Diagnostic and therapeutic algorithm for colorectal peritoneal metastases. A consensus of the peritoneal surface malignancies onco-team of the Italian society of surgical oncology

Aim: the surgical workup for colorectal cancer peritoneal metastases (CRCPM) is complex and should be managed in specialized centers. Diagnostic and therapeutic algorithms (DTA) have been proposed to balance optimal patients management and correct use of resources. Aim of this study was to establish a consensus on DTA for CRCPM patients in Italy. Method: a panel of 18 delegated members of centers afferent to Peritoneal Surface Malignancies Oncoteam of the Italian Society of Surgical Oncology was established. A list of statements regarding the DTA of patients with CRCPM was prepared according to different activities and decision-making nodes with a defined entry and exit point. Consensus was obtained through RAND UCLA methodology. Results: two different DTA were defined and approved according to the modality of presentation of CRCPM (synchronous and metachronous). A consensus was also obtained on 17 of the 19 statements


Introduction
Colorectal cancer (CRC) represents the third most common cancer in men and the second most common in women worldwide [1]. During the last decade, the median overall survival for patients with CRC metastatic disease have improved significantly and is currently estimated around 30 months [2]. Although peritoneum is a rare site of CRC spread (4.8e8.3%) [3e5], peritoneal metastases (PM) are associated with the worst prognosis compared to other metastatic sites (liver, lung, lymph nodes), with a median overall survival of 16,3 months [6]. Systemic chemotherapy, alone or combination with targeted therapy, represents a standardized approach of CRCPM [7,8]. In selected patients, cytoreductive surgery associated (CRS) followed by hyperthermic intraperitoneal chemotherapy (HIPEC) has been shown to be effective in prolonging survival compared to systemic treatment alone [9e12]. Patient selection remains crucial to obtain satisfactory results after CRS-HIPEC. Optimal cytoreduction with less than 2.5 mm (so called cytoreduction grade 0 or 1) and limited tumor diffusion inside the peritoneal cavity (expressed by the peritoneal cancer index, PCI) represent the most important prognostic factors for recurrence and survival in CRCPM treated with CRS-HIPEC [13,14].
Health requirements of the oncologic population becomes more and more complex every. In this scenario, diagnostic and therapeutic algorithms (DTA) allowing the definition of the most viable path within a single organization and network [15] for a particular disease or clinical problem are crucial. In this perspective, a defined DTA for CRCPM is needed for several reasons [16]. Most CRCPM patients are diagnosed outside a tertiary referral center, frequently in an emergency setting. When PM are discovered by colorectal surgeons during abdominal exploration, surgical strategy is neither clear nor standardized [17]. It is possible that a significant number of patients with peritoneal disease who may benefit from a multimodal approach such as CRS-HIPEC are undertreated, and every effort should be made to define the best clinical pathway for diagnosis and treatment of this group of patients. Moreover, CRS-HIPEC is a very complex procedure that is available in a limited number of specialized centers. Evaluation of CRCPM patients in these centers allows to select those who would benefit more from surgery in terms of morbidity and mortality, therefore optimizing the costs of treatment and improving patient quality of life [18,19].
Aim of this study is to establish a consensus on a DTA model viable for CRCPM patients.

Method
The consensus panel included 18 delegated members of centers afferent to Peritoneal Surface Malignancies Onco-team of the Italian Society of Surgical Oncology. The steering committee was charged of constructing a DTA considering the best evidence-based treatment options for CRCPM. During a first meeting (Siena, May 17, 2019), after widespread review of national or society guidelines and results of randomized controlled trials, phase I and II clinical studies, the committee determined that low quality evidence data were of little help to drive decision making. Thus, consensus on list of statements, based on the available literature and expert opinion, and appropriate to create a DTA for CRCPM was designed. DTA was diagrammed into a pathway which covers CRCPM patient care from the initial diagnosis through curative and palliative treatments. DTA was divided in two main clinical scenarios (synchronous and metachronous CRCPM) and organized according to different activities and decision-making nodes with a defined entry and exit point. A list of 19 statements regarding the different activities and decision-making nodes of the DTA were discussed and modified during a second meeting (Cagliari September 9, 2019). Definitive achievement of consensus was obtained through the RAND/University of California Los Angeles Appropriateness Methodology [20]. Appropriateness was scored from 1 (inappropriate) to 9 (completely appropriate). For each statement the appropriateness median score (AMS) was calculated. Statements were classified into three levels of appropriateness: appropriate (AMS, in range 7e9), uncertain (AMS, in range 4e6)) and inappropriate (AMS, in range 1e3). Agreement among panellists on each statement was established according to an Interpercentile Range (IPR) of 0,3-0,7. If the IPR obtained was in range 7e9, this was considered indicative of agreement for that statement. Each panellist ranked anonymously via web each statement for appropriateness as reported above. The same method has been used for obtaining consensus of the diagram of DTA. Statistical analyses (median and IPR) were conducted with SPSS 17.0 (SPSS Statistics, Release 17.0.0).

Results
The statements were related to different decision-making points in the clinical algorithm of CRCPM patients: 7 (1e7) the management of patients diagnosed with synchronous disease, 7 the multidisciplinary evaluation (8e14), 1 (15) the HIPEC regimen, 4 the role of perioperative systemic chemotherapy (16e19) ( Table 1). All the statements proposed for consensus were considered appropriate with an AMS ranging between 7 and 9. Statements 14 and 19, although scored as appropriate (AMS 7,5 and 7 respectively) didn't reach consensus with an IPR lower than limit of 7 (both 6). Statement 13 obtained an AMS of 7 but the IPR was 7e8. The voting results on statements are summarized in Table 2.
The two DTA diagrams for synchronous ( Fig. 1) and metachronous CRCPM (Fig. 2) obtained both an APM of 9 and an IPR of 8e9.

Discussion
The current study reports on a consensus obtained for a DTA in CRCPM patients. In oncology, DTA represents an organizational health pathway, distinct for specific tumors and stages, in which every step of the process is codified and the responsibility for its execution and implementation are clear and defined. The main objectives of DTA is to provide timely patient management, quality of care and patient satisfaction, continuity and equity in terms of access of care and an optimal use of resources. CRCPM is a typical clinical situation in where the complexity of the diagnostic and therapeutic process determines circumstances favouring the heterogeneity of care and facilitate inappropriate behaviors and/or errors, thus causing adverse effects on cancer prognosis and an increase in treatment costs [18,19]. In a Dutch population study, a large variation in treatment approach for synchronous CRCPM was observed at individual hospital level (teaching versus non-teaching center), with significant differences in the chance to undergo CRS-HIPEC and to obtain a favorable survival outcome [21].
The consensus was reached for two main clinical situations, synchronous and metachronous CRCPM (Figs. 1 and 2). In around 10% of cases, PM occur at the time of primary presentation (synchronous) [3,5]. If primary tumor is asymptomatic/uncomplicated, patients with suspected synchronous peritoneal involvement should be directly evaluated by a peritoneal tumor referral center (PTRC) and considered for curative CRS [22] (statement 1). However, PM are more difficult to be detected by conventional imaging (CT or MRI scan) compared to other metastatic sites (liver, lung and lymph nodes) and PM can be an unexpected event (around 4%) during surgery for primary CRC [4]. Intra-operative diagnosis of PM during treatment of primary CRC occurs in most of cases in nonspecialized centers. The correct strategy in this situation is still matter of debate. Small retrospective studies showed that the 5year overall survival of patients treated with surgical resection of the primary and synchronous PM is about 30%, which is lower than that obtained by specialized centers in patients operated with CRS with or without HIPEC [23,24]. Moreover, primary tumor resection seems to determine extended bowel resections and permanent colostomy respect to patients treated with a one-stage procedure [25]. Even if, one stage curative in PTRC is preferable, in presence of nodules limited and close to the tumor, resection of the primary and the peritoneum involved, it is an acceptable option (statement 2) [26,27]. Patients treated for synchronous PM in non-specialized centers, even those with complete resection, should be referred to a PTRC. In a significant quote of patients who underwent resection of PM at the time of primary treatment, a second look surgery allows to discover and treat further peritoneal metastatic nodules with similar outcomes of one-stage curative treatment [28,29]. If the tumor load is high, with multiple and diffuse nodules, a simple histological confirmation after an accurate intra-operative staging (PCI and unresectability causes) is the best choice, before sending the patient to a PTRC (statement 3). In patients with suspected CRCPM and symptoms of obstruction/bleeding, emergency surgery should be performed (statement 4). In case of not deferred surgery for perforation, obstruction or bleeding, radical resection of the primary can be carried out if safe and limited nodules are present, (statement 5). However, in case of emergency surgery and diffuse PM, a limited palliation (primary resection only, stoma formation) with peritoneal biopsy and staging (PCI and unresectability causes) is the best option (statement 6). In those patients treated in an emergency setting but still bearing PM, a procedure of CRS-HIPEC remains a valuable and safe option with a 5-year survival rate similar to that of elective cases [30]. One-stage curative treatment at HIPEC referral centers is still the preferred choice whenever possible [28] and every patient with synchronous PM potentially eligible for CRS-HIPEC should be evaluated by a referral center and discussed in dedicated CRC multidisciplinary meetings (statement 7).
After primary curative treatment, the rate of metachronous PM can reach 19% of cases [31]. Multidisciplinary discussion of CRCPM patients is mandatory for selecting the best therapeutic approach. In this setting, a complete medical history, abdominal/thoracic CT scan and blood tumor markers should be available (statement 8). Every patient with a diagnosis of CRC should undergo CT scan with contrast enhancement medium which represents the gold standard not only for diagnosis but also for staging those with PM (statement 9). Sensitivity and specificity of CT scan in CRCPM is estimated 83 and 86%, respectively [32]. The diagnostic performance of CT scan for PM is dependent on the radiologist's experience and can be suspected in presence of defined radiological features (ascites, organ invasion) [33]. The role of MRI and PET-CT as complementary investigations are still under investigation and should be reserved in selected cases and performed by experienced radiologists in PM diagnosis (statement 10) [32]. Diffusion-weighted MRI is promising for estimation of PCI and prediction of operability but its role is still under study [34,35]. Laparoscopic exploration is an important tool 1 Patients with suspected synchronous peritoneal metastases should be directly evaluated by a peritoneal tumor referral center if the primary is asymptomatic/ uncomplicated 2 PM can be an unexpected event during surgery for primary CRC. In this case, if the nodules are limited and close to the tumor, resection of the primary and the peritoneum involved, it is an acceptable option. However, one-stage curative treatment with HIPEC in referral centers is preferable whenever possible 3 PM can be an unexpected event during surgery for primary CRC. If the tumor load is high, with multiple and diffuse nodules, a simple histological confirmation and an accurate intra-operative staging (PCI and unresectability causes) is the best choice, before sending the patient to a peritoneal referral center 4 If the patient is symptomatic and peritoneal metastases is suspected, surgery should be performed according to the urgency to intervene before sending the patient to a peritoneal referral center 5 In case of obstruction, perforation or bleeding, when radical resection of the primary can be carried out safely and limited nodules are present, resection of the primary combined with peritoneal metastases is an acceptable option before sending the patient to a peritoneal referral center 6 When surgery is emergent and peritoneal metastases are diffuse, a limited palliation (primary resection only, stoma formation) with peritoneal biopsy and staging (PCI and unresectability causes) are indicated, before sending the patient to a peritoneal referral center 7 Every patient potentially eligible for CRS-HIPEC should be evaluated by a referral centre and discussed in dedicated colorectal multidisciplinary meetings 8 During colorectal multidisciplinary meetings a complete medical history, abdominal/thoracic CT scan and blood tumor markers should be available 9 CT scan with contrast enhancement medium represents the gold standard for stage patients with colorectal peritoneal metastases 10 MRI and PET scan should be considered complementary imaging for stage patients with colorectal peritoneal metastases 11 Laparoscopy is a complementary method for stage patients with colorectal peritoneal metastases and It is crucial that is performed by surgeons with experience in CRS-   for confirming the diagnosis of peritoneal involvement, allowing a histological definition of all suspected nodules [29]. Laparoscopic exploration can assess the peritoneal cancer index (PCI) more accurately then preoperative radiological investigations [36]. Moreover, during laparoscopy, any potential cause of unresectability (mesenteric retraction or infiltration of the hepatic hilum, suprahepatic veins and Treitz ligament) can be easily identified under direct vision [37]. Therefore, it is crucial that laparoscopy should be performed by skilled surgeons in peritoneal surgery, it being is the most accurate staging tool in patients selected for surgery (statement 11). Despite notable advances in systemic treatments, patients with isolated PM treated only with cytotoxic/targeted agents show a significantly worse survival (16,3 months) as compared to patients with isolated non-peritoneal sites (liver, lung, lymph nodes) [6]. CRS-HIPEC allows to provide a long-term survival of up to 40 months in selected patients treated in PTRC and is widely adopted worldwide and included in several national guidelines [38e42]. However, the real added value of HIPEC over CRS alone is still under investigation. A recent French multi-institutional randomized controlled trial (Prodige 7) presented at the ASCO annual meeting confirms the relevant role of surgery to treat patients with PM but was not able to show any survival advantage in adding HIPEC to radical surgery [24]. The study is not yet published and it is difficult to assess all aspects and bias, but an important point is that in the subgroup analysis of patients with PCI less than 16, overall survival is significantly better in the HIPEC arm, confirming previous results on the impact of this benchmark in patient prognosis [43]. Therefore, only patients with PCI <16 in which a complete surgery can be obtained (grade of cytoreduction or residual disease zero or less than 2,5 mm) (statement 12).
The role of additional selection factors and their weight within the pre-operative evaluation is unclear and also the panelist didn't reach a consensus on this point. Among these, mutations in the RAS and RAF pathway have a detrimental effect on prognosis but a clear link to peritoneal progression has not yet been clearly demonstrated [44]. However, the selection process for CRS-HIPEC should include always a molecular gene mutation testing (RAS/RAF mutation) and microsatellite status (stable or not) which should be considered important selection factors (statement 13). Other selection factors as performance status (Eastern Cooperative Oncology Group, ECOG), extraperitoneal metastases (liver, lymph nodes), tumor site (right versus left colon) tumor differentiation and signet ring histology ascites, symptoms of obstruction and lymph node status (N2a) of the primary [45], are not considered crucial in the selection process (statement 14).
According to the results of the Prodige 7 trial which showed no survival advantage and more complications (bleeding) in the arm of patients treated with an oxaliplatin based HIPEC, the panelists recommended a Mitomycin C based regimen in patients treated with maximal cytoreduction surgery [24] (statements 15), which is also the most used protocol in Italy. This is in accordance with previous comparative studies which are unable to show any superiority in terms of oncological efficacy of one drug over the other. However, some evidence showed that Oxaliplatin in HIPEC schedule is clearly associated with a worse side effects profile [46,47].
The role of perioperative chemotherapy in eligible patients for CRS-HIPEC is controversial and this topic is still under study [48e51]. A randomized controlled study (CAIRO 6) addressing the role of perioperative systemic chemotherapy (SC) for patients with isolated and resectable CRCPM is ongoing in the Netherlands [52]. After CRS-HIPEC, systemic failure is a relatively common event [53], and postoperative SC can lead to better peritoneal tumor control by eradicating non-visible cells as supposed for lung and liver metastatic patients treated by potentially curative resection [54]. SC after surgery should be indicated according to the recognized risk factors of postoperative recurrence [45]. Before CRS-HIPEC, SC should be considered in all patients with CRCPM (especially synchronous) according to the protocols proposed for metastatic CRC (statement 16). Although a relative unresponsiveness of CRC peritoneal nodules to SC is demonstrated (response rate around 10e20% lower than those reported for liver metastases) [55], SC should be considered for several reasons, especially in presence of synchronous disease. SC can reduce peritoneal tumor load, improving the chance of achieving complete cytoreduction by tumor downsizing and performing a less extensive surgery with lower morbidity [55]. Another theoretical advantage of "neoadjuvant" chemotherapy is that it allows a better selection of patients, by excluding from surgery those who develop early distant progression (liver, lung, lymph nodes). Isolated peritoneal progression during systemic CHT should not be considered an absolute contraindication for CRS-HIPEC, if the other selection criteria are still met [56] (statement 17). In very selected patients with limited metachronous PM and no risk factors, perioperative systemic CHT could be omitted after MDT discussion in favour of a front-line CRS-HIPEC (statement 18). In the sub-group of CRCPM patients who have mismatch repair-deficient (dMMR) and/or microsatellite instability-high (MSI-H), the treatment with immune checkpoint inhibitor is considered a promising option, also as first choice treatment [57]. However, the role of immunotherapy in CRCPM is still under study and although deemed appropriate, should not considered as preferable treatment over CRS-HIPEC if the selection criteria are meet (statement 19).
There is an increasing interest for quality and standardization in surgery, especially in complex procedures requiring specific professional skills which must be consistent at the different levels and steps of care and be delivered in a coordinated manner to obtain the best results [58,59]. Our study has defined a shared algorithm for diagnosis and treatment in a very challenging clinical situation as the occurrence of CRCPM. Additional efforts should be made in this field considering that, for example, no defined criteria are available in most countries and health care systems for the identification of a PTRC. The learning curve is a crucial point to identify centers able to cure these patients with high quality and efficacy. Indeed, the number of procedures performed remains the most objective parameter related to post-operative outcomes (mortality, complications) and oncological results (incomplete cytoreduction, early recurrence) [60,61]. Moreover, further studies will be necessary, which measure the appropriateness and quality of the care path of CRCPM patients through proper QA indicators (relevant, reliable, valid and easily useable), with the primary goal of ameliorating outcomes (survival/quality of life) and cost of treatments of these patients.

Conclusion
A consensus on DTA for CRCPM patients have been reached in Italy. The document will be an important tool for healthcare providers to monitor appropriateness in diagnosis and treatment of CRC patients with isolated peritoneal metastases.
The authors have no affiliation with any organization with a direct or indirect financial interest in the subject matter discussed in the manuscript.

Funding
The study was not funded

Ethical approval for research
No.

Declaration of competing interest
None of the authors have any conflicts of interest.