General
information Research
projects Progress
report Publications
in English Publications
in Finnish Publications
1978-2005 Supervised
MD PhD programs
Progress Report 1996-2005
I. C.trachomatis infections
II Treatment of menorrhagia
III Prevention of preterm birth
IV Prerequisites for HPV vaccination trial
V. Transmission of genital herpes
VI. OtherI. C.trachomatis infections
Screening for Chlamydia trachomatis infections
1. Screening for C.trachomatis by PCR using first void urine (FVU) testing is as effective as testing of cervical swab samples in women (Sex Transm Dis 1997; 24:343). Thus, gynecologic examination is not necessary for the diagnosis of C.trachomatis infection in women.2. We compared the performances of a PCR test (Roche Cobal Amplicor), an LCR test (Abbott Lcx), and conventional cell culture using first-void urine and endocervical and urethral specimens collected from 1,015 patients (J Clin Microbiol 1998; 36:1489). These techniques (PCR, LCR) detected approximately 30% more C.trachomatis infections than the conventional cell culture. The sensitivity of cell culture was 79%, 92% for PCR, and 93% for LCR. PCR and LCR showed excellent consistency with kappa coefficients of 93% and 92%.
3. We developed a decision tree to perform a cost-benefit analysis of a C.trachomatis screening program based on first-void urine testing of asymptomatic women using a PCR test. We found that universal screening for C.trachomatis is cost-beneficial (benefits or outputs exceed costs or inputs) if the prevalence exceeds 3.9% (Obstet Gynecol 1998; 92:292). The conclusion was that socioeconomic studies linking secondary prevention of C.trachomatis infection and infertility or adverse pregnancy outcome are needed to convince public health authorities of the need for and the benefit of such programs.
4. We studied trends of gonococcal and chlamydial infections in STD clinic settings in Helsinki metropolitan area during a 10 year period (Sex Transm Dis 1998; 25:181). While there was a significant decrease in the rates of gonococcal infections, the rates of chlamydial infections have remained high. This suggests that current STD prevention programs (i.e. testing and treating symptomatic individuals) are ineffective against C.trachomatis infections. Thus, secondary prevention by universal screening programs should be implemented in order to limit the current chlamydia epidemic (Genitourin Med 1997;73:103).
5. We defined criteria for selective screening for C. trachomatis infection among approximately 1200 women attending a family planning clinic and a university student health clinic. The overall prevalence of C. trachomatis infection was 3.5%. We found that screening women aged (25 would identify only 29% of all chlamydial infections, while screening all women aged (30 would identify 83% of all infections. By logistic regression, lifetime number of sexual partners, sexual orientation and age were the three variables most predictive of chlamydial infection (Sex Transm Dis 2003; 30:120).
6. During the period 1995-2000, data based on National Infectious Disease Register in Finland showed a significant increase in the incidence of Chlamydia trachomatis infections (from 23.4 per 10,000 to 29.2 per 10,000). The increase was most striking among adolescents and young adults living in non-urban densely populated areas. This core group should therefore be a target for C.trachomatis screening and counseling programmes (Sex Transm Dis 2003; 30:737).
7. The incidence of chlamydial infections has been steadily increasing in Finland and elsewhere (13,357 cases were reported in 2004). The increase has been most striking in the youngest age groups. Efforts to control chlamydial infections have been disappointing suggesting that the current prevention strategies have failed (Duodecim 2005;121:1273). We have demonstrated that the prevalence of chlamydial infections in unscreened asymptomatic women in different European countries ranges from 1.7%-17% (mean 6%) (Human Reprod Update 2002;8:385). In Finland, we have systematically studied the epidemiology and prevalence of asymptomatic chlamydial infections using sentinel clinic network (Sex Transm Dis 2003;30:737) among Student Health Clinic women during 200-2003 (Suomen Lääkäril 2005;60:261), and most recently among 16-17 year old adolescents enrolled in phase 3 HPV vaccination trials (range 1.2%-5.3%).
Sequelae of C.trachomatis infections
1. We used magnetic resonance imaging (MRI) for the diagnosis of pelvic inflammatory disease (PID), and defined the criteria for the MRI diagnosis of PID (Radiology 1999; 210:209). We found that MRI was more accurate than endovaginal ultrasound examination when laparoscopic diagnosis of PID was used as a gold standard. Our paper was considered as a landmark paper since it was published in the top radiology journal.
2. We showed that silent C.trachomatis infection is a common cause of abnormal menstrual bleeding in fertile aged women (Am J Clin Pathol 1999; 112:211). Overall, 27% of such cases had chlamydial upper genital tract infection. Therefore, endometrial sampling is important in the evaluation of young women with bleeding problems.
3. We found no seroepidemiologic evidence of C.trachomatis as a cause of recurrent pregnancy loss (RPL) (Fertil Steril 1999; 72:427). Some, but not all previous studies had suggested that there is a link between past history of C.trachomatis infection and RPL.
4. We studied the efficacy of acute phase operative laparoscopy in 37 women hospitalized for suspected PID. Laparoscopic intervention could be performed in most women with or without PID. We found such patient management highly cost-effective since further hospitalization and future elective operations are often not necessary and prolonged antimicrobial therapy can be avoided (J Am Assoc Gyn Lap 2000; 7:107).
5. We studied the usefulness of Power Doppler transvaginal sonography (TVS) in the diagnosis of PID among 30 women hospitalized for acute PID and 20 control women with proven hydrosalpinx formation. A combined use of specific sonographic findings and Power Doppler findings augmented the clinical diagnosis of PID and allowed simple classification of the severity of the disease. Overall, Power Doppler TVS was 100% sensitive and 95% specific in the diagnosis of acute PID (Ultrasound Obstet Gynecol, 2001;17:233-238) . We also found that power Doppler TVS can distinguish laparoscopically proven acute appendicitis from PID, avoiding the limitations of transabdominal ultrasound examination. Power Doppler TVS seems to be a promising new technique for the diagnosis of appendicitis (Ultrasound Obstet Gynecol 2002; 20:496).
6. We described endometritis as the clinical-pathological syndrome by performing a cross sectional study of 152 women with suspected PID of whom 43 had neither endometritis nor salpingitis, 26 had endometritis alone without salpingitis, and 83 had salpingitis alone. We found that clinical and laboratory abnormalities were associated with endometritis less often than with salpingitis, but more often than with neither salpingitis nor endometritis. The histopathological manifestations of endometritis were associated with clinical manifestations, chlamydial or gonococcal infection, and other specific risk factors (Am J Obstet Gynecol 2002; 6: 690).
7. We tested the intraobserver (are observers likely to agree between themselves?) and interobserver (are observers likely to agree with other observers?) reproducibility and the overall diagnostic accuracy of pelvic inflammatory disease (PID). Based on photo images, the observer reproducibility and the accuracy of the laparoscopic diagnosis of PID were both unsatisfactory when histopathologically proven PID was used as the gold standard. Thus, although the clinical diagnosis of PID remains extremely difficult, laparoscopy seems not to solve the diagnostic problems because of lack of observer agreement and lack of diagnostic sensitivity (Obstet Gynecol 2003; 101:875).
8. Using the Nordic Serum Bank material and cancer registries we found that chlamydial heat shock protein 60-1 antibodies were associated with an increased risk for cervical cancer (Am J Obstet Gynecol 2003;189:1287). This suggests that persistent C.trachomatis infection may contribute to cervical neoplasia.
9. We performed a seroepidemiologic case-control study of an association between past C.trachomatis infection and primary fallopian tube carcinoma among 79 cases diagnosed in 1985-2000 and 156 controls. There was no association between C.trachomatis and fallopian tube carcinoma (Eur J Cancer, in press).
Host immune response to Chlamydia
1. We found evidence of crossreacting humoral antibody response against specific epitopes of chlamydial HSP60 protein and human HSP60 in patients with proven chlamydial PID (J Infect Dis 1998;177:714). This suggests that autoimmunity, based on molecular mimicry of HSP60, plays a role in the immunopathogenesis of chlamydial infection. This is in line with our earlier epitope scanning study in which crossreacting B cell epitopes were detected in infants with chlamydial pneumonitis and human HSP60 (J Infect Dis 1994; 169:908).
2. We studied the relation and reactivity of T lymphocytes of human carotid plaques to C.pneumoniae antigens in tissue specimens obtained from 17 patients who underwent carotid endarterectomy. C.pneumoniae was present in 64% of the cases. T-cell lines were established from all 17 tissue specimens and C.pneumoniae was recognized as a spesific T-cell stimulating antigen in 41% of the cases. Chlamydial 60-kDa heat-shock protein induced specific proliferation in 5 of 7 cases and revealed 2 haplotype binding motifs in the human 60-kDa heat-shock protein. These results suggest that cell-mediated immune response to C.pneumoniae plays a role in the atherosclerotic process and this response may involve autoimmunity (Arterioscler Thromb Vasc Biol 2000; 20:1061).
3. We studied the role of C. trachomatis as a target antigen of endometrial and salpingeal tissue lymphocytes derived from PID patients and tubal factor infertility (TFI) patients. We found that T lymphocytes derived from salpingeal tissue of TFI patients showed specific reactivity to chlamydial HSP60, and responded to C. trachomatis by producing interferon-gamma. This provides further evidence that the specific inflammatory response to C. trachomatis contributes to the pathogenesis of TFI (Hum Reprod 2000; 15:1484).
4. We studied the role of chlamydial HSP60 specific T-lymphocytes in TFI by using salpingeal tissue specimens from TFI patients who underwnet salpingectomy because of hydrosalpinx formation. One third of the chlamydia-specific T cell clones (77/229) recognized chlamydial HSP60 as the target antigen. This suggests that chlamydial HSP60 reactive T-lymphocytes play a key role in the inflammatory response of Chlamydia-induced TFI (Fertil Steril 2002; 77:162).
5. We measured proliferative responses of circulating lymphocytes to chlamydial HSP60 and C. trachomatis elementary body (EB) antigens and determined their differentiation into Th1 (IFN-( secreting) cells or Th2 (IL-10 secreting cells) cells in TFI patients and healthy controls. We found that chlamydial HSP60 stimulation specifically produces Th2-dominant responses, whereas stimulation with Ebs enhance Th1 response. Thus, chlamydal HSP60 plays a central role as a regulator of the immune responses seen in chronic C. trachomatis infection. HLA-DQ alleles and interleukin-10 (IL-10) promoter polymorphism were analyzed in C. trachomatis associated TFI. We found an association between cellular immune response to chlamydial HSP60 and HLA class II alleles and IL-10 promoter genotypes in these patients. Specifically HLA-DQB1*0602 and DQA1*0102 alleles were more prevalent in the TFI group than in controls (Hum Reprod 2002; 17:2073).
6. We studied the role of CHSP60-induced immune response in tubal infertility and found a prominent IL-10 response suggesting that CHSP60 may have a spesific role in regulating immune response and subsequent tubal damage during chlamydial infection (Clin Exp Immunol 2003; 131:299).
7. We showed that by combining tests for humoral and cell mediated immune response to C.trachomatis it is possible to improve the prediction of tubal factor infertility when at least three of the four immunological markers were positive, the risk for TFI was 20-fold (Hum Reprod, in press).
8. We found that impaired cell mediated immune response to C.trachomatis was associated with interleukin-10 genotype. Thus IL-10 may explain the development of chronic C.trachomatis infection in individuals with IL-10 prodicing capacity (Genes Immun, in press). These observations are in line with earlier studies suggesting that individual variation in IL-10 gene polymorphism has been linked to various autoimmune, immunosupresive, and inflammatory conditions.
C. trachomatis in cervical cancer
1. In collaboration with other Nordic countries (Nordic Serum Bank Study Group founded in 1994) and registry linkages between serum banks and national cancer registries we have conducted seroepidemiological studies of the association between C.trachomatis and cervical cancer. Our preliminary results suggest that C.trachomatis is an independent risk factor for cervical squamous cell carcinoma whereas no such association was found for C.pneumoniae. (Int J Cancer 2000; 85:35).
2. We next studied the association between exposure to different C.trachomatis serotypes and subsequent development of cervical carcinoma using the same longitudinal nested case-control study design. Of the specific C.trachomatis serotypes, serotype G was most strongly associated with cervical squamous cell carcinoma (OR 6.6, 95% CI 1.6-27, adjusting for serum antibodies to high risk HPV types and serum cotinine). Other serotypes also associated with cervical squamous cell carcinoma were I (OR 3.8, 95% CI 1.3-11), and D (OR 2.7, 95% CI 1.3-5.6). Exposure to more than one serotype increased the risk (p for trend <0.01). These results suggest that specific serotypes are more virulent than others and perhaps less sensitive to antimicrobial therapy. Further studies are needed of the specific mechanisms of the interaction between C.trachomatis and high risk HPVs (JAMA 2001; 285:47).
3. We studied Pap smears and cervical cancer biopsies by PCR for C. trachomatis DNA and HPV DNA from 118 cases who developed cervical cancer after having a normal Pap smear on average 5.6 years earlier, and from 118 matched controls who did not develop cancer. These cases and controls were women participating in the population-based cervical cancer screening program in Västerbotten county, Sweden. We found that a prior cervical C. trachomatis infection was associated with an increased risk for development of invasive cervical cancer (Int J Cancer 2002; 101: 371).
4. We studied in more detail the joint effects of different HPV and Chlamydia trechomatis infections on the risk of cervical squamons cell cancinoma (Eur J Cancer 2004; 40:1058).
C. trachomatis in ovarian cancer
Our preliminary results suggest that serum antichlamydial micro-IF antibodies consistent with past infection with C.trachomatis interact with ovarian cancer tumor marker CA-125. We found that the performance of CA-125 in the diagnosis of serous ovarian neoplasia (carcinomas and borderline tumors) improved when we excluded patients with serum antichlamydial antibodies. (Manuscript in preparation). Thus, chlamydial antibodies modify the performance of CA-125 as a tumor marker and may at least partly explain why CA-125 is not a valid cancer screening test for asymptomatic women if used alone. Therefore, we will study further the link between past C.trachomatis infection and epithelial ovarian cancer using the Nordic Serum Bank material.
II Treatment of menorrhagia
1. Heavy menstrual blood loss (MBL) is a common reason for women to seek medical care. We performed a randomized controlled trial comparing levanorgestrel releasing intrauterine system (LNG-IUS) and hysterectomy in the treatment of 236 women referred for menorrhagia. All 5 university hospitals in Finland participated in the trial. The health related quality of life and psychosocial well-being improved significantly in both study arms. The overall costs were approximately three times higher in the hysterectomy group than in the LNG-IUS group. Thus, after one year the LNG-IUS proved a cost-effective alternative to hysterectomy in the treatment of essential menorrhagia. However, longer follow-up is needed to confirm these findings (Lancet 2001; 357:273).2. We found that a significant proportion of women with complaints of menorrhagia have their measured menstrual blood loss (MBL) within the normal range. Psychosocial factors turned out to have an impact on the health care seeking behavior of such women. Better understanding of the factors which explain complaints of menorrhagia in women with normal bleeding could improve both medical outcomes and reduce the cost of treatment for menorrhagia (Acta Obstet Gynecol Scand 1998; 77:201, Br J Obstet Gynaecol 2001; 108:1).
3. Vascular factors play an important role in the pathophysiology of menorrhagia but remain poorly understood. Therefore we studied whether there is any association between the flow impedance of uterine arteries and the amount of menstrual blood loss (MBL). A significant inverse correlation was found between uterine artery PI and the mount of MBL suggesting that women with lower uterine flow impedance bleed more. Thus, the correlation between uterine artery PI and amount of MBL suggests that vascular factors are involved in the pathogenesis of menorrhagia (Hum Reprod 1999; 14:186).
4. We used immunohistochemistry to study the distribution of sex steroid receptors and proliferation marker Ki-67 in the endometrial tissue of women with and without menorrhagia, before and after 6 and 12 months of use of levonorgestrel-IUS. In women with menorrhagia, secretory phase endometrium exhibited more proliferative activity than in women without menorrhagia. However, no significant differences in the immunoreactivity of the estrogen or progesterone receptors were found suggesting that other factors are involved in the regulation of endometrial proliferation and MBL. Use of LNG-IUS decreased expression of endometrial Ki-67 declined (Mol Hum Reprod 2000; 6:1013).
5. LNG-IUS use was associated with the development of ovarian cysts, but these were symptomless and showed high rate of spontaneous resolution (Ultrasound Obstet Gynecol 2002; 20: 381).
6. Using the same study design as in II.1., we next analysed the effects of hysterectomy or LNG-IUS on serum FSH levels and menopausal symptoms. One year after the treatment, hysterectomized women had higher FSH levels than women with LNG-IUS (p<.005). Furthermore, hot flushes significantly increased in the hysterectomy group (p<.02). These results suggest that hysterectomy may impair ovarian function (Hum Reprod 2004; 19:378).
7. We have completed five-year follow-up of the randomized trial comparing hysterectomy and LNG-IUS in the treatment of essential menorrhagia. A total of 233 (99%) of the 236 women completed the 5-year follow-up. The study groups did not differ in terms of health related quality of life or psychosocial well-being. Although 42% of the women assigned to the LNG-IUS group eventually underwent hysterectomy, the direct and indirect costs in the LNG-IUS group remained significantly lower than in the hysterectomy group. There was equal satisfaction with the treatment in both groups. Thus, in the treatment of menorrhagia, LNG-IUS is a cost-effective alternative to hysterectomy, and is available in primary care. LNG-IUS offers a genuine improvement in patient choice and ultimately reduces the burden of expensive surgical treatment (JAMA 2004; 291:1436).
8. We are now continuing 10-year follow-up of the randomised trial mentioned above. We are now focusing on the risk of osteoporosis and ovarian function in general using serum inhibin-B analyses and ovarian blood flow monitoring by power Doppler ultrasound (manuscripts in preparation).
III Prevention of preterm birth
1. Since several case control studies have suggested that bacterial vaginosis (BV) is an important risk factor for preterm birth and peripartum infections, we performed a randomized controlled intervention trial of antenatal clinic screening and treatment of BV in early pregnancy among 380 BV positive women, using vaginal clindamycin or placebo. The cure rate was 63% in the clindamycin group and 35% in the placebo group. The overall rate of preterm birth was 5% in the clindamycin group and 4% in the placebo group. The rate of peripartum infectious morbidity was 12% in the clindamycin group and 18% in the placebo group. Thus, vaginal clindamycin did not decrease the rate of preterm deliveries or peripartum infections among women with BV. However, recurrent or persistent BV increased the risk for these complications (Obstet Gynecol 2001; 1997:643).
2. We studied whether of leakage of phosphorylated isoform of insulin-like growth factor binding protein-1 (IGFBP-1, a protein of decidua) into cervical secretions is an indicator of tissue damage at the chorio-decidual interface and a better risk marker for preterm birth. We showed a significant difference between women who had BV and positive cervical phIGFBP-1 and those who had BV and negative cervical phIGFBP-1, suggesting than phIGFBP-1 is a marker for adverse pregnancy outcome among BV positive pregnant women (Lancet 1999; 353:1494).
3. We performed a cost-benefit analysis of antenatal clinic screening for BV. We found that such screening is cost-effective in most situations, including asymptomatic women at low risk for preterm birth. BV screening results in major savings in health care costs (Acta Obstet Gynecol Scand 2004; 83:27)
4. Not only bacterial vaginosis (BV), but also periodontitis has been linked to adverse pregnancy outcome (APO). We performed both oral and gynecologic examinations among 239 women who had discontinued birth control and were planning to get pregnant. Among 105 first women who became pregnant, both periodontitis and BV were associated with an increased risk for APO. Our results suggest that prepregnancy counseling should include both oral and vaginal examinations to rule out periodontitis and BV. Such simple interventions may ultimately have a major impact in antenatal health care, and may prove as novel approaches in order to decrease the risk for unexplained APO. (Infect Dis Obstet Gynecol 2005;13:213-216).
IV HPV vaccination trial
1. We have defined attack rates of HPV16 infection and cervical neoplasia in a potential target population for HPV vaccination, using a population based Serum Bank. Among women less than 25 years, the HPV16 attack rate was 4.5%, and the 10-year cumulative incidence rate of cervical neoplasia (CIN 3 plus invasive cancer) was 0.28%. These attack rates are used for sample size estimates for an HPV vaccine trial/s (Sex Transm Inf 2000; 76:13, J Clin Virol 2000; 19:25, Vaccine 2001; 19:1347).
2. We reviewed the current state of the art of effectiveness of preventive HPV vaccination which shows great promise in the control of cervical cancer worldwide (Int J STD & AIDS 2003;14:787).
3. We have participated in phase 3 HPV VLP vaccination trials by Merck (Quadrivalent HPV 6/11/16/18 VLP vaccine) or GlaxoSmithKline HPV 16/18 VLP vaccine since 2002. Approximately 7,500 16-17 year old women have been enrolled from Finland based on population registry. The enrolment has been extremely successful. We have reported this population based enrolment strategy in a recent article (Int J STD&AIDS 2005, in press). First efficacy results of the phase 2-3 vaccination trials were recently reported at the European Cervical Cancer Conference (ECCO 13, November 2005, Paris). The VLP vaccine turned out to be 100% effective against high grade cervical intraepithelial neoplasia (CIN 2/3) which was used a surrogate marker for cervical cancer risk. Phase 3 efficacy trials are ongoing.
4. In a recent Lancet editorial we emphasised that vaccination against human papillomaviruses shows great promise (Lancet 2004;364:1731).
V. Transmission of genital herpes
1. We analyzed 1,484 heterosexual monogamous couples one of whom had clinical and laboratory confirmed HSV-2 and one of whom was susceptible to HSV-2. The sexual partners with HSV-2 were randomized to once daily 500mg valaciclovir or placebo for 8 months. Valaciclovir treatment of the source partner reduced the risk of transmitting symptomatic genital herpes by 75% and overall acquisition of HSV-2 by 48%. Evidence for acquisition of HSV-2 was found in 3.6% of placebo vs. 1.9% of valaciclovir treated patients. Among the HSV-2 infected source partners, once daily aciclovir reduced the frequency of PCR detectable virus in genital swabs from 10.9% of days to 2.6% of days (p<.001) (N Engl J Med 2004; 350:11).
VI. Other
1. Universal antenatal HIV screening has been offered to all pregnant women in Finland since 1998, and the incidence of HIV infection has remained low. The incidence of HIV infection among women delivering in our hospital district increased from 0.6/10,000 to 4.8/10,000 between 1993 and 2002. All HIV positive pregnant women are systematically followed through the pregnancy and monitored for virological response to antiretroviral medication. We studied pregnancy outcome among 45 HIV infected women delivered 1993-2003, and only 1 vertical transmission was detected in a patient who was found HIV positive postpartum (Acta Obstet Gynecol Scand 2005;84:945).