CLINICAL STUDIES

MULTILOCULATED PYOGENIC BRAIN ABSCESS: EXPERIENCE IN 25 PATIENTS

Thung-Ming Su, M.D.
Department of Neurosurgery, Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan

Chu-Mei Lan, M.S.N.

College of Nursing, Kaohsiung Medical University, Kaohsiung, Taiwan

Yu-Duan Tsai, M.D.

Department of Neurosurgery, Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan

Tao-Chen Lee, M.D.

Department of Neurosurgery, Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan

Cheng-Hsien Lu, M.D.

Department of Neurology, Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan

Wen-Neng Chang, M.D.

Department of Neurology, Chang Gung Memorial Hospital, Kaohsiung Medical Center, Kaohsiung, Taiwan Reprint requests: Cheng-Hsien Lu, M.D., Department of Neurology, Chang Gung Memorial Hospital, 123, Ta Pei Road, Niao Sung Hsiang, Kaohsiung Hsien, Taiwan. Email: chlu99@ms44.url.com.tw Received, June 12, 2002. Accepted, January 9, 2003.

OBJECTIVE: To report our experience in treating multiloculated pyogenic brain abscess and determine whether there are differences in the bacteriology, predisposing factors, treatment choices, and outcomes between multiloculated and uniloculated brain abscesses. METHODS: We studied clinical data collected during a 16-year period from 124 patients with pyogenic brain abscess, including 25 cases of multiloculated abscess. RESULTS: The incidence of multiloculated brain abscess was 20%. In these 25 patients, hematogenous spread from a remote infectious focus was the most common cause of infection, as it was for the cases of uniloculated abscess. Headache and hemiparesis were the most common symptoms in patients with multiloculated abscess. In patients with uniloculated abscess, fever was the most common symptom. Viridans streptococci were the most commonly isolated pathogens. Bacteroides fragilis was the most common anaerobe in multiloculated abscess, and aerobic gram-negative bacilli were the most common pathogens in patients with uniloculated abscess. Of the patients with multiloculated abscess, 21 were treated surgically and 4 were treated with antibiotics only. Overall, eight patients (38%) needed another operation because of abscess recurrence after the initial operation. In uniloculated abscess, the rate of abscess recurrence after initial surgery was 13.1%. Mortality was 16% in multiloculated abscess and 17.1% in uniloculated abscess. CONCLUSION: Multiloculated abscesses accounted for 20% of our patients with pyogenic brain abscess. Excision seems to be the more appropriate surgical choice in multiloculated abscess. Prognosis for patients with multiloculated abscess can be as good as that for patients with uniloculated abscess. However, clinicians must carefully monitor these patients because the possibility of recurrence after surgery is significantly higher in patients with multiloculated abscess than in those with uniloculated abscess.
KEY WORDS: Aspiration, Excision, Multiloculated brain abscess
Neurosurgery 52:1075-1080, 2003
DOI: 10.1227/01.NEU.0000057696.79800.1D

www.neurosurgery-online.com

acterial brain abscess continues to constitute one of the most important neurosurgical emergencies. Since the advent of computed tomography, multiloculated brain abscess has been noted to be a special morphology of brain abscesses. Only a few reports have described the relative incidence of multiloculated brain abscess (4, 6, 7). Nonetheless, these studies did not specifically analyze its bacteriology, predisposing factors, clinical presentations, and therapeutic outcome. To our knowledge, only one report in the literature has focused on multiloculated brain abscess (12). To add to this body of knowledge, we report our experience treating

25 patients with multiloculated brain abscess from 1986 to 2001, including clinical presentation, bacteriology, underlying cause, and therapeutic outcome. We also attempt to determine whether there are differences in the bacteriology, predisposing factors, treatment choices, and outcomes between multiloculated and uniloculated brain abscesses among our patients.

PATIENTS AND METHODS

During a 16-year period (19862001), 124 patients with bacterial brain abscess were treated at Kaohsiung Chang Gung Memorial

NEUROSURGERY

VOLUME 52 | NUMBER 5 | MAY 2003 | 1075

SU

ET AL.

Hospital. Diagnosing criteria for pyogenic brain abscess were: 1) classic clinical manifestations, including headache, fever, localized neurological signs, and/or altered consciousness; 2) computed tomographic and/or magnetic resonance imaging scans revealing characteristic findings of brain abscesses; and 3) evidence of brain abscess observed during surgery or histopathological examination. Multiloculated abscess was defined as an abscess with multiple daughter loculations but without intervening brain parenchyma between these loculations (Fig. 1). Specimens obtained from patients were cultured for aerobic and anaerobic bacteria, mycobacteria, and fungi. Patients were considered to have mixed infections if at least two bacterial organisms were isolated from the initial cultures. In patients with negative cultures, brain abscess was diagnosed according to the classic clinical and neuroradiographic findings and good therapeutic response to combined surgical and antibiotic treatment. For each patient, the presenting symptoms, predisposing factors of infection, site of abscess, therapeutic methods, and outcome were retrospectively reviewed. Predisposing factors included hematogenous spread from a remote infection, contiguous infection from a parameningeal focus, neurosurgical events (head injury or neurosurgical procedures), and unknown factors. Neuroradiographic studies were reviewed carefully, and the number, size, location, and mass effect of brain abscesses were recorded. Mass effect was defined as positive when the abscess caused midline shift or compression of the ventricular system as revealed by imaging. Therapeutic choice was judged according to clinical status, neuroradiographic findings, and therapeutic response. Surgical treatment consisted of either aspiration or excision. Aspiration was defined as aspiration of the contents of the abscess with a ventricular needle or cannula via burr hole or small craniotomy, leaving the capsule intact. Excision was defined as craniotomy and resection of the abscess. In patients with positive cultures, the choice of antibiotic therapy was based on susceptibility tests. If all cultures were negative, antibiotics were selected empirically. Regular neuroradiographic studies were performed to evaluate the therapeutic response. Reoperation cases were defined as patients who required further surgery because the initial operation failed to cure the abscess. The duration of antibiotic treatment depended on the therapeutic response. Mortality was defined as death by any cause occurring during this hospitalization. The differences in the bacteriology, predisposing factors, surgical results, and mortality between multiloculated and uniloculated brain abscesses were noted.

RESULTS
The clinical characteristics of patients with uniloculated and multiloculated abscesses are summarized in Table 1. Comparisons of the bacteriology, predisposing factors, surgical result, and mortality between patients with multiloculated and uniloculated brain abscesses are listed in Table 2. Overall, 25 patients (20%) had multiloculated abscess, including 15 males and 10 females.

Of patients with uniloculated abscess, 78 were males and 21 were females. Of patients with multiloculated abscess, predisposing factors of infection were identified in 19. In the other 6 patients, predisposing factors of infection could not be identified. Hematogenous spread from a remote focus was the FIGURE 1. Magnetic resonance imagmost common source of in- ing scan showing multiloculated abscess, fection in multiloculated ab- defined as a lesion with multiple daughscess, occurring for 10 pa- ter loculations but without intervening tients. Among these 10 brain parenchyma between these patients, underlying sources loculations. included cyanotic congenital heart disease in 5 patients, metastatic spread from a distant infectious focus in 4 patients, and intravenous drug abuse in 1 patient. Five patients acquired the infection from neurosurgical events (craniotomy in 4 patients and penetrating brain injury in 1 patient). Four patients acquired the infection from a parameningeal focus. Similarly, hematogenous spread was the most common source of infection in patients with uniloculated abscess. The percentage of abscesses acquired from neurosurgical events was similar between two groups (Table 2). In patients with multiloculated abscess, headache was the most common presenting symptom. However, fever was the most common symptom in patients with uniloculated abscess. Hemiparesis was noted in 14 patients (56%) with multiloculated abscess and in 40 patients (40.4%) with uniloculated abscess. Eight patients were admitted with altered consciousness in multiloculated abscess. Overall, a higher percentage of patients with uniloculated abscess presented with fever, seizure, and neck stiffness. In this study, viridans streptococci were the most prevalent pathogens, accounting for 8 (32%) of the 25 cases of multiloculated abscess. Bacteroides fragilis was the most common anaerobic pathogen. Staphylococcus aureus was identified in two patients. Two patients had mixed infection. In five patients, no pathogens were isolated. Regarding bacteriology, we found the incidence (8%) of aerobic gramnegative bacilli infection in multiloculated abscess significantly lower than that (25%) in uniloculated abscess (Table 2). However, the incidence (28%) of viridans streptococci infection in multiloculated abscess was higher than that (15%) in uniloculated abscess. The frontal lobe was the most common location of multiloculated and uniloculated abscess. Three patients (12%) with multiloculated abscess had two abscesses. Twenty-one patients (21.2%) with uniloculated abscess had multiple abscesses. In all cases, the abscesses caused significant mass effect. In all patients, the neuroradiographic findings revealed by contrast-enhanced scans were a multiloculated, ringenhancing lesion surrounded by perifocal edema. Of the 25 patients with multiloculated abscess, 21 patients received both surgical and antimicrobial therapy, and 4 re-

1076 | VOLUME 52 | NUMBER 5 | MAY 2003

www.neurosurgery-online.com

MULTILOCULATED PYOGENIC BRAIN ABSCESS

TABLE 1. Characteristics of patients with multiloculated and uniloculated brain abscesses Multiloculated brain abscess (n 25) Sex Male Female Mean age (yr) Clinical manifestations Headache Hemiparesis Fever/chills Altered consciousness Nausea/vomiting Stiff neck Seizure Abscess location Single abscess Frontal lobe Temporal lobe Occipital lobe Parietal lobe Basal ganglion Cerebellum Other Multiple abscesses Uniloculated brain abscess (n 99)

15 (60%) 10 (40%) 37.96 20.42 20 14 8 8 6 3 2 (80%) (56%) (32%) (32%) (24%) (12%) (8%)

78 (78.8%) 21 (21.2%) 43.65 19.04 48 40 63 41 33 36 18 (48.5%) (40.4%) (63.6%) (41.4%) (33.3%) (36.3%) (18.2%)

tients accepted intravenously administered antibiotics for 4 to 6 weeks. In patients with multiloculated abscess, four (16%) died despite aggressive treatment. The overall mortality rate was 16%, which was similar to the observed rate (17%) in patients with uniloculated abscess. Of the patients who died, three had deteriorated consciousness at admission. Severe concomitant medical disease was present in two patients (one with septic shock and one with liver cirrhosis). One patient died because the abscess ruptured into a ventricle and caused ventriculitis. Mortality in patients with uniloculated abscess was 17.1%.

DISCUSSION
The reported incidence of multiloculated brain abscess varies, with most studies reporting an incidence of approximately 10% (6, 7, 12). One study reported the incidence as 43% (5). In our study, 20% of patients had multiloculated brain abscess. To our knowledge, only one published report has focused on multiloculated brain abscess (12). We were unable to identify another article directly comparing multiloculated and uniloculated abscesses. Therefore, our study is the first to report the differences in bacteriology, predisposing factors, surgical response, and outcome between multiloculated and uniloculated abscesses. The clinical features of our patients were similar to those of other patients with brain abscess, and there was nothing specific to suggest the multiloculated lesion. Male predominance was noted in our study, as in the study reported by Stephanov (12). Headache and hemiparesis were the most common symptoms in patients with multiloculated abscess. Although Stephanov (12) reported a high incidence (80%) of fever, only 32% of our patients presented with fever. However, fever was the most common presenting symptom in uniloculated abscess. Stephanov (12) reported that parameningeal infection was the most common origin of infection. However, hematogenous spread from a remote infectious focus was the most common predisposing factor in our study. Abscess encapsulation is influenced by a number of factors, including: 1) the offending organism; 2) the origin of infection (direct extension versus metastatic); 3) the immune status of the host; 4) corticosteroid administration; and 5) antibiotic therapy (2). Britt et al. (3) reported that experimental infection with B. fragilis resulted in multiple daughter abscesses indicative of poor containment of infection. It is known that B. fragilis is capable of producing a variety of enzymatic tissueactive toxins, including a collagenase that can degrade a forming capsule and a hyaluronidase that may further contribute to the development of abscess expansion and surrounding edema (1). In our study, B. fragilis was found in five cases of multiloculated abscess. B. fragilis also was present in three cases of uniloculated abscess, although the incidence was low. It has been demonstrated that abscesses formed by direct inoculation tend to be better encapsulated than metastatic abscesses secondary to septic embolization (13, 14). Metastatic abscesses frequently are associated with vegetative emboli

22 (88%) 12 3 1 1 2 0 3 3 (12%)

78 (78.8%) 31 17 5 4 6 4 11 21 (21.2%)

ceived only antimicrobial therapy. Among the 21 patients who underwent surgical treatment, 12 accepted aspiration as the first surgical method, and 6 (50%) of these patients required another operation to eradicate the recurrent abscess. Nine patients underwent excision as the first surgical method, and two (22%) of these patients required a second operation to eradicate the recurrent abscess. Overall, initial surgery failed in eight patients (38%), necessitating repeat surgery. In patients with uniloculated abscess, 84 patients accepted surgical intervention, and 15 patients were treated with antibiotics alone. Among these patients, 26 accepted aspiration as the first surgical method, and 4 (15.4%) of them required further surgery to eradicate the recurrent abscess. Fifty-eight patients accepted excision as the first surgical method, and seven of them (12%) required further surgery to eradicate the recurrent abscess. Overall, initial surgery failed in 11 patients (13.1%) with uniloculated abscess, necessitating further surgery. Third-generation cephalosporin (i.e., ceftriaxone 4 g/d or ceftazidime 8 g/d) combined with metronidazole 3 g/d was the main empiric antimicrobial treatment for bacterial brain abscess at our institution. The choice of antibiotics was guided by the final culture results. The duration of antibiotic treatment was based on the therapeutic response, and most pa-

NEUROSURGERY

VOLUME 52 | NUMBER 5 | MAY 2003 | 1077

SU

ET AL.

TABLE 2. Comparisons of multiloculated and uniloculated brain abscesses Multiloculated brain abscess (n 25) Predisposing factors Hematogenous spread Neurosurgical events Contiguous infection Unknown Isolated microorganisms Aerobic gram-negative bacilli Klebsiella pneumoniae Pseudomonas species Escherichia coli Other Streptococcus species Viridans streptococci non-A, non-B, and non-D streptococci Staphylococcus species Staphylococcus aureus Other staphylococci Anaerobes Bacteroides fragilis Fusobacterium Peptostreptococcus species Other Corynebacterium species Mixed infection Negative culture Treatment Aspiration Excision Conservative treatment Reoperation casesa Aspiration Excision Mortality
a

Uniloculated brain abscess (n 99)

10 (40%) 5 (20%) 4 (16%) 6 (24%) 2 (8%) 1 0 1 0 7 (28%) 7 0 2 (8%) 2 0 6 (24%) 4 1 1 0 1 (4%) 2 (8%) 5 (20%) 12 (48%) 9 (36%) 4 (16%) 8 (38%) 6 (50%) 2 (22.2%) 4 (16%)

35 (35.4%) 19 (19.2%) 23 (23.2%) 22 (22.2%) 25 (25.3%) 12 3 2 8 16 (16.1%) 15 1 6 (6%) 4 2 12 (12.1%) 3 2 0 7 3 (3%) 15 (15.1%) 22 (22.2%) 26 (26.2%) 58 (58.6%) 15 (15.2%) 11 (13.1%) 4 (15.4%) 7 (12%) 17 (17.1%)

Reoperation cases are recorded according to the initial surgical method. Percentages are as compared with the cases treated surgically according to the method.

that contribute to the formation of microinfarcts. This results in tissue hypoxia that impedes neovascularization and migration of fibroblasts, which form the reticulin precursor of the collagen capsule (8). These factors may result in poor containment of the infection and formation of multiple daughter lesions and may explain why hematogenous spread was the most common cause of infection in our patients with multiloculated abscess. However, hematogenous spread was the most common (35%) predisposing factor in uniloculated abscess, and a neurosurgical event (craniotomy, like direct inoculation) was not a rare (16%) predisposing factor in multiloculated abscess. These facts may suggest that the process of

abscess encapsulation is complex, and no single factor can completely account for the configuration of abscess. Stephanov (12) reported that with aspiration as the first operative method, repeat aspiration was needed to eradicate the abscess completely in four patients. However, among six patients in whom excision was the first surgical method, no further operation was needed. In our patients, there was a higher abscess recurrence rate when aspiration was the first surgical intervention. Thus, excision seems to be the more appropriate surgical choice for multiloculated abscess, as recommended by Loftus et al. (8). Nonetheless, the choice of surgical treatment depends on the correlation of clinical sta-

1078 | VOLUME 52 | NUMBER 5 | MAY 2003

www.neurosurgery-online.com

MULTILOCULATED PYOGENIC BRAIN ABSCESS

tus, neuroradiographic characteristics of the abscess (location, size, and number), and experience of the surgeons. A possible explanation for the high recurrence rate with aspiration in multiloculated abscess is that septations between the loculations may render complete aspiration of the contents impossible. Needle repositioning and repeated aspiration may resolve this problem. However, the risk of bleeding during needle repositioning cannot be overemphasized. Most of our patients accepted 4 to 6 weeks of antibiotic treatment, and this resulted in good therapeutic response. These results suggest that if these patients are treated promptly with a combination of appropriate surgery and antibiotic treatment, they may not require a longer duration of antibiotic treatment than those with uniloculated abscess. We also found a higher rate of abscess recurrence after surgery in multiloculated abscess (38%) as compared with uniloculated abscess (13.1%). This suggests that clinicians must closely monitor these patients both clinically and radiographically to detect abscess recurrence and intervene promptly. The observed mortality of 16% in this report is comparable to the generally reported rate of solitary or multiple brain abscesses (911, 15) irrespective of the abscess configuration. Ersahin et al. (4) noted no significantly higher mortality rate in cases of multiple or multiloculated abscesses. The observed mortality was similar in multiloculated and uniloculated abscesses in this study. This suggests that the prognosis for patients with multiloculated brain abscess can be as good as that for patients with uniloculated abscess if treatment is prompt. In our study, neurological status and concomitant medical disease at presentation were the most important factors influencing the therapeutic outcome. In four patients with multiloculated abscess who died, three were admitted with deteriorated consciousness, and two had severe concomitant medical disease.

2. Britt RH: Brain abscess, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1985, pp 19281956. 3. Britt RH, Enzmann DR, Placone RC Jr, Obana WG, Yeager AS: Experimental anaerobic brain abscess: Computerized tomographic and neuropathological correlations. J Neurosurg 60:11481159, 1984. 4. Ersahin Y, Mutluer S, Guzelbag E: Brain abscess in infants and children. Childs Nerv Syst 10:185189, 1994. 5. Jomma OV, Pennybacker JB, Tutton GK: Brain abscess: Aspiration, drainage or excision? J Neurol Neurosurg Psychiatry 14:308313, 1951. 6. Kiser JL, Kendig JH: Intracranial suppuration: A review of 139 consecutive cases with electron-microscopic observation on three. J Neurosurg 20:494 511, 1963. 7. Le Beau J, Creissard P, Harispe L, Redondo A: Surgical treatment of brain abscess and subdural empyema. J Neurosurg 38:198203, 1973. 8. Loftus CM, Osenbach RK, Biller J: Diagnosis and management of brain abscess, in Wilkins RH, Rengachary SS (eds): Neurosurgery. New York, McGraw-Hill, 1996, pp 32853298. 9. Mamelak AN, Mampalam TJ, Obana WG, Rosenblum ML: Improvement of multiple brain abscesses: A combined surgical and medical approach. Neurosurgery 36:7686, 1995. 10. Mampalam TJ, Rosenblum ML: Trends in the management of bacterial brain abscesses: A review of 102 cases over 17 years. Neurosurgery 23:451458, 1988. 11. Morgan H, Wood MW, Murphey F: Experience with 88 consecutive cases of brain abscess. J Neurosurg 38:698704, 1973. 12. Stephanov S: Experience with multiloculated brain abscesses. J Neurosurg 49:199203, 1978. 13. Wood JH, Doppman JL, Lightfoote WE II, Girton M, Ommaya AK: Role of vascular proliferation on angiographic appearance and encapsulation of experimental traumatic and metastatic brain abscess. J Neurosurg 48:264 273, 1978. 14. Wood JH, Lightfoote WE II, Ommaya AK: Cerebral abscess produced by bacterial implantation and septic embolization in primates. J Neurol Neurosurg Psychiatry 42:6369, 1979. 15. Yang SY: Brain abscess: A review of 400 cases. J Neurosurg 55:794799, 1981.

COMMENTS
u et al. report their 16-year experience in treating patients with multiloculated brain abscesses after hematogenous spread. The outcome and mortality in their patients did not depend on whether the abscess was single or multiloculated. As in most series, poor outcome was related to comorbidities and a poor Glasgow Coma Scale score on arrival at the hospital. Overall, the authors tended to excise abscesses rather than aspirate them. In their series, aspiration was associated with a higher incidence of recurrence in patients with multiloculated abscesses, as compared to those with uniloculated abscesses. In our experience, an alternate form of operative management would be to break down the loculations and place draining catheters under vision but without removal of the capsule. The authors discuss bacterial differences that exist between abscesses arising from otorhinological causes, peripheral sepsis, implantation, and hematogenous spread. They once again confirm that the mortality from brain abscesses is not inconsiderable and that a high level of awareness and early and prompt treatment to irradiate the abscess offers the patient the best outcome. James Van Dellen London, England

CONCLUSION
Multiloculated brain abscess accounted for 20% of our patients with brain abscess. The 8% incidence of aerobic gramnegative bacilli infection in patients with multiloculated abscess was significantly lower than the 25% rate observed in those with uniloculated abscess. Excision seems to be the most appropriate surgical choice in multiloculated abscesses because it is difficult to aspirate these lesions completely. Our results demonstrate that the prognosis for patients with multiloculated brain abscess can be as good as that for patients with uniloculated abscess if treatment is prompt. Nonetheless, clinicians must monitor these patients intensively, because the possibility of recurrent abscess formation after surgery is significantly higher than that in patients with uniloculated abscess.

REFERENCES
1. Alderson D, Strong AJ, Ingham HR, Selkon JB: Fifteen-year review of the mortality of brain abscess. Neurosurgery 8:15, 1981.

NEUROSURGERY

VOLUME 52 | NUMBER 5 | MAY 2003 | 1079

SU

ET AL.

his review of a series of patients with brain abscesses highlights the similarities of uniloculated and multiloculated abscesses. The authors found that there was little difference in etiology or patient characteristics, a difference in bacteriology of uncertain significance, no difference in mortality rate, but an important difference in the chance of recurrence requiring reoperation. The high rate of failure of aspiration as first treatment (6 of 12 patients) led to the recommendation that excision be the first surgical treatment of choice. Although the recommendation cannot be considered definitive, it is sensible on the basis of the limited information available, and it is consistent with the recommendations of others. It is interesting to note that the mortality rate of cerebral abscess, whether uniloculated or multiloculated, remains essentially unchanged since the era of drainage with chicken bones, as reported by William Macewen in the 19th century. Stephen J. Haines Charleston, South Carolina he authors observed a differential bacterial etiology for the 25 cases of multiloculated brain abscess and the 99 cases of uniloculated abscess in their series. Aerobic gram-negative bacilli were present in 8% of multiloculated and in 25% of uniloculated abscesses, anaerobes in 24% of multiloculated

and in 12% of uniloculated abscesses, and streptococci in 28% of multiloculated and in 16% of uniloculated abscesses. It would be interesting to see whether these trends are confirmed in even larger series drawn from different geographic areas. Robert G. Grossman Houston, Texas he authors present an excellent review of multiloculated pyogenic brain abscess treated at their institution during a 16year period. During that time, multiloculated abscesses, as distinct from multiple abscesses, accounted for 20% of patients with pyogenic brain abscess; the prognosis for these patients was as good as for patients with uniloculated abscesses. As expected, this study demonstrated a higher rate of abscess recurrence with aspiration as a first surgical intervention, almost certainly because of the septations between the loculations, which make complete aspiration of contents difficult. Consequently, I would agree with the authors that excision is usually a more appropriate surgical choice in the multiloculated abscess, as a result of the difficulty in complete aspiration of these lesions. Andrew H. Kaye Melbourne, Australia

AESCULAP Prize for Neurosurgical Research of the European Association of Neurosurgical Societies
This prize of US$5000 is offered by the Aesculap Company and awarded annually by the European Association of Neurosurgical Societies (EANS). Those eligible for the prize should be neurosurgeons under the age of 40 at the time of submission, who are either fully trained or still in the course of their training. Applicants should be either a member of one of the national societies of the EANS or working in a department in one of the EANS countries. The basis of the manuscripts submitted should be a research work, either clinical or experimental or both, of relevance in the field of neuroscience. The author should also make a declaration that he has performed most (more than 50%) of the work. Manuscripts that have been published during that calendar year as well as unpublished studies are accepted for evaluation. The format or type of manuscript has to be comparable to that presented for Acta Neurochirurgica. Ten copies of the submitted manuscript together with a brief curriculum vitae should be sent to the Chairman of the EANS Research Committee before December 31st, 2003. The prize will be presented normally during the EANS training course of 2004, and the winner will be invited to attend that meeting and to present his work. The Chairman of the EANS Research Committee is: Professor Dr. Yu cel Kanpolat Inkilap Sokak No: 24/2 Kzlay-06640 Ankara, Turkey Tel: 90/312/417-4078 or 417-4079 Fax: 90/312/419-3684 Email: kanpolat@ada.net.tr

1080 | VOLUME 52 | NUMBER 5 | MAY 2003

www.neurosurgery-online.com