The Role of p16/Ki67 Dual Staining in Cervical Cancer Screening
Round 1
Reviewer 1 Report
This review manuscript provides an overview about various cervical cancer screening approaches and technologies, with a specific focus, but not limited to, p16/Ki-67 dual staining.
Line 65: “The most common type of abnormal cervical smear is ‘atypical squamous cells of unknown significance’ (ASC-US) which is the diagnosed in 59 % of all abnormal smears.”: due to the subjectivity of Pap cytology interpretation and different medico-legal environments, the prevalence of ASC-US results is highly variable. The authors may refer to a specific study when providing an exact number
Line 92: “The disadvantage of cervical cytology lies in its low sensitivity is in the range of 20–70 %”; check the structure of the sentence
Line 92, whole paragraph: the ATHENA, VUSA-Screen and POBASCAM studies had much larger numbers of study participants in which the performance of Pap cytology for the detection of subjects with underlying CIN3 or worse have been evaluated (e.g. more than 47,000 women in ATHENA). The authors should check the numbers provided, which may refer to subcohorts (such as e.g. HPV positive women only)
Line 174: “The majority of HPV infections clear within two years and only a small fraction of persist and represent”; something missing (small fraction of them?)
Line 233: “Almost 100 % of cervical cancers are HPV positive”; while the Walboomers et al. 1999 suggested an almost 100% HPV prevalence in invasive cervical cancers, such high prevalence rates have not been seen in many other studies. Furthermore, with more HPV testing used for screening, the number of non-HPV related cervical cancers may rise further. Therefore, I suggest to state “The vast majority of cervical cancers ….”, instead of, Almost 100%
Line 238 and 242: “while E6 and E7 are involved in interactions with cellular tumor suppressor genes [6,69]”: E6 and E7 oncoproteins interfere with tumor suppressor proteins (p53, pRB), not with their genes
Line 240: “The virus enters the nucleus and integrates into host DNA. This integration occurs in the region of gene E2 and causes the loss of transcriptional control of E6 and E7 through their interaction with tumor-suppressor genes [68].”: this gives the wrong impression that integration of episomal HPV DNA into the host cell’s genomic DNA is an early and frequent event, which is not accurate. Instead, viral DNA integration has been shown to be a rather late event during cervical cancerogenesis, and it is mostly restricted to advanced precancerous and cancerous lesions, following aneuploidy and genomic chaos. Also, viral integration leading to interference with oncogenes or tumor suppressor genes have only been shown in a limited number of cases
Line 256: “Its expression is increased in the basal layer of squamous cells through E7/pRb pathway [16].” The authors may provide a more detailed description of the mechanism leading to over-expression of p16 in epithelial cells undergoing HPV E7 mediated cellular transformation. Also, p16 overexpression in transforming HPV infections is not limited to the basal cell layer of squamous cervical epithelium
Line 273: “Sensitivity of dual staining was comparable to Pap test and specificity was higher.”: this statement is incorrect, the opposite results have been reported for the PALMS trial (Ikenberg et al. 2013 [73]:” …. dual-stained cytology was more sensitive than Pap cytology (86.7% vs 68.5%; P < .001) for detecting CIN2+, with comparable specificity (95.2% vs 95.4%; P = .15)”)
Line 299: “American authors analyzed the 299 utility of p16/Ki67 dual staining” , somewhat weird wording; the authors may want to refer to a study conducted at … in …”
Linde 326: “The authors reported higher sensitivity for dual stain-326 ing but lower specificity.”: not exactly clear which comparison is meant here, given that besides dual staining, two other methods are listed in the previous sentence. Please clarify wording
Line 334: “could be used as triage for HPV positive women [86].”: the authors may want to indicate that dual staining has been approved by the US FDA in March 2020 for the triage of HPV-positive women, either in primary HPV screening, or HPV/Pap cytology co-testing
Line 343: “There are two promising aspects of p16/Ki67 screening which need further research, namely the role of the number of p16/Ki67 positive cells in the detection of high-grade cervical precancerous lesions …”: the authors may want to provide a detailed rationale for their request for additional studies, given the 92 studies the authors refer to which focus on the assessment of the performance of dual staining for the detection of high-grade CIN.
Line 351: The probability of de-351 tecting CIN2+ was 67.3 % with 5 or more p16/Ki-67 positive cells compared to 11.8 % in 352 cases with only one positive cell [91].: this statement is confusing and potentially misleading. First, “probability of detecting CIN2+” may be mixed with PPV but is not the same; however, both are prevalence dependent and thus cannot be generalized beyond the study cohort analyzed. Second, the numbers referred to by Gajsek et al. [91] are very small, e.g. 2 of 17 (1 pos cell), 0 of 10 (2 cells), 1 of 4 (3 cells) etc. Third, for these reasons, it might be better if the authors would refer to sensitivity and specificity results for dual staining dependent on the number of dual stained cells per slide (there are other, larger studies as well which assessed the impact of a different number threshold of dual stained cells on sensitivity and specificity, which the authors may want to cite here as well).
Author Response
This review manuscript provides an overview about various cervical cancer screening approaches and technologies, with a specific focus, but not limited to, p16/Ki-67 dual staining.
Line 65: “The most common type of abnormal cervical smear is ‘atypical squamous cells of unknown significance’ (ASC-US) which is the diagnosed in 59 % of all abnormal smears.”: due to the subjectivity of Pap cytology interpretation and different medico-legal environments, the prevalence of ASC-US results is highly variable. The authors may refer to a specific study when providing an exact number
Response to reviewer’s comment: The reviewer is right. We corrected this sentence and added additional studies to the reference list. The exact number regarding the percentage of ASC-US among all Pap smears has been corrected (Tao et al, Kattoor et al.) (page 2, lines 67-69).
Line 92: “The disadvantage of cervical cytology lies in its low sensitivity is in the range of 20–70 %”; check the structure of the sentence
Response to reviewer’s comment: The sentence has been corrected. It has been highlighted in red in the text (page 4, line 93).
Line 92, whole paragraph: the ATHENA, VUSA-Screen and POBASCAM studies had much larger numbers of study participants in which the performance of Pap cytology for the detection of subjects with underlying CIN3 or worse have been evaluated (e.g. more than 47,000 women in ATHENA). The authors should check the numbers provided, which may refer to subcohorts (such as e.g. HPV positive women only)
Response to reviewer’s comment: Thank you very much for this observation. The respective numbers have been corrected in the text (page 4, line 99).
Line 174: “The majority of HPV infections clear within two years and only a small fraction of persist and represent”; something missing (small fraction of them?)
Line 233: “Almost 100 % of cervical cancers are HPV positive”; while the Walboomers et al. 1999 suggested an almost 100% HPV prevalence in invasive cervical cancers, such high prevalence rates have not been seen in many other studies. Furthermore, with more HPV testing used for screening, the number of non-HPV related cervical cancers may rise further. Therefore, I suggest to state “The vast majority of cervical cancers ….”, instead of, Almost 100%
Response to reviewer’s comment: Thank you very much for this suggestion. The sentence has been corrected according to the reviewer’s suggestion (page 7, line 234).
Line 238 and 242: “while E6 and E7 are involved in interactions with cellular tumor suppressor genes [6,69]”: E6 and E7 oncoproteins interfere with tumor suppressor proteins (p53, pRB), not with their genes
Response to reviewer’s comment: This sentence has also been corrected to include this important observation (page 7, line 248).
Line 240: “The virus enters the nucleus and integrates into host DNA. This integration occurs in the region of gene E2 and causes the loss of transcriptional control of E6 and E7 through their interaction with tumor-suppressor genes [68].”: this gives the wrong impression that integration of episomal HPV DNA into the host cell’s genomic DNA is an early and frequent event, which is not accurate. Instead, viral DNA integration has been shown to be a rather late event during cervical cancerogenesis, and it is mostly restricted to advanced precancerous and cancerous lesions, following aneuploidy and genomic chaos. Also, viral integration leading to interference with oncogenes or tumor suppressor genes have only been shown in a limited number of cases
Response to reviewer’s comment: Thank you very much. The HPV mediated carcinogenesis has been further described to clarify that viral DNA integration is a late event and that in the initial phase of the infection the viral DNA is located in the host cell as episome (page 7, line 241-245).
Line 256: “Its expression is increased in the basal layer of squamous cells through E7/pRb pathway [16].” The authors may provide a more detailed description of the mechanism leading to over-expression of p16 in epithelial cells undergoing HPV E7 mediated cellular transformation. Also, p16 overexpression in transforming HPV infections is not limited to the basal cell layer of squamous cervical epithelium
Response to reviewer’s comment: We sincerely appreciate this observation. The role of p16 in the regulation of cell cycle has been further clarified and the sentences indicating p16 expression only in basal cells have been corrected (page 7, lines 262-267).
Line 273: “Sensitivity of dual staining was comparable to Pap test and specificity was higher.”: this statement is incorrect, the opposite results have been reported for the PALMS trial (Ikenberg et al. 2013 [73]:” …. dual-stained cytology was more sensitive than Pap cytology (86.7% vs 68.5%; P < .001) for detecting CIN2+, with comparable specificity (95.2% vs 95.4%; P = .15)”)
Response to reviewer’s comment: Thank you very much for this correction. The sentence was corrected according to the reviewer’s suggestion (page 8, lines 283-284).
Line 299: “American authors analyzed the 299 utility of p16/Ki67 dual staining” , somewhat weird wording; the authors may want to refer to a study conducted at … in …”
Response to reviewer’s comment: The sentence has been rephrased (page 9, lines 309-312).
Linde 326: “The authors reported higher sensitivity for dual stain-326 ing but lower specificity.”: not exactly clear which comparison is meant here, given that besides dual staining, two other methods are listed in the previous sentence. Please clarify wording
Response to reviewer’s comment: The sentence has been further clarified to explain that higher sensitivity and lower specificity of dual-staining compared with LBC was reported in this study (page 10, line 342).
Line 334: “could be used as triage for HPV positive women [86].”: the authors may want to indicate that dual staining has been approved by the US FDA in March 2020 for the triage of HPV-positive women, either in primary HPV screening, or HPV/Pap cytology co-testing
Response to reviewer’s comment: A sentence has been added to indicate the FDA approval (line 349-352).
Line 343: “There are two promising aspects of p16/Ki67 screening which need further research, namely the role of the number of p16/Ki67 positive cells in the detection of high-grade cervical precancerous lesions …”: the authors may want to provide a detailed rationale for their request for additional studies, given the 92 studies the authors refer to which focus on the assessment of the performance of dual staining for the detection of high-grade CIN.
Response to reviewer’s comment: The reviewer is absolutely right. What we wanted to emphasise is that dual staining may be used in the triage of atypical glandular cells given the difficulties in cytological diagnosis of glandular changes. Atypical glandular changes represent less than 1% of all pathologic smears but are on the other hand associated with cervical precancerous lesions in up to 15%. In addition, there is lower interobserver reproducibility when diagnosing glandular changes compared to squamous cell changes. We found only one study analysing the utility of p16/Ki67 dual staining for the detection of cervical adenocarcinoma. Taking into consideration the promising results of this study we believe that this field is worth investigating further. On the other hand, we agree with the reviewer that larger studies analysed the role of the number of positive p16/Ki67 cells in the detection of high-grade cervical precancerous lesions and additional references related to this topic have been added to the reference list (lines 369-378).
Line 351: The probability of de-351 tecting CIN2+ was 67.3 % with 5 or more p16/Ki-67 positive cells compared to 11.8 % in 352 cases with only one positive cell [91].: this statement is confusing and potentially misleading. First, “probability of detecting CIN2+” may be mixed with PPV but is not the same; however, both are prevalence dependent and thus cannot be generalized beyond the study cohort analyzed. Second, the numbers referred to by Gajsek et al. [91] are very small, e.g. 2 of 17 (1 pos cell), 0 of 10 (2 cells), 1 of 4 (3 cells) etc. Third, for these reasons, it might be better if the authors would refer to sensitivity and specificity results for dual staining dependent on the number of dual stained cells per slide (there are other, larger studies as well which assessed the impact of a different number threshold of dual stained cells on sensitivity and specificity, which the authors may want to cite here as well).
Response to reviewer’s comment: Thank you very much for this comment. The text has been corrected and additional studies (which were already included in the text) were mentioned in this section (lines 369-378).
Reviewer 2 Report
Dear Authors
This is an interesting review about other methods to perform cervical cancer screening
I think this work need some improvement before publication
It is necessary to include a table with sensitivity and ROC curve of all tests. Regarding developing countries, this p16/ki67 dual staining is not available. We need more information about cost-effectiveness about this
Please include in the discussion some paragraphs about cost-effectiveness
Please discuss if vaccination can modify these tests.
Author Response
This is an interesting review about other methods to perform cervical cancer screening
I think this work need some improvement before publication
It is necessary to include a table with sensitivity and ROC curve of all tests. Regarding developing countries, this p16/ki67 dual staining is not available. We need more information about cost-effectiveness about this
Response to reviewer’s comment: We sincerely appreciate this very important observation. We have included a table in the text reporting sensitivity and specificity values of DS for the detection of CIN2+ and CIN3+ from major studies performed to date (pages 9 and 10)
Please include in the discussion some paragraphs about cost-effectiveness
Response to reviewer’s comment: Thank you very much for this comment. Additional references regarding the cost-effectiveness of DS have been added to the reference list (lines 385-397).
Please discuss if vaccination can modify these tests.
Response to reviewer’s comment: Thank you very much for this comment. Additional references have been added to the reference list. There is robust evidence that positive predictive value of high-risk HPV testing and Pap cytology will decrease as a result of lower HPV 16/18 prevalence in the vaccinated population. Additional clarifications have been added to the text, please find a paragraph on page 12. We have not found any studies evaluating the impact of HPV vaccination on HP methylation tests and dual-staining (lines 398-409).
