- Email: [email protected]
Cytological regression and clearance of high-risk human papillomavirus in women with an abnormal cervical smear
RESEARCH LETTERS
Cytological regression and clearance of high-risk human papillomavirus in women with an abnormal cervical smear
Women with high-risk HPV
1·0 0·9
Mild dyskaryosis Moderate dyskaryosis Severe dyskaryosis
0·8 0·7 0·6
Mariëlle A E Nobbenhuis, Theo J M Helmerhorst, Adriaan J C van den Brule, Lawrence Rozendaal, Feja J Voorhorst, P Dick Bezemer, René H M Verheijen, Chris J L M Meijer
0·5 0·4
Only a small proportion of women infected with high-risk human papillomavirus (HPV) will develop cervical intraepithelial neoplasia (CIN) 3 or worse, and most premalignant lesions regress spontaneously.1,2 Therefore most women with abnormal cervical cytology will be overtreated. We used data from an earlier study3 to find out whether clearance of high-risk HPV predicts cytological regression. The previous study was a prospective, nonintervention study in women referred to the colposcopy clinic of University Hospita Vrije Universiteit, Amsterdam, Netherlands, because of an abnormal cervical smear between June, 1990, and December, 1992. Enrolment into the study (baseline) took place about 2 months after referral. No biopsy samples were taken during the study to avoid interference with the natural course of the disease. The women were followed up every 3–4 months by cytology, colposcopy, and testing for high-risk HPV by GP5+/6+ PCR. Study design and participants’ characteristics are presented in the previous report.3 Women were judged to be clear of infection when no high-risk type of HPV from the previous visit was detected at the next visit. Regression of abnormal cervical cytology was defined as a return to normal cytology (at least two consecutive smears read as normal). Time to event ie, cytological regression and high-risk HPV clearance—was taken to be the midpoint between positive and negative test results. Viral clearance and cytological regression rates were calculated by Kaplan-Meier curves. Groups were compared by use of log-rank tests and, if appropriate, were tested for linear trend. Cytological regression was compared with high-risk HPV status at baseline (ie, negative or positive), age in tertile groups, and cervical cytology status at baseline. HPV clearance was compared with cervical cytology status at baseline and age in 5-year intervals. The sequence of high-risk HPV clearance and cytological regression was compared in women who reached both events with paired Student’s t test. This analysis was repeated stratifying for age, cervical smear status at baseline, and high-risk HPV acquisition after initial clearance. 353 women were referred during the study period. The mean age was 32 years (range 18–55), and median follow-
1782
0·2 log-rank p=0·003
0·1 0·0
Women without high-risk HPV Mild dyskaryosis Moderate/severe dyskaryosis
1·0 0·9 0·8 0·7 0·6 0·5 0·4
log-rank p=0·99
0·3 0·2 0·1 0·0
High-risk HPV clearance Normal cytology Very mild dyskaryosis Mild dyskaryosis Moderate dyskaryosis Severe dyskaryosis
1·0 Cumulative rate of HPV clearance
Lancet 2001; 358: 1782–83
Cumulative rate of cytological regression
0·3 We studied the natural course of high-risk human papillomavirus (HPV) infection and cytological regression in women referred for colposcopy because of abnormal cervical smears. We found that high-risk HPV clearance preceded regression of cervical lesions by an average of 3 months. The cumulative 1-year rate of cytological regression was similar in women with mild and moderate dyskaryotic cervical smears. Thus, retesting of high-risk HPV after 6 months in women with mild to moderate dyskaryosis predicts cytological regression.
0·9 0·8 0·7 0·6 0·5 0·4 0·3
log-rank p=0·2
0·2 0·1 0·0 0
12
24
36
48
60
Follow-up time (months) Cytological regression in women with or without high-risk human papillomavirus (HPV) infection, and clearance of highrisk HPV stratified by cytological status at baseline
THE LANCET • Vol 358 • November 24, 2001
For personal use. Only reproduce with permission from The Lancet Publishing Group.
RESEARCH LETTERS
Follow-up (months)
Percentage cumulative incidence (95% CI) Cytological regression
End histology CIN 3
Clinical progression
Mild dyskaryosis (n=67) 6 20·6 (10·4–30·8) 12 37·2 (24·8–49·6) 24 54·9 (41·9–67·9)
3·0 (0·0–7·2) 11·2 (3·2–19·2) 19·8 (9·5–30·1)
.. 6·8 (0·3–13·3) 6·8 (0·3–13·3)
Moderate dyskaryosis (n=51) 6 17·8 (7·0–28·6) 12 33·6 (19·8–47·4) 24 50·5 (34·6–66·4)
4·0 (0·0–9·6) 8·1 (0·3–15·9) 27·0 (14·2–39·8)
.. 2·1 (0·0–6·3) 9·5 (0·4–18·6)
Severe dyskaryosis (n=51) 6 12·4 (2·9–21·9) 12 12·4 (2·9–21·9) 24 18·4 (6·3–30·5)
8·0 (0·3––15·7) 35·2 (21·4–49·0) 59·6 (45·1–74·1)
.. 19·9 (7·3–32·5) 38·5 (21·9–55·1)
Cumulative incidence of cytological regression, end-histology CIN 3, and clinical progression in high-risk HPV-positive women with mild, moderate, or severe dyskaryosis at baseline
up time was 33 months (range 3–74). We found that 218 (62%) women had abnormal cervical smears at baseline, 132 (37%) had normal cytology or very mild dyskaryosis, and three (1%) had an inadequate smear. The cumulative 1-year rate of cytological regression was 38% (95% CI 31–45). Cytological regression was not influenced by age (data not shown). In 169 women positive for high-risk HPV with mild (n=67), moderate (n=51), or severe (n=51) dyskaryosis, the 1-year cumulative rates of cytological regression were 37% (95% CI 25–50), 34% (95% CI 20–47), and 12% (95% CI 3–22), respectively (figure). The corresponding median regression times were 17 months (95% CI 6–27), 24 months (95% CI 1–50), and greater than 60 months (95% CI >60), respectively. In 49 women negative for high-risk HPV with mild (n=41) or moderateto-severe (n=8) dyskaryosis, these rates were 68% (95% CI 54–84) and 53% (95% CI 16–90), respectively (figure). The corresponding median regression times were 5 months (95% CI 1–10) and 6 months (95% CI 1–17), respectively. 230 women with an adequate cervical smear at baseline had a positive high-risk HPV test (61 with normal cytology or very mild dyskaryosis, and 169 with mild, moderate, or severe dyskaryosis). The cumulative 1-year rate of HPV clearance was 35% (95% CI 28–42). Clearance was not influenced by age (data not shown). The 1-year cumulative rates of high-risk HPV clearance in women with normal cytology, very mild, mild, moderate, and severe dyskaryosis were 46% (95% CI 26–66), 50% (95% CI 32–68), 29% (95% CI 18–50), 23% (95% CI 11–35), and 25% (95% CI 12–38), respectively (figure). All women who cleared high-risk HPV did so within 40 months of follow-up. Clearance-rates of high-risk HPV decreased with increasing severity of the lesion (log-rank trend p=0·02). The mean age of 79 women with cytological regression and high-risk HPV clearance was 33 years (range 21–55). Clearance occurred on average 3 months (range ⫺25 to 44; symmetrically distributed) before cytological regression (p<0·05). Age, cervical smear status at baseline, and acquisition of high-risk HPV during follow-up did not influence the outcome of this sequence. Cytological regression occurred without HPV clearance in eight women, among whom a biopsy done at the end of follow-up showed one case each of CIN 1 (mild dysplasia) and CIN 2 (moderate
THE LANCET • Vol 358 • November 24, 2001
dysplasia), and six of CIN 3 (severe dysplasia). High-risk HPV clearance occurred without cytological regression in four women, of whom the biopsy showed two cases of no CIN and one each of CIN 1 and CIN 2. Cytological regression was seen more often in women without high-risk HPV than in women with high-risk HPV infection. The cumulative incidence of cytological regression after 4 years was 100% in high-risk HPV-negative women with mild dyskaryosis, and 85% in women with moderate to severe dyskaryosis. Regression time was longer in women positive for high risk HPV if lesions were more severe. Highrisk HPV clearance occurred on average 3 months before regression time. These findings indicate that regression of cervical lesions is preceded by high-risk HPV clearance. Our regression rates are supported by others who have studied the natural history of cervical lesions.1,2,4 Reported regression rates varied from 30% to 62% in women with mild dysplastic lesions, and from 17% to 54% in those with moderate dysplastic lesions. These studies differ from our study on two points: first, in the other studies, no allowance was made for high-risk HPV status; and second, we defined cytological regression on two consecutive smears. No CIN 3 lesions were found in women who cleared HPV without cytological regression. Six CIN 3 lesions were found in eight women with cytological regression without HPV clearance. Thus, a negative test for high-risk HPV decreases the chance of finding CIN 3.3 Another study shows similar results.5 We suggest that over-treatment in women with abnormal smears could be prevented by implementing a wait-and-see period to allow high-risk HPV clearance and subsequent regression of the lesion. However, the time-interval until retesting should be traded off against the chance of regression and of developing CIN 3 (table). In conclusion, in women with abnormal cervical smears, high-risk HPV clearance predicts cytological regression. 1 2
3
4
5
Östor AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynaecol Pathol 1993; 12: 186–92. Holowaty P, Miller AB, Rohan T, To T. Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst 1999; 91: 252–58. Nobbenhuis MAE, Walboomers JMM, Helmerhorst TJM, et al. Relation of human papillomavirus status to cervical lesions and consequences for cervical-cancer screening: a prospective study. Lancet 1999; 354: 20–25. Nasiell K, Roger V, Nasiell M. Behavior of mild cervical dysplasia during long-term follow-up. Obstet Gynecol 1986; 67: 665–68. Zielinski GD, Snijders PJF, Rozendaal L, et al. High-risk HPV testing in women with borderline and mild dyskaryosis: long term follow-up data and clinical relevance. J Pathol 2001; 195: 300–06.
Departments of Pathology (M A E Nobbenhuis MD, A J C van den Brule PhD, L Rozendaal MD, Prof C J L M Meijer MD), Clinical Epidemiology and Biostatistics (F J Voorhorst MD, P D Bezemer PhD), and Obstetrics and Gynaecology (R H M Verheijen MD), Vrije Universiteit Medical Centre, PO Box 7057, 1007 MB, Amsterdam, Netherlands; and Department of Obstetrics and Gynaecology, University Hospital Rotterdam, Rotterdam, Netherlands (Prof T J M Helmerhorst MD) Correspondence to: Prof Chris J L M Meijer (e-mail: [email protected])
1783
For personal use. Only reproduce with permission from The Lancet Publishing Group.
Cytological regression and clearance of high-risk human papillomavirus in women with an abnormal cervical smear
Women with high-risk HPV
1·0 0·9
Mild dyskaryosis Moderate dyskaryosis Severe dyskaryosis
0·8 0·7 0·6
Mariëlle A E Nobbenhuis, Theo J M Helmerhorst, Adriaan J C van den Brule, Lawrence Rozendaal, Feja J Voorhorst, P Dick Bezemer, René H M Verheijen, Chris J L M Meijer
0·5 0·4
Only a small proportion of women infected with high-risk human papillomavirus (HPV) will develop cervical intraepithelial neoplasia (CIN) 3 or worse, and most premalignant lesions regress spontaneously.1,2 Therefore most women with abnormal cervical cytology will be overtreated. We used data from an earlier study3 to find out whether clearance of high-risk HPV predicts cytological regression. The previous study was a prospective, nonintervention study in women referred to the colposcopy clinic of University Hospita Vrije Universiteit, Amsterdam, Netherlands, because of an abnormal cervical smear between June, 1990, and December, 1992. Enrolment into the study (baseline) took place about 2 months after referral. No biopsy samples were taken during the study to avoid interference with the natural course of the disease. The women were followed up every 3–4 months by cytology, colposcopy, and testing for high-risk HPV by GP5+/6+ PCR. Study design and participants’ characteristics are presented in the previous report.3 Women were judged to be clear of infection when no high-risk type of HPV from the previous visit was detected at the next visit. Regression of abnormal cervical cytology was defined as a return to normal cytology (at least two consecutive smears read as normal). Time to event ie, cytological regression and high-risk HPV clearance—was taken to be the midpoint between positive and negative test results. Viral clearance and cytological regression rates were calculated by Kaplan-Meier curves. Groups were compared by use of log-rank tests and, if appropriate, were tested for linear trend. Cytological regression was compared with high-risk HPV status at baseline (ie, negative or positive), age in tertile groups, and cervical cytology status at baseline. HPV clearance was compared with cervical cytology status at baseline and age in 5-year intervals. The sequence of high-risk HPV clearance and cytological regression was compared in women who reached both events with paired Student’s t test. This analysis was repeated stratifying for age, cervical smear status at baseline, and high-risk HPV acquisition after initial clearance. 353 women were referred during the study period. The mean age was 32 years (range 18–55), and median follow-
1782
0·2 log-rank p=0·003
0·1 0·0
Women without high-risk HPV Mild dyskaryosis Moderate/severe dyskaryosis
1·0 0·9 0·8 0·7 0·6 0·5 0·4
log-rank p=0·99
0·3 0·2 0·1 0·0
High-risk HPV clearance Normal cytology Very mild dyskaryosis Mild dyskaryosis Moderate dyskaryosis Severe dyskaryosis
1·0 Cumulative rate of HPV clearance
Lancet 2001; 358: 1782–83
Cumulative rate of cytological regression
0·3 We studied the natural course of high-risk human papillomavirus (HPV) infection and cytological regression in women referred for colposcopy because of abnormal cervical smears. We found that high-risk HPV clearance preceded regression of cervical lesions by an average of 3 months. The cumulative 1-year rate of cytological regression was similar in women with mild and moderate dyskaryotic cervical smears. Thus, retesting of high-risk HPV after 6 months in women with mild to moderate dyskaryosis predicts cytological regression.
0·9 0·8 0·7 0·6 0·5 0·4 0·3
log-rank p=0·2
0·2 0·1 0·0 0
12
24
36
48
60
Follow-up time (months) Cytological regression in women with or without high-risk human papillomavirus (HPV) infection, and clearance of highrisk HPV stratified by cytological status at baseline
THE LANCET • Vol 358 • November 24, 2001
For personal use. Only reproduce with permission from The Lancet Publishing Group.
RESEARCH LETTERS
Follow-up (months)
Percentage cumulative incidence (95% CI) Cytological regression
End histology CIN 3
Clinical progression
Mild dyskaryosis (n=67) 6 20·6 (10·4–30·8) 12 37·2 (24·8–49·6) 24 54·9 (41·9–67·9)
3·0 (0·0–7·2) 11·2 (3·2–19·2) 19·8 (9·5–30·1)
.. 6·8 (0·3–13·3) 6·8 (0·3–13·3)
Moderate dyskaryosis (n=51) 6 17·8 (7·0–28·6) 12 33·6 (19·8–47·4) 24 50·5 (34·6–66·4)
4·0 (0·0–9·6) 8·1 (0·3–15·9) 27·0 (14·2–39·8)
.. 2·1 (0·0–6·3) 9·5 (0·4–18·6)
Severe dyskaryosis (n=51) 6 12·4 (2·9–21·9) 12 12·4 (2·9–21·9) 24 18·4 (6·3–30·5)
8·0 (0·3––15·7) 35·2 (21·4–49·0) 59·6 (45·1–74·1)
.. 19·9 (7·3–32·5) 38·5 (21·9–55·1)
Cumulative incidence of cytological regression, end-histology CIN 3, and clinical progression in high-risk HPV-positive women with mild, moderate, or severe dyskaryosis at baseline
up time was 33 months (range 3–74). We found that 218 (62%) women had abnormal cervical smears at baseline, 132 (37%) had normal cytology or very mild dyskaryosis, and three (1%) had an inadequate smear. The cumulative 1-year rate of cytological regression was 38% (95% CI 31–45). Cytological regression was not influenced by age (data not shown). In 169 women positive for high-risk HPV with mild (n=67), moderate (n=51), or severe (n=51) dyskaryosis, the 1-year cumulative rates of cytological regression were 37% (95% CI 25–50), 34% (95% CI 20–47), and 12% (95% CI 3–22), respectively (figure). The corresponding median regression times were 17 months (95% CI 6–27), 24 months (95% CI 1–50), and greater than 60 months (95% CI >60), respectively. In 49 women negative for high-risk HPV with mild (n=41) or moderateto-severe (n=8) dyskaryosis, these rates were 68% (95% CI 54–84) and 53% (95% CI 16–90), respectively (figure). The corresponding median regression times were 5 months (95% CI 1–10) and 6 months (95% CI 1–17), respectively. 230 women with an adequate cervical smear at baseline had a positive high-risk HPV test (61 with normal cytology or very mild dyskaryosis, and 169 with mild, moderate, or severe dyskaryosis). The cumulative 1-year rate of HPV clearance was 35% (95% CI 28–42). Clearance was not influenced by age (data not shown). The 1-year cumulative rates of high-risk HPV clearance in women with normal cytology, very mild, mild, moderate, and severe dyskaryosis were 46% (95% CI 26–66), 50% (95% CI 32–68), 29% (95% CI 18–50), 23% (95% CI 11–35), and 25% (95% CI 12–38), respectively (figure). All women who cleared high-risk HPV did so within 40 months of follow-up. Clearance-rates of high-risk HPV decreased with increasing severity of the lesion (log-rank trend p=0·02). The mean age of 79 women with cytological regression and high-risk HPV clearance was 33 years (range 21–55). Clearance occurred on average 3 months (range ⫺25 to 44; symmetrically distributed) before cytological regression (p<0·05). Age, cervical smear status at baseline, and acquisition of high-risk HPV during follow-up did not influence the outcome of this sequence. Cytological regression occurred without HPV clearance in eight women, among whom a biopsy done at the end of follow-up showed one case each of CIN 1 (mild dysplasia) and CIN 2 (moderate
THE LANCET • Vol 358 • November 24, 2001
dysplasia), and six of CIN 3 (severe dysplasia). High-risk HPV clearance occurred without cytological regression in four women, of whom the biopsy showed two cases of no CIN and one each of CIN 1 and CIN 2. Cytological regression was seen more often in women without high-risk HPV than in women with high-risk HPV infection. The cumulative incidence of cytological regression after 4 years was 100% in high-risk HPV-negative women with mild dyskaryosis, and 85% in women with moderate to severe dyskaryosis. Regression time was longer in women positive for high risk HPV if lesions were more severe. Highrisk HPV clearance occurred on average 3 months before regression time. These findings indicate that regression of cervical lesions is preceded by high-risk HPV clearance. Our regression rates are supported by others who have studied the natural history of cervical lesions.1,2,4 Reported regression rates varied from 30% to 62% in women with mild dysplastic lesions, and from 17% to 54% in those with moderate dysplastic lesions. These studies differ from our study on two points: first, in the other studies, no allowance was made for high-risk HPV status; and second, we defined cytological regression on two consecutive smears. No CIN 3 lesions were found in women who cleared HPV without cytological regression. Six CIN 3 lesions were found in eight women with cytological regression without HPV clearance. Thus, a negative test for high-risk HPV decreases the chance of finding CIN 3.3 Another study shows similar results.5 We suggest that over-treatment in women with abnormal smears could be prevented by implementing a wait-and-see period to allow high-risk HPV clearance and subsequent regression of the lesion. However, the time-interval until retesting should be traded off against the chance of regression and of developing CIN 3 (table). In conclusion, in women with abnormal cervical smears, high-risk HPV clearance predicts cytological regression. 1 2
3
4
5
Östor AG. Natural history of cervical intraepithelial neoplasia: a critical review. Int J Gynaecol Pathol 1993; 12: 186–92. Holowaty P, Miller AB, Rohan T, To T. Natural history of dysplasia of the uterine cervix. J Natl Cancer Inst 1999; 91: 252–58. Nobbenhuis MAE, Walboomers JMM, Helmerhorst TJM, et al. Relation of human papillomavirus status to cervical lesions and consequences for cervical-cancer screening: a prospective study. Lancet 1999; 354: 20–25. Nasiell K, Roger V, Nasiell M. Behavior of mild cervical dysplasia during long-term follow-up. Obstet Gynecol 1986; 67: 665–68. Zielinski GD, Snijders PJF, Rozendaal L, et al. High-risk HPV testing in women with borderline and mild dyskaryosis: long term follow-up data and clinical relevance. J Pathol 2001; 195: 300–06.
Departments of Pathology (M A E Nobbenhuis MD, A J C van den Brule PhD, L Rozendaal MD, Prof C J L M Meijer MD), Clinical Epidemiology and Biostatistics (F J Voorhorst MD, P D Bezemer PhD), and Obstetrics and Gynaecology (R H M Verheijen MD), Vrije Universiteit Medical Centre, PO Box 7057, 1007 MB, Amsterdam, Netherlands; and Department of Obstetrics and Gynaecology, University Hospital Rotterdam, Rotterdam, Netherlands (Prof T J M Helmerhorst MD) Correspondence to: Prof Chris J L M Meijer (e-mail: [email protected])
1783
For personal use. Only reproduce with permission from The Lancet Publishing Group.