基因定型點墨法應用於子宮頸抹片樣本之人類乳突瘤病毒DNA檢驗的效能、效度、信度與品質保證

Performance, Validity, Reliability, and Quality Assurance for Blot Typing of Human Papillomavirus DNA in Cervical Swab Samples

林景堉(Ching-Yu Lin)

指導教授 ： 陳建仁教授

國立臺灣大學/公共衛生學院/流行病學研究所/碩士(2007年)

https://doi.org/10.6342/NTU.2007.01823

摘要

本研究的目的是評估EasyChip® HPV Blot應用於子宮頸抹片樣本之人類乳突瘤病毒DNA基因定型之效能、效度與信度，當進行大規模社區HPV篩檢時也導入品質保證系統以確保HPV基因定型結果的品質。以分析靈敏度、分析特異性、靈敏度評估HPV Blot的效能。以39種HPV基因型質體評估HPV Blot的分析靈敏度與分析特異性，以細胞株及子宮頸檢體評估靈敏度。用於評估的39種HPV基因型質體(HPV 6, 11, 16, 18, 26, 31, 32, 33, 35, 37, 39, 42, 43, 44, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 66, 67, 68, 69, 70, 72, 74, 82, CP8061, CP8304, L1AE5, MM4, MM7, and MM8)含有HPV基因組之L1基因片段。每一個質體視為一個基因組相等物。HPV Blot之效度與信度以臨床試驗評估，試驗的黃金標準是20個L1型別專一性PCR。以靈敏度、特異性與準確性評估效度，McNemar’s Test與Kappa統計量評估信度。評估的方式分別針對受試者與個別基因型。品質保證系統包括操作環境的維持、HPV Blot可靠度與再現性的評估及資料判讀。品質保證系統包含品質保證標準，包括操作時7個外部控制組及HPV Blot基因定型時所需的3個內部控制組。每一批次實驗包含89個子宮頸檢體與7個外部控制組。7個外部控制組由Caski細胞株、HeLa細胞株、Jurkat細胞株、男性血液細胞DNA、無菌水及一對配對檢體所組成。而配對檢體是源自相同樣本的分裝液。Caski細胞株、HeLa細胞株、Jurkat細胞株及配對檢體用於評估包括DNA抽取、PCR及HPV基因定型等所有的操作。男性血液細胞DNA與無菌水只用於PCR和HPV基因定型的操作。本研究需要100批次實驗進行分析。Jurkat細胞株與男性血液細胞DNA可當作操作環境中有無潛在污染的外源HPV DNA的指標，而無菌水可當作操作環境中有無潛在污染的外源DNA的指標。
HPV Blot分析靈敏度的偵測極限：HPV 16是1個基因組相等物；HPV 6、26、31、33、37、39、43、44、45、54、55、61、67、68、69、82、CP8061、CP8304、L1AE5、MM7及MM8是10個基因組相等物；HPV 11、18、32、35、42、51、52、53、56、58、59、66、70、74及MM4是20個基因組相等物；HPV 62及72是50個基因組相等物。每一樣本其HPV檢測的分析靈敏度是1~50基因組相等物的拷貝數。HPV Blot的分析特異性是39種辨識不同基因型HPV之個別探針不會與其他基因型HPV的PCR產物相互雜交，無交叉反應。在細胞株檢驗的靈敏度：Caski細胞DNA分別以膠體電泳及HPV Blot檢測的靈敏度為10-3與10-4 ng DNA；HeLa細胞DNA分別以膠體電泳及HPV Blot檢測的靈敏度為10-2與10-3 ng DNA。靈敏度的計算是參照Caski細胞的基因組含有約60~600拷貝的HPV 16基因組，HeLa細胞的基因組含有約10~50拷貝的HPV 18基因組。因此，以細胞株DNA為模板則HPV Blot之檢測極限：HPV 16與HPV 18的靈敏度分別是1~10與2~8的拷貝數。HPV Blot對子宮頸細胞檢體中HPV 16與HPV 18病毒之檢測極限為10個拷貝數。
HPV Blot對受試者的敏感度、特異性與準確性分別為97% (95%CI：94.0% – 98.0%)、96%(95%CI：92.0% – 98.0%)及96.5% (95%CI：94.0%-98.0%)。MY11/GP6+ - HPV Blot對受試者的McNemar’s Test、Kappa統計量與P值分別為0.25、0.93(95%CI：0.89-0.96) 及0.62，顯示HPV Blot與黃金標準的比較是優良的一致性。HPV Blot對個別基因型的敏感度是100%有HPV 6, 11, 31, 35, 45, 59及66；95~99%有HPV 16, 18, 33及58；90~94%有HPV 51, 52, 53及70；80~89%有HPV 39, 56及68；低於79%有HPV 62及CP8061。HPV Blot對個別基因型的特異性是100%有HPV 6, 11, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 62, 66, 68, 70及CP8061；95~99%有HPV 16。HPV Blot對個別基因型的準確性是100%有HPV 6, 11, 35, 45, 59及66；95~99%有HPV 16, 18, 31, 33, 39, 51, 52, 53, 56, 58, 62, 68, 70及CP8061。HPV Blot對個別基因型的一致性是完美的(Kappa=1)有HPV 6, 11, 35, 45及59；優良的(0.99 < Kappa < 0.81)有HPV 16, 18, 31, 33, 39, 51, 52, 53, 58, 66, 68, 70, 62及56；令人滿意的(0.8 < Kappa < 0.61)有HPV CP8061。
所有的實驗均嚴格遵守品質保證系統的四項規範。因此，100批次實驗中每一批次其3個HPV陰性外部控制組的結果都是HPV陰性，表示沒有HPV偽陽性的結果。Caski、HeLa及Jurket細胞株在2000個細胞量的規模中對於HPV 16、HPV 18及GAPDH的基因定型結果有100%的靈敏度與特異性。5個外部控制組的結果符合品質保證系統規範的需求，顯示HPV Blot在操作上有優良的可靠度。批次內與批次間的再現性分別為98%與97%。HPV Blot之基因定型結果需先判斷其三個內部控制組為合格，其基因定型結果才能確認。本研究之基因定型鑑定結果的品質得以保證。
HPV Blot用於HPV感染之臨床與流行病學的研究是一種高靈敏度的、準確的、可靠的及可再現的工具。本研究對於HPV基因定型導入品質保證系統，它包含7個外部控制組與3個內部控制組，對於實驗與實驗室操作有適當的規範並證明有效。由本研究的發現證實了HPV Blot對於臨床之HPV基因定型、專一性型別HPV持續感染的追蹤及專一性型別HPV疫苗的開發有很大的價值。

關鍵字

人類乳突瘤病毒 ； 效能 ； 效度 ； 信度 ； 敏感度 ； 特異性 ； 正確性 ； 品質保證

並列摘要

The aim of this study was to evaluate the performance, validity and reliability of the EasyChip® HPV Blot for Human papillomavirus (HPV) DNA genotyping in cervical swab samples, and assured the quality of HPV typing in large-scale community-based cervical neoplasia screening by conducting a quality assurance system. The performance of HPV Blot was assessed by analytic sensitivity、analytic specificity and sensitivity. The analytic sensitivity and specificity of HPV Blot was assessed by using 39 HPV genotype plasmids, and assessed sensitivity by using cell lines and cervical specimens, respectively. The 39 types of HPV (HPV 6, 11, 16, 18, 26, 31, 32, 33, 35, 37, 39, 42, 43, 44, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 66, 67, 68, 69, 70, 72, 74, 82, CP8061, CP8304, L1AE5, MM4, MM7, and MM8) were assessed using plasmids DNA containing L1 region of HPV genome. Each plasmid was considered as a genome equivalent (geq). A clinical trial was conducted to assess the validity and reliability of HPV Blot. The gold standard of trial was 20 L1 type-specific parallel PCR. The assessment of validity was sensitivity, specificity and accuracy. The assessment of reliability was McNemar’s Test and Kappa statistic. The way of assessment was by subject and by type, respectively. The quality assurance system includes the maintenance of operating environment, assessment of reliability, assessment of reproducibility, and blot interpretation. The quality assurance standards include the assessment of seven extrinsic controls in operation and three intrinsic controls of HPV Blot in genotypes identification. Each batch experiment contained samples from 89 cervical specimens and seven extrinsic controls. The seven extrinsic controls were Caski cells, HeLa cells, Jurkat cells, male human blood cell DNA, sterile water, and paired sibling controls. The replicated aliquots of the same sample were a pair of sibling controls. Caski cells, HeLa cells, Jurkat cells and sibling controls were used simultaneously to assess entire procedures inclusive of DNA extraction, PCR, and HPV typing. Male human blood cell DNA and sterile water were used only in the assessment of PCR and HPV typing. One hundred batch experiments were conducted. Jurkat cells and male human blood DNA were used as an index of potential contamination with exogenous HPV sequences. Sterile water was used as an index of potential contamination with any exogenous DNA sequences.
The detection limits for analytical sensitivities of HPV Blot are 1 geq per PCR for HPV 16, 10 geq per PCR for HPV 6, 26, 31, 33, 37, 39, 43, 44, 45, 54, 55, 61, 67, 68, 69, 82, CP8061, CP8304, L1AE5, MM7 and MM8, 20 geq per PCR for HPV 11, 18, 32, 35, 42, 51, 52, 53, 56, 58, 59, 66, 70, 74 and MM4, and 50 geq per PCR for HPV 62 and 72. The overall analytical sensitivity of HPV detection is 1 to 50 copies of HPV geq for each sample. The analytic specificity of the HPV Blot is each of 39 HPV genotype probes that had no cross-reactivity with amplicons of other HPV genotypes. In cell lines, the detection limits for Caski and HeLa cells were 10–4 and 10–3 ng DNA, respectively, by the HPV Blot and 10–3 and 10–2 ng DNA, respectively, by the gel. The calculated copies are 1~10 for HPV 16 and 2~8 for HPV 18. Both copy numbers were expected on the basis of known HPV 16 and HPV 18 copy numbers in Caski cells (60–600 copies per cell) and HeLa cells (10–50 copies per cell). About 10 copies of HPV 16 and 18 viruses within cervical specimens were detected.
The sensitivity, specificity and accuracy of HPV Blot by subject are 97, 96 and 96.5%, respectively. The McNemar’s Test、Kappa and P value of HPV Blot by subject are 0.25, 0.93 and 0.62, respectively. The sensitivity, specificity and accuracy of HPV Blot by type are 90, 92 and 90.6%, respectively. The McNemar’s Test、Kappa and P value of HPV Blot by type are 2.81, 0.81 and 0.10, respectively. The comparison of HPV Blot and gold standard suggests that is excellent agreement. The sensitivity of HPV Blot by type is 100% for HPV 6, 11, 31, 35, 45, 59 and 66, 99~95% for HPV 16, 18, 33 and 58, 94~90% for HPV 51, 52, 53 and 70, 89~80% for HPV 39, 56 and 68, and less than 79% for HPV 62 and CP8061. The specificity of HPV Blot by type is 100% for HPV 6, 11, 18, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 62, 66, 68, 70 and CP8061, and 99~95% for HPV 16. The accuracy of HPV Blot by type is 100% for HPV 6, 11, 35, 45, 59 and 66, and 99~95% for HPV 16, 18, 31, 33, 39, 51, 52, 53, 56, 58, 62, 68, 70 and CP8061. The type-specific agreement of HPV Blot is perfect (Kappa=1) for HPV 6, 11, 35, 45 and 59, excellent (0.99 < Kappa < 0.81) for HPV 16, 18, 31, 33, 39, 51, 52, 53, 58, 66, 68, 70, 62 and 56, good (0.8 < Kappa < 0.61) for HPV CP8061.
All experiments obeyed 4 criteria of quality assurance system as follows. Thus, all three HPV-negative extrinsic controls were HPV negative in 100 batch experiments. In other words, there was no HPV false-positive result. The HPV typing results for Caski, HeLa, and Jurkat cells indicated that the HPV Blot had 100% sensitivity and specificity for HPV 16, HPV 18, and GAPDH at a scale of 2000 cells. That five extrinsic controls met the requirements of quality assurance system in 100 batch experiments indicate that the HPV Blot had excellent reliability in operation. Intra-batch and inter-batch reproducibility of the HPV Blot were 98% and 97%, respectively. Reader identified three intrinsic controls of HPV Blot, so typing results of HPV Blot were confirmed. In the study, the qualities of identification were qualified.
In conclusion, the HPV Blot is a highly sensitive, accurate, reliable and reproducible tool for clinical and epidemiological studies on HPV infection. This study is conducted a quality assurance system comprising seven extrinsic controls and three intrinsic controls for HPV typing, which has been validated as an appropriate criterion for experiments and laboratory operation. These findings confirm that the HPV Blot is of great value for clinical HPV typing, follow-up of type-specific HPV persistence and development of type-specific HPV vaccines.

並列關鍵字

Human papillomavirus ； Performance ； Validity ； Reliability ； Sensitivity ； Specificity ； Accuracy ； Quality Assurance

參考文獻

Autillo-Touati A, Joannes M, d'Ercole C, Robaglia-Schlupp A, Lambert A, Mazzella E, Blanc B, Seite R. 1998. HPV typing by in situ hybridization on cervical cytologic smears with ASCUS. Acta Cytol 42(3):631-638.

Bulten J, Melchers WJ, Kooy-Smits MM, de Wilde PC, Poddighe PJ, Robben JC, Macville MV, Massuger LF, Bakkers JM, Hanselaar AG. 2002. Numerical aberrations of chromosome 1 in cervical intraepithelial neoplasia are strongly associated with infection with high-risk human papillomavirus types. J Pathol 198(3):300-309.

Burchell AN, Winer RL, de Sanjose S, Franco EL. 2006. Chapter 6: Epidemiology and transmission dynamics of genital HPV infection. Vaccine 24 Suppl 3:S52-61.

Burd EM. 2003. Human papillomavirus and cervical cancer. Clin Microbiol Rev 16(1):1-17.

Castle PE, Lorincz AT, Scott DR, Sherman ME, Glass AG, Rush BB, Wacholder S, Burk RD, Manos MM, Schussler JE, Macomber P, Schiffman M. 2003. Comparison between prototype hybrid capture 3 and hybrid capture 2 human papillomavirus DNA assays for detection of high-grade cervical intraepithelial neoplasia and cancer. J Clin Microbiol 41(9):4022-4030.