National Cancer Database Analysis Research Articles

Abstract 3818: Sociodemographic, access to care and survival outcomes for patients with mantle cell lymphoma: A national cancer database analysis with focus on Hispanics

Abstract Background: Mantle cell lymphoma (MCL), represents 3% of new cases of non-Hodgkin’s lymphomas.(PMID 29348844) MCL affects mainly males; factors like ethnicity, age at presentation can have prognostic implications. Currently, it isn’t clear if Hispanics (HI) have a survival difference compared to Non-Hispanics (NH).(PMID 31029647, PMID 25315847). This National Cancer Database (NCDB) analysis aims to clarify how sociodemographic factors affect presentation and survival in HI vs NH patients with MCL. Methods: Data were analyzed on MCL patients in the United States reported to the NCDB 2004-2019. Demographic and treatment characteristics were compared between ethnic groups. Kaplan-Meier and Cox regression analyses were used to compare overall survival (OS). Multivariate analysis and propensity score matching was performed. Results: Of 32515 of MCL patients, 1729 were HI, and 29309 NH. Male gender and identification as white predominated for both. HI were diagnosed at a younger median age of 65 years (y) vs 68 y for NH, p<0.001. Most patients were diagnosed between years 2016-2019. Stage IV predominated for HI and NH, p<0.001. As to Charlson-Deyo Score ≥2, HI had 6% vs NH 8%. Government sponsored was the most prevalent primary payer, HI (54%) and NH (60%). HI had a higher percentage of unknown/uninsured, 13% vs 4% for NH, p<0.001. In relation to census median income (2008-2012), the most prevalent bracket for HI and for NH was >$63,000, 29% and 37% respectively. When analyzing median income of less than $38,000, HI were 22% vs 13% for NH, p<0.001. For median distance in miles (mi) between the patient’s residence and the hospital that reported the case (Great circle distance), HI lived at a median of 8.2mi, vs NH at median of 11.3mi. Most patients were treated at Academic/Research programs, 50% for HI and 39% for NH, p<0.001. On survival analysis, the survival probability at 2, 5 and 10 for HI was 74%, 57%, 44%, vs for NH was 72%, 54%, 36%, respectively. The median survival (MS) was 7.0 years for HI, vs 5.8 years for NH, with a statically significant difference in OS favoring HI, p=0.00015. On multivariate analysis, not insured status was independently associated with worse OS (HR 1.29, CI 1.08-1.55, p=0,01). The propensity matched analysis showed significant MS difference between HI vs NH (5.8 y vs. 5.6 y). Conclusion: In this analysis, MCL is a malignancy that presents at advanced stages, affecting mainly the elderly white males. HI compared to NH have a statically significant benefit in OS, which might be driven by a younger population with less comorbidities. Despite HI having limited access to insurance and lower income; sociodemographic differences don’t seem to be barriers for HI to access care at academic centers. Newer studies should focus on understanding what genetic and intrinsic disease factors cause HI to have an earlier presentation and a survival advantage over NH. Citation Format: Esteban Toro Velez, Carolina Velez-Mejia, Daniel Rosas, Qianqian Liu, Joel Michalek, Adolfo Enrique Diaz Duque. Sociodemographic, access to care and survival outcomes for patients with mantle cell lymphoma: A national cancer database analysis with focus on Hispanics [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3818.

Abstract 4839: Ethnicity dictates survival benefit in Hispanics diagnosed with diffuse large B-Cell lymphoma survival: A national cancer database analysis

Abstract Background: Diffuse Large B-Cell Lymphoma (DLBCL) is the most common Non-Hodgkin lymphoma in the United States (US) but there is limited evidence showing ethnic disparities affecting overall survival (OS) (Blood10.1182/blood-2022-159201, Blood10.1182/blood-2021-151816). This is the largest nationwide cancer registry analysis evaluating ethnic differences for HI vs Non-Hispanics (NH) with DLBCL in the US. Methods: Data were analyzed using the National Cancer Database from 2004-2019. Sociodemographic characteristics were compared between ethnic groups. Kaplan-Meier and Cox regression analyses were used to compare OS between HI and NH. Multivariate analysis and propensity score matching were performed with adjustment for age, stage, comorbidity score, and insurance status, type of facility, and great circle distance Results: 239,391 patients (HI n=18,290, NH n=221,101) were diagnosed with DLBCL. Male sex predominated for both and the majority of patients were Whites. HI were diagnosed at a median age of 62 years (y) vs 68 y for NH [p<0.001]. For HI and NH, most of the patients were diagnosed from 2016-2019. For both groups, the majority of patients had a Charlson-Deyo Score of 0, stage IV at diagnosis and unknown HIV status. The primary payer at diagnosis was government sponsored for HI and NH. In HI the median income based on the Median Income Quartile for 2000 corresponded to the highest level of >$46,000 and from 2008-2012 it was similarly distributed. For NH, for both time periods the majority of the patient were in the highest bracket. Most patients in the HI and NH were located in metropolitan area and were more likely to be treated at a comprehensive cancer center with a great circle distance (miles) of 7.3 for HI and 9.7 for NH. The median survival for HI was 11 y vs 6.8 y for NH. The survival probability at 2, 5 and 10 y for HI corresponded to 69%, 61% and 52%, while for NH it was 66%, 55% and 41%, respectively. The OS probability at 10 y was statistically significant favoring HI [p<0.0001]. Independently, on multivariate analysis, not insured status was associated with worse OS (HR 1.2, CI 1.13-1.28, [p<0.01]) and private insurance type was associated with better OS (HR 0.81, CI 0.77-0.84, [p<0.01]) Conclusion: This nationwide cancer registry study identified better OS in HI diagnosed with DLBCL in the US. Both cohorts had similar sociodemographic and clinical results; however, HI were diagnosed at younger age. This finding may not only help to explain a unique trait towards development of cancer at an earlier age, but also an enhanced response to therapy. Regarding median income there wasn’t a tendency noted for HI, which support the unique complex interactions in socioeconomic status. Further studies examining intrinsic biologic differences are needed to better understand improved OS for HI, which could also help the rational development of targeted therapies. Citation Format: Carolina Velez-Mejia, Esteban Toro Velez, Daniel Rosas, Qianqian Liu, Joel E. Michalek, Enrique Diaz Duque. Ethnicity dictates survival benefit in Hispanics diagnosed with diffuse large B-Cell lymphoma survival: A national cancer database analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4839.

Abstract 3820: Impact of primary site in early-stage follicular lymphoma on patient outcomes: Analysis of a large national database

Abstract Introduction: Follicular lymphoma (FL) is the most common indolent Non-Hodgkin lymphoma and is associated with better outcomes compared to other types of lymphoma. Despite this, cure is usually elusive and most patients experience multiple relapses within their lifetime. Previous studies have identified risk factors associated with worse outcomes, but little is known about the role of primary disease site in FL. We sought to investigate the outcomes of patients with extranodal early-stage FL compared to nodal disease using a large national cancer database. Methods: The Surveillance, Epidemiology, and End Results (SEER) public database was used to identify and extract data for patients with FL. Early-stage disease was defined as stage I or II by the Ann Arbor staging system. Most patients in the database were diagnosed after 2000 so this was used as the diagnosis year cutoff and also to reflect the impact of recent management strategies on outcomes. Patients were divided into two groups based on their primary disease site. These two groups were nodal (includes spleen) and extranodal. Treatment effect (surgery, radiation, or chemotherapy) was analyzed when available. Survival was estimated using the Kaplan-Meier method and compared using log-rank tests. Cox proportional hazards models were used for adjusted survival analyses. Results: A total of 13,400 patients with FL of all histologic grades diagnosed between 2000 and 2015 were identified in the SEER database. The median age at diagnosis was 65 years (2-100). A majority of patients were Caucasian (81%), had stage I disease (62%), and belonged to the nodal FL group (79%). Median overall survival (OS) for patients with nodal FL was 181 months [95% CI (175-187)]. Median OS for extranodal FL was 189 months [95% CI (179-195)]. Extranodal FL was associated with better OS compared to the nodal group [HR=0.89, 95% CI (0.84-0.96); p-value =0.00012] even after adjusting for age and race [HR=0.84, 95% CI (0.79-0.90); p-value <0.0001]. Conclusions: In this analysis of a large national cancer database, early-stage FL presenting as extranodal disease can be seen in about 20% of patients and is associated with slightly better outcomes. Citation Format: Olivia Davis, Rana Kasht, Andrew Cohoon, Silas Day, Sami Ibrahimi, Adam Asch, Taha Al-Juhaishi. Impact of primary site in early-stage follicular lymphoma on patient outcomes: Analysis of a large national database [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3820.

Race and Odds of Surgery Offer in Small Bowel and Pancreas Neuroendocrine Neoplasms.

Despite existing society guidelines, management of pancreatic (PanNEN) and small bowel (SBNEN) neuroendocrine neoplasms remains inconsistent. The purpose of this study was to identify patient- and/or disease-specific characteristics associated with increased odds of being offered surgery for PanNEN and SBNEN. The Surveillance, Epidemiology, and End Results (SEER) Program database and the National Cancer Database (NCDB) were queried for patients with PanNEN/SBNEN. Demographic and pathologic data were compared between patients who were offered surgery and those who were not. Multivariate logistic regression was performed to identify factors independently associated with being offered surgery. In SEER, there were 3641 patients with PanNEN (54.7% were offered surgery) and 5720 with SBNEN (86.0% were offered surgery). On multivariate analysis of SEER, non-white race was associated with decreased odds of surgery offer for SBNEN [odds ratio (OR) 0.58, p < 0.001], but not PanNEN (p = 0.187). In NCDB, there were 28,483 patients with PanNEN (57.5% were offered surgery) and 42,675 with SBNEN (86.9% were offered surgery). On multivariate analysis of NCDB, non-white race was also associated with decreased odds of surgery offer for SBNEN (OR 0.61, p < 0.001) but not PanNEN (p = 0.414). This study's findings suggest that, in addition to previously reported disparities in surgical resection and surgery refusal rates, racial/ethnic disparities also exist earlier in the course of treatment, with non-white patients being less likely to be offered surgery for SBNEN but not for PanNEN; this is potentially due to discrepancies in rates of referral to academic centers for pancreas and small bowel malignancies.

Increased Pathologic Downstaging with Induction versus Consolidation Chemotherapy in Patients with Locally Advanced Rectal Cancer Treated with Total Neoadjuvant Therapy—A National Cancer Database Analysis

Total neoadjuvant therapy (TNT) is the recommended treatment for locally advanced rectal cancer. The optimal sequence of TNT is debated: induction (chemotherapy first) or consolidation (chemoradiation first)? We aim to evaluate the practice patterns and clinical outcomes of total neoadjuvant therapy with either induction or consolidation regiments in the United States for patients with locally advanced rectal cancer. Methods: This is a retrospective analysis of the National Cancer Database for patients with clinical stage II or stage III rectal cancer, diagnosed between 2006 and 2017, who underwent total neoadjuvant therapy followed by surgery. Results: From 2006 to 2017, we identified 8999 patients and found that the utilization of induction chemotherapy increased from 2.0% to 35.0%. TNT resulted in pathologic downstaging 46.7% of the time and a pathologic complete response 11.6% of the time. Induction chemotherapy lead to higher pathologic downstaging (58% vs. 44.7%, p &lt; 0.001) and pathologic complete responses (16.8% vs. 10.7%, p &lt; 0.001). Similar trends held true in a multivariate analysis and subset analysis of stage II and III disease. Conclusions: These findings suggest that induction chemotherapy may be preferred over consolidation chemotherapy when downstaging prior to oncologic resection is desired. The optimal treatment plan for total neoadjuvant therapy is multi-factorial and requires further elucidation.

Ten-year overall survival of localized lower thoracic and esophagogastric junction carcinoma receiving trimodality therapy: A National Cancer Database analysis.

278 Background: Trimodality therapy, consisting of chemotherapy and radiation followed by esophagectomy, represents the prevailing standard of care for most patients with localized lower thoracic and esophagogastric junction (EGJ) carcinoma. However, robust data reporting the survival of this patient group treated in the real-world setting is lacking. The primary objective of the current study was to comprehensively analyze the long-term survival of this patient population receiving treatment in a real-world setting and the variables that significantly influence the survival. Methods: We identified adult patients in the National Cancer Database with localized (cT1N+M0 or cT2-T4N+/-M0) lower thoracic or EGJ adenocarcinoma (AC) or squamous cell carcinoma (SCC) receiving trimodality therapy between 2004 and 2020 with multi-agent chemotherapy and ≥ 4139 cGy of radiation followed by esophagectomy. The primary endpoint of the analysis was overall survival (OS), estimated using the Kaplan-Meier method. The impact of clinical and demographic features on OS was determined using the log-rank test. We performed a multivariable (MV) Cox analysis to evaluate the independent association of clinical and demographic variables on OS (data presented as hazard ratio [HR] and 95% Confidence Interval [CI]). Results: The analysis included 21,965 patients, characterized by a median age of 63 years, predominance of males (85.9%), non-Hispanic white patients (89.4%), and AC histology (93.6%). Most patients had T2/T3 tumors (87.3%) and regional lymph node (LN) involvement (62.1%). The 10-year OS for the entire cohort was 27.6% (95% CI, 26.8%-28.5%), with a median OS of 40.7 months (95% CI, 39.7-41.6). For patients aged ≤ 65, the 10-year OS was 31.2% (95% CI, 30.2%-32.3%), with a median OS of 44.1 months (95% CI, 42.8-45.5). On MV analysis, the following factors were associated with inferior OS with CI in the parenthesis: age ≥ 66 years (1.17, 1.11-1.22); male gender (1.13, 1.08-1.19); histology, AC vs. SCC (1.19, 1.10-1.28); T4 vs. T1-T3 tumor stage (1.20, 1.07-1.34); lymph node involvement (1.22, 1.18-1.27); Charlson-Deyo score ≥ 2 (1.23, 1.15-1.32); uninsured vs private insurance (1.20, 1.05-1.36); and low-volume facility (1.08, 1.01-1.16). Furthermore, patients treated after 2012 (2013-2019) had modestly improved OS (0.87, 0.84-0.90), perhaps reflecting the impact of the CROSS trial result published in 2012. Conclusions: This study illustrates the grim long-term survival of patients with localized lower thoracic and EGJ cancer undergoing standard trimodality therapy, underscoring the urgent necessity for innovative therapeutic approaches for this patient group. The improvement in OS following the publication of the CROSS trial was modest.

Neoadjuvant chemotherapy vs upfront resection: A challenging choice in resectable pancreatic ductal adenocarcinoma.

620 Background: Pancreatic ductal adenocarcinoma (PDAC) remains an immense challenge within oncology, with only 15-20% of cases initially qualifying as resectable. Based on National Cancer Database analyses there is an increase in neoadjuvant chemotherapy (NACT) use, with potential downstaging effects. Yet, a subset of neoadjuvant-treated patients fail to proceed with curative-intent surgery, leading to worse outcomes. This retrospective analysis aims to uncover factors that predict NACT use and the impact on survival outcomes. Methods: A cohort of 543 consecutive non-metastatic PDAC patients treated at an NCCN-designated comprehensive cancer center between 2015 and 2022 were prospectively enrolled into a REDCap database. Based on NCCN resectability criteria, initial CT staging placed 96 patients as upfront resectable. Survival outcomes were compared using univariate and multivariate cox regression analyses. Results: Within 96 resectable patients, 45 (47%) underwent upfront surgery, 36 (37%) received NACT followed by surgery and 15 (16%) only received chemotherapy. The majority were CA19-9 secretors (94%) with 54% lower values during chemotherapy. While 60% of patients experienced up to 35% tumor size reductions on NACT, 20% had stable disease, and 20% showed progressive disease. The patients who received NACT were significantly younger (p = 0.01), with higher baseline CA19-9 (p &lt; 0.01) and more frequently underwent stent placement (p = 0.01), table 1. Median overall survival for the upfront surgery, NACT+surgery and chemotherapy groups were 32.8, 23.7 and 20.2 months (log-rank test p&lt;0.01) subsequently. On univariate analysis initial presentation with weight loss (p = 0.02, HR = 2.08) and high baseline CA19-9 (log2-scaled, p = 0.01, HR = 1.13) were predictive of worse overall survival. After adjusting for potential confounding variables, patients in chemotherapy group had significantly worse survival compared to the upfront surgery group (P &lt; 0.01, HR = 3.14). These 15 patients failed to undergo surgery due to diverse reasons, including distant progression (7), local progression (3), performance status decline or death during chemotherapy (2), refusal due to chemotherapy side effects (2), and indecision and loss to follow-up (1). Conclusions: These findings underscore the need for more robust selection criteria in resectable PDAC patients to undergo NACT. Leveraging real-world data helps identify predictive factors of treatment decision making and their impact on patient survival outcomes. [Table: see text]

Patterns of surgery and adjuvant therapy in pancreatic carcinoid tumor: An NCDB analysis.

631 Background: Pancreatic Carcinoid Tumor (PCT) is a neuroendocrine tumor (NET) of the gastrointestinal (GI) tract composed of Argentaffin cells. Of the primary pancreatic cancers, NETs are exceedingly rare in comparison to pancreatic adenocarcinoma, accounting for less than 2% of pancreatic cancer diagnoses. PCTs constitute for only a subset of these and hold a better prognosis than adenocarcinoma (What Is a Pancreatic Neuroendocrine Tumor? | American Cancer Society, n.d.). GI carcinoid tumors classically originate in the appendix, ileum, or rectum, and are seldom a primary neoplasm of the pancreas. PCTs generate secretory products with Serotonin being the most frequently isolated. Some cases may produce a carcinoid syndrome with characteristic diarrhea, cutaneous flushing, and bronchospasm (Ha &amp; Tan, 2012). Surgical resection is the hallmark of treatment, though other modalities of management are available. Due to the rarity of PCT, no significant study exists in the literature analyzing the association of survival with receipt of surgery or predictors of receipt of surgery. Methods: The National Cancer Database (NCDB) was used to identify patients diagnosed with PCT from 2004 to 2019 using the histology codes 8240 as assigned by the Commission on Cancer Accreditation program. Kaplan-Meier, ANOVA Chi-Square, Binary Logistic Regression and Cox Proportional Hazards tests were performed. Data was analyzed using SPSS version 29 and statistical significance was set at α = 0.05. Results: 2237 patients with PCT formed the final sample, with an average age of 61 years of age. 724 patients (32.3%) underwent surgery of the primary site, 72 (3.2%) received adjuvant chemotherapy, and 9 (0.4%) received adjuvant radiation. Surgery alone was associated with the longest median survival ( 198.5 months) compared to the other treatments ( Adjuvant chemotherapy=155.3 months, Adjuvant radiation=122.7 months, No surgery=73.95 months; p&lt;0.05). After controlling for covariates, receipt of surgery was independently associated with decreased overall mortality ( HR=.380; p&lt;0.05). Decreased age, lower grade tumors, and fewer positive regional lymph nodes were significantly associated with increased odds of receiving surgery. Additionally, treatment at an academic facility and private insurance status was associated with significantly higher rates of surgery. Patients who did not receive surgery were significantly older, had later stage disease, higher co-morbidity scores, and larger tumor size. Income level and race had no effect on treatment selection. Conclusions: This study found that surgery alone without adjuvant therapy is independently associated with improved survival. Favorable clinical characteristics may be contributing to this discrepancy in survival. Further exploration of treatment strategies in PCT may benefit patients through the creation of definitive treatment guidelines to manage this disease.

Definitive Local Treatment for Metastatic Prostate Cancer: A National Cancer Database Analysis

Introduction: Prostate cancer (PCa) is the second leading cause of cancer mortality in US males, with metastatic disease (mPCa) comprising 5% of cases. This represents an incurable and uniformly lethal disease. Thus, novel therapeutic approaches are needed. Cytoreductive surgery in the metastatic setting has become commonplace for renal cell carcinoma, and definitive local radiotherapy has been employed in the oligometastatic setting for prostate cancer via the STAMPEDE trial. This study further explores the use of definitive local treatment via pelvic radiotherapy or radical prostatectomy along with androgen deprivation therapy (ADT) vs. ADT alone in the setting of metastatic prostate cancer. Methods: The 2019 National Cancer Database (NCDB) was queried to conduct a retrospective cohort analysis of cT1-4N0-3M1 PCa who received local therapy in conjunction with ADT vs. ADT alone. Clinicopathologic variables were compared between the two groups using Wilcoxon signed rank and Chi-square for continuous and categorical variables, respectively. Overall survival (OS) analysis was performed using Cox Proportional Hazards and the Kaplan-Meier method. Comparisons were made between local therapy + ADT vs. ADT alone and between radical prostatectomy + ADT vs. radiation therapy + ADT vs. ADT alone. Results: A total of n=36,635 patients with cM1 were identified, with 3197 (8.7%) patients receiving local therapy + ADT. Among local therapy + ADT, 2884 (90.2%) patients received radiation therapy + ADT, and 313 (9.8%) received radical prostatectomy + ADT. The median follow-up was 2.8 years. Kaplan-Meier analysis showed significant improvement in 5-year OS for patients who received local therapy + ADT vs. ADT alone. For ADT alone, 5-year OS was 31.3% (CI= 30.7-31.8%) vs. 54.2%. (CI= 52.4-56.1%) for local therapy +ADT. Furthermore, comparing the type of local therapy on Kaplan-Meier analysis: radical prostatectomy + ADT showed better 5-yr OS, 74.0% (CI= 67.5-79.1%) vs. 52.2% (CI= 50.2-54.2%) for radiation therapy +ADT (p&lt;0.001). Conclusion: Definitive local treatment, in addition to ADT, may improve 5-year OS for well-selected patients with metastatic prostate cancer. Patient outcomes are significantly improved for those treated with radical prostatectomy vs. radiation therapy. These findings support multimodal treatment for metastatic prostate cancer, and further studies are needed to optimize criteria for patient selection and choice of definitive localized therapy in this setting.

The impact of lymph node ratio on survival in gallbladder cancer: national cancer database analysis

The impact of lymph node ratio on survival in gallbladder cancer: national cancer database analysis

Identifying factors that are associated with survival in patients with solid pseudopapillary tumors of the pancreas: a national cancer database analysis

Identifying factors that are associated with survival in patients with solid pseudopapillary tumors of the pancreas: a national cancer database analysis

Epidemiologic Characteristics, Treatment Patterns, and Survival Analysis of Plasmablastic Lymphoma in the United States: A SEER and NCDB Analysis

BackgroundPlasmablastic Lymphoma (PBL) is a rare aggressive B-cell lymphoma that primarily affects immunocompromised individuals, including those living with HIV. Historically, survival estimates are dismal and range from 8 to 15 months. We aimed to evaluate epidemiologic characteristics, treatment patterns and survival trends on a national scale. Patients and MethodsPatients diagnosed with PBL from 2010 to 2020 were identified in the National Cancer Database (NCDB) and in the Surveillance, Epidemiology, and End Results (SEER) program. Incidence rates were calculated using SEER. Demographic features, treatment characteristics, and overall survival (OS) were identified using the NCDB. ResultsWe identified 1153 patients in the SEER database and 1822 patients in the NCDB. The incidence of PBL is 0.07 cases per 100,000 US population per year. PBL is more common in males (77%), and white patients (77%), with 50% of cases in patients with HIV. Patients who were treated with multiagent chemotherapy had a median OS of 58.6 months. On multivariate Cox regression, we found that HIV status did not have a significant impact on OS. Factors associated with worse OS included advancing age and stage. ConclusionWe present the largest study to date on PBL. Among treated patients, we described a median OS of 58.6 months, greatly improved from previously reported estimates. We found that HIV status did not have a significant impact on OS. While OS remains poor, therapeutic advances over the last decade are promising and highlight the need for continued clinical advances aimed at improving therapeutic options for this rare lymphoma.

Impact of HER2-low status for patients with early-stage breast cancer and non-pCR after neoadjuvant chemotherapy: a National Cancer Database Analysis.

To investigate potential differences in pathological complete response (pCR) rates and overall survival (OS) between HER2-low and HER2-zero patients with early-stage hormone receptor (HR)-positive and triple-negative breast cancer (TNBC), in the neoadjuvant chemotherapy setting. We identified early-stage invasive HER2-negative BC patients who received neoadjuvant chemotherapy diagnosed between 2010 and 2018 in the National Cancer Database. HER2-low was defined by immunohistochemistry (IHC) 1+ or 2+ with negative in situ hybridization, and HER2-zero by IHC0. All the methods were applied separately in the HR-positive and TNBC cohorts. Logistic regression was used to estimate the association of HER2 status with pCR (i.e. ypT0/Tis and ypN0). Kaplan-Meier method and Cox proportional hazards model were applied to estimate the association of HER2 status with OS. Inverse probability weighting and/or multivariable regression were applied to all analyses. For HR-positive patients, 70.9% (n = 17,934) were HER2-low, whereas 51.1% (n = 10,238) of TNBC patients were HER2-low. For both HR-positive and TNBC cohorts, HER2-low status was significantly associated with lower pCR rates [HR-positive: 5.0% vs. 6.7%; weighted odds ratio (OR) = 0.81 (95% CI: 0.72-0.91), p < 0.001; TNBC: 21.6% vs. 24.4%; weighted OR = 0.91 (95% CI: 0.85-0.98), p = 0.007] and improved OS [HR-positive: weighted hazard ratio = 0.85 (95% CI: 0.79-0.91), p < 0.001; TNBC: weighted hazard ratio = 0.91 (95% CI: 0.86-0.96), p < 0.001]. HER2-low status was associated with favorable OS among patients not achieving pCR [HR-positive: adjusted hazard ratio = 0.83 (95% CI: 0.77-0.89), p < 0.001; TNBC: adjusted hazard ratio = 0.88 (95% CI 0.83-0.94), p < 0.001], while no significant difference in OS was observed in patients who achieved pCR [HR-positive: adjusted hazard ratio = 1.00 (95% CI: 0.61-1.63), p > 0.99; TNBC: adjusted hazard ratio = 1.11 (95% CI: 0.85-1.45), p = 0.44]. In both early-stage HR-positive and TNBC patients, HER2-low status was associated with lower pCR rates. HER2-zero status might be considered an adverse prognostic factor for OS in patients not achieving pCR.

Racial and Ethnic Disparities for Primary CNS Lymphoma: A National Cancer Database Analysis with Emphasis on Hispanics

BACKGROUND Primary central nervous system lymphoma (PCNSL) is an aggressive, extra nodal non-Hodgkin lymphoma (NHL), that accounts for ~4% of newly diagnosed central nervous system (CNS) tumors and 1-2% of all NHL. ( Blood PMID 26980727). The majority of PCNSL tumors are of the highly aggressive diffuse large cell subtypes, usually of B-cell phenotypic origin ( J Clin OncolPMID 28640701). Treatment survival remains heterogenous due to biologic differences and racial disparities. Multiple epidemiological studies have evaluated racial differences, with heterogenous results. There are published reports of culturally and ethnically diverse populations from high-income countries, showing statistically significant differences in PCNSL incidence and survival rates among different ethnicities ( Neuro OncolPMID 19273630), but there are no studies comparing Hispanics (HI) and Non-Hispanics (NH). This National Cancer Database (NCDB) analysis aims to provide information on how demographic, clinical, and survival outcomes differ in HI versus NH patients with PCNSL. METHODS Data were analyzed on PCNSL patients in the US reported to the NCDB between 2004 and 2019. Demographic and treatment characteristics were compared between ethnic groups. Kaplan-Meier and Cox regression analyses were used to compare OS between HI and NH populations. Multivariate analysis and propensity score matching was performed with adjustment for age, stage, co-morbidity score, and insurance status, type of facility and great circle distance. RESULTS Of 15908 PCNSL patients, 8% were HI, and 87% NH. 58% of HI were male, compared to 50% in NH. As compared to NH at diagnosis, HI pts were younger (mean 60 years vs. 67 years) (p&amp;lt;0.001); the majority of HI (59%) were diagnosed in the age of &amp;lt;65 vs majority of NH (45%) who were diagnosed in the age bracket of 65-70 (p&amp;lt;0.001). Regarding race, most of HI and NH were whites (92% vs 85%), followed by blacks (0.9% vs 7.3%). When examining insurance type; Government Sponsored was the most prevalent type in HI vs NH (56% vs 61%), followed by private insurance (29% vs 33%). The most non-Insured group was HI (12%) vs NH (2%). Regarding Census Median Income (2008-2012), the most prevalent bracket (29%) for HI was $48,000 - $62,999, and for NH (36%) $63,000 +. When looking at median income of less than $38,000, HI were 25% vs NH 14%. Regarding Charlson-Deyo Score (comorbidities score), HI had 20% =/&amp;gt; 2 score, vs NH 16%. (p&amp;lt;0.001). Academic/Research Programs, including NCI-designated comprehensive cancer centers, was the most prevalent type of facility providing care for both HI (49%) and NH (50%). Regarding the median distance in miles between the patient's residence and the hospital that reported the case (Great Circle Distance), HI lived at a median of 9.7 miles, vs NH at median of 14.7 miles. On survival analysis, the survival probability at 2, 5 and 10 years of HI vs NH were (54% vs 47%), (43% vs 34%), and (32% vs 22%), respectively. The median survival time (MS) was 3 years for HI, vs 1.6 years for NH; and there was an overall survival (OS) difference favoring HI with (p&amp;lt;0.0001). Independently, on multivariate analysis, not insured status was associated with worse OS (HR 1.22, CI 1.02-1.47, p&amp;lt;0.01) and private insurance type was associated with better OS (HR 0.82, CI 0.73-0.93, p&amp;lt;0.01). Interestingly, the propensity matched analysis demonstrated significant MS difference between HI vs NH (median 1.64 years vs. 1.43 years; p&amp;lt;0.01). CONCLUSION In this data analysis from the NCDB between 2004-2019, HI with PCNSL have dissimilar outcomes when compared to NH. There were statistically significant differences between HI and NH in age at diagnosis, median income, and higher Charlson-Deyo Score. There were statistically significant differences in MS, survival at 2, 5, and 10 years and OS favoring the HI cohort. Unique and heterogeneous characteristics from HI may explain why there is an increase in OS after PCNSL diagnosis in this cohort of patients, with differences mainly arising from HI being diagnosed at a younger age and living closer to the treatment facilities as compared to NH. As personalized therapies for PCNSL are being developed, equitable access including clinical trials should be advocated for resource-limited settings as well as aiming at studying biologically characteristics of HI with PCNSL to better understand and improved outcomes in all cohorts.

Impact of Facility Type on Survival in Chronic Myelomonocytic Leukemia: A Propensity Score Matched, National Cancer Database Analysis

Introduction Chronic myelomonocytic leukemia (CMML) is an MDS/MPN overlap syndrome with a heterogenous clinical presentation and limited therapeutic options. Given its rarity, we hypothesize that there may be disparate outcomes based on the type of treatment facility (academic versus community) similar to what has been observed in other myeloid malignancies such as acute myeloid leukemia (VR Bhatt AJH 2017). To that end, we performed an analysis of the National Cancer Database (NCDB) to assess the impact of academic status of the facility on overall survival (OS) in patients with CMML. Methods We identified patients in the NCDB by the ICD-O-3 code for CMML (9945); we excluded patients who did not receive treatment and those who received all or part of their treatment outside of the reporting facility. Importantly, baseline clinical characteristics, laboratory/mutational data, and treatment details are generally not available in the NCDB. Academic centers were defined as those with a peer-reviewed NCI-designated cancer center support grant or university medical school that provide postgraduate medical education in at least four program areas, as defined by the Commission on Cancer Accreditation program. The primary outcome was OS. Propensity score matching (PSM) was performed and after backwards selection, the following variables were used for matching: age, race, median income, urban/rural, education, Charlson comorbidity index (CCI). Cox proportional hazard multivariable regression was employed to identify independent predictors of survival in the PSM population. Results A total of 6290 patients were identified. Baseline characteristics based on facility type are shown in Table 1. Patients treated at an academic center were younger, with a higher prevalence of Black race and lower comorbidity burden. Academic centers also had a patient population with a significantly higher median household income, more private insurance, and higher education level as compared to non-academic centers. The median follow-up time was 14.6 months (95% CI 0.4-89.7). The median OS for the entire cohort was 59 months (95% CI 56.7-62.0) and the median survival was significantly longer in patients treated at an academic center (60.6 months [95% CI 57.5-64.8]) as compared to those treated at a non-academic center (57.3 months [95% CI 53.4-61.8]). In the PSM analysis, a cohort of 1884 patients treated at an academic center were matched to 1895 patients treated at a non-academic center. In an unadjusted analysis, there was no significant difference in median OS (p=0.3) between the academic and non-academic groups. However, in a multivariable analysis of the PSM populations that accounted for age, race, CCI, insurance, distance to center, and median income, patients treated at an academic center had a significantly improved survival as compared to those treated at a non-academic center (HR 0.82, 95% CI 0.71-0.94, p = 0.005, Figure 1). Discussion We found significant differences in the characteristics of CMML patients treated at academic versus non-academic centers. After PSM and multivariable adjustment, survival was significantly improved with treatment at an academic center versus non-academic. Although limited in the depth of information about CMML prognostic features and treatments, this relatively large analysis of CMML patients suggests that management of this rare disease at an academic center may be associated with improved survival outcomes. Future prospective studies to ascertain the underlying causes for survival discrepancy in CMML at academic versus non-academic centers may help identify patients that should be triaged to an academic facility.

Treatment Pattern and Survival Outcome of Limited Stage Mantle Cell Lymphoma: A National Cancer Database Analysis in Pre-BTK Inhibitor Vs BTK Inhibitor Era

Background Limited stage (stage I−II) disease is an uncommon presentation of mantle cell lymphoma (MCL), and its optimal management remains controversial. A British Columbia study (Leitch, Ann Oncol, 2003) and a Surveillance, Epidemiology and End Results (SEER) study (Murthy, Clin Lymphoma Myeloma Leuk, 2014) suggested the importance of radiotherapy (RT). A National Cancer Database (NCDB) study in the pre-BTK inhibitor (BTKi) era showed that chemotherapy (Chemo) plus RT (Chemo+RT) improved overall survival (OS) compared to Chemo or RT alone (Gill, Int J Radiat Oncol Biol Phys, 2015). However, a retrospective International Lymphoma Radiation Oncology Group study (Dabaja, Ann Oncol, 2017) reported comparable OS with Chemo+RT vs Chemo or RT alone. The potential differences in OS with different frontline therapy modalities may be modified by subsequent therapies for relapsed or refractory (R/R) disease. In this study, we used the NCDB data to analyze frontline treatment patterns and outcomes of limited stage MCL in the pre-BTKi vs BTKi eras. Methods The NCDB was queried to identify patients with stage I−II MCL diagnosed in 2004−2020. Because the first BTKi ibrutinib was approved by the FDA for R/R MCL in November 2013, we defined 2004−2013 as pre-BTKi era and 2014−2020 as BTKi era. Frontline treatment modalities were categorized to RT alone, Chemo alone, or Chemo+RT. Survival analysis was performed using the Kaplan-Meier method and the Cox proportional hazards model. Results Among 32,746 patients with newly diagnosed MCL in 2004−2020, 3,836 (11.7%) presented with stage I−II disease, with 2,428 having treatment information available. For patients diagnosed in the pre-BTKi era (n = 2005), 195 (9.7%) received RT alone, 1472 (73.4%) received Chemo alone, and 338 (16.9%) received Chemo+RT. For patients diagnosed in the BTKi era (n = 423), 45 (10.6%) received RT alone, 320 (75.7%) received Chemo alone, and 58 (13.7%) received Chemo+RT. Patients in the RT alone group were older (P &amp;lt;0.001) and had more females (P &amp;lt;0.001), but there were no statistically significant differences in Charlson-Deyo score, international prognostic index, or treating facility type (academic vs non-academic). For patients with stage I disease, in the pre-BTKi era, patients treated with Chemo+RT (n = 182) had improved OS compared to those treated with RT alone (n = 141) or Chemo alone (n = 647), with a 5-year OS of 75.4% vs 71.8% vs 61.4% (P &amp;lt;0.001). Similar results were observed in the BTKi-era, with a 5-year OS of 83.6% vs 69.5% vs 62.9% (P &amp;lt;0.001) for patients treated with Chemo+RT (n = 31), RT alone (n = 30), or Chemo alone (n = 130), respectively. After adjusting for age and sex, Chemo+RT was associated with improved OS compared to Chemo alone (pre-BTKi era, HR = 0.59, 95% CI = 0.46−0.75, P &amp;lt;0.001; BTKi era, HR = 0.34, 95% CI = 0.14−0.79, P = 0.013), but not compared to RT alone (pre-BTKi era, HR = 0.62, 95% CI = 0.23−1.68, P = 0.343; BTKi era, HR = 0.77, 95% CI = 0.56−1.05, P = 0.10) (Figure 1A). For patients with stage II disease, in the pre-BTKi era, patients treated with Chemo+RT (n = 156) had improved OS compared to those treated with RT alone (n = 54) or Chemo alone (n = 825), with a 5-year OS of 70.0% vs 47.3% vs 58.4% (P &amp;lt;0.001). In the BTKi-era, patients treated with Chemo+RT (n = 27) or RT alone (n = 15) both had improved OS compared to those treated with Chemo alone (n = 190), with a 5-year OS of 77.2% vs 78.8% vs 61.9% (P = 0.004) . After adjusting for age and sex, Chemo+RT was associated with improved OS compared to Chemo alone in BTKi era only (pre-BTKi era, HR = 0.85, 95% CI = 0.67−1.07, P = 0.165; BTKi era, HR = 0.49, 95% CI = 0.25−0.95, P = 0.034), but not compared to RT alone (pre-BTKi era, HR = 0.78, 95% CI = 0.53−1.13, P = 0.185; BTKi era, HR = 0.97, 95% CI = 0.35−2.67, P = 0.952) (Figure 1B). Conclusions Treatment pattern for limited stage MCL has not changed significantly from the pre-BTKi era to the BTKi era, with approximately 15% of patients receiving Chemo+RT, 10% RT alone and 75% Chemo alone. In the BTKi era, Chemo+RT appeared to be associated with better OS compared to Chemo alone but not compared to RT alone. Interpretation of the results requires caution because (1) patient numbers are small in Chemo+RT and RT alone groups, (2) longer follow-up may be needed, and (3) selection of therapy was not random, eg, patients receiving Chemo alone may have had higher disease burden such as bulky disease and/or non-contiguous stage II involvement but these details are not available in NCDB.

Machine Learning Can Outperform Ann Arbor Staging in Predicting Survival in Patient with Diffuse Large B-Cell Lymphoma: Analysis of a Large National Cancer Database

Introduction: Diffuse Large B-cell Lymphoma (DLBCL) is the most common lymphoma in the world with usually an aggressive clinical course. The Ann Arbor staging system and International Prognostic Index (IPI) commonly utilized in clinical practice for risk stratification have known limitations. Machine learning (ML) has emerged as a promising tool for more comprehensive and deeper data analysis. We sought to utilize the ability of ML to predict survival in DLBCL compared to Ann Arbor staging system using a large national database. Methodology: We employed the ML algorithm XGBoost on the National Cancer Institute's Surveillance, Epidemiology and End Result (SEER) database to predict overall survival (OS) and the lymphoma specific survival (LSS). For prediction analysis, we transformed the survival labels into a simple Boolean format: “alive” represented as 0, “dead” as 1, and “dead (attributable to this cancer diagnosis)” also as 1. We utilized one-hot encoding to convert categorical features and variables into binary vectors. The data set was divided into two parts: training (80%) and test (20%). Further, we split the training set into the actual training set and validation set by using stratified 5-fold cross validation. Hyper-parameter optimization was done within the validation set. A broad range of attributes were utilized by the model for its prediction algorithm. To understand how each attribute contributes to predictions, we calculated its importance score in XGBoost. Results: A total of 64,912 patients with DLBCL were found and their data were extracted. The majority were Caucasian (78.9%) with a median age range of 60 to 69. The model was able to predict OS and LSS, with an area under the curve (AUC) of 0.89 and 0.75 (Figure 1), respectively. Factors selected by the model for survival prediction included presence or absence of B-symptoms, treatment status, and disease stage. For OS and LSS, the model found B symptoms to be the highest contributing factor with an importance score of 0.205 and 0.167, respectively. Other important factors incorporated by the model included age and stage IV for OS, and stage IV and clinically asymptomatic status for LSS. The least important factors were location of the primary lymphoma site and year of diagnosis (Table 1). Conclusion: Machine learning tools can help predict survival in patients with DLBCL and able to challenge current staging systems. Our results warrant validation in future prospective studies.

A National Cancer Database Analysis of Demographics, Treatment Patterns and Survival Trends of Patients with Plasmablastic Lymphoma in US: Does Ethnicity Predict Enhanced Outcomes?

Background: Plasmablastic lymphoma (PbL) is an exceedingly rare, aggressive non-Hodgkin lymphoma with overlapping features of myeloma and lymphoma diagnosed most commonly in immunocompromised individuals, especially in the setting of human immunodeficiency virus (HIV) and Epstein-Barr virus (EBV) infection ( Int J Lab Hematol PMID 36074710). Pathologically, it is characterized by lacking CD20 and PAX5 expression with the presence of CD38, CD138, MUM1/IRF4, Blimp1, and XBP1 expression ( Blood Lymphat Cancer PMID 31360094). There is no established standard of care ( Blood PMID 25636338). Due to the disease's rarity, limited information is available on PbL, especially when focusing on underrepresented minorities. This study aims to describe the demographics, treatment patterns, and survival outcome differences in Hispanic (HI) versus non-Hispanic (NH) patients in the United States with PbL. Methods: Data was analyzed on patients with PbL in the US reported to the NCDB between 2010 and 2019. Sociodemographic and treatment characteristics were compared between ethnic groups. Kaplan-Meier and Cox regression analyses were used to compare overall survival (OS) between HI and NH populations. Multivariate analysis and propensity score matching were performed with adjustment for age, stage, co-morbidity score, insurance status, type of facility and great circle distance. Results: Between 2010 and 2019, 1624 patients were diagnosed with PbL with 16% identified as HI. HI were diagnosed at a younger median age of 48 years as compared to NH at 58 years. Diagnosis was made predominately in males (78%) with no statistical difference between HI and NH. Most HI and NH were white (92% and 74% in HI and NH, respectively) with increased racial diversity among NH. 17% of all patients identified as black with 1% HI and 21% NH. Regarding income, 32% HI had an income less than $38,000 versus 23% NH; whereas, 31% NH had an income greater than $63,000 versus 20% HI [p=0.017]. 38% HI had a Charlson-Deyo Comorbidity Index score ≥ 2 as compared to 27% NH [p&amp;lt;0.001]. Table 1 provides additional information on demographics. 45% of all patients had stage IV disease at the time of diagnosis. 86% of all patients received treatment with no statistical difference among HI and NH [p=0.246]. 67% of HI were treated at academic/research programs compared to 50% of NH [p=0.004]. The median distance between the patient's residence and treatment facility was 6.9 miles for HI versus 9.8 miles for NH [p=0.001]. The survival probability at 2, 5, and 10 years for HI was 59%, 52%, and 42%, respectively as compared to NH was 47%, 37%, and 31%, respectively. The median survival time (MS) was 9.4 years in HI versus 1.4 years in NH [p &amp;lt;0.001]. Figure 1 shows the Kaplan Meier curve comparing overall survival (OS) between HI versus NH.When adjusted on multivariate analysis, there were no independent variables associated with better or worse OS; moreover, propensity matched analysis showed no MS difference between HI vs NH (1.79 years versus 1.17 years). Conclusion: This hospital-based analysis noted that HI presented with PbL at a younger age than NH; however, HI more frequently had a higher Charlson-Deyo Comorbidity Index score than NH. There was no statistical difference in the stage at diagnosis between HI and NH. Additionally, HI had a lower income and no medical insurance as compared to NH suggesting HI possibly had inferior access to medical resources. Paradoxically, HI had significantly higher OS rates. Further insight into the disease's intrinsic characteristics, as well as biological factors, is needed to gain a better understanding of these poor outcomes.

Timing to Surgery and Lymph Node Upstaging in Gastric Cancer: An NCDB Analysis.

Prior studies have shown tumor specificity on the impact of longer time interval from diagnosis to surgery, however in gastric cancer (GC) this remains unclear. We aimed to determine if a longer time interval from diagnosis to surgery had an impact on lymph node (LN) upstaging and overall survival (OS) outcomes among patients with clinically node negative (cN0) GC. Patients diagnosed with cN0 GC undergoing surgery between 2004-2018 were identified in the National Cancer Database (NCDB) and divided into intervals between time of diagnosis and surgery [short interval (SI): ≥ 4 days to < 8 weeks and long interval (LI): ≥ 8 weeks]. Multivariable regression analysis evaluated the independent impact of surgical timing on LN upstaging and a Cox proportional hazards analysis and Kaplan-Meier curves evaluated survival outcomes. Of 1824 patients with cN0 GC, 71.8% had a SI to surgery and 28.1% had a LI to surgery. LN upstaging was seen more often in the SI group when compared to LI group (82% versus 76%, p = 0.004). LI to surgery showed to be an independent factor protective against LN upstaging [adjusted odds ratio = 0.62, 95% CI: (0.39-0.99)]. Multivariate Cox regression analysis indicated that time to surgery was not associated with a difference in overall survival [hazard ratio (HR) = 0.91, 95% CI: (0.71-1.17)], however uncontrolled Kaplan-Meier curves showed OS difference between the SI and LI to surgery groups (p = 0.037). Timing to surgery was not a predictor of LN upstaging or overall survival, suggesting that additional medical optimization in preparation for surgery and careful preoperative staging may be appropriate in patients with node negative early stage GC without affecting outcomes.

Nationwide Analysis of Locoregional Management for Ductal Carcinoma In Situ in Males: An NCDB Analysis of the Surgical Approach to DCIS in Males.

Limited data exist regarding the optimal locoregional approach for males with ductal carcinoma in situ (DCIS). This study examined trends in management and survival for males with DCIS. The National Cancer Database (NCDB) was queried for males with a diagnosis of DCIS from 2006 to 2017. Patients were categorized by locoregional management. Continuous variables were evaluated by Kruskal-Wallis and categorical variables by chi-square or Fisher's exact test. Univariable and multivariable logistic regressions were performed to evaluate for predictors of patients receiving partial mastectomy (PM) with radiation. Survival was analyzed by Kaplan-Meier. Between 2006 and 2017, 711 males with DCIS were identified. Most received mastectomy alone (57.1%). No change was observed in management approach from 2006 to 2017. Patients who underwent mastectomy alone were mostly hormone-positive (95.9% were estrogen-positive, 90.9% were progesterone-positive), although this cohort was least likely to receive hormone therapy (17.2%). Among those who underwent PM with radiation, only 61% of those who were hormone-positive received hormone therapy. Univariable analysis demonstrated that those of black race had lower odds of receiving PM with radiation (odds ratio [OR], 0.58; 95% confidence interval [CI], 0.36-0.84), which persisted in the multivariable analysis with control for age and tumor size (OR, 0.32; 95% CI, 0.15-0.67). Overall survival did not differ significantly between the four treatment methods (p = 0.08). The management approach to male DCIS did not change from 2006 to 2017. Survival did not differ between treatment methods. Demographic and clinicopathologic features, including race, may influence locoregional treatments received, and further studies are needed to further understand this.