The lymphatic system of the uterine cervix: towards a more precise definition of the drainage routes

Lymph node involvement in cancer of the uterine cervix is a major independent prognostic factor for overall survival. The aim of our study was to examine the lymphatic drainage regions of the different parts of the uterine cervix. An anatomical study of fresh cadavers was conducted by injecting patent blue in the anterior or posterior lip of the uterine cervix and dissecting drainage regions. Furthermore, a retrospective radiological and pathological studies were conducted on patients who were treated for early-stage cancer of the uterine cervix with lymph node involvement. Radiological analysis of pre-therapeutic MRIs and/or pathological analysis of surgical specimens showed the precise location of the tumour. A correlation between the anatomy of the primary lesion site and the lymphatic area involved was established. We administered 15 cadaveric injections: 8 in the anterior lip of the uterine cervix and 7 in the posterior one. For 100% of the anterior lip injections, lymphatic drainage was bilateral ileo-obturator (n = 8/8) combined with bilateral parametrial drainage. For the posterior injections, there was never any ileo-obturator drainage, and 6 of the 7 (75%) posterior injections drained in the posterior uterine serosa. Concerning the clinical study, we included 21 patients. We observed a non-significant tendency towards bilateral lymph node involvement when the tumour of the anterior lip. Physiological lymphatic drainage of the uterine cervix presents anatomical specificities, depending on the structure studied, anterior or posterior lip of the cervix. Better knowledge of these specificities should allow personalized surgery for each patient.


Introduction
Lymph node (LN) involvement still remains the most important prognostic factor for women with uterine cervical cancer [16,26]. However, because early-stage disease has a low risk of LN metastasis, over the past few decades, sentinel lymph node (SLN) mapping has received increased attention. This is a technique that could reduce the morbidity associated with ineffective lymphadenectomy [15]. Development of the SLN technique is relatively recent considering that use of the SLN was first described in 1977 by Dr Cabanas who performed lymphangiography to determine the node(s) to be removed in penile cancer surgery [9]. Somewhat later on, in the 1990s, the SLN technique became widely developed in breast cancer [14]. Doctors only became interested in this technique to treat endometrial and uterine cervical cancers 10 years later, with the intention of de-escalating therapy to reduce surgical morbidity.

3
Since Ballester et al. published their preliminary SLN mapping experiment of uterine cervical injection in women with endometrial cancer, much valuable literature has been produced. There are data supporting the usefulness in detecting micro-metastases through the ultra-staging technique, which provides a lower complication rate than radical pelvic and/or aortic lymphadenectomy [6,21].
In their first study, SENTICOL 1, in 2011, Lécuru et al. demonstrated the feasibility and sensitivity of sentinel node detection in cervical uterine cancer [18]. A trial is currently in progress to determine the oncological safety and improvement in morbidity this procedure provides versus the reference treatment (bilateral ileo-obturator lymphadenectomy) in cancer of uterine cervix [20]. Through this work, we could gain insight into uterine cervical lymphatic drainage as a whole, since cervical injection of isotopic and/or colorimetric products could then be used to locate the SLN, i.e. the first lymphatic connexions. Few authors have studied lymphatic drainage of cervix from an anatomical point of view, studying more specifically the pathways assigned to each part of the cervix. Several injection techniques are described in the literature with regard to the detection of the sentinel lymph node: some authors perform 2 injections in the 9 o'clock and 3 o'clock quadrants and others inject in the 4 quadrants of the cervix: at 12 o'clock, 3 o'clock, 6 o'clock and 9 o'clock. To date, there is a lack of knowledge regarding lymphatic drainage of the cervix in these different anatomical regions: anterior lip or posterior lip. In the area of conservative and minimally invasive management, it is legitimate to better characterize the anatomical lymphatic drainage of the cervix.
The aim of this work was to identify a physiological specificity of the lymphatic drainage region related to the different parts of the cervix.

Materials and methods
First, an anatomical study on fresh cadavers was conducted as follows: the non-injected bodies were warmed up beforehand by leaving them in ambient air for a minimum of 24 h, or by an internal abdominopelvic bath with warm water. With the aid of a speculum, an injection of 2 ml of patent blue was then administered in the anterior (at noon) or posterior ( at 6 o'clock) lip of the uterine cervix, under pressure. A midline laparotomy was then performed at least 6 h after the injection, as this is the minimum time required to obtain lymphatic migration according to previous anatomical work (unpublished data). Drainage regions were then identified by inspection and dissection of the drainage areas (Fig. 1). These areas were inspected from proximal to distal: First, the anterior and posterior uterine serosa and the "parametrium" (regions located laterocervically). Then we dissected the "ilio-obturator areas" defined according to the surgical anatomy recalled by Cibula [10] as the external iliac and obturator areas, between the external iliac axis laterally, the umbilical artery medially, the common iliac artery bifurcation cranially, and the obturator nerve distally. The "common iliac regions" (lateral and ventral to the common iliac vessels) and then the "lumbo-aortic regions" (presacral, lateroaortic, preaortic, laterocaval, precaval and inter-aortic-caval) are then examined. Finally, we reviewed the less classical drainage routes, in particular via the utero-sacral ligaments and the infundibulopelvic ligaments.
In parallel, a retrospective radiological and pathological study was carried out using a database from a previous study [26] (number's register: RnIPH 2020-09, cover by the MR-004). In brief, the retrospectively selected patients had presented with early-stage uterine cervical cancer, i.e. stage IA1, IA2, IB1, IB2 or IIA1 in the 2018 FIGO classification, without extension to the parametrium or neighbouring organs outside the upper 2/3 of the vagina, and for which surgery was performed at the Toulouse and Claudius Régaud University Hospital sites. Surgery consisted of a local procedure (conization, trachelectomy, hysterectomy) and staging by bilateral ilioobturator lymphadenectomy. In our study, the patients who were selected all had lymph node involvement. The precise location of the tumour was determined by MRI analysis, clinical examination data, or information from the pathology report. An analysis of the pathology report of the lymphadenectomies also allowed us to establish the location of the affected lymph node(s). Therefore, a correlation between the anatomy of the primary lesion site and its lymphatic drainage site was established for each patient (Fig. 2).

Results
For the anatomical study, we had 15 fresh bodies of subjects aged 66 to 88 years at the time of death. In 7 cases (4 bodies injected in the anterior margin, 3 bodies in the posterior margin), the bodies had been taken out at room temperature less than 6 h prior to the injection. An internal rewarming technique was then used: after median laparotomy an intra-abdominal bath was administered with water at approximately 37 to 40 °C. The other bodies were taken out at room temperature 12 to 24 h before use. With the aid of a speculum, a 2 ml submucosal injection of patent blue was administered in the anterior lip of cervix in 8 cases, and in the posterior one in 7 cases. After an average waiting period of 15 h (range: 6-24), we identified drainage regions by inspection and anatomical dissection.
In cases where the injection was administered in the anterior lip (n = 8), there was therefore bilateral ileo-obturator drainage in 100% of the cases (8/8). Drainage in the anterior uterine serosa was noted in 87.5% of the cases (7/8), and never in the posterior uterine serosa. In 1 case (12.5%), there was ileo-obturator as well as bilateral common iliac and left infundibulopelvic ligament drainage. In another case there was a drainage pathway in the right parametrium, and on the left side deep in the utero-sacral ligament. Finally, a drainage pathway in the 2 parametria, the 2 ileo-obturator areas as well as the anterior uterine serosa and the 2 infundibulopelvic ligaments demonstrated drainage that is more atypical.
If the injection was administered in the posterior lip of the cervix (n = 7), there was drainage in the parametrium and posterior uterine serosa in 85.7% of the cases (6/7). In 1 case (14.2%), no drainage beyond the parametrium was noted. In 2 cases (28.5%) there was unilateral ileoobturator drainage. Finally, in 3 cases (42.8%), there was bilateral drainage through the utero-sacral ligaments.
For the clinical study, the records of 20 patients managed for a small cancer of the uterine cervix less than 4 cm in size and with lymph node involvement were analysed. Whether the tumour location was on the posterior lip, anterior lip, lateralised or circumferential was determined according to the clinical examination, MRI and/or pathology report. The tumour was located on the anterior lip of the cervix in 6 cases (30%), the posterior lip in 7 cases (35%), and lateral or circumferential in 7 cases (35%). Only ilio-obturator lymphadenectomy was performed for this indication, in accordance with the recommendations. Ilio-obturator lymph node involvement was noted bilaterally in 60% of the cases of anterior tumours (4/6), and only unilaterally in 86% of the cases (6/7) of posterior lip cancer of the cervix. For circumferential or lateralised tumours (n = 7), ileoobturator drainage was also predominantly unilateral (5/7, 71%).

Discussion
Only few studies have been conducted on lymphatic drainage of the uterine cervix. Indeed there is little data, which is sometimes contradictory, and above all variable, depending on the subjects studied, the injection technique used and probably the way in which the drainage pathways concerned are sought. Yet this knowledge of drainage is of major clinical interest. In fact, targeted removal of lymph nodes in a specific drainage region reduces the morbidity of surgical procedures, in particular the risk of lymphoedema [1,8]. Likewise, considering that pelvic lymphadenectomy generates several complications, the sentinel lymph node (SLN) technique has been introduced for 20 years and is increasingly used in cervix cancer. One of the interests of SLN is to highlight the aberrant drainage territories that are not systematically dissected during routine lymphadenectomies [7]. It therefore seemed important to us to better characterize the lymphatic drainage according to the anatomical regions of the cervix to better apprehend the aberrant drainages. Our results show that drainage is mainly in the iliac areas, but through our anatomical examination, we were also able to demonstrate more atypical drainage, in particular in the uterine serosa, the infundibulopelvic ligaments, the utero-sacral ligaments and the common iliac region. However, we visualised no para-aortic drainage. Our work has also provided additional information to refine our knowledge of uterine cervical lymphatic drainage. In our anatomical work we observed so-called "classic" drainage systematically for injections in the anterior lip of the cervix at the ileo-obturator connexion, sometimes associated with more atypical drainages (common iliac, uterine serosa or infundibulopelvic ligament). Drainage of the posterior lip appeared to be less conventional, with much less systematic drainage via the ileo-obturator lymph nodes, but rather via the utero-sacral ligaments and the posterior uterine serosa.
Rouvière [23] gave a definition which is often used as a reference in books on surgical techniques. Drainage of uterine cervix is schematised in 2 regions: the external and the internal iliac. In 1997, Lécuru et al. conducted a preliminary anatomical study for the SENTICOL trials [19]. For this, they injected the cervices of 6 bodies. Even though there were few patients in this study, they were able to observe a vast majority of the drainages under the external iliac veins (40%) and at the roots of the internal iliac veins (60%), which is in keeping with the work by Rouvière. This definition has the advantage of being simple but is probably too simplistic, as we shall see later, and we can already note that amongst the bodies injected by Lecuru, drainage in the infundibulopelvic pathway occurred in 20% of the cases [19].
The SENTICOL trials subsequently provided very interesting data on lymphatic drainage of uterine cervix. Therefore, in 2020, Balaya et al. took back the patients in the SENTICOL 1 and 2 trials who were managed for early-stage cancers of uterine cervix, without preoperative suspicion of lymph node involvement, with identification of SLNs to determine the risk factors of failure of the technique [4]. Most of the sentinel nodes were in the ileo-obturator and external iliac zone (n = 977, or 82.6%), followed by the common iliac region (n = 109, i.e. 9.2%). Drainages were also described from the SENTICOL 1 and 2 patients [5]: 3.9% of the drainage regions were in the parametrium, 1.6% in the promontory, 1.5% in the para-aortic area, and 0.5% in other areas. SENTICOL 3 is currently in the inclusion phase to validate the effect of SLN alone without lymphadenectomy on recurrence-free survival and quality of life. Other studies have confirmed these so-called "atypical" drainages, outside the iliac areas. In 2004, Altgassen et al. reported a certain proportion of para-aortic SLN [3] and in 2006, Marnitz et al., for the AGO (Association of Gynaecological Oncologist), found para-aortic drainage in 4% of the cases, common iliac drainage in 5%, external iliac in 5%, internal iliac in 8%, and parametrial in 7% [22]. They also observed that more para-aortic SLN were found in the event of joint injection of patent blue and a radioactive tracer, which shows the potential variability of results linked to the technique used.
It is of note that finally, few studies have been conducted on lymphatic drainage of uterine cervix. There are few data, which are sometimes contradictory, and above all variable, depending on the subjects studied, the injection technique used and probably the way in which the drainage pathways concerned are sought.
Yet, this knowledge of drainage is of major clinical interest. In fact, targeted removal of lymph nodes in a specific drainage region reduces the morbidity of surgical procedures, in particular the risk of lymphoedema [1,8].
At present, sentinel node detection in uterine cervical cancer is still in the experimental stage to prove oncological safety, as the resulting reduction in morbidity no longer needs to be proven. It is classically done by double detection of migration after injection in the cervix, using a colorimetric method (patent blue), and radioisotope (technetium 99 m). Although this dual detection has resulted in a better detection rate of sentinel nodes, there are still failures of migration and/or detection. In 2020, Balaya et al. reported a detection rate that was at least unilateral in 94.1% of the cases, and bilateral in 80.4% of the 405 patients included in the analysis [4]. Similarly, in their study of 245 patients, Sponholtz et al. reported a unilateral detection rate of 96.3% and a bilateral detection rate of 82% [25].
One technique that is being developed uses indocyanine green fluorescence. In a meta-analysis published in 2016, Ruscito et al. revealed that in uterine cervical cancer, this method showed a better detection rate than patent blue used alone, and a detection rate equivalent to the combination of blue and technetium [24]. However, these immunofluorescence and radio-guided detection techniques require structural equipment that is not always available in developing countries, and are expensive, to the patient's detriment, in regions where the incidence of uterine cervical cancer is highest [27]. The importance of being able to target lymphadenectomy regions according to the location of the cancer could therefore avoid some morbidity by optimising the surgical management of cancer in centres where the SLN technique could not be implemented.
The contribution of PET scans in lymph node staging of uterine cervical cancer remains to be demonstrated. In fact, given the lack of therapeutic value of lymphadenectomy, which is only of interest for predicting the extent of radiotherapy fields, an alternative, non-invasive examination is needed. A review of the literature from 2020 examines the performances of PET scans in predicting lymph node involvement in locally advanced uterine cervical cancer [2]. It appears that the positive and negative predictive values improve when the examination is performed in a population with a higher prevalence of lymph node involvement, i.e. in large volume and/or locally advanced cancers. Therefore, in these populations PET scans are important and, although their exclusive use is not recommended at present, could allow surgery to be avoided. However, in small cancers the performance of FDG PET is less clear. In the Sponholtz study, 103 patients with a median cancer size of 28.0 mm (range 21.0-63.0), the sensitivity of FDG PET/CT alone was 14.8% (95% CI 4.2-33.7%), and the specificity, 85.5% (95% CI 75.6-92.5%) [25]. Therefore, it is difficult to bypass surgical lymph node staging, especially in early-stage cancers of uterine cervix.
The limitations of our work are mainly due to the number of cases presented, which is partly due to the difficulties associated with cadaveric studies, and partly due to the small number of patients with early-stage uterine cervical cancer (for whom the precise location of the tumour is therefore possible) and lymph node involvement (which makes it possible to identify the tumour drainage site). In the clinical part of our study, the limitation of the information collected is also related to the fact that only pelvic lymphadenectomies were performed, and consequently, more atypical drainage areas may have been overlooked. Hence, once again, there is a clinical interest for the SLN in the complete mapping of drainage regions during oncological treatment of uterine cervical cancer.
Likewise, the lymphatic circulation in the corpse is difficult to reproduce which surely explains the absence of drainage highlighted in the para-aortic nodes. Nevertheless, this is a preliminary study, the results of which encourage further work with a module allowing fluids to recirculate. The perfused re-ventilated cadaver model (SIMLIFE model) could contribute considerably to the examination of the lymphatic system, which is fragile and difficult to study as shown by the extremely variable results of previous work. To date, the simlife model is mainly used for pedagogical purposes to learn surgery, allowing ultra-realistic situation in simulation, thanks to a pulsating circulation device and a ventilation device [11][12][13]17]. We are currently performing SIMLIFE procedures for clinical research purposes. Indeed, the lymphatic circulation is done in large part thanks to the mechanism of vis a tergo (residual arterial pressure) and vis a latere (pulsatility of the nearby arteries). We can therefore see here the interest of using this reperfused model with arterial pulsatility to approach a quasi-physiological model of lymphatic drainage.

Conclusion
Better knowledge of the lymphatic drainage of the uterine cervix is necessary to be able to propose "individualised" surgical treatment to patients with uterine cervical cancer, without ignoring more atypical drainage regions according to the precise location of the tumour. Today, there are still large gaps in our knowledge of the anatomy and physiology of the human lymphatic system. With this work, we have been able to provide a better knowledge of the lymphatic drainage of the uterine cervix, knowledge that deserves to be deepened, thanks to the new technologies that are currently being developed. the institutional and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.