Key Points
Question
Is surveillance a successful strategy for detecting new tumors in children and young adults with cancer predisposition syndromes (CPSs)?
Findings
In this cohort study of 274 children and young adults with 35 different CPSs, new tumors identified through surveillance were predominantly localized and amenable to resection, and most were successfully treated. Across all modalities, surveillance was accurate, with minimal false-positive (0.4%) and false-negative (0.3%) findings, and tumors rarely presented symptomatically (1.8%).
Meaning
These findings suggest that standardized surveillance enables early detection of asymptomatic, treatable tumors across a broad range of CPSs.
This cohort study evaluates surveillance outcomes across a wide spectrum of cancer predisposition syndromes in children and young adults.
Abstract
Importance
Pediatric oncology patients are increasingly recognized as having an underlying cancer predisposition syndrome (CPS). Surveillance is often recommended to detect new tumors at their earliest and most curable stages. Data on the effectiveness and outcomes of surveillance for children with CPS are limited.
Objective
To evaluate the performance of surveillance across a wide spectrum of CPSs.
Design, Setting, and Participants
This cohort study reviewed surveillance outcomes for children and young adults from birth to age 23 years with a clinical and/or molecular CPS diagnosis from January 1, 2009, through September 31, 2021. Patients were monitored using standard surveillance regimens for their corresponding CPS at a specialty pediatric oncology center. Patients with hereditary retinoblastoma and bone marrow failure syndromes were excluded. Data were analyzed between August 1, 2021, and December 6, 2023.
Exposure
Cancer predisposition syndrome.
Main Outcomes and Measures
Outcomes of surveillance were reviewed to evaluate the incidence, spectrum, and clinical course of newly detected tumors. Surveillance modalities were classified for accuracy and assessed for common strengths and weaknesses.
Results
A total of 274 children and young adults (mean age, 8 years [range, birth to 23 years]; 144 female [52.6%]) with 35 different CPSs were included, with a median follow-up of 3 years (range, 1 month to 12 years). During the study period, 35 asymptomatic tumors were detected in 27 patients through surveillance (9.9% of the cohort), while 5 symptomatic tumors were detected in 5 patients (1.8% of the cohort) outside of surveillance, 2 of whom also had tumors detected through surveillance. Ten of the 35 tumors (28.6%) were identified on first surveillance imaging. Malignant solid and brain tumors identified through surveillance were more often localized (20 of 24 [83.3%]) than similar tumors detected before CPS diagnosis (71 of 125 [56.8%]; P < .001). Of the 24 tumors identified through surveillance and surgically resected, 17 (70.8%) had completely negative margins. When analyzed across all imaging modalities, the sensitivity (96.4%), specificity (99.6%), positive predictive value (94.3%), and negative predictive value (99.6%) of surveillance were high, with few false-positive (6 [0.4%]) or false-negative (5 [0.3%]) findings.
Conclusions and Relevance
These findings suggest that standardized surveillance enables early detection of new tumors across a wide spectrum of CPSs, allowing for complete surgical resection and successful treatment in the majority of patients.
Introduction
Large-scale sequencing studies have revealed that 5% to 15% of children with cancer harbor an underlying cancer predisposition syndrome (CPS).1,2,3,4,5,6,7 With expanding application of germline sequencing, knowledge of syndrome-specific cancer phenotypes and age-specific cancer risks is increasing. This increase in knowledge presents opportunities to enhance overall outcomes for affected individuals with CPSs through implementation of cancer prevention and surveillance strategies.
Surveillance is the proactive evaluation of individuals at increased cancer risk with the goal of detecting new tumors at the earliest and most treatable stages.8 Surveillance may be considered if an increased risk for tumors is established, safe and accurate methods of tumor detection exist, and effective treatments for identified tumors are available.8,9 Care must be taken to select surveillance modalities that enable tumor detection while minimizing risks such as radiation exposure; repeated sedation; psychological burden; and undue economic impact on patients, families, and health care systems.10,11,12,13,14
Surveillance guidelines for CPSs exist but are rarely data driven. Instead, recommendations are often founded on expert consensus.8 Evidence indicates that adherence to standardized surveillance protocols may improve outcomes for children with particular syndromes. For example, Villani et al15 reported that the multimodal Toronto protocol successfully identified asymptomatic tumors in individuals with Li Fraumeni syndrome (LFS), resulting in an estimated 5-year overall survival rate of 88.8% for those under surveillance compared with 59.6% for those without surveillance. Similarly, Durno et al16 reported that following a surveillance protocol for individuals with constitutional mismatch repair deficiency (CMMRD) resulted in a 4-year overall survival rate of 79% for those undergoing full surveillance vs 15% for those with no surveillance. These studies underscore the success of surveillance in identifying new tumors promptly, leading to early intervention and improved outcomes for affected individuals with CPSs.
Despite these advances, data demonstrating the effectiveness and outcomes of surveillance for children with other CPSs are limited. Moreover, the performance of specific surveillance modalities has not been comprehensively evaluated. An improved understanding of the benefits and limitations of surveillance is imperative to optimize existing protocols. To address these gaps, we scrutinized the incidence and range of cancers identified in patients with CPSs at our institution using standardized surveillance regimens. Additionally, we evaluated the individual and overall performance of surveillance methods across a large cohort of children and young adults with CPS.
Methods
Study Design and Patient Population
In this cohort study, clinical and radiologic data were collected for children and young adults from birth to age 23 years with CPS undergoing surveillance at St Jude Children’s Research Hospital (SJCRH) between January 1, 2009, and September 31, 2021. The SJCRH institutional review board approved this study with a waiver of informed consent due to its retrospective design. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Patients were diagnosed with CPS using established clinical and/or molecular criteria. Patients with hereditary retinoblastoma or bone marrow failure syndromes were excluded, as surveillance was completed by other teams.
Evaluation of Patient Outcomes
Patients were monitored with a diverse range of surveillance tests administered at specific intervals and ages (eTable 1 in Supplement 1). Institutional standardized regimens were established in 2015, coinciding with the establishment of the Cancer Predisposition Program at SJCRH. Regimens are reviewed every 2 to 3 years and updated based on published literature. Prior to 2015, patients were followed up by their primary oncologists, adhering to the surveillance recommendations of that era.
Tumors identified during the study period were documented. Pathology reports were examined and information regarding the method of tumor detection recorded. We defined synchronous tumors as those identified simultaneously, while metachronous tumors were those identified at least 6 months apart. If an individual developed 2 or more tumors of the same histology at a single time point (eg, several synchronous basal cell carcinomas [BCCs]), these were counted as 1 new tumor. If an individual developed 2 or more tumors of different histology or at different time points, these were counted as separate tumors.
Tumor Staging
Following review of pathology and imaging reports, all solid and central nervous system (CNS) cancers and lymphomas were classified as localized (American Joint Committee on Cancer stage I) or advanced (stage II or higher).17 For patients with bilateral synchronous tumors (eg, bilateral Wilms tumor), the highest local stage was used for statistical analyses. The term not applicable was used to classify leukemias, as staging is not comparable.
Classification of Screening Radiographic and Procedure Reports for Accuracy
Reports from magnetic resonance imaging (MRI) (whole body, brain, spine, abdomen, and neck), ultrasonography (abdomen, kidneys, and pelvis), computed tomography (CT) (chest), radiography (chest, skeleton, and dental), echocardiography, esophagogastroduodenoscopy (EGD), and colonoscopy were reviewed. Overall, 1533 surveillance reports were classified as follows:
True positive (TP)—A new tumor was identified and confirmed.
True negative (TN)—No new tumor was identified, and subsequent follow-up did not contradict this result for 1 year.
False positive (FP)—An identified lesion was concerning for tumor, necessitating follow-up interventions, which did not reveal a tumor.
False negative (FN)—A lesion was observed and not felt to be concerning for a tumor, or no lesions were observed. Additional interventions identified a tumor, and in retrospect, it was missed on the categorized scan.
Indeterminant finding—A lesion concerning for a new tumor was detected, but there was insufficient follow-up to determine the final diagnosis.
Incidental finding—A nontumorous finding was detected (eg, renal cysts). Identification of incidental findings did not change the interpretation of the scan as TP, TN, FP, or FN.
Relapse or progression of a known cancer was not counted as a new tumor for this study.
Surveillance reports were initially classified and subsequently reviewed for agreement. For all FP and FN reports and for cases where discrepancies in classification arose, direct radiographic images were reviewed alongside an experienced radiologist. Imaging conducted as part of routine evaluations for existing or recently treated cancers was considered a surveillance scan if done within the relevant timeframe and encompassing the appropriate organs (eg, pelvic MRI for disease follow-up was counted as a surveillance pelvic ultrasonography). Dedicated imaging due to concerning findings on surveillance scans or patient signs and symptoms were not considered routine surveillance. These scans were used for clinical outcomes and to determine the correct classification of the original surveillance scan.
Statistical Analysis
The data analysis was performed from August 1, 2021, to December 6, 2023. Demographic data were collected for descriptive purposes and included age at CPS diagnosis, sex, CPS type, and cancer history prior to CPS diagnosis. Race and ethnicity data were not analyzed in this study. To compare the differences in stage for malignant tumors identified prior to CPS diagnosis vs those identified through surveillance post-CPS diagnosis, the McNemar test of agreement was conducted. The Cohen κ approach was used to estimate the level of agreement between these groups.
To assess the accuracy of surveillance, primary analysis was conducted at the patient level considering that a patient could undergo multiple surveillance tests, followed by a secondary report-based analysis. We computed 4 standard diagnostic metrics (sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]) (eTable 2 in Supplement 1). We used generalized estimating equations with a logit link function and a compound symmetry correlation structure, and each patient was treated as a cluster to account for within-patient correlations across surveillance reports in the same patient.18 We then calculated point estimates for sensitivity, specificity, PPV, and NPV along with robust sandwich standard error estimates. This patient-level approach ensured robust estimation of the uncertainty in the estimated performance metrics. However, when we further examined the metrics by individual surveillance modality, the generalized estimating equation algorithm failed to converge reliably due to reduced sample size. As a result, in the secondary analysis for each modality, we applied a report-based analysis that assumed that each report was generated independently. The incidence of FN and FP results was calculated for all surveillance reports. Statistical analyses were conducted between August 1, 2021, and December 6, 2023, using SAS, version 9.4 software (SAS Institute Inc). A 2-sided P < .05 was considered statistically significant by McNemar test.
Results
Patient Cohort
A total of 274 patients with 35 CPSs underwent surveillance for a median of 3 years (range, 1 month to 12 years) (Table 1). The mean age at CPS diagnosis was 8 years (range, birth to 23 years). The cohort included 144 females (52.6%) and 130 males (47.4%). A total of 139 patients (50.7%) had a prior tumor history, while 135 (49.3%) did not. The most common CPSs included LFS (49 patients [17.9%]), familial adenomatous polyposis (38 [13.9%]), neurofibromatosis type I (28 [10.2%]), and DICER1 syndrome (25 [9.1%]).
Table 1. Description of Patient Cohort.
| Characteristic | No. of patients (%) |
|---|---|
| Total | 274 |
| Sex | |
| Female | 144 (52.6) |
| Male | 130 (47.4) |
| Age at CPS diagnosis, mean (range), y | 8 (Birth-23) |
| CPSa | |
| Li Fraumeni syndrome | 49 (17.9) |
| Familial adenomatous polyposis | 38 (13.9) |
| Neurofibromatosis type 1 | 28 (10.2) |
| DICER1 syndrome | 25 (9.1) |
| Beckwith-Wiedemann spectrum | 20 (7.3) |
| Nevoid basal cell carcinoma (also known as Gorlin syndrome) | 19 (6.9) |
| Rhabdoid tumor predisposition | 15 (5.5) |
| von Hippel-Lindau syndrome | 12 (4.4) |
| Hereditary paraganglioma/pheochromocytoma | 9 (3.3) |
| PTEN hamartoma tumor syndrome | 8 (2.9) |
| Constitutional mismatch repair deficiency | 8 (2.9) |
| Hereditary leiomyomatosis renal cell carcinoma | 7 (2.6) |
| Ollier syndrome | 2 (0.7) |
| Peutz-Jeghers syndrome | 1 (0.4) |
| Otherb | 36 (13.1) |
| Cancer history prior to CPS diagnosis | |
| Nonec | 135 (49.3) |
| Solid tumor | 64 (23.3) |
| CNS tumor | 63 (23.0) |
| Leukemia/lymphoma | 12 (4.4) |
Abbreviations: CNS, central nervous system; CPS, cancer predisposition syndrome.
Five patients were diagnosed with 2 CPSs, and none of these patients developed a tumor during the study.
Other CPSs were diagnosed in fewer than 5 patients, with no patients developing a tumor during the study period: BAP1 (4 patients), juvenile polyposis syndrome (4 patients), LZTR1 (3 patients), multiple endocrine neoplasia 2 (3 patients), neurofibromatosis type 2 (3 patients), AMER1 (2 patients), ataxia telangiectasia (2 patients), Lynch syndrome (MLH1, 1 patient; PMS2, 1 patient), multiple endocrine neoplasia 1 (2 patients), AXIN2 (1 patient), BRCA2 (1 patient), CEBPA (1 patient), FLCN (1 patient), MITF (1 patient), MUTYH (1 patient), PALB2 (1 patient), RECQL4 (compound heterozygous, 1 patient), RUNX1 (1 patient), WT1 (1 patient), trisomy 18 (1 patient), and xeroderma pigmentosa (1 patient).
For the 135 patients without a cancer history, 74 (27%) were diagnosed through targeted familial testing following identification of a CPS in another family member, 50 (18%) presented with a known CPS that had been diagnosed at an outside institution, and 11 (4%) were diagnosed through testing coordinated by the St Jude Cancer Predisposition Program based on the presence of physical features concerning for the presence of an underlying CPS.
Spectrum of Identified Tumors and Incidental Findings
Throughout the study period, 35 asymptomatic tumors (24 malignant, 11 benign) were diagnosed in 27 patients (9.9% of the cohort) through surveillance (Figure 1). Nineteen (54.3%) of these tumors were diagnosed in children with a previously treated cancer, 6 (17.1%) were diagnosed in patients receiving cancer therapy, and 10 (28.6%) were diagnosed in patients with no cancer history. Conversely, 5 symptomatic tumors in 5 children (1.8% of cohort) were not detected through surveillance (1 in a child with a cancer history not receiving therapy, 3 in children actively being treated, and 1 in a child with no cancer history) (eTable 3 in Supplement 1). Two of these 5 children had an additional primary tumor detected through surveillance. Altogether, 40 tumors (24 solid, 13 CNS, 3 hematopoietic) were diagnosed in 30 patients (10.9% of the cohort), of which 35 (87.5%) were identified through surveillance.
Figure 1. Flow Diagram of Patients Undergoing Surveillance and the Number of Lesions Identified.
aThree additional patients were noted to have indeterminant lesions as well as confirmed tumors identified through surveillance. For simplicity and to avoid duplicate patient reporting, they are included with the 30 patients whose tumors were identified through surveillance and not with the patients noted to have indeterminant findings.
Twenty-five tumors (71.4%) identified through surveillance were detected on imaging, 6 (17.1%) on clinical examination, 2 (5.7%) by EGD and/or colonoscopy, and 2 (5.7%) by laboratory evaluation (Table 2). Notably, 10 tumors (28.6%) were detected on first surveillance evaluation and 23 (65.7%) within 2 years of initiating surveillance (Figure 2). Lesions detected through surveillance were primarily solid (21 [60.0%]) or CNS (13 [37.1%]) tumors, with only 1 (2.9%) hematopoietic cancer identified (Hodgkin lymphoma in 1 patient with LFS). Tumors detected outside of surveillance included 2 cases of therapy-associated acute myeloid leukemia and 1 case each of BCC, desmoid tumor, and Sertoli cell tumor. Incidental findings were recorded on 70 scans (4.6% of the 1533 reports evaluated), the most common being cysts and developmental differences (eTable 2 in Supplement 1).
Table 2. Patients Diagnosed With New Tumors Through Surveillance.
| Patient ID | Syndrome | Tumor | Surveillance method |
|---|---|---|---|
| 1 | BWSp | Hepatoblastoma | AFP, abdominal MRI |
| Adrenocortical carcinoma | Abdominal MRI | ||
| 2 | BWSp | Hepatoblastoma | AFP, abdominal MRI |
| 3 | CMMRD | Colorectal adenocarcinoma | Colonoscopy |
| 4 | CMMRD | Astrocytoma | Brain MRI |
| 5 | CMMRD | Melanoma | Skin examination |
| 6 | CMMRD | Glioblastoma | Brain MRI |
| 7 | CMMRD | Glioblastoma | Brain MRI |
| 8 | DICER1 | Pleuropulmonary blastoma, type I | Chest radiograph, CT |
| 9 | DICER1 | Thyroid carcinoma | Thyroid ultrasonography |
| Sertoli-Leydig cell tumor | Abdominal MRI | ||
| 10 | DICER1 | Pleuropulmonary blastoma, type Ir | Chest CT |
| Thyroid carcinoma | Thyroid ultrasonography | ||
| 11 | DICER1 | Thyroid carcinoma | Thyroid ultrasonography |
| 12 | DICER1 | Ciliary body medulloepithelioma | Eye examination |
| 13 | FAP | Colorectal adenocarcinoma | Colonoscopy |
| Thyroid carcinoma | Thyroid ultrasonography | ||
| 14 | Gorlin | Ovarian fibroma | Pelvic ultrasonography |
| 15a | Gorlin | Basal cell carcinomas | Skin examination |
| 16b | Gorlin | Basal cell carcinoma | Skin examination |
| Basal cell carcinoma | Skin examination | ||
| 17 | LFS | Anaplastic astrocytoma | Brain MRI |
| 18 | LFS | Glioneuronal tumor | Brain MRI |
| 19 | LFS | Hodgkin lymphoma | WBMRI |
| 20 | LFS | Ganglioneuroma | WBMRI |
| 21 | NF1 | Plexiform neurofibroma | WBMRI |
| 22 | NF1 | Glioma | WBMRI |
| 23 | NF2 | Acoustic neuroma | Brain MRI |
| Eyelid plexiform schwannoma | Eye examination | ||
| 24 | NF2 | Right vestibular schwannoma | Brain MRI |
| Left vestibular schwannoma | Brain MRI | ||
| 25a | Ollier | Enchondromas | WBMRI |
| 26 | VHL | Hemangioblastoma | Abdominal MRI |
| 27b | VHL | Hemangioblastoma | Brain MRI |
| Hemangioblastoma | Spine MRI |
Abbreviations: AFP, α-fetoprotein; BWSp, Beckwith-Wiedemann spectrum; CMMRD, constitutional mismatch repair deficiency; CT, computed tomography; FAP, familial adenomatous polyposis; LFS, Li Fraumeni syndrome; MRI, magnetic resonance imaging; WBMRI, whole-body magnetic resonance imaging; NF1, neurofibromatosis type 1; NF2; neurofibromatosis type 2; VHL, von Hippel-Lindau syndrome.
Multiple (≥2) tumors of the same histology were identified at a single time point but counted as 1 tumor for this study.
Multiple (≥2) basal cell carcinomas or hemangioblastomas were identified at different time points and in different anatomic locations in patients 16 and 27, respectively. As these tumor types do not generally metastasize, they were considered separate primary tumors for the purpose of this study.
Figure 2. Timing of Tumor Detection Through Surveillance.
The swimmers plot begins with the start of surveillance and continues until the time of last follow-up or end of study, whichever occurred later. If surveillance raised concern for a new tumor but additional follow-up was needed to confirm the diagnosis, the period between the first concerning surveillance result and final diagnosis is denoted with a black arrow. SJCRH indicates St Jude Children’s Research Hospital.
Stage of Identified Malignant Tumors and Surgical Outcomes
The objective of surveillance is to detect tumors at a localized stage when they are most amenable to complete surgical resection. Thus, we performed an internal cohort comparison of the stages of malignant CNS and solid tumors identified through surveillance vs similar tumors diagnosed prior to a patient’s CPS diagnosis. Malignant tumors detected via surveillance were significantly more likely to be localized (ie, stage I, 20 of 24 [83.3%]) compared with those identified before the CPS diagnosis (71 of 125 [56.8%]; P < .001).
We also analyzed the surgical outcomes for tumors detected through surveillance when resection was a treatment option (eTable 4 in Supplement 1). Among 24 tumors treated with surgery, 17 (70.8%) were completely resected with negative margins, 6 (25.0%) had residual microscopic disease, and 1 (4.2%) had distant metastatic disease. Surgery was the only required intervention for 17 tumors (48.6%) diagnosed through surveillance.
Performance of Surveillance Across Imaging and Other Modalities
Evaluation of the overall accuracy of radiologic imaging and diagnostic surveillance procedures using a patient-level approach revealed a high sensitivity (96.4%), specificity (99.6%), PPV (94.3%), and NPV (99.6%) (eTable 2 in Supplement 1). Using a report-based approach for individual modalities, we found that spinal MRI, EGD, and colonoscopy each showed perfect metrics with 100% sensitivity, specificity, PPV, and NPV. In contrast, whole-body MRI maintained excellent sensitivity (100%), specificity (97.7%), and NPV (100%) but showed lower PPV (62.5%) due to 3 FP findings. Abdominal ultrasonography (618 scans) also maintained superb specificity (100%), PPV (100%), and NPV (99.4%), although it had low sensitivity (20%) due to 4 FN findings.
FP, FN, and Indeterminant Findings
Despite the high accuracy of surveillance, instances were observed where results were incorrect or inconclusive (eFigure in Supplement 1). For example, 6 of 1533 (0.4%) surveillance tests in 6 patients (2.2%) revealed lesions concerning for tumor, but on follow-up imaging or laboratory evaluation, the findings were deemed FP (mean number of follow-up scans, 1.7; range, 1-4) (eTable 5 in Supplement 1). No surgical interventions or treatment changes occurred due to an FP report. Similarly, 5 of 1533 (0.3%) surveillance tests in 3 patients (1.1%) revealed FN findings (eTable 6 in Supplement 1). Ten patients (3.6%) underwent imaging that revealed lesions classified as indeterminant and are being monitored closely (eTable 7 in Supplement 1).
Discussion
The increasing integration of germline genetic testing into routine treatment of children with cancer is leading to the recognition of more children with an underlying CPS. Among these individuals exists a subset who faces an augmented risk of developing additional neoplasms. Furthermore, cascade testing of relatives unveils family members who are genetically at increased cancer risk.19,20 Consequently, these individuals may stand to gain from vigilant surveillance aimed at detecting early-stage, potentially curable tumors. However, existing data concerning the effectiveness of surveillance for children with CPS remain restricted, often concentrating on LFS and CMMRD.15,16 To gain deeper insights, we conducted a comprehensive study evaluating the outcomes of surveillance across 274 pediatric patients with CPS in, to our knowledge, one of the largest cohorts to date.
We found that 30 patients (10.9%) developed 1 or more new tumors, with most (87.5%) discovered through surveillance. Notably, 10 new tumors (28.6%) were detected on first surveillance imaging, 23 (65.7%) within the first 2 years of observation, and 6 (17.1%) while patients were still undergoing therapy for a prior cancer. Importantly, tumors identified through surveillance were often localized, and the majority were resected with negative microscopic margins. Only rare FP results arose through surveillance, and no patient required invasive interventions to determine the final etiology of the FP lesions. These data indicate the utility of surveillance; the importance of initiating monitoring as soon as a CPS diagnosis is established, even if it is during treatment for an existing tumor; and the need to coordinate imaging with primary oncologists to reduce exposure to anesthesia and minimize costs from duplicate scans.
All tumors identified through surveillance fell within the phenotypes of the corresponding CPSs, including brain tumors and colon cancer in patients with CMMRD, hepatoblastoma in Beckwith-Wiedemann spectrum, and thyroid cancer in DICER1 syndrome.21,22,23,24,25 Consistent with the high penetrance of CMMRD and LFS, new tumors were among the most prevalent in children with these conditions. As previously reported, asymptomatic low-grade lesions were identified in patients with CMMRD and LFS, underscoring the potential of early detection and resection to reduce the risk of malignant transformation.15,16 Finally, 7 tumors were identified in 5 patients with DICER1 syndrome, all with a history of prior cancer treated with chemotherapy and/or radiation.26,27 This intricate interplay between host genetic factors and exposure to genotoxic agents highlights the necessity of surveillance for this subgroup.
Overall, most imaging modalities showed excellent performance. Whole-body MRI is increasingly integrated into surveillance strategies for conditions including CMMRD, LFS, and neurofibromatosis type 1, among others.28 In this study, whole-body MRI had excellent sensitivity, specificity, and NPV. However, its PPV was low at 62.5%. These findings closely parallel those of Anupindi et al,29 who reported that whole-body MRI in 24 pediatric patients with CPS had 100% sensitivity, 94% specificity, 100% NPV, and 25% PPV. Their lower PPV stemmed from 3 FP bone marrow signal intensities in 3 patients with LFS. This finding is akin to patient 45 in our study, potentially indicating a specific vulnerability of whole-body MRI.
Abdominal ultrasonography had the most FN findings (4 of 618 scans), resulting in low sensitivity. The overlooked findings were small liver lesions and occurred in patients with Beckwith-Wiedemann spectrum with elevated serum α-fetoprotein levels. In these cases, dedicated MRI identified concerning lesions. While MRI boasts heightened sensitivity, the potential risks of recurrent sedation in young patients prompt a cautious approach. Hence, MRI should be reserved for patients with elevated α-fetoprotein levels and negative ultrasonography findings.30,31
Five patients were diagnosed with tumors based on signs and symptoms and not through surveillance tests. For 2 patients, the intervening malignant neoplasm was therapy-associated acute myeloid leukemia, which is notoriously difficult to detect using existing surveillance protocols.32 The other 3 patients developed tumors for which surveillance is not routinely recommended, including Sertoli cell tumor in a patient with Peutz-Jeghers syndrome and BCC in a patient with CMMRD.33,34,35 One patient with familial adenomatous polyposis developed a symptomatic desmoid tumor. Imaging for desmoid tumors generally is recommended only prior to undergoing colectomy.36,37,38 Recognizing the existing gaps inherent to surveillance, it is imperative that clinicians educate patients and families about the signs and symptoms of tumors so that medical attention can be promptly sought for any concerning findings.
Limitations
Despite this study’s positive findings, several limitations warrant consideration. First, this study was not randomized with a control arm in which patients did not undergo surveillance. However, given the rarity of childhood CPS and the increasing acceptance in support of surveillance, it was not deemed feasible or ethical to conduct this study in a randomized fashion. Second, due to the high volume of investigations conducted, initial analysis predominantly relied on reviewing reports rather than the images themselves. Third, data collection spanned a median of 3 years. Had patients been followed up longer, additional lesions may have been detected. Fourth, the examined cohort is heterogeneous with small sample sizes for certain conditions, precluding in-depth statistical evaluation of syndrome-specific surveillance strategies. Nonetheless, report-based point estimates of performance metrics remained unbiased, although confidence intervals may appear narrower than anticipated due to the assumption that reports were generated independently for the same patient.18 Fifth, this study did not examine the psychological, emotional, or economic outcomes of surveillance.39,40 While the cost-effectiveness of surveillance has been explored for patients with LFS, further efforts are needed for other CPSs.34,41 Finally, as SJCRH does not depend on insurance coverage for surveillance care and expenses, it is difficult to assume generalizability to other patient populations and health care systems.
Conclusions
This extensive cohort study underscores that standardized surveillance successfully uncovers new asymptomatic tumors in children and young adults across a broad range of CPSs. Notably, a substantial proportion of tumors were detected within a short evaluation period, and many were localized and effectively treated. These findings provide a solid foundation for future endeavors aimed at refining existing surveillance protocols. Novel radioisotopes, proteomic approaches, and circulating tumor DNA are modern tools being investigated for future surveillance alternatives.42,43,44,45 By partnering with clinicians who come across individuals with CPS, our center and others with similar cancer predisposition programs can provide guidance about when and which patients to refer for a cancer genetics evaluation. For individuals diagnosed with an underlying CPS, we can educate about the associated cancer risks and recommended surveillance protocols. Finally, by working collaboratively, we stand to gain further insights into the natural progression of childhood CPS, the association between germline genotypes and tumor phenotypes, and the overall impact of surveillance for affected children and their families.
eTable 1. Cancer Surveillance Regimens for Children With CPS in This Study
eTable 2. Statistical Performance of Surveillance
eTable 3. Patients Diagnosed With New Tumors
eTable 4. Outcomes of Patients With Tumors Diagnosed Through Surveillance
eTable 5. Patients With False-Positive Results
eTable 6. Patients With False-Negative Results
eTable 7. Patients With Indeterminant Lesions
eFigure. Examples of False-Positive (FP), False-Negative (FN), and Indeterminant Findings
Data Sharing Statement
References
- 1.Fiala EM, Jayakumaran G, Mauguen A, et al. Prospective pan-cancer germline testing using MSK-IMPACT informs clinical translation in 751 patients with pediatric solid tumors. Nat Cancer. 2021;2:357-365. doi: 10.1038/s43018-021-00172-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Mody RJ, Wu YM, Lonigro RJ, et al. Integrative clinical sequencing in the management of refractory or relapsed cancer in youth. JAMA. 2015;314(9):913-925. doi: 10.1001/jama.2015.10080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Newman S, Nakitandwe J, Kesserwan CA, et al. Genomes for kids: the scope of pathogenic mutations in pediatric cancer revealed by comprehensive DNA and RNA sequencing. Cancer Discov. 2021;11(12):3008-3027. doi: 10.1158/2159-8290.CD-20-1631 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Parsons DW, Roy A, Yang Y, et al. Diagnostic yield of clinical tumor and germline whole-exome sequencing for children with solid tumors. JAMA Oncol. 2016;2(5):616-624. doi: 10.1001/jamaoncol.2015.5699 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhang J, Walsh MF, Wu G, et al. Germline mutations in predisposition genes in pediatric cancer. N Engl J Med. 2015;373(24):2336-2346. doi: 10.1056/NEJMoa1508054 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chang W, Brohl AS, Patidar R, et al. Multidimensional clinomics for precision therapy of children and adolescent young adults with relapsed and refractory cancer: a report from the Center for Cancer Research. Clin Cancer Res. 2016;22(15):3810-3820. doi: 10.1158/1078-0432.CCR-15-2717 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Everett JN, Mody RJ, Stoffel EM, Chinnaiyan AM. Incorporating genetic counseling into clinical care for children and adolescents with cancer. Future Oncol. 2016;12(7):883-886. doi: 10.2217/fon-2015-0022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Malkin D, Nichols KE, Schiffman JD, Plon SE, Brodeur GM. The future of surveillance in the context of cancer predisposition: through the murky looking glass. Clin Cancer Res. 2017;23(21):e133-e137. doi: 10.1158/1078-0432.CCR-17-2026 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Teplick A, Kowalski M, Biegel JA, Nichols KE. Educational paper: screening in cancer predisposition syndromes: guidelines for the general pediatrician. Eur J Pediatr. 2011;170(3):285-294. doi: 10.1007/s00431-010-1377-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Tobias JD. Tolerance, withdrawal, and physical dependency after long-term sedation and analgesia of children in the pediatric intensive care unit. Crit Care Med. 2000;28(6):2122-2132. doi: 10.1097/00003246-200006000-00079 [DOI] [PubMed] [Google Scholar]
- 11.Weber E, Shuman C, Wasserman JD, et al. “A change in perspective”: exploring the experiences of adolescents with hereditary tumor predisposition. Pediatr Blood Cancer. 2019;66(1):e27445. doi: 10.1002/pbc.27445 [DOI] [PubMed] [Google Scholar]
- 12.Ross LF, Moon MR. Ethical issues in genetic testing of children. Arch Pediatr Adolesc Med. 2000;154(9):873-879. doi: 10.1001/archpedi.154.9.873 [DOI] [PubMed] [Google Scholar]
- 13.Dixon RB, Hertelendy AJ. Interrelation of preventive care benefits and shared costs under the Affordable Care Act (ACA). Int J Health Policy Manag. 2014;3(3):145-148. doi: 10.15171/ijhpm.2014.76 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Halpern MT, Romaire MA, Haber SG, Tangka FK, Sabatino SA, Howard DH. Impact of state-specific Medicaid reimbursement and eligibility policies on receipt of cancer screening. Cancer. 2014;120(19):3016-3024. doi: 10.1002/cncr.28704 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Villani A, Shore A, Wasserman JD, et al. Biochemical and imaging surveillance in germline TP53 mutation carriers with Li-Fraumeni syndrome: 11 year follow-up of a prospective observational study. Lancet Oncol. 2016;17(9):1295-1305. doi: 10.1016/S1470-2045(16)30249-2 [DOI] [PubMed] [Google Scholar]
- 16.Durno C, Ercan AB, Bianchi V, et al. Survival benefit for individuals with constitutional mismatch repair deficiency undergoing surveillance. J Clin Oncol. 2021;39(25):2779-2790. doi: 10.1200/JCO.20.02636 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Amin MB, Edge SB, Greene FL, et al. , eds. AJCC Cancer Staging Manual. 8th ed. Springer; 2017. [Google Scholar]
- 18.Zeger SL, Liang KY. Longitudinal data analysis for discrete and continuous outcomes. Biometrics. 1986;42(1):121-130. doi: 10.2307/2531248 [DOI] [PubMed] [Google Scholar]
- 19.Katz SJ, Abrahamse P, Hodan R, et al. Cascade genetic risk education and testing in families with hereditary cancer syndromes: a pilot study. JCO Oncol Pract. 2023;19(6):e848-e858. doi: 10.1200/OP.22.00677 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Garcia C, Sullivan MW, Lothamer H, et al. Mechanisms to increase cascade testing in hereditary breast and ovarian cancer: impact of introducing standardized communication aids into genetic counseling. J Obstet Gynaecol Res. 2020;46(9):1835-1841. doi: 10.1111/jog.14366 [DOI] [PubMed] [Google Scholar]
- 21.Brodeur GM, Nichols KE, Plon SE, Schiffman JD, Malkin D. Pediatric cancer predisposition and surveillance: an overview, and a tribute to Alfred G. Knudson Jr. Clin Cancer Res. 2017;23(11):e1-e5. doi: 10.1158/1078-0432.CCR-17-0702 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Schultz KAP, Williams GM, Kamihara J, et al. DICER1 and associated conditions: identification of at-risk individuals and recommended surveillance strategies. Clin Cancer Res. 2018;24(10):2251-2261. doi: 10.1158/1078-0432.CCR-17-3089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bakry D, Aronson M, Durno C, et al. Genetic and clinical determinants of constitutional mismatch repair deficiency syndrome: report from the constitutional mismatch repair deficiency consortium. Eur J Cancer. 2014;50(5):987-996. doi: 10.1016/j.ejca.2013.12.005 [DOI] [PubMed] [Google Scholar]
- 24.Durno CA, Sherman PM, Aronson M, et al. ; International BMMRD Consortium . Phenotypic and genotypic characterisation of biallelic mismatch repair deficiency (BMMR-D) syndrome. Eur J Cancer. 2015;51(8):977-983. doi: 10.1016/j.ejca.2015.02.008 [DOI] [PubMed] [Google Scholar]
- 25.Sobel Naveh NS, Traxler EM, Duffy KA, Kalish JM. Molecular networks of hepatoblastoma predisposition and oncogenesis in Beckwith-Wiedemann syndrome. Hepatol Commun. 2022;6(8):2132-2146. doi: 10.1002/hep4.1972 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Oue T, Inoue M, Kubota A, Kuwae Y, Kawa K. Pediatric thyroid cancer arising after treatment for pleuropulmonary blastoma. Pediatr Blood Cancer. 2008;50(4):901-902. doi: 10.1002/pbc.21265 [DOI] [PubMed] [Google Scholar]
- 27.de Kock L, Sabbaghian N, Soglio DBD, et al. Exploring the association between DICER1 mutations and differentiated thyroid carcinoma. J Clin Endocrinol Metab. 2014;99(6):E1072-E1077. doi: 10.1210/jc.2013-4206 [DOI] [PubMed] [Google Scholar]
- 28.Greer MC, Voss SD, States LJ. Pediatric cancer predisposition imaging: focus on whole-body MRI. Clin Cancer Res. 2017;23(11):e6-e13. doi: 10.1158/1078-0432.CCR-17-0515 [DOI] [PubMed] [Google Scholar]
- 29.Anupindi SA, Bedoya MA, Lindell RB, et al. Diagnostic performance of whole-body MRI as a tool for cancer screening in children with genetic cancer-predisposing conditions. AJR Am J Roentgenol. 2015;205(2):400-408. doi: 10.2214/AJR.14.13663 [DOI] [PubMed] [Google Scholar]
- 30.Brioude F, Kalish JM, Mussa A, et al. Expert consensus document: clinical and molecular diagnosis, screening and management of Beckwith-Wiedemann syndrome: an international consensus statement. Nat Rev Endocrinol. 2018;14(4):229-249. doi: 10.1038/nrendo.2017.166 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Kalish JM, Doros L, Helman LJ, et al. Surveillance recommendations for children with overgrowth syndromes and predisposition to Wilms tumors and hepatoblastoma. Clin Cancer Res. 2017;23(13):e115-e122. doi: 10.1158/1078-0432.CCR-17-0710 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Porter CC, Druley TE, Erez A, et al. Recommendations for surveillance for children with leukemia-predisposing conditions. Clin Cancer Res. 2017;23(11):e14-e22. doi: 10.1158/1078-0432.CCR-17-0428 [DOI] [PubMed] [Google Scholar]
- 33.Tabori U, Hansford JR, Achatz MI, et al. Clinical management and tumor surveillance recommendations of inherited mismatch repair deficiency in childhood. Clin Cancer Res. 2017;23(11):e32-e37. doi: 10.1158/1078-0432.CCR-17-0574 [DOI] [PubMed] [Google Scholar]
- 34.Achatz MI, Porter CC, Brugières L, et al. Cancer screening recommendations and clinical management of inherited gastrointestinal cancer syndromes in childhood. Clin Cancer Res. 2017;23(13):e107-e114. doi: 10.1158/1078-0432.CCR-17-0790 [DOI] [PubMed] [Google Scholar]
- 35.Lefevre H, Bouvattier C, Lahlou N, Adamsbaum C, Bougnères P, Carel JC. Prepubertal gynecomastia in Peutz-Jeghers syndrome: incomplete penetrance in a familial case and management with an aromatase inhibitor. Eur J Endocrinol. 2006;154(2):221-227. doi: 10.1530/eje.1.02085 [DOI] [PubMed] [Google Scholar]
- 36.Syngal S, Brand RE, Church JM, Giardiello FM, Hampel HL, Burt RW; American College of Gastroenterology . ACG clinical guideline: genetic testing and management of hereditary gastrointestinal cancer syndromes. Am J Gastroenterol. 2015;110(2):223-262. doi: 10.1038/ajg.2014.435 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Nieuwenhuis MH, Mathus-Vliegen EM, Baeten CG, et al. Evaluation of management of desmoid tumours associated with familial adenomatous polyposis in Dutch patients. Br J Cancer. 2011;104(1):37-42. doi: 10.1038/sj.bjc.6605997 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sinha A, Tekkis PP, Gibbons DC, Phillips RK, Clark SK. Risk factors predicting desmoid occurrence in patients with familial adenomatous polyposis: a meta-analysis. Colorectal Dis. 2011;13(11):1222-1229. doi: 10.1111/j.1463-1318.2010.02345.x [DOI] [PubMed] [Google Scholar]
- 39.Alderfer MA, Lindell RB, Viadro CI, et al. Should genetic testing be offered for children? the perspectives of adolescents and emerging adults in families with Li-Fraumeni Syndrome. J Genet Couns. 2017;26(5):1106-1115. doi: 10.1007/s10897-017-0091-x [DOI] [PubMed] [Google Scholar]
- 40.Bui KT, Liang R, Kiely BE, Brown C, Dhillon HM, Blinman P. Scanxiety: a scoping review about scan-associated anxiety. BMJ Open. 2021;11(5):e043215. doi: 10.1136/bmjopen-2020-043215 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Frankenthal IA, Alves MC, Tak C, Achatz MI. Cancer surveillance for patients with Li-Fraumeni syndrome in Brazil: a cost-effectiveness analysis. Lancet Reg Health Am. 2022;12:100265. doi: 10.1016/j.lana.2022.100265 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Bag AK, Wing MN, Sabin ND, et al. 11C-methionine PET for identification of pediatric high-grade glioma recurrence. J Nucl Med. 2022;63(5):664-671. doi: 10.2967/jnumed.120.261891 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Szymanski JJ, Sundby RT, Jones PA, et al. Cell-free DNA ultra-low-pass whole genome sequencing to distinguish malignant peripheral nerve sheath tumor (MPNST) from its benign precursor lesion: a cross-sectional study. PLoS Med. 2021;18(8):e1003734. doi: 10.1371/journal.pmed.1003734 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Hackshaw A, Cohen SS, Reichert H, Kansal AR, Chung KC, Ofman JJ. Estimating the population health impact of a multi-cancer early detection genomic blood test to complement existing screening in the US and UK. Br J Cancer. 2021;125(10):1432-1442. doi: 10.1038/s41416-021-01498-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Phallen J, Sausen M, Adleff V, et al. Direct detection of early-stage cancers using circulating tumor DNA. Sci Transl Med. 2017;9(403):eaan2415. doi: 10.1126/scitranslmed.aan2415 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Cancer Surveillance Regimens for Children With CPS in This Study
eTable 2. Statistical Performance of Surveillance
eTable 3. Patients Diagnosed With New Tumors
eTable 4. Outcomes of Patients With Tumors Diagnosed Through Surveillance
eTable 5. Patients With False-Positive Results
eTable 6. Patients With False-Negative Results
eTable 7. Patients With Indeterminant Lesions
eFigure. Examples of False-Positive (FP), False-Negative (FN), and Indeterminant Findings
Data Sharing Statement
