A Question-and-Answer System to Extract Data From Free-Text Oncological Pathology Reports (CancerBERT Network): Development Study.

BACKGROUND: Information in pathology reports is critical for cancer care. Natural language processing (NLP) systems used to extract information from pathology reports are often narrow in scope or require extensive tuning. Consequently, there is growing interest in automated deep learning approaches. A powerful new NLP algorithm, bidirectional encoder representations from transformers (BERT), was published in late 2018. BERT set new performance standards on tasks as diverse as question answering, named entity recognition, speech recognition, and more. OBJECTIVE: The aim of this study is to develop a BERT-based system to automatically extract detailed tumor site and histology information from free-text oncological pathology reports. METHODS: We pursued three specific aims: extract accurate tumor site and histology descriptions from free-text pathology reports, accommodate the diverse terminology used to indicate the same pathology, and provide accurate standardized tumor site and histology codes for use by downstream applications. We first trained a base language model to comprehend the technical language in pathology reports. This involved unsupervised learning on a training corpus of 275,605 electronic pathology reports from 164,531 unique patients that included 121 million words. Next, we trained a question-and-answer (Q&A) model that connects a Q&A layer to the base pathology language model to answer pathology questions. Our Q&A system was designed to search for the answers to two predefined questions in each pathology report: What organ contains the tumor? and What is the kind of tumor or carcinoma? This involved supervised training on 8197 pathology reports, each with ground truth answers to these 2 questions determined by certified tumor registrars. The data set included 214 tumor sites and 193 histologies. The tumor site and histology phrases extracted by the Q&A model were used to predict International Classification of Diseases for Oncology, Third Edition (ICD-O-3), site and histology codes. This involved fine-tuning two additional BERT models: one to predict site codes and another to predict histology codes. Our final system includes a network of 3 BERT-based models. We call this CancerBERT network (caBERTnet). We evaluated caBERTnet using a sequestered test data set of 2050 pathology reports with ground truth answers determined by certified tumor registrars. RESULTS: caBERTnet's accuracies for predicting group-level site and histology codes were 93.53% (1895/2026) and 97.6% (1993/2042), respectively. The top 5 accuracies for predicting fine-grained ICD-O-3 site and histology codes with 5 or more samples each in the training data set were 92.95% (1794/1930) and 96.01% (1853/1930), respectively. CONCLUSIONS: We have developed an NLP system that outperforms existing algorithms at predicting ICD-O-3 codes across an extensive range of tumor sites and histologies. Our new system could help reduce treatment delays, increase enrollment in clinical trials of new therapies, and improve patient outcomes.

Vet-ICD-O-Canine-1, a System for Coding Canine Neoplasms Based on the Human ICD-O-3.2.

Cancer registries are fundamental tools for collecting epidemiological cancer data and developing cancer prevention and control strategies. While cancer registration is common in the human medical field, many attempts to develop animal cancer registries have been launched over time, but most have been discontinued. A pivotal aspect of cancer registration is the availability of cancer coding systems, as provided by the International Classification of Diseases for Oncology (ICD-O). Within the Global Initiative for Veterinary Cancer Surveillance (GIVCS), established to foster and coordinate animal cancer registration worldwide, a group of veterinary pathologists and epidemiologists developed a comparative coding system for canine neoplasms. Vet-ICD-O-canine-1 is compatible with the human ICD-O-3.2 and is consistent with the currently recognized classification schemes for canine tumors. It comprises 335 topography codes and 534 morphology codes. The same code as in ICD-O-3.2 was used for the majority of canine tumors showing a high level of similarity to their human counterparts (n = 408). De novo codes (n = 152) were created for specific canine tumor entities (n = 126) and topographic sites (n = 26). The Vet-ICD-O-canine-1 coding system represents a user-friendly, easily accessible, and comprehensive resource for developing a canine cancer registration system that will enable studies within the One Health space.

Pancreatic Carcinosarcoma Clinical Outcome Analysis of the National Cancer Institute Database.

BACKGROUND: Pancreatic carcinosarcomas (PCS) are rare aggressive biphasic malignancies with a poor prognosis. We aimed to improve the understanding of PCS by analyzing variables that influence the mortality of PCS patients. METHODS: The Surveillance, Epidemiology, and End Results database was queried for cases of PCS from 1973 to 2016. Cases were analyzed for patient demographics, tumor characteristics, and surgical intervention. Kaplan-Meier and Cox regression analyses were applied to investigate the overall survival (OS) and prognostic factors. RESULTS: Thirty-nine cases of PCS were identified along with the disease demographics and characteristics. The majority of patients had a regionally invasive or metastatic disease. There was a significant decrease in OS with the increase of the tumor extension. Conversely, surgery showed to improve OS in the crude analysis, including patients that underwent lymphadenectomy. In addition, the unadjusted Cox regression results showed decreased hazard ratios with a local disease versus distant metastasis and with cancer-directed surgery versus no surgery. Nevertheless, the adjusted Cox regression results revealed that metastatic disease was the only significant predictor of survival. CONCLUSIONS: This population-based study provides some insight to a very rare disease by analyzing 39 cases of PCS. Our finding suggests considering PCS as a nonsurgical disease and reserving surgery solely for patients with a localized disease in combination or after neoadjuvant therapy. Consequently, there is a need to further investigate novel therapies for this aggressive malignancy.

A comprehensive epidemiological review of spinal astrocytomas in the United States.

OBJECTIVE: Spinal cord astrocytoma (SCA) is a rare tumor whose epidemiology has not been well defined. The authors utilized the Central Brain Tumor Registry of the United States (CBTRUS) to provide comprehensive up-to-date epidemiological data for this disease. METHODS: The CBTRUS was queried for SCAs on ICD-O-3 (International Classification of Diseases for Oncology, 3rd edition) histological and topographical codes. The age-adjusted incidence (AAI) per 100,000 persons was calculated and stratified by race, sex, age, and ethnicity. Joinpoint was used to calculate the annual percentage change (APC) in incidence. RESULTS: Two thousand nine hundred sixty-nine SCAs were diagnosed in the US between 1995 and 2016, resulting in an average of approximately 136 SCAs annually. The overall AAI was 0.047 (95% CI 0.045-0.049), and there was a statistically significant increase from 0.051 in 1995 to 0.043 in 2016. The peak incidence of 0.064 (95% CI 0.060-0.067) was found in the 0- to 19-year age group. The incidence in males was 0.053 (95% CI 0.050-0.055), which was significantly greater than the incidence in females (0.041, 95% CI 0.039-0.044). SCA incidence was significantly lower both in patients of Asian/Pacific Islander race (AAI = 0.034, 95% CI 0.028-0.042, p = 0.00015) and in patients of Hispanic ethnicity (AAI = 0.035, 95% CI 0.031-0.039, p < 0.001). The incidence of WHO grade I SCAs was significantly higher than those of WHO grade II, III, or IV SCAs (p < 0.001). CONCLUSIONS: The overall AAI of SCA from 1995 to 2016 was 0.047 per 100,000. The incidence peaked early in life for both sexes, reached a nadir between 20 and 34 years of age for males and between 35 and 44 years of age for females, and then slowly increased throughout adulthood, with a greater incidence in males. Pilocytic astrocytomas were the most common SCA in the study cohort. This study presents the most comprehensive epidemiological study of SCA incidence in the US to date.

How cancer registries can detect neoplasms in pathology laboratories that code with SNOMED CT terminology? An actual, simple and flexible solution.

BACKGROUND: Pathology laboratories are one of the main information sources for cancer registries and have traditionally been coded with SNOMED; some of them are migrating to SNOMED CT (SCT). Cancer registries encode topography and morphology of neoplasms by the International Classification of Diseases for Oncology (ICD-O). ICD-O updates morphology with WHO Classification of Tumors (Blue-Books). Morphological codes of the ICD-O, Blue-Books and SNOMED (former SNOMEDID) have always coincided. In 2017, SCT removed the SNOMEDID. OBJECTIVES: to define neoplastic and topographic subsets in SCT and map them to ICD-O-3.1/Blue-Books; reduce the original number of SCT concepts; correctly identify neoplasms in the laboratories in accordance with international cancer registry rules. METHODOLOGY: SCT neoplastic concepts were identified by manual revision and SCT resources ("is a", "Associated morphology" relationships; Simple Map Reference Set). Topographic concepts were extracted from the body structure hierarchy of SCT. Both subsets were mapped to ICD-O-3.1/Blue-Books, afterwards. Updating algorithms were designed to automate and update each subset with every SCT release. The process of neoplasms identification was validated in a sample of 5212 specimens with 7378 records from 8 Catalan hospitals. RESULTS: The number of concepts in neoplastic and topographic subsets (16,448 and 32,278) was reduced after the mapping to ICD-O-3.1/Blue-Books (2115 and 330, respectively). Neoplastic subset classified the specimens correctly in the 98.6% of the specimens. CONCLUSIONS: This article presents a flexible tool to exhaustively identify neoplasms in pathology laboratories that code with SCT, following international PBCRs standards and in line with the pathologists, oncologists and epidemiologists' needs.

Tumours in European Shorthair cats: a retrospective study of 680 cases.

OBJECTIVES: Naturally occurring tumours in domestic cats are less common than in dogs and represent the leading cause of death among older animals. The main objective of this study was to analyse a large data set of histologically diagnosed tumours to highlight the most common World Health Organization (WHO) tumour histotypes, the effect of age and sex, and the International Classification of Diseases for Oncology (ICD-O) topographical site predilections of feline breed-specific tumours. METHODS: A total of 680 feline tumours diagnosed in European Shorthair cats by three veterinary diagnostic laboratories located in central Italy from 2013 to 2019 were collected. Data on age, sex and topography of lesions were recorded. Samples were morphologically and topographically coded using the WHO and the ICD-O-3 classification system. RESULTS: Skin and soft tissue neoplasms comprised 55.9% of all tumours, followed by mammary gland (11%), alimentary tract (7.9%), oral cavity and tongue (7.3%), nasal cavity and middle ear (6%), lymph node (3.1%), bone (1.8%) and liver/intrahepatic bile duct (1.3%) tumours. Squamous cell carcinoma (SCC), sarcoma, lymphoma and basal cell tumours were the most diagnosed neoplasms. Malignant tumours were 82.9% of the total and the topographical sites mainly involved were skin (C44), connective/subcutaneous/other soft tissues (C49), mammary gland (C50), small intestine (C17), nasal cavity and middle ear (C30), and gum (C03). CONCLUSIONS AND RELEVANCE: This study aimed to provide an in-depth evaluation of spontaneous feline tumours in the European Shorthair cat breed. Results identify SCC as the most commonly represented skin neoplasm. It is likely that the analysed feline population, living in southern latitudes, was more subject to prolonged exposure to ultraviolet light, explaining the discrepancy with previous studies in which SCC was less represented.

The epidemiology of spinal schwannoma in the United States between 2006 and 2014.

OBJECTIVE: Spinal schwannoma remains the third most common intradural spinal tumor following spinal meningioma and ependymoma. The available literature is generally limited to single-institution reports rather than epidemiological investigations. As of 1/1/2004, registration of all benign central nervous system tumors in the United States became mandatory after the Benign Brain Tumor Cancer Registries Amendment Act took action, which provided massive resources for United States population-based epidemiological studies. This article describes the epidemiology of spinal schwannoma in the United States from January 1, 2006, through December 31, 2014. METHODS: In this study, the authors utilized the Central Brain Tumor Registry of the United States, which corresponds to 100% of the American population. The Centers for Disease Control and Prevention's National Program of Cancer Registries and the National Cancer Institute's Surveillance Epidemiology and End Results program provide the resource for this data registry. The authors included diagnosis years 2006 to 2014. They used the codes per the International Coding of Diseases for Oncology, 3rd Edition: histology code 9560/0 and site codes C72.0 (spinal cord), C70.1 (spinal meninges), and C72.1 (cauda equina). Rates are per 100,000 persons and are age-adjusted to the 2000 United States standard population. The age-adjusted incidence rates and 95% confidence intervals are calculated by age, sex, race, and ethnicity. RESULTS: There were 6989 spinal schwannoma cases between the years 2006 and 2014. The yearly incidence eminently increased between 2010 and 2014. Total incidence rate was 0.24 (95% CI 0.23-0.24) per 100,000 persons. The peak adjusted incidence rate was seen in patients who ranged in age from 65 to 74 years. Spinal schwannomas were less common in females than they were in males (incidence rate ratio = 0.85; p < 0.001), and they were less common in blacks than they were in whites (IRR = 0.52; p < 0.001) and American Indians/Alaska Natives (IRR = 0.50; p < 0.001) compared to whites. There was no statistically significant difference in incidence rate between whites and Asian or Pacific Islanders (IRR = 0.92; p = 0.16). CONCLUSIONS: The authors' study results demonstrated a steady increase in the incidence of spinal schwannomas between 2010 and 2014. Male sex and the age range 65-74 years were associated with higher incidence rates of spinal schwannomas, whereas black and American Indian/Alaska Native races were associated with lower incidence rates. The present study represents the most thorough assessment of spinal schwannoma epidemiology in the American population.

Changes in life expectancy for cancer patients over time since diagnosis.

The aims of this study were to provide life expectancy (LE) estimates of cancer patients at diagnosis and LE changes over time since diagnosis to describe the impact of cancer during patients' entire lives. Cancer patients' LE was calculated by standard period life table methodology using the relative survival of Italian patients diagnosed in population-based cancer registries in 1985-2011 with follow-up to 2013. Data were smoothed using a polynomial model and years of life lost (YLL) were calculated as the difference between patients' LE and that of the age- and sex-matched general population. The YLL at diagnosis was highest at the youngest age at diagnosis, steadily decreasing thereafter. For patients diagnosed at age 45 years, the YLL was above 20 for lung and ovarian cancers and below 6 for thyroid cancer in women and melanoma in men. LE progressively increased in patients surviving the first years, decreasing thereafter, to approach that of the general population. YLL in the long run mainly depends on attained age. Providing quantitative data is essential to better define clinical follow-up and plan health care resource allocation. These results help assess when the excess risk of death from tumour becomes negligible in cancer survivors.

Anaplastic meningioma: an analysis of the National Cancer Database from 2004 to 2012.

OBJECTIVE Anaplastic meningiomas represent 1%-2% of meningioma diagnoses and portend a poor prognosis. Limited information is available on practice patterns and optimal management. The purpose of this study was to define treatment patterns and outcomes by treatment modality using a large national cancer registry. METHODS The National Cancer Database was used to identify patients diagnosed with anaplastic meningioma from 2004 to 2012. Log-rank statistics were used to compare survival outcomes by extent of resection, use of adjuvant radiotherapy (RT), and use of adjuvant chemotherapy. Least-squares linear regression was used to evaluate the utilization of RT over time. Logistic regression modeling was used to identify predictors of receipt of RT. Cox proportional hazards modeling was used to evaluate the effect of RT, gross-total resection (GTR), and chemotherapy on survival. RESULTS A total of 755 adults with anaplastic meningioma were identified. The 5-year overall survival rate was 41.4%. Fifty-two percent of patients received RT, 7% received chemotherapy, and 58% underwent GTR. Older patients were less likely to receive RT (OR 0.98, p < 0.01). Older age (HR 1.04, p < 0.01), high comorbidity score (HR 1.33, p = 0.02), and subtotal resection (HR 1.57, p = 0.02) were associated with increased risk of death on multivariate modeling, while RT receipt was associated with decreased risk of death (HR 0.79, p = 0.04). Chemotherapy did not have a demonstrable effect on survival (HR 1.33, p = 0.18). CONCLUSIONS Anaplastic meningioma portends a poor prognosis. Gross-total resection and RT are associated with improved survival, but utilization of RT is low. Unless medically contraindicated, patients with anaplastic meningioma should be offered RT.

Oligodendroglioma resection: a Surveillance, Epidemiology, and End Results (SEER) analysis.

OBJECTIVE The available evidence suggests that the clinical benefits of extended resection are limited for chemosensitive tumors, such as primary CNS lymphoma. Oligodendroglioma is generally believed to be more sensitive to chemotherapy than astrocytoma of comparable grades. In this study the authors compare the survival benefit of gross-total resection (GTR) in patients with oligodendroglioma relative to patients with astrocytoma. METHODS Using the Surveillance, Epidemiology, and End Results (SEER) Program (1999-2010) database, the authors identified 2378 patients with WHO Grade II oligodendroglioma (O2 group) and 1028 patients with WHO Grade III oligodendroglioma (O3 group). Resection was defined as GTR, subtotal resection, biopsy only, or no resection. Kaplan-Meier and multivariate Cox regression survival analyses were used to assess survival with respect to extent of resection. RESULTS Cox multivariate analysis revealed that the hazard of dying from O2 and O3 was comparable between patients who underwent biopsy only and GTR (O2: hazard ratio [HR] 1.06, 95% confidence interval [CI] 0.73-1.53; O3: HR 1.18, 95% CI 0.80-1.72). A comprehensive search of the published literature identified 8 articles without compelling evidence that GTR is associated with improved overall survival in patients with oligodendroglioma. CONCLUSIONS This SEER-based analysis and review of the literature suggest that GTR is not associated with improved survival in patients with oligodendroglioma. This finding contrasts with the documented association between GTR and overall survival in anaplastic astrocytoma and glioblastoma. The authors suggest that this difference may reflect the sensitivity of oligodendroglioma to chemotherapy as compared with astrocytomas.

Integrating cancer diagnosis terminologies based on logical definitions of SNOMED CT concepts.

In oncology, the reuse of data is confronted with the heterogeneity of terminologies. It is necessary to semantically integrate these distinct terminologies. The semantic integration by using a third terminology as a support is a conventional approach for the integration of two terminologies that are not very structured. The aim of our study was to use SNOMED CT for integrating ICD-10 and ICD-O3. We used two complementary resources, mapping tables provided by SNOMED CT and the NCI Metathesaurus, in order to find mappings between ICD-10 or ICD-O3 concepts and SNOMED CT concepts. We used the SNOMED CT structure to filter inconsistent mappings, as well as to disambiguate multiple mappings. Based on the remaining mappings, we used semantic relations from SNOMED CT to establish links between ICD-10 and ICD-O3. Overall, the coverage of ICD-O3 and ICD10 codes was over 88%. Finally, we obtained an integration of 24% (203/852) of ICD-10 concepts with 86% (888/1032) of ICD-O3 morphology concepts combined to 39% (127/330) of ICD-O3 topography concepts. Comparing our results with the 23,684 ICD-O3 pairs mapped to ICD-10 concepts in the SEER conversion file, we found 17,447 pairs of ICD-O3 concepts in common among which 11,932 pairs were integrated with the same ICD-10 concept as the SEER conversion file. The automated process leverages logical definitions of SNOMED CT concepts. While the low quality of some of these definitions impacted negatively the integration process, the identification of such situations made it possible to indirectly audit the structure of SNOMED CT.

Pediatric spinal cord astrocytomas: a retrospective study of 348 patients from the SEER database.

OBJECTIVE Intramedullary spinal cord tumors comprise 1%-10% of all childhood central nervous system neoplasms, with astrocytomas representing the most common subtype. Due to their rarity and poor prognosis, large population-based studies are needed to assess the epidemiology and survival risk factors associated with these tumors in the hope of improving outcome. The authors undertook this retrospective study to explore factors that may influence survival in pediatric patients with spinal cord astrocytomas. METHODS Utilizing the Surveillance, Epidemiology, and End Results (SEER) database, a prospective cancer registry, the authors retrospectively assessed survival in histologically confirmed, primary spinal cord astrocytomas in patients 21 years of age and younger. Survival was described with Kaplan-Meyer curves, and a multivariate regression analysis was used to assess the association of several variables with survival while controlling for confounding variables. RESULTS This analysis of 348 cases showed that age (hazard ratio [HR] 1.05, 95% CI 1.01-1.09, p = 0.017), nonwhite race (HR 1.74, 95% CI 1.11-2.74, p = 0.014), high-grade tumor status (HR 14.67, 95% CI 6.69-32.14, p < 0.001), distant or invasive extension of the tumor (HR 2.37, 95% CI 1.02-5.49, p = 0.046), and radiation therapy (HR 3.74, 95% CI 2.18-6.41, p < 0.001) were associated with decreased survival. Partial resection (HR 0.37, 95% CI 0.16-0.83, p = 0.017) and gross-total resection (HR 0.39, 95% CI 0.16-0.95, p = 0.039) were associated with improved survival. CONCLUSIONS Younger age appears to be protective, while high-grade tumors have a much worse prognosis. Early diagnosis and access to surgery appears necessary for improving outcomes, while radiation therapy has an unclear role. There is still much to learn about this disease in the hope of curing children with the misfortune of having one of these rare tumors.

Building a model for disease classification integration in oncology, an approach based on the national cancer institute thesaurus.

BACKGROUND: Identifying incident cancer cases within a population remains essential for scientific research in oncology. Data produced within electronic health records can be useful for this purpose. Due to the multiplicity of providers, heterogeneous terminologies such as ICD-10 and ICD-O-3 are used for oncology diagnosis recording purpose. To enable disease identification based on these diagnoses, there is a need for integrating disease classifications in oncology. Our aim was to build a model integrating concepts involved in two disease classifications, namely ICD-10 (diagnosis) and ICD-O-3 (topography and morphology), despite their structural heterogeneity. Based on the NCIt, a "derivative" model for linking diagnosis and topography-morphology combinations was defined and built. ICD-O-3 and ICD-10 codes were then used to instantiate classes of the "derivative" model. Links between terminologies obtained through the model were then compared to mappings provided by the Surveillance, Epidemiology, and End Results (SEER) program. RESULTS: The model integrated 42% of neoplasm ICD-10 codes (excluding metastasis), 98% of ICD-O-3 morphology codes (excluding metastasis) and 68% of ICD-O-3 topography codes. For every codes instantiating at least a class in the "derivative" model, comparison with SEER mappings reveals that all mappings were actually available in the model as a link between the corresponding codes. CONCLUSIONS: We have proposed a method to automatically build a model for integrating ICD-10 and ICD-O-3 based on the NCIt. The resulting "derivative" model is a machine understandable resource that enables an integrated view of these heterogeneous terminologies. The NCIt structure and the available relationships can help to bridge disease classifications taking into account their structural and granular heterogeneities. However, (i) inconsistencies exist within the NCIt leading to misclassifications in the "derivative" model, (ii) the "derivative" model only integrates a part of ICD-10 and ICD-O-3. The NCIt is not sufficient for integration purpose and further work based on other termino-ontological resources is needed in order to enrich the model and avoid identified inconsistencies.

Health disparities and impact on outcomes in children with primary central nervous system solid tumors.

OBJECTIVE Health disparities in access to care, early detection, and survival exist among adult patients with cancer. However, there have been few reports assessing how health disparities impact pediatric patients with malignancies. The objective in this study was to examine the impact of racial/ethnic and social factors on disease presentation and outcome for children with primary CNS solid tumors. METHODS The authors examined all children (age ≤ 18 years) in whom CNS solid tumors were diagnosed and who were enrolled in the Texas Cancer Registry between 1995 and 2009 (n = 2421). Geocoded information was used to calculate the driving distance between a patient's home and the nearest pediatric cancer treatment center. Socioeconomic status (SES) was determined using the Agency for Healthcare Research and Quality formula and 2007-2011 US Census block group data. Logistic regression was used to determine factors associated with advanced-stage disease. Survival probability and hazard ratios were calculated using life table methods and Cox regression. RESULTS Children with advanced-stage CNS solid tumors were more likely to be < 1 year old, Hispanic, and in the lowest SES quartile (all p < 0.05). The adjusted odds ratios of presenting with advanced-stage disease were higher in children < 1 year old compared with children > 10 years old (OR 1.71, 95% CI 1.06-2.75), and in Hispanic patients compared with non-Hispanic white patients (OR 1.56, 95% CI 1.19-2.04). Distance to treatment and SES did not impact disease stage at presentation in the adjusted analysis. Furthermore, 1- and 5-year survival probability were worst in children 1-10 years old, Hispanic patients, non-Hispanic black patients, and those in the lowest SES quartile (p < 0.05). In the adjusted survival model, only advanced disease and malignant behavior were predictive of mortality. CONCLUSIONS Racial/ethnic disparities are associated with advanced-stage disease presentation for children with CNS solid tumors. Disease stage at presentation and tumor behavior are the most important predictors of survival.

The effect of pregnancy on survival in a low-grade glioma cohort.

OBJECTIVE The impact of pregnancy on survival in female patients with low-grade glioma (LGG) is unknown and controversial. The authors designed a retrospective cohort study on prospectively collected registry data to assess the influence of pregnancy and child delivery on the survival of female patients with LGG. METHODS In Norway, the reporting of all births and cancer diagnoses to the Medical Birth Registry of Norway (MBRN) and the Cancer Registry of Norway (CRN), respectively, is compulsory by law. Furthermore, every individual has a unique 11-digit identification number. The CRN was searched to identify all female patients with a histologically confirmed diagnosis of World Health Organization (WHO) Grade II astrocytoma, oligoastrocytoma, oligodendroglioma, or pilocytic astrocytoma who were 16-40 years of age at the time of diagnosis during the period from January 1, 1970, to December 31, 2008. Obstetrical information was obtained from the MBRN for each patient. The effect of pregnancy on survival was evaluated using a Cox model with parity as a time-dependent variable. RESULTS The authors identified 65 patients who gave birth to 95 children after an LGG diagnosis. They also identified 281 patients who did not give birth after an LGG diagnosis. The median survival was 14.3 years (95% CI 11.7-20.6 years) for the entire study population. The effect of pregnancy was insignificant in the multivariate model (HR 0.71, 95% CI 0.35-1.42). CONCLUSIONS Pregnancy does not seem to have an impact on the survival of female patients with LGG.

Malignant peripheral nerve sheath tumors of the eighth cranial nerve arising without prior irradiation.

OBJECTIVE Malignant peripheral nerve sheath tumors (MPNSTs) of the eighth cranial nerve (CN) are exceedingly rare. To date the literature has focused on MPNSTs occurring after radiation therapy for presumed benign vestibular schwannomas (VSs), while MPNSTs arising without prior irradiation have received little attention. The objectives of the current study are to characterize the epidemiology, clinical presentation, disease course, and outcome using a large national cancer registry database and a systematic review of the English literature. Additionally, a previously unreported case is presented. METHODS The authors conducted an analysis of the Surveillance, Epidemiology, and End Results (SEER) database, a systematic review of the literature, and present a case report. Data from all patients identified in the SEER database with a diagnosis of MPNST involving the eighth CN, without a history of prior radiation, were analyzed. Additionally, all cases reported in the English literature between January 1980 and March 2015 were reviewed. Finally, 1 previously unreported case is presented. RESULTS The SEER registries identified 30 cases between 1992 and 2012. The average incidence was 0.017 per 1 million persons per year (range 0.000-0.0687 per year). The median age at diagnosis was 55 years, and 16 (53%) were women. Thirteen cases were diagnosed upon autopsy. Of the 17 cases diagnosed while alive, the median follow-up was 118 days, with 3 deaths (18%) observed. When compared with the incidence of benign VS, 1041 VSs present for every 1 MPNST arising from the eighth CN. Including a previously unreported case from the authors' center, a systematic review of the English literature yielded 24 reports. The median age at diagnosis was 44 years, 50% were women, and the median tumor size at diagnosis was 3 cm. Eleven patients (46%) reported isolated audiovestibular complaints typical for VS while 13 (54%) exhibited facial paresis or other signs of a more aggressive process. Treatment included microsurgery alone, microsurgery with adjuvant radiation, or microsurgery with chemoradiation. Sixty-one percent of patients receiving treatment experienced recurrence, 22% of which were diagnosed with drop metastases to the spine. Ultimately, 13 patients (54%) died of progressive disease at a median of 3 months following diagnosis. The ability to achieve gross-total resection was the only feature that was associated with improved disease-specific survival. CONCLUSIONS MPNSTs of the eighth CN are extremely rare and portend a poor prognosis. Nearly half of patients initially present with findings consistent with a benign VS, often making an early diagnosis challenging. In light of these data, early radiological and clinical follow-up should be considered in those who elect nonoperative treatment, particularly in patients with a short duration of symptoms or atypical presentation. These data also provide a baseline rate of malignancy that should be considered when estimating the risk of malignant transformation following stereotactic radiosurgery for VS.

Short-term mortality following surgical procedures for the diagnosis of pediatric brain tumors: outcome analysis in 5533 children from SEER, 2004-2011.

OBJECT Thirty-day mortality is increasingly a reference metric regarding surgical outcomes. Recent data estimate a 30-day mortality rate of 1.4-2.7% after craniotomy for tumors in children. No detailed analysis of short-term mortality following a diagnostic neurosurgical procedure (e.g., resection or tissue biopsy) for tumor in the US pediatric population has been conducted. METHODS The Surveillance, Epidemiology and End Results (SEER) data sets identified patients ≤ 21 years who underwent a diagnostic neurosurgical procedure for primary intracranial tumor from 2004 to 2011. One- and two-month mortality was estimated. Standard statistical methods estimated associations between independent variables and mortality. RESULTS A total of 5533 patients met criteria for inclusion. Death occurred within the calendar month of surgery in 64 patients (1.16%) and by the conclusion of the calendar month following surgery in 95 patients (1.72%). Within the first calendar month, patients < 1 year of age (n = 318) had a risk of death of 5.66%, while those from 1 to 21 years (n = 5215) had a risk of 0.88% (p < 0.0001). By the end of the calendar month following surgery, patients < 1 year (n = 318) had a risk of death of 7.23%, while those from 1 to 21 years (n = 5215) had a risk of 1.38% (p < 0.0001). Children < 1 year at diagnosis were more likely to harbor a high-grade lesion than older children (OR 1.9, 95% CI 1.5-2.4). CONCLUSIONS In the SEER data sets, the risk of death within 30 days of a diagnostic neurosurgical procedure for a primary pediatric brain tumor is between 1.16% and 1.72%, consistent with contemporary data from European populations. The risk of mortality in infants is considerably higher, between 5.66% and 7.23%, and they harbor more aggressive lesions.

Treatment of pediatric Grade II spinal ependymomas: a population-based study.

OBJECT: Grade II spinal cord ependymomas occurring in pediatric patients are exceptionally rare neoplasms. In this paper the authors use a national cancer database to determine patient demographics, treatment patterns, and associated outcomes of this cohort. METHODS: The Surveillance Epidemiology and End Results (SEER) database was used to analyze subjects younger than 18 years with histologically confirmed diagnoses of Grade II spinal cord ependymoma from the years 1973 to 2008. Descriptive data on the demographic characteristics of this cohort and the associated treatment patterns are shown. The Kaplan-Meier method was used to estimate overall survival at 1, 2, 5, and 10 years. RESULTS: This cohort comprised 64 pediatric subjects with Grade II spinal ependymoma. The median age was 13 years, nearly half of the patients were male, and most were white (84%). The median follow-up was 9.2 years. Overall survival at 5 and 10 years was 86% and 83%, respectively. Gross-total resection was achieved in 57% of subjects, and radiation therapy was administered to 36%. Radiation therapy was administered to 78% of subjects after subtotal resection but only to 19% of patients after gross-total resection; this difference was significant (p < 0.001). In a multivariate regression model analyzing sex, age at diagnosis, year of diagnosis, radiotherapy, and extent of resection, female sex was found to be an independent predictor of decreased mortality (HR 0.15 [95% CI 0.02-0.94], p = 0.04). CONCLUSIONS: These data show long-term outcomes for pediatric patients with Grade II spinal ependymoma. Radiotherapy was more likely to be administered in cases of subtotal resection than in cases of gross-total resection. Female sex is associated with decreased mortality, while other demographic or treatment modalities are not.

Descriptive epidemiology of pituitary tumors in the United States, 2004-2009.

OBJECT: Pituitary tumors are abnormal growths that develop in the pituitary gland. The Central Brain Tumor Registry of the United States (CBTRUS) contains the largest aggregation of population-based data on the incidence of primary CNS tumors in the US. These data were used to determine the incidence of tumors of the pituitary and associated trends between 2004 and 2009. METHODS: Using incidence data from 49 population-based state cancer registries, 2004-2009, age-adjusted incidence rates per 100,000 population for pituitary tumors with ICD-O-3 (International Classification of Diseases for Oncology, Third Edition) histology codes 8040, 8140, 8146, 8246, 8260, 8270, 8271, 8272, 8280, 8281, 8290, 8300, 8310, 8323, 9492 (site C75.1 only), and 9582 were calculated overall and by patient sex, race, Hispanic ethnicity, and age at diagnosis. Corresponding annual percent change (APC) scores and 95% confidence intervals were also calculated using Joinpoint to characterize trends in incidence rates over time. Diagnostic confirmation by subregion of the US was also examined. The overall annual incidence rate increased from 2.52 (95% CI 2.46-2.58) in 2004 to 3.13 (95% CI 3.07-3.20) in 2009. Associated time trend yielded an APC of 4.25% (95% CI 2.91%-5.61%). When stratifying by patient sex, the annual incidence rate increased from 2.42 (95% CI 2.33-2.50) to 2.94 (95% CI 2.85-3.03) in men and 2.70 (95% CI 2.62-2.79) to 3.40 (95% CI 3.31-3.49) in women, with APCs of 4.35% (95% CI 3.21%-5.51%) and 4.34% (95% CI 2.23%-6.49%), respectively. When stratifying by race, the annual incidence rate increased from 2.31 (95% CI 2.25-2.37) to 2.81 (95% CI 2.74-2.88) in whites, 3.99 (95% CI 3.77-4.23) to 5.31 (95% CI 5.06-5.56) in blacks, 1.77 (95% CI 1.26-2.42) to 2.52 (95% CI 1.96-3.19) in American Indians or Alaska Natives, and 1.86 (95% CI 1.62-2.13) to 2.03 (95% CI 1.80-2.28) in Asians or Pacific Islanders, with APCs of 3.91% (95% CI 2.88%-4.95%), 5.25% (95% CI 3.19%-7.36%), 5.31% (95% CI -0.11% to 11.03%), and 2.40% (95% CI -3.20% to 8.31%), respectively. When stratifying by Hispanic ethnicity, the annual incidence rate increased from 2.46 (95% CI 2.40-2.52) to 3.03 (95% CI 2.97-3.10) in non-Hispanics and 3.12 (95% CI 2.91-3.34) to 4.01 (95% CI 3.80-4.24) in Hispanics, with APCs of 4.15% (95% CI 2.67%-5.65%) and 5.01% (95% CI 4.42%-5.60%), respectively. When stratifying by age at diagnosis, the incidence of pituitary tumor was highest for those 65-74 years old and lowest for those 15-24 years old, with corresponding overall age-adjusted incidence rates of 6.39 (95% CI 6.24-6.54) and 1.56 (95% CI 1.51-1.61), respectively. CONCLUSIONS: In this large patient cohort, the incidence of pituitary tumors reported between 2004 and 2009 was found to increase. Possible explanations for this increase include changes in documentation, changes in the diagnosis and registration of these tumors, improved diagnostics, improved data collection, increased awareness of pituitary diseases among physicians and the public, longer life expectancies, and/or an actual increase in the incidence of these tumors in the US population.

A proposal for creating a guideline for cancer registration of the fibromatosis, PEComa group, malignant lymphoma In Situ and dendritic cell tumors (III).

BACKGROUND: Understanding the biologic behavior of a tumor is a prerequisite for tumor registration code assignment. The aim of this report was to propose appropriate behavior codes of the International Classification of Disease Oncology 3 (ICD-O3) to rare, yet pathologically interesting hematopoietic and soft tissue tumors. METHODS: The Study Group for Hematopathology, the Bone and Soft Tissue Pathology Study Group, and the Cancer Registration Committee prepared the questionnaire containing provisional behavior codes of selected diseases. RESULTS: In situ lesions of mantle cell and follicular lymphomas, dendritic cell tumors, and neoplasms with perivascular epithelioid cell differentiation (PEComa), not otherwise specified were classified as malignant (-/3). The fibromatosis group, with the exception of lipofibromatosis, was proposed as benign (-/0). Lipofibromatosis and several diseases that belong to the PEComa group were proposed as uncertain malignant potential (-/1). For the hematologic and soft tissue tumors, 274 and 288 members of the Korean Society of Pathologists, respectively, provided opinions through questionnaire, and most responders showed agreement with the provisional behavior code proposed. CONCLUSIONS: The determination of behavior codes for the rare diseases described in this study, especially those of the PEComa group or malignant lymphoma, could be viewed as impractical and premature, but this study provides the basis for future research on this topic.