Objective
The aim of this study was to assess the effect of the excision margin after cone for stage Ia1 cervical cancer on long-term outcomes.
Study Design
Retrospective observational study. Patients were divided into 3 groups. Group A underwent immediate reflex hysterectomy; group B had cervical intraepithelial neoplasia (CIN) at the margins but were followed up; group C had clear margins.
Results
We identified 111 women: 19 (17.1%) in group A; 29 (26.1%) in group B; and 63 (56.8%) in group C. Women in group A were older (median, 40 years vs 35 years; P = .0001) with higher rate of endocervical margin involvement (89.5 vs 48.1%, P = .007) than in group B. The women had been followed for a total of 960 woman-years with median follow-up of 398 weeks (quartiles: 258,612). Women with clear margins in the initial excision were more likely to remain free of disease than those with involved ( P < .0001). Further surgery was required due to abnormal cytology in 9 (31.0%) women from group B and 7 (11.1%) from group C ( P = .04). The cumulative rate of recurrent CIN2+ was 6.4% in group B and 2.7% in group C ( P = .17). In group B, recurrences were more common in positive endocervical rather than ectocervical margins (66.6% vs 33.4%, P < .05); all had high-grade CIN at the margins.
Conclusion
The risk of posttreatment CIN2+ is substantially reduced when complete excision is achieved at first treatment. Conservative management is contraindicated in women with microinvasion at the margin. When CIN involves the margin, there is a greater risk of residual disease and of further treatment.
The incidence of early stage screen-detected cervical cancers is increasing. The risk of lymph node metastasis and paracervical tissue extension in stage Ia1 disease is minimal. Local excision or simple hysterectomy is therefore recommended. Commonly, these women are young and wish to preserve the potential for future childbearing. Fertility-sparing treatment interventions aim to minimize reproductive morbidity without adversely affecting recurrence and/or survival. Current recommendations issued by the National Health Service Cervical Screening Programme and the European Federation of Colposcopy and published reports confirm the safety of fertility-sparing conservative management of stage Ia1 disease with local excision followed by close surveillance, provided that the excision margins are free of preinvasive and invasive disease. In cases of positive margins, further surgery is advised.
Disease at the resection margins is recognized as a significant prognostic indicator for residual/recurrent disease following cervical intraepithelial neoplasia (CIN) treatment. Although some carefully selected women treated for CIN with involved excision margins can be managed conservatively, the evidence for women with stage Ia1 disease is very limited. Repeat excision or hysterectomy is usually advised in women with involved margins because of the risk of residual invasive disease.
Because the repeat histology sample often appears to contain no residual invasive or preinvasive disease, and because repeat local excision substantially increases the risk of future preterm birth, the authors of 3 small studies suggested that close surveillance may be a safe alternative in selected women with stage Ia1 disease with only CIN at the excision margins. However, these 3 studies included a total of only 33 women of whom at least 19 had only ectocervical margins involved. A larger study of 140 women of whom 55 underwent repeat excision included women with only early stromal invasion <1 mm deep. Most (67%) women undergoing repeat excision had incomplete excision at the deep margins. The 27% of these women with residual disease included 3 with stromal invasion. In another series, half of the patients (51%) had residual CIN and 2 of 64 (3%) residual multifocal microinvasive cancer in the repeat conization/hysterectomy sample.
There is a need to further assess the value of additional treatment together with the safety of adopting a more conservative approach in selected cases. The aim of this study was to evaluate the long-term outcomes for women treated for stage Ia1 squamous cell carcinoma and to assess how recurrence-free survival correlates to the excision margins.
Materials and Methods
This retrospective observational study was conducted in 2 North West London hospitals (Queen Charlotte’s and St Mary’s University Hospitals) between 1986 and 2011.
Inclusion and exclusion criteria
We included all patients that were treated primary with local excision that diagnosed stage Ia1 squamous cell cervical carcinoma and had at least one follow-up visit after treatment. Women diagnosed with more advanced disease at the second conisation or reflex hysterectomy were included. Women that had been lost to follow-up immediately after treatment and women with more advanced disease were excluded. Women with positive margins were followed up only if they had CIN at the margins; cases with microinvasion or invasion at the margins were treated with hysterectomy. All women had primary treatment by local excision with needle, knife, or large loop excision of the transformation zone. The decision to proceed to reflex treatment with hysterectomy was taken after primary treatment with local excision.
Interventions
Women were managed with hysterectomy, repeat cone, or close surveillance on an individual basis. Women postconization were followed-up at 4-6 months intervals for the 2 years, 12 monthly for 3 years, and then annually for at least another 10 years with their general practitioner, originally in the colposcopy clinic and subsequently in primary care with cytology.
Data source
The full clinical details of all treated women have routinely been stored on a dedicated computer program. The records of all potentially eligible women were extracted to a new, specially designed database. This database included treatment information, colposcopy findings, cytology, and histology results from every follow-up appointment in the colposcopy clinic and in the community, the last having been obtained from the Exeter database. The Exeter database holds screening data for all women eligible for cervical screening and registered with a general practitioner. The Exeter record contains the dates of smears taken as well as codes for the results and follow-up recommendations. Information obtained from the Exeter database using OPEN EXETER, a web browser interface, was added to the follow-up records in our database. Data were not available for women who had moved out of the region.
Outcomes and statistical analysis
The study population was divided into 3 groups. Group A consisted of women who underwent a hysterectomy as part of their initial treatment. Group B consisted of women who had some form of local excision with CIN at the margins of excision who were followed-up closely. Some of these had more than 1 local excision. Group C were women who had clear margins after 1 or more local excisions.
We assessed differences in characteristics between the different groups and calculated the rates of residual disease in the hysterectomy specimens in group A. We computed the invasive and preinvasive recurrence rates for women in group A, B, and C and correlated these with the original treatment characteristics (clear vs involved margins, endo- vs ectocervical margins, low-grade vs high-grade lesions). We defined as recurrence the histological diagnosis of CIN 2 or worse (CIN2+) in the repeat treatment sample (excisional cone or hysterectomy).
Recurrence-free survival was defined as the time from diagnosis to recurrence or death from any cause, whichever occurred first. Patients who were recurrence-free and alive at the end of the follow-up period were censored. We plotted Kaplan-Meier curves of recurrence-free survival for each group (time to event outcome). We used the Wilcoxon, Scheffé, Fisher exact, Fisher-Freeman-Halton exact, and Peto-Prentice tests to assess differences between groups. Statsdirect (version 1.9.5, StatsDirect statistical software, 2013; StatsDirect Ltd., Altrincham Cheshire, UK) software was used for statistical analysis. Results with a 2-sided P < .05 were considered significant.
Results
We identified 114 women, 3 of whom were excluded because they had no follow-up. Of the remaining 111, 19 (17.1%) were in group A; 29 (26.1%) in group B; and 63 (56.8%) in group C. The details of the demographics and treatment characteristics are shown in Table 1 . The majority of the population was referred with high-grade squamous or glandular lesions (92.7%). Only 2 patients (1.8%) presented purely with symptoms, whereas 18 of 108 (16.6%) with abnormal cytology also had symptoms. All women in group B that were followed-up conservatively had CIN at the margins; there were no cases with microinvasion at the margin.
| Characteristic | n (%) |
|---|---|
| Mean age in years (range) | 36.5 (23–75) |
| Reason for referral | |
| Severe dyskaryosis or severe dyskaryosis ? invasion | 85 (76.5) |
| Moderate dyskaryosis | 16 (14.4) |
| Mild dyskaryosis | 5 (4.5) |
| High-grade glandular abnormality | 2 (1.8) |
| Unsatisfactory cytology | 1 (0.9) |
| Symptoms | 2 (1.8) |
| History of previous cone | |
| Yes | 10 (9.0) |
| No | 101 (91.0) |
| Treatment method | |
| LLETZ | 42 (37.8) |
| NETZ | 41 (36.9) |
| Laser excision | 19 (17.1) |
| CKC | 9 (8.1) |
| Anesthetic | |
| Local anesthesia | 44 (39.6) |
| General anesthesia | 67 (60.3) |
| Margins in initial cone | |
| Involved with CIN | 58 (52.2) |
| Involved with invasive disease | 0 (0) |
| Clear | 53 (47.8) |
| Uncertain | 0 (0) |
Two of the hysterectomies were performed after a second cone had failed to excise the lesion completely. Two were modified radical hysterectomies because of a concern that the lesion might be more than stage Ia1. The initial treatment consisted of 2 excisional cones in 2 of the women in group B and in 10 of the women in group C. The initial local excisions were performed using a cold knife conization in 9 women, laser in 19, large loop excision of the transformation zone in 42, and needle excision of the transformation zone in 41. There was no significant difference between the methods in respect to the margin status (data not shown). In all cases, the sample was removed as a single specimen. None of the cases receiving second conization or reflex hysterectomy were diagnosed with more advanced disease (more than stage Ia1). The median depth of the cone was 17.6 mm for group A (5 women), 14 mm in group B (18 women), and 16.8 mm in group C (20 women) (data available for 43 patients). The median number of follow-up smears was 8.2 for group A (vault smears), 10.9 for group B and 8.3 for group C.
Women in group A were older than those in group B (median age, 40 vs 35 years P = .0001; Scheffé) and those in group C (median age, 40 vs 34 years P = .0001; Scheffé). They were also more likely to have a deep margin involved than those in group B (17/19, 89.5% vs 15/29, 48.1%, P = .007, Fisher exact test). Although CIN 3 at the margins was more common in group A than in group B (17/19, 89.5% vs 19/29, 65.5%, P = .07, Fisher exact test), the difference was not statistically significant. Approximately half of both groups A and B had at least 1 child. Ten women had previous treatment to the cervix. Seven of these were in group A, none in group B, and 3 in group C ( P = .0002, Fisher-Freeman-Halton exact test) ( Table 2 ).
| Variable | Involved margins | Clear margins | P value | |
|---|---|---|---|---|
| Group A: hysterectomy (n = 19) n (%) a | Group B: close surveillance (n = 29) n (%) a | Group C: close surveillance (n = 63) n (%) a | ||
| Median age (IQR) | 45 y (40–57) | 35 y (30–40) | 34 y (30–38) | .0001 b |
| Previous cone biopsy | 7 | 0 | 3 | .0002 c |
| Location of involved margin | ||||
| Endocervical | 9 (47.4) | 7 (24.1) | N/A | .03 c |
| Ectocervical | 2 (10.5) | 14 (48.3) | N/A | |
| Both margins | 8 (42.1) | 8 (27.6) | N/A | |
| Lesion at margin | ||||
| CIN 3 | 17 (89.5) | 19 (65.5) | N/A | .08 c |
| CIN 2 | 0 (0) | 2 (6.9) | N/A | |
| CIN 1 | 0 (0) | 4 (13.8) | N/A | |
| Unknown | 2 (10.5) | 4 (13.8) | N/A | |
| Histology of reflex treatment | ||||
| Stage Ia1 SCC | 3 (15.8) | N/A | N/A | |
| CIN 2/3 | 4 (21.1) | N/A | N/A | |
| CIN 1 | 1 (5.3) | N/A | N/A | |
| No residual | 9 (47.4) | N/A | N/A | |
| Not known (treated elsewhere) | 2 (10.5) | N/A | N/A | |
| Abnormal cytology leading to further surgery | 0 (0) | 9 (31.0) | 7 (11.1) | .04 c |
| Histology of further surgery | ||||
| CIN2+ | 0 (0) | 3 (10.3) | 2 (3.2) | .17 d |
| Stage Ia1 SCC | 0 (0) | 0 (0) | 1 (1.6) | |
| CIN3 | 0 (0) | 1 (3.4) | 0 (0) | |
| CIN2 | 0 (0) | 2 (6.9) | 1 (1.6) | |
| CIN1 | 0 (0) | 1 (3.4) | 2 (3.2) | |
| Normal | 0 (0) | 5 (17.2) | 3 (4.8) | |
| Median interval initial to further surgery (IQR) | N/A | 68 wk (27–183) | 49 wk (27–79) | |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
