Wai-Keong Wong

Singapore General Hospital, Tumasik, Singapore

Are you Wai-Keong Wong?

Claim your profile

Publications (48)141.97 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: Background: Recent studies have demonstrated that the neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) are prognostic for various malignancies; however, there are limited data to date demonstrating their usefulness for gastrointestinal stromal tumors (GISTs). The aim of this study was to determine whether NLR and PLR are prognostic for GIST. Methods: Three hundred patients who underwent operative resection for primary localized GIST with preoperative results for neutrophil, platelet, and lymphocyte counts available were retrospectively reviewed. Optimal cutoff values for high NLR (≥3.0) and PLR (≥275) in predicting recurrence-free survival (RFS) were determined. Prognostic factors of RFS were determined using univariate and multivariate Cox regression analyses for the 266 patients who did not receive adjuvant imatinib. Results: On univariate analyses, tumor size, mitotic count, location, and both a high NLR and PLR were significant prognostic indicators of decreased RFS (hazard ratio [HR], 2.89 [95% CI; 1.609-5.179 P < .001] and HR, 3.572 [95% CI, 2.094-6.096; P < .001], respectively). On multivariate analyses, tumor size, mitotic count, tumor location, and both high NLR and PLR were independent prognostic factors of RFS in GIST. Both a high NLR and PLR were significant prognostic factors for GISTs within the National Institutes of Health (NIH) and Armed Forces Institute of Pathology (AFIP) high-risk categories. Addition of NLR or PLR to the NIH or AFIP improved the accuracy of these systems. Conclusion: High NLR and PLR were independent prognostic factors of RFS in GIST. NLR and PLR would be useful as a preoperative prognostic tool and its incorporation into current prognostication systems improved their accuracy.
    No preview · Article · Dec 2015 · Surgery
  • [Show abstract] [Hide abstract]
    ABSTRACT: To validate the Memorial Sloan Kettering Cancer Center (MSKCC) prognostic nomogram in a single-institution cohort of patients with gastrointestinal stromal tumors (GISTs), and to compare its predictive accuracy against other established risk classification systems, including the National Institutes of Health (NIH), Armed Forces Institute of Pathology (AFIP), and Joensuu criteria. We retrospectively reviewed 289 patients who underwent surgical resection for primary localized GISTs without adjuvant imatinib therapy and compared the actuarial recurrence-free survival (RFS) with the predicted RFS. Tumors >5 cm in size, with high mitotic index, and which had ruptured were significantly associated with recurrent disease. The 2-year RFS was 77.2 % [95 % confidence interval (CI) 71.6-81.8], and the 5-year RFS was 67.9 % (95 % CI 61.7-73.4). The concordance probability of the nomogram of 2-year RFS was 0.71 (SE 0.02), and 5-year RFS was 0.71 (SE 0.19). The 2-year and 5-year MSKCC nomogram probability calculations and the AFIP criteria gave a better estimation of RFS compared to the NIH (p < 0.001) and Joensuu (p < 0.001) criteria. There was no significant difference between the predictive accuracy of the nomogram compared to the AFIP criteria. The MSKCC nomogram slightly underestimated the probability of RFS after surgical resection of GISTs. It was associated with a significantly better predictive accuracy compared to the NIH and Joensuu. This study suggests that there is a wider than expected prognostic divergence between gastric GISTs versus GISTs arising from the small intestine.
    No preview · Article · Feb 2015 · Annals of Surgical Oncology
  • [Show abstract] [Hide abstract]
    ABSTRACT: Background The Sendai Consensus Guidelines(SCG) was formulated in 2006 to guide management of mucinous cystic lesions of the pancreas (CLP)and was updated in 2012 (ICG 2012). This study aims to evaluate the clinical utility of the ICG 2012 with the SCG based on initial cross-sectional imaging findings Methods One-hundred and fourteen patients with mucinous CLP were reviewed and classified according to the ICG 2012 as high risk(HRICG2012), worrisome(WICG2012),low risk(LRICG2012) and according to the SCG as high risk(HRSCG) and low risk(LRSCG) Results On univariate analysis; symptoms, obstructive jaundice, elevated serum CEA/CA 19-9, solid component, MPD ≥ 10 mm and MPD ≥ 5mm was associated with high grade dysplasia(HGD)/invasive carcinoma in all mucinous CLP. Increasing number of HRSCG or HRICG2012 features was associated with an significantly increased likelihood of malignancy. The PPV of HRSCG and HRICG2012 for HGD/invasive carcinoma was 46% and 62.5% respectively. The NPV of both LRSCG and LRICG2012 was 100%. Conclusion Both the guidelines were useful in the initial cross-sectional imaging evaluation of mucinous CLP. The ICG2012 guidelines were superior to the SCG
    No preview · Article · Aug 2014 · American journal of surgery
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The Sendai Consensus Guidelines (SCG) were formulated in 2006 and updated in Fukuoka in 2012 (FCG) to guide management of cystic mucinous neoplasms of the pancreas. This study aims to evaluate the clinical utility of the SCG and FCG in the initial triage of all suspected pancreatic cystic neoplasms. Overall, 317 surgically-treated patients with a suspected pancreatic cystic neoplasm were classified according to the SCG as high risk (HR(SCG)) and low risk (LR(SCG)), and according to the FCG as high risk (HR(FCG)), worrisome (W(FCG)), and low risk (LR(FCG)). Cystic lesions of the pancreas (CLP) were classified as potentially malignant/malignant or benign according to the final pathology. The presence of symptoms, proximal lesions with obstructive jaundice, elevated serum carcinoembryonic antigen/carbohydrate antigen 19-9 (CEA/CA 19-9), size ≥3 cm, presence of solid component, main pancreatic duct dilatation, thickened enhancing walls, and change in ductal caliber with distal atrophy were predictive of a potentially malignant/malignant CLP on univariate analyses. The positive predictive value (PPV) and negative predictive value (NPV) of HR(SCG) and HR(ICG2012) for a potentially malignant/malignant lesion was 67 and 88 %, and 88 and 92.5 %, respectively. There were no malignant lesions in both LR groups but some potentially malignant lesions such as cystic pancreatic neuroendocrine neoplasms with uncertain behavior were classified as LR. The updated FCG was superior to the SCG for the initial triage of all suspected pancreatic cystic neoplasms. CLP in the LR(FCG) group can be safely managed conservatively, and those in the HR(FCG) group should undergo resection.
    Full-text · Article · Feb 2014 · Annals of Surgical Oncology
  • [Show abstract] [Hide abstract]
    ABSTRACT: This study is a systematic review and meta-analysis that compares the short- and long-term outcomes of laparoscopic gastric resection (LR) versus open gastric resection (OR) for gastric gastrointestinal stromal tumors (GISTs). Comparative studies reporting the outcomes of LR and OR for GIST were reviewed. A total of 11 nonrandomized studies reviewed 765 patients: 381 LR and 384 OR. A higher proportion of high-risk tumors and gastrectomies were in the OR compared with LR (odds ratio, 3.348; 95 % CI, 1.248-8.983; p = .016) and (odds ratio, .169; 95 % CI, .090-.315; p < .001), respectively. Intraoperative blood loss was significantly lower in the LR group [weighted mean difference (WMD), -86.508 ml; 95 % CI, -141.184 to -31.831 ml; p < .002]. The LR group was associated with a significantly lower risk of minor complications (odds ratio, .517; 95 % CI, .277-.965; p = .038), a decreased postoperative hospital stay (WMD, -3.421 days; 95 % CI, -4.737 to -2.104 days; p < .001), a shorter time to first flatus (WMD, -1.395 days; 95 % CI, -1.655 to -1.135 days; p < .001), and shorter time for resumption of oral intake (WMD, -1.887 days; 95 % CI, -2.785 to -.989 days; p < .001). There was no statistically significant difference between the two groups with regard to operation time (WMD, 5.731 min; 95 % CI, -15.354-26.815 min; p = .594), rate of major complications (odds ratio, .631; 95 % CI, .202-1.969; p = .428), margin positivity (odds ratio, .501; 95 % CI, .157-1.603; p = .244), local recurrence rate (odds ratio, .629; 95 % CI, .208-1.903; p = .412), recurrence-free survival (RFS) (odds ratio, 1.28; 95 % CI, .705-2.325; p = .417), and overall survival (OS) (odds ratio, 1.879; 95 % CI, .591-5.979; p = .285). LR results in superior short-term postoperative outcomes without compromising oncological safety and long-term oncological outcomes compared with OR.
    No preview · Article · Jun 2013 · Annals of Surgical Oncology
  • [Show abstract] [Hide abstract]
    ABSTRACT: The aims of the study were to assess factors responsible for the reduction of preoperative anxiety in patients undergoing breast and abdominal surgeries. In particular, we investigated whether question prompt lists (QPL), patients' knowledge, or the communication skills of surgeons had effects on anxiety reduction. Patients were randomly assigned to QPL and control groups. Anxiety was assessed on the State Trait Anxiety Inventory. Both groups showed significant reduction in anxiety between initial consultation and one day prior to surgery, with QPL patients showing a trend towards a greater reduction of anxiety after surgery and a significant reduction at the first outpatient follow-up. Satisfaction with consultation and the doctor's ability to answer questions concerning diagnosis, and treatment were significantly associated with anxiety reduction. Effective anxiety reduction hinged on doctors' communication abilities and patients' satisfaction with the consultation.
    No preview · Article · Oct 2011 · Asian Journal of Surgery
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The sentinel lymph node biopsy (SLNB) is now an accepted alternative to the axillary lymph node dissection for pathologic evaluation of the axilla in patients with early breast cancer. The use of SLNB after neoadjuvant chemotherapy (NAC) is controversial. This meta-analysis aims to determine the feasibility and the accuracy of SLNB in the population of patients who are clinically node-negative after NAC for breast cancer.
    Full-text · Article · Mar 2011 · Journal of Surgical Oncology
  • [Show abstract] [Hide abstract]
    ABSTRACT: Presently, several systems for the prognostication of pancreatic endocrine neoplasms (PENs) exist and the most appropriate classification system has not been clearly defined. This study aims to validate the performance of the 2004 World Health Organization (WHO), European Neuroendocrine Tumor Society (ENETS), Memorial Sloan-Kettering Cancer Center (MSKCC), American Joint Committee for Cancer (AJCC) TNM staging and Bilimoria criteria in a cohort of patients with PENs who underwent surgery at a single institution. This study is a retrospective review of 61 consecutive patients who underwent surgical treatment for PEN. Actuarial disease-specific survival (DSS) of all 61 patients and recurrence-free survival (RFS) of 53 patients who had curative resection were analysed. On univariate analyses, tumour size ≥50 mm, non-curative resection, lymph node involvement, presence of distant metastases, presence of necrosis, mitotic count ≥2/10 hpf and poor differentiation were associated with decreased DSS. Tumour size ≥50 mm, lymph node involvement, lymphovascular invasion, presence of necrosis and mitotic count ≥2/10 hpf were associated with decreased actuarial RFS. All five staging systems were useful in stratifying the 61 patients according to actuarial DSS. However, the MSKCC grading and ENETS grading systems were not statistically significant in stratifying DSS in the 61 patients. In the 53 patients who underwent curative resection, the WHO, ENETS, MSKCC, AJCC staging and the MSKCC grading systems were successful in stratifying the patients according to actuarial RFS. However, the Bilimoria scoring and ENETS grading systems were not useful in prognosticating these 53 patients. All five classification systems were useful for the prognostication of surgically treated PENs in our patient cohort.
    No preview · Article · Jan 2011 · ANZ Journal of Surgery
  • [Show abstract] [Hide abstract]
    ABSTRACT: The present study is designed to determine the feasibility and impact of the introduction of laparoscopic wedge resection as a surgical option for the treatment of suspected small/medium-sized (<7 cm) gastric gastrointestinal stromal tumors (GISTs). The study involved a retrospective review of 53 consecutive patients who underwent laparoscopic or open wedge resection of a suspected gastric GIST. It was divided into two consecutive time periods wherein laparoscopic resection was a surgical option only in the latter period. Comparisons were made between the outcomes of patients who underwent laparoscopic versus open wedge resection and the outcomes of patients treated during the two consecutive time periods (to determine the impact of the introduction of laparoscopic wedge resection), Fourteen patients (26%) underwent laparoscopic wedge resection with 1 conversion. The pathological exam showed that 41 patients (77%) had a GIST. Laparoscopic resection was significantly associated with a longer operative time, an earlier return of bowel function, earlier resumption of liquid and solid diet, decreased duration of parenteral or epidural analgesia use, and shorter postoperative hospitalization compared to open resection. There was no statistical difference in the rate of R1 resection and actuarial recurrence-free survival for the two approaches. Comparison between the two time periods demonstrated that the introduction of the laparoscopic approach in the latter period resulted in an earlier return of bowel function, earlier resumption of liquid and solid diet, and decreased duration of parenteral or epidural analgesia. Laparoscopic wedge resection for gastric GIST can be safely adopted. It is associated with a more favorable perioperative outcome than the open approach. Its introduction as a surgical option has resulted in an improvement in perioperative outcomes without compromising oncologic safety at our institution.
    No preview · Article · Aug 2010 · World Journal of Surgery
  • [Show abstract] [Hide abstract]
    ABSTRACT: Complete tumor resection with clear margins including adjacent organs is the treatment of choice for gastrointestinal stromal tumors (GISTs). However, true tumor invasion of adjacent organs has been reported to be rare. Concomitant distal pancreatectomy (DP) for suspected tumor infiltration is not infrequently performed during resection of large gastric GISTs. This study aims to determine the true frequency of adjacent organ involvement by large gastric GISTs with particular attention to the pancreas and compares the outcome after curative resection with and without a concomitant DP in order to determine if DP is truly necessary. A retrospective review of 37 patients who underwent curative resection of large (>or=10 cm) gastric GISTs was conducted. Wedge resections were performed in 22, partial gastectomies in nine, and total gastrectomies in six patients. The median operative time was 180 min (range, 60-330 min), and the patients had a median postoperative stay of 8 days (range, 4-29 days). Overall, there were eight (22%) morbidities including two (5%) mortalities. Nineteen (51%) had concomitant adjacent organ resection, and these included 15 (41%) DPs with splenectomies. Direct organ invasion was demonstrated in 5/19 patients (26%) and 7/30 organs (23%) resected. Only 1/15 (6.7%) DP specimens demonstrated tumor infiltration. Comparison between the patients with and without a concomitant DP demonstrated that performance of a DP was associated with a longer operation time [225 min (range, 105-305 min) vs 158 min (60-330 min), P=.002)], increased postoperative stay [9 days (range, 7-29 days) vs 7.5 days (4-19 days), P=.042], and increased postoperative morbidity [6 (40%) vs 2 (9%), P=.025]. The DP cohort also had a statistically significant poorer 5-year recurrence free survival (22% vs 60%, P=.017). Although adjacent organ involvement is not uncommon with large gastric GISTs, concomitant DP is usually unnecessary as direct pancreatic invasion is rare. Furthermore, concomitant DP with splenectomy is associated with an increase in postoperative morbidity.
    No preview · Article · Apr 2010 · Journal of Gastrointestinal Surgery
  • Brian K.P. Goh · Sittampalam M Kesavan · Wai-Keong Wong

    No preview · Article · Jul 2009 · Gastroenterology
  • [Show abstract] [Hide abstract]
    ABSTRACT: Presently, the need for and choice of preoperative localization tests for insulinomas remain controversial. We report the results from a single institution experience whereby the management policy adopted was that of accurate preoperative localization before surgical exploration. From 1990 to 2008, 17 patients with a clinical and biochemical diagnosis of an insulinoma who underwent surgery were retrospectively reviewed. The diagnosis of all insulinomas were confirmed pathologically. All tumors were localized preoperatively and an average of 2.2 preoperative localization studies including 1.4 noninvasive studies and 0.8 invasive studies were utilized per patient. Invasive localization modalities were more sensitive (92%) than noninvasive modalities in localizing insulinomas (71%). Intra-arterial calcium stimulation with hepatic venous sampling was the most sensitive invasive modality (100%), whereas magnetic resonance imaging was the most sensitive noninvasive modality (63%). Fifteen of 17 tumors (88%) were localized intraoperatively via inspection/palpation and/or intraoperative ultrasonography. Both insulinomas which were not localized intraoperatively were localized correctly to the distal pancreas via preoperative transhepatic portal venous sampling. None of the patients required a blind resection or surgical reexploration for failed localization. All 17 patients underwent complete surgical resection which included eight enucleations and nine distal pancreatectomies with a cure rate of 94% (16/17) at a median follow-up of 35 (range, 1-217) months. The postoperative morbidity and long-term outcome of enucleation was similar to distal pancreatectomy despite a higher rate of microscopic margin involvement. Accurate preoperative localization of insulinomas is useful as it eliminates the need for blind distal pancreatectomy and avoids reoperation. Complete surgical resection is the treatment of choice, and whenever possible, a pancreas-sparing approach such as enucleation should be adopted.
    No preview · Article · Apr 2009 · Journal of Gastrointestinal Surgery
  • [Show abstract] [Hide abstract]
    ABSTRACT: Gastrointestinal stromal tumors (GISTs) arising from outside the gut wall also termed extragastrointestinal stromal tumors (EGISTs) are reported to be rare. Presently, their pathogenesis remains controversial, and recently, it has been proposed that EGISTs may be the result of extensive extramural growth of GISTs which lose contact with the gut wall. This study presents a single-institution experience with eight EGISTs and compares their clinicopathological features with mural GISTs in order to determine further insight to their possible origin. Between 1997 and 2008, 156 patients with pathologically proven CD117-positive primary GISTs were retrospectively reviewed. Eight tumors were identified as EGISTs, 104 were gastric GISTs, and 44 were small-bowel GISTs. Mural GISTs were classified as extramural or intra/transmural according to their gross pattern of growth. There were five male and three female patients with a median age of 58 years (range, 42-81 years). All patients were symptomatic, and the tumors were located in the greater omentum (n = 2), lesser sac (n = 2), lesser omentum, retroperitoneum, small-bowel mesentery, and pancreas. The median tumor size was 140 mm (range, 55 to 220 mm). Seven of eight EGISTs were found to be in close association to the adjacent gut wall. Pathological examination demonstrated that two tumors demonstrated focal involvement of the muscularis propria of the adjacent gut wall. Four tumors demonstrated tumor abutting or adherent to the serosa but no muscle involvement and one tumor was separated from the serosa. Comparison between the clinicopathological features of EGISTs with extramural GISTs and intra/transmural GISTs demonstrated that EGISTs were significantly larger [140 range (55-220) mm vs 80 (5-260) mm vs 50 (15-190) mm, P = 0.049, P < 0.001 respectively]. The occurrence of true EGISTs is rare. Most cases demonstrate some form of communication or contact with the gut wall, and EGISTs are significantly larger than extramural or intra/transmural GIST. These observations suggest that most, if not all, cases of EGISTs are likely to represent mural GISTs with extensive extramural growth with eventual loss of contact with the muscle layer of the gut.
    No preview · Article · Mar 2009 · Journal of Gastrointestinal Surgery
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Gastric volvulus is a rare condition. Presenting acutely, mesenteroaxial gastric volvulus has characteristic symptoms and may be easily detected with upper gastrointestinal contrast studies. In contrast, subacute, intermittent cases present with intermittent vague symptoms from episodic twisting and untwisting. Imaging in these cases is only useful if performed in the symptomatic interval. We describe a patient with a long history of intermittent chest and epigastric pain. An earlier barium meal was not diagnostic. Diagnosis was finally secured during the current admission by a combination of (1) serum investigations, (2) endoscopy, and finally (3) computed tomography. Non-specific and misleading symptoms and signs may delay the diagnosis of intermittent, subacute volvulus. Imaging studies performed in the well interval may be non-diagnostic. Elevated creatine kinase and aldolase of a non-cardiac cause and endoscopic findings of ischaemic ulceration and difficulty in negotiating the pylorus may raise the suspicion of gastric volvulus. In this case, abdominal computed tomography with spatial reconstruction was crucial in securing the final diagnosis.
    Preview · Article · Dec 2008 · Journal of Medical Case Reports
  • [Show abstract] [Hide abstract]
    ABSTRACT: To critically analyze a large single-institution experience with distal pancreatectomy (DP), with particular attention to the risk factors, outcome, and management of the postoperative pancreatic fistula (PF). Retrospective study. Tertiary referral center. A total of 232 consecutive patients with pancreatic or extrapancreatic disease necessitating DP over 21 years. Twenty-one patients underwent spleen-preserving DP, 117 underwent DP with splenectomy, and 94 underwent DP with multiorgan resection. The perioperative and postoperative data of patients who underwent DP were analyzed. This included factors associated with postoperative morbidity with particular attention to the PF (defined by the International Study Group of Pancreatic Fistula) and changing trends in operative and perioperative data during the study period. The overall operative morbidity and mortality were 47% (107 patients) and 3% (7 patients), respectively. During the study period, the rates of resection increased from 3 cases to 23 per year, and increasingly these were performed for smaller and incidental lesions. The morbidity rate remained unchanged, but there was a decline in postoperative stay and the need for care in the intensive care unit. Pancreatic fistulas occurred in 72 patients (31%); 41 (18%) were grade A, 13 (6%) grade B, and 18 (8%) grade C. Increased weight, higher American Society of Anesthesiologists score, blood loss greater than 1 L, increased operation time, decreased albumin level, and sutured closure of the stump without main duct ligation were associated with a postoperative PF on univariate analysis. A DP with splenectomy was associated with a higher incidence of grade B or C PF and non-PF-related complications. Ninety-two percent of PFs were successfully managed nonoperatively. Clinical outcomes correlated well with PF grading, as evidenced by the progressive increase in outcome measures such as postoperative stay, readmissions, reoperations, radiologic interventions, and non-PF-related complications from grade A to C PFs. Pancreatic fistula is the most common complication after DP and its incidence varies depending on the definition applied. Several risk factors for developing a PF were identified. Splenic preservation after DP is safe. The grade of a PF correlates well with clinical outcomes, and most PFs may be managed nonoperatively.
    No preview · Article · Nov 2008 · Archives of surgery (Chicago, Ill.: 1960)
  • [Show abstract] [Hide abstract]
    ABSTRACT: Pancreaticoduodenectomy (PD) for locally advanced stomach cancer involving duodenum or/and pancreatic head was controversial and rarely carried out. It was mainly reported from the Japanese institutions. A review of prospective database from January 2003 to December 2006 of patients who had locally advanced stomach cancer involving duodenum or/and head of pancreas that precluded curative subtotal gastrectomy who underwent diagnostic laparoscopy or exploratory laparotomy to exclude peritoneal metastatic disease. Patients were advised to undergo neoadjuvant chemotherapy before PD. Seven patients underwent PD during the above-mentioned period. Only four patients had neoadjuvant chemotherapy before PD. The median operative time was 8 h (range 6-9 h). Five patients had combined tranverse colectomy done. There was no 30-day operative mortality or re-operation. Three patients developed controlled pancreatic leaks and fistulas that were successfully treated with conservative measures. The length of hospital stay was 10-53 days (median 15 days). Median survival was 13 months and 2-year survival rate was 60%. Patients who received neoadjuvant chemotherapy seemed to have better survival rate (P = 0.039). Our initial experience has shown that with careful and stringent patients selection, PD for locally advanced stomach cancer can be carried out with acceptable morbidity and mortality. Early results for patients who received neoadjuvant chemotherapy showed trend towards prolonged survival. However, longer follow up and further patient recruitment are needed to confirm our initial optimistic findings.
    No preview · Article · Oct 2008 · ANZ Journal of Surgery
  • [Show abstract] [Hide abstract]
    ABSTRACT: This study aims to validate and compare the performance of the National Institute of Health (NIH) criteria, Huang modified NIH criteria, and Armed Forces Institute of Pathology (AFIP) risk criteria for gastrointestinal stromal tumors (GISTs) in a large series of localized primary GISTs surgically treated at a single institution to determine the ideal risk stratification system for GIST. The clinicopathological features of 171 consecutive patients who underwent surgical resection for GISTs were retrospectively reviewed. Statistical analyses were performed to compare the prognostic value of the three risk criteria by analyzing the discriminatory ability linear trend, homogeneity, monotonicity of gradients, and Akaike information criteria. The median actuarial recurrence-free survival (RFS) for all 171 patients was 70%. On multivariate analyses, size >10 cm, mitotic count >5/50 high-power field, tumor necrosis, and serosal involvement were independent prognostic factors of RFS. All three risk criteria demonstrated a statistically significant difference in the recurrence rate, median actuarial RFS, actuarial 5-year RFS, and tumor-specific death across the different stages. Comparison of the various risk-stratification systems demonstrated that our proposed modified AFIP criteria had the best independent predictive value of RFS when compared with the other systems. The NIH, modified NIH, and AFIP criteria are useful in the prognostication of GIST, and the AFIP risk criteria provided the best prognostication among the three systems for primary localized GIST. However, remarkable prognostic heterogeneity exists in the AFIP high-risk category, and with our proposed modification, this system provides the most accurate prognostic information.
    No preview · Article · Jun 2008 · Annals of Surgical Oncology
  • [Show abstract] [Hide abstract]
    ABSTRACT: Intraabdominal schwannomas are rare, benign tumors. This study presents a single institution experience with 12 such tumors. Between 1991 to 2006, 12 patients with a pathologically proven intraabdominal schwannoma were identified from a series of 216 mesenchymal tumors and were reviewed retrospectively. There were nine females and three male patients with a median age of 58 years (range 35-88 years). Eleven patients were symptomatic, and the tumors were located in the stomach (n = 8), jejunum, colon, rectum, and lesser sac. Multiple preoperative investigations including endoscopies with biopsies and computed tomography (CT) scans were performed, but none yielded a correct definitive preoperative diagnosis. The median tumor size was 52 mm (range 18-95 mm). Pathological examination demonstrated the 11 gastrointestinal tract (GIT) schwannomas to be solid homogenous tumors, which were highly cellular and were composed of spindle cells with positive staining for S100 protein. The pathological appearance of the lesser sac schwannoma was distinct as it demonstrated cystic degeneration with hemorrhage and Antoni A and B areas on microscopy typical of soft tissue schwannomas. All 12 patients were disease-free at a median follow-up of 22 months (range 1-120 months). Intraabdominal schwannomas are rare tumors, which are most frequently located within the GIT. GIT schwannomas are difficult if not impossible to diagnose preoperatively as endoscopic and radiologic findings are nonspecific. The treatment of choice is complete surgical excision because of diagnostic uncertainty, and the long-term outcome is excellent as these lesions are uniformly benign.
    No preview · Article · May 2008 · Journal of Gastrointestinal Surgery
  • [Show abstract] [Hide abstract]
    ABSTRACT: Present surgical opinion is divided regarding the optimal method for the treatment of duodenal gastrointestinal stromal tumor (GIST) with some supporting the selective use of limited resection (LR) versus others who prefer pancreaticoduodenectomy (PD). A retrospective review of 22 patients who underwent surgery for suspected GIST involving the duodenum. There were 15 GISTs, 1 leiomyosarcoma and 6 other non-GIST benign submucosal tumors. Seven patients underwent LR and seven underwent PD for GIST. The median follow-up was 42 (range, 2-174) months. Patients who underwent LR versus PD had similar mean disease-specific survival [144 (95% CI, 92-196) vs. 130 (95% CI, 82-127) months, P = 0.808] and recurrence rates (14% vs. 29%, P = 0.515). All recurrences occurred at distant sites. Comparison between LR versus PD demonstrated that LR was associated with a significantly shorter operation time [125 (range, 50-305) vs. 350 (range, 210-465) min., P = 0.001] but similar morbidity rate (23% vs. 43%, P = 0.357). Comparison between GIST and other benign tumors demonstrated that size was the only statistically significant distinguishing factor [8.5 (range, 2.5-18.0) vs. 2.5 (range, 1.5-8.0) cm, P = 0.014]. Benign non-GIST tumors may be distinguished from duodenal GIST as they are smaller in size. LR is a viable treatment option for suspected GIST involving the duodenum.
    No preview · Article · May 2008 · Journal of Surgical Oncology
  • [Show abstract] [Hide abstract]
    ABSTRACT: This study aims to determine the use of preoperative clinical, biochemical, and cross-sectional imaging features for predicting malignancy in cystic lesions of the pancreas (CLP). Two hundred twenty patients who underwent operations for CLP or suspected CLP were reviewed. Patients were divided into two groups, patients undergoing operations for pseudocysts and patients undergoing operations for suspected cystic neoplasms. The predictive effect of various preoperative factors on the malignant potential of CLP was evaluated. Forty-four patients with a preoperative diagnosis of pseudocysts underwent operations for complications of pseudocyst. Forty-two were confirmed pathologically to have pseudocysts, but two were found, unexpectedly, to harbor malignant lesions. One hundred seventy-six patients underwent operations for suspected pancreatic cystic neoplasms. There were 70 benign, 51 potentially malignant, and 55 malignant CLP. On multivariate analysis, three factors, ie, elevated serum carcinoembryonic antigen (CEA) or carbohydrate antigen 19-9; cyst size > 3 cm; and presence of one or more of three morphologic features, such as solid component; peripheral calcification; and main duct dilation on cross-sectional imaging were independent predictors of malignancy. Presence of two or three of these factors had a positive predictive value of 88% in predicting a premalignant or malignant CLP. Most pancreatic pseudocysts can be accurately diagnosed preoperatively. In patients with suspected pancreatic cystic neoplasms, elevated serum CEA or carbohydrate antigen 19-9, cyst size > 3 cm, and presence of suspicious morphologic features on imaging are predictors of potentially malignant or malignant CLP. Patients with a high likelihood of a potentially malignant or malignant lesion based on these three factors should undergo operation without additional investigations.
    No preview · Article · Feb 2008 · Journal of the American College of Surgeons

Publication Stats

1k Citations
141.97 Total Impact Points


  • 2002-2015
    • Singapore General Hospital
      • • Department of General Surgery
      • • Department of Surgery
      Tumasik, Singapore
  • 2013
    • Duke-NUS Graduate Medical School Singapore
      Tumasik, Singapore
  • 2005-2011
    • National Cancer Centre Singapore
      • Department of Surgical Oncology