Despite initial passion and enthusiasm among surgeons and gastroenterologists, natural orifice translumenal endoscopic surgery (NOTES) has not succeeded in attracting continuous interest from doctors and industry. Its potential use has been demonstrated in a wide variety of preclinical studies, but this has resulted in only limited use in clinical settings [1, 2]. Although the reasons for this slow acceptance remain to be fully addressed, technical difficulty in using flexible endoscopes and instruments in the abdominal cavity is no doubt one of the biggest issues.

As one of the practical solutions to the problems in flexible instrumentation, use of rigid endoscopes and instruments has been proposed by some NOTES researchers [2–4]. Use of rigid instruments has been encouraged especially for transvaginal (TV) procedures, mainly due to their simplicity in achieving access and closure [2–4]. However, the safety and efficacy of using rigid/straight instruments via the vagina have not been fully evaluated for each abdominal organ with different positional relationship to the vagina. Some complications in using new techniques and technologies do not fully depend on the skill of the surgeons, and may be inherent to the surgical instruments themselves. The purpose of this study is to obtain baseline anatomical data necessary for safer and more effective use of rigid TV instruments, by three-dimensional (3-D) radiologic measurements.

Methods

This was a retrospective study using radiologic data compiled in hospital data archives. The study protocol below was approved by the Osaka University Institutional Review Board (folio no. 08282 dated 18 February 2009).

Consecutive female Japanese patients with aortic aneurysm who underwent whole-body multidetector computed tomography (MDCT, TSX-301A; Toshiba Medical Systems Corp., Tochigi, Japan) from January 2008 to December 2008 as their preoperative evaluation were identified in the clinical database. Patients’ demographic and anthropometric data including age, body weight, and height were obtained from their medical charts. The 3-D images were reconstructed from the MDCT images that were obtained in a standard fashion, and were displayed using a dedicated software package (Aquarius NET DICOM viewer; TeraRecon, Inc., Tokyo, Japan). First, five anatomical landmarks were identified on the reconstructed 3-D images of each patient: the posterior fornix of the vagina (V), the gallbladder fundus (G), the esophagogastric junction (E), the upper pole of the spleen (S), and the sacral promontory (P). Patients with unsuccessful demonstration of any of these landmarks were excluded from the study. Then, the following three measurements were performed using corresponding MDCT slices:

Straight-line distance from the vagina

The straight-line distance was obtained between V and each target: G, E, and S (Fig. 1).

Fig. 1
figure 1

Straight-line distance measured from the vagina to each target on a 3-D image. V vagina, G gallbladder fundus, E EG junction, S spleen, upper pole

Full size image

Transverse deviation from the midline

The angle formed by the “V–target” line and the midline was obtained on most appropriate transverse slices (Fig. 2). The slices which gave the clearest demonstrations of the targets were selected.

Fig. 2
figure 2

Transverse deviation from the midline obtained as an angle formed by the “V–target” line and the midline

Full size image

Sagittal deviation from the “V–P” line

The line between V and P (V–P line) was set on the most appropriate sagittal slice, and the angle formed to the “V–target” line was obtained (Fig. 3). When the target is “above” the V–P line, the angle is positive, whereas when located below the V–P line, the angle becomes negative.

Fig. 3
figure 3

Sagittal deviation obtained as an angle formed by the V–P line and V–target line. P sacral promontory

Full size image

Data are given as median (range). Fisher’s exact test was used whenever indicated for statistical analysis using SAS 9.1.3 (SAS Institute Inc., Cary, NC, USA).

Results

Ten patients were excluded from the study due to unsuccessful demonstrations of the posterior fornix of the vagina (seven cases) and/or incompleteness of whole-body scanning (three cases). As a result, 41 patients were enrolled in the study. Table 1 presents patients’ demographic and anthropometric data. Due to the nature of their original disease, the group was rather elderly with median age of 77 years. Median body height was 152 cm, and weight was 49.8 kg (body mass index, BMI = 21.5 kg/m2). The series included nine scoliosis patients, of whom three were diagnosed with “severe” scoliosis based on Cobb angle assessment. Severe kyphosis was seen in five patients based on kyphotic index assessment.

Table 1 Patient demographics
Full size table

Straight-line distance from the vagina

The straight-line distance from the vagina was 26.1 (19.8–32.2) cm for the gallbladder fundus, 30.6 (25.9–36.4) cm for the esophagogastric junction, and 31.1 (24.7–37.9) cm for the upper pole of the spleen (Fig. 4).

Fig. 4
figure 4

Straight-line distance. The images represent the distance between V and G. The actual distance was obtained on a 3-D image

Full size image

Transverse deviation from the midline

The deviation from the midline was 17.7 (10.7–30.7) degrees for the gallbladder fundus, 7.0 (3.2–15.3) degrees for the esophagogastric junction (Fig. 5), and 12.9 (6.7–20.5) degrees for the upper pole of the spleen (Fig. 5).

Fig. 5
figure 5

Transverse deviation. The images represent the deviation of V–G line (left) and V–E line (right)

Full size image

Sagittal deviation from the “V–P” line

The deviation from the V–P line was 7.6 (−7.5 to 36.6) degrees for the gallbladder fundus, −7.0 (−15.5 to 19.1) degrees for the esophagogastric junction, and −10.3 (−23.4 to 5.04) degrees for the upper pole of the spleen. The percentage of “negative angle” cases, meaning that the target is located below the V–P line, was 9.8 % (gallbladder), 88 % (esophagogastric junction), and 88 % (spleen). The differences among these three angles were statistically significant (p < 0.001). Figure 6 demonstrates typical cases with positive and negative sagittal angles. There was no correlation between sagittal deviation and severity of kyphosis (data not shown).

Fig. 6
figure 6

Sagittal deviation. The positive angle case (left) indicates that the V–G line is above the V–P line. The negative angle case (right) suggests potential collision of rigid devices with the sacral promontory

Full size image

Discussion

In the beginning, the concept of NOTES was based on use of flexible endoscopes and instruments in the abdominal cavity [5]. Within several years after its initial appearance in the literature, a wide variety of surgical procedures were reported as “feasible” using the original NOTES techniques. Researchers, however, also soon realized that pure NOTES on a flexible platform is highly technically demanding and has associated potential risks [1, 2, 4]. The vision is unstable, instruments are at an early stage of development, and application of retraction is limited or only possible with use of instruments introduced via abdominal trocars [4]. As a result, almost all human NOTES cases have been performed with the assistance of transabdominal rigid instruments [1].

A more practical shift has been seen in TV NOTES. The pioneering work was mostly carried out by German researchers, performing TV cholecystectomy using rigid endoscopes and instruments via the vagina [3, 4]. They recently reported a large series of TV NOTES on a rigid platform, suggesting that TV NOTES with rigid endoscopes/instruments can be applied in routine clinical practice [2]. Based on their website-based registry, more than 2,500 cases of NOTES have been performed, mainly with rigid endoscopes/instruments [6]. The series include cholecystectomy, appendectomy, and more recently bariatric surgery and colon resection [6].

In contrast to the above rapid progress of NOTES in “German style,” optimization of rigid instruments to be passed through the vagina has been delayed. Buess proposed that surgical instruments would need a length of “around 40 cm” from the entrance of the vagina to the area of the diaphragm, and “another 20 cm” so that the instruments can be easily balanced over the pivot point of the sealing system [4]. Zornig et al. [1] used similar “extralong” instruments, and reported that TV cholecystectomy required significantly longer operating time than conventional laparoscopic cholecystectomy, mainly due to difficulty in handling such long instruments. Horgan et al. [7] reported a successful use of longer articulating grasper with approximately 75 cm length. Most of the procedures included in these series were cholecystectomies. Is 60 cm sufficient to gain access to the gallbladder? Do we need much longer instruments for other upper abdominal organs?

Our anatomical measurements first clearly demonstrated that the effective, i.e., “intra-abdominal,” length of TV instruments for upper abdominal procedures should be more than 35 cm in Japanese population. When the grasper is used for more cephalad application, e.g., retraction of the gallbladder fundus for cholecystectomy, an additional intra-abdominal length would be required. This means that the current laparoscopic instruments with standard length are too short to be used in TV surgery. In other words, use of “extralong” (60–75 cm) instruments in German-style NOTES is considered anatomically reasonable.

The pathway from the vagina to the upper abdominal organ may pose a risk of injury to midway organs [8]. One of the most obstructing anatomical structures has been known to be the sacral promontory, especially when using rigid and straight tools [9, 10]. Authors have also reported difficulty in overcoming the sacral promontory with a rigid and straight stapling device to complete TV NOTES partial gastrectomy in humans [10]. For this reason, we first measured the deviation of each representative anatomical landmark from the midline (transverse deviation), with each relation to the vagina–promontory line (sagittal deviation). As shown by the data, the gallbladder is widely deviated from the midline and therefore can be safely approached even with rigid and straight TV instruments, without being obstructed by the sacral promontory. Access to more midline and distant targets, e.g., esophagogastric junction, however, may suffer from interference by the sacral promontory. Continuous compression of this area with rigid and straight instruments may be potentially dangerous in terms of risk of compression injury, e.g., vascular thrombosis and hypogastric nerve palsy.

There are several limitations to this study. One major concern is that our group of patients might not be a good representation of the normal female population, since the group had advanced age and included a significant number of scoliotic/kyphotic patients. In addition, there exists a clear difference in stature between Japanese and Western female populations. Although not matched in age, the height of our patients (152 cm) is obviously shorter than that of German (165 cm, >18 years of age), French (163 cm, 18–70 years of age), and American females (162 cm, >20 years of age) [11]. All these factors may affect the anatomical measurements shown in our study. Another limitation is that it does not consider any impact of patient positioning in the operating room. The entry point of TV surgery, i.e., posterior fornix of the vagina, could be shifted ventrally when the patient is placed in steep Trendelenburg position. Each target organ could also be displaced with repositioning of the patient during surgery. Further accumulation of preoperative anatomical and anthropometric data, ideally with intraoperative measurement of representative anatomical landmarks, is needed to develop safer and more effective TV instrumentation.

Conclusions

Based on our anatomical measurements using 3-D reconstructed MDCT images, the intra-abdominal length of TV instruments should be more than 35 cm in Japanese population. Although widely deviated organs, e.g., gallbladder, can be safely approached with rigid and straight instruments from the vagina, additional care should be taken in approaching more midline and distant targets, since the TV pathway to those organs may be obstructed by the sacral promontory, therefore posing a risk of compression injury with rigid and straight instruments.