This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schneider-Stock, R. Articles by Günther, T. Search for Related Content PubMed PubMed Citation Articles by Schneider-Stock, R. Articles by Günther, T. Social Bookmarking                            What's this?

High Prognostic Value of p16^INK4 Alterations in Gastrointestinal Stromal Tumors

Regine Schneider-Stock, Carsten Boltze, Jerzy Lasota, Markku Miettinen, Brigitte Peters, Matthias Pross, Albert Roessner, Thomas Günther

From the Department of Pathology, Department of Biometrics, Department of General Surgery, Otto-von-Guericke University, Magdeburg, Germany; and Department of Soft Tissue Pathology, Armed Forces Institute of Pathology, Washington, DC.

Address reprint requests to Regine Schneider-Stock, PhD, Department of Pathology, Otto-von-Guericke University, Leipziger Str 44, 39120 Magdeburg, Germany; email: regine.schneider-stock{at}medizin.uni-magdeburg.de.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Purpose: Gastrointestinal stromal tumors (GISTs) represent a distinctive (but histologically heterogeneous) group of neoplasms, the malignant potential of which is often uncertain. To determine the prognostic relevance of p16^INK4 alterations in GISTs, we investigated a larger group of GISTs and correlated the genetic findings with clinicopathological factors and patient survival.

Material and Methods: We evaluated the methylation status of the promotor by methylation-specific polymerase chain reaction (PCR), the presence of mutations by PCR-SSCP-sequencing, the loss of heterozygosity at the p16^INK4 locus (using the c5.1 marker), and the immunohistochemical expression of p16^INK4 protein in 43 GISTs in 39 patients.

Results: p16^INK4 alterations were found in 25 of 43 GISTs (58.1%), with benign, borderline, or malignant GISTs showing no differences in the type and frequency of alteration. p16^INK4 alterations were correlated with a loss of p16^INK4 protein expression (P < .01). Patients who had tumors with p16^INK4 alterations had a poorer prognosis than patients with tumors without such alterations (P = .02). There was a high predictive value for p16^INK4 alterations only in the group of benign and borderline GISTs (P < .01) with regard to clinical outcome. Univariate Cox’s proportional hazard regression analysis revealed a strong correlation between p16^INK4 alterations, tumor size, mitotic index, and overall survival (P < .02), whereas multivariate Cox’s analysis confirmed only p16^INK4 alterations as an independent prognostic factor.

Conclusion: We believe that the evaluation of p16^INK4 alteration status is a helpful prognosticator, particularly in the benign and borderline groups of GISTs.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
GASTROINTESTINAL STROMAL tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract. These tumors represent a distinctive (but histologically heterogeneous) group of neoplasms.^1,2 Immunohistochemically, GISTs are defined as KIT (CD117, stem cell factor receptor)-positive tumors.^3–5 Specific c-kit mutations leading to ligand-independent activation of KIT tyrosine kinase have been documented in GISTs.³ The majority of these mutations have been found in the juxtamembrane domain of the gene;^6,7 however, mutations in KIT extracellular and kinase domains have also been reported,^8–10 although the prognostic significance of c-kit mutations^6,7,11 is controversial.^10,12

It is often difficult to predict the malignant behavior of GISTs. Tumor stage at presentation, tumor size, and mitotic activity evaluated in the context of tumor location and its histological features constitute the most important clinicopathological markers. Small tumors with low mitotic activity have an excellent prognosis, whereas large tumors with a high mitotic rate usually show malignant behavior. However, a definite percentage of GISTs have uncertain malignant potential.¹³ High Ki-67 index, as well as high expression of BCL2^14,15 and c-Myc proteins,¹⁵ and telomerase activity¹⁶ are other markers, indicating a poorer prognosis for GISTs. Comparative genomic hybridization-based studies have shown that gains and high level amplifications at 5p and 20q, as well as losses at 9p, are highly specific for malignant and metastatic GISTs.¹⁷ Consistent losses of chromosome 9 have also been revealed in malignant GISTs by cytogenetic¹⁸ and interphase fluorescent in situ hybridization studies.¹⁹

A cyclin-dependent kinase (CDK) 4 inhibitor (p16^INK4) gene located at 9p21 has been shown to be inactivated in a variety of tumors^20–22 by homozygous deletions, point mutations,^23,24 or de novo methylation of its promoter region.^22,25 As a critical G1/S-cell cycle regulator, p16^INK4 is involved in the pathway that converges in the tumor suppressor protein Rb.²⁶

Little is known about p16^INK4 tumor suppressor gene alterations in GISTs, although homozygous deletion of p16^INK4 gene has recently been documented in two malignant tumors.²⁷ It might be of great practical value for GISTs to check additionally the prognostic relevance of p16^INK4 alterations. It is a generally recognized problem that, in contrast to other sarcomas, a final consensus on the grading of GISTs has not yet been reached, and the biologic behavior often remains unclear. Because this situation leads to several problems in the management of these tumors, it is helpful to add every new molecular marker that may shed light on the matter.

In this study, we evaluated the status of p16^INK4 tumor suppressor gene in 43 GISTs. We correlated the methylation status of the promotor, the presence of inactivating mutations, the loss of heterozygosity (LOH) at p16 locus, and the expression of p16^INK4 protein with clinicopathological parameters and established the prognostic significance of p16^INK4 alterations in GISTs.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Clinical and Morphological Features
Patients. The clinical and morphological data are summarized in Table 1. We analyzed 39 primary GISTs, three recurrent GISTs, and one metastatic GIST obtained from 39 patients. We studied both the primary tumor and the metastasis in one patient and the primary tumor and the corresponding recurrence in three patients. There were 25 males (64.1%) and 14 females (35.9%), and their ages ranged from 9 to 85 years (mean, 61.9 years; standard error of means, 13.4). Primary tumors originated in the stomach (n = 28), small intestine (n = 6), colon (n = 5), and mesenterium (n = 3); liver metastasis occurred in one case. Tumors were histologically classified as predominantly spindled (n = 27), epithelioid (n = 13), or mixed-spindled epithelioid (n = 3). Tumor size ranged from 0.4 to 30 cm in greatest dimension (mean 10.2 ± 6.5 cm). Tumors were retrieved from the files of the Armed Forces Institute of Pathology, Washington, DC, and the Department of Pathology, Otto-von-Guericke University, Magdeburg, Germany.

View larger version (115K): [in this window] [in a new window]   Fig 1. CD117 staining (A–C); Ki-67 (Mib1) (D–F); and p16INK4 (G–I) in benign (tumor 9), borderline (tumor 16), and malignant gastrointestinal stromal tumors (GISTs; tumor 36). CD117 was equally expressed. The brownish stained nuclei (D–F) show the highest proliferation rate (Ki-67) in the malignant GIST (F); and p16INK4 expression was higher in the nuclei and cytoplasm (red color) of the benign GIST (G).

There was a significant correlation between morphological and clinical classification categories (P = .025). There was disconcordance in five of 37 cases (13.3%). Three of these five patients (patients 2, 3, and 8) had morphologically benign tumors at surgery that showed clinically malignant behavior, and the patients died in less than 2 years. Two morphologically malignant tumors were still clinically benign (patients 33 and 37). These two patients were older than age 60 years at surgery, and they were followed up for 40 months. Eight of 11 tumors histologically classified as borderline GISTs showed clinically malignant behavior. Patients 16, 17, and 18 had borderline tumors at the morphological level and had follow-up of less than 24 months; these patients belonged in the prognostically indeterminate group. All three patients had developed large tumor masses (13, 18, and 16.5 cm, respectively) at primary surgery.

Immunohistochemistry
In general, for tumors less than 5 cm, we evaluated two sections for immunohistochemistry. For tumors ranging in size from 5 to 10 cm, we assessed five sections. For tumors larger than 10 cm, we checked 10 sections.

Immunohistochemically, we used the following antibodies for diagnostic evaluation: CD117 (KIT, WAK Chemie, Berlin, Germany), dilution 1:1,000, microwave pretreatment 450W, 2 x 10 minutes, EDTA, pH 8.4; CD34 (Biogenex, Hamburg, Germany), dilution 1:30, microwave pretreatment 450W, 2 x 8 minutes, EDTA, pH 8.4; -smooth muscle actin (Immunotech, Hamburg, Germany), dilution 1:20, no pretreatment; desmin (Immunotech, Hamburg, Germany), dilution 1:150, microwave pretreatment 450W, 2 x 10 minutes, EDTA, pH 8.4; S100 (DAKO, Hamburg, Germany), dilution 1:500, 14 minutes, and protease pretreatment (Ventana, Strasbourg, France).

All tumors were strongly positive for CD117, and 78% the cases showed coexpression of CD34 (29.6% were focally positive for alpha-smooth muscle actin and 5.1% were focally positive for S100 protein). All but five tumors were negative for desmin. Ki-67 staining (DAKO, Hamburg, Germany; dilution 1:50, microwave pretreatment 450W, [2 x 10 minutes], EDTA, pH 8.4) ranged from 0% to 4% in benign tumors, from 1% to 30% in borderline tumors, and from 1% to 19% in malignant tumors (primary tumors). The Mib1 labeling index equals the percentage of positively stained tumor cell nuclei. In the three recurrences and the only liver metastasis, the Mib1 index ranged from 5% to 60%.

For p16 immunohistochemistry, a monoclonal mouse antibody to p16 (1:100 dilution; Quartett, Hamburg, Germany) and antigen retrieval, using microwave heating (three times for 10 minutes; 10 mmol/L citrate buffer, pH 6.0), were used after inhibition of endogeneous peroxidase activity. The primary antibody was incubated for 1 hour at 37°C. The slides were subsequently incubated with a 1:10 dilution of normal swine serum (Vector Laboratories, Inc, Burlingame, CA). After washing in phosphate-buffered saline (pH = 7.4), the samples were incubated with a 1:200 dilution of biotinylated antigoat secondary antibody (Vector Laboratories) for 30 minutes at room temperature. The detection of bound antibody was accomplished using the avidin-biotin complex method (Dianova, UniTect A.B.C. System XHC1 Kit, Hamburg, Germany). A 5% solution of New-Fuchsine (red) was used as a chromogen. Specificity for immunostaining was checked by omitting single steps in the immunochemical protocol and by replacing primary antibody with nonimmune serum.

Estimating 10 HPFs, a section was considered to be immunohistochemically positive for p16^INK4 if tumor nuclei were stained (with or without cytoplasmic staining), according to a 4-point semiquantitative scale, as follows: no staining, 0% to 5% (0); weak staining, 6% to 20% (1); moderate staining, 21% to 50% (2), and strong staining, more than 50% (3). Nontumorous stromal cells showing nuclear reactivity served as an internal control.

Surgery and Follow-Up
The median follow-up time was 34 months (range, 8 to 204 months). Survival analysis was carried out in 34 patients (Table 1). All primary, recurrent, and metastatic tumors were treated solely by surgical resection. All cases were R0-resected (local or multivisceral) with histologically confirmed margins (one exception was case 18, which was R2-resected). There was no follow-up (<1 month) for three patients (patients 7, 15, and 27). Patient 39 was excluded because she received chemotherapy (imatinib mesylate) and was in complete remission. None of the other patients examined in this study received oncologic medicamentous therapy. Patient 5 was excluded from survival analysis because she died of colorectal cancer.

c-kit Mutations
The c-kit gene mutational status of exons 9, 11, 13, and 17 was evaluated by polymerase chain reaction (PCR) amplification of genomic DNA, using previously described procedures.^5,7

Evaluation of p16^INK4 Alterations
DNA from frozen tissue was extracted by the standard phenol-chloroform-proteinase K protocol.²⁸ DNA from formalin-fixed, paraffin-embedded tissue was prepared using the Nucleo Spin Tissue kit (Macherey and Nagel, Düren, Germany).

p16^INK4 Promotor Methylation
The methylation status of the p16^INK4 promotor was determined by methylation-specific PCR (MSP).²⁵ MSP requires an initial bisulfite reaction, using 200 to 500 ng genomic DNA (CpGenome DNA Modification Kit, QBiogene, Illkirch, France), in which all unmethylated cytosines are deaminated and converted to uracils, whereas 5-methylcytosines remain unaltered. Modified DNA was used as a template for MSP, using primers specific for either methylated or modified unmethylated sequences. CpGenome universal methylated DNA (QBiogene) was used as a positive control for methylated alleles. DNA from normal lymphocytes was used as a negative control for methylated alleles.

The primer sequences for the unmethylated reactions were sense 5'-TTA TTA GAG GGT GGG GTG GAT TGT-3' and antisense 5'-CAA CCC CAA ACC ACA ACC ATA A-3', which amplify a 151-base pair (bp) product. The primer sequences for the methylated reaction were sense 5'-TTA TTA GAG GGT GCG GAT CGC-3' and antisense 5'-GAC CCC GAA CCG CGA CCG TAA-3', which amplify a 150-bp product. The two sense primers were labeled with FAM dye.

PCR was performed, using one tenth of bisulfite reaction and a MasterAmp Optimization Kit (Biozym, Madison, WI) in an automated PTC 2000 thermocycler (Biozym). A 50-µL reaction contained 50 mmol/L TRIS/HCl (pH 8.3), 50 mmol/L KCl, 200 µmol/L each deoxynucleotide triphosphate (dNTP), 3 mmol/L MgCl₂, 4x MasterAmp PCR enhancer, 25 pmol each primer, and 0.2 units Amplitherm polymerase (Biozym). Reactions were hot-started at 95°C for 5 minutes, followed by 30 cycles of 95°C for 1 minute, 60°C (methylated PCR) or 58°C (unmethylated PCR) for 1 minute, and 72°C for 2 minutes, and they were finished by one single step at 72°C for 10 minutes. Reactions were analyzed on an ABI310 sequencer; injection time was between 20 and 30 seconds, using a Pop 4 dRhodamine Matrix standard ABI Prism (Perkin Elmer, Weiterstadt, Germany).

p16^INK4 Gene Mutations
Mutations were analyzed by PCR-single-stranded conformational polymorphism sequencing. Exons 1 and 2 of the p16^INK4 gene were amplified using the following primers: exon 1, sense 5'-CGG AGA GGG GGA GAG CAG GCA G-3' and antisense 5'-GCT ACC TGA TTC CAA TTC CCC TG-3', amplifying a 270-bp product; and exon 2, sense 5'-ACC CTG GCT CTG ACC ATT CTG TTC-3' and antisense 5'-CCG GGC TGA ACT TTC TGT GCT-3', amplifying a 370-bp fragment. Both sense primers were biotinylated. PCR was performed using a MasterAmp Optimization Kit (Biozym). A 50-µL PCR mix contained 50 mmol/L TRIS-HCL (pH 8.3), 50 mmol/L KCl, 200 µmol/L each dNTP, 3.5 mmol/L (exon 1) or 2.5 mmol/L (exon 2) MgCl₂, 4x (exon 1) or 6x (exon 2) MasterAmp PCR enhancer, 25 pmol each primer, and 0.2 units Amplitherm polymerase (Biozym). A PTC 200 automated thermocycler was used for amplification (Biozym).

For SSCP analysis, exon 2 fragments were digested with SmaI, yielding products less than 250 bp (227 bp and 143 bp). PCR products were diluted at a ratio of 1:1 with a solution containing 100% formamide, 0.05% xylene cyanol, and 0.05% bromophenol; the solution was denatured at 96°C for 7 minutes and chilled on ice. PCR fragments were run at 4°C on ultrathin nondenaturing mutation detection enhancement gels (AT Biochem, Malvern, PA) for exon 1 and on T10C1 gels (Amersham Biosciences, Uppsala, Sweden) for exon 2, followed by silver staining, according to Goldman and Merril.²⁹ PCR products showing altered banding patterns of single strands were directly sequenced on an automated ALF Express fluorescence sequencer (Amersham Biosciences), using a T7-sequencing kit (Amersham Biosciences). Sequence primers used for exon 1 were 5'-GCC ATC CCC TGC TCC CGC TGC-3'; sequence primers used for exon 2 were 5'-GAA TGC TCT GAG CTT TGG AAG-3'. Deletions were confirmed by PCR amplification of a 171–bp fragment of the housekeeping gene phenylalaninhydroxylase (PAH), using the sense primer, 5'-CCA TGC CAC TGA GAA CTC TCT-3' and the antisense primer, 5'-TCT TAA GCT GCT GGG TAT TGT C-3'.

p16^INK4 Loss of Heterozygosity
For LOH analysis, the sequence-tagged site marker c5.1 (localized between coding regions of exon 2 and exon 3) was investigated, as previously described.³⁰

Statistical Analysis
Statistical analyses were carried out using the ² test or Fisher’s exact test in cross tables, and one-way analysis of variance (ANOVA) for comparison of means was used to assess the relationship between p16^INK4 alterations and clinicopathological factors. All statistical tests were two-sided. To describe the correlation between p16^INK4 alterations and p16^INK4 protein expression, the regression function was assessed. Total survival curves were drawn according to the Kaplan-Meier method and compared by using the log-rank test. Multivariate analysis was based on Cox’s regression analysis. A P value of less than .05 was considered statistically significant. Calculations were carried out by means of SPSS-9.0 software (SPSS, Chicago, IL).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Table 1 provides an overview of the 43 GISTs investigated, including selected clinical data of the patients, histopathological diagnoses, and molecular genetic findings.

p16^INK4 Alterations: Frequency and Alteration Type
The p16^INK4 promotor was methylated in 13 of 40 cases (32.5%) (Fig 2). In three cases, the sparse amount of DNA did not allow us to perform MSP. p16^INK4 gene mutations were identified in nine of 43 cases (20.9%) (Table 2). There were three cases showing sense mutations, with two tumors (tumors 3 and 18A) showing the same base substitution in codon 6. Four other mutations were missense mutations (tumors 7, 8, 32, and 39). Tumor 18 and the corresponding recurrence showed T-insertion, causing a reading frame shift that created an early stop codon and that subsequently confirmed a truncated p16^INK4 protein at the immunohistochemical level. In one sample (tumor 21) showing a clear band shift in exon 1 SSCP-gel, we failed to sequence the PCR product successfully. All patients that had tumors with p16^INK4 mutations died of disease, one patient was lost for follow-up, and one patient underwent treatment therapy. p16^INK4 mutations were accompanied by simultaneous promotor methylation in two GISTs and by LOH in four cases (Table 3).

View larger version (15K): [in this window] [in a new window]   Fig 2. Analysis of p16INK4 promotor methylation on ABI310 Prism. Methylated (left arrow) and unmethylated (right arrow) signals can be detected between the 150-bp and the 160-bp matrix standards. (A) Unmethylated gastrointestinal stromal tumor (GIST; tumor 10); (B) highly methylated GIST (tumor 3); and (C) low methylated GIST (tumor 6).

Correlation of p16^INK4 Protein Expression and p16^INK4 Alteration Status
We observed loss of p16^INK4 protein expression in 27 of 43 GISTs (62.8%). In all 43 GISTs, p16^INK4 alterations (methylation, mutation, and LOH) correlated with the absence or lowering of protein expression (P < .01) (Table 1, Fig 1). Of 25 tumors harboring p16^INK4 methylation, mutation, deletion, or LOH, only one tumor (tumor 6, benign with low p16^INK4 methylation) showed highly positive staining for p16. Five cases showed p16^INK4 protein only in a small number of tumor cells (tumors 7, 16, 28, 35, and 38). Of 18 GISTs without p16^INK4 alterations, 15 GISTs (83.3%) contained p16^INK4-immunoreactive tumor cells and three GISTS were p16^INK4-immunonegative (tumors 5, 14A, and 29).

p16^INK4 Alteration: Clinicopathological Factors
There were no significant correlations between p16^INK4 alterations and tumor size, mitotic index, sex, age, localization, histological subtype, and Mib1 proliferation index, although there was a tendency for tumors with p16^INK4 alterations to be larger, to show a higher mitotic index (Fig 1) and to more frequently display spindle histology (Table 4).

Patients who had tumors with p16^INK4 alterations had a poorer prognosis than those who had tumors without p16^INK4 alterations. After 5 years, patients who had tumors with and without p16^INK4 alterations had a total survival of 9% and 75%, respectively (P = .02) (Fig 3). Univariate Cox’s regression analysis revealed a strong correlation between p16^INK4 alterations, tumor size, mitotic index, and overall survival (Table 5). Patients who had tumors with p16^INK4 alterations had a five-fold increased risk of dying of disease. Multivariate Cox’s regression analysis confirmed only p16^INK4 alterations as an independent prognostic factor (P < .01).

View larger version (18K): [in this window] [in a new window]   Fig 3. Kaplan-Meier curve for patients with primary GISTs regarding p16INK4 alterations. Patients having tumors without p16INK4 alterations (n = 17) have a better prognosis than patients having tumors with p16INK4 alterations (n = 17).

Predictive Value of p16^INK4 Alterations in the Morphological Groups of Gastrointestinal Stromal Tumors
Analyzing the two groups separately according to their morphological classification, we found differences in the prognostic relevance for p16^INK4 alterations. In the malignant group, there was no significant correlation with overall survival (P = .45). In the group of borderline tumors, six of seven patients who had tumors with p16^INK4 alterations died of disease, whereas five of six patients who had tumors without p16^INK4 alterations were still alive (P = .024). In the benign group of GISTs, all six patients who had tumors without p16^INK4 alterations were still alive, and three of four patients who had tumors with p16^INK4 alterations died of disease (P = .028). Considering the benign and the borderline GISTs together (which corresponds to the group of GISTs with mitotic index 10 mitoses per HPF), there was a high predictive value for p16^INK4 alterations on the clinical outcome of the patients (Fig 4; P < .01). Thus, p16^INK4 alterations seem to be useful for defining a subgroup with unfavorable prognosis in the benign and borderline groups with GISTs.

View larger version (18K): [in this window] [in a new window]   Fig 4. Kaplan-Meier curve for patients with primary benign and borderline gastrointestinal stromal tumors regarding p16INK4 alterations. Patients having tumors with p16INK4 alterations (n = 10) have a significantly poorer prognosis than patients having tumors without p16INK4 alterations (n = 13).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
To the best of our knowledge, this is the first large-scale study investigating p16^INK4 alterations in GISTs. We found p16^INK4 alterations in more than one half of tumors, with benign, borderline, and malignant GISTs showing no differences in the frequency and type of alteration. Of interest, p16^INK4 alterations provide additional prognostic information about GIST patients, particularly in the group of benign and borderline cases.

Frequency and Alteration Type
p16^INK4 alterations were found in 25 of 43 (58.1%) GISTs. With a few exceptions, p16^INK4 alterations were accompanied by p16^INK4 protein loss. It is noteworthy that for six of eight identified mutations and for both partial deletions, the wild-type p16^INK4 allele was inactivated by methylation in two tumors and by deletion (LOH) in six cases, which supports the hypothesis that these mutations are of functional relevance. Accordingly, in these cases, complete loss of p16^INK4 protein was observed, whereas we found a small number of p16^INK4-immunoreactive cells in tumors with low p16^INK4 promotor methylation. With regard to the three p16^INK4-immunonegative GISTs without p16^INK4 alterations, we think that deletions exist that do not involve exons 1 and 2 or that point mutations in noncoding regions possibly affect protein expression. Furthermore, p16^INK4 might be downregulated by transcriptional or post-transcriptional mechanisms.

GISTs without p16^INK4 downregulation may harbor alterations of other genes involved in the p16-CDK4-cyclin D-pRb-mediated cell cycle checkpoint. However, with regard to GISTs, we can only speculate about such alterations because only a few studies have investigated cell cycle regulator genes in this tumor entity. Although Cunningham et al¹⁴ and Panizo-Santos et al¹⁵ reported a high rate of BCL2 protein expression in GISTs, this factor was without prognostic relevance. Immunohistochemical studies revealed a low frequency of p53-positive GISTs.^14–16 Our recent findings of a high frequency of telomerase-positivity in GISTs³¹ and the herein reported high frequency of p16^INK4 alterations in GISTs seem to be in accordance with the findings of Serrano et al,²⁶ who reported that p16^INK4 loss facilitates immortalization in many cellular systems. Investigating breast tumors, Landberg et al³² reported the highest telomerase activity in tumors with low expression of the p16^INK4 gene, suggesting that downregulation of p16^INK4 expression, in combination with other cell cycle defects, might contribute to maximum telomerase activity.

p16^INK4 blocks progression through the cell cycle by binding to either CDK4 or CDK6, thus inhibiting the action of cyclin D.²³ The major function of cyclin D is to drive the cell cycle forward by binding to CDKs and forming a catalytically active complex that phosphorylates the pRb protein, which results in the release of E2F and new transcription of important cell cycle genes.³³ Thus, p16^INK4 is considered a tumor suppressor gene, and its major biochemical effect is to halt cell cycle progression at the G1/S boundary.²⁶ Loss of p16 function may lead to cancer progression by unregulated cellular proliferation. With regard to GISTs, we believe that p16^INK4 alterations are not responsible for proliferation dysregulation because we found that p16^INK4 alterations occurred in benign and malignant tumors with nearly the same frequency. Irrespective of this observation, malignant tumors showed a higher Mib1 proliferation index than did benign GISTs, possibly because of an accumulation of additional genetic alterations.

In total, there were no differences in frequency or p16^INK4 alteration type between benign, borderline, and malignant GISTs. Because we also found p16^INK4 alterations in benign GISTs with high frequency, we think that p16^INK4 alterations occur early in the development of this tumor. The fact that p16^INK4 alterations are associated with tumor progression suggests that they maintain the GIST phenotype. Indeed, all GISTs with 100 mitoses per 50 HPF showed p16^INK4 inactivation. The tendency of p16^INK4 LOH to occur more frequently in malignant than in benign GISTs seems to be in accordance with the CGH study of El-Rifai et al,¹⁷ who found frequent losses in 9p, particularly in malignant GISTs. Moreover, the authors reported that losses in chromosome 9p occur more frequently in metastatic than in nonmetastatic GISTs, suggesting a correlation with an aggressive course of the disease.

Although we observed a high frequency of p16^INK4 alterations, our finding is not unique for GISTs. Such changes have also been reported for several other tumors; nevertheless, the type of p16^INK4 inactivation varied markedly among different tumor types (p16^INK4 LOH in esophageal squamous cell carcinomas and pancreatic ductal adenocarcinomas;³⁴ no mutations and no promotor methylation in prostate cancer;³⁵ no gene mutations, but high frequency of promotor methylation in colon cancer,³⁶ cervical cancer,³⁷ and gastrinomas;³⁸ and p16^INK4 deletions, but no promotor methylation, in squamous cell carcinoma of the bladder³⁹). The special thing about GISTs is that there seems to be no preference for one of the genetic alterations: all types of p16^INK4 inactivation were observed with nearly the same frequency,

Molecular Prognostic Factors in Gastrointestinal Stromal Tumors
Using Cox’s regression model, we were the first to demonstrate a high prognostic significance for p16^INK4 alterations in GISTs. Patients who had tumors with p16^INK4 alterations show a poor clinical outcome and they have a five-fold increased risk of dying. P16^INK4 alterations were confirmed as an independent prognostic factor for GISTs. In particular, in patients with benign and borderline tumors, both p16^INK4 alterations and the classical morphological criteria seem to be useful predictors for clinical outcome. Because the genetic analyses are time and labor intensive, our data allow us to recommend the sole estimation of the p16^INK4 protein expression status in benign and borderline GISTs as prognostically practicable in clinical settings.

p16^INK4 alterations are also of prognostic relevance in other tumors. In general, the presence of p16^INK4 hypermethylation seems to predict shorter survival in lung cancer⁴⁰ and colorectal cancer.⁴¹ Only a few studies have demonstrated the prognostic value of molecular factors in GISTs. El-Rifai et al¹⁷ showed that gains at 8q, 17q, and 20q, as well as losses at 1p, 9p, and 13q, are associated with malignancy in GISTs. Furthermore, the survival rate of patients with telomerase-positive GISTs was significantly lower than that of patients with telomerase-negative GISTs.¹⁶ In the literature, multivariate analyses yielded extremely divergent results. According to some studies, telomerase activity,¹⁶ mitotic index,¹⁴ c-Myc, tumor size, and Mib1 index¹⁵ can be handled as prognostic determinants. In our study, the Mib1 proliferation index reached borderline significance in univariate Cox’s analysis, but it was omitted in the subsequent multivariate calculation. p16^INK4 alteration was found to be the strongest variable.

In summary, we suggest that p16^INK4 alterations play an important role in GISTs. The determination of the p16^INK4 alteration status, particularly of the immunohistochemical protein expression status, might be a helpful prognosticator, especially for the borderline and benign groups of GISTs, where the classical histological evaluation comes up against limiting factors.

	NOTES

 
Supported in part by Mildred-Scheel Foundation grant no. 10-1501.

We thank Hiltraud Scharfenort and Antje Schinlauer for their expert technical assistance.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Miettinen M, Lasota J: Gastrointestinal stromal tumors-definition, clinical, histological, immunohistochemical, and molecular genetic features and differential diagnosis. Virchows Arch 438:1–12, 2001[CrossRef][Medline]

2. Fletcher CD, Berman JJ, Corless C, et al: Diagnosis of gastrointestinal stromal tumors: A consensus approach. Hum Pathol 33:459–465, 2002[CrossRef][Medline]

3. Hirota S, Isozaki K, Moriyama Y, et al: Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279:577–580, 1998[Abstract/Free Full Text]

4. Kindblom LG, Remotti HE, Aldenborg F, et al: Gastrointestinal pacemaker cell tumor (GIPACT): Gastrointestinal stromal tumors show phenotypic characteristics of the interstitial cell of Cajal. Am J Pathol 152:1259–1269, 1998[Abstract]

5. Miettinen M, Sobin LH, Sarlomo-Rikala M: Immunohistochemical spectrum of GISTs at different sites and their differential diagnosis with a reference to CD117 (KIT). Mod Pathol 13:1134–1142, 2000[CrossRef][Medline]

6. Ernst SI, Hubbs AE, Przygodzki RM, et al: KIT mutation portends poor prognosis in gastrointestinal stromal/smooth muscle tumors. Lab Invest 78:1633–1636, 1998[Medline]

7. Lasota J, Jasinski M, Sarlomo-Rikala M, et al: Mutations in exon 11 of c-Kit occur preferentially in malignant versus benign gastrointestinal stromal tumors and do not occur in leiomyomas or leiomyosarcomas. Am J Pathol 154:53–60, 1999[Abstract/Free Full Text]

8. Lux ML, Rubin BP, Biase TL, et al: KIT extracellular and kinase domain mutations in gastrointestinal stromal tumors. Am J Pathol 156:791–795, 2000[Abstract/Free Full Text]

9. Lasota J, Wozniak A, Sarlomo-Rikala M, et al: Mutations in exons 9 and 13 of KIT gene are rare events in gastrointestinal stromal tumors. Am J Pathol 157:1091–1095, 2000[Abstract/Free Full Text]

10. Rubin BP, Singer S, Tsao C, et al: KIT activation is a ubiquitous feature of gastrointestinal stromal tumors. Cancer Res 61:8118–8121, 2001[Abstract/Free Full Text]

11. Taniguchi M, Nishida T, Hirota S, et al: Effect of c-kit mutation on prognosis of gastrointestinal stromal tumors. Cancer Res 59:4297–4300, 1999[Abstract/Free Full Text]

12. Corless CL, McGreevey L, Haley A, et al: KIT mutations are common in incidental gastrointestinal stromal tumors one centimeter or less in size. Am J Pathol 160:1567–1572, 2002[Abstract/Free Full Text]

13. Miettinen M, El-Rifai W, Sobin LH, et al: Evaluation of malignancy and prognosis of gastrointestinal stromal tumors: A review. Hum Pathol 33:478–483, 2002[CrossRef][Medline]

14. Cunningham RE, Abbondanzo SL, Chu WS, et al: Apoptosis, bcl-2 expression, and p53 expression in gastrointestinal stromal/smooth muscle tumors. Appl Immunohistochem Mol Morphol 9:19–23, 2001[Medline]

15. Panizo-Santos A, Sola I, Vega F, et al: Predicting metastatic risk of gastrointestinal stromal tumors: Role of cell proliferation and cell cycle regulatory proteins. Int J Surg Pathol 8:133–144, 2000[Abstract/Free Full Text]

16. Sakurai S, Fukayama M, Kaizaki Y, et al: Telomerase activity in gastrointestinal stromal tumors. Cancer 83:2060–2066, 1998[CrossRef][Medline]

17. El-Rifai W, Sarlomo-Rikala M, Andersson LC: DNA sequence copy number changes in gastrointestinal stromal tumors: Tumor progression and prognostic significance. Cancer Res 60:3899–3903, 2000[Abstract/Free Full Text]

18. Gunawan B, Bergmann F, Hoer J, et al: Biological and clinical significance of cytogenetic abnormalities in low-risk and high-risk gastrointestinal stromal tumors. Hum Pathol 33:316–321, 2002[CrossRef][Medline]

19. Debiec-Rychter M, Lasota J, Sarlomo-Rikala M, et al: Chromosomal aberrations in malignant gastrointestinal stromal tumors: Correlation with c-KIT gene mutation. Cancer Genet Cytogenet 128:24–30, 2001[CrossRef][Medline]

20. Esteller M, Herman JG: Cancer as an epigenetic disease: DNA methylation and chromatin alterations in human tumours. J Pathol 196:1–7, 2002[CrossRef][Medline]

21. Esteller M, Corn PG, Maylin SB, et al: A gene hypermethylation profile of human cancer. Cancer Res 61:3225–3229, 2001[Abstract/Free Full Text]

22. Merlo A, Herman JG, Mao L, et al: 5' CpG island methylation is associated with transcriptional silencing of the tumour suppressor p16/CDKN2/MTS1 in human cancers. Nat Med 1:686–692, 1995[CrossRef][Medline]

23. Kamb A, Gruis NA, Weaver-Feldhaus J, et al: A cell cycle regulator potentially involved in genesis of many cancer types. Science 264:436–440, 1994[Abstract/Free Full Text]

24. Okamoto A, Demetrick DJ, Spillare EA, et al: Mutations and altered expression of p16^INK4 in human cancer. Proc Natl Acad Sci U S A 91:11045–11049, 1994[Abstract/Free Full Text]

25. Herman JG, Graff JR, Myöhänen S, et al: Methylation-specific PCR: A novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci U S A 93:9821–9826, 1996[Abstract/Free Full Text]

26. Serrano M: The tumor suppressor protein p16^INK4a. Exp Cell Res 237:7–13, 1997[CrossRef][Medline]

27. Kim NG, Kim JJ, Ahn JY, et al: Putative chromosomal deletions on 9p, 9q and 22q occur preferentially in malignant gastrointestinal stromal tumors. Int J Cancer 85:633–638, 2000[CrossRef][Medline]

28. Sambrook J, Fritsch EF, Maniatis T: Molecular cloning: A Laboratory Manual (ed 4). Cold Spring Harbor, NY, Cold Spring Harbor Laboratory, 1989

29. Goldman D, Merril CR: Silver staining of DNA in polyacrylamide gels: Linearity and effect of fragment size. Electrophoresis 3:24–26, 1982[CrossRef]

30. Schneider-Stock R, Radig K, Roessner A: Loss of heterozygosity on chromosome 9q21 (p16 gene) uncommon in soft-tissue sarcomas. Mol Carcinog 18:63–65, 1997[CrossRef][Medline]

31. Gunther T, Schneider-Stock R, Hackel C, et al: Telomerase activity and expression of hTRT and hTR in gastrointestinal stromal tumors in comparison with extragastrointestinal sarcomas. Clin Cancer Res 6:1811–1818, 2000[Abstract/Free Full Text]

32. Landberg G, Nielsen NH, Nilsson P, et al: Telomerase activity is associated with cell cycle dysregulation in human breast cancer. Cancer Res 57:549–554, 1997[Abstract/Free Full Text]

33. Sherr CJ: The INK4a/ARF network in tumour suppression. Nat Rev Mol Cell Biol 2:731–737, 2001[CrossRef][Medline]

34. Liu Q, Yan YX, McClure M, et al: MTS-1 (CDKN2) tumor suppressor gene deletions are a frequent event in esophagus squamous cancer and pancreatic adenocarcinoma cell lines. Oncogene 10:619–622, 1995[Medline]

35. Maruyama R, Toyooka S, Toyooka KO, et al: Aberrant promotor methylation profile of prostate cancers and its relationship to clinico-pathological features. Clin Cancer Res 8:514–519, 2002[Abstract/Free Full Text]

36. Burri N, Shaw P, Bouzourene H, et al: Methylation silencing and mutations of the p14ARF and p16INK4a genes in colon cancer. Lab Invest 81:217–229, 2001[Medline]

37. Wong YF, Chung TK, Cheung TH, et al: Methylation of p16INK4a in primary gynecologic malignancies. Cancer Lett 136:231–235, 1999[CrossRef][Medline]

38. Serrano J, Goebel SU, Peghini PL, et al: Alterations in the p16INH4a/CDKN2A tumor suppressor gene in gastrinomas. J Clin Endocrinol Metab 85:4146–4156, 2000[Abstract/Free Full Text]

39. Tsutsumi M, Tsai YC, Gonzalgo ML, et al: Early acquisition of homozygous deletions of p16/p19 during squamous cell carcinogenesis and genetic mosaicism in bladder cancer. Oncogene 17:3021–3027, 1998[CrossRef][Medline]

40. Seike M, Gemma A, Hosoya Y, et al: Increase in the frequency of p16INK4 gene inactivation by hypermethylation in lung cancer during the process of metastasis and its relation to the status of p53. Clin Cancer Res 6:4307–4313, 2000[Abstract/Free Full Text]

41. Esteller M, Gonzalez S, Risques RA, et al: K-ras and p16 aberrations confer poor prognosis in human colorectal cancer. J Clin Oncol 19:299–304, 2001[Abstract/Free Full Text]

Submitted August 15, 2002; accepted February 6, 2003.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				K. Kikuta, M. Gotoh, T. Kanda, N. Tochigi, T. Shimoda, T. Hasegawa, H. Katai, Y. Shimada, Y. Suehara, A. Kawai, et al. Pfetin as a Prognostic Biomarker in Gastrointestinal Stromal Tumor: Novel Monoclonal Antibody and External Validation Study in Multiple Clinical Facilities Jpn. J. Clin. Oncol., October 7, 2009; (2009) hyp125v1. [Abstract] [Full Text] [PDF]

				S. Romeo, M. Debiec-Rychter, M. Van Glabbeke, H. Van Paassen, P. Comite, R. Van Eijk, J. Oosting, J. Verweij, P. Terrier, U. Schneider, et al. Cell Cycle/Apoptosis Molecule Expression Correlates with Imatinib Response in Patients with Advanced Gastrointestinal Stromal Tumors Clin. Cancer Res., June 15, 2009; 15(12): 4191 - 4198. [Abstract] [Full Text] [PDF]

				S. Sulong, A. V. Moorman, J. A. E. Irving, J. C. Strefford, Z. J. Konn, M. C. Case, L. Minto, K. E. Barber, H. Parker, S. L. Wright, et al. A comprehensive analysis of the CDKN2A gene in childhood acute lymphoblastic leukemia reveals genomic deletion, copy number neutral loss of heterozygosity, and association with specific cytogenetic subgroups Blood, January 1, 2009; 113(1): 100 - 107. [Abstract] [Full Text] [PDF]

				Y. Suehara, T. Kondo, K. Seki, T. Shibata, K. Fujii, M. Gotoh, T. Hasegawa, Y. Shimada, M. Sasako, T. Shimoda, et al. Pfetin as a Prognostic Biomarker of Gastrointestinal Stromal Tumors Revealed by Proteomics Clin. Cancer Res., March 15, 2008; 14(6): 1707 - 1717. [Abstract] [Full Text] [PDF]

				L. A. Meza-Zepeda, S. H. Kresse, A. H. Barragan-Polania, B. Bjerkehagen, H. O. Ohnstad, H. M. Namlos, J. Wang, B. E. Kristiansen, and O. Myklebost Array Comparative Genomic Hybridization Reveals Distinct DNA Copy Number Differences between Gastrointestinal Stromal Tumors and Leiomyosarcomas. Cancer Res., September 15, 2006; 66(18): 8984 - 8993. [Abstract] [Full Text] [PDF]

				L Tornillo and L M Terracciano An update on molecular genetics of gastrointestinal stromal tumours. J. Clin. Pathol., June 1, 2006; 59(6): 557 - 563. [Abstract] [Full Text] [PDF]

				F. Haller, B. Gunawan, A. von Heydebreck, S. Schwager, H.-J. Schulten, J. Wolf-Salgo, C. Langer, G. Ramadori, H. Sultmann, and L. Fuzesi Prognostic Role of E2F1 and Members of the CDKN2A Network in Gastrointestinal Stromal Tumors Clin. Cancer Res., September 15, 2005; 11(18): 6589 - 6597. [Abstract] [Full Text] [PDF]

				J. Martin, A. Poveda, A. Llombart-Bosch, R. Ramos, J. A. Lopez-Guerrero, J. G. del Muro, J. Maurel, S. Calabuig, A. Gutierrez, J. L. G. de Sande, et al. Deletions Affecting Codons 557-558 of the c-KIT Gene Indicate a Poor Prognosis in Patients With Completely Resected Gastrointestinal Stromal Tumors: A Study by the Spanish Group for Sarcoma Research (GEIS) J. Clin. Oncol., September 1, 2005; 23(25): 6190 - 6198. [Abstract] [Full Text] [PDF]

				R. Schneider-Stock, C. Boltze, J. Lasota, B. Peters, C. L. Corless, P. Ruemmele, L. Terracciano, M. Pross, L. Insabato, D. Di Vizio, et al. Loss of p16 Protein Defines High-Risk Patients with Gastrointestinal Stromal Tumors: A Tissue Microarray Study Clin. Cancer Res., January 15, 2005; 11(2): 638 - 645. [Abstract] [Full Text] [PDF]

				C. L. Corless, J. A. Fletcher, and M. C. Heinrich Biology of Gastrointestinal Stromal Tumors J. Clin. Oncol., September 15, 2004; 22(18): 3813 - 3825. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Schneider-Stock, R. Articles by Günther, T. Search for Related Content PubMed PubMed Citation Articles by Schneider-Stock, R. Articles by Günther, T. Social Bookmarking                            What's this?