Treatment Response Evaluation in Patients With Non-resectable Colorectal Liver Metastases A Feasibility Study

Overview

More than 4200 new cases of colorectal cancer (CRC) are diagnosed each year in Denmark, and 30.000 patients live with the diagnosis. Up to 40% of CRC patients will have synchronous liver metastases (LM) at the time of the diagnosis or will develop metachronous LM during the course of their disease. CRC-LM are treated with a combination of chemotherapy and liver surgery, but less than 25% of the referred patients with CRC-LM may be treated with curative intend. If looking at population-based data this figure drops to less than 5%. During pre-operative chemotherapy the treatment response is monitored by CT and MR scans, and the patients are then discussed on multidisciplinary team (MDT) conferences. However, monitoring is inaccurate since the simple measurement of size of the liver lesions cannot provide reliable evidence of the treatment response. The cancer cells may actually have been replaced by scar tissue but without any shrinkage. The question is how may we improve the evaluation of treatment response? With the goal of improving the assessment of the response to chemotherapy, and thereby only treat the patients that will benefit from chemotherapy?

Clinical study Evaluation of oncological treatment response in patients with non-resectable colorectal liver metastases based on microscopic tissue analyses obtained by a minimal invasive technique. Background The incidence of colorectal cancer (CRC) in Denmark is 4.200 cases and the prevalence is approximately 30.000. One in five patients will have synchronous liver metastases (CRC-LM) and a similar number will develop metachronous CRC-LM during the course of the disease. Surgery is the cornerstone in treatment of CRC-LM with a curative intent, most often in a combination with chemotherapy. There are no randomized studies comparing surgery to chemotherapy alone, but data indicate that there is an early advantage in progression free survival when surgery is combined with chemotherapy in comparison to surgery alone, although this might not translate into an overall survival benefit. The 5 year survival rate for patients treated with liver resection is approximately 40 % and even higher in selected series. Clinical series suggest that one in four patients with CRC-LM may be resected, whereas the few population based studies report of resection rates as low as 4%. In order to improve the overall treatment of cancer patients in Denmark, the Danish Health and Medicines Authority has launched a series of National Integrated Cancer Pathways ("Kræftpakker"). A pathway for CRC-LM was issued by the end of 2008 in order to reduce processing time (e.g. faster diagnosis and quicker onset of treatment), and the compulsory use of a multidisciplinary team conference (MDT) evaluation in all patients with CRC-LM was stated in this document. Thus, all patients in Denmark with CRC-LM must be evaluated at an MDT in one of the four regional centers treating these patients (Aalborg, Aarhus, Odense or Copenhagen). However, epidemiological studies from the Region of Southern Denmark on gastro-esophageal and pancreatic cancer clearly indicate that a substantial number of patients are not getting any active treatment and/or are not even presented and evaluated on the regional MDT. This could also apply to patients with CRC-LM, and this may explain some of the differences in the reported treatment rates between population based and clinical based studies. When the patients with CRC-LM are evaluated on the regional MDT, they will be divided into five groups based on MR/CT scans and an assessment of co-morbidity and performance status: 1. Patients with resectable liver metastases 2. Patients with potentially resectable liver metastases. These patients will need "downstaging" in order to undergo resection 3. Patients with non-resectable liver metastases. 4. Patients with disseminated extra-hepatic disease 5. Patients unfit for surgery and chemotherapy Patients in group 1-4 will be offered chemotherapy in different settings. The re-evaluation after chemotherapy of the patients in group 2, 3 and 4 is crucial to the next treatment step in these patients: Should chemotherapy be continued, stopped or should another chemotherapy regimen be used? Unfortunately, this re-evaluation is uncertain or even wrong, and this may have a significant negative impact on the patient's situation. In general, the major problem in studies on chemotherapy in CRC-LM (as in many other cancer diseases) is the objective evaluation of the treatment response, and thus the effect of different treatment strategies. Worldwide, this evaluation is based on non-invasive imaging and surrogate markers like the RECIST 1.1 criteria (e.g. measurement of tumor diameter. Complete radiological response (e.g. no visual evidence of disease) is observed in 4-6% of patients with locally advanced or metastatic CRC treated by modern chemotherapy regimens, whereas 5-38% of the patients present with one or more disappearing lesions in the liver after receiving chemotherapy. However, a complete pathological response as low as in 20% in these disappearing lesions underlines the imaging problem. The presence of complete calcifications on CT scans may indicate response, but it does not imply sterilization of the malignancy. Thus, resection of calcified lesions and the areas of the disappearing lesions are recommended, but this strategy may prove difficult when based on therapeutic imaging. At the other end of the response scale, it is estimated, that up to 15% of the patients will progress during systemic therapy, and this problem must also be addressed during re-evaluation scans. Thus, a long list of problems relates to the objective evaluation of treatment response in patients with CRC-LM. At the moment, the best way to evaluate this response would be a detailed intra-operative ultrasonographic (IOUS) mapping with histological comparison of the metastases before and after treatment. Obviously, this method would face both ethical and practical problems, but using a setup with minimal invasive, pre-therapeutic fiducial marking followed by laparoscopic biopsy of the marked lesions before and after treatment may be the first step towards a reliable (e.g. histologically controlled) re-evaluation of these patients. In addition, this approach would enable an estimate of the potentially damaging effect of chemotherapy on normal liver tissue. Hypothesis The hypothesis of the clinical study part is that we will be able to devise a minimal invasive, safe and histologically controlled setup for the evaluation of the treatment effect following chemotherapy in patients with initially non-resectable CRC-LM. Aim of the clinical study To devise and test a new minimal invasive setup for the evaluation of treatment effect following chemotherapy in patients with initially non-resectable CRC-LM. More specific, the clinical study will address the following questions: 1. Is it technically feasible to perform laparoscopic ultrasound (LUS) guided marking of CRC-LM? 2. Is it safe to perform LUS guided marking of CRC-LM? 3. Is it possible to locate the markings on conventional non-invasive imaging and during LUS following chemotherapy? 4. Is it possible to obtain a representative laparoscopic (LAP) specimen/biopsy of the marked tissue after chemotherapy? 5. Is it possible to monitor treatment response by LAP inspection and LUS evaluation? 6. Is it possible to monitor histological response based on the obtained biopsies/specimens? 7. How well does imaging response (e.g. RECIST) correlate with histological response? 8. Which side-effects on the non-neoplastic liver does the chemotherapeutic regimens have, when comparing the pre-chemotherapy non-neoplastic liver biopsy with a post-chemotherapy non-neoplastic liver biopsy? Method and design for the clinical study The clinical study will include patients with CRC-LM in group 3, only (see Ethics and above). The main study will be preceded by a pilot study with the main focus on aim a) through d). These endpoints are merely observations of the technical feasibility and safety of the LUS guided marking technique and an evaluation of the obtained biopsies (+/- sufficient for pathological analysis). The pilot study includes 8 patients. Successful marking, repeated biopsies and a complete primary and secondary evaluation, including histological evaluation, in six patients without complications are considered necessary before proceeding to the main study. The main study will include 30 patients completing both histological evaluations (including the 8 patients from the pilot study). There are no available data to support the calculation of the number of patients necessary to answer the questions e) - h). Thus, the number of included patients is based on an estimate of the number needed to establish a reliable minimal invasive technique balanced against an estimated minimal number of liver biopsies necessary to evaluate treatment response as well as the impact of chemotherapy on non-neoplastic liver tissue. Patient inclusion All patients diagnosed with CRC-LM are evaluated on the regional MDT which is held twice a week. Patients in group 3 with non-resectable CRC-LM (based on pre-therapeutic imaging) are approached for inclusion in the study. The incidence of CRC in The Region of Southern Denmark is approximately 1100. Approximately 400 patients will present with CRC-LM each year, and at least 200 patients will belong to group 3-5. A high proportion of these patients will be candidates for this study. Laparoscopy, LUS, mapping and marking Before palliative chemotherapy is initiated, these patients will undergo laparoscopy (LAP) and laparoscopic ultrasonography (LUS) in general anesthesia with a detailed mapping of all metastatic lesions (size/lesions/segment, +/- superficial). During this procedure, two metastases (preferably in the left liver lobe, segment 2 or 3) are biopted followed by a LUS guided fiducial marking. Moreover, to be able to evaluate the appearance of the non-metastatic liver prior to chemotherapy, another liver biopsy from non-metastatic liver will be obtained. This biopsy will aid as "baseline", to which the post-treatment non-neoplastic biopsy will be compared. All patients will be treated and monitored (blood samples, CT/MR scans) according to the normal standards for palliative chemotherapy in patients with non-resectable CRC-LM during the study. Thus, apart from the perioperative admission, the patients are monitored and treated by the Department of Oncology, Odense University Hospital. After two months of standard palliative chemotherapy and standard CT/MR follow up (RECIST and morphology evaluation(16), a second LAP/LUS is performed, and the two marked lesions are laparoscopically biopsied along with another biopsy from non-neoplastic liver. In order to further optimize a future minimal invasive setup, an attempt will be made to perform percutane ultrasound guided biopsy of the marked metastases. The LAP/LUS will be carried out right after as described above in order to ensure safety and adequate specimen retrieval. Evaluation of biopsies and resected specimens When possible the patients are also subdivided according to chemotherapy regimens during the pilot study. This subdivision includes KRAS mutant (KRASm) and KRAS wild type (KRASwt) patients in order to evaluate the effect of bevacizumab-/cetuximab-based treatment regimens. There are several methods to evaluate the histological response of liver metastases to chemotherapy (see point A below). Moreover, the evaluation of the potential negative side effects of chemotherapy in the non-neoplastic liver tissue (e.g. sinusoidal obstruction syndrome (SOS) and Chemotherapy-Associated SteatoHepatitis (CASH,) is clinically relevant (see point B below). A) Analysis of the pre- and post-therapeutic biopsies from liver metastases 1. Tumor Regression Grading (TRG). This scoring system will be applied to evaluate the histological effect of chemotherapy on the biopsy from the liver metastases. The type of necrosis will also be recorded. If the necrosis is of the "infarct-like" or "mucin-like" type (ILN), then also a modified TRG (mTRG) grade will be assigned for these cases. In brief, usual necrosis (UN) will be regarded as indicative of a lack of response, whereas infarct-like necrosis (ILN) will be regarded as fibrosis, i.e. as indicative of response to chemotherapy. 2. Tumor thickness measured at the Tumor-Normal liver tissue Interface (TNI) will be used, if the TNI is clearly present in the biopsy. This method has been validated as a prognostic factor for therapy response and survival outcome in patients with resected CRC-LM when applied upon resected metastases. B: Analyses of pre- and post-therapeutic biopsies from the non-neoplastic liver parenchyma These biopsies will be evaluated for the presence/absence of CASH and SOS.