Encephaloceles have generally been classified according

Transcription

1 J J Neurosurg Pediatrics 7: , 7:88 93, 2011 Encephalocele in Uganda: ethnic distinctions in lesion location, endoscopic management of hydrocephalus, and survival in 110 consecutive children Clinical article Benjamin C. Warf, M.D., 1 Vita Stagno, M.D., 2 and John Mugamba, M.D. 3 1 Department of Neurosurgery, Children s Hospital Boston, and Harvard Medical School, Boston, Massachusetts; 2 Department of Neurological Sciences, Universita degli Studi di Napoli Federico II, Naples, Italy; and 3 CURE Children s Hospital of Uganda, Mbale, Uganda Object. This study characterizes the first clinical series of encephalocele (EC) from East or Central Africa, and is the largest reported from the African continent. The authors explored survival, the efficacy of primary endoscopic management of associated hydrocephalus, and ethnic differences in EC location. Methods. One hundred ten consecutive children presented to CURE Children s Hospital of Uganda for treatment of EC over a 9-year period. Clinical data, including patient demographic information, birth date, lesion type (sincipital, parietal, or occipital), operative data, and subsequent course had been entered prospectively into a clinical database. Home visits to update the status of those lost to follow-up were done when possible. With appropriate institutional approvals, the database was reviewed for this retrospective study. Two-tailed probability values calculated using the Fisher exact test were used to assess the significance of differences among groups, with p < 0.05 being considered significant. The Kaplan-Meier method was used for analysis of survival and treatment success probabilities. Results. There were 53 (48%) occipital, 33 (30%) sincipital, and 24 (22%) parietal lesions. Occipital lesions were significantly more common among children of Bantu origin (p = 0.02). Nilotes demonstrated a roughly equal distribution among sincipital, parietal, and occipital locations. The female/male ratio was 1.2, with no difference between EC types (range , p = ). Of 110 patients, 108 (98%) underwent surgical repair at a median age of 1 month (mean 15.7 months), whereas 2 had treatment for hydrocephalus only. Wound revision was required in 13% of cases. Surgery-related mortality was 3%. One-year and 5-year survival rates were 87% (95% CI ) and 61% (95% CI ), respectively. Hydrocephalus required treatment in 32%, and was equally common among the 3 EC types. Thirteen patients were treated with combined endoscopic third ventriculostomy/choroid plexus cauterization (ETV/CPC) and 2 with ETV alone, whereas 18 patients received primary shunt placement. Predicted treatment success at 1 year was 79% for ETV or ETV/CPC (95% CI ) and 47% for shunt placement (95% CI ). Conclusions. Analysis of this first EC series from this region suggests that sincipital lesions are 3 times more common in East than in West Africa. Occipital lesions predominate in patients of Bantu origin, but not among those of Nilotic descent. Hydrocephalus incidence was equally common among different EC types, and endoscopic treatment was more successful (79%) than shunting (47%) at 1 year. The 5-year mortality rate was similar to that for infants with myelomeningocele in Uganda, and more than twice that for their unaffected peers. (DOI: / PEDS10326) Key Words encephalocele outcome hydrocephalus Uganda Africa Bantu Nilote ethnicity endoscopic third ventriculostomy choroid plexus cauterization Abbreviations used in this paper: CCHU = CURE Children s Hospital of Uganda; EC = encephalocele; ETV/CPC = endoscopic third ventriculostomy/choroid plexus cauterization; VP = ventriculoperitoneal. Encephaloceles have generally been classified according to location as occipital (including suboccipital), parietal, sincipital (including frontal and nasofrontal), and basal. 1 There are well-known geographic variations in the predominant EC type, with occipital lesions predominating in North America, Europe, and the Middle East, 5,7,11,19,23,25,27 and sincipital lesions predominating in Southeast Asia. 3,16,22,23,27,28 Series of ECs have been described in the West African countries of Nigeria This article contains some figures that are displayed in color on line but in black and white in the print edition. 88 J Neurosurg: Pediatrics / Volume 7 / January 2011

2 Encephalocele in Uganda and Senegal where, as in North America, occipital lesions predominate and sincipital lesions are rare. 1,2,6,14 To date, there have been no clinical series of EC reported from an East or Central African country. We present the largest surgical series of ECs to be reported from the African continent, and the first from East Africa. This study provides the first report of a higher incidence of sincipital EC in East Africa, and suggests a difference in this regard between peoples of Bantu and Nilotic descent. We also provide the first report of combined ETV/CPC as the primary treatment for hydrocephalus associated with EC. Methods Clinical Management Infants presenting to CCHU with EC underwent surgical repair, the timing of which was based on the condition of the child and whether CSF leakage was present or imminent. In the latter case, closure was performed urgently (Fig. 1). Preoperative ultrasonography or CT scanning was performed to image the sac contents, delineate the anatomy, and assess for ventriculomegaly. The surgical goal for occipital and parietal lesions was excision of the sac, watertight reconstruction of the dura mater, and skin closure. In the case of sincipital lesions, a bicoronal scalp flap for a limited glabellar craniotomy and intradural repair was typically used, as has been described by others. 18,29 In this approach, the sac contents can be emptied and the dural margins developed for primary closure. In very young children, a modest-sized sac was given time to regress spontaneously, to avoid an incision on the face (Fig. 2). For very large sacs that required excision, particularly among those in whom the sac obstructed vision or in whom there was a compromised skin covering or frank CSF leak, an initial effort was made to repair the EC extracranially from the anterior approach prior to proceeding with craniotomy (Fig. 3). If dural closure was not satisfactory from this approach, we then proceeded with the intradural repair described above. Postoperatively, the patients were followed for the development of progressive hydrocephalus by head circumference measurement, clinical symptoms and signs, and surveillance imaging with cranial ultrasonography Fig. 2. Left: Preoperative photograph of 2.5-month-old infant with sincipital EC. Right: Postoperative result at approximately 18 months after the intracranial repair. or CT scanning. The treatment of infant hydrocephalus in our institution evolved over time. Placement of a VP shunt was the primary treatment from January to June of Endoscopic third ventriculostomy was attempted as the primary treatment in all cases after June 2001, and bilateral CPC was subsequently combined with ETV and was demonstrated to increase ETV success significantly in infants younger than 1 year of age. 30 In the event an ETV could not be performed for technical reasons, a VP shunt was typically placed during the same procedure. Clinical Database Clinical data, including patient demographic information, birth date, lesion type, operative data, and subsequent course, were entered prospectively into our clinical database. When patients were lost to follow-up, attempts were made to update their status by home visits. With approval from the Internal Review Board of CCHU and the Children s Hospital Boston Committee on Clinical Investigations (protocol number M ), the database was reviewed for this retrospective study. Statistical Methods Chi-square analysis performed using the Fisher 2- Fig. 1. Photograph of an infant with occipital EC and compromised integrity of the skin on the EC sac. J Neurosurg: Pediatrics / Volume 7 / January 2011 Fig. 3. Left: Photograph of a newborn baby with sincipital EC and exposed extracranial cerebral tissue. Right: The same infant immediately following extracranial repair. 89

3 B. C. Warf, V. Stagno, and J. Mugamba tailed test was used to test the significance of differences among groups, with p < 0.05 being considered significant. The Kaplan-Meier method was used for analysis of survival and treatment success probabilities. Results Surgical Results Between January 2001 and March 2010, 110 patients presented to CCHU for treatment of EC. Of these, 61 patients were female and 49 were male, a female/male ratio of 1.2. Of these, 108 (98%) underwent surgical repair at a median age of 1 month (mean 15.7 months). Two patients, both with sincipital EC, underwent treatment of hydrocephalus only. One of these patients underwent palliative shunt placement for severe hydrocephalus when virtually no cerebral tissue was noted on CT scans (Fig. 4). The other patient underwent ETV/CPC as initial treatment for hydrocephalus, with the intention of subsequent surgical repair of the EC, but this child died of malaria 3 months later. Wound revision was required in 14 lesions (13%), 8 of which were sincipital, 4 occipital, and 2 parietal in location. There were 3 deaths within 30 days of the operation (surgery-related mortality rate of 3%). Encephalocele Location The EC location was classified as sincipital, parietal, or occipital, with no basal lesions being identified. There were 53 (48%) occipital, 33 (30%) sincipital, and 24 (22%) parietal lesions. The female/male ratio among the different presentations was similar (occipital 1.4, sincipital 1.2, and parietal 1.0), with no significant differences among groups (p = ). Ethnic Origin and Seasonal Pattern Because CCHU serves people of both Nilotic and Bantu origins (see Discussion), a relationship between ethnic heritage and EC location was investigated. Of 110 patients, 62 were identified by district and family name as being of Bantu origin, and 48 were of Nilotic origin. The comparative distribution of lesion locations is presented in Table 1. Occipital lesions compared with those in other TABLE 1: Location of EC and patient ethnicity in 110 cases in Uganda Ethnic Group EC Location (%) Sincipital Occipital Parietal Bantu 16 (26) 36 (58) 10 (16) Nilote 17 (35) 17 (35) 14 (29) total 33 (30) 53 (48) 24 (22) locations were significantly more common among children of Bantu origin (p = 0.02). Nilotes demonstrated a roughly equal distribution among sincipital, parietal, and occipital locations, whereas occipital lesions predominated among Bantu. Uganda is a tropical country with distinct rainy and dry seasons. Typically, the rainfall peaks in April and October, with minimums in January and July. Because increased incidence of sincipital EC had been reported during the dry season in the tropical country of Cambodia, 22 we reviewed the date of birth for all patients and selected those born in either the 2 peak rainy season months or the 2 peak dry season months for analysis (Table 2). There was no apparent relationship between season of birth and number of sincipital versus other lesions (p = 0.48) or occipital versus other lesions (p = 0.33). Treatment for Hydrocephalus Thirty-five patients (32%) required treatment for hydrocephalus, representing 32% of occipital, 36% of sincipital, and 25% of parietal lesions. There was no significant difference in the incidence of hydrocephalus among groups (p = ). In 31 patients, hydrocephalus was treated at a mean of 87.9 days (median 32 days) following EC repair, including 4 patients who were treated at the time of repair, and 2 who required treatment more than 1 year after EC repair. In 2 patients hydrocephalus was treated 7 and 14 days prior to EC repair, and in 2 other patients, treatment for hydrocephalus was the only operation, as noted above. Ventriculoperitoneal shunts were placed in 16 patients, and subdural-peritoneal shunts in 2. The combined Fig. 4. Head CT scans of an infant with sincipital EC, severe hydrocephalus, and a paucity of cerebral tissue. 90 J Neurosurg: Pediatrics / Volume 7 / January 2011

4 Encephalocele in Uganda TABLE 2: Location of EC and season of birth in 37 patients born during the peak seasons in Uganda Season EC Location Sincipital Parietal Occipital ETV/CPC procedure was performed in 13 patients, and ETV alone in 2. One patient each had CPC alone and CPC with aqueductoplasty because of technical failure to establish a third ventriculostomy. Both of these children ultimately required shunt placement. Of the 15 whose hydrocephalus was treated by ETV or ETV/CPC, 2 had less than 1 month of follow-up. Of the 13 with sufficient follow-up, treatment failed in 2, requiring an additional operation (1 shunt placement and 1 successful repeat ETV), and 11 (85%) were successfully treated by ETV or ETV/CPC, with mean and median follow-up durations of and 437 days, respectively. Kaplan-Meier survival analysis indicates a 1-year success probability of 79% (95% CI ). Of the 18 patients who initially underwent shunt placement (16 VP shunts and 2 subdural-peritoneal shunts), 3 were lost to followup at less than 1 month, and 9 (60%) of the 15 required another shunt operation at a mean of days (median 42 days), with all but one failure occurring within the 1st year of placement. One patient underwent ETV after a second shunt failure, with a successful follow-up of more than 5 years. The probability of success at 1 year for primary shunt placement was 47% (95% CI ). Patient Survival Of 110 patients, 3 died (surgery-related mortality rate of 3%) and 20 were lost to follow-up at less than 30 days. An additional 16 patients are known to have died later, after a mean clinical follow-up of 666 days following surgery (median 475 days). The 71 surviving patients had a mean follow-up of 614 days (median days). Survival analysis for the entire group of 110 patients gives survival probabilities of 87% at 1 year (95% CI ), 73% at 2 years (95% CI ), and 61% at 5 years (95% CI ). Discussion Incidence of EC in Uganda This is the largest African EC series, and the first from an East or Central African country of which we are aware. The incidence of EC at birth in Uganda is unknown. Using known population and birth rate data for Mbale District (where CCHU is located) and the number of infants born in Mbale District with EC presenting for treatment, one can derive an absolute minimum birth incidence. During the 9 years from 2001 to 2009, there were 310,284 anticipated births in Mbale District. During that period, 10 infants were born with EC in that district and presented to CCHU for treatment. This would then give a minimum birth incidence of 0.3/10,000 live births. J Neurosurg: Pediatrics / Volume 7 / January 2011 Total dry rainy In the same way, we derived a very similar result for the neighboring district of Kumi (6 patients for an anticipated 168,336 births, or 0.35/10,000). Such children, however, often may not be taken for treatment. We previously estimated that only 25% of infants with myelomeningocele born in the immediate region of our hospital may come to CCHU for treatment. 34 Based on a similar assumption, an approximation of EC birth incidence, then, could be as high as 1.2 per 10,000 live births. This would be within the range of 0.8 to 4 per 10,000 live births reported from other regions of the world, 23 although not as high as the 5/10,000 births reported from Nigeria by Adetiloye and coauthors 2 in their hospital-based study. Encephalocele Location and Ethnic Predisposition Encephalocele location varies by geography, with occipital lesions being much more common in the Western Hemisphere and sincipital lesions generally being more common in the East. 23,27 Occipital lesions account for 66% 95% of ECs in North America, Western Europe, and the Middle East. 5,7,11,19,23,25,27 Sincipital lesions predominate in Thailand, Malaysia, Burma, Cambodia, the Philippines, and the Pacific Islands, 3,4,8,16,20,22, 27,28 comprising 20 of 21 EC cases in one report. 3 Monteith and coauthors 16 reported ethnic distinctions in New Zealand, with patients of Pacific Island descent having a much higher incidence of sincipital lesions (44%) than those of European origin. Suwanwela 27 reviewed the world s literature up to 1970, and also found a very high incidence of sincipital lesions reported from Russia (86%). In the same review, he reported relatively high incidences from South Africa (47%), and Morocco in northern Africa (39%). There have been few studies of EC originating in sub-saharan Africa. Suwanwela 27 uncovered a single small Nigerian series of EC from 1967 in which 19% of lesions were sincipital. Several more recent studies from the western African countries of Nigeria and Senegal have reported only 8% 12% of lesions as being sincipital, with the great majority being occipital. 1,2,6,14 The 30% sincipital location in our East African population of Uganda is 3 times greater. Present-day Uganda was positioned at the boundary of two major population group expansions. The progenitors of the Bantu people originated from Nigeria and Cameroon in West Africa, expanding east and south as agriculturalists and displacing or incorporating huntergatherer peoples in the process. 9,12,24 They settled in western and southern Uganda by the 1st century of the Common Era. Nilotes are thought to have more diverse origins from the Nile River valley, the Sudan, and the Horn of Africa. They had settled into the northern part of present-day Uganda after the 1st millennium. Recent studies indicate a low genetic admixture between the two groups. 10 Given the contrast in the prevalence of sincipital EC between our population and that reported from several centers in Nigeria, we hypothesized a possible ethnic difference in this regard between children of Bantu and non-bantu origin. In a similar way, it is recognized that Nilotes have a higher incidence of oral-facial clefts, 26 an anomaly known to be associated with sincipital EC. 4,21 We found that infants of Bantu origin had a signifi- 91

5 B. C. Warf, V. Stagno, and J. Mugamba cantly higher proportion of occipital lesions than did those of Nilotic origin. For Bantu, occipital lesions comprised the majority, whereas there was an equal distribution between occipital and sincipital locations among Nilotes. Unlike the experience reported from Cambodia, 22 we did not see a seasonal pattern for sincipital EC. Incidence of Associated Hydrocephalus Our overall 32% incidence of hydrocephalus requiring treatment is very similar to that reported by others, 7,13 as is our 32% incidence among patients with occipital EC. However, whereas others have reported hydrocephalus to be much less common among children with frontal lesions, we found it to be at least as common in this group. Among sincipital lesions in our series, 36% required treatment for hydrocephalus, in contrast to the approximately 9% 17% reported by others. 15,23 Endoscopic Treatment of Hydrocephalus We have previously reported the unique difficulties of shunt dependence in the context of rural Africa Because of this, it was our philosophy to treat hydrocephalus endoscopically whenever possible, regardless of the patient s age or the cause of hydrocephalus. 31 Because ETV alone was least successful in infants younger than 1 year of age, we developed the technique of combined ETV/CPC, which proved to be more successful than ETV alone in this age group. 30,32 We subsequently demonstrated that early neurocognitive development among infants with myelomeningocele treated by ETV/CPC was as good as those whose hydrocephalus had been shunted. 33 This is the first report of results for ETV/CPC as treatment for hydrocephalus associated with EC. Treatment by ETV alone was previously reported in 2 patients with occipital EC, of whom 1 was treated successfully. 17 The success rate of 79% that we report here is similar to the 76% success we previously reported for this procedure among infants with myelomeningocele. 30,32 It is noteworthy that among those requiring treatment for hydrocephalus, 2 (13%) of 15 primarily treated by ETV or ETV/CPC underwent a second operation related to hydrocephalus, compared with 9 (50%) of 18 in whom shunting was the primary treatment, and that the 1-year probabilities of treatment success were 79% and 47% for ETV/CPC and VP shunt, respectively. Patient Survival The 1- and 5-year survival rate in these patients was 87% (95% CI ) and 61% (95% CI ), respectively. We recently reported an 80% 1-year and 63% 5-year survival rate among infants with myelomeningocele treated at CCHU, compared with an expected 84% childhood (5-year) survival for the general population (unpublished data). This suggests that the survival in these two populations is very similar in Uganda. Significantly, we found that deaths in the myelomeningocele population were most commonly secondary to treatable infectious diseases, and that survival was significantly higher (close to that of their unaffected peers) for children with access to a community-based rehabilitation program. This suggested that disabled children in East Africa may have more restricted access to general health care than other children in the absence of a proactive community-based support program. Of the 16 patients with EC known to have died after the initial postoperative period, the cause of death was identified in 7 (malaria in 3, sepsis in 2, breathing failure in 1, and heart failure in 1). Unless EC is associated with other life-threatening problems (such as cardiac anomalies), an uncomplicated closure and successful management of hydrocephalus should lead to childhood survival rates similar to those for the general population. The mortality rate in a recent Canadian series of 85 children treated over a 12-year period was 5%. 13 However, the reality for these children in rural East Africa is a 5-year mortality rate of approximately 40%; more than twice that anticipated for their unaffected Ugandan peers. Conclusions This is the first East African series of EC to be reported, and the largest series from the continent. Sincipital lesions comprised one-third, and were 3 times more common than those reported from several series in West Africa. Occipital lesions predominated among children of Bantu origin, whereas sincipital lesions were as common as occipital in the Nilotic population. Hydrocephalus was equally common among infants with anterior and posterior lesions, requiring treatment in one-third. Endoscopic treatment of hydrocephalus was successful in 79%; a result very similar to that for the myelomeningocele population. Primary CSF shunting had more than twice the failure rate at 1 year. The 5-year survival rate of 61% mirrored what we have previously reported for Ugandan infants with myelomeningocele. As we have shown in the spina bifida population, community-based advocacy for these children may be crucial to improving their survival. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Warf. Acquisition of data: all authors. Analysis and interpretation of data: all authors. Drafting the article: Warf. Critically revising the article: Warf. Reviewed final version of the manuscript and approved it for submission: Warf. Statistical analysis: Warf. Administrative/technical/ material support: Warf. Study supervision: Warf. References 1. Adeleye AO, Olowookere KG: Central nervous system congenital anomalies: a prospective neurosurgical observational study from Nigeria. Congenit Anom (Kyoto) 49: , Adetiloye VA, Dare FO, Oyelami OA: A ten-year review of encephalocele in a teaching hospital. Int J Gynaecol Obstet 41: , Agthong S, Wiwanitkit V: Encephalomeningocele cases over 10 years in Thailand: a case series. BMC Neurol 2:3, Arshad AR, Selvapragasam T: Frontoethmoidal encephalocele: treatment and outcome. J Craniofac Surg 19: , Asindi A, Al-Shehri A: Neural tube defects in the Asir Region of Saudi Arabia. Ann Saudi Med 21:26 29, J Neurosurg: Pediatrics / Volume 7 / January 2011

6 Encephalocele in Uganda 6. Ba MC, Kabre A, Badiane SB, Ndoye N, Ly Ba A, Gueye EM, et al: [Fronto-ethmoïdal encephaloceles in Dakar. Report of 9 cases.] Dakar Med 48: , 2003 (Fr) 7. Baradaran N, Nejat F, Baradaran N, El Khashab M: Cephalocele: report of 55 cases over 8 years. Pediatr Neurosurg 45: , Barone CM, Jimenez DF, Helling ER, Laskey AL: What is the incidence of hypertelorism in Filipino frontonasal encephalocele patients? J Craniofac Surg 18: , Diamond J, Bellwood P: Farmers and their languages: the first expansions. Science 300: , Gomes V, Sánchez-Diz P, Amorim A, Carracedo A, Gusmão L: Digging deeper into East African human Y chromosome lineages. Hum Genet 127: , Ingraham FD, Swan H: Spina bifida and cranium bifidum. I. A survey of five hundred and forty-six cases. N Engl J Med 228: , Kajumbula H, Wallace RG, Zong JC, Hokello J, Sussman N, Simms S, et al: Ugandan Kaposi s sarcoma-associated herpesvirus phylogeny: evidence for cross-ethnic transmission of viral subtypes. Intervirology 49: , Lo BWY, Kulkarni AV, Rutka JT, Jea A, Drake JM, Lamberti- Pasculli M, et al: Clinical predictors of developmental outcome in patients with cephaloceles. Clinical article. J Neurosurg Pediatr 2: , Mabogunje OA: Cranium bifidum in northern Nigeria. Childs Nerv Syst 6:95 98, Macfarlane R, Rutka JT, Armstrong D, Phillips J, Posnick J, Forte V, et al: Encephaloceles of the anterior cranial fossa. Pediatr Neurosurg 23: , Monteith SJ, Heppner PA, Law AJ: Encephalocoele epidemiological variance in New Zealand. J Clin Neurosci 12: , Moorthy RK, Rajshekhar V: Management of hydrocephalus associated with occipital encephalocoele using endoscopic third ventriculostomy: report of two cases. Surg Neurol 57: , Pinzer T, Lauer G, Gollogly J, Schackert G: A complex therapy for treatment of frontoethmoidal meningoencephalocele in a developing third world country: neurosurgical aspects. J Neurosurg 104 (5 Suppl): , Raja RA, Qureshi AA, Memon AR, Ali H, Dev V: Pattern of encephaloceles: a case series. J Ayub Med Coll Abbottabad 20: , Richards CG: Frontoethmoidal meningoencephalocele: a common and severe congenital abnormality in South East Asia. Arch Dis Child 67: , Rittler M, López-Camelo JS, Castilla EE, Bermejo E, Cocchi G, Correa A, et al: Preferential associations between oral clefts and other major congenital anomalies. Cleft Palate Craniofac J 45: , Roux FE, Oucheng N, Lauwers-Cances V, Draper L, Cristini C, Collogly J, et al: Seasonal variations in frontoethmoidal meningoencephalocele births in Cambodia. Clinical article. J Neurosurg Pediatr 4: , Rutka J, Carlotti C, Iantosca M: Encephaloceles, in Winn HR (ed): Youmans Neurological Surgery, ed 5. Philadelphia: WB Saunders, 2004, pp Salas A, Richards M, De la Fe T, Lareu MV, Sobrino B, Sánchez-Diz P, et al: The making of the African mtdna landscape. Am J Hum Genet 71: , Simpson DA, David DJ, White J: Cephaloceles: treatment, outcome, and antenatal diagnosis. Neurosurgery 15:14 21, Spritz RA, Arnold TD, Buonocore S, Carter D, Fingerlin T, Odero WW, et al: Distribution of orofacial clefts and frequent occurrence of an unusual cleft variant in the Rift Valley of Kenya. Cleft Palate Craniofac J 44: , Suwanwela C: Geographical distribution of fronto-ethmoidal encephalomeningocele. Br J Prev Soc Med 26: , Suwanwela C, Hongsaprabhas C: Fronto-ethmoidal encephalomeningocele. J Neurosurg 25: , Tobias JD, Mateo C, Ferrer MJ, Jimenez DF, Barone CM, Reyes de Castro L: Intrathecal morphine for postoperative analgesia following repair of frontal encephaloceles in children: comparison with intermittent, on-demand dosing of nalbuphine. J Clin Anesth 9: , Warf BC: Comparison of endoscopic third ventriculostomy alone and combined with choroid plexus cauterization in infants younger than 1 year of age: a prospective study in 550 African children. J Neurosurg 103 (6 Suppl): , Warf BC: Hydrocephalus in Uganda: the predominance of infectious origin and primary management with endoscopic third ventriculostomy. J Neurosurg 102 (1 Suppl):1 15, Warf BC, Campbell JW: Combined endoscopic third ventriculostomy and choroid plexus cauterization as primary treatment of hydrocephalus for infants with myelomeningocele: long-term results of a prospective intent-to-treat study in 115 East African infants. Clinical article. J Neurosurg Pediatr 2: , Warf BC, Ondoma S, Kulkarni A, Donnelly R, Ampeire M, Akona J, et al: Neurocognitive outcome and ventricular volume in children with myelomeningocele treated for hydrocephalus in Uganda. Clinical article. J Neurosurg Pediatr 4: , Warf BC, Wright EJ III, Kulkarni AV: Factors affecting survival of infants with myelomeningocele in southeastern Uganda. Clinical article. J Neurosurg Pediatr [in press], 2010 Manuscript submitted July 26, Accepted September 29, Address correspondence to: Benjamin C. Warf, M.D., Department of Neurosurgery, Children s Hospital Boston, 300 Longwood Avenue, Boston, Massachusetts benjamin.warf@ childrens.harvard.edu. J Neurosurg: Pediatrics / Volume 7 / January