou G, Levine AJ: Two distinct mechanisms alter p53 in breast cancer: Mutation and nuclear exclusion. Proc Natl Acad Sci U S A 89:7262-7266, 1992 16. Sambrook J, Fritsch EF, Maniatis T: Molecular Cloning: A Laboratory Manual (ed 2, book 1, chap 7). Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press, 1989 17. Orita M, Suzuki Y, Sekiya T, et ah Rapid and sensitive detection of point mutations and DNA polymorphism using the polymerase chain reaction. Genomics 5:874-879, 1989 18. Davidoff AM, Humphrey PA, Iglehart, et al: Genetic basis for p53 overexpression in human breast cancer. Proc Natl Acad Sci U S A 88:5006-5010, 1991 19. Moll UM, LaQuaglia M, B~nardJ, et al: Wild-type p53 protein undergoes cytoplasmic sequestration in undifferentiated neuroblastomas but not in differentiated tumors. Proc Natl Acad Sci USA 92:44074411, 1995 20. Fritsche M, Haessler C, Brandner G: Induction of nuclear accumulation of the tumor suppressor protein p53 by DNA-damaging agents. Oncogene 8:307-318, 1993 21. Harris JR, Lippman ME, Veronesi U, et al: Breast cancer: Review in three parts. N Engl J Med I:327:319-328; II: 327:390-398; III: 327:473-480, 1992 22. Bonadonna G, Brusarnolino E, Valagussa P, et ah Combination chemotherapy as an adjuvant treatment in operable breast cancer. N EnglJ Med 294:405410, 1976 23. Fisher B, Carbone P, Economon SG, et al: 1-Phenylalanine mustard (L-PAM) in the management of primary breast cancer: A report of early findings. N EnglJ Med 292:117-122, 1975 24. Controlled trial of tamoxifen as a single adjuvant agent in the management of early breast cancer: Analysis at eight years by "Nolvadex" Adjuvant Trial Organization. Br J Cancer 57:608-611, 1988 25. Mauriac L, Durand M, Avril A, et al: Effects of primary chemotherapy in conservative treatment of breast cancer patients with operable tumors larger than 3 cm: Results of a randomized trial in a single center. Ann Oncol 2:347-354, 1991 26. Elledge RM, Allred DC: The p53 tumor suppressor gene in breast cancer. Breast Cancer Res Treat 32:39-47, 1994 27. Cattoretti G, Rilke F, Andreola S, et al: P53 Expression in breast cancer. I n t J Cancer 41:178-183, 1988 28. BartekJ, BartkovaJ, Vojtesek B, et al: Patterns of expression of the p53 tumor suppressor in human breast tissues and tumors in situ and in vitro. I n t J Cancer 46:839-844, 1990 29. VarleyJM, Brammar WJ, Lane DP, et al: Loss of chromosome 17p13 sequences and mutation of p53 in human breast carcinomas. Oncogene 6:413-421, 1991 30. Davidoff AM, Kerns BM, Pence JC, et al: P53 Alterations in all stages of breast cancer. J Surg Oncol 48:260-267, 1991 31. Borresen A-L, Hovig E, Smith-Sorensen B, et ah Constant denaturant gel electrophoresis as a rapid screening technique for p53 mutations. Proc Natl Acad Sci U S A 88:8405-8409, 1991 32. Osborne RJ, Merlo GR, Mitsudomi T, et al: Mutations in the p53 gene in primary human breast cancers. Cancer Res 51:6194-6198, 1991 33. Domagala W, Harezga BA, Szadowska A, et al: Nuclear p53 protein accumulates preferentially in medullary and high grade ductal but rarely in lobular breast carcinomas. Am J Pathol 142:669-674, 1993 34. Lipponen P, Ji H, Aaltomaa S, et al: P53 protein expression in breast cancer is related to histopathological characteristics and prognosis. I n t J Cancer 55:51-56, 1993 35. Martinazzi M, Crivelli F, Zampatti C, et al: Relationship between p53 expression and other prognostic factors in human breast carcinoma. A m J Clin Pathol 100:213-217, 1993 36. Thor AD, Moore DH II, Edgerton SM, et al: Accumulation of p53 tumor suppressor gene protein: An independent marker of prognosis in breast cancers. J Natl Cancer Inst 84:1109-1114, 1992 37. Riou G, Le M, Travagli JP, et ah Poor prognosis of p53

nuclear overexpression and mutation in inflammatory breast carcinoma. J Natl Cancer Inst 85:1765-1767, 1993 38. Friedrichs K, Gluba S, Eidtmann H, et al: Overexpression of p53 and prognosis in breast cancer. Cancer 72:3641-3647, 1993 39. Friend S: P53: A glimpse at a puppet behind a shadow play. Science 265:334-335, 1994 40. Shaulsky G, Goldfinger N, Peled A, et al: Involvement of wild-type p53 protein in the cell cycle requires nuclear localization. Cell Growth Differ 2:661-667, 1991 41. Shaulsky G, Goldfinger N, Ben-Ze'ev A, et al: Nuclear accumulation of p53 protein is mediated by several nuclear localization signals and plays a role in tumorigenesis. Mol Cell Biol 10:6565-6577, 1990 42. Shaulsky G, Goldfinger N, Tosky M, et al: Nuclear localization is essential for the activity of p53 protein. Oncogene 6:20552065, 1991 43. Zambetti GP, Levine AJ: A comparison of the biological activities of wild-type and mutant p53. FASEB J 7:855-865, 1993 44. Isola J, Visakorpi T, Holli K, et al: Association of overexpression of tumor suppressor protein p53 with rapid cell proliferation and poor prognosis in node-negative breast cancer patients. J Natl Cancer Inst 84:1109-1114, 1992 45. Allred DC, Clark GM, Elledge R, et al: Association of p53 protein expression with tumor cell proliferation rate and clinical outcome in node negative breast cancer. J Natl Cancer Inst 85:200206, 1993 46. Dittmer D, Pad S, Zambetti GP, et al: Gain of function mutations in p53. Nat Genet 4:42-46, 1993 47. Chin KV, Ueda K, Pastan I, et al: Modulation of activity of the promoter of the human MDR-1 gene by r a s and p53. Science 255:459462, 1992 48. Lowe SW, Bodis S, McClatchey A, et al: p53 Status and the efficacy of cancer therapy in vivo. Science 266:807-810, 1994 49. Tishler RB, Calderwood SK, Coleman CN, et al: Increase in sequence specific DNA binding by p53 following treatment with chemotherapeutic and DNA damaging agents. Cancer Res 53:22122216, 1993 50. Tannock IF: Tumor growth and cell kinetics, in Tannock IF, Hill RP (eds) : The Basic Science of Oncology. New York: Pergamon Press, 1987, p 140 51. Fan S, Smith ML, Rivet DJ I1, et al: Disx-uption of p53 function sensitizes breast cancer MCF-7 cells to cisplatin and pentoxyfylline. Cancer Res 55:1649-1654, 1995 52. Merlo GR, Basolo F, Fiore L, et al: P53-dependent and p53independent activation of apoptosis in mammary epithelial cells reveals a survival function of EGF and insulin..1 Cell Biol 128:11851196, 1995 53. McIlwrath AJ, Vasey PA, Ross G: Cell cycle arrest and radiosensitivity of human tumor cell lines: Dependence on wild-type p53 for radiosensitivity. Cancer Res 54:3718-3722, 1994 54. Koechli OR, Schaer GN, Seifert B, et al: Mutant p53 protein associated with chemosensitivity in breast cancer specimens. Lancet 344:1647-1648, 1994 55. Allred DC, Glark GM, Fuqua SAW, et ah Overexpression of p53 in node-positive breast cancer. Breast Cancer Res Treat 27:131, 1993 (abstr) 56. Makris A, Powles TJ, Dowsett M, et ah P53 protein overexpression and chemosensitivityin breast cancer. Lancet 345:1181-1182, 1995 57. Mathieu M-C, Koscielny S, Le Bihan M-L, et al: P53 protein overexpression and chemosensitivity in breast cancer. Lancet 345:1182, 1995 58. Heimdal K~ Lothe RA, Lystad S, et ah No germline TP 53 mutations detected in familial and bilateral testicular cancers. Genes Chromosom Cancer 6:92-97, 1993 59. Gaidano G, Ballerini P, Gong JZ, et al: p53 Mutations in human lymphoid malignancies: Association with Burkitt lymphoma and chronic lymphocytic leukemia. Proc Natl Acad Sci U S A 88:54135417, 1991 60. Felix CA, Nau MM, Takahashi T, et al: Hereditary and acquired p53 gene mutations in childhood acute lymphoblastic leukemia. J Clin Invest 89:640-647, 1992 61. Yeargin J, Cheng J, Yu AL, et al: P53 mutation in acute lymphoblastic leukemia is of somatic origin and is stable during establishment of T cell acute lymphoblastic leukemia cell lines. J Clin Invest 91:211-217, 1993

1301