Key words
LLETZ - specimen - conization - artifacts - colposcopic experience
Schlüsselwörter
LLETZ - Probe - Konisation - Artefakte - kolposkopische Erfahrung
Introduction
Large loop excision of the transformation zone (LLETZ) is a commonly used technique
for managing women with pathological changes on the uterine cervix, which has gained
worldwide acceptance due to its numerous advantages [1], [2], [3], [4]. Unlike ablative methods, LLETZ allows histological examination of the excised tissue
[1], [2], [3], [4].
However, LLETZ causes thermal damage of specimens [5], due to the use of electricity, which simultaneously cuts and coagulates the tissue
[6]. Thermal artifacts interfere with the pathological analysis [7]. Furthermore, there is the problem of multiple excisions, which are necessary to
manage large ectocervical lesions. This causes specimen fragmentation that hampers
assessment of the margins [8]. All those factors may have a negative impact on the further management of women
in reproductive age, who need an accurate and minimally invasive treatment for their
future fertility.
LLETZ effectiveness and cure rates, adverse effects, complications, both short-term
and long-term clinical outcomes and its influence on fertility have been widely investigated
[1]. However, the risk factors for obtaining inadequate specimen and subsequent difficulties
in histological interpretation, particularly in terms of surgeonsʼ skills and experience
in colposcopy, are yet to be identified [3], [9].
The aim of this study was to perform the quality assessment of LLETZ procedure in
women of reproductive age, by investigating the colposcopic experience of the surgeons,
and its influence on the presence, diagnostic significance and type of artifacts and
their influence on histological interpretation and specimen adequacy.
Materials and Methods
Study design
A retrospective cohort study was performed in a university-affiliated hospital over
a four-year period, between January 2010 and December 2013. The Institutional Review
Board approved the study.
Our colposcopy service provides consultations for patients that are referred for abnormal
Pap smears and/or colposcopies. Women suspected of having cervical dysplasia or HPV
lesions following a Pap test and a colposcopy are either referred to a biopsy or a
LLETZ procedure, based on the judgment of the attending colposcopist and the patientʼs
wishes. In addition, our colposcopy clinic provides consultative service for patients
included in the IVF procedures covered by the national health care system, as well
as for the patients from national transplantation clinics. In such patients, a follow-up
is not acceptable and immediate LLETZ treatment of any abnormalities is requested.
LLETZ indications further included discrepancies between cytology, colposcopy and
biopsy results, persistent (> 2 years) low-grade abnormalities, according to the national
protocol and patientsʼ request for maximum oncological safety. Some of the patients
were treated with LLETZ procedure upon their request, as some women were not willing
to comply with the other treatment modalities or/and follow-up visits. In our culture,
patients prefer LLETZ conization to other treatments, as they traditionally consider
this approach to be oncologically safer. This is probably due to the absence of well-organized
screening programs and a relatively high mortality of invasive cervical cancer in
our country. Thus, some of the patients in our study had been submitted directly to
LLETZ instead of to diagnostic biopsy. In such patients, the indications for the LLETZ
treatment included abnormal colposcopy and/or cytology.
LLETZ excision procedure
All patients gave informed consent for the LLETZ. All the procedures were performed
under general anesthesia using Valley Lab Force Triad Electrosurgical Unit (Covidien
Ltd, Dublin, Ireland). Loop electrodes were chosen by surgeonsʼ preference, either
20 × 15 mm or 10 × 10 mm curvilinear loop and ball 3 mm or 5 mm in diameter. Unit
was calibrated from 35 watt to 45 watt.
Selection of patients
The exclusion criteria were pregnancy, age over 45, menopausal status, invasive or
micro invasive carcinoma of the cervix, adenocarcinoma in situ, and re-treatment for
positive margins after previous surgery, and missing data relevant for the study.
The following data from the pathological records were taken: age, parity, and previous
cervix treatment, indication for surgery, surgeonʼs level of colposcopic experience,
definitive histological diagnosis and margin involvement, specimen fragmentation,
presence and type of artifacts and their interference with the histological interpretation.
The LLETZ indications were divided into three categories: abnormal Pap smear, colposcopy,
or histology (in cases of prior biopsy). According to biopsy results, histological
diagnoses were divided into subcategories: condyloma, CIN 1, CIN 2 and CIN 3. Surgeons
were grouped according to their experience in colposcopy into three categories: expert,
experienced and junior colposcopists. The descriptors of the groups by experience
are presented in [Table 1]. A total of 266 patients were divided into three groups: group A – 75 patients operated
by juniors; group B – 74 patients operated by experienced, and group C – 117 patients
operated by expert colposcopists.
Table 1 The descriptors of the groups by experience.
Category
|
Definition
|
Expert colposcopist
|
Gynecologists certified in colposcopy, with more than 20 years in experience, performing
more than 30 colposcopies per week and performing colposcopy consultations for other
groups, also involved in both lectures and practical education at the National School
of Colposcopy
|
Experienced colposcopist
|
Gynecologists certified in colposcopy, with more than 10 and less than 20 years of
experience, performing 15 to 30 colposcopies per week, who are involved in practical
education at the National School of Colposcopy
|
Junior colposcopist
|
Gynecologists certified in colposcopy, with less than 10 years of experience, performing
up to 15 colposcopies per week, who are not involved in any kind of education at the
National School of Colposcopy
|
Histopathological procedure
The specimens were collected in plastic boxes, separately fixed, labeled and microscopically
evaluated. Pathologists proficient in gynecological pathology, who provide pathological
reviewing for national referral service, assessed histological diagnosis and margin
involvement. They were not blinded to the surgery indications. Margin status was reported
for all the pieces in cases of specimen fragmentation. The artifacts determined by
the pathologist were grouped according to their influence on histological diagnosis
as absent, present but not interfering with the histological diagnosis and present
as well as interfering with the histological diagnosis. The types of artifacts assessed
were categorized as thermal damage, fragmentation and multiple artifacts. Mechanical
damage of the specimen was not described as an isolated artifact in any of the cases.
Statistical analysis
The data were analyzed using SPSS 17.0 (Chicago, USA), with a significance level set
at 0.05. For continuous variables, the differences between groups were compared by
1-way analysis of variance (ANOVA). ANOVA was performed with a Brown-Forsythe adjustment
for heteroscedasticity and with post-hoc Tukey or Dunnettʼs T3 procedure for multiple
comparisons. The Pearson χ2 and likelihood ratio χ2 as appropriate were used for proportions.
Results
Characteristics of investigated patients
There were no differences between the patients in the defined groups analyzing the
age, age at menarche, number of previous abortions and deliveries, previous cervix
treatment and LLETZ indications. The majority of the patients, 177 (66.54 %) were
nulliparous. Patientsʼ demographic and clinical data are presented in [Table 2].
Table 2 Patientsʼ demographic and clinical data.
Data
|
Total (n = 266)
|
Group A (n = 75)
|
Group B (n = 74)
|
Group C (n = 117)
|
p
|
a ANOVA was performed with a Brown-Forsythe adjustment for heteroscedasticity and with
post-hoc Tukey or Dunnettʼs T3 procedure for multiple comparisons of unequal variances;
b Between-groups comparison with equal variance was performed by 1-way ANOVA; c Likelihood ratio χ2; d Pearsonʼs χ2 test.
|
Patientsʼ characteristic (mean ± SD)
|
|
|
|
|
|
Age (years)
|
31.58 ± 5.66
|
31.81 ± 5.79
|
31.65 ± 6.32
|
31.38 ± 5.16
|
0.871 a
|
Menarche (years)
|
13.09 ± 1.39
|
13.19 ± 1.34
|
13.02 ± 1.35
|
13.07 ± 1.45
|
0.796b
|
Number of abortions
|
0.31 ± 0.66
|
0.29 ± 0.59
|
0.46 ± 0.80
|
0.24 ± 0.60
|
0.081 a
|
Number of deliveries
|
0.46 ± 0.81
|
0.50 ± 0.84
|
0.46 ± 0.91
|
0.44 ± 0.73
|
0.893b
|
Clinical data n (%)
|
|
|
|
|
|
Previous treatment of the cervix
|
|
|
|
|
|
|
247 (92.86)
|
72 (96.00)
|
70 (94.59)
|
105 (89.74)
|
0.451c
|
|
3 (1.13)
|
0 (−)
|
1 (1.35)
|
2 (1.71)
|
|
|
16 (6.02)
|
3 (4.00)
|
3 (4.05)
|
10 (8.55)
|
|
Indication for LLETZ
|
|
|
|
|
|
|
135 (50.75)
|
41 (54.67)
|
38 (51.35)
|
56 (47.86)
|
0.287 d
|
|
46 (17.29)
|
8 (10.67)
|
17 (22.97)
|
21 (17.95)
|
|
|
85 (31.95)
|
26 (34.67)
|
19 (25.68)
|
40 (34.19)
|
|
Characteristics of assessed specimens
There was a significant difference among the groups with respect to specimen adequacy
for pathological assessment. Overall, 238 specimens were adequate for pathological
assessment, most frequently in group C (94.87 %). Out of 28 inadequate specimens,
53.57 % were in group A. Groups exhibited differences in terms of both the presence
and the diagnostic significance of the artifacts. Artifacts were most commonly present
in group A, and absent in group C. Diagnostically significant artifacts, interfering
with the histological diagnosis, were present in 28 (10.53 %) of the cases, with significant
differences among the groups. Artifacts precluding histological diagnosis were most
common in group A (53.57 % of the cases). There was no difference among the groups
regarding the type of artifacts. The most common artifacts were thermal (11.28 %),
followed by multiple artifacts (9.40 %), while fragmentation (4.14 %) was the least
present. Multiple artifacts were most common in group A. Of the 266 specimens, 176
(66.17 %) were obtained in one piece, without differences between the groups (p = 0.160).
However, it is noteworthy that the number of fragments was the highest in group A
(data not shown). The data about the quality and adequacy of the specimen are listed
in [Table 3].
Table 3 Specimen adequacy and quality.
Parameter
|
Total n = 266 (%)
|
Group A n = 75 (%)
|
Group B n = 74 (%)
|
Group C n = 117 (%)
|
p
|
a Pearsonʼs χ2 test; b Likelihood ratio χ2.
|
Specimen adequacy
|
|
|
|
|
|
Adequate
|
238 (89.47)
|
60 (80.00)
|
67 (90.54)
|
111 (94.87)
|
0.004a
|
Inadequate
|
28 (10.53)
|
15 (20.00)
|
7 (9.46)
|
6 (5.13)
|
|
Presence of artifacts
|
|
|
|
|
|
Absent
|
200 (75.19)
|
52 (69.33)
|
51 (68.92)
|
97 (82.91)
|
0.036a
|
Present
|
66 (24.81)
|
23 (30.67)
|
23 (31.08)
|
20 (17.09)
|
|
Diagnostic significance of artifacts
|
|
|
|
|
|
Absent
|
200 (75.19)
|
52 (69.33)
|
51 (68.92)
|
97 (82.91)
|
0.004a
|
Present and diagnostically not significant
|
38 (14.29)
|
8 (10.67)
|
16 (21.62)
|
14 (11.97)
|
|
Present and diagnostically significant
|
28 (10.53)
|
15 (20.00)
|
7 (9.46)
|
6 (5.13)
|
|
Type of artifacts
|
|
|
|
|
|
Absent
|
200 (75.19)
|
53 (70.67)
|
51 (68.92)
|
96 (82.05)
|
0.080b
|
Thermal damage
|
30 (11.28)
|
6 (8.00)
|
13 (17.57)
|
11 (9.40)
|
|
Fragmentation
|
11 (4.14)
|
5 (6.67)
|
4 (5.41)
|
2 (1.71)
|
|
Multiple artifacts
|
25 (9.40)
|
11 (14.67)
|
6 (8.11)
|
8 (6.84)
|
|
Specimen fragmentation
|
|
|
|
|
|
One piece
|
176 (66.17)
|
43 (57.33)
|
51 (68.92)
|
82 (70.09)
|
0.160a
|
Two and more pieces
|
90 (33.83)
|
32 (42.67)
|
23 (31.08)
|
35 (29.91)
|
|
Histology and margin status of specimens
There was no difference among the groups regarding definitive histological diagnosis,
although it was close to a statistical significance (p = 0.088). In terms of margin
status, there was a significant difference among the groups. Endocervical and ectocervical
margins were most frequently inconclusive in group A. Out of 27 specimens with both
margins inconclusive, 55.56 % were in the group A. The incidence of positive ectocervical
margin throughout the study was 8.27 %, with the lowest frequency in group C (7 of
117, i.e. 5.98 %). Out of 184 specimens with both margins negative, 45.11 % were in
group C, 30.43 % were in group B and 25.46 % were in group A. Definitive histological
diagnoses and margin status of the specimens are listed in [Table 4].
Table 4 Histological diagnoses and margin status of the specimen.
Parameter
|
Total n = 266 (%)
|
Group A n = 75 (%)
|
Group B n = 74 (%)
|
Group C n = 117 (%)
|
p
|
a Pearsonʼs χ2 test; b Likelihood ratio χ2.
|
Histological diagnosis
|
|
|
|
|
|
Normal
|
80 (30.08)
|
19 (25.33)
|
25 (33.78)
|
36 (30.77)
|
0.088a
|
Condyloma
|
34 (12.78)
|
12 (16.00)
|
14 (18.92)
|
8 (6.84)
|
|
CIN 1
|
30 (11.28)
|
8 (10.67)
|
10 (13.51)
|
12 (10.26)
|
|
CIN 2
|
27 (10.15)
|
6 (8.00)
|
9 (12.16)
|
12 (10.26)
|
|
CIN 3
|
95 (35.71)
|
30 (40.00)
|
16 (21.62)
|
49 (41.88)
|
|
Margin status
|
|
|
|
|
|
Ectocervical margin
|
|
|
|
|
|
|
217 (81.58)
|
53 (70.67)
|
59 (79.73)
|
105 (89.74)
|
0.005a
|
|
22 (8.27)
|
7 (9.33)
|
8 (10.81)
|
7 (5.98)
|
|
|
27 (10.15)
|
15 (20.00)
|
7 (9.46)
|
5 (4.27)
|
|
Endocervical margin
|
|
|
|
|
|
|
188 (70.68)
|
47 (62.67)
|
57 (77.03)
|
84 (71.79)
|
0.011a
|
|
50 (18.80)
|
13 (17.33)
|
10 (13.51)
|
27 (23.08)
|
|
|
28 (10.53)
|
15 (20.00)
|
7 (9.46)
|
6 (5.13)
|
|
Margin condition (both margins)
|
|
|
|
|
|
|
184 (69.17)
|
45 (60.00)
|
56 (75.68)
|
83 (70.94)
|
0.007b
|
|
18 (6.77)
|
5 (6.67)
|
7 (9.46)
|
6 (5.13)
|
|
|
27 (10.15)
|
15 (20.00)
|
7 (9.46)
|
5 (4.27)
|
|
|
36 (13.53)
|
10 (13.33)
|
4 (5.41)
|
22 (18.80)
|
|
|
1 (0.38)
|
0 (−)
|
0 (−)
|
1 (0.85)
|
|
Discussion
Adequate management of cervical dysplasia is an important step in cervical cancer
prevention [10]. The cure rates following LLETZ have been broadly investigated with respect to patientsʼ
age, parity, size and grade of lesion, cone size and volume, as well as the use of
various LLETZ technique modifications [2], [4], [11], [12], [13]. However, data on surgeonsʼ colposcopic experience and its influence on specimen
quality, the presence and diagnostic significance of the artifacts have been inadequately
studied. The difficulties in margin assessment due to the presence of artifacts are
also described, particularly from the pathologistsʼ point of view [8]. Surgical textbooks and publications often underline the significant influence of
surgical experience on surgery results [6], [14]. Nevertheless, surgical experience is not the only determining factor of LLETZ specimen
quality. The issue of sufficient colposcopic experience to verify oneʼs certification
for the LLETZ has not been evaluated enough. LLETZ is performed by surgeons experienced
in both oncology and colposcopy, as well as by general practitioners and nurses, sometimes
with very limited experience in colposcopy. This is due to the belief that LLETZ has
an easier learning curve in comparison to the other conization techniques [2].
Inadequate specimens increase the treatment failure rate; they impair the patientʼs
psychological well-being and cause additional costs. Besides, publications about possible
adverse perinatal outcomes after LLETZ pose a serious question regarding its application
in women of reproductive age by surgeons inexperienced in colposcopy [15], [16]. Thus, inadequate treatment of cervical dysplasia in women of the reproductive age
influences both the risk of future cervical cancer development and perinatal complications
[2].
Adequate histological interpretation of the LLETZ specimen requires complete assessment
of the margins. Literature data suggest that several patient and surgeon characteristics
may influence sample adequacy [3]. Patient related factors include habitus, relaxation and vaginal compliance, age,
parity, indication for the surgery and morphology and grade of the disease. Surgeon
related factors include experience and colposcopic impression lesion morphology, loop
size shape and speed of passage through the tissue, current settings and colposcopic
interposition [7], [17].
Our study groups did not differ in relation to demographic and clinical data. We documented
a lower incidence of artifacts with the higher the level of colposcopic expertise.
Furthermore, colposcopy experience significantly influenced margin status, as well
as the presence and diagnostic significance of the artifacts. Both inadequate specimens
and artifacts were the most frequent in group A. Artifacts precluding margin assessment
were the most frequent in group A. The type of artifacts was not influenced by colposcopy
experience, although multiple artifacts were most frequent in group A. LLETZ procedure
is associated with thermal artifacts at all the margins, which could limit the adequacy
and quality of the specimen and preclude margin assessment [7], [8], [13]. The published incidence rates are up to 48 % [18]. Wright et al. [19] found the range of thickness of the coagulated zone in LLETZ cones to be 0.150 to
0.830 mm, with a mean of 0.396 mm. Baggish et al. [20] found the depth of coagulation at the ectocervical margin to be 0.187 mm and at
the endocervical margin 0.295 mm. Nevertheless, the presence of thermal injury influences
the interpretation of the margins only in specimens with no healthy tissue between
the dysplastic and coagulated area. The frequency of thermal artifacts in our study
corresponds to the previously published results, and its incidence is confirmed to
be surgeon dependent [21], [22]. Messing et al. [8] evaluated the specimens obtained by residents and found that 26.09 % of them had
severe thermal artifacts precluding the evaluation for both the grade of dysplasia
and the margins. Thermal artifacts are determined by the loop size, speed of cutting,
electrical energy and tissue conductivity [3], [8], [17]. In cases of repeated excisions, artifacts are present on all sides of the specimens,
except on the ectocervical surface [8]. Incidence of thermal artifacts and depth of the coagulation zone could be reduced
with appropriate loop size, adequate power settings and improvement of surgeonsʼ colposcopic
skills. All the above-mentioned factors, except tissue conductivity are surgeon dependent
and influenced by experience [8].
Specimen fragmentation mainly results from poor surgical technique and it increases
diagnostic uncertainty [8]. In cases of large ectocervical lesions, multiple excisions are inevitable as loop
size limits the size of the cone, which is not true for cold knife or laser conization.
Based on the recommendations of the National Health Service Cervical Screening Programme
from the United Kingdom, at least 80 % of the LLETZ procedures should result in single
piece specimen, which is a marker of good practice [23]. Still, the rate of single fragment LLETZ cones in literature ranges from 49 to
92 % [2], [6]. The trainee status in colposcopy was previously established to be a substantial
risk factor for specimen fragmentation [3]. Cure rates are also lower in cases of lesions treated with multiple LLETZ excisions.
Papoutsis et al. [2], [24] found that increased incidence of multiple excisions, even when performed by experienced
colposcopists, did not result in a higher rate of negative margins. Bharathan et al.
[3] documented an influence of specimen fragmentation on both the margin involvement
and inconclusive margins, while trainee status was related to specimen fragmentation,
which was present in 25 % of patients. Despite failing to prove that colposcopy experience
influenced fragmentation, we found that the rate of fragmented specimens was inversely
related to colposcopic experience.
The incidence of positive ectocervical margins may be reduced by performing the excision
under colposcopic guidance. However, this cannot reduce the incidence of positive
ectocervical margins, if a surgeon is not properly trained in colposcopy. An incidence
of positive margins of 39.96 % was reported in a study where the majority of LLETZ
procedures were performed by residents [8]. We documented the lowest incidence of negative margins in group A. Miroshnichenko
et al. [25] found that the incidence of positive margins was 38 % in LLETZ procedures performed
by experienced gynecologic oncologists. Out of 96 specimens, 85 % had entirely interpretable
margins. Urlich et al. [9] established that certification in colposcopy influences the rate of positive margins
in a study including 334 CKCs and 25 LLETZ. None of the specimens obtained by gynecologists
certified in colposcopy had uncertain margins, so they concluded that certification
and experience in colposcopy improves the rate of free margins, which corresponds
to our results. Apart from colposcopic experience, the lowest rate of inconclusive
endocervical margins in group C can be explained by the lowest rate of fragmented
samples in this group. Although we failed to prove a difference in the incidence of
specimen fragmentation among the groups, this does not necessarily mean that fragmentation
has no clinical significance on the occurrence of inconclusive margins. Montz et al.
[7] documented a lesser depth of coagulation artifacts at the endocervix in cases of
single specimens as compared to fragmented ones. Boardman et al. [26] found an incidence of inconclusive margins of 27 % in women operated by physicians
in training, which is similar with the incidence of inconclusive margins in group
A.
The status of margins determines both the postoperative follow-up and management.
In cases of inconclusive margins, the lesion may be both entirely excised and not
excised. Such patients require further close follow-up, diagnostics or even repeated
treatment, which increases healthcare costs, and physiciansʼ anxiety, and disrupts
patientsʼ quality of life. In women of reproductive age, this outcome could postpone
pregnancy. Thus, for women desiring pregnancy, we suggest that expert colposcopists
should perform LLETZ, because LLETZ could also lead to unfavorable perinatal outcomes
[15], [16].
This study has several drawbacks. The analyzed data were generated from pathological
reports, which may limit their quality. Nevertheless, a retrospective study design
allowed us to study the LLETZ practice without bias, by not allowing surgeons to adjust
their technique and patient selection according to the nature of investigation. This
eliminates the Hawthorne effect, and, therefore, we believe that the obtained results
could be clinically useful to those who treat women of reproductive age with cervical
dysplasia.
The high rate of cone specimens free of disease is another limitation of our study,
which is influenced by the indications for the procedure. Nevertheless in our setting,
regardless of the histological specimen diagnosis, pathologists in their reports describe
as part of routine practice the quality of the specimen, margin status and presence
of artifacts, as well as the number of fragments. This is customary both in cases
with cervical dysplasia and where it is absent. As the aim of the study was to evaluate
the quality of specimens, in terms of the presence of artifacts that could possibly
interfere with pathological assessment, the authors believe that the surgical technique
was not influenced by the specimen histology, thus enabling us to compare the performance
of various surgeons in all the cases included.
Our investigation is lacking data on managing patients with positive and inconclusive
margins, as well as follow-up data. The results are not applicable to all women, given
that we have only studied women of reproductive age treated with LLETZ. Although our
results represent the experience of one teaching hospital, we must underline that
they also represent the practice of 48 surgeons with different levels of colposcopic
experience. In the hands of surgeons with less experience in colposcopy, as we demonstrated,
LLETZ is compromised by the following disadvantages: a high frequency of artifacts
and inconclusive margins.
Conclusions
We believe that our study will provide valuable information for the education of the
medical personnel managing cervical dysplasia by LLETZ. Although LLETZ is considered
a minor surgery, this study indicates that skills in colposcopy are an essential prerequisite
for optimal results.