International Journal of Colorectal Disease

, Volume 20, Issue 1, pp 57–61

Brain metastasis from colorectal cancer

Authors

    • Department of Surgery and Surgical Basic Science, Graduate School of MedicineKyoto University
  • Satoshi Nagayama
    • Department of Surgery and Surgical Basic Science, Graduate School of MedicineKyoto University
  • Tsuyoshi Tachibana
    • Department of Surgery and Surgical Basic Science, Graduate School of MedicineKyoto University
  • Akihisa Fujimoto
    • Department of Surgery and Surgical Basic Science, Graduate School of MedicineKyoto University
  • Masayuki Imamura
    • Department of Surgery and Surgical Basic Science, Graduate School of MedicineKyoto University
Original Article

DOI: 10.1007/s00384-004-0631-3

Cite this article as:
Onodera, H., Nagayama, S., Tachibana, T. et al. Int J Colorectal Dis (2005) 20: 57. doi:10.1007/s00384-004-0631-3

Abstract

Purpose

The mechanism of brain metastasis is not well understood, but the affinity between cancer cells and neural tissues may be involved in the process. The aim of our study is to elucidate the involvement of neural cell adhesion molecule (NCAM) and therapeutic parameters in patients with brain metastasis from colorectal cancer.

Methods

We retrospectively identified 17 patients with brain metastasis from colorectal cancer. Data were collected with regard to patients’ characteristics, location, and stage of primary tumor, and extent and location of metastatic disease. NCAM histochemical staining was undertaken using a paraffin block, and compared with 56 Dukes C patients and 13 Dukes D patients.

Results

Neural cell adhesion molecule expression was significantly higher in the primary tumors of the brain metastasis patients than in the lesions of the Dukes C and Dukes D control groups (p=0.0004). Patients whose tumor was managed by radiosurgery survived longer than patients who had had whole brain radiation or those who had been left untreated.

Conclusion

The fact that NCAM expression was high in the primary tumors of brain metastasis patients suggests that the affinity of cancer cells to a particular organ is important for circulation-mediated metastasis. Controlling local tumors using radiosurgery is certainly going to play an important role in extending survival and improving the patient’s quality of life (QOL).

Keywords

NCAM expressionBrain metastasisColorectal cancerRadiosurgeryQOL

Introduction

Due to the advancement of cancer therapies and resultant prolongation of life expectancy, metastasis of cancers to other organs has become a frequent problem. In particular, brain metastasis results in neurological disorders and can end in death, if left untreated. The disease often progresses within a short period of time after the appearance of initial symptoms such as headache, nausea, and convulsive seizure. All types of cancers can spread to the brain by metastasis, and brain metastasis is found in an estimated 10–15% of patients who die of cancer. One important aspect is that once cancer metastasizes to the brain, the metastasized brain tumor becomes the direct cause of mortality in almost all cases. Therefore, the treatment of brain metastasis is considered very important.

Lung and breast cancers are the top two primary tumors that most frequently metastasize to the brain [1]. Most seriously, about 30% of lung cancer patients suffer from brain metastasis, and diagnosis and treatment of brain metastasis has become an important aspect of lung cancer therapy. Brain metastasis is thought to be less common in association with cancers of the digestive system, but once they metastasize to the lung, the risk of subsequent metastasis to the brain is high. The mechanism of brain metastasis is not well understood, but the affinity between cancerous cells and neuronal tissues may be involved in the process [2].

In this study, we examined the brain metastasis of patients with colorectal cancers. We focused our study on the involvement of neuronal affinity factor neural cell adhesion molecule (NCAM) and therapeutic parameters.

Patients and methods

Patients

We selected 17 cases of clinically identified brain metastasis among 1,077 colorectal cancer patients on whom surgery was performed in our hospital between 1979 and 1998. For the staining control of NCAM, we selected 56 Dukes C patients and 13 Dukes D patients who were hospitalized and surgically treated in 1997 and 1998. For all these patients, we have conducted more than 5 years of follow-up observations.

Histochemical analysis of NCAM

Using the labeled streptavidin biotin method (DAKO LSAB kit; DAKO Japan, Kyoto, Japan), the immunohistochemical staining of NCAM (B-A19; Invitrogen Japan, Tokyo, Japan) was performed on paraffin-embedded tumor specimens. One block containing the cancer tissue showing the deepest invasion was thin-sectioned serially to 5 μm after a review of H&E-stained slides of the surgical specimens. The sections were deparaffinized in xylene and rehydrated. After quenching endogenous peroxidase activity with 3% hydrogen peroxide, nonspecific binding was blocked by preincubation with normal bovine. Normal mouse IgG, diluted to an equivalent protein concentration, was used as a negative control in place of the primary antibody.

The mouse monoclonal anti-NCAM antibody, diluted 1:100 in phosphate-buffered saline (PBS), was supplemented by 0.1% bovine serum albumin and 5% normal horse serum. After incubation with biotinylated antimouse secondary antibody for 1 h, staining was performed. 3-Amino-9-ethyl carbazole was used as chromogen, and the slides were counter-stained with hematoxylin.

The staining status was evaluated by counting the frequency of immunoactive tumor cells in three fields (×200 magnification), each containing roughly 300 cells. As there is a slight heterogeneity between the invasive margin and the center of the tumor, we adopted the invasive margin to secure the stability, and the mean ratio of the staining among the three fields represents a positive score. This count was expressed using the following criteria: negative, less than 5%; positive, 5–50%; and strongly positive, more than 50% (Fig. 1). Analysis of the microscopy was performed by two observers who had no knowledge of patient status (HO and SN).
Fig. 1

Immunohistochemical staining of the primary colorectal cancer with brain metastasis with anti-neural cell adhesion molecule (NCAM) antibody (×200). a Negative, b positive, c strongly positive. Staining was seen at the membrane and within the cytoplasm of the tumor cells

DNA ploidy analysis

This method has been described elsewhere [3]. Briefly, two or three 50-μm thick sections were obtained from the same paraffin block. Sections were dewaxed with three changes and rehydrated through decreasing concentrations of ethanol. Nuclei were released by cytoplasmic digestion for 30 min at 37°C, with 0.5% Pronase E in physiological saline at pH 7.5. The release of nuclei was enhanced by intermittent vortex mixing and forcing through nylon mesh. The nuclei were stained with propidium iodide (PI). Using Epics Profile apparatus (Coulter, Hialeah, FL, USA), two cytograms and two histograms were obtained by measuring forward light scatter, PI integra, and PI peak. Aneuploidy was defined by excluding diploid cells.

Statistical analysis

Statistical analysis was conducted using the Kruskal–Wallis test. The survival rate was calculated using the Kaplan–Meyer method, and the difference in survival rate was evaluated by the log rank test. We considered statistical differences significant when the p value was less than 0.05.

Results

Clinical pathology of colorectal cancer patients with brain metastasis

The clinical pathology of the 17 brain metastasis patients is summarized in Table 1. These represent 1.6% of the total number of cases of surgically treated colorectal cancer. In all patients, primary tumors were localized to the sigmoid colon or rectum. The incidence of brain metastasis among rectosigmoid colon cancer was 2.1% (17 out of 795). The average age of the patients was 58.6 years, and the male-to-female ratio was 3.25 to 1. Lung and liver metastasis was found in 80% and 50% of the patients respectively. The patients were treated as follows: no treatment, 7 cases; whole brain radiation therapy, 4 cases; local therapy including radiosurgery, 6 cases.
Table 1

Clinical details of patients with brain metastasis. WB radiation whole brain radiation

Patient number

Age/gender

Initial location

Dukes staging

NCAM expression

Combined metastasis

Type of therapy

Differentiation

DNA ploidy

Interval from primary cancer (months)

Previous chemo

1

33, male

Rectum

B

+

Lung

None

Well

Aneuploid

45

Oral 5Fu

2

77, female

Sigmoid

C

++

Lung, liver

None

Mod

Diploid

72

Oral 5Fu

3

53, male

Rectum

C

Lung

Radiosurgery

Well

Diploid

57

5Fu infusion

4

49, female

Rectum

D

++

Lung

WB radiation

Moderate

Aneuploid

3

5Fu+MMC

5

40, male

Rectum

C

+

Lung, liver

WB radiation

Well

Aneuploid

21

5Fu+MMC

6

54, male

Rectum

C

+

Lung, bone

WB radiation

Moderate

Aneuploid

47

5Fu+MMC

7

55, male

Rectum

C

+

Lung, liver

Resection

Well

Diploid

49

5Fu+MMC

8

80, male

Sigmoid

C

++

Liver, bone

None

Moderate

Aneuploid

6

Oral 5Fu

9

63, male

Rectum

C

+

Lung

None

Moderate

Aneuploid

16

5Fu+MTX

10

66, female

Rectum

C

None

None

Poor

Aneuploid

11

Oral 5Fu

11

69, male

Rectum

D

++

None

Radiosurgery

Moderate

Aneuploid

9

5Fu+MTX

12

68, male

Sigmoid

D

+

Lung, liver

Radiosurgery

Well

Aneuploid

2

5Fu+MTX

13

57, male

Rectum

B

++

Lung, bone

None

Well

Diploid

55

Oral 5Fu

14

54, male

Sigmoid

D

+

Liver

WB radiation

Poor

Aneuploid

6

5FU+LV

15

57, male

Rectum

C

++

Lung, liver, bone

Radiosurgery

Moderate

Aneuploid

12

5FU+LV

16

67, male

Sigmoid

D

++

Lung, liver

Radiosurgery

Poor

Aneuploid

24

5FU+LV

17

55, female

Rectum

D

+

Lung

Radiosurgery

Well

Diploid

52

5FU+LV

The difference in NCAM expression

The expression of NCAM in cancer cells was demonstrated by diffuse staining in cytoplasm. Frequencies of NCAM-expressing cells were significantly higher in the primary tumors of the brain metastasis patients than in the lesions of the Dukes C and Dukes D control groups (p=0.0004; Table 2). There were no significant correlations between NCAM expression and DNA ploidy (data not shown).
Table 2

Summary of immunohistochemical findings for neural cell adhesion molecule (NCAM) with percentages in parentheses

 

Immunoreactivity*

+

++

Brain metastasis (17)

2 (11.7)

8 (47.1)

7 (41.2)

Dukes C (56)

27 (48.2)

27 (48.2)

2 (3.6)

Dukes D (13)

5 (38.5)

7 (5.8)

1 (7.7)

*Kruskal-Wallis test, p=0.0004

DNA ploidy analysis

The diploid/aneuploid pattern was 5 out of 12 in the brain metastasis group, 14 out of 42 in the Dukes C group, and 3 out of 10 in the Dukes D group. There were no significant differences in DNA ploidy between the brain metastasis patients and the Dukes C and Dukes D patients (Table 3).
Table 3

Summary of DNA ploidy pattern with percentages in parentheses

 

DNA ploidy

Diploid

Aneuploid

Brain metastasis (17)

5 (29.4)

12 (70.63)

Dukes C (56)

14 (25.0)

42 (75.0)

Dukes D (13)

3 (31.3)

10 (76.9)

Prognosis of brain metastasis patients

The median length of survival was 4.5 months. Patients whose tumor was managed by local therapy such as radiosurgery or surgery survived longer than those who were treated by whole brain radiation or those who were untreated (Fig. 2). There were no significant differences in survival according to the NCAM expression, DNA ploidy, time after operation, or tumor differentiation.
Fig. 2

Survival rates after diagnosis of brain metastasis from colorectal cancer according to the type of therapy

Discussion

Metastasized brain tumors represent about 12% of total brain tumors. The primary tumor is most frequently lung cancer (50%), followed by breast cancer (12%) [1]. Colorectal cancer is relatively rare as a primary tumor of metastasized brain cancer, and is found in only 2–3% of autopsy cases and 1–4% of clinically diagnosed cases. Although less frequent, it is still important to understand how colorectal cancer metastasizes to the brain. In order to understand the overall clinical and pathological features as well as to evaluate therapeutic approaches, this study reported on our initiative to analyze the brain metastasis of colorectal cancers that were treated in our hospital.

For the pathological analyses, we focused our examination on the expression of neural affinity factor, NCAM. NCAM is a glycoprotein purified by Edelman and colleagues in 1977 [4]. It is involved in the formation of the neuronal network and synapses between neurons and muscles [58]. The extracellular domain of NCAM contains five immunoglobulin-like motifs and two type-III fibronectin-like motifs. NCAM is expressed in nearly all neuronal cells, and their function as adhesion receptors of NK cells has also been identified. Linneman and colleagues reported that melanoma cells that have reduced levels of NCAM expression are more frequently metastasized through circulation [9].

Interestingly, the current study demonstrated that the NCAM expression level is significantly higher in the primary tumors of brain metastasis cases than in the tumors of advanced cancers without metastasis. In general, the metastasis pathway from colorectal cancer to the brain is thought to be:
  1. 1.

    Through circulation

     
  2. 2.

    Through the lymphatic system

     
  3. 3.

    Through neuronal and cerebrospinal fluid

     
Furthermore, Cascino and colleagues specifically listed the following three pathways as potential routes of metastasis through circulation [10]:
  1. 1.

    From the rectal venous plexus to the inferior vena cava

     
  2. 2.

    Through the Baston spinal venous plexus

     
  3. 3.

    Through the portal vein, liver, and lung

     
to reach the brain. In only 10% of our cases the metastasis was limited to the brain, suggesting that metastasis through the portal vein, and subsequently through the liver and lungs, could be the major pathway. The fact that NCAM expression was high in the primary tumors of brain metastasis cases suggests that the affinity of cancer cells to a particular organ is important for circulation-mediated metastasis.

The localization of the primary tumors was the distal colon in all 17 cases (rectum in 11 cases and sigmoid colon in 7 cases). In the analysis of 100 cases by Hammoud et al. [11], 65% were found to be localized to the rectum and sigmoid colon. The distal colon tends to be more frequently involved in metastasis in other reports as well. These observations suggest that distal colon cancer is metastasized to the lung not only by the portal system but also by the greater circulation system.

In our study, we found cancers in the lung (13 cases, 76%), liver (8 cases, 47%), and bone (4 cases, 23%) as complications. Two cases (12%) metastasized to the brain only. In terms of localization in the brain, 5 cases (29%) were found to be subtentorial or intracerebellar, while 13 cases (76%) showed multiple tumors in the brain (average 3.5 tumors per brain). In many reports, colorectal cancer patients with brain metastasis were reported to have a poorer prognosis shown by median length of survival and 1-year survival rate [1214]. Wronski and Arbit suggested that the causes of the poor prognosis are frequent metastasis to other organs and frequent metastasis to intracerebellar and subtentorial regions of the brain [15].

Our analysis of prognosis showed that the 50% survival duration was 4.5 months, which is an equally poor result compared with the durations reported by Wronski and Arbit [15] and Alden et al. [16] We showed the prognoses of metastasized brain tumor patients who were treated in our hospital using survival curves comparing various therapeutic methods. Due to the small number of cases, we cannot identify any statistically significant differences. Nevertheless, the prognosis appeared to be better using local therapies including radiosurgery than using whole brain radiation or no treatment at all.

Recently, an increasing number of patients have been choosing palliation by radiosurgery [17]. The advantages of this treatment are as follows:
  1. 1.

    Treatment of the tumor can be carried out without a craniotomy

     
  2. 2.

    It is less invasive and a lower risk operation compared with surgery

     
  3. 3.

    A shorter hospital stay is expected

     
Because of these advantages, further application of this technique is greatly anticipated.

Managing metastasized brain tumors requires careful consideration of many factors, including the patient’s quality of life (QOL). Thornton and Harsh [18] propose to determine the appropriate therapy based on an evaluation of coexisting metastasized tumors, the number and size of metastasized brain tumors, and accessibility to the tumor. Brain metastasis from colorectal cancer is the most advanced status of the disease, and all treatment options need to be carefully thought through. Many patients have a poor prognosis as they have cancerous tumors in other organs; it is, therefore, important to consider the balance between the treatment and prognosis on an individual basis. In this regard, controlling local tumors using techniques such as radiosurgery is certainly going to play an important role in extending survival and improving the patient’s QOL.

Copyright information

© Springer-Verlag 2004