Selection Criteria for Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy With Special Emphasis on Laparoscopy as an Efficient Tool

Abstract

Background/Aim: The aim of the study was to assess diagnostic tools implemented in selecting candidates for cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) with special emphasis on diagnostic laparoscopy.

Patients and Methods: A total of 229 patients underwent laparoscopy between January 2011 and December 2015 with peritoneal carcinomatosis and were analyzed retrospectively to study the role of laparoscopy in selecting patients for CRS and HIPEC.

Results: After initial laparoscopy, 101 (44%) of patients underwent CRS and HIPEC with 128 (56%) being aborted. The most prevalent overall exclusion criterion was extensive small bowel disease with a prevalence of 107 (47%) patients. The rate of complete cytoreduction (CCR0 and CCR1) in patients who were deemed initially operable during laparoscopy was 96%. The most prevalent primary tumor site was the stomach with a prevalence of 36,7%. The overall median peritoneal cancer index (PCI) score of the excluded patients were 16.8 and 4.6 in the CRS and HIPEC group. Port-site metastasis was 3.1% and bowel perforation rate 3.0%.

Conclusion: Laparoscopy is a proper selection tool in patients with peritoneal metastases for multimodal treatment. Sensitivity for detecting non-resectable patients is high and the morbidity and mortality low.

Peritoneal carcinomatosis (PC) is generally considered as palliative situation when extensive spread is present and has a mean survival of 2-6 months (1). There is however strong evidence for prolonged survival and in well selected cases a means of a cure through a multimodal therapy for selected patients consisting of cytoreductive surgery (CRS) followed by hyperthermic intraperitoneal chemotherapy (HIPEC). Cytoreductive surgery involves complete removal of all visible tumor mass. Following resection, the heated chemotherapeutic agent can be administered in high concentrations locally without the systemic side effects and destroys remaining microscopic tumor cells in the abdominal cavity.

The peritoneal cancer index (PCI) and completeness of cytoreductive score (CC-score) are used as assessment tools of the operative site before and after surgical resection and correlates with long-term prognosis (2). A crucially important precondition for achieving macroscopically complete or nearly complete cytoreduction (CC0/CC1) is the appropriate patient selection prior surgery. Tumor location and distribution has a strong impact on the resulting CC-score. Diffuse tumor spread of the small bowel and mesentery is the main limiting factor for completion of CC0 resections and concomitantly the most frequent cause for suboptimal resection (1).

Recently diagnostic laparoscopy has been widely used to assess intra-abdominal tumor extent and distribution in order to select the appropriate candidates for CRS and HIPEC, and at the same time to reduce the ratio of nontherapeutic laparotomies. However, because not all patients with peritoneal metastasis can be considered for diagnostic laparoscopy, a preliminary selection based on radiological imaging techniques is necessary. Therefore, there is more of a complementary multidisciplinary patient selection chain involving radiologists, medical oncologists, gastroenterologist and oncologic surgeons. Indeed, the addition of diagnostic laparoscopy to computer tomography (CT) showed a clinically relevant reduction of "open-and-close" frustrating laparotomy in a recent retrospective study (3).

The current study will highlight the positives and negatives with each diagnostic tool in the patient selection, emphasizing on the laparoscopy and its role in selecting patients with the presentation of our experience with its implementation in our high-volume tertiary center.

Patient selection and instruments. This is crucial for the achievement of a survival benefit in selected patients with peritoneal metastases. Basically, selection includes treatment related factors for assessment of chance to achieve a complete surgical cytoreduction via CT, MRI, PET-CT, and laparoscopy as well as center qualities (expertise, certification, register study). Furthermore, tumor related factors as tumor primary, PCI, histology, and response on systematic chemotherapy. Finally, patients related factors like age, performance status, motivation and expected quality of life after surgery.

However, PC cannot be considered as one disease entity with correlating prognosis, biology, and response, therefore patients will not profit equally from the treatment. It´s nevertheless clear that the likelihood of prolonged survival, regardless of the tumor histology is associated with complete cytoreduction. This poses the importance of tumor morphology and spread in patient selection (4). Age is considered a relative contraindication based on the general condition of patient and is not reported to have a higher incidence of severe morbidity and mortality (5).

Many different diagnostic tools can be used for patient selection. Currently the general consensus is the use of contrast-enhanced multi-sliced computer tomography (MDCT), magnetic resonance imaging (MRI), positron emission tomography (PET), serum tumor markers and laparoscopy (6).

Computed tomography. There are different imaging modalities to evaluate the extent of peritoneal carcinomatosis, however computed tomography (CT) is generally considered as the modality of choice. Qualities that make it popular are fast image acquisition time, high spatial resolution, less movement artifacts caused by bowel peristalsis and good accessibility (7). The advancement of multiple detectors has increased the sensitivity and accuracy of the CT scan to detect peritoneal carcinomatosis, which varies from 60% to 90% (8,9). The detection rate of nodules greater than 5 cm is considered 60% to 90%, contrary to nodules smaller than 1 cm that showed a 10% to 30% chance of detection (6,10,11). Regions afflicted with small sized PC with low contrast flow such as small bowels, mesentery, and porta hepatis have low detection rates and in case of dissemination a negative outcome for completion of cytoreductive surgery with resulting poor prognosis. The value of preoperative CT-imaging is essential in oncologic surgery, however, presents limitations in selection of patients when used as a single diagnostic tool (12).

MRI/DWI-MRI. Magnetic resonance imaging (MRI) is not as widely used as the CT for abdominal imaging, although its superior contrast resolution offers a very precise evaluation. MR imaging has a better sensitivity than CT for detecting tumor less than 1 cm (85%-90%) and overall for all nodules (84%) (13).

The DWI-MR (diffusion weighted imaging-magnetic resonance) has proven itself incredibly valuable in staging of abdominal and pelvic tumors, particularly in otherwise difficult locations such as bowel serosa, mesentery, perihepatic and peripancreatic region. Fujii et al. reported a sensitivity and specificity of 90% and 95.5% (14). The technique does not require extra scanner hardware, excessively prolonged scanning times or intravenous contrast material administration and can be easily incorporated into already existing MR imaging protocols (10,15). Disadvantages of the MR-imaging approach are generally long imaging time, high costs, contraindication with certain implants and lack of availability.

FDG-PET/CT. This imaging technique combines a positron emission tomography scanner and a computer tomography scanner, acquiring sequential imaging from both in the same session resulting in superimposed images (13). Unlike the MDCT and MRI imaging methods, the major advantage of the positron emission tomography-CT (PET-CT) is the coverage of the whole body. Schmidt et al. detected metastatic disease, not been seen on the MRI or MDCT in 20% of patients undergoing PET-CT (16). Pfannenberg et al. showed a significant improve of the accuracy of staging compared with MRI or MDCT alone (11). Published data on PC detection through FDG-PET/CT is limited, however sensitivities of 78%-100% have been reported (13). Others, such as Schmidt et al. showed PET/CT having a slightly lower sensitivity for detecting PC than MDCT and MRI (16). A proposed cause might be due to physiologic respiratory and bowel movements during the acquisition time which can cause blurring artifacts, leading to an either false-negative or false-positive results (16). A common obstacle of the widespread use of PET/CT is the expensive and difficult to produce radiopharmaceuticals used for PET imaging, such as FDG (16).

As of now, there are no clear guidelines outlining the use of imaging in PC, however the limitation of the various modalities is known. Even if PET/CT or DWI-MRI might be more accurate than the MDCT, the direct visualization through laparoscopy or laparotomy remains the gold standard in final therapeutic planning. These imaging modalities can however facilitate patient selection for CRS and HIPEC and remain a vital tool.

Tumor markers. Tumor markers may be used as indicators targeting patients with probable unfavorable prognosis who potentially could benefit from a perioperative therapy. Their most common use is however in the follow-up phase as indicators of disease recurrence (12).

Laparoscopy. One of the major problems of handling patients with PC is acquiring correct preoperative assessment of the extent and location before engaging the CRS and HIPEC, a procedure associated with a severe morbidity rate (Clavien-Dindo Grade III and IV) of 12-52% and a mortality rate ranging between 0.9% and 5.8% (17).

Studies show that as many as 20% to 40% of patients who underwent imaging are excluded at direct exploratory laparotomy during the planned procedure (18,19). Nontherapeutic laparotomies that are associated with significant risk of morbidity and mortality, which in turn lead to delay in initiation of systemic therapy, can be avoided with the use of diagnostic laparoscopy excluding patients deemed inoperable (20). However, these unwarranted laparotomies still affect approximately 40% of all patients currently (18). Esquivel et al. studied patients with advanced disease who underwent exploratory laparotomy as single procedure and reported morbidity rates of 12-23% (21).

Application of diagnostic laparoscopy prior to resection has been shown to be effective in excluding non-operable disease where radiological information is not conclusive. Yet its success rate in completely assessing the abdomen in patients with PC, which may have a hostile abdomen and adhesions, remains a hot topic.

The diagnostic laparoscopy is a minimally invasive surgery, enabling an access to the abdominal cavity for staging. Other applications are diagnosis of carcinomatosis of unknown origin by taking biopsy, restaging after neoadjuvant therapy, and restaging during follow up in case of suspected recurrence (22). The benefits of the procedure compared to the laparotomy are shorter length of stay (LOS), shorter time to recovery with less pain and less morbidity. There are however possible factors of concern regarding the laparoscopy.

One factor might be port-site metastasis (PSM). Most studies agree that the appearance of PSM should be assessed as a negative prognostic factor, suggesting aggressive tumor biology, with a decreased overall survival (22-25).

The advanced stage of disease is associated with excessive adhesions, leading to a potentially inaccurate assessment or increased risk of complications such as visceral perforation. We investigated the safety and feasibility of the diagnostic laparoscopy for selecting potential candidates for CRS and HIPEC.

Patients and Methods

We included 229 patients with known or suspected PC undergone diagnostic laparoscopy between January 2011 and December 2015. Tumor, when known, had to be deemed locally resectable. The current study includes anonymized prospectively gathered patient data. Informed consent was obtained from all participants included in this study.

PCI was used to evaluate the distribution and volume of PC. It quantitatively combines the extent of disease throughout 13 abdominopelvic regions by scoring the size of lesions. Two sagittal and two transverse planes divide the abdomen into 9 regions. These are numbered in a clockwise fashion, with 0 at the umbilicus and 1 defining the quadrant under the right diaphragm. The other 4 define the small bowel, with 10 and 11 being upper and lower jejunum, followed by 12 and 13 being upper and lower ileum. The regions are scored after the largest measured tumor implant. The lesions size (LS) is scored respectively; LS-0 indicates no visible implants, LS-1 indicates lesions up to 0.5 cm, LS-2 indicates lesions up to 5 cm, LS-3 indicates lesions above 5 cm or confluence of PC. These regions are summated and a score between 0 and 39 is given (26). The PCI is not only useful for estimating the likelihood of completion of cytoreduction, but also as a prognostic indicator. Sugerbaker et al. showed this by a 5-year survival study on patients with peritoneal metastasized colon carcinoma. Patients with PCI <10 had a 5-year survival of 50%, however patients with a PCI over 20 had a 5-year survival of 0% (27).

Diagnostic laparoscopy was performed almost exclusively with an open technique, otherwise with the Veress needle. Patients were under general anesthesia in the supine position. An approximately 2 cm incision was made in the umbilical region, with subsequent exposition and opening of the abdominal fascia. The underlying parietal peritoneum was carefully dissected, and the peritoneal cavity entered. The abdominal wall was explored with a finger in 360 degrees for adherent visceral organs or obstacles to the introduction of the 10 mm trocar. After the introduction of a trocar, pneumoperitoneum was established, ensued with the introduction of the optic. A careful abdominal and pelvic inspection was carried out, starting with in most cases with the abdominal viscera. Additional trocars (5 mm) were usually placed for complete visualization of the small intestine or taking biopsies. Preferred positions of the trocars were the left or right iliac fossa, depending on preoperative imaging depicting tumor mass or adhesions. All peritoneal surfaces and folds, liver, periportal area, greater and smaller omentum, stomach, small bowel loops and mesentery were carefully inspected. The pelvis was mostly visualized through use of Trendelenburg position and when necessary, retraction of small bowel. Biopsies were taken from all suspicious lesions, along with cytology samples. The fascia at the 10 mm port-site was sutured.

Data from preoperative, intraoperative, and postoperative courses were obtained from reviewing medical records. Operative details about the procedure, intraoperative PCI, possible complications, and completeness of cytoreduction score were gathered from the surgical reports.

Results

The results of the 229 patients, patient characteristics and intraoperative data are described in Table I and Table II. The retrospective study included 115 men and 114 women. Median age was 53 years, range=19-86 years. Laparoscopy was performed on all patients with 101 patients ultimately undergoing the CRS and HIPEC. A total of 97 of the remaining 128 patients were excluded due to massive tumor infiltration of the small intestine or its mesentery, 81 of these during laparoscopy. This finding validates a presumed high sensitivity of detecting small intestinal metastasis through laparoscopy.

Table I. Data collected from the laparoscopy and subsequent exploratory laparotomy.

CRS: Cytoreductive surgery; PCI: peritoneal cancer index; CUP: cancer of unknown primary; HIPEC: hyperthermic intraperitoneal chemotherapy; LOS: length of stay; CCR score: completeness of cytoreduction score.

Table II. Pre- and postoperative features and data of enrolled patients.

CRS: Cytoreductive surgery; CUP: cancer of unknown primary; HIPEC: hyperthermic intraperitoneal chemotherapy; LOS: length of stay.

The most prevalent primary tumor site was gastric with 84 patients. A total of 42.5% of these patients underwent a CRS and HIPEC. Second most common site was colon cancer with 55 patients. A total of 47.2% of these patients were resectable. Our third most common cancer site was the appendix with 33 patients. 60% of all appendicular tumors were resected. Sixteen of patients had pseudomyxoma peritonei and 6 of them a peritoneal mesothelioma.

A higher median intraoperative PCI was seen in patients with appendix carcinoma (median of 9) than with gastric carcinoma (median of 4). However more patients with appendicular cancer were considered operable.

A total of 106 patients had intraabdominal adhesions. Adhesions were classified based on surgical reports, as none, minimal and extensive. Our results were sub-grouped respectively into 102, 28 and 78 patients respectively. A total of 185 patients had previous abdominal surgery, with 31.4% of them having extensive abdominal adhesions. Table III depicts the raising number of aborted CRS and HIPEC with increasing amount of adhesions. There was a total of 78 patients with extensive adhesions, 22 of them were not adequately assessable with the laparoscopy and needed ultimately a laparotomy.

Table III. Adhesions affected rate of completed cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC).

A total of 128 patients (55.9%) were excluded from the multimodal therapy. The most prevalent overall exclusion criterion was extensive small bowel disease with 84.4%. The second most prevalent was non-operable findings such as tumor growth into vessels and structures rendering the operation technically not feasible. Third most common was obtained biopsy results (5.2%) diagnosing a tumor deemed not suitable for CRS and HIPEC.

The overall median PCI score of excluded patients was 16.8, compared to 4.6 of the CRS and HIPEC group. A total of 26 patients (11.3%) were aborted during the exploratory laparotomy with intention of CRS and HIPEC. The median PCI of these patients was estimated to be 8.5 during the diagnostic laparoscopy and 17.1 during the exploration. This differs significantly when compared to patients who underwent the CRS and HIPEC, with median PCI of laparoscopy and explorative laparotomy of 3.3 and 5.4 respectively. One factor that sets the groups apart was the number of patients with extensive adhesions. The CRS and HIPEC group had 25.6%, the exclusion group 65.4%. Of the 101 patients deemed suitable of undergoing a CRS and HIPEC, 77% underwent a CCR0 resection. A total of 18.8% had a CCR1 and 2% either a CCR2 or CCR3 resection. A total of 80.6% of all aborted CRS and HIPECs were due to exclusion during the laparoscopy.

In regard to safety of laparoscopy, the complications we experienced were iatrogenic damage of abdominal viscera. Our cohort had 6 small bowel injuries (2.6%) and 1 colon perforation (0.4%) with all cases converted and treated with primary suture. An extensive amount of adhesions were recorded in 6 of the 7 perforations, indicating increased morbidity in this group. Nonetheless, remaining 72 patients that had extensive amounts of adhesions recorded no complications, further supporting the safe use of laparoscopy in PC patients. Conversion occurred in 13 patients (5.7%), with 6 due to an inaccessible abdomen. We witnessed 7 port-site metastases accounting for 3.1% of our patients. Length of hospital stay was significantly shorter with 5.1 days after aborted cytoreduction compared to 18.3 days after CRS and HIPEC.

Discussion

Selecting patients with peritoneal metastases for an aggressive, multimodal treatment strategy is a difficult task. It is combined with a long learning curve and needs to consider a vast number of different aspects, as previous described. One of the key-points is the resectability. Assessing resectability will be mainly possible prior to surgery using imaging tools and laparoscopy.

The CT imaging is in our opinion the modality of choice (using oral, rectal, and intravenous contrast) with detection rates varying from 60% to 90% (8,9). The detection rate however drops when the lesion size is under 1 cm, drastically compromising the sensitivity with detection rates of 10% to 30% (6,10,11). Added sensitivity with MRI/DWI-MRI and FDG-PET increases the detection rate of PC, particularly in those small lesions. However, the accompanying time and cost consumption deems them inappropriate as standardized tools.

A problem for patients with PC is the high rate of explorative laparotomies performed due to insufficient imaging for small lesions. Due to the wide access, one must take the increased morbidity, mortality, and LOS in consideration. Laparoscopy has therefore been introduced as a method of patient selection prior to CRS and HIPEC if there were any doubts regarding the completeness of cytoreduction.

We showed that the laparoscopy could be considered a sensitive tool in patients with aggressive tumor biology (poorly differentiated or with signet ring cells), ruling out military disease. As seen by our findings, more than half of our patients had this kind of histology. Most of these patients do not have large tumor volumes, e.g., patients with gastric cancer had a median PCI of 4, as compared to appendiceal malignancies with PCI of 9. A low median PCI in patients with gastric PC deemed eligible for CRS and HIPEC is described by others as well (28).

One possible limitation for laparoscopy may be adhesions due to previous abdominal surgery. This was the case in 80% of our assessed patients, which is similar to other published data (18,29). About one third of our patients had even extended amounts of adhesions. However, even if during laparoscopy not all quadrants were explored, an assessment of the small bowel was possible in 93% of all patients, which was the limiting factor for resectability. Therefore, some authors recommend an adhesiolysis during laparoscopy for a complete visualization (30). Our result show however that tumor spread in the peritoneal cavity, particularly the involvement of small intestine, can be accurately examined in patients with adhesions without any lysis. Subsequently 56% of all patients undergoing laparoscopy were not eligible for a CRS and HIPEC, avoiding unnecessary laparotomies. This is concordant with other published data (29,31). The rate of complete cytoreduction CCR0 and CCR1 was 96% in patients who went through the CRS and HIPEC. These results are similar to what other groups have shown (29,32).

A total of 11.3% (n=26) of patients were primarily deemed resectable and in turn were considered inoperable during exploratory laparotomy. This rate of PCI underestimation has been shown by others and in some cases as Tabrizian et al. showed underestimated in 30% of patients (20). Moreover Yurttas et al. reported in a recent analysis that at 27 of 43 patients (62.8%) PCI was underestimated during laparoscopy compared to open exploration. Further 9 patients (21%), who were considered suitable for CRS and HIPEC during prior laparoscopy, ended up having a non-therapeutic laparotomy (33). The authors concluded that laparoscopy is imprecise in the assessment of eligibility for CRS and HIPEC (33).

A total of 7% of our patients could not be thoroughly assessed with laparoscopy due to adhesions, correlating with what others have shown (18). Interestingly, 3.1% of our patients had a bowel perforation and 5.7% needed a conversion to explorative laparotomy. This shows that laparoscopy may have its own morbidity, even if low. Other groups studied the morbidity of diagnostic laparoscopy, showing complication rates between 0,4% and 2% (20,22,25,29), similar to ours. This rate is significantly lower than for laparotomies, which has been shown by others as high as 20% (21).

Conversion occurred in 13 patients (5.6%), giving a success rate of 94.4%. Tabrizian et al. reported a success rate of conducting the laparoscopy of 92.6% and a conversion rate of 7.4% (20). Iversen et al. reported in 2013 a success rate of 95.5% (18), while Hentzen et al. report a conversion rate of 13% (31). Port-site metastases (PSM) were rare (3%), most of them in patients with malignant peritoneal mesothelioma or gastric cancer, as reported by others (20,28). Vergote et al. reported a 42% rate of PSM in patients who underwent a complete port-site resection during the CRS (34). However, only 24% of the cases were clinically diagnosed during the operation. The remaining where established during the pathological examination, leaving the question of how many go undetected. Many studies however, such as Garofalo et al. reported no or very low incidence <2% of PSM, similar to the rate of midline metastases after open surgery (22-25).

Noteworthy was the time between laparoscopy and exploratory laparotomy, which was less than 14 days in most cases, possibly being too short for visible tumor growth at port sites.

Laparoscopy is a safe, sensitive, and easily applicable tool in patients with PC. It enables to avoid an unnecessary explorative laparotomy in more than half of the patients that appear potential candidates for CRS and HIPEC. Most frequent reason is a disseminated small bowel disease. Our results validate laparoscopy as a proficient tool in ruling out small intestine or mesentery involvement, even in patients with previous abdominal surgeries.

Even with laparoscopy, the tumor extent may be underestimated. In patients considered suitable for CRS and HIPEC, a completeness of cytoreduction (CCR0) could be achieved in 77%. The procedure can be performed even in patients with extensive adhesions showing relatively low complication rates.

Conflicts of Interest

All Authors have declared no potential conflicts of interest.

Authors’ Contributions

Conceived and designed the experiments: MA, AD. Performed the study: AD, HL, PP Analyzed the data: MA, AD, SB, HL. Contributed materials/analysis tools: HL, PP. Wrote the paper: MA, AD. Designed the software used in the analysis: AD. Reviewed the final version of the article: PP. All Authors have read and agreed to the published version of the manuscript.