S100� as a predictor of brain metastases: Brain versus cerebrovascular damage

817

S100␤ As a Predictor of Brain Metastases Brain versus Cerebrovascular Damage

Michael A. Vogelbaum, M.D., Thomas Masaryk, M.D.2,3 Peter Mazzone, M.D., M.P.H.4 Tarek Mekhail, M.D.4 Vincent Fazio, M.S.1,2 Sally McCartney, M.S., M.T.1 Nicola Marchi, Ph.D.1,2 Andrew Kanner, M.D.1,2 Damir Janigro, Ph.D.1,2

Ph.D.

1

1

Brain Tumor Institute, The Cleveland Clinic Foundation, Cleveland, Ohio.

2

Cerebrovascular Research Center, The Cleveland Clinic Foundation, Cleveland, Ohio.

3

Department of Radiology, The Cleveland Clinic Foundation, Cleveland, Ohio.

4

Department of Pulmonary Medicine and Hematology/Oncology, The Cleveland Clinic Foundation, Cleveland, Ohio.

BACKGROUND. The identification of brain metastases in patients with malignant disease has important implications for determining their treatment and prognosis. Asymptomatic metastatic brain tumors may be detected by surveillance imaging techniques, but longitudinal follow-up of patients who are at risk is sporadic primarily due to cost. Because the development of brain metastases is accompanied and detected by extravasation of contrast agents across the blood-brain barrier (BBB), the authors hypothesized that peripheral analysis of the BBB indicator S100␤ may be useful as a screening tool for brain metastases in patients who have no neurologic symptoms. METHODS. Thirty-eight patients were enrolled for the current study. All patients had newly diagnosed lung carcinoma and had no neurologic symptoms or known history of brain metastasis. Patients underwent an initial magnetic resonance imaging (MRI) scans and S100␤ blood tests. S100␤ tests were repeated in a subset of patients at the time of routine follow-up MRI scans. RESULTS. Based on imaging studies and on serum S100␤ analyses, the patients were divided in 3 categories: 1) patients with normal S100␤ levels (0.08 ⫾ 0.02 ␮g/L; n ⫽ 22 patients) and normal MRI scans; 2) patients with elevated S100␤ levels (0.5 ⫾ 0.28 ␮g/L; n ⫽ 8 patients) and pronounced microvascular changes on MRI scans but with no metastases; and 3) patients with elevated S100␤ levels (0.28 ⫾ 0.19 ␮g/L; n ⫽ 7 patients) and metastatic brain tumor(s) on MRI scans. CONCLUSIONS. Because of the significant overlap in S100␤ levels between patients with cerebral microvascular diseases and patients with brain metastases, the authors concluded that the serum S100␤ level may be used as a surveillance tool to predict or detect brain metastases if appropriate prescreening radiologic tests are obtained and if patients who are candidates for false-positive results are identified and excluded. Cancer 2005;104:817–24. © 2005 American Cancer Society.

KEYWORDS: blood-brain barrier, brain metastases, cerebral microvascular disease, lung carcinoma, S100B, vascular leakage.

T

Supported by NIH grants RO1 HL51614, RO1 NS43284, R01 NS0465NIH13, and NIH-RO1 NS38195 to D.J. Address for reprints: Damir Janigro, Ph.D., Cerebrovascular Research Center, Department of Neurological Surgery, NB20, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195; Fax: (216) 445-1466; E-mail: janigrd@ ccf.org Received December 3, 2004; revision received February 28, 2005; accepted March 22, 2005.

he identification of brain metastases in patients with newly diagnosed or established lung carcinoma has important implications for determining their prognosis and treatment options. Estimates of the incidence of brain metastases in these patients vary widely from about 2% to nearly 15%, depending on the stage of disease (ranging from newly diagnosed, limited disease to established, widespread disease).1,2 This feature contrasts with most of the rest of the body, where metastatic spread is much more common. Accordingly, there has been some controversy about the role of surveillance imaging in patients with lung carcinoma who do not have neurologic symptoms.3,4 Because of the high costs associated with a routine schedule of magnetic resonance image (MRI) studies over multiple years of surveillance, most clinicians opt instead only to obtain imaging stud-

© 2005 American Cancer Society DOI 10.1002/cncr.21220 Published online 21 June 2005 in Wiley InterScience (www.interscience.wiley.com).

818

CANCER August 15, 2005 / Volume 104 / Number 4

ies when indicated according to the presence of neurologic symptoms or as a requirement for entry into an investigational protocol. However, evidence indicates that the overall incidence of brain metastasis is rising in patients with lung carcinoma due to improved therapy for systemic disease.2 The treatment of brain metastases has improved over the past 3 decades to the point where a patient with a limited number of brain metastases has a prognosis that is more dependent on the status of their extracranial disease than on the presence of brain metastases.5 Furthermore, brain metastases are treated more effectively when they are detected early; consequently, there is a premium on the discovery of brain metastases before they produce neurologic symptoms. Loss of blood-brain barrier (BBB) function is a hallmark of many neurologic diseases. Invasion is the process by which a tumor gains access to the local tissue around it and gradually replaces it. This step is the first in the metastatic process by which tumor cells seed to remote sites.6,7 Consequent neovascularization produces a tumor vascular system that does not have a fully functional BBB. For diagnosis as well as planning and evaluation of therapy, imaging techniques, such as computed tomography (CT) and contrast-enhanced MRI (CE-MRI), are invaluable tools for detecting lesions that have a disrupted BBB. In CEMRI, contrast enhancement of tumor tissue is secondary to extravasation and subsequent accumulation in the interstitium of small contrast agents, such as gadolinium diethylenetriaminepenta-acetic acid (GdDTPA). Tumor vessels generally are leakier than normal vessels and, thus, will permit faster extravasation.7–9 Hence, dynamic CE-MRI measurements allow identification of suspected malignancies by highlighting regions with increased rates of wash-in and wash-out of contrast agents. An alternative approach for the detection of BBB abnormalities was developed recently that relies on the detection of changes in a serum marker that indicates BBB disruption. This marker, S100␤, is a protein that is found in astrocytes and is released into serum only when the BBB is breached.9 –13 S100␤ primarily is synthesized in the brain by the endfeet process of the astrocytes and is released quickly from the brain in the blood when the BBB is disrupted.13–16 S100␤ also has been found in other tissues but at lower concentrations.17–19 Although the appearance of S100␤ in plasma correlated well with BBB openings, it has been shown that S100␤ increases in plasma, cerebrospinal fluid, or both as a consequence of other pathologies that are not limited to the central nervous system (CNS). S100␤ also may detect brain damage or may

indicate advanced metastasis in patients with melanoma.20 –24 The fact that S100␤ can increase in serum independent of brain (or neuronal) damage was demonstrated indirectly in a study of the effects of boxing and other high cardiovascular output activities on the levels of S100␤ in serum. It is noteworthy that, in that study, a significant increase in S100␤ was observed in the serum of individuals who undertook activities that involved repetitive, jarring movement or contact to the head (such as boxing, sparring, running, and jogging); however, essentially, no increase was observed in individuals who exerted themselves through exercise on a stationary bicycle. Clearly, these activities neither caused nor promoted brain damage, but the rise in S100␤ protein in running activities may be due to astroglial activation, astroglial destruction, BBB disruption, or a combination of the 3. This finding also indicates that the source of S100␤ was not influenced greatly by secretion of the protein from extracranial tissue. We hypothesized that, if BBB leakage is a hallmark of brain metastasis, then peripheral markers of BBB function may be useful, noninvasive, and inexpensive tools to confirm or rule out brain invasion by highly malignant tumors. To test this, we measured S100␤ in serum from patients who presented with lung carcinoma, a malignancy that is characterized by a relatively high propensity for CNS metastasis. A corollary of this study was to further examine the usefulness of serum markers as an alternative to enhancementbased neuro-imaging.

MATERIALS AND METHODS This prospective study included patients who presented to the Cleveland Clinic with newly diagnosed lung carcinoma. Potential candidates for this study were identified by a pulmonologist (P.M.) or a medical oncologist (T.M.). To be included in the study, patients had to have newly diagnosed nonsmall cell lung carcinoma with no neurologic symptoms or signs and with no known brain metastases. Many patients with newly diagnosed lung carcinoma at the Cleveland Clinic undergo screening MRIs as part of their initial routine evaluation. All patients signed an informed consent according to institutional review protocols at The Cleveland Clinic Foundation. Radiologic examination was performed within 3 weeks of the time of when the blood sample was obtained. These patients underwent diagnostic or volumetric MRI studies at The Cleveland Clinic Foundation in 2003–2004. S100␤ was measured with an enzyme-linked immunosorbent assay, as described elsewhere.25 Routine MRI studies included T1-weighted sequences with and without gadolinium (Gd-DTPA), T2-weighted im-

S100␤ and Brain Metastases/Vogelbaum et al.

ages, and fluid-attenuated inversion recovery (FLAIR) images. Contrast-enhanced, 3-dimensional volume acquisition with 1-mm slice interval and supplementary 2-mm spin echo images through the area of interest were obtained. All MRIs were analyzed for contrast enhancement.

Classification of Imaging Studies Routine MRI studies included T1-weighted sequences with and without gadolinium (Gd-DTPA), T2-weighted sequences, and FLAIR images. Scan results were interpreted by a neuroradiologist, and routine paradigms were used to determine the presence or absence of vascular changes.

Statistical Analysis Analyses of variance were used to determine statistical significance. A difference with P ⬍ 0.05 was considered statistically significant.

RESULTS We obtained blood samples from 38 patients during their visits to the Cleveland Clinic Foundation. No patients underwent more than one blood draw or imaging study. A summary of the clinical characteristics, imaging results, and S100␤ levels are listed in Table 1. First, we explored the dependency of serum S100␤ on patient age or body weight. The results are shown in Figure 1A,B. No significant correlation was found between S100␤ and body weight or patient age (P ⫽ 0.54 and P ⫽ 0.86, respectively). Similarly (Fig. 1C), S100␤ levels were not dependent on gender (P ⬎ 0.3). Analysis of the imaging studies revealed that the patient population could be divided in three groups. Approximately 58% of patients (22 of 38 patients) had no sign on initial MRI or CT studies, suggesting the presence of brain metastasis or chronic cerebrovascular disease (“normal”). Approximately 24% of patients (9 of 38 patients) had no metastasis but had significant chronic vascular changes on their initial imaging studies (“cerebral microvascular changes”). The remaining patients (7 of 38 patients; 18%) had at least 1 brain metastasis on their initial imaging study (“metastasis”). Analysis of serum S100␤ levels showed that patients who had normal imaging studies had lower levels of this serum marker of BBB function compared with patients who had imaging studies that showed cerebral microvascular changes or metastasis (Fig. 1D) (mean normal, 0.07 ␮g/mL; mean vascular, 0.49 ␮g/ mL; mean metastasis, 0.28 ␮g/mL). However, no significant difference was observed in the degree of serum S100␤ elevation in the vascular change group and

819

the metastasis group (P ⬎ 0.10). Figure 1E shows the distribution of S100␤ values for each imaging group. Based on previous experience and published work,10,13 we assumed S100␤ levels of 0.12 ␮g/mL as the upper limit for normal; this is indicated by the shaded area in Figure 1E. Figure 2 shows examples of patients’ MRIs and their associated serum S100␤ levels. Figure 2A shows a representative MRI from a patient who was classified with cerebral microvascular changes but who did not have a brain metastasis. This patient’s S100␤ level was 0.21 ␮g/L, which is considered elevated. Figure 2B shows a representative MRI from a patient who had no evidence of vascular disease, but this patient also had an elevated S100␤ level (0.46 ␮g/L). This scan initially was read as normal, but subsequent examination revealed the presence of a small metastatic lesion (see Fig. 2B, circled area). Based on these results, we calculated the positive predictive value (PPV) and negative predictive value (NPV) of an S100␤ level ⬎ 0.12 ␮g/mL on the presence of a brain metastasis. The NPV was 1.00; no patient with a “normal” S100␤ level was had either brain metastasis or cerebral microvascular changes. The PPV of an elevated S100␤ level was 0.471 for the entire series or 0.875 if the patients who had a cerebral microvascular pattern were excluded. The fi (fitness of analysis ⫽ PPV ⫻ NPV ⫻ 1000) was 471 if patients with vascular disease were grouped together with patients who had metastasis and 875 if the patients with vascular disease were excluded. The sensitivity of the S-100␤ ELISA, as reported by Diasorin, is 0.03 ␮g/L (B0 ⫹ 3 SD). The intra-assay variation of the assay was calculated by ANOVA to be ⬍ 10%, with an inter-assay variation of ⬍ 15% in the range of concentration from 0.18 to 4.0 ␮g/L. The main finding of our study is that the elevated, 100% negative predictive value of serum S100␤ warrants is further evaluation and use as an alternative to contrast-based MRI scans routinely used to evaluate BBB integrity and possible CNS metastases in lung carcinoma patients. This finding, together with previous reports on other patients affected by neurological disorders or brain tumors9,10,13,26,27 suggest that a simple blood test may be used to screen patients prior to more extensive, and expensive, MRI evaluations.

DISCUSSION Many neurologic disorders and lesions are associated with increased BBB permeability, including primary and metastatic brain tumors, ischemia, hypertension, dementia, epilepsy, infection, multiple sclerosis, and trauma.28 The BBB is composed primarily of microvascular endothelial cells linked by tight junctions,

820

CANCER August 15, 2005 / Volume 104 / Number 4

TABLE 1 Summary of the Clinical Characteristics, Imaging Results and S100␤ Levels in Patients with Newly Diagnosed Lung Carcinoma Patient ID

S100␤ (␮g/L)

Gender

Weight (kg)

Age (yrs)

History

Metastasis group LT112702-1130 LT120502-1300 LT123102-0830 LT020303-1445 LT040903-1530 LT041403-1230 LT100203-1430

0.71 0.15 0.29 0.13 0.22 0.26 0.21

M M F M M M F

67.7 85.3 37.1 81.7 75.0 85.0 95.2

72.0 52.9 70.3 75.2 67.6 65.1 65.8

Newly diagnosed, extensive-stage small cell lung CA, left shoulder weakness Malignant neoplasm bronchi/lung; anxiety, numbness, and weakness in right leg Other lung disease, lung mass.; sided chest pain, diverticulitis, hysterectomy, seizures during pregnancy. Nonsmall cell lung CA, hypertension, carotid stenosis, hypothyroidism, abnormal gait Other lung disease, lung mass, hypertension, high lipids, gout, diabetes mellitus, cataracts Other lung disease, osteoarthritis, diabetes mellitus (type 2), hypertension, atrial fibrillation, hyperlipidemia Nonsmall cell CA, lobectomy (adenocarcinoma 6/2002; hysterectomy, 1978), pulmonary nodules, dry cough

Vascular group LT112602-1245 LT121202-0930 LT041603-1400 LT062303-1230

1.08 0.52 0.32 0.35

M M M F

67.1 83.9 124.0 72.0

59.3 69.5 79.3 66.6

LT100603-1130 LT102403-1630

0.68 0.48

F F

65.6 64.7

66.3 68.5

LT110603-1130 LT020504-1240

0.47 0.14

M M

95.3 86.9

66.0 74.3

LT020604-1345

0.34

M

86.9

72.2

Primary lung CA with lymphatic spread, upper GI bleed, hiatal hernia, hypertension, tuberculosis Benign hypertension, coronary atherosclerosis, MI, hyperlipidemia, malignant neoplasm bronchi/lung Nonsmall cell lung CA, hernia repair, partial thyroidectomy, hypertension, glaucoma, osteoarthritis, hemoptysis Stage IIIA nonsmall cell lung CA, chronic sinusitis, difficulty swallowing, colostomy (25 yrs ago), pneumonia, RU lung lesion. Nonsmall cell lung CA, carotid endardectomy, asthma, fever with palpitations, cataracts, and hypertension Squamous cell CA (10/2003), pneumonia, SVT ablation, diverticulitis, hematuria, hysterectomy, Legionnaire disease Nonsmall cell lung CA, shortness of breath, hypertension, stroke, knee surgery, noninsulin-dependent diabetes Bronchogenic malignancy, coronary artery bypass grafting, hepatitis B, hyperlipidemia, mitral valve disorder, atrial fibrillation, bloody sputum Malignant neoplasm bronchi/lung, heart transplant (1994), congestive heart failure, chronic renal insufficiency, bacterial pneumonia, neoplasm larynx, small vessel disease

Normal group LT111402-1330 LT111802-0915 LT111902-1425

⬍ 0.02 0.02 0.07

M F M

73.0 55.8 99.2

80.3 41.5 65.7

LT112602-0900 LT120302-1230 LT120602-1230 LT121602-1440 LT012003-1420 LT012703-1140

0.10 0.08 ⬍ 0.02 ⬍ 0.02 0.10 0.12

M F F F M F

98.4 58.4 52.2 70.1 77.7 67.6

83.1 78.6 49.5 58.6 75.0 74.8

LT021303-1430 LT031803-1500

0.11 ⬍ 0.02

M F

103.0 68.0

59.5 58.5

LT040203-1230

0.06

M

176.0

61.4

LT050503-1430

⬍ 0.02

M

82.3

73.2

LT051203-1020 LT060503-1400 LT062503-1200 LT062703-1430

0.03 0.03 0.04 0.06

M M F M

87.0 86.0 117.0 72.3

78.1 66.3 56.9 85.4

LT070103-1600 LT071703-0930

0.59 0.05

M M

95.9 93.8

62.1 61.7

LT072503-1400

⬍ 0.02

F

57

76.5

LT082603-1500 LT082803-1200

0.03 0.06

M M

96.6 80.1

63.3 74.5

Other lung disease (NEC), aortocoronary bypass, MI, hyperlipidemia Nonsmall cell lung CA (surgery 4/2002), lung CA malignant neoplasm bronchi/lung Malignant neoplasm bronchi/lung, diabetes mellitus (type 2), COPD, bladder tumor, high lipids, cardiothoracic bypass surgery Nonsmall-cell lung CA, MI (1998), angioplasty, diabetes mellitus (type 2), hypertension, white lung Other lung disease (NOC), hypertension, chronic hematuria, fibromyalgia, pneumonia, upper lobe lung nodule Nonsmall cell lung CA, ovarian cyst removed, arthroscopic knee surgery, appendectomy, joint pain Bronchogenic malignancy, left upper lobe lung mass, breast lesion Adenocarcinoma of 1 upper lobe, hernia, deafness, arthritis, asthma, pulmonary TB Chronic airway obstruction, chronic heart failure, asthma, benign hypertension, pulmonary nodules, motor vehicle collision (1995) Newly diagnosed lung CA, CAD, lower GI bleeding, erectile dysfunction, hyperlipidemia Other lung disease, partial hysterectomy (1991), knee surgery, tonsillectomy, chest pain, right side chest pleurisy, abnormal chest X-ray. Metastatic nonsmall cell lung CA, Hodgkin disease, neuropathy, endarterectomy, upper lobe lesion, bilateral adrenal nodules Other lung disease, peptic ulcer, diabetes mellitus, TB, hemorrhoids, hypertension, cataracts, skin CA, nocturia, weight loss, lung mass Lung CA, gall bladder surgery, chest discomfort, abnormal X-ray Other lung disease, cardiac disease, atrial fibrillation, coronary atherosclerosis, hemoptysis Malignant neoplasm bronchi/lung, hypertension, hypothyroid, cholecystectomy, shortness of breath, coughing Other lung disease, prostatectomy (benign), CAD with bypass, abdominal aortic aneurysm, hypertension, hypothyroidism Squamous cell CA, angioplasty, osteoarthritis, RU lobe lung lesion Other lung disease, hypertension, high lipids, kidney trauma and repair, appendectomy, fatigue, lung lesions, liver lesions Malignant neoplasm bronchi/lung, incontinence, partial hysterectomy, vein stripping, foot neuroma, hemoptysis, chest wall pain Chest swelling, mass/lump, hypertension, CAD Other lung disease, hyperplasia of prostate, varicose veins (leg), hyperlipidemia, benign neoplasm of large bowel, osteoporosis, chest pain

ID: identification; M: male; F: female; CA: carcinoma; GI: gastrointestinal; MI: myocardial infarction; RU: right upper; SVT; superficial vein thrombosis; NEC: necrotizing enterocolitis; COPD: chronic obstructive pulmonary disease; NOC: not otherwise classified; TB: tuberculosis; CAD: coronary artery disease.

S100␤ and Brain Metastases/Vogelbaum et al.

821

FIGURE 1. S100␤ levels in the patients enrolled in this study. (A) No significant correlation was found between S100␤ levels and patient age. (B,C) No significant correlations were found between S100␤ levels and body weight (Wt) (B) or between S100␤ level and gender (C). (D) Three groups of patients could be distinguished radiologically (for imaging studies, see Fig. 2). A statistically significant difference was found between S100␤ levels in patients who had with normal magnetic resonance imaging/computed tomography findings and patients who were diagnosed either with metastatic brain tumor (Metastasis) or with vascular changes (Vascular). No significant difference was found between the two latter groups. Single asterisk, P ⬍ 0.05; double asterisks, P ⬍ 0.005 (analyses of variance). (E) This is a box-plot diagram of the data shown in D (max: maximum; min: minimum). The curves fitting the data points were fitted by using a Lorentzian fitting routine in Origin 6.0 software.

which largely prevent molecular communication between blood and the brain. Astrocytes and their processes invest ⬎ 90% of endothelial capillaries, and their endfeet are projected tightly around the endothelial cells.29 This relation between the astrocytes and endothelial cells is essential for the formation of a fully functional BBB. Astrocytic proteins are synthesized and released next to capillaries; however, due to the negligible transendothelial permeability to proteins, they extravasate into the serum only when the BBB is breached. Under these conditions, S100␤ is expected to appear in the systemic circulation.9 –13,30 –34 Although several studies clearly demonstrated that serum S100␤ is a marker of BBB function, others demonstrated a positive correlation with brain damage.35–37 How could these seemingly contrasting two findings be explained? Modeling of these 2 phenomena predicted that high levels of serum S100␤ would

be correlated with brain damage, whereas lesser increases above normal values would be associated with BBB leakage in the absence of parenchymal damage.13,30 The results of this prospective study confirmed the presence of a correlation between imaging evidence of a loss of BBB integrity and elevated levels of serum S100␤. Furthermore, these results suggest that serum S100␤ may be used to screen for brain metastases in a certain group of patients who have a known systemic malignancy. Because no patient with cerebral metastases was missed by the test (NPV, 100%), although nearly 50% of the whole sample had normal values, we propose the use of S100␤ as a screening test for the presence of asymptomatic cerebral metastases in patients who have a malignancy that has a known predilection for the development of brain metastases. An elevated S100␤ level did not always correlate

822

CANCER August 15, 2005 / Volume 104 / Number 4

FIGURE 2. These radiologic imaging studies from typical patients exemplify the criteria used to construct the data illustrated in FIGURE 1. Scans from normal participants are not shown. (A–C) These magnetic resonance images (MRI) and computed tomography (CT) scans were obtained from patients with small metastases who had serum S100␤ levels of 0.21 ␮g/L (A), 0.46 ␮g/L (B), and 0.34 ␮g/L (C) at the time of the imaging studies (Gd: gadolinium). (D) These MRI studies were obtained from a patients with obvious, large metastases who had a serum S100␤ level of 0.71 ␮g/L at the time of the imaging studies. (E,F) These MRI studies were obtained from patients who had vascular leakage without metastases and who had serum S100␤ levels of 0.68 ␮g/L (E) and 0.35 ␮g/L (F) at the time of the imaging studies (for details, see text).

with the presence of brain metastasis, however. In the current series of patients, an elevated S100␤ level was associated either with brain metastasis or with the presence of imaging changes suggestive of chronic, diffuse cerebral microvascular disease. This finding makes the use of S100␤ as a sole marker for the detection of brain metastasis problematic. Clearly, patients who have an elevated S100␤ level will need to

undergo neuroimaging studies to confirm the diagnosis of brain metastasis. Patients who have evidence of chronic cerebrovascular disease likely will receive no further benefit from routine screening of their serum S100␤ level and, instead, will need to be followed clinically. Further study of this group of patients may reveal that the onset of a brain metastasis is associated with a further elevation in the absolute level of their

S100␤ and Brain Metastases/Vogelbaum et al.

serum S100␤, although this remains to be determined. Conversely, patients who have normal MRI studies in the face of an elevated serum S100␤ levels may benefit from closer follow-up imaging studies. We are investigating other serum markers actively that may distinguish between the two groups described here (patients with cerebral microvascular and patients with brain metastasis). Early identification of brain metastases, before neurologic symptoms or signs develop, provides patients with a wider set of treatment options and likely will enhance their quality of life. Small, asymptomatic brain metastases most often can be treated successfully with stereotactic radiosurgery and, in certain circumstances, without the need for whole-brain irradiation. It has been shown that aggressive treatment of brain metastases improves overall survival5 to the point where prognosis is more dependent on the status of the extracranial disease. It has to be underscored that S100␤ had a 100% negative predictive accuracy. Thus, all patients who presented with brain metastasis had elevated S100␤ values. In one patient (see Fig. 2A), S100␤ was increased above normal values, although the MRI scans revealed metastatic lesions only after careful examination. In another patient, a negative scan accompanied by an elevated S100␤ level was followed by the development of a metastasis, suggesting that blood tests sometimes may be more accurate than radiologic screening. One possible application of the test could be envisioned as follows: After the diagnosis of a systemic tumor with a demonstrated propensity toward CNS metastasis but with no obvious metastases at that time, a first S100␤ test is performed. A positive result above normal values is interpreted as a sign of cerebrovascular disease consistent with BBB leakage. A negative S100␤ value, conversely, is an indicator of an intact BBB. Both signals are used as a baseline, and additional, repetitive tests are used to detect a departure from the patient’s “S100␤ score.” This event is interpreted as a sign of possible metastatic growth that will be confirmed or disproved by a radiologic examination. This approach may optimize the timing of radiologic examination, increasing the possibility of the early detection of CNS changes consistent with the spread of systemic carcinoma to the brain. Additional prospective studies currently are being designed to evaluate this possibility. In conclusion, the current results demonstrate that detection in the serum of the astrocytic protein S100␤ is a potential candidate protein for the nonradiologic, inexpensive screening of brain metastasis in lung carcinoma patients. We propose that S100␤ be

823

used as a screen to indicate who should receive an MRI scan for brain metastases. This conclusion is based on the fact that S-100␤ had a negative predictive value of 100% in this study and in our preliminary investigations published elsewhere.10

REFERENCES 1.

2. 3.

4. 5.

6.

7. 8. 9.

10.

11.

12.

13.

14.

15.

16. 17.

Walker AE, Robins M, Weinfeld FD. Epidemiology of brain tumors: the national survey of intracranial neoplasms. Neurology. 1985;35:219 –226. Posner JB, Chernik NL. Intracranial metastases from systemic cancer. Adv Neurol. 1978;19:579 –592. van de PM, van Oosterhout AG, Wilmink JT, ten Velde GP, Twijnstra A. MRI in detection of brain metastases at initial staging of small-cell lung cancer. Neuroradiology. 1996;38: 207–210. Pugatch RD. Radiologic evaluation in chest malignancies. A review of imaging modalities. Chest. 1995;107:294S–297S. Andrews DW, Scott CB, Sperduto PW, et al. Whole brain radiation therapy with or without stereotactic radiosurgery boost for patients with one to three brain metastases: Phase III results of the RTOG 9508 randomised trial. Lancet. 2004; 363:1665–1672. Subramanian A, Harris A, Piggott K, Shieff C, Bradford R. Metastasis to and from the central nervous system—the ‘relatively protected site.’ Lancet Oncol. 2002;3:498 –507. Rapoport SI. Blood-brain barrier in physiology and medicine. New York: Raven Press, 1976. Evelhoch JL. Key factors in the acquisition of contrast kinetic data for oncology. J Magn Reson Imaging. 1999;10:254 –259. Fazio V, Bhudia SK, Marchi N, Aumayr B, Janigro D. Peripheral detection of S100beta during cardiothoracic surgery: what are we really measuring? Ann Thorac Surg. 2004;78:46 – 52. Kanner AA, Marchi N, Fazio V, et al. Serum S100beta: a noninvasive marker of blood-brain barrier function and brain lesions. Cancer. 2003;97:2806 –2813. Kapural M, Krizanac-Bengez L, Barnett G, et al. Serum S-100beta as a possible marker of blood-brain barrier disruption. Brain Res. 2002;940:102–104. Marchi N, Fazio V, Cucullo L, et al. Serum transthyretin as a possible marker of blood-to-CSF barrier disruption. J Neurosci. 2003;23:1949 –1955. Marchi N, Rasmussen PA, Kapural M, Fazio V, Cavaglia M, Janigro D. Peripheral markers of brain damage and bloodbrain barrier dysfunction. Restorative Neurol Neurosci. 2003; 21:109 –121. Buccoliero AM, Caldarella A, Noccioli B, Fiorini P, Taddei A, Taddei GL. Brain heterotopia in pharyngeal region. A morphological and immunohistochemical study. Pathol Res Pract. 2002;198:59 – 63. Dyck RH, Van Eldik LJ, Cynader MS. Immunohistochemical localization of the S-100 beta protein in postnatal cat visual cortex: spatial and temporal patterns of expression in cortical and subcortical glia. Brain Res Dev Brain Res. 1993;72: 181–192. Mercier F, Hatton GI. Immunocytochemical basis for a meningeo-glial network. J Comp Neurol. 2000;420:445– 465. Jonsson H, Johnsson P, Alling C, Backstrom M, Bergh C, Blomquist S. S100B after coronary artery surgery: release pattern, source of contamination, and relation to neuropsychological outcome. Ann Thorac Surg. 1999;68:2202–2208.

824

CANCER August 15, 2005 / Volume 104 / Number 4

18. Mrak RE, Flanigan S, Collins CL. Malignant acoustic schwannoma. Arch Pathol Lab Med. 1994;118:557–561. 19. Rickmann M, Wolff JR. Modifications of S100-protein immunoreactivity in rat brain induced by tissue preparation. Histochem Cell Biol. 1995;103:135–145. 20. Brochez L, Naeyaert JM. Serological markers for melanoma. Br J Dermatol. 2000;143:256 –268. 21. Chakrabarty A, Franks AJ. Meningioangiomatosis: a case report and review of the literature. Br J Neurosurg. 1999;13: 167–173. 22. Grocott HP, Arrowsmith JE. Serum S100 protein as a marker of cerebral damage during cardiac surgery. Br J Anaesth. 2001;86:289 –290. 23. Ingebrigtsen T, Romner B, Marup-Jensen S, et al. The clinical value of serum S-100 protein measurements in minor head injury: a Scandinavian multicentre study. Brain Injury. 2000;14:1047–1055. 24. Jonsson H, Johnsson P, Alling C, Westaby S, Blomquist S. Significance of serum S100 release after coronary artery bypass grafting. Ann Thorac Surg. 1998;65:1639 –1644. 25. Georgiadis D, Berger A, Kowatschev E, et al. Predictive value of S-100beta and neuron-specific enolase serum levels for adverse neurologic outcome after cardiac surgery. J Thorac Cardiovasc Surg. 2000;119:138 –147. 26. Biberthaler P, Mussack T, Wiedemann E, et al. Rapid identification of high-risk patients after minor head trauma (MHT) by assessment of S-100␤ ascertainment of a cut-off level. Eur J Med Res. 2002;7:164 –170. 27. Biberthaler P, Mussack T, Wiedemann E, et al. Elevated serum levels of S-100␤ reflect the extent of brain injury in alcohol intoxicated patients after mild head trauma. Shock. 2001;16:97–101. 28. Grant GA, Janigro D. The blood-brain barrier. In: Winn HR, editor. Youmans neurological surgery, vol 1. Philadelphia: Saunders, 2004:153–174. 29. Emmi A, Wenzel HJ, Schwartzkroin PA, et al. Do glia have

30.

31.

32.

33.

34.

35.

36.

37.

heart? Expression and functional role for ether-a-go-go currents in hippocampal astrocytes. J Neurosci. 2000;20:3915–3925. Marchi N, Cavaglia M, Bhudia S, Hallene K, Janigro D. Peripheral markers of blood-brain barrier damage. Clin Chim Acta. 2004;342:1–12. Janigro D, Fazio V, Cucullo L, Marchi N. Markers of BBB damage and inflammatory processes. Presented at the 32nd annual meeting, Society for Neuroscience, Orlando, Florida, November 2–7, 2002. Ding M, Haglid KG, Hamberger A. Quantitative immunochemistry on neuronal loss, reactive gliosis and BBB damage in cortex/striatum and hippocampus/amygdala after systemic kainic acid administration. Neurochem Int. 2000; 36:313–318. Herrmann M, Vos P, Wunderlich MT, de Bruijn CH, Lamers KJ. Release of glial tissue-specific proteins after acute stroke: a comparative analysis of serum concentrations of protein S-100B and glial fibrillary acidic protein. Stroke. 2000;31: 2670 –2677. D’Ambrosio R, Maris DO, Grady MS, Winn HR, Janigro D. Selective loss of hippocampal long-term potentiation, but not depression, following fluid percussion injury. Brain Res. 1998;786:64 –79. Grocott HP, Arrowsmith JE. Serum S100 protein as a marker of cerebral damage during cardiac surgery. Br J Anaesth. 2001;86:289 –290. Ingebrigtsen T, Waterloo K, Jacobsen EA, Langbakk B, Romner B. Traumatic brain damage in minor head injury: relation of serum S-100 protein measurements to magnetic resonance imaging and neurobehavioral outcome. Neurosurgery. 1999;45:468 – 475. Romner B, Ingebrigtsen T, Kongstad P, Borgesen SE. Traumatic brain damage: serum S-100 protein measurements related to neuroradiological findings. J Neurotrauma. 2000; 17:641– 647.