卵巢癌及宫颈癌中17p13.3的杂合性丢失

目的探讨染色体１７ｐ１３．３的杂合性丢失（ＬＯＨ）与卵巢癌、宫颈癌发生及发展之间的相关性。方法采用ＰＹＮＺ．２２探针做Ｓｏｕｔｈｅｒｎ印迹技术,检测２４例卵巢癌、９例宫颈癌及１３例妇科非癌患者手术切除组织染色体１７ｐ１３．３的ＬＯＨ。结果１２例卵巢癌（包括１例交界性粘液性囊腺癌）和４例宫颈癌发生１７ｐ１３．３的ＬＯＨ,丢失频率分别为５０．０％和４４．４％。１３例非癌组织中,仅１例（７．７％）发生丢失,该例经病理证实为宫颈上皮内瘤变Ⅲ级,属癌前期病变（Ｐ＜０．０１）。结论染色体１７ｐ１３．３的ＬＯＨ可能与宫颈癌和卵巢癌的发生相关,检测１７ｐ１３．３的杂合性丢失将有助于深入了解卵巢癌和宫颈癌发生及发展的分子基础。     

人卵巢癌3号染色体短臂杂合性丢失的研究

目的探讨3号染色体短臂(3p14)等位基因杂合性丢失(10ss 0f heterozygosity,LOH)与人卵巢癌发生及发展之间的相关性研究.方法采用聚合酶链反应并结合二核苷酸重复序列多态性方法,分别对31例卵巢癌及24例卵巢良性肿瘤患者的组织标本DNA中3p14上3个微卫星位点(D3S1234、D3S1300、D3S1312)杂合性丢失(LOH)进行检测,同时还随机地检测31例卵巢癌中21例患者的血清DNA中3p14的LOH.结果31例卵巢癌组织DNA中21例(67.7%)至少在1个微卫星位点中出现杂合性丢失,11例卵巢癌患者(35.5%)有2个以上微卫星位点出现LOH.癌组织DNA中基因杂合性丢失频率与癌细胞分化程度呈正相关,与肿瘤病理类型及FIGO分期无关.24例卵巢良性肿瘤组织及血清DNA中均未出现3p位点上的杂合性丢失.21例卵巢癌血清DNA与肿瘤组织DNA 3p14基因3个微卫星位点杂合性丢失率之间存在明显的相关性(P＜0 05).结论鉴于人卵巢癌3p14出现杂合性丢失率与其癌细胞分化恶性程度相关以及卵巢癌患者血清DNA与癌组织DNA中3p14出现LOH相一致,故本文实验结果提示3p14 LOH的检测可能具有对卵巢癌患者的临床诊断潜在性价值.     

Comparison of chromosome 3p deletions between cervical precancers synchronous with and without invasive cancer

Cervical cancers are considered to originate from a series of pre-malignant lesions (cervical intra-epithelial neoplasia, CIN). The mechanisms behind these events are unknown. In addition to HPV infection, deletions of chromosome 3p have been found to be a frequent event in cervical cancer and likely play an important role in the transition of CIN to invasive cancer. To classify the potential role of 3p deletions in early-stage cervical carcinogenesis, we analyzed LOH of 3p in cervical precancers. Thirty cases with single or multiple CIN lesions were selected for the study, including 20 cases without and 10 cases with synchronous invasive cancers. Allelic losses on 1 or more 3p loci were recorded in 33% (3/9) of CIN II and 36% (5/14) of CIN III lesions from 20 cases without co-existing invasive cancer, whereas an increasing percentage of LOH was observed in the 10 precancerous lesions synchronous with invasive cancer, with 71% (5/7) CIN II and 76% (13/17) CIN III lesions. This result implies that 3p deletions have selective roles in early transition of pre-malignancy to invasive cancer. Comparing the LOH patterns between the 2 groups, genetic deletions in cases with invasive cancers involved extensive regions of 3p but were more localized in precancer cases without concomitant invasive cancer. Two interstitial regions, 3p22-21.3 around marker D3S1260 and 3p21.1 around markers D3S1289 and D3S1076, were most frequently deleted in both groups, suggesting that these 2 regions are novel tumor-suppressor loci which may play a role in early transition of cervical precancer to invasive cancer. Identical LOH patterns between multiple CIN lesions and synchronous invasive cancer in the same case suggests that different cervical precancers and invasive cancer are genetically linked and most likely originate from a single precursor cell.     

Detailed genetic and physical mapping of tumor suppressor loci on chromosome 3p in ovarian cancer.

Hemizygosity and homozygosity mapping studies show that many common sporadic cancers including lung, breast, kidney, cervical, ovarian, and head and neck cancer display deletions on the short arm of chromosome 3. For ovarian cancer, monochromosomal transfer suppression studies have identified three candidate regions for chromosome 3p ovarian cancer tumor suppressor genes (OCTSGs). To accurately map OCTSG candidate regions, we analyzed 70 ovarian tumors for loss of heterozygosity (LOH) at 20 loci on chromosome 3p that were selected to target those regions proposed to contain tumor suppressor genes for common sporadic cancers. All samples were informative for at least five markers. In 33 (52%) tumors without microsatellite instability, LOH was observed for at least one 3p marker. Analysis of 27 ovarian tumors demonstrating both loss and retention of 3p markers enabled us to define four nonoverlapping minimal deletion regions (OCLOHRs): (a) OCLOHR-1 mapped distal to D3S3591 at 3p25-26; (b) OCLOHR-2 mapped between D3S1317 and D3S1259 at 3p24-25; (c) OCLOHR-3 mapped between D3S1300 and D3S1284, an area that includes the FHIT locus at 3p14.2; and (d) OCLOHR-4 mapped between D3S1284 and D3S1274 at 3p12-13, a region known to contain overlapping homozygous deletions in lung and breast tumor cell lines. However, microsatellite markers from the chromosome 3p21.3 interval homozygously deleted in lung cancer cell lines did not identify a distinct OCLOHR. The frequency and extent of 3p LOH correlated with tumor stage such that LOH at two or more OCLOHRs was present in 53% (16 of 30) of stage III tumors but only 26% (5 of 19) of stage I/II tumors (P = 0.08). To determine the relationship between the OCLOHRs and the three candidate ovarian cancer suppression regions (OCSRs) identified previously by monochromosome transfer studies, we performed detailed genetic and physical mapping studies to define the extent of the three candidate OCSRs and to establish YAC contigs covering each region. OCSR-A at 3p25-26 and OCSR-B at 3p24 were shown to overlap with OCLOHR-1 and OCLOHR-2, respectively, providing further evidence for OCTSGs in these regions. We also show that OCSR-C overlaps with a locus at 3p21.3 previously implicated in lung and breast cancer.     

Allelic loss in ovarian cancer

Loss of heterozygosity (LOH) was examined at 86 loci distributed on every chromosomal arm in 50 human ovarian tumors. Frequent allele losses were observed on chromosomes 13q (42%), 17p (42%), 17q (45%), and Xp (41%). Deletion mapping on chromosome 17 revealed a candidate gene on the long arm distal to D17S41/S74 for ovarian cancer which is distant from the locus for early onset breast cancer. LOH on chromosome 17q was found to be concordant with LOH on chromosomes 3p, 13q, 17p and Xp suggesting that it may be an early event in neoplastic development. These findings indicate that multiple tumor-suppressor genes for ovarian cancer possibly exist on chromosomes 13q, 17, and/or Xp and provide the basis for the identification of candidate gene(s) associated with ovarian cancer. The chromosomal mechanisms resulting in allele losses in ovarian cancer include deletion, deletion/duplication, mitotic recombination and monosomy, in concordance with the developed genetic model.     

Allelotype of non-small cell lung cancer

Loss of heterozygosity (LOH) studies have been used extensively to identify regions on chromosomes that may contain putative tumour suppressor genes. We have undertaken extensive allelotyping of 45 specimens of non-small cell lung cancer (NSCLC) using 92 polymorphic microsatellite markers on 39 chromosome arms. The most frequent allelic imbalances were found on chromosome arms 3p, 9p and 17p. Significant allelic imbalance was found on other chromosome arms including, 5q (21%), 8p (19%), 13q (24%) and 17q (18%). The LOH data on 3p was subdivided into the four chromosomal regions considered to contain putative tumour suppressor genes 3p25-p24 (10%), 3p21 (10%), 3p14 (25%) and 3p13-p12 (22%). The frequency of loss in the different regions on 9p were: 9pter-p23 (31%), 9p23-p22 (45%) and 9p21-cent (30%). LOH on 17p was separated into three regions: 17pter-p13 (9%), 17p13 (33%) and 17p13-cent (22%). No correlation was found between LOH on any of the chromosomal arms and any of the clinicopathological parameters such as pathology, level of differentiation, TNM staging or alcohol intake. Only one significant association was found between LOH and tumour types. A significant difference was found between LOH on 17q in adenocarcinomas and squamous cell carcinomas (p=0.037). The fractional allele loss (FAL) values for this group of 45 NSCLC gave a median value of 0.9 (range 0-0.45). No correlation was found between FAL and nodes at pathology (p>0.05) and between FAL and tumour grade (p>0.05). No correlation was found between p53 or ras mutations in these NSCLC specimens and their FAL values. Accumulated genetic damage, as provided by this allelotype analysis, provides a useful molecular parameter by which to assess NSCLC and may, in time, assist in the determination of the clinical behaviour and clinical outcome of these tumours.     

Partial allelotype of schistosomiasis-associated bladder cancer

In Egypt and other regions of the Middle East where the trematode Schistosoma haematobium is endemic, bladder cancer is the most common adult cancer. Unlike bladder cancers in Western countries, which are predominantly transitional-cell carcinoma (TCC), these schistosomiasis-associated bladder cancers are predominantly squamous-cell carcinoma (SCC). Our aim was to assess a large series of schistosomiasis-associated bladder tumours for genetic alterations commonly found in TCC in the United Kingdom and the United States. We have carried out a partial allelotype of 70 tumours from patients with schistosomiasis. LOH was found on all chromosome arms studied (3p, 4p, 4q, 8p, 9p, 9q, 11p, 11q, 13q, 14q, 17p, 18q). The most frequent regions of LOH were 9p (65%), 17p (58%), 3p (40%), 9q (39%) and 8p (37%). LOH on 17p, where the TP53 gene is located, was more common in Egyptian TCC than in SCC. Similarly, 8p LOH was more common in TCC than SCC. The most striking difference between this group of tumours and TCCs from the United Kingdom and the United States was the high frequency of 9p LOH in the region of the CDKN2 gene (65%) and the relatively low frequency of 9q LOH (39%); 15 of 43 tumours with LOH of at least one marker on chromosome 9 showed LOH of 9p only. This suggests that a 9p gene, possibly CDKN2, may contribute to the development of the majority of schistosomiasis-associated bladder tumours but that genes on 9q play a much less important role.     

Frequent loss of heterozygosity at the DCC locus in gastric cancer.

We examined 28 cases of surgically resected gastric cancer, excluding the diffuse type, for loss of heterozygosity (LOH) on 12 chromosomal arms using polymorphic DNA markers. LOH on chromosome 18q was detected in 61% (14 of 23) of the cases by the probes OLVIIA8, OLVIIE10, p15-65, SAM 1.1, and OS-4, and a putative common region showing LOH included the locus of the DCC tumor suppressor gene. LOH on chromosome 17p was also frequently found (8 of 19 or 42% of the cases) by the probes p10-3 and pHF12-1, and in 5 of these 6 cases the LOH on chromosome 17p was accompanied by LOH on chromosome 18q. On the other hand, the incidence of LOH was 30% or less using probes pHRnES, pHF12-65, p-c-mybE2.6, NJ3 3.2, pHF12-8, pHINS6.0, p9D11, hp2-alpha, pCMM6, and P1A5 on chromosomes 1q, 5, 6q, 7q, 9, 11p, 13q, 16q, 20, and 22q, respectively. LOH on chromosome 18q was frequent irrespective of the depth of tumor invasion, whereas the incidence of LOH on chromosome 17p was higher in the cases in which the tumor invaded beyond the muscularis propria than in those in which tumor invasion was limited to the submucosa and muscularis propria. These results suggest that LOH on chromosome 18q occurs at an earlier stage than LOH on chromosome 17p and that the inactivation of tumor suppressor genes located on chromosome 17p and 18q (e.g., the p53 and DCC genes) is critically involved in the development of the majority of gastric cancers. While alteration of the p53 gene is observed in various human cancers, that of the DCC gene is considered to occur more selectively in gastrointestinal cancers.     

Frequent loss of heterozygosity on chromosomes Xp and 13q in human ovarian cancer.

Loss of heterozygosity (LOH) was examined at 27 loci on chromosomes 3p, 6q, 11p, 13q, 17 and X in 42 human ovarian tumors. LOH was detected in 12 of 26 (46%) and 5 of 12 (42%) informative cases at 2 chromosome 13q loci, D13S32 and D13S34 respectively. On chromosome Xp, tumor-specific allele loss was observed in 9 out of 15 informative cases (60%) at the ornithine transcarbamylase (OTC) gene locus. Examination of 12 additional Xp and 13q loci has mapped the common deletion regions to Xp21.1-->p11.4 and 13q33-->q34. The observation of significant LOH on Xp represents a strong indication of genetic changes in the X chromosome in a human malignancy. The allele losses on 13q which have been reported for other cancers suggest that chromosome 13, in addition to the retinoblastoma gene, may contain other growth-regulating gene(s) important in the development of several tumor types, including ovarian malignancies.     

Comparative analysis of loss of heterozygosity of specific chromosome 3, 13, 17, and X loci and TP53 mutations in human epithelial ovarian cancer

We previously reported the identification of three minimal regions of deletion on the short arm of chromosome 3 (3p) in epithelial ovarian tumor specimens, suggesting that the inactivation of tumor-suppressor genes in these regions may be important in terms of ovarian tumorigenesis. Another previous study of ovarian cancer observed that allele loss of chromosome 179 was frequently found in ovarian tumors that also showed loss of heterozygosity (LOH) of chromosomes 3p, 13q, 17p, and Xp. In an independent study, we also reported a high frequency of LOH for selected chromosome 17 loci in high-grade and late-stage ovarian tumors. We have extended our LOH analysis of chromosome 3p to include 102 ovarian tumor specimens (29 and 73 samples were previously examined for LOH of chromosome 3p and 17 markers, respectively), using additional polymorphic markers, to assess the coordinate LOH of loci representing the three chromosome 3p minimal regions of deletions [von Hippel-Lindau syndrome (VHL), thyroid hormone receptor beta, and fragile histidine triad (FHIT)] and LOH of other important loci [tumor protein 53 (TP53), breast cancer 1 early onset (BRCA1), breast cancer 2 early onset, retinoblastoma 1, ornithine carbamoyltransferase, and androgen receptor] or somatic mutations in TP53. There was a significant association between LOH of any chromosome 3p marker and LOH of any chromosome 17 marker (P = 0.026). The frequency of LOH at the TP53 locus was higher in the group of samples that displayed LOH of a 3p marker (P = 0.019), as was the frequency of LOH at the BRCA1 locus (P = 0.014). LOH of chromosome 3p was noted in four specimens that did not display LOH of either the BRCA1 or the TP53 locus, indicating that LOH of these loci need not precede LOH of the chromosome 3p loci. We found a significant association between LOH of the VHL (3p25) locus and LOH of any chromosome 17 marker (P = 0.005), suggesting that there may be an important relationship, in the tumorigenesis of epithelial ovarian cancer, between a gene at 3p25 and a gene located on chromosome 17. Our results indicate that inactivation of p53 by somatic mutation is unlikely to be a prerequisite to chromosome 3p LOH, because we found no significant association between mutations in TP53 and LOH of the three chromosome 3p loci. The frequency of LOH at the FHIT locus at 3p14 increased significantly with advancing age at diagnosis (P = 0.018), as did the frequency of somatic TP53 mutations (P = 0.008).     

Comparison of human papillomavirus DNA levels in gynecological cancers: implication for cancer development.

We have previously demonstrated the presence of human papillomavirus (HPV) DNA in several gynecological cancers using conventional PCR. In the present study, to further understand the role of HPV in malignant transformation of these cancers, the infection rates and viral loads of HPV 16 and 18 in gynecological cancers were analyzed using real-time quantitative PCR (qPCR). HPV 16 DNA was detected in 61.0% (58/95), 15.2% (7/46) and 32.1% (18/56) of cases of cervical, endometrial and ovarian cancers, respectively. On the other hand, HPV 18 DNA was detected in 23.2% (22/95) of cervical cancers, 1.8% (1/56) of ovarian cancers, and in no cases of endometrial cancer. Thus, HPV 16 is much more prevalent than HPV 18 in malignancies of the female genital tract. We also found that both HPV 16 and 18 were significantly (p < 0.05) less frequently present in endometrial and ovarian cancers than in cervical cancer. The median copy numbers of HPV 16 DNA in endometrial and ovarian cancers were 3,500 and 7,590 copies/microg DNA, respectively. These amounts were also significantly (p < 0.05) lower than HPV 16 DNA in cervical cancer (492,800 copies/microg DNA). Thus, HPV 16 could be detected in all three types of gynecological cancer, whilst HPV 18 is extremely rare in endometrial and ovarian cancers. The lower HPV 16 infection rates and lower copy numbers when compared with cervical cancer tend to suggest that HPV plays a less essential role in the development of endometrial cancer and ovarian cancer.     

A statistical analysis of chromosome 11 and 17 loss of heterozygosity in epithelial ovarian cancer

Many regions of the genome exhibit loss of heterozygosity (LOH) in epithelial ovarian cancer (EOC) suggesting sites of recessive genetic elements such as tumor suppressor genes. We performed detailed LOH studies of chromosomes 17 and 11 using 24 microsatellite repeat markers in a population of 47 patients with EOC. Univariate statistical analysis revealed that significant co-losses of chromosomal loci occurred between 17p and 17q whole arms (p=0.0003), NME1 (17q21) with D11S922 (11p15.5) (p=0.0067) and D11S912 (11q24) with D11S935 (11p13) (p=0.0073). Statistical analysis of the relationship between LOH on particular chromosomal arms and clinicopathological factors revealed a significant association between serous histological subtype of ovarian adenocarcinoma and chromosome 17p (p=0.0052) and telomeric 17q (p=0.0007) LOH. An analysis of specific polymorphic chromosomal loci demonstrated that adverse survival was significantly associated with LOH at 11q24 (p=0.0067) and 17q21 (p=0.0076). There were nonsignificant trends suggesting a relationship between chromosome 17p LOH and poorly differentiated (p=0.025) and advanced FIGO stage (p=0.031) tumours. Considering these statistical associations, a preliminary multistep model for involvement of chromosomes 11 and 17 in ovarian neoplasia can be constructed.     

Chromosome 3p Alterations in Northeastern Thai Women with Cervical Carcinoma

The purpose of this study was to determine the incidence of the loss of heterozygosity (LOH) among normal ‍cervixes, cervical intraepithelial neoplasias (CINs) and invasive cervical cancers (ICCs). DNA samples (136) were ‍obtained from 31 normal cervixes, 49 CINs and 56 ICCs. Four polymorphic microsatellite markers (D3S1300, ‍D3S1351, D3S1478 and D3S4103) covering the chromosome 3p arm, were employed. LOH at one or more loci were ‍identified in: 9/31 (8.1%) normal cervixes, 17/49 (14.6%) CINs and 26/56 (22.1%) invasive cancers. The incidence of ‍the LOH at 3p varied for each locus and ranged from 5.6% for D3S1351 to the highest rate of 16.6% for D3S1300. ‍We thus found that LOH of chromosome 3p can occur in normal cervixes and that incidences increase in CINs and ‍ICCs. Deletion in the 3p14.2 (D3S1300) and 3p21.2 (D3S1478) regions might be an early event and, in fact, necessary ‍for cervical cancer progression. The loss of function of tumor suppressor genes (TSGs) located in these regions may ‍have a sequential effect in cervical cancer carcinogenesis.     

Fractional allele loss data indicate distinct genetic populations in the development of non-small-cell lung cancer.

Allelic imbalance or loss of heterozygosity (LOH) has been widely used to assess genetic instability in tumours, and high LOH on chromosome arms 3p, 9p and 17p has been considered to be a common event in non-small-cell lung cancer (NSCLC). We have investigated allelic imbalance in 45 NSCLCs using 92 microsatellite markers on 38 chromosome arms. LOH of 38% was observed on 3p using nine markers, 58% on 9p using 15 markers and 38% on 17p using five markers. Fractional allele loss (FAL) has been calculated for each tumour (FAL is the number of chromosome arms showing LOH/number of informative chromosome arms) and a median FAL value of 0.09 was obtained in the 45 NSCLCs studied. The LOH data were examined on the basis of FAL scores: low FAL (LFAL) (0.00-0.04), medium FAL (MFAL) (0.05-0.13) and high FAL (HFAL) (0.14-0.45) based symmetrically around the median FAL value of 0.09. Tumours with HFAL values showed a very clear polarisation of the LOH data on chromosome arms 3p, 9p and 17p, such that 80% showed loss on 3p, 80% on 9p and 73% on 17p. These incidences of LOH were significantly higher than would be expected, since overall genetic instability in these HFAL tumours ranged from 14% to 45% LOH. Nine of the 14 patients in the LFAL group were found to have no LOH on 3p, 9p or 17p, but five of these had LOH at other sites: i.e. LOH on 5p, 5q, 8p, 13q, 16q and 19q. These results indicate that LFAL patients form a new subset of NSCLC tumours with distinct molecular-initating events, and may represent a discrete genetic population.     

Genetic analysis of benign ovarian tumors

Ovarian cancer represents a major cause of cancer death among women and yet remarkably little is known about its etiology. The paradigm established by the colorectal carcinogenesis model would suggest that ovarian cancers are likely to arise through malignant transformation of benign ovarian tumors. However, molecular genetic data that could answer this important question is lacking. In our study, we analyzed 80 benign ovarian tumors for TP53 and K-ras mutations and for LOH on chromosomes 6, 7, 9, 11 and 17 using 56 microsatellite markers. Twenty-five percent (5/20) of non-epithelial tumors and 73% (44/60) of epithelial tumors exhibited LOH on at least 1 chromosome arm. A particularly high frequency of LOH was detected among the epithelial tumors on chromosome arms 6q (17%), 7p (17%), 7q (27%) and 11p (18%), which are also regions of frequent LOH among ovarian carcinomas. No K-ras mutations were detected in any tumor but somatic TP53 mutations were detected in 2/34 (6%) serous and 1/26 (4%) mucinous epithelial tumors. In contrast to most previous studies our data is derived from a relatively large number of microdissected tumors and is likely to represent a more accurate picture of the frequency of alterations in these tumors. We conclude that LOH is common in benign ovarian tumors, suggesting that inactivation of tumor suppressor genes are pivotal in their development. The high frequency of alterations is consistent with their being precursors to malignant disease but does not unequivocally prove this continuum. It does however provide a framework for future analysis of the molecular genetic etiology of ovarian tumorigenesis.     

Correlation of allelic losses and clinicopathological factors in 504 primary breast cancers

BACKGROUND: We have defined 18 chromosomal regions in which allelic losses were frequent among breast cancers. We examined whether specific allelic losses might correlate with any clinicopathological factors. METHODS: We tested DNA from matched normal and tumor tissues for loss of heterozygosity (LOH) at 18 microsatellite loci from a cohort of 504 patients who had undergone surgery for breast cancer. RESULTS: LOH at 3p14.3 correlated with a larger size of tumor (greater than 2 cm). LOH at 1p22, 3p25.1, 3p14.3, or 17q21.1 correlated with loss of estrogen receptors. LOH at as many as eleven regions correlated with loss of progesterone receptor, suggesting that these represent general phenomena associated with progression of cancer. Above all, allelic losses at 11q23-24, 13q12, 17p13.3, or 22q13 significantly correlated with lymph-node metastasis (11q23-24, p= 0.0042; 13q12, p=0.0207; 17p13.3, p=0.0478; 22q13, p=0.0162). CONCLUSION: These results suggest that some clinical characteristics of breast cancers are determined by loss of tumor suppressor genes present at specific chromosome regions. Especially, LOH at 11q23-24, 13q12, 17p13.3, and 22q13 is a significant predictor of lymph-node metastasis for patients who have undergone surgery for breast cancer, and may serve as a negative prognostic indicator.     

Allelic loss and prognosis in carcinoma of the uterine cervix

Cervical carcinomas develop as a result of multiple genetic alterations. As the genetic alterations are the cause of malignant transformation, it is likely that specific genetic alterations lead to specific clinical behaviour. The aim of this study was (i) to localise chromosome arms that harbour likely tumour-suppressor genes, by analysing loss of heterozygosity (LOH) and (ii) to study the association of LOH with clinicopathological parameters. To define the regions of interest, we studied the presence of loss of heterozygosity at all chromosomes in 67 cervical carcinomas (stages IB and IIA) with 81 polymorphic markers. In addition, all frequent allelic imbalances were correlated with HPV status and clinicopathologic parameters including survival, FIGO-stage, lymph-node metastasis, tumour size, number of mitoses, vaso-invasion and histologic type. LOH at a frequency over 25% was observed at sites on 9 chromosome arms: 3p21, 4p16.1-15, 6p, 6q22.3-23.1, 11q22-24, 15q11-21.1, 17p13.3, 18q22-qter and Xq. LOH of chromosome 6q14-16.2, 6p22 and 17p13 correlated marginally with HPV-16 positivity. LOH on chromosome 3p21 was weakly correlated with high mitotic activity, while LOH on chromosomes 11q23.3, 15q21.1 and 17p13 correlated with low mitotic activity. LOH at chromosome 17p13 associated marginally with FIGO stage I, while LOH at chromosome 15q associated weakly with FIGO stage II. When chromosome 18q showed LOH in the tumour, the patients had decreased survival (p = 0.024). We conclude that, in carcinoma of the uterine cervix, a novel tumour-suppressor gene may be present on chromosome 15q21 and that patients with LOH on chromosome 18q have relatively poor survival (p = 0.025). Int. J. Cancer (Pred. Oncol.) 79:411–417, 1998. © 1998 Wiley-Liss, Inc.     

Early loss of heterozygosity on 17q in ovarian cancer. The Abe Ovarian Cancer Genetics Group.

We have studied 146 ovarian tumours (94 carcinomas, 22 tumours of low malignant potential and 30 benign tumours) for evidence of allele loss on chromosome 17p and 17q sufficient to imply the proximity of a tumour-suppressor gene. We have examined two polymorphic loci (YNZ22.2 and BHP53) on 17p13 and one on chromosome 17q (17q23-qter). Loss of heterozygosity (LOH) was detected in 34/63 (54%) informative malignant tumours at YNZ22.2 and 22/47 (47%) at BHP53; on 17q, 45/64 (70%) had LOH. Allele loss was detected in a small number of benign and borderline tumours. There was a statistically significant difference between the patterns of allele loss in serous and endometrioid groups of tumours, and allele loss occurred with significantly greater frequency on 17q than on 17p. Comparison of all malignant tumours presenting with either localized (FIGO stage I/II) or widespread (FIGO stage III/IV) disease showed that, particularly on 17q, allele loss increases in the more advanced stages. The p53 tumour-suppressor gene is implicated in ovarian carcinogenesis, and our findings suggest that an important tumour-suppressor gene may be located in the region 17q23-qter. Loss of function in this gene may be responsible for the frequently observed rapid progression of serous-type adenocarcinomas to an advanced stage.     

Heterogeneity for allelic loss in human breast cancer.

BACKGROUND: Loss of heterozygosity (LOH) at specific chromosomal regions in the tumor cells implicates the presence of tumor suppressor genes. However, it is also possible for an LOH to be randomly acquired and irrelevant to tumor development. PURPOSE: To determine whether a particular LOH in human breast carcinomas represents a loss of tumor suppressor gene or merely a random loss, we analyzed untreated primary breast cancers for LOH. METHODS: Ninety-eight primary human breast cancers from previously untreated patients were analyzed for LOH at 12 chromosomal regions including five randomly selected regions and seven regions previously reported in other cancer types and/or breast cancers. RESULTS: The baseline incidence of LOH was five out of 124 tests (4%) using randomly selected probes on chromosomes 1p, 2p, 4p, 11q, and Xq. Incidences of LOH significantly greater than background were seen in the following chromosomal regions: 22q (10 of 26 cases, 38%); 18p (five of 24 cases, 21%); 17p (30 of 59 cases, 51%); 13q (five of 14 cases, 36%); 3p (13 of 28 cases, 46%); and 1q (18 of 70 cases, 26%). In contrast to previous reports, the incidence of LOH was not significantly different from background for 11p15. In all cases, results were the same for metastatic lymph nodes and primary tumors, suggesting that the losses occurred early in the malignant progression. In cases with LOH at more than one locus, the same DNA sample often varied in degree of signal reduction for the missing alleles. CONCLUSION: These observations indicate the presence of both intertumor and intratumor heterogeneity for LOH.     

Whole-genome allelotyping identified distinct loss-of-heterozygosity patterns in mucinous ovarian and appendiceal carcinomas.

PURPOSE: Mucinous adenocarcinoma of the ovary is one of the common histologic types of ovarian cancer. Its pathogenesis is largely unknown. In addition, the differential diagnosis of metastatic mucinous carcinomas to the ovaries, particularly those originating from the appendix, remains challenging. The purpose of this study is to identify molecular biomarkers for mucinous ovarian adenocarcinoma and compare them with those of appendiceal origin. EXPERIMENTAL DESIGN: Genome-wide loss-of-heterozygosity (LOH) analysis was done on DNA isolated from 28 microdissected primary mucinous ovarian carcinomas and five appendiceal adenocarcinomas. Markers from high-loss regions were selected for further analysis on a total of 32 ovarian and 14 appendiceal cancers. RESULTS: High levels of LOH rates (>40%) were detected on chromosome arms 9p, 17p, and 21q in mucinous ovarian carcinoma cases. The frequency of allelic loss was similar between high-grade and low-grade mucinous ovarian carcinoma cases but was significantly higher in ovarian versus appendiceal cases. In addition, LOH rates on five chromosomal loci were statistically different between ovarian and appendiceal carcinomas. CONCLUSION: A high frequency of LOH can be found in mucinous ovarian adenocarcinomas independent of grade. Despite histologic similarities between mucinous ovarian carcinomas and metastatic appendiceal carcinomas, they have distinct LOH profiles, which may be used for distinguishing the two diseases.