Keywords endometriosis - surgery time - #Enzian - AAGL2021 - rASRM
Schlüsselwörter Endometriose - Operationsdauer - #Enzian - AAGL2021 - rASRM
Introduction
Endometriosis is a chronic gynecological disease that affects approximately 10% of
women of reproductive age. It is defined as the presence of endometrial tissue outside
the uterus, which may manifest as peritoneal lesions, ovarian cysts (known as endometriomas),
or deep endometriosis [1 ]
[2 ]
[3 ].
Commonly used classification systems for endometriosis are the rASRM, the #Enzian,
and the AAGL2021 classification. The rASRM score is the most widely used classification
system, but hardly describes deep endometriosis [4 ]
[5 ]. The #Enzian classification describes superficial and deep endometriosis, as well
as ovarian endometriosis, adhesions and adenomyosis [6 ]. The AAGL2021 classification is a 4-stage system analogous to the rASRM classification
but has been developed to better reflect surgical complexity of endometriosis surgery
by using a more detailed mapping [7 ].
The treatment options for endometriosis vary depending on the symptomatology and results
of non-invasive diagnostics, with surgery being one of the primary treatment pillars,
along with hormonal treatment. Preoperative planning plays a critical role in optimizing
health resource management, particularly in the context of endometriosis surgery,
where operative times can vary significantly based on disease extent and complexity.
Optimal estimation of the planned surgery duration can achieve improved operating
room efficiency, more accurate scheduling, reduced patient waiting times and better
resource allocation [8 ]. To the best of our knowledge, no publication has evaluated the duration of endometriosis
surgery in relation to the stage of all three classification systems mentioned above
within the same patient cohort.
The aim of this study was to evaluate the correlation between the rASRM, #Enzian,
and AAGL2021 classifications and the duration of surgery. The main intention was to
ascertain which of these classifications most accurately reflects the duration of
surgery and should be employed to plan operating room resources.
Material and Methods
Participants in this retrospective data analysis were all patients who underwent laparoscopic
endometriosis surgery at the Department of Gynecology, Obstetrics and Gynecological
Endocrinology at the Kepler University Hospital in 2021 and 2022. Eligible patients
were identified using the digital surgery book of the hospital information system.
The following exclusion criteria were formulated: lack of histological evidence of
endometriosis, incomplete treatment of endometriosis, as well as additional surgical
indications besides endometriosis with an influence on the duration of surgery (e.g.,
simultaneous hysterectomy, ovarian cysts of a different entity). However, for some
frequently performed additional surgical procedures (e.g., hysteroscopy), a time factor
was added to enable inclusion in the study (see [Table 1 ] below). These specific correction time factors are based on evaluations of procedure
duration in the operating room and have been rounded up to five-minute thresholds.
The time factors were subtracted from the total operative time resulting in a corrected
operative time that was then used for further calculations.
Table 1
Time factors for additional intraoperative measures.
Additional intraoperative measures
Frequency [n]
Time factor [min]
min = minutes
Hysteroscopy + chromopertubation
10
10
Hysteroscopy + curettage + thermal balloon endometrial ablation
1
20
Conization
2
10
Prophylactic salpingectomy
4
5
Hysteroscopy
3
10
All surgeries were performed by gynecologic surgeons with substantial experience in
endometriosis treatment; when necessary, in multidisciplinary collaboration (e.g.,
colorectal surgeon). The extent of endometriosis was assessed intraoperatively using
the rASRM, #Enzian and AAGL2021 classifications. However, the AAGL2021 score was only
applied intraoperatively to patients from November 2021 onwards regarding to its publication
date. For patients operated before November 2021, the AAGL2021 score was assigned
retrospectively based on the surgical report and stored intraoperative images. This
partially retrospective staging of the AAGL2021 classification was performed or supervised
by the last author (S-H.E.).
The study was approved by the Institutional Ethics Committee of the Johannes Kepler
University (submission number 1188/2022, date of first approval 08/10/2021).
Statistical analysis
Descriptive statistics, including frequencies, means, medians, interquartile ranges
(IQR) and standard deviations, were calculated to summarize the data. Kendall’s tau
correlation was used to determine if there was a correlation between the operative
time and the rASRM/AAGL2021 classifications. Two tests were used to calculate the
correlation between the #Enzian compartments and operative time: Spearman’s correlation
and the Mann-Whitney U test (for the F compartments). Generalized linear models for
regression analysis were performed to explore whether there were significant associations
between the individual compartments of the #Enzian classification and the duration
of surgery.
The level of significance was set to p < 0.05, and no adjustment for multiple testing
was used. Data were analysed using SPSS Statistics 29.0.2.0.
Results
Origin and description of dataset
248 patients underwent laparoscopic surgery for endometriosis at the Department of
Gynecology, Obstetrics, and Endocrinology at the Kepler University Hospital (KUK)
between 2021 and 2022. After applying the exclusion criteria, 139 patients were included
in the study – 64 in 2021 and 75 in 2022. The flowchart of the inclusion process is
shown in [Fig. 1 ].
Fig. 1
Flowchart of inclusion process.
The median age in the dataset is 32 years (IQR = 28–37) with a median BMI of 22.86 kg/m2 (IQR = 20.31–25.71). Of the patients included, 50.4% had undergone at least one previous
abdominal surgery, and eight patients had undergone three or more abdominal surgeries.
The most prevalent subtype of endometriosis in the sample was peritoneal endometriosis
(85.6%), followed by ovarian endometriosis (49.6%), deep endometriosis (46.8%), and
adenomyosis (22.3%). Rectal involvement was observed in 10.1% of patients. According
to the AAGL2021 classification, 43.2% of patients were classified as stage 1, 30.9%
as stage 2, 5.0% as stage 3 and 20.9% as stage 4, while according to the rASRM classification,
37.5% of study participants were in rASRM stage 1, 12.5% in stage 2, 36.8% in stage
3 and 13.2% in stage 4. The distribution of affected compartments according to the
#Enzian classification is provided in the appendix (Table S1 online).
The median “actual” operative time – after subtracting time factors for additional
intraoperative measures – was 68.50 minutes (IQR = 46.75–97), with a minimum of 19
minutes and a maximum of 402 minutes. In two cases rectal shaving was required, eight
patients had an intestinal resection without stoma and one patient had an intestinal
resection with stoma. In one case, the surgical procedure was converted from laparoscopic
surgery to laparotomy.
AAGL2021 and rASRM in relation to surgery time
As illustrated in [Table 2 ], a comparison has been made of the operative times of the rASRM and AAGL2021 classifications
across different stages. The median operative time increased with stage in both the
rASRM classification and the AAGL2021 classification. Using Kendall’s tau correlation,
a significant positive correlation (p < 0.001) has been identified between the stages
and the duration of surgery in both classifications.
Table 2
Operative times for the different stages of the rASRM and AAGL2021 classifications.
rASRM classification
AAGL2021 classification
Stage
n
min
max
Median (IQR)
Kendall’s tau correlation
n
min
max
Median (IQR)
Kendall’s tau correlation
Operative times in minutes
IQR = interquartile range; max = maximum; min = minimum
I
51
19
156
47 (34–63)
ρ = 0.465
p < 0.001
60
19
156
47 (35.25–63.75)
ρ = 0.572
p < 0.001
II
17
34
287
68 (58.5–78)
43
25
189
72 (54–90)
III
49
25
351
79 (55–124.5)
7
62
290
79 (64–145)
IV
17
45
402
110 (97–232.5)
28
69
402
124 (97.75–221.25)
#Enzian classification in relation to surgery time
Due to the more complex structure and the lack of stages, the evaluation of the #Enzian
classification is more complex. [Table 3 ] shows the descriptive statistics for operative time in relation to the compartments
of the #Enzian classification. The longest median operative time can be found for
surgeries coded C3. Also, A3 and the involvement of compartments FI and FU showed
long median operative times. Spearman’s correlation for individual compartments of
the #Enzian classification shows significant positive correlations (p < 0.001) between
surgery duration and all compartments. A regression analysis was carried out to identify
associations between the #Enzian compartments and the operative time as a whole (Table S2 online). Several parameters are statistically significant, including O2right, C3,
C1, FA, and FU, indicating these factors have a significant impact on the operative
time in the dataset.
Table 3
Operative times for the different compartments of the #Enzian classification.
#ENZIAN classification
0
pos
1
2
3
Correlation/
Data as Median [IQR] in minutes, IQR = interquartile range
“pos” refers to all positive parts of the compartment (1–3 of the corresponding compartment)
Correlation: Spearman Correlation for P/O/T/A/B/C compartments and Mann-Whitney U
test for F compartments
* significant in the regression analysis (details in Table S2 )
** Upper quartile could not be calculated due to small sample size.
Comp P
71.5 [49.75–101.5]
68 [46–96.25]
51 [39–77.25]
75 [63.35–124.5]
122.5 [85.5–168]
ρ = 0.283
Comp O left
58.5 [42.5–78.75]
88 [63.75–123.75]
96 [66–116.5]
80 [47–127]
89.5 [67–125.25]
ρ = 0.312
Comp O right
58 [41–83]
81 [63–122.5]
75 [50.5–119.5]
94.5 [70.75–125.75]*
86 [67.25–117.25]
ρ = 0.355
p < 0.001
Comp T left
55 [41.75–74.25]
97 [77.25–134.5]
92 [70.5–116]
128.5 [58.5–169.5]
103 [84–255]
ρ = 0.482
p < 0.001
Comp T right
59.5 [44–86.25]
104 [84.5–232.5]
104 [77–209]
90 [90–90]
104 [84.5–232.5]
ρ = 0.410
p < 0.001
Comp A
60 [41–80]
103.5 [67–143.25]
68.5 [52.75–126.25]
117 [85.5–130]
287 [103–351]
ρ = 0.434
p < 0.001
Comp B left
58.5 [39–78]
86.5 [58–131.5]
67.5 [49.25–83]
103 [78–255]
156 [129–225]
ρ = 0.439
p < 0.001
Comp B right
59.5 [41–86.25]
97 [64–139]
80 [58–124]
107 [67.75–213.75]
131 [123–NA]**
ρ = 0.351
p < 0.001
Comp C
62.5 [44.25–86.75]
162 [122–322.75]
144 [86.25–178.5]*
116 [104–NA]**
324.5 [203.5–363.75]*
ρ = 0.444
p < 0.001
Comp FA
59.5 [43.5–89]
97 [68–225]*
NA
NA
NA
p < 0.001
Comp FB
68.5 [46.25–96.75]
160 [65–NA]**
NA
NA
NA
p = 0.328
Comp FI
67.5 [46–95]
222 [124.5–365.25]
NA
NA
NA
p < 0,001
Comp FU
68 [46–96]
255 [79–NA]*, **
NA
NA
NA
p = 0.028
Discussion
This study represents a pioneering effort to compare operative times for endometriosis
surgery according to codings by the #Enzian, rASRM and AAGL2021 classifications. The
findings indicate a significant correlation between extended endometriosis and the
length of surgical procedures in all three classifications, whether by stages in the
rASRM and AAGL2021 classifications or by compartments in the #Enzian classification.
For both four-stage classification systems (rASRM and AAGL2021), the median operative
time extended with increasing stages. In the existing literature, there is a lack
of explicit studies investigating operative times according to rASRM stage. An indirect
correlation was observed in a study by Nicolaus et al. reporting that higher rASRM
stages are associated with more complications, leading to longer operative times [9 ]. However, the rASRM classification does not adequately capture deep endometriosis,
which may be a contributing factor why this system alone does not fully reflect the
actual surgical effort for more complex cases [10 ]. Furthermore, the applicability of the rASRM classification in a preoperative non-invasive
setting is limited, thereby restricting its use for preoperative planning of operating
room resources.
Similar to rASRM, there is a shortage of studies reporting surgical duration according
to the AAGL2021 stage. A comprehensive search of the PubMed database failed to identify
any publications that addressed operating times in relation to the AAGL2021 classification.
Even in the original publication on the AAGL2021 classification, no operating times
are presented in the surgical validation cohort [7 ]. However, a study by Abrao et al. on the applicability of ultrasound concluded that
patients can be reliably identified preoperatively with the correct AAGL2021 stage
[11 ]. This statement, in conjunction with the presents study’s results, underscores the
potential for incorporating AAGL2021 staging into preoperative planning tools.
The #Enzian classification, with its detailed coding of endometriosis by affected
compartments and size, revealed a more complex relationship with the duration of surgery.
Specifically, the compartments C, FI, and FU were associated with the longest median
operative times, highlighting the significant impact of deep endometriosis on surgical
complexity and duration, particularly in cases of bowel involvement. This finding
is in line with the study of Haas et al., in which a multiple regression analysis
for predicting operating time based on the Enzian classification showed an extension
of the operation time when C3 is involved [12 ]. Another study by Thomassin-Naggara et al., based on preoperative Enzian codings
by MRI, demonstrated that the operating time for Enzian coding C3 was significantly
longer than for C2 [13 ]. In both studies an earlier version of the Enzian classification was used, but its
basic structure has been retained in the current #Enzian classification. The regression
analysis of the #Enzian compartments in our study further identified compartment O
(ovarian endometriomas) and FA (adenomyosis) as relevant influencers on surgery duration.
A study of Banerjee et al. described that the presence of endometriomas is a good
indicator for the presence of rectosigmoid disease and involvement of the posterior
cul-de-sac [14 ]. Other studies have demonstrated a strong association between adenomyosis and deep
endometriosis [15 ]
[16 ]. Whether adenomyosis plays an independent role or is merely a surrogate marker of
accompanying deep endometriosis cannot be determined by the study’s data.
The #Enzian classification can be applied either intraoperatively or by using imaging
modalities such as ultrasound or MRI. By precisely mapping endometriosis across anatomical
compartments and lesion sizes, the #Enzian demonstrates strong potential as a predictive
tool for surgical complexity and duration, particularly through its associations with
deep infiltrating compartments like C, FI, and FU, as well as ovarian involvement
and adenomyosis. In contrast to rASRM, the AAGL2021 and #Enzian classifications offer
preoperative utility, with #Enzian particularly suited for algorithmic tools due to
its structured and compartmental format. However, high-quality preoperative mapping
and use of detailed classification systems requires in-depth anatomical knowledge
and targeted training to ensure consistent and correct use. Systems like #Enzian or
AAGL2021 should be regarded as a crucial quality standard to promote continuous improvement
in preoperative planning.
Recent studies have reported alternative scores for predicting the length of surgery.
For instance, the preoperative applicability of the NMS-E (Numerical Multi-Scoring
System of Endometriosis) was described in a recent paper [17 ]. This score demonstrated a robust correlation with the rASRM score and surgical
outcomes, and was useful in predicting surgical duration. The Deep Pelvic Endometriosis
Index magnetic resonance imaging score (dPEI score) is a newly published preoperative
score for women with deep endometriosis based on MRI images. This score is helpful
in predicting operating time, length of hospitalisation and postoperative complications
[18 ]. Ebrahimi et al. evaluated a preoperative five-domain ultrasound-based endometriosis
staging system (UBESS), which was compared postoperatively with the RANZCOG/AGES system.
The results showed that preoperative ultrasound using the UBESS staging system was
able to predict the complexity of laparoscopic surgery [19 ]. The list demonstrates that new endometriosis classifications are springing up,
yet a universally accepted classification remains elusive. Fadinger et al. described
in their study a first attempt to convert the #Enzian classification into a 4-stage
scoring system comparable to the AAGL classification, demonstrating an overlapping
of both classification systems. In their opinion, there should be a universal classification
in the future that incorporates the strengths of existing classifications [20 ].
The strength of the study lies in its direct comparison of three distinct endometriosis
classifications within a single data set of a diverse endometriosis population. Consistent
surgical procedures by physicians at an endometriosis center certified by EuroEndoCert
ensure high standards of care and expertise [21 ]. However, limitations of the study must be acknowledged including its retrospective
study design and single-center data collection. Experience in endometriosis surgery
and individual surgical technique plays an important role in the duration of surgery,
making it challenging to extrapolate the time data to other centers. Additionally,
a larger study population would facilitate a more nuanced analysis of the complex
#Enzian coding system, maybe enabling a preoperative planning tool. While certain
compartments in the #Enzian classification (e.g., C3, FI, FU) showed statistically
significant associations with prolonged operative time in the regression analysis,
these findings must be interpreted with caution due to the small sample sizes in those
subgroups, increasing the risk of overfitting or type I error. However, it is important
to emphasize that the statistically significant findings in our study align well with
clinical experience and plausibility. The time factors employed for minor additional
intraoperative interventions may introduce inaccuracies in the actual operative time,
though this is presumed to have no substantial impact on the study’s findings.
Conclusion
All three classification systems used in our study (#Enzian, AAGL2021 and rASRM) show
a considerable impact on operative time, with an increase in duration corresponding
to higher stage or specific coding. This underscores the importance of non-invasive
classification systems as predictors of surgical complexity and duration. In contrast
to the rASRM classification, the AAGL2021 and #Enzian classifications can be used
as preoperative diagnostic tools, with the #Enzian offering a more diverse coding.
Future research should focus on prospective studies with larger, multi-center datasets
to develop algorithms incorporating codings from the #Enzian classification for enhanced
preoperative planning. Such advancements could lead to more accurate predictions of
surgery duration and better resource allocation.
Appendix – Supporting Information
Appendix – Supporting Information
