Treatment of patients with metastatic colorectal carcinoma is surgical resection. Only 10-15% of the patients will be candidates for curative resection. After response to chemotherapy this figure rises 10-13% more. To perform the surgery it is necessary to have a sufficient remnant liver volume (RLV), which allows maintaining optimal liver function after resection. If the estimated RLV is insufficient preoperatively, portal venous embolization site (PVE) is performed for compensatory hypertrophy, thus increasing the number of resections 19%. Still, in 20% of these patients surgery can not be performed because RLV is not achieved or because the disease progresses while waiting for growth. Therefore, it is necessary to improve liver regeneration without promoting tumor growth. Studies on liver regeneration, have determined that cells (CD133 +) are involved in the liver hypertrophy that occurs after hepatectomy. CD133 + have been used to induce liver hypertrophy with encouraging results. This population of CD133 +, can be selected from peripheral blood after stimulation with Granulocyte colony-stimulating factor (G-CSF), being able to obtain a large number of them. The investigators propose to treat patients who do not meet criteria for surgery because of insufficient volume <40%, with CD133 + and portal embolization in order to carry out a surgical resection in a second place.

Inclusion criteria:

1. Men and women between ≥ 18 years and ≤ 80. Women in fertile age should use contraceptive methods recommended by the Clinical Trial Facilitation Group (CTFG).

2. ECOG (Eastern Cooperative Oncology Group) quality of life scale ≤ 2.

3. Patients with hepatic metastases of colorectal carcinoma and insufficient hepatic remnant function to perform a major hepatectomy. This volume would be calculated by Positron emission tomography-Computed tomography (PET-CT) images or 64-channel multidetector CT.

   In patients who have received preoperative chemotherapy, the estimate residual liver volume to be included in the study will be <40%.

   Patients who have preserved liver function (Child ≤ B7 and International Normalized Ratio (INR) ≤ 2 in non-anticoagulated patients) and those who have not received chemotherapy, needs < 30% of residual volume to be included in the study.

4. Patients should have signed informed consent.

Exclusion criteria:

1. Pregnancy or lactation period.
2. Any condition that the investigators consider an unjustifiable risk in the patient.
3. Severe comorbidities: American Society of Anesthesiologists (ASA) ≥ 4.
4. Alterations in the hemogram and morphological alterations, evaluated by hematologist prior to administration of G-CSF.
5. Liver function: Child ≥ B7 and INR ≥ 2 in non-anticoagulated patients. In anticoagulated patients the values will have to be reverse, prior to surgery.
6. Patients who have not received research drugs in the last 30 days or in the period of 5 elimination half-life.
7. In patients on chemotherapy treatment, G-CSF can not be administered until 48 hours after the last administration of chemotherapy. Patients who have received Bevacizumab must wait 30 days since the last administration.
8. In addition, all those aspects that prevent the patient from being part of the study, understand the rules, follow the instructions given, or other aspects.