Surgeons have always wanted to look inside the human body but rigid and primitive instruments, poor illumination, and inadequate anaesthesia originally confined such procedures to the natural orifices. General anaesthesia and, at the end of the nineteenth century, the introduction of electric light and glass telescopes first made cystoscopy feasible and later bronchoscopy and laparoscopy. An angled telescope allowed the surgeon to see around a corner but the continuous curves of the bowel remained a barrier to further progress in gastrointestinal endoscopy until the development of flexible glass fibres. Gynaecologists were the first to appreciate the advantages of laparoscopy. Parallel developments in instrumentation and illumination soon meant that some therapeutic procedures could be performed although there were limitations because the operator was the only one who could see what he was doing and one of his hands was occupied holding the endoscope itself. Nevertheless laparoscopic sterilization and transurethral resection of the prostate have been standard operations for many years.

Minimally invasive or endoscopic surgery for a much wider range of operations has become possible as a result of the development of miniature television cameras which can be attached to any suitable telescope. The cameras are unobtrusive, light in weight, and provide a perfect and magnified colour image of the operative field on a television monitor. The surgeon has both hands free to manipulate instruments and any number of assistants can watch what is being done on separate monitors and operate as well. New instruments have been developed to overcome the limitations on access and movement within the body cavities and new ways of performing old and well established operations have also been devised. In that respect this chapter will be out of date by the time it is published but even so the basic principles of endoscopic surgery are already well established.

Some endoscopic operations, such as those on the nasal sinuses, take place within a natural cavity so that no special methods are needed to obtain a space in which to work. But for operations within the chest or the abdomen it is first necessary to create a pneumothorax or a pneumoperitoneum. The most convenient way to introduce gas is to use the spring loaded Veress needle inserted through a small stab incision in the skin (Fig. 1) 844. The needle has a sharp but flat bevelled point within which there is a blunt, hollow stylet through which gas flows from a distal side hole. As soon as the point of the needle penetrates the peritoneum or the pleura the stylet springs forward and pushes away the lung or the bowel thus lessening the risk of penetrating either structure.

The Veress needle can be introduced anywhere through the chest wall although the umbilicus is the most popular site in the abdomen. It is best to keep well away from any previous incisions as there are likely to be local adhesions. When creating a pneumoperitoneum it is wise, first, to be sure the bladder is empty and, after insertion, to test that the point of the needle is in the correct place. A hanging drop of saline in the hub of the Veress needle should fall easily into the peritoneum when the abdominal wall is lifted and once gas flow commences the intra-abdominal pressure should initially remain low whilst the gas flow rate should be high. Some surgeons prefer to introduce the first cannula under direct vision through a small skin incision and then to inflate the cavity. A gas-tight seal is obtained with a purse-string suture tied around the port. This is certainly a safe method if local adhesions are to be expected.

Carbon dioxide is the best gas to use; it is rapidly soluble should an embolism occur and it does not support combustion. The lung collapses spontaneously once gas is introduced into the pleural space and it is then only necessary to maintain a slight positive intrathoracic pressure. By contrast, maintenance of a pneumoperitoneum requires a constant intra-abdominal pressure of about 15 mmHg. Pressures in excess of this may lead to difficulties with ventilation whilst a fall in pressure will lead to a collapse of the pneumoperitoneum and the operative field will disappear from view. Modern high flow, electronically controlled, insufflators will maintain a preset intra-abdominal pressure and automatically correct minor leaks as well as the more substantial loss of gas that occurs every time an instrument is taken in or out of a port.

Artificial cavities can be created almost anywhere in the body by the forced insufflation of carbon dioxide into a natural tissue plane. The fibrous bands that cross these artificial spaces can be a nuisance and an alternative, in some circumstances, is to create the cavity by insufflating a balloon at an appropriate site within the tissues. The balloon is then removed but the cavity is maintained by the continued insufflation of carbon dioxide gas under pressure.

The surgeon must be able to see what he is doing. Modern endoscopes are designed to carry light to illuminate the operative field and to pass the image back to the external television camera all within a single cylindrical metal tube (Fig. 2) 845. The endoscope and the necessary instruments gain access to the operation site through cylindrical cannulae, or ports, placed through the body wall (Fig. 3) 846. Where the cavity is only maintained by positive pressure, as in the abdomen, these ports must be provided with valves to retain the gas within the cavity.

The telescope, the light source, and the television camera are essential for endoscopic surgery and are just as sophisticated as the operations for which they are used. Modern equipment is reliable and reusable but breakdowns do occur and will require expert repair so that spare equipment should be available in any large theatre suite. Most of the instruments are made either as reusable items that can be repeatedly sterilized or as single use disposable items. A few instruments, particularly those used to staple tissue together, are only available as disposable items. From the practical point of view either type of instrument is perfectly satisfactory and the choice is dictated purely by preference and cost.

The proposed operation and the shape and size of the individual patient dictate the precise placement of the cannulae. The basic layout is two working ports and one port for the telescope. In general it is best if these three ports are placed at the apices of an equilateral triangle with the operative field midway along one side of the triangle (Fig. 4) 847. Sometimes two sides have to be longer than the third but they should be equal and the endoscope should lie at the apex of the two long sides with the operative field midway along the short side. If this arrangement is followed then the tips of the working instruments will always lie at right angles across the field of view when they are being used. Additional ports, which are mostly used for retraction, should be placed outside the basic triangle.

The most common problem is for the operative field to lie further away from the endoscope than the midpoint between the two operating ports. If this happens the telescope and the instruments may touch each other which can make it impossible to work and to see what you are doing at the same time. Even if the endoscope and the instruments do not touch it may still be impossible to see the tips of the instruments as they are being used. This is inherently dangerous because the two-dimensional image makes the perception of depth very difficult. If scissors are being used it is a good rule that the manoeuvre should be abandoned unless the tips of the scissors can be clearly seen and it is absolutely certain that nothing unintended will be cut. An oblique viewing endoscope and instruments with curved blades can sometimes overcome both these difficulties.

Laparoscopic cannulae come in two standard sizes with an outside diameter of 5.5 mm and 11 mm (Fig. 3) 846. They are designed to take 5 mm and 10 mm instruments respectively. Larger and smaller sizes are made for special purposes and reducing sleeves allow the use of small instruments through the larger ports. Most cannulae incorporate a gas-tight valve. Disposable cannulae and the smaller reusable ports use ball or flap valves whilst sliding trumpet valves are fitted to the larger reusable cannulae. Each cannula is fitted with a trocar which should have a sharp hollow-ground pyramidal point. When the trocar is rotated the sharp edges and the pyramidal shape of the point force the tissues apart and also cut some of the fibres so that less force is required to gain entry into a cavity.

To insert a cannula the whole assembly is held in the palm of the hand with the index finger lying along the shaft to prevent excessive penetration (Fig. 5) 848. Each trocar and cannula is introduced through a small skin incision and then wound through the body wall with minimal pressure, thereby reducing the risk of a sudden uncontrolled entry into the cavity. Unless an open technique is used the first cannula is inserted blind and therefore carries the greatest risk of damage to underlying structures. Thereafter the entry of each subsequent cannula can be observed directly (Fig. 6) 849 although it is still wise to angle the trocar away from any important structures.

An external cold light source with an output of at least 250 W is essential for endoscopic surgery (Fig. 7) 850. Light is conducted to the endoscope through a flexible fibreoptic cable and is automatically controlled by a feedback from the video camera so that more light is available as the field of view enlarges. Even so there are physical limits on the amount of light that can be delivered and losses should be reduced to a minimum. This means taking care not to break the glass fibres within the fibreoptic cable and removing blood, which absorbs light, from the operative field.

The standard rigid endoscope is 10 mm in external diameter and contains the light cable and a glass rod-lens telescope (see Fig. 9 852). The illumination and the field of view, which is directly ahead from the tip of the telescope (0° forward viewing), are exactly matched. A wide-angle lens at the tip of the telescope is the most useful for general purposes.

There are many variations in this basic design and forward-oblique viewing angles (30 and 45°) are certainly helpful in certain circumstances (Fig. 8) 851. Endoscopes as fine as 0.7 mm in outside diameter are available for intravascular work and for some operations the inclusion of an operating channel alongside the endoscope is an advantage. Endoscopes must be kept clean and undamaged. The camera and the endoscope should be carefully dried before they are put together to eliminate condensation and warming the endoscope and antifog solutions applied to the distal lens will reduce misting.

Endoscopes with a flexible end are already made and in the future the video camera may be placed at the tip of the larger instruments. Even stereoscopic endoscopes are under development.

The colour television camera is externally mounted directly on to the end of the endoscope without the use of a beam splitter (Fig. 9) 852. Any of the modern microchip cameras designed for endoscopic work are suitable although there are minor differences in colour reproduction and ease of use. All of them have a marker to indicate the orientation of the image and correct alignment of the camera and the endoscope is essential when a forward-oblique instrument is used. The cable from the camera leads back to a control box at the side of the operating table (Fig. 7) 850. The electronic image is then fed to two or more monitors but it can also be recorded or transmitted over any distance, even to the other side of the world. Immersion in glutaraldehyde is the usual method of sterilization although the camera and the cable end must be protected from water. An alternative is to place the unsterile camera inside a sterile plastic sheath. This avoids the use of glutaraldehyde and the risk of water damaging the camera or its cable.

The basic endoscopic instrument is either 5 mm or 10 mm in diameter and between 30 and 35 cm in length (Fig. 10) 853. Shorter and thinner instruments are available for paediatric work and some stapling devices are 12 mm or more in diameter. Originally the working ends were simple adaptations of a conventional design which were attached by a long shaft to an ordinary handle but many of these instruments were not very easy to use in practice. A plethora of new instruments have now been developed specifically for endoscopic work although the demand for a long, narrow, and circular design does impose quite severe limitations on the engineer.

In an open operation there is complete freedom to move in any direction but an endoscopic instrument has only 3 degrees of movement. The shaft can move in or out (translation), the tip can rotate about the axis (axial rotation), or the whole instrument can rotate around the point of entry (relative rotation). These limitations and the fixed relationship of the working ends to the handle in most surgical instruments can impose impossible demands on the surgeons' hands and a better arrangement is to allow the shaft of the instrument to rotate on the handle (Fig. 11) 854. There are also new designs of handle that can be used wherever they lie in the hand. Holding tissues in order to work on them is as essential in an endoscopic operation as it is in an open one but it can be difficult when the hinge of the forceps lies very close to the working end of the instrument. So far no entirely satisfactory design has been developed to overcome this problem but most of the available endoscopic graspers rely on refinements to the working surface of the blades usually by increasing the number or the size of the teeth in contact with the tissues. Conventional ratchets for securing an instrument to the tissues are almost impossible to release easily when they are positioned through an endoscopic cannula and screw threads on the handles (Fig. 12) 855 or a trigger grip are better alternatives.

The need to pass every instrument down a straight cannula imposes a limit on the curvature of the tips. This is a serious disadvantage when it is combined with the limitations on movement since it is commonly necessary to surround a structure before it is occluded and divided. It is true that the angle of approach can be infinitely varied if the curved tip can rotate on the shaft but it is still only possible to approach the structure from one general direction. A clever solution to this problem takes advantage of the memory that some metal alloys possess. The working end of the instrument is housed within a shaft of suitable diameter. When the end is pushed out of the housing the extension with the working end attached assumes a natural curve which makes dissection around a structure much easier. The end is then withdrawn into the housing before the instrument is removed.

In conventional surgery most instruments are designed for a single purpose and an efficient scrub nurse has always ensured that the surgeon has the instrument he needs immediately available. This is not so easy in endoscopic operations because every instrument has to be passed in and out of a cannula. This takes time and so there has been a move to devise instruments which do more than one thing. The most obvious adaptation has been the addition of diathermy to every suitable instrument so that both dissection and haemostasis can be accomplished without the need to change instruments (Fig. 13) 856.

There are two fundamental parts to any surgical operation. Tissues and structures are first divided and are then joined back together again either in a different way or with part of the structure missing. In conventional surgery a scalpel and a pair of scissors are the basic tools for dissection and restoration is achieved with a needle and thread. Unfortunately neither the needle nor the knife adapt well to endoscopic work and alternative techniques have had to be developed.

Dissection with a knife does not work in endoscopic surgery. Depth perception is difficult, the incisions always bleed and a sharp unprotected, and often invisible, blade could easily cause damage. Scissor dissection is not ideal either although it is sometimes essential. The tiny blades do not remain sharp for long particularly if they are also used for diathermy and the hinges become loose so that the tissues are not cut cleanly.

Gently tearing the tissues apart along the lines of least resistance either with a jet of water (aquadissection) or with a blunt instrument is useful. The dissection can be both accurate and precise because of the magnified image on the monitor even though the technique is not particularly attractive. Blunt instruments are safer when the tips cannot be seen and any small vessels can be occluded with diathermy as the dissection proceeds. Some oozing is inevitable and blood, which absorbs light, should be immediately sucked or washed away with warm saline solution. Even small amounts of blood can obscure the view during an endoscopic operation and so, whatever method of dissection is used, haemostasis is important.

Coagulation of the tissues at the same time as they are divided is ideal and monopolar and bipolar diathermy, laser light, and ultrasound are all useful tools for dissection and adapt easily to the endoscopic environment. They minimize oozing and will stop haemorrhage from a visible vessel. When monopolar diathermy is used there should be a visible effect on the tissue the moment the current is passed (Fig. 14) 857. If nothing happens there may be a short circuit and therefore the possibility of damage to another structure. The most common error is contact between the metal of an instrument and an uninsulated cannula outside the field of view. Lasers are also potentially dangerous if improperly used. The staff must wear eye protection and backstops may be needed on the instruments to ensure that structures beyond the intended target are not damaged by the beam.

Bipolar diathermy will occlude vessels the size of the appendicular artery but anything larger than this requires some form of mechanical closure. It is not practical to tie a conventional ligature through an endoscopic cannula but slip knots which can be placed around a structure and then pulled tight in a single movement are useful. The original Roeder knot (Fig. 15) 858 was designed to be used with thick chromic catgut and there are modifications suitable for use with polyglactin and polydioxanone. All these knots are safe and reliable provided that they are properly constructed and correctly laid down. Every endoscopic surgeon should be able to tie them.

A more convenient method for occluding an artery or a duct is to apply a clip. Titanium metal clips are familiar and are available in multifire guns (Fig. 16) 859 especially designed for endoscopic work. Polydioxanone clips (Fig. 17) 860, which will eventually dissolve, have a locking mechanism which ensures that they are applied correctly. They can be difficult to close and have to be applied one at a time. Only one size of clip of either type is available at the present time.

A vascular pedicle can be occluded and divided in one manoeuvre with a disposable stapler. The design is similar to the instruments used for intestinal anastomosis except that the staples are smaller and are placed closer together to ensure good haemostasis. Their use is limited by their cost.

Endoscopic suturing requires time, constant practice, and considerable dexterity. Special ski needles are available (Fig. 18) 861 and the knots can be tied externally and tightened with a pusher or internally with instruments. Quicker and easier but more expensive are the mechanical stapling devices that are becoming available (Fig. 19) 862. Endoscopic adaptations of the larger stapling instruments which are in common use in conventional surgery are available although they are expensive. They will, in time, become the standard method for joining tissues together in an endoscopic operation.

Most, but not all, endoscopic procedures require the removal of a specimen at the end of the operation. Small pieces of tissue can be easily removed through a large cannula. Larger specimens are more difficult. Some can be reduced in size by removing the contents and then the specimen either through a cannula or through the incision when the cannula is withdrawn. Large solid specimens are more difficult still. A few can be extracted through a natural orifice which is opened during the course of the operation but most must be reduced to smaller pieces and removed through one of the cannula sites. The specimen is placed inside a tough plastic bag introduced into the body cavity (Fig. 20) 863. The bag is brought to the surface and the tissue is morcellated or liquefied with a special instrument within the bag inside the body. The pieces of tissue are removed or sucked away and the bag is then removed as well. The specimen is effectively destroyed so far as histological examination is concerned and suitable biopsies must be taken first.

Finally one of the cannula sites can be enlarged or a separate incision made specifically to remove the specimen. This obviously loses some of the advantages of the minimally invasive approach although the incisions are often smaller, more discreet, and less morbid than they would otherwise be.

Most surgeons are not familiar with endoscopic techniques but they can be used in every branch of surgery and in the future every surgeon will need endoscopic skills. Some initial training outside the operating theatre is useful for learning the basic manoeuvres that are needed. Flexible endoscopists are familiar with the two-dimensional image and the limitations of manipulation through an endoscope but everyone finds the reversed movement of endoscopic instruments as a result of the pivoting action of the body wall very strange to begin with.

Although more sophisticated training systems are now available, the most useful practice in our experience is to learn to peel a tangerine with scissors and forceps inside a box whilst watching the procedure on a television monitor (Fig. 21) 864. This requires a team of two people, one to hold the camera and the other to operate on the fruit. It is easy to extend the idea to include learning to sew, tying a knot, or placing a clip. Once the basic technique is mastered it is then a question of assisting an experienced surgeon at a number of laparoscopic procedures and later starting to do parts of an endoscopic operation under supervision as with conventional surgical training. Every operation should be recorded for it is then easy to review and reconsider any difficulties that have been encountered.

Anaesthesia for endoscopic surgery is little different to conventional anaesthesia for the same operation. Relaxation is needed for abdominal operations and the lung must be collapsed for procedures within the thorax. Excessive pressure within a body cavity will cause undesirable physiological effects but these are rare with the modern electronically controlled insufflators, although there is always the risk of a gas embolism. Some carbon dioxide will be absorbed during the course of a long operation but it is easily removed by increasing the ventilation rate. By its very nature there is less need for analgesia after an endoscopic procedure but it is worthwhile injecting bupivacaine around the small incision sites.

Endoscopic procedures are popular with most theatre staff because everyone can easily see what is being done and there are obvious benefits for the patient. However, the scrub nurse can feel less involved than usual because there are fewer instruments to pass although the individual instruments often need particular attention. Hinge screws require regular tightening and the endoscope needs constant cleaning to ensure a clear image and there are many other similar little tasks so that an attentive and interested nurse will always make an operation run more smoothly.

Staff in the sterile supply department are involved too for the instruments are both delicate and hard to clean. They often contain fiddly little pieces which are essential for proper function but are easily lost when the instrument is taken apart. For example, trumpet valves must be oiled every time they are sterilized otherwise they will stick in use. Once again attention to detail can make all the difference in the operating theatre.

The development of endoscopic surgery can be compared in importance to the introduction of anaesthesia into surgical practice and the impact will be felt well into the next century. The enthusiasm for the endoscope now is in marked contrast to the rejection of the laparoscope in the past. It is true that gynaecologists have always used laparoscopy but few general surgeons embraced the technique, mostly because only a limited number of organs could be seen and not many abdominal operations are possible with one eye and one hand occupied with a laparoscope. It is the development of video equipment that has made the revolution possible.

It was also fortuitous and fortunate that the first operation to make endoscopic surgery a household word should have been cholecystectomy. The endoscopic operation is identical to an open procedure, it is not particularly difficult to do in most cases, and the benefits to the patient were, and still are, dramatic. There has been some evidence of overenthusiasm. A few obsolete operations have been resurrected because they are easy to do laparoscopically and a few serious complications have occurred but overall the advantages are obvious and patients themselves have been pressing to have an endoscopic operation where possible.

The intention in this section is to illustrate what is presently possible in detail for some operations and in outline for others. Many more endoscopic operations will have been described by the time this chapter is published and we shall leave the reader to discover more information either elsewhere in this book or in a specialist text.

The significant disadvantages of a major abdominal wound and the success of laparoscopic cholecystectomy has ensured the rapid development of endoscopic abdominal surgery. The retroperitoneal organs, with the exception of the kidney, are inaccessible at the moment and most of the operations, like cholecystectomy, are endoscopic adaptations of a conventional operation. A few procedures, for example hernia repair, explore new ideas and more of these can be expected.

Laparoscopic cholecystectomy was first performed in 1987 by Phillipe Mouret in Lyons, France. The operation was immediately accepted by the surgical community and has been followed by a worldwide wave of enthusiasm for endoscopic surgery. The laparoscopic operation reduces the access required to an absolute minimum whilst at the same time achieving the same objective as a conventional open cholecystectomy.

Every patient who needs a cholecystectomy and is fit for a general anaesthetic should be suitable for the laparoscopic technique although pregnancy and portal hypertension might make the operation difficult. However, some of these patients will also have stones in the bile duct and, at the present time, laparoscopic exploration of the ducts is not widely available. It is important to identify these patients in advance and we select patients for preoperative endoscopic cholangiography on the basis of a history of jaundice, abnormal liver function tests, and dilatation of the bile ducts on ultrasound examination. Normal liver function tests virtually exclude the presence of ductal stones and only half the patients with an abnormal ultrasound will have an abnormal ERCP (endoscopic retrograde cholangiopancreatography). Any stones that are found on the preoperative cholangiogram can be removed, when appropriate, through an endoscopic sphincterotomy and the patient can then have a laparoscopic cholecystectomy at a later date. Preoperative intravenous cholangiography is preferred in some centres.

There are no absolute contraindications to a laparoscopic operation although the discovery of some unsuspected disease or problem during a preliminary laparoscopy may require a full laparotomy. Obesity is not a contraindication. The view of the operative field is just as good in a fat patient as it is in a thin one and the absence of an abdominal incision reduces the incidence of complications. Dense peritoneal adhesions, acute cholecystitis, severe chronic cholecystitis, and difficulty in identifying the anatomy in Calot's triangle can make continuing with a laparoscopic operation dangerous. Conversion to an open operation in these circumstances, which is necessary in about 5 per cent of patients, demonstrates good surgical judgement and must not be regarded as a failure.

The principles of the laparoscopic operation are the same as those in an open procedure. The cystic duct and the cystic artery are dissected out, occluded and divided, and the gallbladder is then dissected out of the liver bed. Diathermy haemostasis is used where necessary and the gallbladder is extracted from the abdomen through either the umbilical or the epigastric port.

In contrast to the open operation the common bile duct is not always seen and retrograde dissection is not really feasible because upward retraction of the fundus of the gallbladder is essential for exposure of Calot's triangle. Cholangiography can be done before the cystic duct is ligated and divided if necessary.

General anaesthesia with full abdominal relaxation is essential and patients should receive the usual prophylaxis against wound infection and venous thrombosis. A urethral catheter is unnecessary provided that the patient empties his or her bladder before coming to the theatre and we only insert a nasogastric tube if the stomach is distended with gas. The patient lies flat on the operating table and the surgeon stands on the left side with the television monitors and the associated electronic equipment on either side at the head of the table (Fig. 22) 865. Some surgeons prefer the patient in the Lloyd Davies position and stand between the legs to operate. The exposure of Calot's triangle is sometimes improved by elevating the head of the table and rotating the patient to the left.

The Veress needle and the first 11 mm cannula are placed through the umbilicus (Fig. 23) 866. We do not hesitate to adopt an open technique if there is any difficulty and we always do so if there is a paraumbilical hernia as this can be repaired at the end of the operation. The abdomen is inspected as soon as the forward viewing endoscope is inserted although only the gallbladder, the stomach, and the liver are well seen in most circumstances.

Three further ports are needed and they should be placed under direct vision and after careful assessment of the anatomy of the individual patient. An 11 mm port is placed in the epigastrium and the tip should come through the right leaf of the falciform ligament. Two 5.5 mm ports are placed immediately below the right costal margin (Fig. 23) 866. The medial one should complement the epigastric port and lie in the correct position in relation to the operative field whilst the lateral one is placed over the fundus of the gallbladder.

Occasionally in fat patients a fifth port to depress the duodenum and the colon is helpful and then a 5.5 mm cannula is inserted just to the left of the midline halfway between the umbilical and the epigastric ports. Cholangiography catheters and t-tube drains either come with their own special ports or are inserted through separate small incisions placed below the costal margin.

A grasping forceps is placed through the right lateral port and is fixed to the fundus of the gallbladder. The gallbladder and the attached liver are then retracted upwards towards the patient's right shoulder. This opens up the subhepatic space (Fig. 24) 867. Any adhesions to the gallbladder are divided and a second grasping forceps is attached to the neck of the gallbladder and pulled laterally to expose Calot's triangle.

Dissection through the epigastric port with the dolphin nosed forceps (Fig. 25) 868 starts at the junction of the gallbladder with the cystic duct. The peritoneum, fat, and loose areolar tissue around the gallbladder is grasped and gently torn down towards the bile duct. This continues until the inferior aspect of the cystic duct is clearly seen. Any troublesome bleeding should be stopped at once with diathermy but care must be taken not to burn the bile duct. The superior aspect of the cystic duct is cleaned in the same way and eventually the duct is surrounded. Curved forceps can be useful for this (Fig. 26) 869. The cystic duct is therefore clearly identified as a direct extension from the neck of the gallbladder and this is important because the junction of the cystic duct with the common bile duct is not always seen. Most of this dissection is done from the front but by varying the position and the placement of the forceps on the neck of the gallbladder it is often possible to expose the posterior aspect of Calot's triangle. The cystic artery is then exposed in the same way (Fig. 27) 870. Both structures are clipped and divided (Fig. 28) 871. Sometimes the cystic duct must be divided before the cystic artery can be safely dissected. Occasionally the cystic duct is too large to fit inside a clip. It must then be tied off with a Roeder knot.

The dissection is not always easy, particularly if there is a lot of fibrosis from chronic inflammation. The loss of tactile sensation is a disadvantage but there are some substitutes. Significant structures, such as an artery or a duct, contain elastic tissue and will twang if they are tweaked. Fibrous tissue has no such elasticity. As always it is important not to rush a difficult dissection because time often reveals the anatomy. Furthermore the variations in anatomy are no different from an open operation and the same rules therefore apply. No significant structure should be divided until the surgeon and his assistant are completely happy that they both understand the anatomy.

The technique is identical to open cholangiography although it is rarely needed if a policy of preoperative cholangiography is adopted (Fig. 29) 872. A cholangiogram will delineate the anatomy but if this is in doubt it is probably safer to convert to an open operation.

The cystic duct is occluded distally with a clip and a small proximal incision is made in the cystic duct with fine scissors. A catheter is guided into the duct and held there with a special instrument or a loosely applied titanium clip. Air bubbles must be excluded and the ports, which are radio-opaque, should be held out of the way while the radiographs are taken. We usually use a 1.3 mm umbilical catheter inserted through an intravenous cannula (size 14) although specially designed cholangiography catheters each with its own introducer are available.

Once the cystic duct and the cystic artery are divided the free edge of the lesser omentum falls away. The midcostal grasper is repositioned just above the cystic duct clip and it is pulled upwards and away from the liver to expose the plane between the liver and the gallbladder. The gallbladder is then dissected off the liver bed with scissors, diathermy, or a laser beam starting at the neck (Fig. 30) 873. All the bleeding points on the liver bed are coagulated as dissection proceeds until the gallbladder is only just attached to the liver at the fundus. Any bile or blood that has accumulated is sucked away and, if necessary, a drain is placed in the gallbladder bed before this final strand is divided.

Dissection at the fundus is often difficult and simply requires patience and constant changes of position. If the gallbladder is perforated, and this is preferable to entering the liver itself, the bile and as many stones as possible should be removed. Nevertheless stones are often left behind in the peritoneal cavity where, somewhat surprisingly, they very rarely cause complications.

Once the gallbladder is free the cystic duct is drawn into one of the 11 mm ports and the port and the gallbladder are withdrawn together through the abdominal wall. Usually the fundus of the gallbladder and any contained stones remain within the abdomen and they must be manipulated through the wound. Crushing any large stones, sucking away the bile or extending the incision are sometimes necessary and all of these are easier to do if the gallbladder is inside a plastic bag.

A final inspection of the abdomen is appropriate before all the carbon dioxide gas is allowed to escape and the cannulae are removed. We only suture the abdominal wall if a wound has been enlarged and we prefer to close the skin with adhesive tape rather than with sutures.

The best management of bile duct stones in this new endoscopic era is not yet clear. We prefer to remove any stones from the bile duct at an ERCP procedure before embarking on a laparoscopic operation but this may not be appropriate and, sometimes, is impossible.

Large stones, which tend to occur in older patients with grossly dilated bile ducts, always cause difficulty at endoscopy but they are relatively easy to remove at a laparoscopic exploration of the bile ducts. In young patients it is best not to divide the ampullary sphincter and the duct must be explored surgically. The choice lies between a conventional open operation or one of the newer laparoscopic techniques.

Stones can be extracted from the bile duct through the cystic duct if it is first dilated with a suitable balloon. Instruments, including a narrow cholangioscope, can then be passed into the bile duct and the stones retrieved or broken up under direct vision. Stones in the hepatic ducts are difficult to find, particularly if the cystic duct enters the common duct low down behind the pancreas. A fine tube can be left in the cystic duct at the end to act as a drain and to permit later cholangiography. The precise place of this technique in endoscopic practice is not yet clear.

Direct endoscopic exploration of the bile duct is quite straightforward. A 30° forward oblique telescope gives the best view of the bile duct and a supraduodenal choledochotomy is easily made in the anterior surface of the duct with the diathermy hook and scissors. Stones can be washed out of the duct or removed with a balloon or a basket (Fig. 31) 874. The ducts are then inspected with a choledochoscope and any remaining stones are removed. At the end a t-tube is sutured into the bile duct.

If stones are discovered in the bile ducts unexpectedly during the course of a laparoscopic cholecystectomy the surgeon is left in something of a dilemma. Small stones can probably be safely left to pass spontaneously. Larger stones need to be removed either immediately and surgically or at a later ERCP.

All three choices have disadvantages and deciding between them depends on the findings in the individual patient, the equipment available, and the skill and experience of the surgeon. If stones are left behind in the ducts it is probably wise to leave a drain down to the cystic duct stump in case it leaks. An ERCP should be done within a day or two of operation to remove any residual stones and as soon as retained stones are discovered.

Inserting a cannula or the Veress needle can cause spectacular damage and invisible retroperitoneal aortic injury is one cause of an acute collapse during a laparoscopic operation. Such injuries are fortunately very rare.

Sudden severe haemorrhage is always a problem and it is best avoided in the first place. A torn cystic artery or a liver laceration are the common causes and haemorrhage from either can be difficult to stop. The usual general principles apply. Blood and clot are sucked or washed away and the best possible exposure is obtained whilst the surgeon applies local pressure to the bleeding area. The scrub nurse prepares the diathermy and a titanium clip applier and when everything is ready the pressure is removed and the bleeding point is identified and occluded. This is much more difficult to achieve with the limitations on movement in an endoscopic environment and there is sometimes no alternative to a laparotomy.

Every surgeon fears damaging a bile duct. In large published series the accident happens once in every 300 laparoscopic cholecystectomies. This is slightly more frequent than during an open operation and probably reflects the introduction of a new technique rather than a fundamental flaw in the operation. Great care should be taken to identify the anatomy correctly and to place any clips properly. It is very easy for a titanium clip to be placed so that the tips partially occlude the bile duct without the surgeon appreciating what has happened and any doubts about the anatomy should lead to an immediate laparotomy. The next most important aspect of a bile duct injury is that it should be recognized at the time. Late presentation of biliary damage is always associated with a poorer outcome. Exactly what is done depends upon the injury and when it is recognized but a laparotomy will usually be needed.

Reactionary haemorrhage is rare and generally requires a laparotomy although with increasing experience a repeat laparoscopy may be sufficient. Bile leaks from the cystic duct stump or the liver bed present with pain in the right upper quadrant, fever and a subhepatic collection on ultrasound. The bile is removed with a percutaneous drain and the leak is identified at an ERCP and is stopped by placing a stent in the bile duct. The drain remains until the drainage ceases and the stent is removed not less than 2 months later. Surprisingly a simple sphincterotomy will not allow the leak to close. A few patients present with a previously unidentified bile duct stone within a few months of their operation and the stone should be removed through an endoscopic sphincterotomy.

The mortality rate following a laparoscopic cholecystectomy is about 0.1 per cent and the complication rate is about 4 per cent. Both figures are much better than those for an open operation and the cosmetic benefits of four small incisions are important to many patients. At the moment about 1 in every 20 patients needs a laparotomy although we can expect this figure to improve as experience increases. Patients must be warned about this beforehand and sign an appropriate consent form.

Most patients are mobile and can eat and drink later the same day. Some patients are troubled by nausea and vomiting but very few need opiate analgesia. Any drain can be removed after about 12 h and most patients can then go home. The majority are back at work within 2 weeks.

Laparoscopic appendicectomy combines the advantages of diagnosis and treatment in one procedure. The usual preoperative preparation is necessary and prophylactic antibiotics are given on induction of general anaesthesia. The patient must consent to an open operation should it be needed.

The patient lies flat on the operating table and the bladder is emptied. After creating the pneumoperitoneum an 11 mm port and endoscope are placed through the umbilicus and the diagnosis is confirmed. Two further ports are needed on either side of the abdomen and should be placed in relation to the position of the appendix. An 11 mm port is placed in the right iliac fossa and a 5.5 mm port in the left iliac fossa.

Any local adhesions are gently divided and the tip of the appendix is grasped and drawn into the port in the right iliac fossa. The appendix mesentery is occluded with bipolar diathermy or a ligature around the appendicular artery and then divided. The base of the appendix is secured with a Roeder knot and then occluded beyond the ligature with bipolar diathermy. The appendix is divided across the burnt area and is removed through the right iliac fossa port. It is not necessary to bury the appendix stump. Free peritoneal fluid or pus can be sucked away and the peritoneal cavity washed although it is important not to flood infected fluid into the pelvis or the subphrenic spaces. It is easy to place a drain to the appendix stump if necessary and the pneumoperitoneum is then released and the ports removed.

Most patients have less pain from the wounds than after a conventional appendicectomy but the recovery time is only slightly improved. Many patients are toxic from the infection and take time to recover whatever method is used to remove the appendix.

The only complication specific to the laparoscopic procedure is a low incidence of reactionary haemorrhage as a result of failure to occlude the appendicular artery properly. Septic complications occur just as with any other appendicectomy although wound infection is less serious.

Laparoscopic colonic resection is already possible for the techniques involved are no different to those used in a laparoscopic cholecystectomy or appendicectomy. However, tying and dividing the mesentery can be very tedious and the resection may not include all the important lymph nodes. It also takes a long time and considerable skill to construct a hand-sewn intraperitoneal anastomosis and although stapling guns can help they are presently too expensive for routine use. So a complete laparoscopic colonic resection is not really a practical option at the moment although this situation is unlikely to persist as new technological developments are introduced.

Nevertheless endoscopic techniques can be useful during colonic resection in certain patients. Division of the mesentery under laparoscopic vision is fairly straightforward (Fig. 32) 875 and the segment of bowel can be delivered through a small discreetly placed incision, the resection and anastomosis being completed outside the abdominal cavity. The bowel is returned to the abdomen and any necessary drains are placed before all the ports are withdrawn and the incision closed. This approach minimizes the trauma to the patient and overcomes the problem of removing the specimen from the abdominal cavity.

Once the ideas and the concepts of an endoscopic operation are accepted it takes very little imagination to transfer almost every operation to an endoscopic procedure. Indeed most abdominal operations have already been done laparoscopically somewhere in the world.

Operations around the stomach, such as fundoplication and pyloromyotomy in neonates, are relatively easy to do although there is little demand for vagotomy. Small bowel resections are mostly required during emergency operations and although endoscopic techniques are described most surgeons will be naturally hesitant to embark on such procedures at the moment. The same remarks apply to splenectomy. Liver resections are clearly impractical at present but it is easy to deroof a liver cyst with a diathermy hook and bleeding from a liver biopsy site can often be controlled with a laparoscopic technique, thus avoiding a laparotomy. In both instances it is quite possible to suck away blood or other fluids lying in the peritoneal cavity.

Emergency laparoscopic surgery is still being developed. A perforated anterior duodenal ulcer is easy to close with an omental patch through the laparoscope as any surgeon who has done a laparoscopic cholecystectomy will appreciate. Laparoscopy for intestinal obstruction is difficult because of the distension but it is possible to decompress the bowel and then to divide an adhesive band. Some reports have also suggested that laparoscopy and laparoscopic surgical techniques are useful in patients with abdominal trauma.

This is the only laparoscopic procedure to remove a retroperitoneal organ although the kidney is approached across the peritoneum. The operation can take some time to do and at the moment it is probably best reserved for benign diseases of the kidney.

The patient is placed in the traditional lateral position and the endoscope is inserted at the lateral border of the rectus abdominis muscle level with the umbilicus. Three or four further ports are needed. The peritoneum on the lateral side of the colon is divided and the bowel falls away under gravity. The ureter is identified at the pelvic brim and traced up to the hilum of the kidney. The vascular pedicle is divided with a special stapling gun which inserts six rows of fine vascular staples and divides the tissues down the middle all in one action. The kidney is then freed from its surrounding attachments and removed from the abdomen within a laparoscopic organ bag.

This is a straightforward day-case procedure that divides the testicular veins just above the internal inguinal ring. The endoscope is inserted through the umbilicus in the usual way and two 5.5 mm ports are required on either side of the umbilicus. The testicular vessels are easily seen through the peritoneum just inside the internal inguinal ring and it is a simple matter to divide the peritoneum with scissors, dissect out the testicular veins from the testicular artery, clip the veins, and excise a segment (Fig. 33) 876. Even if the testicular artery is inadvertently included in a clip the testicle should still survive provided that the branches to the spermatic cord from the inferior epigastric artery are not damaged.

Peritoneal dialysis is widely used in the long-term treatment of chronic renal failure but the Silastic catheter is easily blocked by displacement of the tube out of the pelvis or with omentum wrapped around the side holes. It is easy to insufflate the abdomen through the catheter, remove any omentum adherent to the tubing, and to replace the coiled tip in the pelvis using the laparoscope and a suitable grabber passed though one 5.5 mm port. Transplanted kidneys are sometimes surrounded by collections of lymph (lymphocele) and if percutaneous drainage fails fenestration of the lymphocele into the peritoneal cavity is the best method of treatment. It is easy to cut a hole in the peritoneal wall of a lymphocele with a diathermy hook (Fig. 34) 877. In both instances the benefit to the patient of avoiding a laparotomy is obvious.

In some respects endoscopic surgery has not advanced so rapidly in gynaecology. This is partly because transcervical resection of the endometrium has reduced the need for hysterectomy but also because a laparoscopic hysterectomy, which is perfectly possible, is regarded by some gynaecologists as a rather more complicated version of a vaginal hysterectomy. Endoscopic operations on the fallopian tubes and the ovaries certainly have a place and are particularly valuable in preserving tubal function after the removal of an ectopic pregnancy for example. Laser ablation of endometriosis and the division of pelvic adhesions have been standard for many years and pelvic lymphadenectomy, which is a popular procedure in the United States, is an ideal operation for the endoscopist. For a detailed description of all these operations the reader is referred to a specialist text.

Most inguinal hernias are repaired through a groin incision and approach the inguinal canal by division of the external oblique aponeurosis. The posterior or preperitoneal approach is also popular but it still involves an incision through the abdominal wall muscles. Either operation is decidedly uncomfortable and very few patients are back at work in less than a month. A nylon darn is the most popular method of repair but synthetic mesh is also used and seems to be particularly successful when it is placed in the preperitoneal plane. Laparoscopic hernia repair avoids the inguinal incision and combines the advantages of plastic mesh and the preperitoneal approach. As in paediatric surgery, where endoscopic hernia repair is not yet established, laparoscopy also identifies the patient with a hernia on the other side.

We describe below a transabdominal extraperitoneal mesh repair of an inguinal hernia which is one technique among many. We do not yet know which is the best, but it is certain that the operative anatomy is strange, the access is awkward, and the iliac vessels are alarmingly close to the operation site.

The basic objective of a laparoscopic inguinal hernia repair is to deal with the peritoneal sac and then to staple a sheet of plastic mesh over the muscular defect. A direct sac and a small indirect one can normally be easily pulled back into the abdominal cavity. A large indirect sac is best transected at the internal inguinal ring and the residual peritoneum left in the inguinal canal. The peritoneum must also be removed from a femoral hernia but it is then sufficient to plug the femoral canal with a roll of plastic mesh.

General anaesthesia with good relaxation of the abdominal wall is required and all our patients are given prophylactic antibiotics. The patient lies flat on the table although a little head-down tilt is sometimes helpful. A 10 mm endoscope is inserted through the umbilicus and the operation site inspected. Two further ports are needed, one on each side of the umbilicus at the lateral edge of the rectus abdominis. A large 12 mm port to take the stapler is placed on the side of the hernia and a 5.5 mm port on the other side (Fig. 35) 878.

An indirect hernia is very different in appearance from a direct one. The direct sac is a wide necked defect straight through the anterior abdominal wall flanked on the lateral side by the inferior epigastric vessels and medially by the umbilical ligament. An indirect hernia clearly runs obliquely through the abdominal wall muscles and the end of the sac cannot usually be seen. The testicular vessels are visible in the inferolateral corner of the sac at the internal inguinal ring whilst the vas appears as a white cord slipping out of the inferomedial corner and running down over the iliac vessels into the pelvis (Fig. 36) 879.

The operation starts well lateral and superior to the hernia sac by dividing the peritoneum. The peritoneal incision continues horizontally across the top of the internal ring and up to, and sometimes into, the umbilical ligament. The peritoneum is then dissected downwards and is thus pulled out of the hernia although sometimes the sac is best circumcised at its neck.

The testicular vessels and the vas are often rather adherent and the dissection can be somewhat difficult at this stage. The abdominal wall muscles and the inferior epigastric vessels should then be cleaned of any fatty tissue.

At this point the anatomy of the whole area should be apparent. Medially the pubic tubercle should be palpable with Cooper's ligament running away posteriorly and laterally. Anteriorly and laterally the posterior aspect of the conjoint tendon should be visible spreading behind the inferior epigastric vessels. The inguinal ligament is rarely identified. A 10 cm by 5 cm sheet of Prolene mesh is rolled up and passed into the peritoneal cavity through the 12 mm port. It is unravelled, manipulated to cover the margins of the hernia defect, and stapled in place. Medially the inferior margin of the mesh can be stapled to Cooper's ligament but laterally great care must be taken not to damage the testicular vessels, the vas, and the femoral nerve. An angled stapling gun helps to place the staples correctly (Fig. 37) 880.

Finally the original peritoneal layer is lifted back up over the mesh and stapled back to the peritoneum on the anterior abdominal wall. Deflating the abdomen can make this easier. The ports are then removed and the muscles at the site of the 12 mm port are sutured if necessary.

Most patients are up and about later the same day and many of them can go home. There is no reason to restrict their activity in any way as the repair does not depend on muscles healing together and although most patients experience some discomfort in the groin they do not need much analgesia. Most patients are back at work within 2 weeks.

One alarming development is gross gaseous distension of the scrotum at the end of the operation. Fortunately the carbon dioxide is rapidly absorbed with no serious consequences. An inguinal haematoma is a nuisance and damage to the femoral nerve is a cause of persistent pain in the wound. Infection around the mesh is rare. So far as recurrence is concerned this type of laparoscopic operation appears to have recurrence rates very comparable to the open operation, but only time will tell if the conjecture is true.

All the conventional open approaches to the cervical sympathetic trunk are unsatisfactory, either because of poor access and inadequate exposure, or the proximity of important structures. In contrast, the endoscopic operation is easy and safe and the surgeon has an excellent view of the sympathetic trunk. The procedure itself is simple, elegant, and quick.

A pneumothorax with carbon dioxide is created with the patient fully anaesthetized and lying flat on the operating table with the arms abducted to 60°. A 5 mm endoscope is introduced into the chest in the fourth intercostal space at the anterior axillary line and the cervical sympathetic trunk is immediately visible under the parietal pleura on the necks of the 2nd, 3rd, 4th, and 5th ribs (Fig. 38) 881. The stellate ganglion is sometimes seen superiorly. Through a separate port in the fifth intercostal space it is a simple matter to divide the sympathetic trunk just below the stellate ganglion and either to excise a segment or obliterate the chain with diathermy over the second, third, and fourth ribs. A special port with a channel for the endoscope and one for the instruments is available and means that the operation can be done with only one puncture. Both sides can be done at the same time and most patients can go home the next day.

Any technique that avoids a painful thoracotomy is to be welcomed and the new stapling instruments make any form of endoscopic pulmonary resection relatively easy although the specimen has to be removed within an organ bag.

Mediastinal operations and oesophageal resections are also feasible. There are two approaches to oesophageal resection. In one technique the abdominal and cervical dissections take place through conventional incisions but the thoracic oesophagus is mobilized endoscopically across either pleural cavity. The oesophagus and stomach are pulled up into the cervical incision, the tumour resected, and then continuity is restored.

An alternative approach is to mobilize the oesophagus within the mediastinum by dissection under endoscopic vision using a specially designed instrument which is inserted through a cervical incision and slowly advanced down the mediastinum alongside the oesophagus. The stomach is mobilized laparoscopically and, once again, everything is drawn up into the neck where the resection and the anastomosis are performed.

Endoscopic techniques are already widely used in other surgical specialties that we have not discussed. Arthroscopy has revolutionized operations on the knee in orthopaedics, for example, and nasal operations are much more accurate when viewed through an endoscope. Every branch of surgery will be influenced in due course and even now fine endoscopes are being made that can be inserted through a tiny burr hole into the subarachnoid space to examine the surface of the brain. It will not be long before neurosurgical operations are possible without the need for a craniotomy.

New operations and new ways of performing old operations are being developed with astonishing speed and endoscopic surgery is still in its infancy. Many future developments will depend on advances in technology but they will also depend on surgical skill. It is also of critical importance that proper evaluation of these developments takes place continually, not only in assessing the quality of the results of endoscopic surgery, but also the cost effectiveness of these procedures.

Alain JL, Grousseau D, Terrier G. Extramucosal pyloromyotomy by laparoscopy. J Pediatr Surg 1991; 26: 1191–2.

Barkun JS, et al. Randomised controlled trial of laparoscopic versus mini cholecystectomy. Lancet 1992; 340: 1116–9.

Byrne J, Walsh TN, Hederman WP. Endoscopic transthoracic electrocautery of the sympathetic chain for palmar and axillary hyperhidrosis. Br J Surg 1990; 77: 1046–9.

Coptcoat MJ, Wickham JEA. Laparoscopy in urology. Min Invas Ther 1992; 1: 337–42.

Cuschieri A, Berci G. Laparoscopic biliary surgery. Oxford: Blackwell Scientific Publications, 1990

Cuschieri A, et al. The European experience with laparoscopic cholecystectomy. Am J Surg 1991; 161: 385–7.

Cuschieri A, Buess G, Perissat J, eds. Operative manual of endoscopic surgery. Berlin: Springer-Verlag, 1992.

Dandy DJ. The present state of arthroscopy. Min Invas Ther 1991; 1: 51–6.

Gotz F, Pier, A, Bacher C. Modified laparoscopic appendectomy in surgery. A report on 388 operations. Surg Endosc 1990; 4: 6–9.

Holohan TV. Laparoscopic cholecystectomy. Lancet 1991; 338: 801–3.

Hunter JG. Laparoscopic transcystic common bile duct exploration. Am J Surg 1992; 163: 53–8.

Larson GM, et al. Multipractice analysis of laparoscopic cholecystectomy in 1983 patients. Am J Surg 1992; 163: 221–6.

Loh A, Taylor RS. Laparoscopic appendicectomy. Br J Surg 1992; 79: 289–90.

Macintyre IMC. Laparoscopic herniorrhaphy. Br J Surg 1992; 79: 1123–4.

Martin IG, et al. Laparoscopic cholecystectomy as a routine procedure for gallstones: results of an ‘all-comers’ policy. Br J Surg 1992; 79: 807–10.

Masters A, Rennie JA. Endoscopic transthoracic sympathectomy for upper limb hyperhidrosis. Min Invas Ther 1992; 1: 325–8.

Miller SS. Laparoscopic operations in paediatric surgery. Br J Surg 1992; 79: 986–7.

Myers WC, et al. A prospective analysis of 1518 laparoscopic cholecystectomies. N Engl J Med 1991; 324: 1073–8.

Paterson-Brown S. Emergency laparoscopic surgery. Br J Surg 1993; 80: 279–83.

Sackier J, Berci G, Phillips E, Carroll B, Shapiro S, Paz-Partlow M. The role of cholangiography in laparoscopic cholecystectomy. Arch Surg 1991; 126: 1021–6.

Sutton C. Operative laparoscopy. Br J Hosp Med 1993; 49: 312–27.

Wilson RG, Macintyre IMC, Nixon SJ, Saunders JH, Varma JS, King PM. Laparoscopic cholecystectomy as a safe and effective treatment for severe acute cholecystitis. Br Med J 1992; 305: 394–6.