Background and Aims

Diagnostic imaging is essential in children with unclear cause of cervical lymphadenopathy. Recently, ultrasound elastography was presented as a promising technology to distinguish between malignant and benign lymph nodes. We aimed to assess the accuracy of elastography in detecting pediatric malignant cervical lymph nodes, and if this modality can obviate the need for surgical biopsies.

Methods

A prospective study from September 2017 to September 2020 included 64 children with persistent cervical lymphadenopathy. Patients were evaluated by meticulous history and physical assessment. B-mode ultrasound, color doppler, and sonoelastography were conducted thereafter. Elastography scans were classified into five patterns, and patterns from 3-5 were considered as malignancies. All children underwent open biopsies followed by pathological examination. Results of tissue diagnosis were compared to patterns of elastography to determine its accuracy.

Results

Twenty-eight patients (43.8%) had malignant nodes and the remaining 36 (56.2%) were due to benign causes. Elastography patterns of 1-2 were documented in 30 patients, and all of them were diagnosed as benign lesions. Patterns of 3-5 were demonstrated in 34 patients. Out of them, 28 were confirmed as malignancies, whilst 6 children were of benign nature (false-positive). All false-positive results occurred in pattern 3 of elastography. There were no complications encountered regarding surgical biopsies, and all patients were discharged on the same day of the procedure. Ultrasound elastography achieved 100% and 85.7% sensitivity and specificity, respectively, and an overall accuracy of 90.6% in the differentiation between malignant and benign entities. The overall accuracy of B-mode and color doppler were 75% and 82.2%, respectively.

Conclusions

Elastography is a useful tool that should be added to ultrasound modalities and diagnostic algorithm for pediatric cervical lymphadenopathy. Surgical biopsy in eligible patients is imperative to commence proper therapy or to discharge the child. Despite favorable results of elastography, it cannot replace surgical biopsy or change its indications.

Background and Aims

Paediatric Hodgkin Lymphoma (HL) is a major cause of morbidity and mortality in children. In low to middle-income countries like South Africa, there is a need to understand the clinical practices surrounding diagnosis and surveillance of HL to reduce the burden on health systems. Understanding the clinical utility of PET/CT scans may decrease repeated tissue biopsies during disease surveillance. This study aimed to compare the utility of the two investigations.

Methods

The study is a retrospective cohort study of patients aged less than 18 years, treated for Hodgkin Lymphoma at Chris Hani Baragwanath Academic Hospital from January 1^st 2009 to December 31^st 2018. Ethics HREC No. M210732.

Results

Among the 54 patients included in the study, there was a male to female ratio of 5:1 and a mean age at diagnosis of 9 years old (σ: ±3.7). Nodular Sclerosing HL was the most frequent subtype at diagnosis and relapse with frequencies of 42.59% (n=23) and 44.44% (p=0.372) respectively. Seventy percent of patients (n=38) received a PET/CT and tissue biopsy during their initial diagnostic workup, and 37% (n=17) of patients received a PET/CT scan during surveillance. Tissue biopsy and PET/CT, showed moderate agreement of 48.2% (κ = 0.14) in diagnosing relapsed disease during surveillance. The false negative rate for tissue biopsy during surveillance was 42.9%. No patients had a negative PET/CT during surveillance with a positive tissue biopsy result (n=18).

Conclusions

This study is the first cohort to explore the clinical utility of PET/CT scans and tissue biopsies in a low-resourced setting. Our findings showed moderate agreement between the modalities in diagnosing relapsed disease during surveillance. A portion of this discordance can be attributed to false negative tissue biopsy results. While the sample is limited, our findings are consistent with the high negative predictive value of PET/CT scans reported in the literature.

Background and Aims

Central venous access device (CVAD) insertion is one of the most common paediatric surgical procedures. Our aim was to identify the failure rate and the associated risk factors in our centre.

Methods

A prospective audit was undertaken between June 2020-December 2021 for CVAD insertions in our centre. Failure was defined as both mechanical (block, break or dislodge) and infection, necessitating unplanned removal and calculated per 1000 line days. Statistical analysis was performed using Kaplan Meier survival and p value <0.05 was deemed significant.

Results

A total of 462 CVADs (301 patients) were inserted (234 Hickmans, 101 Portacaths, 98 Peripherally Inserted Central Catheters, 29 Haemocaths). Chemotherapy was the most common indication for insertion (39%). The majority were USS guided percutaenous insertions (94.5%) with a mean age of 68(SD 63.5) months and weight of 22.2(SD 19.4) kg. The IJV was the most frequent vessel accessed (68%) and most line tips were located at the SVC/RA junction (63.5%). The total number of line days was 88248, with an overall failure rate of 1.6 per 1000 line days, of which 58.5% were due to mechanical reasons (0.9 per 1000 line days). Portacaths were least likely to fail (p<0.001). Increased rates of failure were associated with more lumens (p<0.001), open insertion (p=0.036), emergency setting (p=0.042) and age less than 60 months (p=0.021). Weight was not a significant factor (p=0.466).

Conclusions

We identified multiple factors associated with line failure. Mechanical failure represents a significant proportion of line failures in children, which historically may have been overlooked compared to infection. As expected, failure is higher in external lines when compared to portacaths. Future consideration for studies investigating the value of fixation devices and postoperative care bundles to reduce potentially preventable causes for failure are important, alongside robust ways to capture and monitor this data.

Background and Aims

Postoperative complications in children with solid tumors are a major cause of morbidity and mortality. The Clavien – Dindo classification is validated across various surgical specialties and thus may be also used to objectify complications in pediatric solid tumors. We aimed to assess the surgical complications in pediatric solid tumors objectively and validate this classification.

Methods

This is a retrospective study over 5 years conducted in a tertiary care centre from January 2017 to December 2021. All patients with pediatric solid tumors, except brain and bone tumors, were included. Postoperative complications were assessed and graded on the basis of Clavien- Dindo classification system. Outcome measures like operative time, blood loss, length of hospital stay, and redo surgery were evaluated.

Results

A total of 188 patients were included. there was a statistically significant relation between grades of postoperative complications and duration of hospital stay with a mean postoperative stay of 29 days in children with grade 4b complications. There was a significant correlation with operative time with a mean operative time of 150 minutes in children with grade 4 complications as compared to just 75 to 100 minutes in low-grade complicated children. Mean intraoperative blood loss was more than 300 ml in grade 4 complicated children as compared to just 50 to 100 ml in low-grade complicated children. More than 50% of children with grade 4 complications had recurrence as compared to children with lower grades of complications. Redo surgery was done in 8% of children.

Conclusions

Clavien – Dindo classification can be used in pediatric surgical oncology to assess postoperative complications in an objective way. Children with a higher grade of complications have increased morbidity in form of longer hospital stays, blood transfusions, and a higher rate of redo operations. This also leads to subsequent delays in adjuvant treatment.

Background and Aims

Since its introduction in children in 2001, robotic surgery (RS) has been proposed by experienced pediatric laparoscopic surgeons in a continuously expanding range of procedures. The technical capabilities of the robot may be ideal for complex pediatric surgical cases that require intricate dissection, such as in pediatric cancer surgery. Although the comfortable age and weight for RS candidates are progressively lowering and new potential indications are currently being explored, some limits must be acknowledged and possibly recognized during surgery, to avoid complications or emergency conversion. This study aims to investigate these issues through a retrospective analysis of a single-center experience.

Methods

Data were collected from patients affected by cancer treated with a robotic approach in two periods (2015-2016 and 2019-2021). Surgical indication and approach were discussed within the multidisciplinary tumor board. Descriptive statistics, data on conversions, and complications are reported.

Results

Thirty-eight patients were operated on at a median age of 5 years (4 months - 24 years), with a median weight of 22 Kg (6 - 96 Kg). Median console time was 76 minutes. In the majority of patients, 4 robotic trocars were used. In the 4 younger patients, 3 robotic trocars were used. The median distance between trocars was 4.5 cm (4 - 10cm). Eleven patients (28.9%) required conversion to open surgery due to difficult vascular management (6), bleeding (2), undefined margins (2), and adhesions (1). Two (5%) intraoperative complications (bleeding, gallbladder injury) and 3 (7.9%) postoperative complications (incisional hernia, bowel obstruction, pelvic fluid collection) were reported. The average hospital stay was 5 days.

Conclusions

RS allows a more comfortable and precise minimally invasive treatment in pediatric cancer surgery. An adequate learning curve is mandatory to better address surgical indications, reduce complications, and optimize trocars positioning. The potential risk for vascular injury is the main reason for conversion to open surgery.

Background and Aims

Robotic surgery in pediatric oncology is rare. Disponibility of robotics in our pediatric surgical department since 4 years offer the new possibility of oncological resection by robotics.

Pediatric oncological surgery by robotics started in june 2019 with and we report the experience of a single team during a period of 2 years and a half.

Methods

A prospective study was started in june 2019 and collected data were reviewed in februar 2022. After an experience of classical robotic procedures in children, we started with pediatric oncological surgery beginning with selected cases approved by a certified tumor board and increasing the indications as we were at the end of the study.

Results

24 tumors were resected in this period and 22 were malignant.

The median age at surgery was 5.5 years (range 3 mnths to 15 years old). Renal tumors were the largest group (n=14) including 11 Wilms tumors, 1 mesoblastic tumour and 2 carcinoma. Neuroblastic tumours was the second largest group with 6 neuroblastomas. The other tumors were 2 germ cell tumors, 1 pseudopapillary pancreatic tumour and 1 adrenal adenoma.

There was no conversion to an open approach. No intraoperative tumour rupture occured.The median hospital stay was 3 days. One boy suffering of metastatic medullary carcinoma died of his metastatic disease.

Conclusions

Robotic surgery for pediatric tumours can be safely started in a team with a heavy laparoscopic experience and could rapidly be "the procedure of choice" for selected cases and after a certified tumour board meeting approbation.

Background and Aims

Children with suspected malignancy who present with central (mediastinal or abdominal) lesions not amenable to percutaneous biopsy pose a significant diagnostic challenge. Very few studies have evaluated the role of endoscopic ultrasound (EUS)-guided fine needle aspiration (FNA) for the diagnosis of malignancy in paediatric patients. We describe a series of seven children with solid tumours in whom EUS-FNA was utilised for diagnosis.

Methods

We performed a retrospective review of children underdoing EUS biopsy in a single principal treatment centre from January 2020 -January 2022. Demographic, diagnostic and treatment data were collated along with a record of complications. All EUS-FNA samples were taken under general anaesthetic, using a 22G Echo 3 needle.

Results

Seven children at our centre had suspected malignant disease not amenable to percutaneous biopsy and underwent EUS-FNA during this two-year period. Age and weight ranges were 1 year 9 months -16 years and 14-50kg. Four were trans-oesophageal biopsies of mediastinal masses and three were trans-gastric biopsies of abdominal masses. 6/7 children had diagnostic biopsies. Three had Hodgkin lymphoma; the remaining three were diagnosed with extra-gonadal yolk-sac tumour, mediastinal paraganglioma and gastrointestinal stromal tumour.

One biopsy was undertaken on Paediatric Intensive Care, the others in an operating theatre. 4/7 were combined with other procedures under the same anaesthetic. 6/7 were uncomplicated. One child with an extremely large spleen infiltrated with disease underwent splenectomy two days after trans-gastric biopsy of splenic hilar nodes, although it was not clear whether the biopsy contributed directly to the small amount of splenic bleeding noted at laparotomy.

Conclusions

EUS FNA biopsy has a high diagnostic yield and is a safe option for children with solid tumours not amenable to percutaneous biopsy. We have utilised it successfully at our centre in children from 21 months of age and 14 kg in size.

Background and Aims

The Enhanced Recovery After Surgery (ERAS) is a pathway designed to achieve ealy recovery for patients undergoing major surgery using a multidisciplinary team approach. In the pas two decades, ERAS was improved for adults and proved efficiency for decrasing length of stay (LOS) without increasing post operative complications. The effect of RAS for children are not well known. We wanted to asess the feasability of enhanced recovery protocols for children undergoing in oncological surgery.

Methods

We collected the medical records of consecutive patients who underwent oncological surgery since 2015 to 2021.The primay aim was the feasibility of ERP and children were divided into 2 groups : Group A : ERAS was feasible and Group B ERAS was impossible. criteriae were : drain > 5 days , urinary catheter > 5 days length of stay > 5 days

Data were analysed according to the Clavien Dindo classification and 30 day readmission rate. Statistical analysis was univariate and multivariate logistic regression to wether RRP were predictive of LOS

Results

50 patients were included followed up for 224 days (31-660) Group A : 37 patients were elligible Group B : 13 patients coudn't benefit of ERAS .

Median Age of surgery was 6.4 years group A : 5 and group B : 12 p>0.2

Median weight was 19 kgs Group A : 18 and Group B : 21 p>0.2

74 % of the patients reached the target of hospital discharge of the fith post operative day . Few complications ocurred : 16% without any association with the feasibility of ERAS . One month readmission rate was 3/50 (6%) only in the ERAS group (p>0.2)

Conclusions

These preliminary results suggest that ERAS implementation for selected oncological surgery in chidren is suitable , associated with benefits and reduction of overall costs without increasing complications and readmissions.

Background and Aims

Standard sentinel lymph node procedure (SNP) in pediatric cancer consists of a preoperative injection with^99mTechnetium-nanocolloid in combination with an optional intraoperative injection with blue dye for visual guidance. However, blue dye has disadvantages and the visual detection rate is low with only 60% of the sentinel lymph nodes (SLNs) staining blue. In adult oncology, near-infrared (NIR) fluorescence imaging using indocyanine green (ICG) has been shown to be a safe and accurate method for intra-operative visual detection of SLNs, with a higher sensitivity (up to 97%) compared to blue dye. Therefore, our aim is to determine the feasibility of the addition of ICG to ^99mTechnetium-nanocolloid for visual detection of the SLN in pediatric patients.

Methods

Twelve pediatric melanoma, squamous cell carcinoma and sarcoma patients were prospectively included. Intra-operative, the SLN was first detected with fluorescence and the gamma probe. Postoperatively, fluorescence was quantified by tumor-to-background ratio (TBR) and surgeons evaluated the use of ICG using a standardized questionnaire.

Results

In all patients at least 1 SLN was detected with a median of 2. In 9/12 (75%) patients, the SLNs were visible transcutaneous. Twenty-five SLNs were identified on pre-operative imaging, of which 22 (88%) were detected with fluorescence, while 24 (96%) could be detected based on radioactivity. Of all intraoperative detected SLNs, 27/29 (93%) were fluorescent and 29/29 (100%) were radioactive. Furthermore, ICG led to the identification of five additional SLNs as compared to preoperative imaging. These additional SLNs were also radioactive. The average TBR in vivo of the SLNs was 7.2 and no adverse events occurred. The surgical evaluation showed that ICG assisted in SLN detection and was easy to use.

Conclusions

ICG combined with ^99mTechnetium-nanocolloid for the SNP is a feasible procedure in pediatric patients. It showed an accurate detection rate, was helpful for visual guidance, and did not lead to adverse events.

Background and Aims

Lymph node sampling is critical to Wilms tumour staging with failure associated with an increased risk of relapse.

Fluorescent-guided nodal mapping (FGNM) using ICG is used in some adult cancer patients to identify lymph nodes during resection surgery.

No technique is currently used for paediatric tumour nephrectomy, and we describe the early results of paediatric FGNM.

Methods

Two tertiary level referral centers independently began FGNM. Ethical approval was obtained as per institutional practice. Consecutive patients fulfilling eligibility criteria were enrolled.

ICG was injected during minimally invasive (MIN) and open (ON) radical nephroureterectomy in patients with unilateral renal tumours managed on International Society of Paediatric Oncology (SIOP), or Children’s Oncology Group (COG) protocols.

In MIN patients, an intra-parenchymal (IP) weight dependent ICG dose was injected. In ON patients a single peri-hilar (PH) dose of 5mg of ICG was injected.

Successful lymph node mapping was defined as presence of fluorescence signal in draining lymph nodes.

Results

Eight patients(4M,4F) of median 2(range 2-5) years were included. Six had localised and two metastatic disease at presentation. Seven patients had Wilms tumour (WT) and one had clear cell sarcoma.

All patients underwent intra-operative real-time FGNM with no complications from the ICG injection. Fluorescence was observed in draining nodes at 5-15 minutes following injection.

Seven patients had successful FGNM with 8 (3-15) nodes sampled. Left-sided tumours drained to hilar and periaortic lymphatic and right-sided tumours to hilar, retro-caval, inter-aortocaval, and precaval lymphatic zones.

Three patients had positive lymph nodes of which two were fluorescent.

There was one conversion to open surgery and one child developed a caecal volvulus.

Conclusions

ICG injection is feasible without upstaging patients on COG or SIOP protocols based on this small set of patients. Early results reveal adequate nodal harvest in WT without protocol violation or complications although a larger randomised study is required.

Background and Aims

Near InfraRed Fluorescence(NIRF) using IndoCyanine Green(ICG) has been available for use for several years and has recently transferred to the paediatric surgical sphere. There is increasing interest in its use in paediatric oncology surgery due to the Enhanced Permeability and Retention effect. We wanted to ascertain if Minimally Invasive Fluorescent Guided Surgery (MIFGS) could give better identification of oncological lesions; allow more precise tumour dissection and improve nodal sampling.

Methods

This is a single site, open label, prospective feasibility study of patients receiving ICG as part of a research trial in MIFGS. Funding was obtained from Children’s Cancer and Leukaemia Group, UK. All patients undergoing MIS paediatric oncology surgery were eligible for enrolment.

Patient demographics, dose and timing of ICG injection, intra-operative appearances, histology and other data were reviewed. The pathologist was blinded to ICG use.

Results

Thirteen patients were enrolled but two were excluded prior to surgery. The remaining 11 patients underwent 12 procedures. Diagnoses were Wilms tumour in 4, Neuroblastoma in 3, osteosarcoma, Hodgkins lymphoma, Inflammatory myofibroblastic tumour and suspected thoracic sarcoma. All patients underwent treatment according to SIOP protocols

Those patients having nodal mapping received ICG intra-operatively. All others received ICG intravenously at induction of anaesthesia.

Three patients underwent MIFGS metastectomy with histologically clear margins. Two patients underwent MIFGS resection of which both had clear margins. Four patients underwent nodal mapping during MIS tumour nephrectomy a median of 8(3-9) nodes sampled One tumour was unsuitable for resection. Two patients had tumours which did not take up ICG.

Conclusions

ICG can be given during induction of anaesthesia rather than 24 hours beforehand.

In this small series MIFGS shows promising results in paediatric oncology surgery.

A multicentre, multinational, open label, single blind, randomised controlled trial(GLOSurgery) is in the setup phase which should give us an idea about the efficacy of this approach,

Background and Aims

Fluorescence-guided surgery (FGC) is a method that makes it possible to identify structures during surgery. The basic principle involves the injection of indocyanine green and subsequent visualization with an infrared camera The aim of this study is to describe the initial experience with FGC.

Methods

Retrospective study of patients undergoing FGC from January 2020 to March 2022. Classification of the type of FGC, the dose of indocyanine, the time when the indocyanine was infused, and what was the usefulness of the FGC, classifying according to a modified Likert scale from 1 to 5. 1: irrelevant; 2: little relevant; 3: did not change behavior but could be helpful in a similar case; 4: partially relevant and 5: totally relevant

Results

Sixteen FGC were performed. The age ranged from 20 days to 20 years. The oncological diagnosis was: liver tumor in 3 patients, sacrococcygeal teratoma in 2, ovarian tumor in 3, and others. FGC was used as arteriography in 6 patients, visualization of liver tumor in 3, cholangiography in 3, cystography in 2, and lymphography in 2. Indocyanine green was injected intravenously in the arteriography, inside the cyst on cystography, one to two hours before the procedure for cholangiography, two to three days before the procedure for liver tumors, and in the testis for Palomo surgery. The surgeries performed were: three regulated hepatectomies, one non-regulated hepatectomy, two sacrococcygeal teratoma resections, two peri-aortic lymphadenectomies, two cholecystectomies, one salpingo-oophorectomy, a laparoscopic Palomo surgery, a thoracoscopic thoracic duct ligation, and a hepatic hilum lymphadenectomy. The procedures were classified using a Likert scale: one in category 2, three in category 3, eight in category 4, and three in category 5.

Conclusions

The FGC is a tool to assist in various procedures. Comparative studies may demonstrate the impact of this technology.

Background and Aims

Surgical control of pulmonary metastases, the most common site of metastasis for patients with osteosarcoma, is considered necessary for long term survival. Resection has been the mainstay of surgical management and can be performed as an open or thoracoscopic procedure. Radiologically detected lesions are sometimes difficult to identify peroperatively, especially when no palpation is possible during thoracoscopy. In the adult population, near-infrared fluorescence imaging using indocyanine green (ICG) has been shown to be a safe method for intra-operative visual identification of pulmonary metastases. Our aim is to determine the feasibility of using ICG for visual identification of pulmonary metastases in pediatric osteosarcoma patients.

Methods

Seven patients received an intravenous dose of ICG 24 hours preoperatively. To determine the most optimal dosage, a dose escalation scheme was followed consisting of three patients in every dosage group. We started with 0.5mg/kg, followed by 1.0mg/kg. Intra-operatively and post-operatively, the lesions were visualized with a near-infrared camera system that could be used in open and minimally invasive surgical setting. Fluorescence was quantified by tumor-to-background ratio (TBR) and surgeons evaluated the use of ICG for metastasis identification.

Results

Two (29%) patients underwent a thoracoscopy and five (71%) underwent a thoracotomy. Three (43%) patients had a fluorescent metastasis. In four (57%) patients there were no fluorescent metastases visible during surgery. In two cases the metastasis turned out to be necrotic, and in the other two visibility may have been hindered by the depth (>5mm) of the metastases. Four patients received an ICG dosage of 0.5mg/kg and three patients received 1.0mg/kg. 1.0mg/kg did not improve the TBR significantly. No adverse events occurred.

Conclusions

ICG for pulmonary osteosarcoma metastasectomy in the pediatric population is a feasible procedure, given 24 hours pre-operatively using a dose of 0.5mg/kg. It does not seem to be useful for deep lying and necrotic metastatic disease.

Background and Aims

Indocyanine green (ICG) administration is gaining popularity in pediatric surgery for a broad spectrum of procedures. ICG has also been proved to be useful in metastases detection, lymph node evaluation and tumor resection.

We aim to describe the results of the use of ICG administration as a potential tool to guide the surgical approach in our pediatric cancer patients.

Methods

A prospective non-randomized study was conducted including all pediatric patients with a solid tumor who required a surgical evaluation at our institution between March 2021 and March 2022. Inclusion criteria included the presence of a solid tumor requiring surgery for diagnosis or definitive treatment and absence of iodine contrast allergies. Depending on diagnosis and intended surgical procedure, ICG was administered: intravenously (1,5mg/kg) 24 hours prior surgery (group A), intraoperatively (group B) or subcutaneously (5mg) intraoperatively (group C). Parental consent was obtained before ICG administration. Near infrared camera (Storz IMAGE1 S™ RUBINA) was used to assess ICG uptake. Data regarding demographics, diagnosis, ICG uptake and histological reports were recorded.

Results

ICG was administered to 33 patients (15 males - 18 females) with a median age of 9 years old [range, 10 days – 16 years]. In group A (n=24), 9 patients showed ICG uptake: 2 neurogenic tumors primarily excised, 4 patients with lung metastases, 1 lymphoma undergoing biopsy, 1 undifferentiated sarcoma and 1 Wilms tumor. Group B (n= 7) included hepatic tumors (n=3), thyroid surgery for parathyroid evaluation (n=3) and ovarian sparing surgery (n=1). Patients included in group C (n=2, rhabdomyosarcoma) were assessed for sentinel lymph node biopsies, showing ICG uptake in both cases.

Conclusions

Our study showed the benefits of ICG administration in pediatric surgical oncology, specifically in lymph node biopsies, lung metastases and hepatic tumors. However, further investigations are needed to unify patient selection criteria and indications.

Background and Aims

18F-fluorodeoxyglucose (FDG) positron emission tomography (PET) scans are essential in the diagnosis and monitoring of pediatric cancer. Gamma probe (PET probe) localization of small FDG-avid lesions has been described in a few adult studies with only modest results. We sought to assess the utility and limitations of the PET probe and present the first use of the PET probe in pediatric patients.

Methods

The PET probe was first assessed in a PET scan calibration model (phantom) using FDG-dense spheres (lesions) situated in either saline or a dilute FDG-solution. Prior to PET probe survey, a PET scan was performed. In a clinical model, pediatric patients with one or more occult FDG-avid lesions underwent excisional biopsy utilizing IV FDG and PET probe survey. Prospective data collection of gamma counts from lesions and adjacent background tissue were recorded.

Results

Lesions in the phantom showed an SUV of 4 when the model contained background FDG. The PET probe was able to localize these lesions when the background solution was saline but not in dilute FDG. Similarly, in clinical use (nine operations in eight patients), the PET probe could not be used to guide dissection due to significant background detection. After lesion excision by standard methods, the PET probe did help confirm increased gamma counts ex vivo. In total, 17 lesions were excised. Adequate tissue for diagnosis was present in 16 (94%) of the specimens and positive for malignancy in 12 (71%). The average SUV on pre-op imaging was 8.4 and average tumor size was 2.0 cm.

Conclusions

Use of the PET probe was safe and feasible. The probe’s specificity for FDG, however, was poor in both the phantom and in clinical use. Tumor-targeted, radiolabeled agents are likely necessary to allow the PET probe to identify occult pediatric tumors in vivo with sufficient specificity.