Human papillomavirus genotype in cervical cancer: A population-based study

Authors

  • Chyong-Huey Lai,

    Corresponding author
    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    • Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital Linkou Medical Center, 5 Fu-Shin St. Kueishan, Taoyuan 333, Taiwan
    Search for more papers by this author
    • Fax: +886-3-328-8252.

  • Huei-Jean Huang,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Swei Hsueh,

    1. Department of Pathology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Angel Chao,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Cheng-Tao Lin,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Shang-Lang Huang,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Fang-Yu Chao,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Jian-Tai Qiu,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Ji-Hong Hong,

    1. Department of Radiation Oncology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Hung-Hsueh Chou,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Ting-Chang Chang,

    1. Department of Obstetrics and Gynecology, Chang Gung Memorial Hospital and Chang Gung University College of Medicine, Taoyuan, Taiwan
    Search for more papers by this author
  • Chee-Jen Chang

    1. Graduate Institutes of Basic Medical Sciences, Chang Gung University, Taoyuan, Taiwan
    Search for more papers by this author

  • The abstract of this article was presented in the European Research Organization on Genital Infection and Neoplasia, Paris, France, April 22–26, 2006.

Abstract

Our aim was to investigate the human papillomavirus (HPV) genotype distribution and correlation between HPV parameters and clinicopathological variables in cervical carcinoma treated in a large tertiary referral medical center in Taiwan. Consecutive patients treated for cervical carcinoma (Stages I–IV according to the International Federation of Gynecology and Obstetrics) between 1993 and 2000 were included. HPV genotyping using SPF1/GP6+ PCR was performed, followed by hybridization with a genechip (Easychip® HPV Blot, King Car, Taiwan). E6 type-specific PCR was performed to validate multiple-type. HPV-negative samples were further verified by type-specific PCR and a repeat HPV Blot. A total of 2,118 patients were eligible for analysis. HPV DNA sequences were detected in 96.6% (95% CI, 95.8–97.4%) of the specimens, among which 82% harbored single-type and 18% contained multiple-type HPV sequences. Thirty-five types of HPV were identified and the leading 8 were HPV16 (50.0%), HPV18 (17.8%), HPV58 (16.3%), HPV33 (8.7%), HPV52 (6.8%), HPV39 (3.0%), HPV45 (2.5%) and HPV31 (2.3%). HPV58 or 33 or 52 was detected in 30.3% (641/2,118). By multivariate analysis, HPV58- or 33- or 52-infection was significantly associated with older age (p < 0.001) and primary radiotherapy or concurrent chemoradiation (RT/CCRT) (p < 0.001). Among HPV-positive cases, multiple-type was more frequently seen in those receiving primary RT/CCRT (p < 0.001). The knowledge of HPV genotype distribution will form a basis for guidelines in HPV-based cervical cancer screening and cost-effective multivalent HPV vaccine policy in Taiwan and in the world. The association between HPV parameters and clinicopathological variables warrants further investigations. © 2007 Wiley-Liss, Inc.

Cervical cancer is one of the most common female genital cancers worldwide.1 There is strong epidemiological and molecular evidence indicating that human papillomavirus (HPV) infection is a necessary event in the development of cervical intraepithelial lesions and subsequent invasive carcinoma.2, 3, 4, 5 Anogenital HPVs have been divided into risk groups according to the frequency of their presence in cervical carcinoma. Low-risk HPV types include HPV6, 11, 40, 42, 43, 44, 54, 61, 70, 72, 81 and CP6108. High-risk HPV types are represented essentially by HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68 and 82 (MM4). The third group, called probably carcinogenic types, is represented by HPV26, 53 and 66.2

With the advancement of technology such as polymerase chain reaction (PCR), the sensitivity of detecting HPV DNA has been possible for as low as 1–50 copies of virus in a sample. Several type-specific PCR primer sets and general (consensus) primer sets such as MY09/MY11 and GP5+/GP6+ are widely used. Although the sensitivity of in situ hybridization is limited, it permits localization of HPV DNA in the sample.6 However, even with the combinations of type-specific and general primer sets, the sensitivity is limited especially in the setting of paraffin-embedded tissue.7 Generally, the efficiency of PCR primers generating a smaller product is higher than those yielding a larger amplimer.6 A study found that GP5+/GP6+ exhibited poor sensitivity for HPV52.8 Results of an international collaborative study on the proficiency of various technologies in the detection of HPV DNA showed that quality assays were generally consistent across the 29 participating laboratories, but HPV31 was found to be the least accurately detected.9

In an international multicenter study of 3,607 cervical cancers, the 15 most common types were, in descending order of frequency, HPV16, 18, 45, 31, 33, 52, 58, 35, 59, 56, 39, 51, 73, 68 and 66.10 A meta-analysis including 85 studies, using PCR to estimate HPV prevalence in invasive cervical cancer (n = 10,058), also reached a similar conclusion in all regions except Asia (n = 3,091), where HPV Types 58 and 52 were more frequently identified.11

The relation of HPV genotype to clinical parameters of invasive cervical cancer has been an issue of controversy. Some found that HPV18-positive tumors were associated with poorer prognosis when compared with HPV16-positive ones.12, 13, 14 Some reported that HPV31 related types predicted better survival,15 whereas others did not find a prognostic value for HPV genotype.5 Obviously, further studies with adequate sample size and sensitivity in detecting HPV DNA are necessary to draw a conclusion.

Knowledge of population-based HPV genotype distribution is important for the development of HPV-based cervical cancer screening and for the assessment of the effect of prophylactic vaccination on HPV infections, but these data are limited for many world regions. In this study, we investigate the distribution of HPV genotypes and correlation between HPV genotype and clinicopathological variables in consecutive patients with cervical cancer (Stages I–IV according to the International Federation of Gynecology and Obstetrics (FIGO)) treated in a tertiary referral medical center in Taiwan.

Methods

Study population and samples

Through a search of the disease code database (International Classification of Diseases of Oncology (ICDO)) and Systematized Nomenclature of Medicine (SNOMED) code for cervical cancer and primary treatment from August 1993 through May 2000 at Chang Gung Memorial Hospital, all the consecutive cases with FIGO Stages I–IV cervical cancer primarily treated with surgery (usually radical hysterectomy and pelvic lymphadenectomy [RH-PLND]) or definitive radiotherapy or concurrent chemoradiation (RT/CCRT) were checked for medical records and pathological diagnosis. The institutional review board of the Chang Gung Memorial Hospital approved this study. Formalin-fixed paraffin-embedded tissue specimens were used for DNA analysis. Those who had a wrong diagnosis or incomplete medical records, with missing paraffin blocks, without tumor cells left in the histology slides or blocks, and with inadequate DNA quality were eliminated, and the remaining patients were eligible for the current study (Fig. 1).

Figure 1.

Study profile.

DNA extraction

Two paraffin embedded tissue sections of every specimen (10 μm thick) were deparaffinized through xylenes and rehydrated through serial alcohol. Sections were finely minced with disposable razor blades and digested in a solution of proteinase K (1 mg/ml) in 50 mmol/l Tris buffer (pH 7·8) for 16–20 hr at 55°C. Proteinase K was denatured by heating samples at 95°C for 10 min. Samples were further purified through DNeasy Tissue Kit (QIAGEN, Valencia, CA). Finally, 50 μl of DNA solution was eluted and 1 μl of the aliquot was used for PCR amplification.16, 17

SPF1/GP6+ PCR

The SPF1/GP6+16, 17 consensus primers (Table I) were used to amplify a fragment of ∼184 bp in the L1 open reading frame. Primer GP6+ was biotinylated at 5′ end. PCR conditions were as follows: preheating for 10 min at 94°C, followed by 40 cycles of 1 min at 94°C, 1 min at 45°C, and 1 min at 72°C and a final extension of 5 min at 72°C. Each PCR experiment was performed with several positive and negative controls. Routine precautionary procedures were applied to avoid carrying-over or contamination.16, 17

Table I. Primer Sequences and Sizes of Amplimers
PrimerAmplimer (bp)Sequences (5′–3′)
SPF1A GCICAGGGICACAATAATGG
SPF1B GCICAGGGICATAACAATGG
SPF1C GCICAGGGICATAATAATGG
SPF1D GCICAAGGICATAATAATGG
GP6+ GAAAAATAAACTGTAAATCATATTC
GAPDHF136GGCAGCAGCAAGCATTCCT
GAPDHR GCCCAACACCCCCAGTCA
11E6–F252CGTGTGCCTGTTGCTTAGAA
11E6–R CTTCCATGCATGTTGTCCAG
16E6–F204TTGCTTTTCGGGATTTATGC
16E6–R CAGGACACAGTGGCTTTTGA
18E6–F258CACTTCACTGCAAGACATAG
18E6–R CACCGCAGGCACCTTATTA
31E6–F202CACACGGAGTGTGTACAAAATGT
31E6–R TGGAATCGTTTCTTTTTATCCAA
32E6-F246TGCCTCATCACAGCCAAGTA
32E6-R TGCCAAAATGCTGATCTGTC
33E6–F191AAACCACGAACATTGCATGA
33E6–R GATAAGAACCGCAAACACAG
35E6–F285ACAGCGGAGTGAGGTATATG
35E6–R CACCGTCCACCGATGTTATG
39E6–F178TTACTCGGACTCGGTGTATG
39E6–R TCGACACTGTCCTGTATAGC
42E6–F228GCACTGCGACACTACGAAAG
42E6–R TTAGGGTAGGCGTCTCTCCA
45E6–F212GCAACATTGGAACGCACAGA
45E6–R CGCAGGCACCTTATTAACAAAT
51E6–F283TCCATATGCAGTATGCAAACAAT
51E6–R TTACACTTGGGTTTCGTTACGTT
52E6–F198GCGTGTGTATTATGTGCCTACG
52E6–R TATGAAATCGCTTGTTTGCATT
53E6–F262ACGGGTATCCGTATGGAGTG
53E6–R TGTGTGTCTCCAGCATGTCA
55E6-F139TTTGACTTCGCGGGATATGC
55E6-R ATCGCGCCTTCTCCAATATG
56E6–F232AGAACTAACACGTGCTGAGG
56E6–R CGGAGTTAACGGACTTTGAC
58E6–F205ATGGAAATCCATTTGCAGTATGT
58E6–R CTTTTGTTTAAATCCACATGCCT
59E6–F258CAAACTGCCTGATTTGAGCA
59E6–R AACGGTGTCTTGGTTTCAGC
67E6–F240TGGGGTATGTAAGCAATGCC
67E6–R CAACACACTGAACACCGTCC
68E6–F205GGACACTACATTGCATGACGTT
68E6–R TGTAGTTGCATACACCGATTCC
69E6–F257CCCAGAGAAAGACCACGAAC
69E6–R GCTTCCAGTGTTGCACCATA
70E6–F232AGAACGGCCATACAAATTGC
70E6–R CACCGAGTTCGAATAATGCC
74E6-F280TTGCAGGAAAACCTTGTCTGT
74E6-R AAATCTTGCCCTTTCCACAAT
82E6–F189CCATATGCAGCATGCAAAAA
82E6–R GTCGTCCACCACCTTTTGC
CP8061E6-F233AAAGTGGTTTCCGTTTGCTG
CP8061E6-R GCCAATGGCCTGCTATTTTA
CP8304E6-F223AGAGAGCTAAACTTGGTGTGGC
CP8304E6-R GGCCTGTCACAAGATACTCCTT

HPV genotying by genechip

Fifteen microliters of the resultant amplimers were then hybridized with an Easychip® HPV Blot (King Car, I-Lan, Taiwan) (hereafter HPV Blot) membrane. Thirty-eight types of 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, 71 [CP8061], 72, 74, 81 [CP8304], 82 [MM4], 83 [MM7], 84 [MM8], L1AE5) oligonucleotide probes of 20- to 30-mer with an ∼100- to 200-mer poly-T tail were immobilized on a nylon membrane in a single reaction.16, 17

Hybridization procedures were described previously.16, 17 After hybridization, the membranes were washed and incubated in 500 μl of buffer containing Avidex-AP™ (alkaline phosphatase conjugates and biotinylated antibodies, 1:1,000 dilution) for 40 min. Then 70 μl of substrate NBT/BCIP (5-bromo-4-chloro-3-indolyl-phosphate and nitroblue tetrazolium) was added and incubated for 30 min at room temperature. After drying, the results were determined by the eye according to the HPV type format on the HPV Blot membranes.16, 17

HPV genotyping by direct sequencing analyses

PCR products amplified by SPF1/GP6+ were isolated from agarose gels. Sequencing reaction of double-stranded PCR products were performed with a dye-labeled terminator cycle sequencing kit (ABI PRISM® dRhodamine Terminator Cycle Sequencing Ready Reaction Kit, Applied Biosystems) The DNA sequences obtained from the patient samples were compared with the GenBank sequences by using BLAST Program at the National Center for Biotechnology Information.16, 17

Type-specific PCR

Type-specific primer sequences and amplimer sizes of the 25 HPV genotypes are listed in Table I. Type specific PCR was performed for 50 cycles of 1 min at 94°C, 1 min at 55°C and 1 min at 72°C. A final extension of 5 min at 72°C was added after 50 cycles. Total yield of each PCR product was analyzed by electrophoresis on 2% agarose gels stained with ethidium bromide. Because direct sequencing usually failed to resolve multiple infections, E6 type-specific PCRs were performed to validate multiple types on HPV Blot.

External control GAPDH PCR

Quality of isolated DNA was checked with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) PCR, by which a 136-bp product was amplified. Forty cycles were performed. Because coamplification of house-keeping gene primers could reduce the efficiency of target HPV DNA amplification,17 GAPDH PCR was performed when HPV-negative results were obtained at the first round (Fig. 1).

Validation of HPV-negative results and rare types

If GAPDH PCR was also negative, the histology slides were checked for presence of adequate tumor cells, and another block of the hysterectomy or conization specimens was used when appropriate. In case of GAPDH-positive and HPV-negative by both type-specific PCR of the leading 8 types (determined by the preliminary data) and a repeat of HPV Blot, the final diagnosis was designated HPV-negative. For types detected in not more than 5 cases (called rare types), type-specific PCR (for multiple types) and direct sequencing (for single type) was performed (Fig. 1). In case of discordance between the direct sequencing and HPV Blot results, a repeat of HPV Blot and type-specific PCR was performed to resolve the discrepancy.

Statistical analysis

The data were analyzed by the SPSS 11.0 statistical package. Pearson's χ2 test was used to evaluate the proportion differences between designated groups. Their odds ratios and 95% confidence intervals (CI) were calculated. Multiple logistic regression was used for those covariates selected in univariate analyses with a p-value less than 0.05. Continuous covariates were compared between groups using both parametric and nonparametric approaches, mainly, the Student's t-test and Mann–Whitney U test when appropriate. All tests were two-sided, and a p-value of <0.05 was considered statistically significant.

Results

Between August 1993 and May 2000, there were 2,446 cases with FIGO stages I–IV cervical cancer primarily treated with surgery or definitive RT/CCRT at Chang Gung Memorial Hospital. Of the 2,446 patients, 161 had a wrong diagnosis or incomplete medical records, 154 had missing paraffin blocks. One hundred forty-six patients had an initial HPV-negative result, while 1,985 cases had a HPV-positive result. Of these 146 HPV-negative patients, 24 were GAPDH-negative, among which 13 did not have tumor cells left in the histology slides (received conization outside and hysterectomy specimens were free of residual tumor) or blocks (those available did not contain tumor), while the other 11 specimens were reamplified on another block which showed GAPDH-positive. After repeat HPV Blot and E6 type-specific PCR, 72 of the 133 cases remained HPV-negative and 61 were HPV-positive. A total of 2,118 patients (72 HPV-negative and 2,046 HPV-positive) were eligible for the subsequent analysis (Fig. 1).

The clinicopathological characteristics of the study patients are summarized in Table II. Of the 2,118 eligible patients, HPV DNA sequences were detected in 2,046 (96.6% [95% CI, 95.8–97.4%]) of the specimens, among which 82% harbored single-type and 18% contained multiple-type HPV sequences. HPV genotype distribution is depicted in Table III. The leading 10 types were HPV16 (50.0%), HPV18 (17.8%), HPV58 (16.3%), HPV33 (8.7%), HPV52 (6.8%), HPV39 (3.0%), HPV45 (2.5%), HPV31 (2.3%), HPV51 (1.2%) and HPV70 (1.1%). Especially, HPV58 or 33 or 52 was detected in 30.3% (641 of 2,118) of Stages I–IV cervical cancer tissue. About 293 of the 2,118 (13.8%) cervical cancer specimens contained double types, 64 (3.0%) triple types, 10 (0.5%) quadruple types, and 1 case each of 5 and 7 types (Table IV). Among the leading 10 types, HPV16 was significantly associated single type (83.6%) than the other 9 types (34.8%–68.7%) (Table III). HPV16/18, HPV16/58 and HPV18/58 were the most common double types, while HPV16/18/33 and HPV16/18/58 were the most common triple types (Table IV).

Table II. Patient Characteristics (N = 2,118)
 Number (%)
  1. FIGO, International Federation of Gynecology and Obstetrics.

Age (median (range)) (years)55 (26–97)
 <35101 (4.8)
 35–50723 (34.1)
 >501,294 (61.1)
FIGO stage
 I1,146 (54.1)
 II656 (31.0)
 III288 (13.6)
 IV28 (1.3)
Histologic type
 Squamous carcinoma1,847 (87.2)
 Adenoadenosquamous carcinoma259 (12.2)
 Small cell carcinoma12 (0.6)
Primary treatment
 Surgery1,072 (50.6)
 Radiotherapy or chemoradiation1,046 (49.4)
Table III. HPV Genotype Distribution (N = 2,118)
 No. of patients (%)1Multiple typeSingle type
  • NA, not applicable.

  • 1

    Same woman can be counted more than once because of multiple infections.

  • 2

    HPV 58 or 33, or 52 was identified in 30.3% (641/2,118) of the study samples.

Total2,118 (100)  
HPV (−)72 (3.4)NANA
HPV (+)2,046 (96.6)18.0% (369/2,046)82.0% (1,677/2,046)
HPV161,059 (50.0)16.4% (174/1,059)83.6% (885/1,059)
HPV18376 (17.8)44.4% (167/376)55.6% (209/376)
HPV582346 (16.3)41.0% (142/346)59.0% (204/346)
HPV332184 (8.7)41.3% (76/184)58.7% (108/184)
HPV522144 (6.8)31.3% (45/144)68.7% (99/144)
HPV3964 (3.0)59.4% (38/64)40.6% (26/64)
HPV4554 (2.5)48.1% (26/54)51.9% (28/54)
HPV3148 (2.3)41.7% (20/48)58.3% (28/48)
HPV5125 (1.2)60.0% (15/25)40.4% (10/25)
HPV7023 (1.1)65.2% (15/23)34.8% (8/23)
HPV5622 (1.0)77.3% (17/22)22.7% (5/22)
HPV5922 (1.0)50.0% (11/22)50.0% (11/22)
HPV3519 (0.9)31.6% (6/19)68.4% (13/19)
HPV5315 (0.7)46.7% (7/15)53.3% (8/15)
HPV4214 (0.7)78.6% (11/14)21.4% (3/14)
HPV84 (MM8)11 (0.5)81.8% (9/11)18.2% (2/11)
HPV82 (MM4)10 (0.5)50.0% (5/10)50.0% (5/10)
HPV619 (0.4)88.9% (8/9)11.1% (1/9)
HPV679 (0.4)44.4% (4/9)55.6% (5/9)
HPV687 (0.3)57.1% (4/7)42.9% (3/7)
HPV697 (0.3)28.6% (2/7)71.4% (5/7)
HPV265 (0.2)20.0% (1/5)80.0% (4/5)
HPV665 (0.2)60.0% (3/5)40.0% (2/5)
HPV445 (0.2)80.0% (4/5)20.0% (1/5)
HPV544 (0.2)75.0% (3/4)25.0% (1/4)
HPV L1AE52 (0.1)0% (0/2)100% (2/2)
HPV372 (0.1)50% (1/2)50% (1/2)
HPV621 (0.05)0% (0/1)100% (1/1)
HPV721 (0.05)0% (0/1)100% (1/1)
HPV114 (0.2)100% (4/4)0% (0/4)
HPV321 (0.05)100% (1/1)0% (0/1)
HPV553 (0.1)100% (3/3)0% (0/3)
HPV71 (CP8061)1 (0.05)100% (1/1)0% (0/1)
HPV743 (0.1)100% (3/3)0% (0/3)
HPV81 (CP8304)2 (0.1)100% (2/2)0% (0/2)
Table IV. Distribution of HPV Multiple Infections (N = 369)
 No. of patients
  1. The rare types existent only as multiple types are highlighted using block letters, and the cases not verified by E6 type-specific PCR is annotated with a “*”.

Double types293
 HPV16/18, HPV16/58, HPV16/33, HPV16/52,44, 37,15, 7
 HPV16/45, HPV16/31, HPV 16/39, HPV16/51, HPV16/56,6, 5, 5, 4, 3
 HPV 16/11, HPV16/35, HPV16/37, HPV16/42, HPV16/44,2, 1, 1, 1, 1,
 HPV16/53, HPV16/55, HPV16/59, HPV16/61, HPV16/70,1, 1, 1, 1, 1
 HPV16/81(CP8304), HPV16/82 (MM4), HPV16/84 (MM8)1, 2, 1
 HPV18/58, HPV18/33, HPV18/52, HPV18/4521, 13, 10, 8
 HPV18/31, HPV18/39, HPV18/51, HPV18/53, HPV18/56,3, 3, 2, 2, 2
 HPV18/61, HPV 18/11*, HPV18/35, HPV18/55, HPV18/59,2, 1, 1, 1, 1
 HPV18/84 (MM8)2
 HPV58/39, HPV58/52, HPV58/33, HPV58/31,12, 5, 5, 4
 HPV58/35, HPV58/42, HPV58/51,3, 3, 3
 HPV58/45, HPV58/56, HPV58/70, HPV58/82 (MM4)2, 2, 2, 2
 HPV58/44, HPV58/67, HPV58/68, HPV58/691, 1, 1, 1
 HPV52/33, HPV52/42, HPV52/53, HPV53/565, 1, 1, 1
 HPV33/39, HPV33/452, 2
 HPV33/42, HPV33/56, HPV33/66, HPV33/70, HPV33/74*1, 1, 1, 1, 1
 HPV31/67, HPV31/70, HPV31/84 (MM8)1, 1, 1
 HPV39/45, HPV39/59, HPV39/84 (MM8)1, 1, 1
 HPV51/33, HPV51/52, HPV51/591, 1, 1
 HPV53/59, HPV53/681, 1
 HPV42/71 (CP8061), HPV52/84 (MM8), HPV54/70,1, 1, 1
 HPV56/68, HPV61/701, 1
Triple types64
 HPV16/18/33, HPV16/18/58, HPV16/18/45, HPV16/39/587, 6, 6, 3
 HPV16/18/52, HPV16/45/58, HPV16/18/562, 2, 1
 HPV16/31/61, HPV16/33/51, HPV16/33/67, HPV16/42/591, 1, 1, 1
 HPV18/33/58, HPV18/33/39, HPV18,33/52, HPV18/26/334, 2, 2, 1
 HPV18/31/52, HPV18/31/58, HPV18/33/56, HPV18/33/591, 1, 1, 1
 HPV18/39/58, HPV18/51/70, HPV18/52/81 (CP8304)*1, 1, 1
 HPV18/54/74, HPV18/56/70, HPV18/58/691, 1, 1
 HPV11/18/581
 HPV31/32/52, HPV31/44/511, 1
 HPV33/39/58, HPV33/52/61, HPV33/58/74*1, 1, 1
 HPV35/39/581
 HPV39/42/58, HPV39/52/58, HPV39/56/58, HPV39/58/671, 1, 1, 1
 HPV52/56/581
 HPV58/61/701
 HPV59/70/84 (MM8)1
Quadruple types10
 HPV16/18/33/45, HPV16/18/39/70, HPV16/18/42/581, 1, 1
 HPV18/33/54/58, HPV18/33/58/ 84 (MM8), HPV18/33/58/591, 1, 1
 HPV18/52/56/58, HPV18/53/66/681, 1
 HPV52/58/70/84 (MM8)1
 HPV58/59/61/701
5 types1
 HPV18/44/59/66/701
7 types1
 HPV16/18/42/52/55*/56/581

A total of 35 types of HPV were detectable in this study. Of the 35 types, 21 types were demonstrable by HPV Blot in more than 5 cases. The other 8 types (HPV26, 66, 44, 54, L1AE5, 37, 62 and 72) demonstrable in not more than 5 cases were regarded as true existent because they were present as single-type both by HPV Blot and direct sequencing, while the remaining 6 types detectable in not more than 5 cases (HPV11, 32, 55, 71[CP8061], 74, 81[CP8304]) were only coexistent with the other more prevalent types (Tables III and IV). The samples containing the latter 6 types were not all verified by E6 type-specific PCR in each, but their existence was confirmed (Table IV). Because they did not exist as a single type, the carcinogenic potential of the last 6 types was unproven in the current study.

Associations of HPV parameters with various clinicopathological characteristics are summarized in Table V. By Pearson's χ2 analysis, age (cutoff at 35 years, p = 0.010; cutoff at 50 years, p = 0.007), FIGO stage (p = 0.009), and mode of primary treatment (p< 0.001) were significantly related to HPV status (positive or negative), and HPV18-positivity was significantly more common in adeno-adenosquamous (37.5%) or small cell carcinoma (33.3%) than squamous carcinoma (14.9%) (p < 0.001). HPV58- or 33- or 52-infection was significantly associated with age (≤35 vs. >35: 10·9% vs. 31·2%, p < 0.001; ≤50 vs. >50: 16.5% vs. 39.0%, p < 0.001), stage (I–II vs. III–IV: 27.0% vs. 48.7%, p < 0.001), histologic type (squamous vs. adeno-adenosquamous vs. small cell: 31.9% vs. 17.8% vs. 41.7%, p < 0.001) and mode of primary treatment (surgery vs. RT: 16.4% vs. 44.5%, p < 0.001). Among those who had PLND at primary surgery (n = 1,066), the incidence of lymph node metastasis was not different according to HPV status, presence of HPV18, or presence of 58 or 33 or 52 (Table V).

Table V. Association of HPV Parameters with Clinicopathological Variables (N = 2,118)
 HPV genotype
HPV(+) n (%)HPV(–) n (%)PHPV18 (+) n (%)HPV18 (–) n (%)pHPV52 or 58 or 33 (+)1n (%)HPV52& 58& 33 (–)2n (%)p
  • FIGO, International Federation of Gynecology and Obstetrics; RT, radiotherapy; CCRT, concurrent chemoradiation.

  • 1

    HPV58 or 52 or 33 positive.

  • 2

    HPV58, 52, and 33 negative.

  • 3

    Tumor size estimated by histopathology.

Age (years)
 ≤3593 (92.1)8 (7.9)0.01013 (12.9)88 (87.1)0.18811 (10.9)90 (89.1)<0.001
 >351,953 (96.8)64 (3.2) 363 (18.0)1654 (82.0) 630 (31.2)1387(68.8) 
 ≤50785 (95.3)39 (4.7)0.007162 (19.7)662 (80.3)0.067136 (16.5)688 (83.5)< 0.001
 >501,261 (97.4)33 (2.6) 214 (16.5)1080 (83.5) 505 (39.0)789 (61.0) 
Primary tumor size (n = 991)3
 <2 cm439 (96.1)18 (3.9)0.24682 (17.9)375 (82.1)0.38585 (18.6)372 (81.4)0.089
 2-4 cm459 (94.1)29 (5.9) 73 (15.0)415 (85.0) 66 (13.5)422 (86.5) 
 >4 cm45 (97.8)1 (2.2) 6 (13.0)40 (87.0) 6 (13.0)40 (87.0) 
Cell type
 Squamous1,788 (96.8)59 (3.2)0.256275 (14.9)1572 (85.1)<0.001590 (31.9)1257 (68.1)<0.001
 Adeno-adenosquamous246 (95.0)13 (5.0) 97 (37.5)162 (62.5) 46 (17.8)213 (82.2) 
 Small cell12 (100.0)0 (0.0) 4 (33.3)8 (66.7) 5 (41.7)7 (58.3) 
FIGO stage
 I–II1,733 (96.2)69 (3.8)0.009318 (17.6)1484 (82.4)0.761487 (27.0)1315 (73.0)<0.001
 III–IV313 (99.1)3 (0.9) 58 (18.4)258 (81.6) 154 (48.7)162 (51.3) 
Primary treatment
 Surgery1,019 (95.1)53 (4.9)<0.001176 (16.4)896 (83.6)0.104176 (16.4)896 (83.6)<0.001
 RT/CCRT1,027 (98.2)19 (1.8) 200 (19.1)846 (80.9) 465 (44.5)581 (55.5) 
Lymph node metastasis (n = 1,066)
 Yes94 (96.9)3 (3.1)0.37211 (11.3)86 (88.7)0.16416 (16.5)81 (83.5)0.997
 No919 (94.8)50 (5.2) 163 (16.8)806 (83.2) 160 (16.5)809 (83.5) 

Among HPV-positive cases (n = 2,046), multiple-type infection was more frequently seen in older patients (>35 vs. ≤35 or >50 vs. ≤50-years old), advanced stage (Stages III–IV vs. I–II or IIB–IV vs. I–IIA), those receiving primary RT/CCRT and absence of lymph metastasis among those receiving primary RH-PLND (n = 1,013) (Table VI). By multivariate logistic regression analysis (age, FIGO stage, histologic type and primary treatment included), HPV58- or 33- or 52-infection was significantly associated with age (>50 vs. ≤50 years: odds ratio 2.29 [1.83–2.88], p < 0.001) and mode of primary treatment (RT/CCRT vs. surgery: odds ratio 3.03 [2.42–3.80], p < 0·001). Among HPV-positive cases, multivariate logistic regression analysis including age, stage and primary treatment found that multiple-type infection was significantly associated with primary RT/CCRT (odds ratio: 3.32 [2.51–4.39], p < 0.001). Lymph node metastasis was not included in the multivariate analysis for multiple HPV infections because only those who received primary surgery with PLND had data for status of lymph node (Table VII).

Table VI. Single or Multiple HPV Infection According to Clinicopathological Variables in HPV-Positive Patients (N = 2,046)
 Single type n (%)Multiple type n (%)p1Odds ratio (95% CI)2
  • FIGO, International Federation of Gynecology and Obstetrics; RT, radiotherapy; CCRT, concurrent chemoradiation.

  • 1

    By Pearson's χ2 test.

  • 2

    By univariate logistic regression.

  • 3

    Tumor size estimated by histopathology.

Age (years)
 ≤3587 (93.5)6 (6.5)0.0031
 >351590 (81.4)363 (18.6) 3.31 (1.44–7.63)
 ≤50675 (86.0)110 (14.0)<0.0011
 >501002 (79.5)259 (20.5) 1.59 (1.24–2.02)
Primary tumor size (n = 943)3
 <2 cm393 (89.5)46 (10.5)0.0901
 ≥2 cm467 (92.7)37 (7.3) 0.68 (0.43–1.07)
Cell type
 Squamous1461 (81.7)327 (18.3)0.4331
 Nonsquamous216 (83.7)42 (16.3) 0.87 (0.61–1.24)
FIGO stage
 I–II1448 (83.6)285 (16.4)<0.0011
 III–IV229 (73.2)84 (26.8) 1.86 (1.41–2.47)
 I–IIA1191 (85.7)198 (14.3)<0.0011
 IIB–IV486 (74.0)172 (26.0) 2.11 (1.68–2.67)
Primary treatment
 Surgery922 (90.5)97 (9.5)<0.0011
 RT/CCRT755 (73.5)272 (26.5) 3.42 (2.66–4.40)
Lymph node metastasis (n = 1013)
 No825 (89.8)94 (10.2)0.0271
 Yes91 (96.8)3 (3.2) 0.29 (0.09–0.93)
Table VII. Multivariate Analysis of the Association of Clinicopathological Covariates and HPV Parameters
 Multiple HPV infections (N = 2,046)HPV58 or 52 or 33 positive (N = 2,118)
OR (95% CI)pOR (95% CI)p
  1. RT, radiotherapy; CCRT, concurrent chemoradiation; CI, confidence interval; OR, odds ratio.

Age (>50 vs. ≤50)1.07 (0.82–1.39)0.6192.29 (1.83–2.88)<0.001
FIGO stage (III–IV vs. I–II)1.03 (0.77–1.40)0.8271.29 (0.99–1.69)0.063
Primary treatment (RT/CCRT vs. surgery)3.32 (2.51–4.39)<0.0013.03 (2.42–3.80)<0.001

Discussion

The work of genotyping with PCR-based methods usually involves restriction fragment length polymorphism (RFLP),18, 19 hybridization with multiple oligonucleotide probes by either dot blot, southern blot,13, 20in situ hybridization,6, 21or enzyme-linked immunoassay,22 direct sequencing,5, 6, 17, 23, 24 or denaturing high performance liquid chromatography,25which are all laborious. Multiple infections cannot be identified simultaneously using direct sequencing,17, 24 and restriction enzyme analysis is also limited in resolving multiple types.18 Type-specific PCR requires the use of multiple PCR reactions and is usually carried out on a limited number of genotypes.6, 21 There are many investigators using general PCR followed by revert-blot or multiplex probe assay, and multiple HPV infections are much more frequently detected in both invasive cervical cancer and cervical intraepithelial neoplasia than other methods.5, 6, 17, 26, 27, 28

In our previous studies, SPF1/GP6+ PCR followed by revert blotting with HPV Blot was found sensitive and reproducible for HPV genotyping in cervical swab specimens16 as well as paraffin-embedded tissues.17 In the current study, HPV DNA sequences were detected in 96.6% of the specimens. A total of 35 types of HPV were detectable in this study with 28 of these 35 types existing as single type and the other 6 types only detectable along with the other more prevalent types. The leading 10 types were listed in decreasing order of frequency as HPV16, 18, 58, 33, 52, 39, 45, 31, 51 and 70. HPV70 was classified as low-risk type in a study using pooled data from 11 case-controlled studies,2 yet it ranked 10th in Taiwan.

Regarding the distribution of HPV genotypes worldwide, HPV16 is consistently the most prevalent type, while the frequencies of other genotypes vary in different geographic areas and ethnic groups.2, 3, 11 Sebbelov et al.22 found HPV16 to be the most prevalent type: 78.8% in Alaska natives, 96.3% in Greenland natives and 82.8% in Danish Caucasians. In contrast, a study of HPV genotype of Zimbabwean cervical carcinoma found that HPV33 was identified in 39%.23

The previous molecular epidemiologic studies of HPV in cervical cancer from Taiwan, Hong Kong, or China were controversial.8, 15, 16, 17, 20, 24, 29, 30, 31, 32, 33 Some of them indicated a possible importance of HPV52 and 58 in the Chinese population.24, 30, 31, 32, 33 Huang et al. investigated biopsy specimens of 40 cervical carcinomas from Chinese women by MY09/MY11 PCR followed by hybridization with a consensus probe detecting 45 types and 10 type-specific probes or sequencing, in which 87.5% were HPV-positive (7.5% HPV16, 10% HPV18, 20% HPV16/18, 15% HPV52, 15% HPV58 and 12.5% HPV52/58).24 Chen et al. used MY09/MY11 PCR and RFLP and noted that HPV-positive tumors were associated with increased risk of pelvic node metastasis.20 In contrast, pelvic node metastasis was associated with lower multiple infection rates and was unrelated to HPV status (positive or negative), HPV18-positive, or other genotype combinations analyzed (data not shown) in our series. Lai et al.31 used a consensus PCR-RFLP method and noted that HPV 52 or 58 or 33 was positive in 23.4% (22/94) invasive cervical cancer from Taiwan. In our series, HPV58 or 52 or 33 is detected in 30.3% of Stages I–IV cervical cancer (n = 2,118). A large series (n = 809) from 5 regions in China32 found that HPV16 was present in 79.6%, HPV18 in 7.5%, HPV52 in 2.6% and HPV58 in 3.8% of all HPV-positive specimens, and only 0.8% were multiple types. Another series (n = 541) from China noted only 1.97% multiple infection.33 In our series, 17.4% (369/2,118) of cervical carcinoma samples harbored multiple-type HPV infection. Studies relying on direct sequencing for genotyping tend to underestimate the frequency of multiple infection and bias the genotype distribution.24 It is understandable that results of distribution of HPV types or rates of multiple HPV infections (from 0.8% to 43.7%) vary extensively because of different specimens and techniques used.6, 8, 21, 32, 33

Our hospital cares approximately one-fifth of newly diagnosed invasive cervical cancers from all over Taiwan; therefore, the current study represents population-based data. We contributed the largest cervical cancer patient population of all participating institutions in the 2001 FIGO annual report.34 The current study confirms that HPV58 or 52 or 33 is detected in 30.3% of Stages I–IV cervical cancer of Taiwanese women. However, it is intriguing that HPV58 or 33 or 52 infection is significantly associated with older age and primary RT/CCRT, and multiple-type infection is significantly associated with primary RT/CCRT. The unanswered questions, why HPV58 or 33 or 52 and presence of multiple infections are associated primary RT and whether HPV genotype and/or multiple infections could be related to prognosis, will be analyzed in our future reports.

The implementation of HPV vaccines will be the future trend of primary prevention for cervical cancer.35, 36, 37, 38 The value of HPV testing as an adjunct to Pap smear is promising, though it has not yet been proven by prospective randomized controlled trials.39, 40 The knowledge of HPV genotype distribution will be a basis of guidelines to HPV-based cervical cancer screening and cost-effective prophylactic HPV vaccine policy and assessment of the effect of vaccination on HPV infection in each geographic area.

Acknowledgements

The authors thank Ms. Hsinni Yeh for her excellent assistance in the preparation of the artwork of this manuscript.

Ancillary