Prevalence of HPV in premalignant and malignant cervical lesions in Greenland and Denmark: PCR and in situ hybridization analysis on archival material

O INSTITUTPASTEUR/Et.sEVlER Paris 1994

Res. Virol. 1994, 145, 83-92

Prevalence of HPV in premalignant and malignant cervical lesions in Greenland and Denmark: PCR and in situ hybridization analysis on archival material A.M. Sebbelov (i), C. Svendsen (t), H. Jensen (3), S.K. Kjaer (2) and B. Norrild 0)(*) (1) DNA Tumor Virus Laboratory, Institute o f Medical Microbiology, University o f Copenhagen, (2) The Danish Cancer Society, Danish Cancer Registry, Institute o f Cancer Epidemiology, Copenhagen, and (3) The Department o f Pathology, Rigshospitalet, Copenhagen

SUMMARY

Archival formalin-fixed paraffin-embedded cervical specimens from 125 women in Greenland and 139 women in Denmark who had CIN I-III or cervical cancer diagnosed between 1983 and 1987 were analysed for human papillomavirus type 16 (HPV-16) by in situ hybridization and for HPV-16, 18, 31, 33, 35 and 45 by PCR. In situ hybridization analysis showed an HPV-16 prevalence of 17 % (16/95) and 23 % (24/105) in the premalignant lesions from Greenland and Denmark, respectively. The HPV-16 prevalence rate in the cancer specimens was 10 % (3/30) in the samples from Greenland and 29 % (10/34) in the Danish specimens. A total of 82 Greenlandic and 107 Danish specimens were {3-globin-positive by PCR reaction. HPV-16-specific PCR on these samples showed 63 % (34/54) of the Greenlandic and 68 % (50/74) of the Danish preinvasive lesions to be positive. The corresponding HPV-16-positive rates for the inv'asive cancers were 82 % (23/28) for Greenland and 70 % (23/33) for Denmark. This study of patient samples supports our previous population-based studies in which similar HPVdetection rates were found among random samples of women from Greenland and Denmark, although Greenland has a 4-5-fold higher cervical cancer incidence. Key-words: Papillomavirus, Tumour, Cervix; PCR, In situ hybridization, Greenland, Denmark.

INTRODUCTION Human papillomaviruses (HPV) cause a spectrum of h u m a n diseases, and more than 70 different virus types have been isolated from various lesions. About 25 types have specific af-

Submitted August 23, 1993, accepted January 19, 1994. (*) For correspondence.

finity for the ano-genital mucosa, and due to their suspected oncogenic potential, some of these types have been extensively studied (De Villiers, 1989; Munoz et al., 1992; Munoz and Bosch, 1992; Pfister, 1992). HPV-16 is the most prevalent type in ano-genital squamous cell le-

84

A . M . S E B B E L O V E T AL.

sions in the western world, whereas HPV-18 seems to prevail in genital adenocarcinomas (Bosch et al., 1989). The incidence rate o f cervical cancer varies at least 20-fold between different parts o f the world. Greenland constitutes a high-risk area with an age-standardized incidence o f 63.7 per 100,000 (1983-1985) (Nielsen et al., 1988 ; Prener et al., 1991) compared to Denmark, where the incidence is 16.4 per 100,000 (1988) (Storm et al., 1988).

graphical area is presented in table I. There was a tendency for Greenlandic women to be slightly younger than Danish women in all three diagnostic groups, but the age distribution was not significantly different in the two study populations. In situ hybridization

The study population consisted of women from Greenland and Denmark (Copenhagen) with a diagnosis of cervical intraepithelial neoplasia (CIN) or invasive cancer. All Greenland cancer patients are referred to the University Hospital in Copenhagen (Rigshospitalet), and the Danish patient group was also chosen from this hospital. All lesions were diagnosed from 1983 to 1987, and the diagnosis was verified on the material sectioned for analysis in the present study. From the computerized pathological register at Rigshospitalet, a sample of women in the eligible diagnostic groups was drawn at random among Greenlanders and Danes. A small proportion of women from both areas had to be excluded because no archival tissue material was available. From Denmark, 139 patients were finally included with the following diagnoses : 52 (37 07o)CIN I/II, 53 (38 07o) CIN III, and 34 (25 07o)invasive cervical cancers. A total of 125 patients from Greenland were included, 49 (39 070)CIN I/II, 46 (37 070)CIN III and 30 (24 070) invasive carcinomas. The age distribution of the women in the different diagnostic groups by geo-

In situ hybridization was done under highstringency conditions using a biotinylated, nicktranslated HPV-16 probe from a commercially available "Biohit" typing kit (Biohit OY, Helsinki, Finland). In each round of hybridization, Caski cells or a patient specimen previously found positive for HPV-16 was included as a positive control. Two tissue sections were hybridized from each patient, one with the biotinylated HPV-16 probe and one with a biotinylated pBR322 probe as a negative control, q-he stringency of the hybridization was adjusted so that no cross-hybridization occurred when sections from an HPV-16-positive control sample (patient sample or Caski cells) were hybridized with an HPV-31 nicktranslated probe (Biohit kit). The specificity of the HPV-16 probe was tested by hybridization of the HPV-16 probe to HPV-6-positive tissue. Hybridization at 50°C followed by washing at 60°C was found to be the optimal condition. The hybridization protocol was slightly modified from that recommended by the supplier. In brief, the sections were baked at 60°C overnight and deparaffinized in two changes of xylene followed by hydration in graded ethanols. The slides were air-dried and treated for 15 min at 37°C in a humidity chamber with 0.5 mg/ml proteinase-K (Boehringer-Mannheim). The slides were soaked in PBS at room temperature, dehydrated in graded ethanols and air-dried. The hybridization mixture was supplemented with dextran sulphate and thus contained 50 °70 deionized formamide, l0 070 w/v dextran sulphate, 2 x SSC, 0.4 mg/ml sonicated, denatured herring sperm DNA (Promega) and 0.5-1.0 lzg biotinylated HPV-16 probe. This was 1/10th of the amount recommended by the supplier, but because of the addition of dextran sulphate, the reduced probe concentration gave optimal hybridization. Slides with the hybridization mixture were heated in a 110°C oven for 5-6 min to denature the sample DNA, coverslips were sealed with silicone and slides hybridized in a humidifying chamber overnight at 50°C. Following hybridization, the coverslips were gently removed and the slides soaked in 2 × SSC and washed 3 times at 60°C in 0.2 x SSC and once in 2 x S S C at room temperature. To visualize the

CIN HPV

PCR SDS

The purpose o f the present study was to extend the population-based studies carried out in Greenland and Denmark with an analysis of the H P V type 16 prevalence in preinvasive cervical lesions and in cancers occurring in the two geographical areas. In addition, the four diagnostic categories were analysed for the presence o f the less prevailing H P V types 18, 31, 33, 35 and 45.

MATERIALS AND METHODS Study population

PBS

= cervicalintraepithelialneoplasia. = humanpapillomavirus. = phosphate-bufferedsaline.

= polymerasechain reaction. = sodiumdodecylsulphate.

HPV IN PREMALIGNANT AND MALIGNANT

CERVICAL LESIONS

85

Table I. Age distribution by histopathoiogicai diagnosis and geographical area.

Age 24 25-34 /> 35

CIN I/II Denmark Greenland 17 22 13 52

(33 °70) (42 070) (25 070) (100 °70)

20 16 13 49

(41 07o) (33 °7o) (26 070) (100 °7o)

Age ~< 24 25-34 /> 35

CIN IlI Denmark Greenland 10 (19 070) 30 (57 070) 13 (24 °7o) 53 (100 07o)

hybridization, an immunohistochemical procedure was used as described for the "Biohit" kit. The slides were counterstained with carbo-fucin diluted 1:200 and mounted with " D e p e x " mounting medium. The slides were read by two of the investigators and scored positive only when several nuclei stained positive. PCR analysis The paraffin-embedded tissue was cut in 5-~tm sections and, depending on the size of the tissue, 1 to 3 sections were pooled in a 1.5-ml Eppendorf tube. The microtome was cleaned in alcohol between each block, and after every six patients a neutral section from either heart muscle or pure paraffin was cut as a contamination control. The tissue sections were treated according to two protocols. In protocol 1, sections were deparaffinized in 500 ~tl xylene, centrifuged and subsequently dispersed in 500 ~.196 070 ethanol to remove any left-over xylene. Following centrifugation, the ethanol was discarded, and 1 ~tg/ml tRNA and 800 ~tl 70 070 ethanol were added before storage of the samples at - 20°C until use. After removal of the ethanol and addition of 100 ~tl 5 mM Tris pH 7.5, tissue extracts were ready for PCR analysis (Shibata et al., 1988). In protocol 2, 100 izl 5 mM Tris pH 7.5 were added to the sections and samples were heated to melt the paraffin and spun down. Protocol 2 was more simple, with less risk of contamination. The deparaffinized tissue sections were treated with 20 ~g proteinase K (Boehringer-Mannheim, no. 161519) at 55°C for 3 h. After heating to 95°C for 1 min, the tissue extracts were centrifuged and stored at - 2 0 ° C until PCR analysis (Manos et al., 1988). PCR reaction A positive ~3-giobin PCR reaction was used as a criterion for the identification of tissue extracts qualified for further PCR analysis.

12 23 11 46

(26 070) (50 07o) (24 07o) (100 07o)

Age ~ 44 45-54 >f 55

Invasive Denmark Greenland 16 6 12 34

(47 070) 18 (60 070) (18 o7o) 5 (17 °70) (53 o7o) 7 (23 °70) (100 070) 30 (100 07o)

Amplification was carried out in 500-~1 tubes (Eppendorf, safe-lock, no. 0030121.023). A 100-~tl reaction volume contained 15 ~1 of the tissue extract, 100 pmoles of each primer, 2.5 units enzyme (Amplitaq, Perkin-Elmer, Cetus) and buffer as recommended by the manufacturer. A drop of mineral oil was added (Sigma, M3516), and a " H y b a i d " thermal reactor was used for thermal cycling. The program used for the HPV-type-specific primers was 35 cycles beginning with 94°C for 5 min, followed by 40 s at 54°C, 40 s at 72°C and 40 s at 92°C for 35 cycles ending with a 5-min extension at 72°C after the last annealing. The program used for consensus primers was modified from the program published by Gregoire et aL, 1993. Each PCR was run, including a control of the proteinase K digestion buffer, a negative control and a positive control containing a standard amount of HPV DNA. The primers used for [3-globin-specific amplification were PCO3 and PCO4 (Saiki et al., 1985), which amplified a 100-bp sequence. The specific probe was the 19A sequence (Saiki et aL, 1986). The primers used for HPV-type-specific amplification of HPV-16 and -18 had been previously published by Young et al. (1989). Specific primers to HPV-31 and HPV-33, as listed in table II, were identified for the present study (A.M. Sebbelov and R. Albrectsen). In order to study the prevalence of HPV-35 and -45, the consensus primers gp5 and gp6 were used (Sneijders et al., 1990) and the amplified products were analysed by Southern blotting. Southern blot analysis of the amplification products PCR reaction product (20 ~.1)was electrophoresed in a 2.5 070 agarose gel (BRL, no. 540-5510UB) for band separation. The gels were ethidium-bromidestained, photographed and denatured before transfer to " N y t r a n " nylon membranes (Schleicher & Schuell) using standard procedures. Hybridization to specific [3-globin and HPV-16 PCR blots were with alpha 32P-dATP (Amersham, no. PB10235) -endlabelled oligonucleotide probes (19A and HPV-16

86

A.M. SEBBELOV ET AL.

probes, table II). Optimal analytical stringency conditions for the chosen probe size were obtained when hybridization was performed at 37°C in 6xSSC, 10 x Denhardt solution with 50 izg/ml sheared calf thymus DNA (Boehringer-Mannheim, no. 104175) and the blots were washed at 37°C in 6 x SSC with 0 . 1 % SDS. Probes labelled by random-priming with alpha 32p-dCTP (Dupharma ApS, NEG 013H) using PCR products as templates were made as described by Sneijders et al. (1990) and used for HPV-35 and -45 analysis. Hybridization and the first wash were performed at low stringency (Tm -33°C). After autoradiography, a second wash was performed at high stringency (Tm - 10°C). For autoradiography "Kodak T-MAT G5500" film was used.

13 formalin-fixed, paraffin-embedded specimens which were negative by the in situ hybridization test. They were chosen randomly a m o n g the samples which were HPV-16-positive by PCR analysis; one section was hybridized with the standard a m o u n t of probe used in the present study and the other with the concentration recommended by the supplier. Samples that were negative using our standard conditions remained negative in the in situ hybridization test carried out with the increased amount of probe; the two positive controls showed increased intensity o f the colour reaction with a high amount of probe, but the n u m b e r o f HPV-positive cells was unchanged (data not shown).

RESULTS

PCR analyses In situ hybridization analysis HPV-16 (fig. 1) was detected in 23 % o f the preinvasive lesions in D e n m a r k ( 2 1 % positive CIN I / I I and 25 070 CIN III). A similar positivity rate was observed in the Greenlandic premalignant lesions (17 %), with 14 070 positive CIN I / I I and 20 07o positive CIN III lesions (table III). Only 29 % and 10 070 o f the invasive carcinomas were HPV-16-positive in D e n m a r k and Greenland, respectively. This difference reached borderline significance (p <~ 0.06) (table III). To eliminate the possibility that the HPV-16 probe was used at limiting concentrations, two sections were made from each o f

Tissue sections from all specimens in the study were tested by PCR for [3-globin positivity. Only [3-globin-positive specimens qualified for HPVspecific PCR analyses (table III). Nearly all the invasive carcinomas were [3-globin-positive (Denm a r k : 99 °7o; G r e e n l a n d : 93 07o). A m o n g CIN III lesions, 64 °/o and 65 07o were positive in D e n m a r k and Greenland, respectively. A total o f 77 07o o f the Danish CIN I/II lesions, but only 49 o7o o f the Greenlandic CIN I/II lesions, were [3-globin-positive, with the difference being statistically significant (p < 0.005). Using PCR, HPV-16 D N A was found in 53 %, 85 07o and 70 % o f the Danish CIN I/II, CIN III and in-

Table II. Primers used for detection of HPV DNA by PCR analysis. HPV type (oligonucleotide)

Sequence

Size of Genome position amplificate ( 5 ' . . . 3')

HPV-31 sense primer HPV-31 antisense primer

TGT GCA AAC CTA CAG ACG CC

HPV-33 sense primer HPV-33 antisense primer

ACT GAG GAA AAA CCA CGA AC CAT ATT CCA AAT GGA TTT CC

E6:296 to 277

HPV-16 probe

ATC TGG ACA AAA AGC AAA GAT TCC ATA ATA TAA GGG GTC

E6:456 to 494

E6:88 to 107 205 bp

ACT CCG TGT GGT GTG TCG TC

HPV genome sequence Goldsborough et al. (1989)

E6:292 to 273 E6:121 to 140 176 bp

Cole and Streeck (1986)

Seedorf et al. (1985)

HPV IN PREMALIGNANT AND MALIGNANT CERVICAL LESIONS

Fig. 1. In situ hybridization with HPV-16 or pBR322 biotinylated probes used in a concentration of 0.5-1.0 ~.g. A carcinoma in situ sample hybridized with HPV-16 (A) or pBR322 (B) ; an invasive cancer specimen hybridized with HPV-16 probe (C) and pBR322 (D).

87

88

A.M. SEBBELO V E T AL.

Table III. Prevalence of HPV by histopathological diagnosis and geographical area using in situ hybridization or PCR. PCR (070 of [3-globin-positive) Histology

Total HPV-16 no. in situ hyb.

13-globinpositive

HPV-positive HPV-16

HPV-18

HPV-31

HPV-33

Total

2 (5 %)0) 0 0

4 (I0 070)0) 10 (29 070) 2 (6 070)

20 (50 070) 31 (91 070) 23 (70 070)

3 (16 °7o)t2) 0 7 (14 070) 24 (49 070) 13 (54 °7o) 0 9 (20 070) 30 (65 070) 21 (70 070) 1 (4 070)t3) 2 (7 °7o)t4) 3 (12 070)°) 3 (10 070) 28 (93 070) 23 (82 °70) 0 0 1 (4 070)

14 (64 070) 19 (63 070) 23 (82 070)

Denmark:

CIN I/II CIN III Invasive

52 53 34

11 (21 07o) 40 (77 070) 21 (53 070) 0 13 (25 070) 34 (64 070) 29 (85 070) 0 10 (29 07o) 33 (99 07o) 23 (70 070) 0

Greenland:

CIN I/II CIN III Invasive

49 46 30

(1) ! sample not tested for HPV-31 and 33; (2) 5 samples not tested for HPV-31 ; (3) 4 samples not tested for HPV-18 and 31 ; (4) 2 samples not tested for HPV-31.

vasive lesions, respectively. In Greenland, 54 % CIN I/II, 70 % CIN III and 82 % invasive lesions were positive. Only for CIN III was the difference between Denmark and Greenland statistically significant (p < 0.025). If CIN I-III is considered as a single category ("premalignant lesions"), then the prevalence of HPV-16 was 68 % in Denmark and 63 % in Greenland. No statistically significant differences in the prevalence of HPV-31 and 33 were observed between the two areas, but the prevalence of HPV-33, especially in the Danish CIN II! lesions, should be noted. HPV-33 is most often observed in tissue also infected with HPV-16, and of the 20 HPV-33-positive samples, 16 were double and 2 triple infections. Only one case of HPV-18 was found and no cases of HPV-35 or -45 were observed (Nuovo et al., 1991) (table III).

Comparison of PCR and in situ hybridization The correlation between the results obtained by P C R and by in situ hybridization is shown in table IV. There is agreement between the results obtained by the two methods in 22, 13 and 14 of the specimens from the Danish CIN I/II, CIN III and invasive cases, respectively. In the Greenland diagnostic categories, agreement was obtained in 13, 13 and 8 specimens, respec-

tively. As expected, many of the specimens which were negative by the in situ test turned out to be positive by P C R analysis. This increased the number of HPV-positive cases, with 15, 21 and 16 in the three Danish diagnostic categories, and 10, 15 and 20 in the corresponding Greenland diagnostic groups. It should be noted that a few cases which were HPV-positive by in situ hybridization were negative when analysed by P C R (2 samples among the Danish CIN I/II, 3 among the Danish invasive cases, l and 2 among the Greenlandic CIN I/II and CIN III cases, respectively).

DISCUSSION The present study was des!gned to compare the prevalence of H P V DNA in premalignant and malignant cervical lesions from a high-risk area (Greenland) and a low-risk area (Denmark) for cervical cancer. When the in situ hybridization test was used for the demonstration of H P V DNA, a low HPV-16 prevalence was found in all the diagnostic categories in Denmark and Greenland. This may be due to a low copy number of HPV in the study samples; for the Greenland cancers, it is probably also related to the lower level of differentiation in these as compared to the Danish cancers (fig. 2). The in situ

HPV IN PREMALIGNANT AND MALIGNANT

89

CERVICAL LESIONS

Table IV. Comparison of in situ hybridization and PCR in detection of HPV-16 DNA by histopathological diagnosis and geographical area in [3-globin-positive samples. In situ hybridization

CIN I/II PCR + PCR-

Denmark Positive Negative (') Greenland Positive Negative

CIN III PCR + PCR-

Invasive PCR + PCR-

6 15

2 16

8 21

0 5

7 16

3 7

3 10

1 10

6 15

2 7

3 20

0 5

(*) One sample not tested by PCR is omitted.

% 40

the quality of the test. The low sensitivity of the in situ hybridization test was unexpected, and other studies refer to a much higher HPV prevalence using similar techniques (Nuovo, 1991).

. . . .

30

2O

Microinv.

High

Med.

Med./Low

low

Unknown

Fig. 2. C o m p a r i s o n o f levels o f t u m o u r differentiation between Greenlandic and Danish w o m e n . Black columns = Greenland; hatched columns = Denm a r k . Med. = m e d i u m .

hybridization test was thus much less sensitive than PCR analysis for diagnosis of HPV. This was not caused by limitations in the concentration of the HPV-16 probe, since a 10-fold increase in the concentration of the probe gave the same results. We found that the addition of dextran sulphate to the hybridization mixture led to a reduction in the quantity of probe used. This reduced the cost significantly without reducing

Since most of the specimens were suitable for PCR analysis, as measured by [3-globin PCR, HPV-specific PCR was performed on 76 °70 of the Danish and 66 e/0 of the Greenland samples. We expected a higher success rate for the [3globin PCR because the tissue blocks were all less than 10 years old, and to our knowledge, all specimens were fixed in buffered formalin as recommended by Wright et al. (1~90). In another study, we had analysed paraffin-embedded tissue blocks with a success rate of 95 070. These specimens were analysed in the laboratory during the same time period as samples in the present study. We are therefore quite confident that PCR was done under optimal conditions; in the present study, however, we were unable to reach a success rate corresponding to the 94 GT0 reported for archival smears (Smits et al., 1992). The observed prevalence of HPV-16 appears comparable in the Danish and Greenlandic CIN I-III cases, as well as in the cancers. However, it should be taken into consideration that a significantly lower percentage of Greenland sampies in the CIN I/II categories were suitable for PCR analysis compared to Danish samples ; indeed, only 49 % of the Greenlandic samples were [3-globin-positive, in contrast to 77 °7o of the corresponding samples from Denmark. This observation indicates that the poor quality of the

90

A.M. S E B B E L O V E T AL.

Greenland samples might be related to less optimal fixation of the tissue. The observed difference may have influenced the measurement of HPV prevalence. HPV-16 is found to be the most prevalent HPV type in both Denmark and Greenland; types 18, 31, 33, 35 and 45 are not found in high numbers of specimens. HPV type 16 isolated from the native population in Greenland is the same as the most common type found in the western world, which was confirmed in a recent study in which the different HPV-16 isolates were DNA-sequenced (Ho et al., 1993). At present, we have no explanation for the incidence of simultaneous detection of HPV types 16 and 33 in the two populations. We find it unlikely that the phenomenon is due simply to contamination or cross-reactivity. We have computer tested the reliability of the HPV-33 primer sequencies by matching to relevant known HPV genomes, and performed PCR on DNA samples from HPV types 6/11, 16, 18, 31, 35 and 45, as well as on human tissue. There seemed to be no cross-annealing to any of the DNA tested, including the HPV 16 prototype. However, it cannot be totally excluded that HPV 16 variants exist which may cause falsepositive results. It should be mentioned that double infections with HPV-16 and 33 are even more pronounced in cancers from Alaskan native women (Sebbelov, to be published). The results of the present study should be compared to those of previous population-based studies of random samples of women from the general population in Greenland and Denmark. These showed no difference in HPV prevalence among the two populations independently of whether filter in situ hybridization, ViraPap/ ViraType or PCR was used (Kjaer et al., 1988, 1993). By contrast, in studies conducted in Columbia and Spain, which also represent highand low-risk areas, respectively, for cervical cancer, HPV prevalence correlated strongly with cancer risk. This correlation was observed only when HPV DNA was measured by PCR, and it should be noted that it was not found by the ViraPap method. In the cancers from Columbia and Spain, the HPV prevalence was comparable, and this was also observed in the present study (Munoz et al., 1992). Studies per-

formed by both Kjaer et al. (1993) and Munoz et ai. (1992) conclude that HPV is a persistent infection, and it was observed that the HPV gene copy number is higher in the high-risk population (Munoz et al., 1992). The present study could not confirm a high copy number of HPV-16 in the Greenland study group; a high copy number would be reflected in higher HPV prevalence in Greenland than in Denmark using the in situ hybridization test. In conclusion, our data supplement results obtained in previous studies from Greenland and Denmark and indicate that HPV is equally prevalent in premalignant and malignant cervical lesions from these areas. In addition to HPV, other factors would seem to be necessary in order for premalignant lesions to become malignant, and it is well documented that the two populations differ with respect to, for example, smoking habits and the frequency of other sexually transmitted diseases (Kjaer et ai., 1993). Whether these influence the progression of HPV lesions to cancer is unknown.

Acknowledgement

The kind donation of HPV types 16 and 18 from Prof. H. Zur Hausen, HPV-31 from Prof. W. Lancaster. HPV-33 from Prof. G. Orth, HPV-35 from Dr. A. Lorincz and HPV-45 from Prof. K. Shaw is gratefully acknowledged. The kind help and instructions from Prof. Stina Syrj~inen are highly acknowledged. We are thankful for the skilled technicalassistanceof L. Christensen, H. Haslund, P. Christensenand H. GreveJensen, and for the experienced computer assistance of Svend.E.Kofod. This work was supported by grants from the Danish Cancer Society no. 92-005 and no. 91-062, F.L. Foghts Foundation and the NOVO Foundation.

Prevalence du virus du papillome humain dans les I~sions cervicales malignes et pr~malignes au Groenland et au Danemark: analyse de specimens par hybridation in situ et PCR

Des sp6cimens cervicaux, fix6s par le formol et inclus en paraffine, de 125 Groenlandaises et 139 Danoises porteuses d'un cancer intra6pith61ial du col ut6rin (CIN I-III) diagnostiqu6 entre 1983 et 1987 ont ~t6 analys6s pour la recherche du type 16 du papillomavirus humain (HPV-16) par hybridation in

HPV IN PREMALIGNANT AND MALIGNANT CERVICAL LESIONS situ et par P C R pour ies types 16, 18, 31, 33, 35 et 45. L'hybridation r6v~le une pr6valence de HPV-16 respectivement de 17 070 (16/95) et de 23 07o (24/105) dans les 16sions pr6canc6reuses de femmes du Groenland et du Danemark. Dans les 16sions malignes, le taux de pr6valence de HPV-16 est de 10 07o (3/30) dans les sp6cimens groenlandais et de 29 07o (10/34) darts les specimens danois. La P C R montre que 82 sp6cimens groenlandais et 107 danois sont [3globine positifs et la P C R sp~cifique de HPV-16 r6v~le que 63 07o (34/54) sont positifs pour le Groenland et 68 °7o (50/74) pour le Danemark ; les taux de positivit6s des cancers invasifs sont respectivement de 82 (23/28) et de 70 o7o (23/33). Ces r~sultats confirment nos pr~c6dentes 6tudes effectu~es darts la population et qui ont montr6 des pr~valences similaires de H P V chez les Groenlandaises eties Danoises, bien que I'incidence du cancer cervical au Groenland for de 4 h 5 fois sup6rieur. Mots-clds: Tumeur, Col de l'ut6rus, Papillomavirus; P C R , Hybridation in situ, Groenland, Danemark.

References Bosch, F.X. & Munoz, N. (1989), Human papillomavirus and cervical neoplasia: a critical review of the epidemiological evidence, in "Human papillomavirus and cervical cancer" iN. Munoz, F.X. Bosch & O.M. Jensen), 94 (pp. 135-151). IARC Scientific Publ., Lyon. Cole, S.T. & Streeck, R.E. (1986), Genome organization and nucleotide sequence of human papillomavirus type 33, which is associated with cervical cancer. J. Virol., 58, 991-995. De Villiers, E. (1989), Heterogeneity of the human papilIomavirus group. J. Virol., 63, 4898-4903. Goldsborough, M.D., Di Dilvestre, D., Temple, G.F. & Lorincz, A.T. 0989), Nucleotide sequence of human papillomavirus type 31 : a cervical neoplasia-associated virus. Virology, 171, 306-311. Gergoire, L., Arella, M., Campione-Piccardo, J. & Lancaster, E.D. (1989), Amplification of human papilIomavirus DNA sequences by using conserved primers. J. Clin. Microbiol., 27, 1660-2665. Ho, L., Chan, S.-Y., Burk, R.D., Das, B.C., Fujinaga, K., lcenogle, J.P., Kahn, T., Kiviat, N., Lancaster, W., Mavromara, P., Mitrani-Rosenbaum, S., Norrild, B., Pillai, M.R., Stoerker, J., Syrjaenen, K., Syrjaenen, S., Tay, S.-K., Villa, L., Wheeler, C.M., WiUiamson, A.-L. & Bernard, H.-U. (1993), The genetic drift of human papillomavirus type 16 is a means of reconstructing prehistoric viral spread and movement of ancient human populations. J. Virol. (in press). Kjaer, S.K., De Villiers, E.-M., Haugaard, B.J., Christensen, R.B., Teisen, C., MCller, K.A., Poll, P., Jensen, H., Vestergaard, B.F., Lynge, E. & Jensen, O.M.

91

(1988), Human papillomavirus, herpes simplex virus, and cervical cancer incidence in Greenland and Denmark ; a population-based cross-sectional study. Int. J. Cancer, 41, 518-524. Kjaer, S.K., De Villiers, E.-M., Caglayan, H., Svare, E., Haugaard, B.J., Engholm, G., Christensen, R.B., MCller, K.A., Poll, P., Jensen, H., Verstergaard, B.F., Lynge, E. & Jensen, O.M. (1993), Human papillomavirus, herpes simplex virus and other potential risk factors for cervical cancer in a high-risk area (Greenland) and a low-risk area (Denmark) - - a second look. Brit. J. Cancer, 67, 830-837. Manos, M.M., Ting, Y., Wright, D.K., Lewis, A.J., Broker, T.R. & Wolinsky, S.M. (1989), Use of polymerase chain reaction amplification for the detection of genital human papillomavirus. Cancer Celles, 7, 209-214. Munoz, N. & Bosch, F.X. (1992), HPV and cervical neoplasia: review of case-control and cohort studies, in "The epidemiology of cervical cancer and human papillomavirus" (N. Munoz, F.X. Bosch, K.V. Shah & A. Meheus), ll9 (pp. 251-261). IARC Scientific Puhl., Lyon. Munoz, N., Bosch, F.X., de Sanjose, S., Tafur, L., lzarzugaza, I., Gili, M., Viladiu, P., Navarro, C., Martos, C., Ascunce, N., Gonzalez, L.C., Kaldor, J.M., Guerrero, E., Lperincz, A., Santamaria, M., Alonso de Ruiz, P., Aristizabal, N. & Shah, K. (1992), The causal link between human papillomavirus and invasive cervical cancer: a population-based case-control study in Colombia and Spain. Int. J. Cancer, 52, 743-749. Nielsen, H.N., Jensen, H. & Hansen, P.K. (1988), Cervical cytology in Greenland and occurrence of cervical carcinoma in situ, and dysplasia. Extent and impact of uncoordinated screaning activity 1976-1985. Arct. Med. Res., 47, 179-188. Nuovo, G.J. (1991), Comparison of Bouin solution and buffered formalin fixation on the detection rate by in situ hybridization of human papillomavirus DNA in genital tract lesions. J. Histotechnol., 14, 13-17. Nuovo, G.J., Darfler, M.M., lmpraim, C.C. & Bromley, S.E. (1991), Occurrence of multiple types of human papillomavirus in genital tract lesions. Am. J. PathoL, 138, 53-58. Pfister, H. (1992), Papillomaviren, Krebs and Immunsuppression. AIFO Heft, 3, 115-124. Prener, A., Nielsen, N.H., Storm, H.H., Hansen, J.P.H. & Jensen, O.M. (1991), Cancer in Greenland 1953-1985. APMIS (suppl. 20), 99, 38-40. Saiki, R.K., Bugawan, T.L., Glenn, T.H., Mullis, K.B. & Erlich, H.A. (1986), Analysis of enzymatically amplified beta-globin and HLA-DQ alpha DNA with allele-specific oligonucleotide probes. Nature (Lond.), 324, 163. Saiki, R.K., Scharf, S., Faloona, F., Mullis, K.B., Horn, G.T., Erlich, H.A. & Arnheim, N. (1985), Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science, 230, 1350-1354. Seedorf, K., Kraemmer, G., Duerst, M., Suhai, S. & Roewekamp, W.G. (1985), Human papillomavirus type 16 DNA sequence. Virology, 145, 181-185. Shibata, D.K., Arnheim, N. & Martin, W.J. (1988), Detection of human papilloma virus in paraffin-

92

A.M. SEBBELOV ET AL.

embedded tissue using the polymerase chain reaction. J. Exp. Med., 167, 225-230. Smits, H.L., Tieben, L.M., Tjong-A-Hung, S.P., Jebbink, M.F., Minnaar, R.P., Jansen, C.L. & ter Schegget, J. (1992), Detection and typing of human papillomaviruses present in fixed and stained archival cervical smears by a consensus polymerase chain reaction and direct sequence analysis allow the identification of a broad-spectrum of human papillomavirus types. J. Gen. Virol., 73, 3263-3268. Snijders, P.J.F., Van den Brule, A.J.C., Schrijnemakers, H.F.J., Snow, G., Meijer, C.J.L.M. & Wallboomers, J.M.M. (1990), The use of general primers in the polymerase chain reaction permits the detection of a broad

spectrum of human papillomavirus genotypes. J. Gen. ]tirol., 71, 173-181. Storm, H.H., Manders, T., Friis, S. & Bang, S. (1988), Cancer incidence in Denmark 1988. Danish Cancer Society, Copenhagen. Wright, D.K. & Manos, M.M. (1990), PCR protocools: a guide to methods and applications (M.A. lnnis, D.H. Gelfand, J.J. Sninsky & T.J. White) (pp. 153-168). Academic Press, London, New York. Young, L.S., Bevan, I.S., Johnson, M.A., Blomfield, P.I., Bromidge, T., Maitland, N.J. & Woodman, C.B.J. (1989), The polymerase chain reaction: a new epidemiological tool for investigating cervical human papillomavirus infection. Brit. Med. J., 298, 14-18.