- Email: [email protected]
Response to ctla-4 gene variations in southern Iranian patients with cervical cancer
Letters to the Editor
641
events in OCEM group: Pattern I was that only cancer tissue demonstrated LOH, while the endometriosis and the transitional area remained intact (18 events); Pattern II was that endometriosis, transitional and cancer tissues all demonstrated LOH (13 LOH events). Fujii et al. reported homogeneous and heterogeneous patterns of allelic losses in their series studies including gynecological carcinosarcomas [2–4]. They described various heterogeneous LOH patterns in their elegant studies. We think our findings are consistent with their reports. Our classification of type I and type II for LOH status is similar to their heterogeneity and uniform in definition, respectively [2]. Watanabe et al. pointed out that “However, they examined only one sample of each histological feature in each patient.”. It is not fully true. We have stated in our article “Two representative areas were microdissected for each different pathological tissue from each specimen.” And all the cases with LOH were re-amplified and analyzed to ensure reproducibility. Our findings of common LOH events in solitary ovarian endometriosis, although occurring at a lower frequency than in OCEMs, support the possibility that these two diseases may have a common genetic lineage. Based on our results, it may be reasonable to propose a possibility for the malignant transformation of endometriosis. The initial genetic abnormalities on some chromosomes such as sporadic LOH may first increase the predisposition to endometriosis. Because most endometriosis cases remain benign, these genetic defects may only be involved in the initiation and maintenance of endometriosis. Accumulated LOH on more loci and further genetic alterations involving tumour suppressor genes and/or oncogenes may confer invasive features to the cells and lead to the malignant transformation of endometriosis into cancer. We believe that our data may supply evidence that both clonal progression and genetic heterogeneity are involved in malignant transformation of endometriosis.
“Letter to editor” in lieu of article published below in Gynecol Oncolo. 2010 Oct;119(1):136–9 ctla-4 gene variations in southern Iranian patients with cervical cancer
Conflict of interest statement The authors have no conflict of interests to declare.
References
This case–control study depicts the association of ctla-4 gene polymorphisms, and susceptibility to cervical cancer. Although the study provides preliminary evidence to consider ctla-4 polymorphisms as a risk factor for cervical cancer, there are a few important issues that need to be clarified. Sample size remains a major issue in genetic case–control studies analyzing the association of polymorphisms with disease susceptibility. The study should obtain adequate statistical power to estimate a significant association accurately, which remains a primary criterion to perform such studies. Underpowered studies usually lead to falsepositive associations and misinterpretations [1,2]. The authors have not mentioned the incidence rate of cervical cancer in their study population. Whether the number of patients and controls recruited in the present study achieved the required statistical power is questionable. The authors did not discuss, as to why they have found risk in wild genotype and allele. Wild type genotype and allele is always taken as a reference. All these points suggest a thorough examination of the associations observed in this study, and should be clarified before concluding that ctla-4 gene polymorphisms are potential markers of cervical cancer. Conflict of interest statement None.
[1] Hattersley AT, McCarthy MI. What makes a good genetic association study? Lancet Oct 8 2005;366(9493):1315–23. [2] Delucchi KL. Sample size estimation in research with dependent measures and dichotomous outcomes. Am J Public Health Mar 2004;94(3):372–7.
References [1] Xu B, Hamada S, Kusuki I, Itoh R, Kitawaki J. Possible involvement of loss of heterozygosity in malignant transformation of ovarian endometriosis. Gynecol Oncol 2011;120:239–46. [2] Fujii H, Inagaki M, Kasai S, Miyokawa N, Tokusashi Y, Gabrielson E, Hruban RH. Genetic progression and heterogeneity in intraductal papillary-mucinous neoplasms of the pancreas. Am J Pathol 1997;151:1447–54. [3] Fujii H, Ajioka Y, Kazami S, Takagaki T, Gong Zhu X, Hirose S, Watanabe H. Loss of heterozygosity in the clonal evolution of flat colorectal neoplasms. J Pathol 2002;197:298–306. [4] Fujii H, Yoshida M, Gong ZX, Matsumoto T, Hamano Y, Fukunaga M, Hruban RH, Gabrielson E, Shirai T. Frequent genetic heterogeneity in the clonal evolution of gynecological carcinosarcoma and its influence on phenotypic diversity. Cancer Res 2000;60:114–20.
Bing Xu Department of Obstetrics and Gynecology, The Affiliated Hospital of Medical College, Qingdao University, Qingdao, P.R. China Jo Kitawaki Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine, Graduate School of Medical Science, Kyoto, Japan Corresponding author at: Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine, 465 Kajii-cho, Kamigyo-ku, Kyoto 602-8566, Japan. Fax: +81 75 212 1265. E-mail address: [email protected].
26 January 2011 doi:10.1016/j.ygyno.2011.01.034
To the Editor:
Raju K. Mandal Rama Devi Mittal⁎ Department of Urology and Renal Transplantation, Sanjay Gandhi Post, Graduate Institute of Medical Sciences, Raebareli Rd., Lucknow-226014, India E-mail addresses: [email protected], [email protected] (R.D. Mittal). ⁎Corresponding author. 15 December 2010 doi:10.1016/j.ygyno.2010.12.349
Response to ctla-4 gene variations in southern Iranian patients with cervical cancer
To the Editor: In a letter to “Gynecologic Oncology” editor, Raju et al. raised the issues regarding the sample size and the power of our study recently published in Gynecologic Oncology [1] . Unfortunately we ignored to indicate, in our article, the low incidence of cervical cancer in Iran which makes it practically very difficult to collect a big sample size with the time limit we had for our project. In the south and southwest area of Iran, from where we collected our samples, the incidence of cervical cancer is 1.1 per 100,000 or less according to different
642
Letters to the Editor
published information [2,3]. In our study [1], 55 patients and 110 healthy controls were enrolled and genotyped for four polymorphic loci; −1722(T/C), −1661(A/G), −318(C/T) and +49(A/G) in ctla-4 gene. After Bonferroni correction for multiple comparisons, no differences were found in the distribution of genotypes and alleles between patients and controls. The differences in two haplotypes, however, stand Bonferroni correction. Although this sample size is not appropriate to detect small differences between controls and cases, the bigger differences might always be present and can be detected with a relatively good power even when the sample size is not too large. We agree with Raju et al. that a bigger sample size could have provided us a better power of study; however, based on our findings in Tables 1 and 2 of the article [1] and using the algorithms in the Power and Precision 4 software package (Biostat, Inc., Englewood, NJ), the required sample size for detection of some small differences, even with a power of 0.50, exceeded more than 4000 which is practically impossible to be collected in a population with such a low incidence of cervical cancer. The reason for the low incidence of cervical cancer in our population is a different issue to be discussed by the experts; however, the human papillomavirus (HPV) types 16 and 18, which are known as one of the causal factors in cervical cancer, seem not to be frequent in the population from the south of Iran [4,5]. As we emphasized at the end of the discussion, our data provided preliminary results which need to be verified in different ethnic groups and/or through meta-analysis studies and, of course, a bigger sample size. In their letter to the editor, Raju et al. also asked for a reason why we have found risk in wild genotype and allele. As noted before, the differences observed in genotypes and alleles, including − 318 wild type C allele and CC genotype, did not stand the Bonferroni correction which means that one cannot consider a definite significant association, but two haplotypes consisted of four investigated SNPs were associated with susceptibility and protection. Despite this, it is not a strange finding if a wild type allele/genotype is associated with the risk of a disease especially when we are talking about a regulatory molecule like CTLA-4. CTLA-4 is the negative regulator of the immune system. Theoretically, CTLA-4 must always be expressed at a balanced level. This ensures a balance between an uncontrolled immune response and an immune compromised response. The first one can be pathologically associated with autoimmunity and the later with an immune-suppressed situation like cancer. If a wild type allele, for instance, is associated with the increase in CTLA-4 expression, it might be associated with health when we are talking about autoimmunity, but at the same time, it may render susceptibility to cancer in a predisposed individual. Additionally, considering only one SNP in gene–disease association study might sometimes be misleading. SNPs are better be seen in the context of the haplotypes constructed from the several related SNPs. Information from haplotypes, in which a mutant allele of an SNP
might always be seen adjacent to the wild type allele of the nearby SNP, are more reliable. It is actually the whole genetic sequence of a cis-acting element which defines the net regulatory effect on the gene expression and haplotypes provided us more information regarding the variations in sequence than a single SNP. One should also note that in the SNP studies, an allele is usually considered as wild type or mutant based on the frequency of the allele in the population, i.e., the higher abundant allele is usually considered as wild type and the other as the mutant form. The abundance of an allele may differ between different populations which means, in SNP studies, the terms “mutant” and “wild type” are not too strict. Taken together, it is not strange at all to find the risk for a disease associated with a wild type genotype and allele. Conflict of interest Authors declare no conflict of interest.
References [1] Rahimifar S, Erfani N, Sarraf Z, Ghaderi A. ctla-4 gene variations may influence cervical cancer susceptibility. Gynecol Oncol 2010;119(1):136–9 Oct. [2] Mehrabani D, Tabei SZ, Heydari ST, Shamsina SJ, Shokrpour N, Amini M, et al. Cancer occurrence in Fars province, southern Iran. Iranian Red Crescent Medical Journal 2008;10(4):314–22. [3] Farahmandbeigi M, Kadivar MR. The incidence rate of registered cancer in Fars province. In: Sciences, SUoM, editor. Disease Control Unit. Shiraz: Shiraz University Press; 2000. [4] Farjadian S, Asadi E, Doroudchi M, Dehaghani AS, Tabei SZ, Kumar VP, et al. High risk HPV types in southern Iranian patients with cervical cancer. Pathol Oncol Res 2003;9(2):121–5. [5] Safaei A, Khanlari M, Momtahen M, Monabati A, Robati M, Amooei S, et al. Prevalence of high-risk human papillomavirus types 16 and 18 in healthy women with cytologically negative pap smear in Iran. Indian J Pathol Microbiol 2010;53(4): 681–5 Oct-Dec.
Nasrollah Erfani Shiraz Institute for Cancer Research, Shiraz University of Medical Sciences, Shiraz, Iran E-mail address: [email protected]. Abbas Ghaderi Shiraz Institute for Cancer Research, Shiraz University of Medical Sciences, Shiraz, Iran Department of Immunology, Shiraz University of Medical Sciences, Shiraz, Iran Corresponding author at: Shiraz Institute for Cancer Research, Shiraz University of Medical Sciences, P.O. Box 71345-3119, Shiraz, Iran. Fax: +98 711 2304952. E-mail address: [email protected]. 27 January 2011 doi:10.1016/j.ygyno.2011.01.035
641
events in OCEM group: Pattern I was that only cancer tissue demonstrated LOH, while the endometriosis and the transitional area remained intact (18 events); Pattern II was that endometriosis, transitional and cancer tissues all demonstrated LOH (13 LOH events). Fujii et al. reported homogeneous and heterogeneous patterns of allelic losses in their series studies including gynecological carcinosarcomas [2–4]. They described various heterogeneous LOH patterns in their elegant studies. We think our findings are consistent with their reports. Our classification of type I and type II for LOH status is similar to their heterogeneity and uniform in definition, respectively [2]. Watanabe et al. pointed out that “However, they examined only one sample of each histological feature in each patient.”. It is not fully true. We have stated in our article “Two representative areas were microdissected for each different pathological tissue from each specimen.” And all the cases with LOH were re-amplified and analyzed to ensure reproducibility. Our findings of common LOH events in solitary ovarian endometriosis, although occurring at a lower frequency than in OCEMs, support the possibility that these two diseases may have a common genetic lineage. Based on our results, it may be reasonable to propose a possibility for the malignant transformation of endometriosis. The initial genetic abnormalities on some chromosomes such as sporadic LOH may first increase the predisposition to endometriosis. Because most endometriosis cases remain benign, these genetic defects may only be involved in the initiation and maintenance of endometriosis. Accumulated LOH on more loci and further genetic alterations involving tumour suppressor genes and/or oncogenes may confer invasive features to the cells and lead to the malignant transformation of endometriosis into cancer. We believe that our data may supply evidence that both clonal progression and genetic heterogeneity are involved in malignant transformation of endometriosis.
“Letter to editor” in lieu of article published below in Gynecol Oncolo. 2010 Oct;119(1):136–9 ctla-4 gene variations in southern Iranian patients with cervical cancer
Conflict of interest statement The authors have no conflict of interests to declare.
References
This case–control study depicts the association of ctla-4 gene polymorphisms, and susceptibility to cervical cancer. Although the study provides preliminary evidence to consider ctla-4 polymorphisms as a risk factor for cervical cancer, there are a few important issues that need to be clarified. Sample size remains a major issue in genetic case–control studies analyzing the association of polymorphisms with disease susceptibility. The study should obtain adequate statistical power to estimate a significant association accurately, which remains a primary criterion to perform such studies. Underpowered studies usually lead to falsepositive associations and misinterpretations [1,2]. The authors have not mentioned the incidence rate of cervical cancer in their study population. Whether the number of patients and controls recruited in the present study achieved the required statistical power is questionable. The authors did not discuss, as to why they have found risk in wild genotype and allele. Wild type genotype and allele is always taken as a reference. All these points suggest a thorough examination of the associations observed in this study, and should be clarified before concluding that ctla-4 gene polymorphisms are potential markers of cervical cancer. Conflict of interest statement None.
[1] Hattersley AT, McCarthy MI. What makes a good genetic association study? Lancet Oct 8 2005;366(9493):1315–23. [2] Delucchi KL. Sample size estimation in research with dependent measures and dichotomous outcomes. Am J Public Health Mar 2004;94(3):372–7.
References [1] Xu B, Hamada S, Kusuki I, Itoh R, Kitawaki J. Possible involvement of loss of heterozygosity in malignant transformation of ovarian endometriosis. Gynecol Oncol 2011;120:239–46. [2] Fujii H, Inagaki M, Kasai S, Miyokawa N, Tokusashi Y, Gabrielson E, Hruban RH. Genetic progression and heterogeneity in intraductal papillary-mucinous neoplasms of the pancreas. Am J Pathol 1997;151:1447–54. [3] Fujii H, Ajioka Y, Kazami S, Takagaki T, Gong Zhu X, Hirose S, Watanabe H. Loss of heterozygosity in the clonal evolution of flat colorectal neoplasms. J Pathol 2002;197:298–306. [4] Fujii H, Yoshida M, Gong ZX, Matsumoto T, Hamano Y, Fukunaga M, Hruban RH, Gabrielson E, Shirai T. Frequent genetic heterogeneity in the clonal evolution of gynecological carcinosarcoma and its influence on phenotypic diversity. Cancer Res 2000;60:114–20.
Bing Xu Department of Obstetrics and Gynecology, The Affiliated Hospital of Medical College, Qingdao University, Qingdao, P.R. China Jo Kitawaki Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine, Graduate School of Medical Science, Kyoto, Japan Corresponding author at: Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine, 465 Kajii-cho, Kamigyo-ku, Kyoto 602-8566, Japan. Fax: +81 75 212 1265. E-mail address: [email protected].
26 January 2011 doi:10.1016/j.ygyno.2011.01.034
To the Editor:
Raju K. Mandal Rama Devi Mittal⁎ Department of Urology and Renal Transplantation, Sanjay Gandhi Post, Graduate Institute of Medical Sciences, Raebareli Rd., Lucknow-226014, India E-mail addresses: [email protected], [email protected] (R.D. Mittal). ⁎Corresponding author. 15 December 2010 doi:10.1016/j.ygyno.2010.12.349
Response to ctla-4 gene variations in southern Iranian patients with cervical cancer
To the Editor: In a letter to “Gynecologic Oncology” editor, Raju et al. raised the issues regarding the sample size and the power of our study recently published in Gynecologic Oncology [1] . Unfortunately we ignored to indicate, in our article, the low incidence of cervical cancer in Iran which makes it practically very difficult to collect a big sample size with the time limit we had for our project. In the south and southwest area of Iran, from where we collected our samples, the incidence of cervical cancer is 1.1 per 100,000 or less according to different
642
Letters to the Editor
published information [2,3]. In our study [1], 55 patients and 110 healthy controls were enrolled and genotyped for four polymorphic loci; −1722(T/C), −1661(A/G), −318(C/T) and +49(A/G) in ctla-4 gene. After Bonferroni correction for multiple comparisons, no differences were found in the distribution of genotypes and alleles between patients and controls. The differences in two haplotypes, however, stand Bonferroni correction. Although this sample size is not appropriate to detect small differences between controls and cases, the bigger differences might always be present and can be detected with a relatively good power even when the sample size is not too large. We agree with Raju et al. that a bigger sample size could have provided us a better power of study; however, based on our findings in Tables 1 and 2 of the article [1] and using the algorithms in the Power and Precision 4 software package (Biostat, Inc., Englewood, NJ), the required sample size for detection of some small differences, even with a power of 0.50, exceeded more than 4000 which is practically impossible to be collected in a population with such a low incidence of cervical cancer. The reason for the low incidence of cervical cancer in our population is a different issue to be discussed by the experts; however, the human papillomavirus (HPV) types 16 and 18, which are known as one of the causal factors in cervical cancer, seem not to be frequent in the population from the south of Iran [4,5]. As we emphasized at the end of the discussion, our data provided preliminary results which need to be verified in different ethnic groups and/or through meta-analysis studies and, of course, a bigger sample size. In their letter to the editor, Raju et al. also asked for a reason why we have found risk in wild genotype and allele. As noted before, the differences observed in genotypes and alleles, including − 318 wild type C allele and CC genotype, did not stand the Bonferroni correction which means that one cannot consider a definite significant association, but two haplotypes consisted of four investigated SNPs were associated with susceptibility and protection. Despite this, it is not a strange finding if a wild type allele/genotype is associated with the risk of a disease especially when we are talking about a regulatory molecule like CTLA-4. CTLA-4 is the negative regulator of the immune system. Theoretically, CTLA-4 must always be expressed at a balanced level. This ensures a balance between an uncontrolled immune response and an immune compromised response. The first one can be pathologically associated with autoimmunity and the later with an immune-suppressed situation like cancer. If a wild type allele, for instance, is associated with the increase in CTLA-4 expression, it might be associated with health when we are talking about autoimmunity, but at the same time, it may render susceptibility to cancer in a predisposed individual. Additionally, considering only one SNP in gene–disease association study might sometimes be misleading. SNPs are better be seen in the context of the haplotypes constructed from the several related SNPs. Information from haplotypes, in which a mutant allele of an SNP
might always be seen adjacent to the wild type allele of the nearby SNP, are more reliable. It is actually the whole genetic sequence of a cis-acting element which defines the net regulatory effect on the gene expression and haplotypes provided us more information regarding the variations in sequence than a single SNP. One should also note that in the SNP studies, an allele is usually considered as wild type or mutant based on the frequency of the allele in the population, i.e., the higher abundant allele is usually considered as wild type and the other as the mutant form. The abundance of an allele may differ between different populations which means, in SNP studies, the terms “mutant” and “wild type” are not too strict. Taken together, it is not strange at all to find the risk for a disease associated with a wild type genotype and allele. Conflict of interest Authors declare no conflict of interest.
References [1] Rahimifar S, Erfani N, Sarraf Z, Ghaderi A. ctla-4 gene variations may influence cervical cancer susceptibility. Gynecol Oncol 2010;119(1):136–9 Oct. [2] Mehrabani D, Tabei SZ, Heydari ST, Shamsina SJ, Shokrpour N, Amini M, et al. Cancer occurrence in Fars province, southern Iran. Iranian Red Crescent Medical Journal 2008;10(4):314–22. [3] Farahmandbeigi M, Kadivar MR. The incidence rate of registered cancer in Fars province. In: Sciences, SUoM, editor. Disease Control Unit. Shiraz: Shiraz University Press; 2000. [4] Farjadian S, Asadi E, Doroudchi M, Dehaghani AS, Tabei SZ, Kumar VP, et al. High risk HPV types in southern Iranian patients with cervical cancer. Pathol Oncol Res 2003;9(2):121–5. [5] Safaei A, Khanlari M, Momtahen M, Monabati A, Robati M, Amooei S, et al. Prevalence of high-risk human papillomavirus types 16 and 18 in healthy women with cytologically negative pap smear in Iran. Indian J Pathol Microbiol 2010;53(4): 681–5 Oct-Dec.
Nasrollah Erfani Shiraz Institute for Cancer Research, Shiraz University of Medical Sciences, Shiraz, Iran E-mail address: [email protected]. Abbas Ghaderi Shiraz Institute for Cancer Research, Shiraz University of Medical Sciences, Shiraz, Iran Department of Immunology, Shiraz University of Medical Sciences, Shiraz, Iran Corresponding author at: Shiraz Institute for Cancer Research, Shiraz University of Medical Sciences, P.O. Box 71345-3119, Shiraz, Iran. Fax: +98 711 2304952. E-mail address: [email protected]. 27 January 2011 doi:10.1016/j.ygyno.2011.01.035