Rationale
It has been shown that often the prognosis following the complete resection of a tumor that invades an adjacent organ is similar to that of a tumor which simply breaches the serosa. In gastric cancer, Bozzetti et al. examined the outcome of 143 tumors with invasion of an adjacent structure. The 5-year survival following complete resection of a pT4 N0 tumor was 29%, whereas this was 21% for pT3 N0 tumors.3 A possible explanation might be that extension into a neighboring structure prevents shedding of cancer cells from the serosal surface and peritoneal carcinomatosis. Local invasion is certainly less ominous than loco-regional or distant spread. In the series of Bozzetti et al. lymph node involvement had a quite substantial negative impact on survival-the 5-year survival for pT3 N1 and pT4 N1 tumors were respectively only 2 and 5%. This is analogous to what has been shown in nonsmall cell lung cancer where much larger series are available. The 5-year survival following complete resection of a pT3 N0 tumor with chest wall invasion is 40-50%, which is similar to that of a pT1/2 N1 lesion and considerably better than the only 25-40% 5-year survival of patients with true minimal N2 disease. The 5-year survival of patients with completely resectable (as opposed to true minimal) N2 disease is only 9%.
Indications
Intraoperative Discovery
Despite careful preoperative evaluation the intraoperative staging, which is the mandatory first step in any cancer resection, will show invasion into an adjacent structure in a certain number of cases. There are areas where the normal close approximation of structures can make it impossible to accurately assess juxtaposition and invasion by CT or MRI. Again this situation is much more codified in thoracic surgery than in general surgery. In thoracic oncology (nonsmall cell lung cancer) it is commonly thought that in an ideal world the surgeon should have a 5% “open and shut” rate, i.e., cases where the intraoperative staging has shown a more advanced stage than predicted leading to the decision not to proceed with the planned resection. This prevents some patients being denied curative surgery due to over-interpretation of the preoperative imaging. A higher rate denotes an excessively cavalier attitude on the surgeon’s behalf. In the abdomen these areas of juxtaposition vs. invasion, which can be delicate to assess on imaging, are the bile duct and the portal vein, the gastric antrum and the head of the pancreas, and the unciform process of the pancreas and the confluence of the mesenteric and portal veins. If this situation is anticipated or encountered it is essential that the patient should be cared for by a surgeon who feels comfortable performing these complex operations. Otherwise the patient should be referred to a tertiary referral center for the resection. What is essential is that the resection should be complete and that during the dissection the areas of tumor adhesion must not be meddled with surgically but encompassed by adequate resection margins.
Surgery after Induction Therapy
The term primary (chemo-, radio-, or chemo-radio) therapy implies that the object is to downstage the tumor so as to render an irresectable tumor resectable. Neo-adjuvant therapy is intended to improve the 5-year survival of a technically operable tumor which has an inherently poor prognosis. Induction therapy comprises both primary and neo-adjuvant therapy. However, the terms are often used interchangeably.
When primary tumors of the esophagus,4–6 the stomach,7–11 the ovary,12–14 the rectum,15,16 or the pancreas are irresectable at the time of presentation induction, therapy is often considered to try to downstage the tumor.17–19 If there is a complete or partial response resection can be considered. Again it is worthwhile to remember that the definition of a partial response in many protocolsis a reduction in size of >50% for the sum of the two greatest diameters. In some cases a high (fivefold, for example) reduction of the SUV on the PET scan can be used as a surrogate marker for tumor response to induction therapy. If the dissection planes between the organ of origin of the tumor and the involved second organ are not clear cut and clearly tumor free then a multiorgan resection should be performed.
In every case these operations require careful planning so that an appropriate team and instrumentation are available at the time of surgery. If a vascular substitute may be required (greater saphenous vein, internal jugular vein for example) the patient is positioned, prepped, and draped in preparation for this after ensuring that the anesthesiologist has not attempted to put lines into the vessel. Likewise at times ureteral catheters will need to be placed by cystoscopy prior to starting the operation itself.
Induction therapy followed by surgery is a potentially morbid combination that requires fit patients. The upper age in most trials is usually 70 years and this is worth remembering even outside of a trial setting. Even more than for multiorgan resections alone, patient selection is of paramount importance to obtain acceptable results.
Tumors which are Poorly Responsive to Chemo- and Radiotherapy
Some tumors, such as bile duct adenocarcinomas and retroperitoneal sarcomas do not respond to induction therapy in a satisfactory or reliable manner. This expose the patient to a significant risk of tumor progression during induction therapy, which could mean progression to a completely irresectable stage. These are always very difficult situations to deal with both in terms of treatment decisions and the technical aspect of the resection itself. These are the operations which are liable to present the greatest technical challenges and therefore require the highest amount of preoperative planning and preparation. It is also essential that during the consent process the patient (and his/her family) is made aware not only of the risks of surgery, but also of the not insignificant risk of finding an irresectable tumor at the time of surgery which mandates retreat without resection (the so-called “open and shut” case).
Conditions
Below is the check-list of conditions which must be fulfilled for a patient to be a candidate for a multiorgan resection.
The Patient’s General Condition
These are long, complex, occasionally hemorrhagic, sometimes staged procedures which put considerable physiologic stress on the patients. A careful and complete history is the mandatory first step in the evaluation of the patient and it is essential to quantify or rule out major organ dysfunction. Obviously routine blood work will be obtained to assess liver and kidney function. The threshold to obtain specialized investigations such as full pulmonary function tests, echocardiograms, thallium scans, or even coronary angiograms should be low. The overall performance score should be assessed. We use the Eastern Cooperative Oncology Group (ECOG) score which goes from 0 (no limitations whatsoever) to 4 (fully bed-bound).20 Candidates for these complex operations will ideally all be in categories 0-2.
Distant Metastases
Appropriate investigations must be performed to rule out distant metastases. These will include CT of the chest, abdomen and pelvis, MRI of the brain (or at least CT if MRI is not readily available), PET scan if appropriate. If a PET scan is performed a bone scan is not necessary (PET has both higher sensitivity and specificity than the bone scan for bone metastases if the primary tumor is FDG avid). As a general rule the presence of distant metastases is an absolute contraindication to multiorgan resection. However, there is at least one report of thermoablation of limited liver metastases with radiofrequency in association with the resection of abdominal tumors.21 This this is still controversial though and should only be considered in the setting of a prospective trial.
Peritoneal Carcinomatosis
There are at least two scoring systems for peritoneal carcinomatosis. Sugarbaker has described the Peritoneal Cancer Index.22 The abdomen is divided up into 13 regions and a score from 0–3 is attributed to each one (0 = no tumor; 1 = tumor 0.5 cm; 2 = tumor 5 cm; 3 = tumor >5 cm or confluent tumor nodules). The total score can thus be between 0 and 39. This score can be used to predict the probability of complete tumor cytoreduction.
The Japanese Research Society for Gastric Cancer has divided peritoneal carcinomatosis into three stages23: stage P1 is limited to peritoneal metastases on the adjacent peritoneum, stage P2 denotes a few scattered distant peritoneal implants, and stage P3 means there are multiple distant peritoneal tumor nodules. They have reported the expected prognostic implications of increasing P stage following gastrectomy with wide systematic nodal dissection—with P1 carcinomatosis the median survival was 21.7 months. This figure was respectively 10.4 and 12.8 months for stage P2 and P3 disease.24
Diffuse peritoneal carcinomatosis precludes multiorgan resection. In favorable cases surgery might still be considered if there is very limited and completely resectable stage P1 disease in a very fit patient.
Peritonectomy and its different aspects are an important topic. That is why Chapter 13 is devoted to this subject, even if these techniques are not in the strict sense part of the topic of multiorgan resections. The evolution (reduction or disappearance) of peritoneal carcinomatosis following induction therapy can influence the decision regarding multiorgan resection. The findings at diagnostic laparoscopy before and after induction therapy will allow the peritoneum to be assessed and the final decision to be made.25
Predicted Survival
Multiorgan resection must always be done with curative intent. However, as shown by Martin et al. it is more the T and N stage that are the predominant predictors of survival rather than the number of resected structures.26 In their series of 1133 gastrectomies for cancer, 286 patients underwent the resection of an adjacent organ. The median survival following gastrectomy alone was 63 months, whereas it was 32 months for gastrectomy in association with the resection of an adjacent organ. However, the authors reported that this was due more to the higher N stage of this second category of tumors than the multiorgan resection per se.
Peritoneal carcinomatosis has an impact on the survival following multiorgan resections: 5-year survival can be achieved for ovarian cancer but does not extend beyond 2 years for gastric cancer.
Requirement for a Stoma
Quality of life issues must also be taken into account in determining the indication for these resections. Especially in the lower abdomen these operations can require one or, at times, even two stomas to be constructed. These will influence the life style of patients in 80% of cases and their sexual activity will be curtailed in 40% of cases.27 An ileostomy can allow better quality of life than a colostomy and this should be taken into account in deciding the best operative strategy.28 Most centers that will be performing this type of surgery will have specialized stoma nurses who can have a considerable impact on lessening the negative impact of urinary or digestive stomas on the patient’s quality of life.29
Paralyzing Sequelae
Pelvic and sacral resections have the risk of causing paresis or paralysis of the lower limbs as well as anorectal and/or urinary dysfunction, not to forget sexual dysfunction.
The resection of a sarcoma from the iliac fossa puts the femoral nerve at risk of temporary or permanent injury. This will limit flexion of the thigh. Usually the patient can get used to this limitation and with time the impact becomes fairly limited.
Resection of sacral nerves disturbs urinary, anorectal, and sexual function. If these are below the level of S3 the functional consequences are usually quite limited. Unilateral resections of the S1–S3 roots cause minor urinary dysfunction. Bilateral S1–S2 resections will often submit the patient to the requirement for occasional self-catheterization.
Chronic Lower Limb Edema
Preoperative lower limb edema is usually the harbinger of extensive retroperitoneal involvement and thus implies a poor prognosis. The indication for surgery must be evaluated very carefully and every effort must be made to image the retroperitoneal extension of the tumor. MRI can at times be helpful in this context. If extensive retroperitoneal involvement is found, palliative therapy in association with appropriate physiotherapy should be offered.30
Retroperitoneal Tethering
Extensive retroperitoneal tethering usually represents a contraindication to surgery because it often means that the celiac axis or the root of the mesentery is infiltrated. Tumors of the body of the pancreas are the most common example of this condition. That is why these tumors are only resectable in one quarter of all cases.31 These tumors often invade the celiac axis, the retroperitoneal lymphatics, and the celiac plexus early in their course.
Diagnostic Laparoscopy
Laparoscopy is one of the most powerful staging tools available today. It can allow the extent of the tumor to be ascertained as well as tethering, loco-regional lymph nodes can be sampled for histologic verification, the peritoneum can be inspected for distant spread, and parts of the liver can be assessed for metastases. It can assist in the determination of the indication for surgery and/or induction therapy.
Esophagus
Transesophageal ultrasound with ultrasound guided fine needle aspiration biopsies (FNAB) has made the preoperative staging of the esophagus much more precise and in many cases offers the required information to make the appropriate management decisions (however transesophageal ultrasound alone without FNAB is an insufficiently precise tool to make a decision on suitability for resection). There are, however, cases where it will not resolve all the staging issues in a satisfactory manner. In these cases minimally invasive staging can be performed with laparoscopy and/or thoracoscopy. The sensitivity of thoracoscopy for mediastinal nodal staging could be as high as 93%. The same sensitivity (>93%) is seen for celiac axis nodes sampled by laparoscopy.32 Thus, minimally invasive surgical staging can be precious in determining the indication for trimodal therapy.33,34
Stomach
Laparoscopy is better than CT and ultrasound in determining the precise stage of gastric carcinomas.35 Diagnostic laparoscopy is performed in a reverse TNM sequence. Firstly, distant metastases (peritoneal) are ruled out, the loco-regional nodes are sampled and finally the tumor itself is assessed.36 When combined with laparoscopic ultrasonography the sensitivity of the method is greater than 95%.37,38 On the other hand the technical resectability cannot be determined with the same level of accuracy.39
Pancreas
Laparoscopy will allow liver and/or peritoneal metastases which were not detected to be seen, avoiding an unnecessary laparotomy. On the other hand the true resectability was accurately assessed in only 35% of cases. This is why staging laparoscopy cannot be considered to be a routine staging procedure for carcinoma of the pancreas at this time.40,41
References
2. Gross GE. The role of the tumor board in a community hospital. CA Cancer J Clin. 1987;37:88–92.