Preservation of multiple oncogenic human papillomavirus types in recurrences of early-stage cervical cancers

Preservation of multiple oncogenic human papillomavirus types in recurrences of early-stage cervical cancers Alexander F. Bnrnett, MD, Edward C. Grendys, MD, Gerry D. Willett, MD, Jacqueline C. Johnson, MD, James F. Barter, MD, and Willard A. Barnes, MD Washington, D.C. OBJECTIVES: Our purpose was to determine the relationship between human papillomavirus genotypes contained in primary early stage cervical cancers and those contained in the respective recurrences. STUDY DESIGN: Six early-stage cervical cancers that were considered cured by surgical extirpation subsequently recurred within 21 months of the original surgery. The primary tumors and the recurrences underwent polymerase chain reaction for human papillomavirus typing with confirmation of types performed by means of diagnostic restriction fragments. RESULTS: All primary tumors and recurrences contained human papillomavirus, with all primary tumors positive for multiple types. The concordance rate between the primary tumors and recurrences for specific types was 73% (11/15). Among the highly oncogenic types 16 and 18 there was 100% concordance between primary and recurrent tumors. CONCLUSIONS: Highly oncogenic types of human papillomavirus are preserved between primary tumors and their recurrences in cervical cancers. This further supports the role of oncogenic types in the maintenance of the malignant state and supports the clonogenic nature of cervical cancer recurrence. (AM J OasTET GVNECOL 1993;170:1230-3.)

Key words: Early-stage cervical cancer, human papillomavirus, polymerase chain reaction

The association of human papillomavirus (HPV) with cervical cancer has been documented in numerous studies. I -3 Specific HPV genotypes have been associated with varying degrees of oncogenic potential! The most highly oncogenic types in cervical cancer are HPV 16 and 18, which have been identified in 70% of tumors. 5 Intermediate risk types include HPV 31, 33, and 35.< It is felt that integration of the E6/E7 open reading frames of the papillomavirus deoxyribonucleic acid (DNA) into the host genome is part of the carcinogenic process in cervical cancer. 6, 7 Prior studies with in situ hybridization have documented the presence of HPV 16 in primary cervical tumors and in the recurrent specimens from the same patients. 8 This evidence supports the clonogenic theory of carcinogenesis with maintenance of specific HPV DNA from primary to recurrent cancer. More sensitive assays are now available, including polymerase chain reaction, which is capable of detecting as low as a single copy number of HPV DNA in cancer cells. 9 The polymerase chain reaction technique has the added benefit From the Diviszon of Gynecologic Oncology, Georgetown University Medical Center. Received for publication October 4, 1993; revised October 29, 1993; accepted November 22, 1993. Reprint requests: Alexander F. Burnett, MD, Dzvison of Gynecologic Oncology, 3 PHC, Georgetown UniverSity Medical Center, 3800 Reservozr Road, NW, Washington, DC 20007. Copyright © 1994 by Mosby-Year Book, Inc.

0002-9378/94 $3,00

1230

+ 0 6/1/53076

of adequately detecting HPV in archival tissue. lo With the more sensitive technique the presence of multiple HPV types within cancers has been shown; detecting copy numbers too low to be identified with hybridization techniques. II Documentation of infection with multiple HPV types in both primary and recurrent cancers that are identical would lend further support to the clonogenic nature of tumor recurrence. This study examines the HPV types found in recurrences of early-stage cervix cancers and compares them with the primary tumor. By means of the highly sensitive polymerase chain reaction this report documents the preservation of multiple HPV types from primary tumor to recurrences, suggesting that there may be integration of multiple HPV types within cervical cancers and that this may be a risk factor for recurrence. We examined patients with early stage-cervical cancers who had complete removal of tumor by radical hysterectomy and who had negative nodes.

Material and methods Six stage IB cervical cancers with subsequent recurrence were examined for HPV type. All tumors were confined to the cervix, and all had negative margins and negative lymph nodes. All patients had been treated with a type III radical hysterectomy with bilateral pelvic and paraaortic lymphadenectomy.12 Histopathologic analysis of primary tumor specimens, lymph nodes, margins, and recurrent tumors was performed in

Burnett et al.

Volume 170, Number 5, Part 1 Am J Obstet Gynecol

1231

Table I. Polymerase chain reaction primers and diagnostic restriction enzymes

HPVtype

6 11

16 18 31 33 35

Polymerase chain reaction product length

Primer sequence (5' to 3') CAC CTA AAG GTC CTG TIT GAA CCG CGC CTIGGTTAG CGC AGA GAT ATA TGC ATA TG AGT TCT AAG CAA CAG GCA CA CCC AGA AAG TIA CCA CAG TAC TAT GCA TAAATCCCG GAA ACG GTI GAA TCC AGC GTI CCT GTC GTGCTCGGT GAC CTC GGA AAT TGC ATG TGT TIC TGT TAACTGACC GTA TAT AGA GAG GGAAATTAAAGC TIT TITAACTGT ACA AGA ATI ACA GCA GAG TAA CTG TIT GTIGCATIG

a blinded fashion; tumor confinement to the cervix was document for all cases. All patients were considered to have been definitely treated with the surgical procedures, and none received adjuvant therapy after radical hysterectomy. Tissue was extracted from paraffin-embedded blocks for all six cervical cancers and their respective biopsies of recurrences. Deparafinization was performed with xylene and ethanol in a standard fashion. The tissue was further prepared for polymerase chain reaction analysis by overnight digestion with proteinase K. Polymerase chain reaction was performed as in previous reports}' with oligonucleotide primers for HPV types 6, 11, 16, 18, 31, 33, and 35 developed from the E6/E7 open reading frames (Table I). Oligonucleotide primers were initially tested with plasmid HPV DNA and were capable of specifically amplifying femtomolar concentrations of DNA. Polymerase chain reaction technique. Tissue cuts were processed initially with xylene and alcohol deparaffinization and wash. They were then placed at 37° C overnight with the addition of 50 fLl of proteinase K (5 mg/ml). Ten microliters of sample was incubated in a total volume of 100 ml in solution containing the following: potassium chloride 50 mmol/L; Tris-hydrochloric acid 10 mmoI/L (pH 8.0); magnesium chloride 3 mmol/L; 0.1% (wt/vol) gelatin; 200 mmol/L each deoxyadenosine triphosphate, deoxycytidine triphosphate, deoxyguanosine triphosphate, and deoxythymidine triphosphate; and 1 ng/ml oligonucleotide primers. This solution was heated to 99° C for 10 minutes to destroy further protease activity before 1 unit of Taq polymerase (Perkin-Elmer Cetus, Norwalk, Conn.) was added. This mixture was overlaid with mineral oil and subjected to 50 cycles of polymerase chain reaction with denaturation for 1.2 minutes at 94° C, followed by 0.6 minutes at 52° C (reannealing), and 1.2 minutes at 72° C (extension).

Diagnostic restnction enzyme

Restriction fragment lengths

230

Nde I

198, 32

89

Nde I

74,15

134

Ssp I

79, 55

119

Dde I

93, 26

97

Alu I

73, 24

132

Tth III, II

99, 33

186

sph I

148, 38

Polymerase chain reaction product was then electrophoresed on agarose gels with ethidium bromide intercalating the DNA for visualization of bands under ultraviolet light. Positive results were then verified with diagnostic restriction enzyme analysis of the known product sequence. The restriction fragments were then electrophoresed; band size shifts were confirmatory. All polymerase chain reaction samples were run concomitantly with samples from sterile paraffin blocks and from the MCF-7 breast cancer cell line to detect any contamination of reagents. In addition, all samples were also run with primers for human I3chain hemoglobin to document adequate DNA extraction. Recurrent tumors were typed for all of the above HPV types and were then compared with the HPV types found in the primary tumors.

Results The average age of the patients at time of original cancer diagnosis was 47.5 years (range 33 to 68 years). Five of the tumors were squamous cell carcinomas, and one was adenocarcinoma. Four of the primary tumors exhibited lymph-vascular space involvement. The mean depth of invasion for the primary tumors was 1.3 cm, with a mean tumor diameter of 3.3 cm. Three of the tumors were moderately differentiated, two were poorly differentiated, and one was well differentiated. All recurrences were pelvic, with a mean time to recurrence of 12.1 months (range 2 to 21 months). The results of HPV typing of primary tumors and the respective recurrences are presented in Table II. HPV 18 was identified in five primary tumors, HPV 16 in four, HPV 31 in three, HPV 35 in two, and HPV 33 in one. All six primary cancers were positive for more than one type of HPV and had either HPV 16 or HPV 18. Three tumors were infected with both HPV 16 and HPV 18. In the recurrences 73% (11/15) of the HPV types

1232

May 1994

Burnett et al.

Am J Obstet Gynecol

Table II. HPV types of primary and recurrent tumors Tumor

Primary HPV types

A B C D E F

16, 35 16,18,35 16, 18 18, 31, 33 16, 18,31 18,31

Recurrence HPV types 16 16, 16, 18, 16, 18

18 18 33 18,31

found in the primary tumors were identified. Infection with the highly oncogenic types HPV 16 and HPV 18 was identical in every recurrence compared with its primary. In two cases HPV 35 was identified in the primary tumor but not the recurrence, and in two additional cases HPV 31 was also identified in the primary tumor but not the recurrence. No HPV types were identified in the recurrences that were not found in the primary cancers.

integration of multiple oncogenic HPV types confers a higher risk for recurrence in surgically excised early cervical cancer. The relative lack of preservation for intermediately oncogenic HPV types (33% preservation of types 31, 33, and 35) suggests that integration of these types into the cancer genome may not be critical for preservation of the oncogenic phenotypes; however, the numbers in this series are small and warrant further investigation.

Comment Early-stage cervical cancer amenable to surgical therapy has a low recurrence rate. Factors that have been identified in .node-negative patients as risks for recurrences include large tumor volume, deep invasion of tumor, and presence of lymph-vascular space involvement. 14 More recently infection with multiple oncogenic HPV types and infection with HPV genotype 18 have been suggested as risk factors for recurrence." This report documents the preservation of infection with oncogenic types and infection with multiple types of HPV in recurrences. Previously detection of a single type of HPV identical in primary and recurrent cancers has been performed with in situ hybridization. 8 This study used a far more sensitive technique (polymerase chain reaction) and was able to detect multiple types of oncogenic HPV in primary and recurrent cancers. Polymerase chain reaction has been shown to detect HPV DNA in up to 96% of cervical cancers.15 In addition, it has demonstrated multiple types of HPV within a single cervical tumor, with one study demonstrating HPV types 16, 18, or both in all cancers studied. II This study failed to detect HPV types 6 or 11 in the cancers examined. In a prior report we documented cervical cancers infected with a single type of papillomavirus. 13 In that study nine of 15 (60%) of early cervix cancers that did not recur were infected with a single HPV type. Preservation in cervical cancer recurrences of multiple HPV types identical to the primary tumor further support the role of HPV in maintaining the malignant state of cervical cancer. The preservation of the detection pattern of the highly oncogenic types (HPV 16 and HPV 18) in all patients suggests that both these types are integrated into the tumor cell genome, and perhaps

1. Reeves WC, Brinton LA, Garcia M, et al. Human papillomavirus infection and cervical cancer in Latin America. N Engl] Med 1989;320:1437-41. 2. Kulski ]K, Demeter Mutavdzic S, Sterrett GF, Mitchell KM, Pixley EC. Survey of histologic specimens of human cancer for human papillomavirus types 6/11/16/18 by filter in situ hybridization. Am] Clin Pathol 1990;94:566-70. 3. King LA, Tase T, Twiggs LB, et al. Prognostic significance of the presence of human papillomavirus DNA in patients with invasive carcinoma of the cervix. Cancer 1989;63: 897-900. 4. Lorincz AT, Reid R, Jenson AB, et al. Human papillomavirus infection of the cervix: relative risk associations of 15 common anogenital types. Obstet Gynecol 1992;79:32837. 5. Walker ], Bloss ]D, Uao SY, Berman M, Bergen S, Wilczynski SP. Human papillomavirus genotype as a prognostic indicator in carcinoma of the uterine cervix. Obstet Gynecol 1989;74:781-5. 6. zur Hausen H. Human papillomaviruses in the pathogenesis of anogenital cancer. Virology 1991;184:9-13. 7. Takebe N, Tsunokawa Y, Nozawa S, et al. Conservation of E6 and E7 regions of human papillomavirus types 16 and 18 present in cervical cancers. Biochem Biophys Res Commun 1987;143:837-44. 8. Holloway RW, Farrell MP, Castellano C, et al. Identification of human papillomavirus type 16 in primary and recurrent cervical cancer following radiation therapy. GynecolOncol 1991;41:123-8. 9. ] enkins A, Kristiansen BE, Ask E, et al. Detection of genital papillomavirus types of polymerase chain reaction using common primers. APMIS 1991;99:667-73. 10. Shibata DK, Arnheim N, Martin WJ. Detection of human papilloma virus in paraffin-embedded tissue using the polymerase chain reaction.] Exp Med 1988;167:225-30. 11. Griffin NR, Bevan IS, Lewis FA, Wells M, Young LS. Demonstration of multiple HPV types in normal cervix and in cervical squamous cell carcinoma using the polymerase chain reaction on paraffin wax embedded material. ] Clin Pathol 1990;43:52-6. 12. Piver MS, Rutledge FN, Smith PJ. Five classes of extended hysterectomy of women with cervical cancer. Obstet Gynecol 1989;44:265-72. 13. Burnett AF, Barnes WA, Johnson ]C, et al. Prognostic significance of polymerase chain reaction detected human

REFERENCES

1:

Volume 170, Number 5, Part 1 Am J Obstet Gynecol

papillomavirus of tumors and lymph nodes in surgically treated stage IB cervical cancer. GynecolOncol 1992;47: 343-7. 14. Delgado G, Bundy B, Zaino R, Sevin B, Creasman WT, Major F. Prospective surgical-pathological study of disease-free interval in patients with stage IB squamous cell

Arias

carcinoma of the cervix: a Gynecologic Oncology Group study. Gynecol Oneol 1990;38:352-7. 15. Yoshikawa H, Kawana T, Kitagawa K, et al. Detection and typing of multiple genital human papillomaviruses by DNA amplification with consensus primers. Jpn J Cancer Res 1991 ;82:524-31.

Delayed delivery of multifetal pregnancies with premature rupture of membranes in the second trimester Fernando Arias, MD, PhD St. Louis, Missouri OBJECTIVE: Premature rupture of membranes in mutifetal gestations during the second trimester has an ominous prognosis and the majority of the fetuses die after preterm delivery. STUDY DESIGN: We used cervical cerclage, tocolysis, and antibiotic therapy after vaginal delivery of the fetus with ruptured membranes in eight patients with multifetal pregnancies to extend the intrauterine life and improve the outcome of the remaining fetuses. RESULTS: Survival of six fetuses was achieved in five of the eight pregnancies (four originally twins and one originally triplets). The mean ± SO gestational age of the fetuses first delivered was 19.6 ± 2.6 weeks, and it was 26.7 ± 6.8 weeks in the fetuses with delayed delivery (exact two-tailed p = 0.01). The mean ± SO birth weight of the fetuses delivered first was 306 ± 149 gm and it was 1029 ± 947 gm for the fetuses who had delayed delivery (exact two-tailed p = 0.05). The mean ± SO prolongation of pregnancy was 48.8 ± 42.06 days (range 8 to 114 days). CONCLUSIONS: Intervention with tocolysis, antibiotics, and cervical cerclage after delivery of the first fetus is a reasonable option for some patients with multifetal pregnancies and premature rupture membranes in the second trimester. (AM J OBSTET GYNECOL 1994;170:1233-7.)

Key words: Twins, cerclage, premature rupture of membranes, triplets, delayed delivery Premature rupture of membranes before 26 weeks of gestation affects 1.37% of twin gestations. l A similar incidence probably occurs in other multifetal pregnancies of higher fetal order. Mter a latent period with a mean duration of 1.1 days, 1 the fetus with ruptured membranes is delivered, and this is followed by the spontaneous or induced delivery of the other fetuses. Delivery of the fetuses with intact sacs is usually the result of uterine activity that does not subside after delivery of the fetus with ruptured membranes. In some cases the uterus ceases to contract once the first fetus is delivered, and in this situation, because of the possibil-

From the Department of Obstetrics and Gynecology, St. John'S Mercy Medical Center. Received for publication June 29, 1993; revised November 1, 1993; accepted November 12, 1993. Repnnt requests: Fernando Anas, MD, Department of Obstetrics and Gynecology, St.John'S Mercy Medical Center, St. Louls, MO 63141. Copynght © 1994 by Mosby-Year Book, Inc. 0002-9378/94 $3.00 + 0 6/1/52865

ity of severe intrauterine infection when the delivery is not completed, induction of labor with oxytocin is used to accelerate the delivery of the remaining fetuses. There is an increasing number of cases described in the recent literature in which efforts have been made to prolong the interval between deliveries after preterm premature rupture of membranes and delivery of the first fetus in multifetal pregnancies. A literature review on this subject, published in 1992; summarized 26 articles reporting on 28 pregnancies with 65 fetuses, 9 from triplets and 19 from twin pregnancies. Since that review, five more articles have appeared,3-7 increasing the total number of reported attempts to delay delivery in this situation to 41,9 with triplets and 32 with twins. In 1984, a 32-year-old white woman, gravida 5, para 0, was first seen at 20.5 weeks with premature rupture of membranes and twin pregnancy. Mter 10 days of expectant management umbilical cord prolapse of the first twin occurred, followed by spontaneous labor. Consent was obtained for attempting to delay delivery 1233