Metastasizing colorectal cancer (CRC) is one of the leading causes of tumor-related death 1. At the time of diagnosis, about 20% of the tumors have synchronous non-resectable liver metastases 23. In these patients, the aim of treatment is to prolong survival and to maintain an acceptable quality of life. Modern chemotherapy may serve this purpose 24. Depending on the regimen used, response rates vary between 20–70% 56. Most phase III trials using combination therapy have reported a response rate on the order of 40% 7. In 20–30%, stability of the metastases can be achieved 89. Mean survival time in patients with advanced stage CRC varies between 12–21 months after diagnosis 10.

With the development of modern chemotherapy and endoscopic techniques, new therapeutical options for the treatment of advanced metastasized CRC have emerged. The 5-FU-leucovorin-based regimen combined with oxaliplatin or irinotecan was approved for the first-line treatment of advanced CRC and has become the treatment standard 11. Combination with antibody therapy, such as the anti-VEGF-antibody bevacizumab 8 or the anti-EGFR-antibody cetuximab have become further options 12.

However, surgical treatment of the primary tumor in non-curatively resectable stage IV disease is still a matter of discussion 131415. In the past, investigators have recommended elective resection of the primary tumor to prevent the need for urgent surgical procedures arising from local complications 16. More recently, some authors have suggested elective resection of asymptomatic colorectal cancers at least in a subset of patients with less advanced stage IV disease 1718. Others have suggested deferring the resection of minimally symptomatic CRC, as in most of these patients death is related to systemic disease progression rather than complications related to the primary lesion 219. In the setting of incurable stage IV CRC, in which colonic resection has further been associated with a high mortality rate 1920 no survival advantage is gained by resection of an asymptomatic primary lesion 21.

As metastatic disease progression determines the outcome of these patients, efforts have been concentrated on the surveillance of tumor-related metastasis, typically measured by imaging techniques at scheduled intervals 222324. In case of non-resected asymptomatic CRC, surveillance does generally not include colonoscopies at tight intervals. Local inspection of the primary may be repeated to select patients for operation from a primarily non-operative setting.

Encouraging preliminary studies 142526 prompted us to observe the local tumor behaviour in patients with non-curatively resectable stage IV CRC during chemotherapy treatment. We are aware that modern treatment of these patients is based on individual treatment concepts which need to be re-evaluated during the course of therapy. Sometimes, resection of metastasis is possible so that palliative chemotherapy becomes a later therapy option. We here report on the local tumor behaviour, the time-to-progression of the primary and the overall survival of 20 patients treated with modern pharmacotherapy with or without endoscopic tumor debulking.

Tumor staging was undertaken before therapy and every three to four months after the beginning of treatment. After the initial staging and a possible initial tumor debulking, the patients underwent chemotherapy treatment.

Endoscopic tumor debulking of the primary tumor was performed if the tumor mass resulted in a luminal obstruction of about 80% or above as detected endoscopically during staging, or if bleeding or obstruction symptoms were reported. The endoscopic treatment was done either by using a standard polypectomy snare technique alone or in combination with argon plasma coagulation. If the metastasis was considered operable during ongoing treatment, operation of both the metastasis and of the primary was offered.

Staging included the measurement of the tumor marker CEA (carcino embryonic antigen) and the imaging of the metastasis (CT scan or ultrasound) and, if considered necessary, the inspection of the local tumor by sigmoidoscopy or colonoscopy. Endoscopy was initially performed every three to four months in non-symptomatic patients. In symptomatic patients, endoscopic tumor debulking was performed at, on average, two sessions (range 1–5) in the first two months.

Since endoscopic examinations can be associated with complications, we comment that there is a question as to how often they should be performed. For the patients in this study, the high degree of luminal obstruction initially suggested regular examination and where appropriate, debulking.

Complete response of the primary was defined as macroscopically intact mucosa and histologically confirmed absence of malignant cells in the biopsy.

Partial response of the primary was defined as macroscopic tumor regression with histologically confirmed tumor cells.

Endoscopic regression of the primary has not previously been defined in the literature. The inspection of the gut lumen was done by experienced endoscopists (TA, JGS, HS, GR), and included the description of bowel motility and the possibility of passage of the tumor with a colonoscope. Depending on the macroscopic appearance, the decision had to be taken whether endoscopic intervention was needed. Complete luminal occlusion – no passage possible with the thinnest endoscope (4 mm) – was rated 100%. If the tumor could just be passed with the colonoscope (14 mm), luminal occlusion was considered to be 85%. We are of course aware this is a rather subjective method.

For most patients with CRC, surgical resection with or without chemo(-radio)therapy is the standard treatment approach. The treatment of surgically non-curable metastasis of colorectal cancer is palliative chemotherapy 2829.

Taking into account that in many cases bleeding of the tumor or obstructive disturbances can now be treated endoscopically 252730, the indication for surgery of the primary tumor in patients with unresectable metastatic CRC appears more controversial 132131. The fact that after primary-directed-surgery in stage IV disease a significant proportion of patients do not get palliative chemotherapy; e.g. in the study of Temple et al. 40% of the patients above 65 years of age did not receive chemotherapy after surgery of the primary tumor 32, argues against initial surgery of the primary. Supporting this argument, several retrospective studies have shown that first-line chemotherapy is safe 21 and more effective than it was before 2002 31. In addition, the rate of gastrointestinal complications related to the unresected primary is low and comparable to those observed in resected patients 1421. When surgical resection of the primary was compared to conservative treatment in the setting of incurable stage IV disease, no survival advantage was gained by the resection of an asymptomatic primary 21. It is understood that most of these patients die of their systemic disease before the development of a major complication related to an intact primary lesion 2133.

Although the number of patients in our study is limited, it represents a further attempt to show that chemotherapeutics combined with interventional endoscopy avoid the need for resection of the primary in unresectable stage IV CRC, even when the primary is symptomatic (obstructive or bleeding). From our study we are unable to comment on the relative effectiveness of chemotherapy or endoscopic debulking in isolation. However, interventional endoscopy might not be necessary in asymptomatic patients under effective chemotherapy.

The survival rate of our patients is similar to that of other studies in which the primary tumor had been surgically resected and where subsequent chemotherapy was possible 10. We are aware that the results of our study should be considered as tentative and hopefully will encourage future larger studies. In mono-centric studies, large numbers cannot be obtained for this indication.

Patients' characteristics in our study were comparable to other studies performed in a patient collective with metastasizing CRC as regards the location of the primary and the site of metastasis 812. However, they were distinct for age and performance status at the beginning of the study. The median age in our study was, at 67 years, older than the 65 years reported in other studies. Two Patients were above 80, six patients were between 70 and 80 years old. Six of the ten patients with macroscopic tumor regression (group III) were older than 70, and could not receive complete chemotherapy courses. These patients are also those who might not receive chemotherapy after primary-directed surgery 32. Old age is a negative prognostic factor for general outcome, making preventive medical check-ups inevitable. Early administration of 100% of the calculated chemotherapy dose with completed chemotherapy cycles was associated with a significantly better OS and TTP of the primary in advanced disease.

In the present study, combined endoscopic and pharmacological therapy was found to be effective and safe not only in asymptomatic, but also in symptomatic advanced stage IV CRC.

Our treatment improved both the initial symptoms of all patients and reduced or stabilized the size of the primary tumor. We therefore confirm and further expand the data by Chau et al., and others who reported rapid symptomatic response after neoadjuvant treatment in patients with colorectal cancer 133134. The survival rate of our patients is similar to that of other studies in which the primary tumor had been operated 10. Furthermore, this study shows that anti-cancer drugs combined with endoscopic therapy may achieve a local response in the majority of the patients: The primary tumor seems to respond more effectively to chemotherapy than its metastases.

CEA: (carcinoembryonic antigen); CRC: (colorectal cancer); ECOG: (Eastern Cooperative Oncology Group Performance Status); OS: (overall survival); TTP: (time-to-progression).

The authors declare that they have no competing interests.

SC has prepared the manuscript and is responsible for the data collection. DH supervised the chemotherapy of the patients and collected the data for chemotherapy. She performed the Kaplan Meier analysis of the treatment outcome. TM did some of the documentation of the endoscopic images of the tumors. TA, JGS, HS and GR did the endoscopic tumor debulking and were responsible for the treatment choice of the patients. LF performed the histopathological analysis of the tumors. GR contributed to the main ideas of this work, selected the patients and helped to realize the paper. All authors have read and approved the manuscript.

We are thankful to the collegues and the nurses of the department of gastroenterology and endocrinology who participated in the treatment of the patients. We wish to thank all the patients for allowing us to summarize the clinical data for publication.