United States

Title: Accuracy Benchmark Highlights AI&#39;s Potential to Transform Medical Education and Curricula Development

Background: The impressive capabilities of modern large language models (LLMs) like ChatGPT have recently been demonstrated in the context of the United States Medical Licensing Examination (USMLE). The National Board of Medical Examiners (NBME) released a study showing that these models can pass actual USMLE exams, with numerous studies showcasing improvements with successive models. However, a comprehensive analysis of these models&#39;, including ChatGPT 4 Omni, performance across specific medical content domains remains absent, limiting our understanding of their full potential in medical education. Furthermore, establishing a benchmark for accuracy is crucial to underline the reliability of these models in providing accurate and dependable information.

Methods: Utilizing 750 clinical vignette-based multiple-choice questions (MCQs) provided by medical schools, the performance of successive iterations of the leading LLM, ChatGPT—namely, ChatGPT 3.5 (GPT-3.5), ChatGPT 4 (GPT-4), and ChatGPT 4 Omni (GPT-4o)—were evaluated across various USMLE disciplines, clinical clerkships, and clinical skills such as diagnostics and management. Accuracy assessments were conducted using a standardized protocol, with rigorous statistical analyses employed to compare and analyze the reliability of these models in responding to the MCQs.

Results: GPT-4o achieved a remarkable correct response rate of 90.4%, significantly outperforming both GPT-4 (81.1%), GPT-3.5&#39;s (60.0%), and the medical student average of 59.3% (95% CI: 58.3%-68.3%). In preclinical disciplines, GPT-4o performed best in Social Sciences (95.5%), Microbiology (92.3%), and Immunology (92.9%). GPT-4&#39;s highest accuracy was in Behavioral and Neuroscience (86.5%), and GPT-3.5&#39;s highest was in Behavioral and Neuroscience (76.9%). In Clinical Clerkship categories, GPT-4o attained perfect scores in Family Medicine (100%) and Internal Medicine (100%), while GPT-4&#39;s highest was in Internal Medicine (95.5%), and GPT-3.5&#39;s highest was in Neurology (69.5%).

Conclusion: This study highlights the exceptional performance of the ChatGPT series across USMLE preclinical and clinical disciplines, showcasing the transformative potential of AI in medical education. The advancements seen in ChatGPT 4 Omni underscore its utility as an educational tool, suggesting a future where AI enhances learning and competency in medical fields. This study benchmarks groundbreaking accuracy, ushering in newfound reliability in the accuracy of AI. As AI integration in education grows, our findings emphasize the necessity for a formal curriculum to guide proper usage, ensuring oversight and validation to maximize benefits while maintaining educational standards.


AMA Research Challenge 2024 

Accuracy Benchmark Highlights AI&#39;s Potential to Transform Medical Education and Curricula Development

Title: Accuracy Benchmark Highlights AI's Potential to Transform Medical Education and Curricula Development

Background: The impressive capabilities of modern large language models (LLMs) like ChatGPT have recently been demonstrated in the context of the United States Medical Licensing Examination (USMLE). The National Board of Medical Examiners (NBME) released a study showing that these models can pass actual USMLE exams, with numerous studies showcasing improvements with successive models. However, a comprehensive analysis of these models', including ChatGPT 4 Omni, performance across specific medical content domains remains absent, limiting our understanding of their full potential in medical education. Furthermore, establishing a benchmark for accuracy is crucial to underline the reliability of these models in providing accurate and dependable information.

Methods: Utilizing 750 clinical vignette-based multiple-choice questions (MCQs) provided by medical schools, the performance of successive iterations of the leading LLM, ChatGPT—namely, ChatGPT 3.5 (GPT-3.5), ChatGPT 4 (GPT-4), and ChatGPT 4 Omni (GPT-4o)—were evaluated across various USMLE disciplines, clinical clerkships, and clinical skills such as diagnostics and management. Accuracy assessments were conducted using a standardized protocol, with rigorous statistical analyses employed to compare and analyze the reliability of these models in responding to the MCQs.

Results: GPT-4o achieved a remarkable correct response rate of 90.4%, significantly outperforming both GPT-4 (81.1%), GPT-3.5's (60.0%), and the medical student average of 59.3% (95% CI: 58.3%-68.3%). In preclinical disciplines, GPT-4o performed best in Social Sciences (95.5%), Microbiology (92.3%), and Immunology (92.9%). GPT-4's highest accuracy was in Behavioral and Neuroscience (86.5%), and GPT-3.5's highest was in Behavioral and Neuroscience (76.9%). In Clinical Clerkship categories, GPT-4o attained perfect scores in Family Medicine (100%) and Internal Medicine (100%), while GPT-4's highest was in Internal Medicine (95.5%), and GPT-3.5's highest was in Neurology (69.5%).

Conclusion: This study highlights the exceptional performance of the ChatGPT series across USMLE preclinical and clinical disciplines, showcasing the transformative potential of AI in medical education. The advancements seen in ChatGPT 4 Omni underscore its utility as an educational tool, suggesting a future where AI enhances learning and competency in medical fields. This study benchmarks groundbreaking accuracy, ushering in newfound reliability in the accuracy of AI. As AI integration in education grows, our findings emphasize the necessity for a formal curriculum to guide proper usage, ensuring oversight and validation to maximize benefits while maintaining educational standards.


Accuracy Benchmark Highlights AI's Potential to Transform Medical Education and Curricula Development

poster

Welcome to the **largest national, multi-specialty research** event featuring research from more than 750 medical students, residents, fellows, and international medical graduates.

Use the buttons below or navigation menu to access the event.
 

[![](https://assets.underline.io/markdown_image/1/image/20e1e77aa954efeac7a15d84c85eb383.jpg)](https://underline.io/events/468/sessions?searchGroup=all)
 
[![](https://assets.underline.io/markdown_image/1/image/068522d81aca0b68a369787c7dfccd1b.jpg)](https://underline.io/events/468/posters?searchGroup=all)
 
[![](https://assets.underline.io/markdown_image/1/image/d3971ece79d5e5575b178bdc2c8bbfae.jpg)](https://underline.io/events/468/help-desk)
 

#### Event Details

The 2024 American Medical Association Research Challenge includes a poster symposium hall and a semifinals hall with the top scored research. **Log in at any time throughout the event to:**

* **View posters and presentations** on a variety of topics and specialties
* **Score posters in the semifinals** to help decide the top five finalists (AMA member exclusive opportunity)
* **Connect with participants** and provide feedback to help advance their research

At the conclusion of this event, the AMA will announce the top five finalists who will advance to the AMA Research Challenge finals, present their research to an elite panel of judges, and compete to win a $10,000 grand prize presented by Laurel Road. 

### Grand Prize Sponsor

AMA would like to thank Laurel Road for sponsoring a $10,000 grand prize for the winner of the AMA Research Challenge. Laurel Road is not involved in winner selection process. The winner will be announced on February 20, 2025 during the AMA Research Challenge finals. Payment of the grand prize will be directly from Laurel Road. See the AMA Research Challenge Official Rules for terms and conditions.

[![](https://assets.underline.io/uploads/markdown_image/1/image/611e331d4a0eba0920f240cfc04b4d77.png)](https://www.laurelroad.com/)

**There is no registration fee for this event! **

If you have used the AMA Research Challenge platform previously, please click the ‘Log-in’ button below. If you do not remember your password, please click “forgot password” and follow prompts to reset your password. After resetting your password, please refresh your webpage to access the platform.

Register to explore research in a variety of topics. AMA members get exclusive access to score posters in the semifinals. 

**Please use this link [here](https://www.ama-assn.org/member-benefits/events/ama-research-challenge) to register for free.**

Registration or Log-in Is Required

Join the AMA Research Challenge poster symposium and semifinals to explore top research posters and score posters in the semifinals (an AMA member exclusive opportunity).

INTRODUCTION 

American football is renowned for being one of the most contact-intensive sports, frequently resulting in head injuries, with over 40% of retired National Football League (NFL) players showing evidence of long-term sequelae. Contrastingly, epidemiological data on head injuries in rugby, a sport with shared origins and similarities to American football, is lagging. Rugby has grown significantly in the United States over the past few decades. Despite the level of contact rivaling that of football, players wear little to no protective gear. Given the rising popularity of rugby in American sports, it is crucial to examine the rates and mechanisms of injury to better protect these high-performance athletes. 

METHODS 

The National Electronic Injury Surveillance System (NEISS) was queried for injuries related to playing rugby while using body part codes for head injuries between 2004-2023. NEISS offers estimates of national injuries reported to emergency departments across the United States. Narratives were screened for relevance, specific diagnosis, and mechanisms of injury. Non-cranial head injuries were excluded from the narratives, and nonspecific narratives only indicating a ‘head injury’ or ‘head trauma’ were only included if a loss of consciousness was noted. Weighted survey analysis determined incidence rates and 95% confidence intervals. Weighted binomial regression was used to analyze relationships between age and other binary variables. 

RESULTS

A total of 910 narratives produced a national estimate of 29,250 emergency department (ED) visits for traumatic brain injuries (TBI) in the United States from playing rugby between 2004-2023. Of all the estimated cases of TBI, 63.3% were concussions (Figure 1), of which 17.6% were accompanied by a loss of consciousness. The majority of those who sustained a TBI were male (70%), and the average age was 20 years (Figure 2). Male rugby players with TBI were older on average than females (P = 0.04). A total of 565 narratives (national estimate: 17,781) indicated the mechanism of injury, with tackles (27.4%) resulting in more injuries than any other observed mechanism.

The prevalence of head injuries among rugby players is a pressing issue, compounded by the lack of comprehensive research on injury frequency and on-field signs like motor incoordination. These injuries pose immediate and long-term risks, particularly for younger, inexperienced athletes who may have poorer tackling form. Addressing these challenges requires ongoing research, education, and improved injury prevention strategies, such as tailored protective equipment and training in safe play practices. Investing in these measures enhances player safety and supports the long-term health and performance of young athletes, highlighting the need for further studies into rugby players' neurologic health.



Traumatic Brain Injuries Amongst Rugby Players in the US


Authors: Hamid George Bangura, BS; Chelsea Willie, MD; Jasmine Dowell, MD

Background: Multiple studies have shown that Social Determinants of Health (SDoH) are associated with disparities in health outcomes in children and adults admitted to intensive care units. Studies have also shown that adults disproportionately suffered from severe illnesses with COVID-19. This study investigates patient-level characteristics to determine associations between SDoH and Pediatric Intensive Care Unit (PICU) admission patterns for critically ill children with COVID-19.

Methods: Retrospective analysis of acute COVID-19-related admissions to the Children’s Wisconsin (CW) PICU between January 1, 2020, and March 31, 2022, from any Emergency Department (Direct Admission) or the CW Acute Care Ward (Indirect Admission). Exclusion criteria included patients for whom initial admission decisions were made outside the CW context. The study was deemed non-human subjects research and exempted from full IRB application. All data were retrieved from the Virtual Pediatric Systems database. Multivariable statistical analysis was completed with biostatistical support from the Clinical and Translational Science Institute of Southern Wisconsin.

Results: 215 cases of COVID-19 hospitalization met inclusion criteria for analysis. The racial composition of this cohort was 39.4% Non-Hispanic White (NHW), 36.2% Non-Hispanic Black (NHB), and 19.2% Hispanic. Neighborhood Stress Score (NSS7), a measure of socioeconomic deprivation, was the primary influencer of direct admission to the PICU after adjusting for race, sex, and severity of illness (OR 1.02, 95% CI [1.02, 1.22]). Secondly, indirectly admitted NHB patients had significantly longer than expected PICU stays compared to indirectly admitted NHW patients (p=0.049).

Conclusion: Socioeconomically deprived children critically ill with COVID-19 were more likely to be directly admitted to the PICU. NHB children admitted to the General Pediatric ward prior to transferring to the PICU required longer hospitalization in the PICU than predicted for their level of illness. Taken together, these findings indicate SDoH influenced admission patterns and length of stay outcomes for children with COVID-19. Our findings are in line with current research
indicating NHBs are disproportionately affected by health disparities that influence health outcomes. Our research highlights why consideration of social determinants in healthcare is important when evaluating healthcare and disease, especially for centers that provide care to minority or socioeconomically disadvantaged patients.

Social Determinants of Health and Pediatric Intensive Care Unit Admission Patterns in Children with COVID-19

Background
The residency application period involves high levels of uncertainty, stress, and significant financial burden for trainees. Recent national workgroup recommendations have encouraged the development of professional standards for communication with applicants. Here, the authors explored the common but under-reported phenomenon of students submitting applications to residency programs but never receiving a response, termed “application ghosting.”

Methods
A cross-sectional survey with multiple-choice and open-ended questions was shared with recent medical graduates across a national medical student listserv from March-May 2023. Ghosting was defined as the proportion of applications submitted by students that never received a response. Personal and academic factors, as well as preference signaling, were then examined for their potential contributions to ghosting. Thematic analysis was also conducted on open-ended responses to explore the emotional impact and other qualitative outcomes of application ghosting.

Results
6,006 applications to 21 specialties were submitted by 115 applicants. Applicants submitted an average of 52.2±34.8 applications. Amongst all applications, 2,900 (48.3%) were ghosted. Individuals who identified as men (p=0.004), from disadvantaged backgrounds (p=0.011), from lower class quartiles (p=0.015), or applying to the most competitive specialties (p<0.001) experienced higher ghosting rates. Preference signaling and geographical ties to the program’s region reduced ghosting. Open-ended responses highlighted feelings of uncertainty, false hope, professionalism double standards, and hidden expectations associated with application ghosting.

Conclusion
Application ghosting has a negative impact on trainees and contributes to uncertainty within the residency selection process. To mitigate these undesired effects, residency programs should adopt standardized communication practices.

Exploring the impact of application ghosting in the residency selection process

Background/Introduction: 
Formal hand-off process is essential for safe patient transition of care. Structured hand-off at our institution was implemented using the IPASS module. In addition to the in-person sign-out process, a shared spreadsheet was implemented to the existing GME admission tracker document in-order to augment and reinforce the follow-up by the residents on night-float for “watcher” and “unstable” patients. 

Methods: 
Review of the transition of care and types of hand-offs along with the IPASS module were presented as a lecture during the orientation week of the class of 2026. Full participation of all GME internal medicine residents was required for this lecture. Verbal sign-out was reinforced through faculty periodic monitoring of residents during day-to-night sign out. Multiple e-mail reminders were distributed by the chief residents. Evaluation of deficiencies were discussed as part of townhall meetings. Assessment of Pre/Post implementation was done through the ACGME survey as well as a post-implementation survey which was distributed to PGY2/PGY3 residents. Descriptive statistics were used to evaluate the success of the implementation. 

Results: 
ACGME annual survey 2023/2024 (pre/post IPASS implementation) responses to “information not lost during shift changes, patient transfers, or the hand-over process” revealed (65%, 3.7/5.0; N=69/72) and (77%, 3.9/5.0; N=70/71). Survey responses to PGY2 and PGY3 revealed a mean rating of 3.61/5.0. 

Discussion: 
The transition of care process was successfully improved through the implementation of the IPASS module to structure sign out at our program. Furthermore, application of a shared spreadsheet was useful in the follow-up for watcher and unstable patients by residents covering night-float. This year, expansion of the QI initiative will implement the IPASS as part of the bootcamp for the incoming residents, the lecture reviewing the transition of care process will also be included in this years orientation series.

IPASS Module Integration to Resident Tracker– An ACGME Survey directed QI initiative​

Background: Resistant hypertension (RHTN) is defined as consistent blood pressure above 130/80 mmHg despite being on three or more antihypertensive medications. RHTN carries higher cardiovascular morbidity and mortality than those with hypertension (HTN), even when adjusted for blood pressure.
RHTN is usually associated with increased arterial stiffness (AS), a direct consequence of chronic inflammation. An inflammatory response develops from the innate upregulation of several proinflammatory mediators (PiMs) in response to stressors, followed by resolution. The resolution process is complex, but specialized pro-resolving mediators (SPMs) play an active role. It has been shown that in obesity, an inefficient SPM response and elevated PiMs lead to a persistent low-grade proinflammatory state. This causes dysregulation in the inflammation-resolution process, leading to impaired vascular repair and increased AS, subsequently predisposing to RHTN.
Herein, we assess the role of PiMs and SPMs in the development of AS among non-diabetic patients with obesity-related RHTN.

Methods: Patients seen in the University of Alabama at Birmingham Hypertension Referral Clinic underwent AS measurement and analysis of PiMs and SPMs. AS was measured via carotid-femoral pulse wave velocity (SphygmoCor). PiMs and SPMs were measured by targeted lipidomics analysis by LC/MS in combination with quantification of the lipid mediators. Plasma from 15 individuals was included in the final results. Healthy normotensive patients served as control.

Results: We analyzed 40 bioactive lipid mediators as potential PiMs and SPMs contributing to AS. In this cohort of 15 individuals, the mean age of the hypertensive group was 57.5±3.5 years, BMI was 34.85±2.7 kg/m2, blood pressure was 146.39±8.7/88.9±5.8 mmHg, and AS was 10.7±1.9 m/s (p =0.0031). LC/MS analysis showed PiMs such as prostaglandin F1(PGF1)-alpha was significantly higher among the obesity-related HTN group with increased AS compared to the control group (p=0.01), and PGF2-alpha was also higher when compared to the control group (p=0.06). SPMs such as Resolvin-D1 did not show a significant difference between the two groups, indicating a possible complex dysequilibrium in the resolution of chronic inflammation in the vasculature.

Conclusion: Our study shows that PiMs play a significant role in AS in obesity-related RHTN, and dysregulation in the SPM pathway leads to continued low-grade inflammation in the vasculature. Further studies with a larger patient population can help us target the inflammation-resolution pathway for novel therapies to reduce AS and subsequently prevent RHTN and cardiovascular disease.

When Good Inflammation Turns Bad: Role of Bioactive Lipids in Obesity, Chronic Inflammation and Resistant Hypertension

Abstract Title
Large Saccular Intracranial Aneurysm in a Child with RASA1-Associated Capillary Malformation AVM Syndrome

Background
Large cerebral aneurysms are much less common in children than in adults. Thus, when they present, these lesions require careful surgical evaluation and comprehensive genetic testing. RASA1-associated capillary malformation-arteriovenous malformation (RASA1-CM-AVM) syndrome is a rare disorder of angiogenic remodeling known to cause port-wine stains and arteriovenous fistulas but not previously associated with pediatric aneurysms.

Case Presentation
The authors report a case of a previously healthy 6-year-old boy who presented with seizure-like activity. Imaging demonstrated a lesion in the right ambient cistern with compression of the temporal lobe. Imaging characteristics were suggestive of a thrombosed aneurysm versus an epidermoid cyst. The patient underwent craniotomy, revealing a large saccular aneurysm, and clip ligation and excision were performed. The patient recovered well from surgery with no postoperative complications or neurological deficits. Postoperative histology was consistent with a saccular cerebral aneurysm. Postoperative genetic analysis revealed a RASA1-CM-AVM syndrome. 

Discussion
This represents the first documented case of a RASA1-associated pediatric cerebral aneurysm in the neurosurgical literature. This unique case highlights the need for maintaining a broad differential diagnosis as well as the utility of genetic testing for detecting underlying genetic syndromes in young children presenting with cerebral aneurysms.

Large Saccular Intracranial Aneurysm in a Child with RASA1-Associated Capillary Malformation AVM Syndrome

Background
Congenital heart disease presents unique challenges in pregnant patients due to the physiological changes of pregnancy. This case discusses a pregnant patient with incomplete D- loop miso/dextrocardia with situs solitus and Shone's complex who presented after a cardiac arrest.

Case Presentation
A 33-year-old pregnant female with a past medical history of congenital heart disease incomplete D-loop, meso/dextrocardia with situs solitus, and Shone's complex), paroxysmal atrial fibrillation, and schizophrenia was brought in following PEA cardiac arrest with ROSC achieved in the field by EMS personnel. Cardiac workup was largely benign, with no signs of ischemia on EKG and initially elevated but quickly downtrending troponin levels. Transthoracic echocardiogram was stable from baseline and showed a reduced left ventricular ejection fraction of 25-30% with global hypokinesis of the left ventricle, bioprosthetic aortic valve, congenital transposition of the great vessels, and perimembranous VSD. ICU course was complicated by refractory status epilepticus secondary to anoxic brain injury, necessitating a Cesarean section at 34 weeks, to allow for the introduction of additional anti-epileptic drugs (AEDs). After a lengthy hospital stay of 2 months, she was discharged to an LTAC.

Discussion
The management of adult individuals with congenital cardiac abnormalities is quite difficult as they are quite rare, and the variant anatomy may respond atypically to classical treatments. This case was complicated by pregnancy and the number of congenital cardiac conditions which are discussed next. An incomplete d-loop which refers to how the normal inflow portion of the right ventricle lies to the right of the left ventricle but in this patient, only partially does. Dextrocardia is the normal position of the heart but this patient instead partially suffered from a midline heart, or mesocardia. To qualify for Shone’s syndrome, at least 3 of 8 criteria must be satisfied: this patient had a bicuspid aortic valve, hypoplastic left ventricle, small aortic arch, and supramitral ring. While no clear etiology for the cardiac arrest was determined, it was presumed that heavy alcohol use with pregnancy placed too much of a strain on heart with these many abnormalities. Pregnancy also played a role in management of seizures as certain AEDs are extremely teratogenic and initially limited the 4 AED regiment. The case underscores the importance of a multidisciplinary approach involving numerous specialties and tailored management plans for patients with extremely complicated medical histories.

Cardiac Arrest in a Pregnant Patient with Congenital Heart Disease and Shone’s Complex

Background: With improving technology, children are surviving with increasingly complex medical conditions, and most are living at home. Given the shortage of home-based clinicians for children with medical complexity (CMC), parents often provide nearly 24/7 support with few if any breaks, resulting in poor long-term mental and physical health outcomes. Studies have found that respite – supports and resources that provide direct care so parents can rest – can foster greater resilience and longevity for those parents. This qualitative study aimed to identify how parents of CMC find respite and recharge. We anticipated that key themes would include a lack of opportunities for rest and recharge, including inadequate home care.

Methods: This multi-site, cross-sectional qualitative study included semi-structured interviews with parents of CMC receiving care from the pediatric palliative care team or the pediatric complex care team at two academic children’s hospitals in Boston, Massachusetts and with home-based clinicians (HBC) working in those children’s households. Interviews explored 4 domains: work of care, rest and sleep, recharge, and engagement with the existing respite care system, including home care. Qualitative data was analyzed using rapid qualitative analysis and mixed inductive-deductive thematic analysis. Data were analyzed by an interprofessional research team with expertise in pediatric palliative care, pediatric complex care, and qualitative methods.

Results: Twenty parents of 19 children and 8 HBC participated. Parents confirmed the incessant work of care and the lack of respite, with three themes related to barriers to home care: (1) overall inaccessibility of HBC, (2) mismatch between families’ needs (e.g., number of hours’ coverage, time of day, intensity of care tasks) and clinician availability, and (3) burden on parents to train home-based clinicians on the unique needs of their child. HBC reported overwhelming case-loads and inadequate incentivization for HBC. Both parents and HBC advocated for improved resource navigation to match children and their parents with appropriate home care providers.

Conclusion: Parents of CMC lack opportunities for rest and recharge. In concept, home health care can be a key tool for providing effective respite services, yet parents and home-based clinicians report barriers to clinicians providing effective respite in the home. The findings from this qualitative study should inform improved resource development as well as reform of pediatric home care policy, including Medicaid and CHIP.

Respite Support for Family Caregivers of Children with Medical Complexity: Preliminary Findings

The Dermatology Medical Literacy and Advocacy (DMLA) program aims to advance diversity within dermatology by engaging underrepresented undergraduate students and equipping them with essential knowledge and skills. Through a blend of lectures, hands-on workshops, and discussions on health disparities, DMLA fosters a more inclusive pipeline into the field while preparing participants to address the needs of underserved communities. The program emphasizes not only clinical competence but also the importance of community engagement, empowering students to share their knowledge and advocate for dermatological health within their communities. 
This study analyzes pre- and post-course survey data to evaluate improvements in participants’ knowledge of dermatological conditions, confidence in clinical skills, and understanding of health disparities. Results show significant gains in these areas, reflecting the effectiveness of the program’s structure. While progress in community engagement was modest, the program’s evolving focus on outreach aims to better prepare participants for future roles as advocates. These findings highlight the DMLA program’s role in promoting both diversity and health equity, with future iterations focused on expanding outreach opportunities and enhancing participant readiness for community-based advocacy.


Dermatology Medical Literacy and Advocacy Program: Fostering Diversity in Dermatology while Empowering Communities


Background
Infection with human immunodeficiency virus (HIV) increases risk for developing cancer.1 However, clinical presentations of malignancies may be subtle. Clinical suspicion for cancer in patients living with HIV (PLWH) should remain high. 

Case Presentation
A 50-year-old male with history of HIV with undetectable viral load on ART and unilateral orchiectomy for testicular cancer presented to the emergency department for evaluation of left eye blurriness, ptosis, and ophthalmoplegia. History was notable for emergency department and outpatient ophthalmology visits one month prior resulting in a diagnosis of cranial nerve six palsy. Magnetic resonance angiography (MRA) did not reveal any acute findings. He developed difficulty opening his left eye and reported having weight loss over the past month. 

On exam, vital signs were within normal limits. Labs notable for uric acid 8.0 mg/dL and WBC 9.34 K/uL. Repeat orbital imaging showed infiltrative soft tissue throughout the cavernous sinuses. Magnetic resonance imaging of the brain and positron emission tomography revealed widely distributed lesions and lymphadenopathy throughout the brain, chest, abdomen, pelvis, and skeleton suspicious for malignancy. Bone marrow and lymph node biopsy revealed B cell lymphoma, CD10+/CD5- lambda monotypic. Cytogenetic analysis revealed IGH::MYC rearrangement. Pathology confirmed diagnosis of Epstein-Barr virus (EBV) negative Burkitt lymphoma (BL). Patient began chemotherapy, EPOCH-R, and follow up 6 months later shows continued response to therapies. 

Discussion
HIV is a known carcinogen for eleven different types of cancer, including so-called “AIDS-defining cancers” (ADC), and non-AIDS defining cancers (NADC).1 A 2017 study including 448,258 HIV-infected individuals showed PLWH have a standardized incidence ratio of 1.69 for cancer compared to the general population (ADC 14.0, NADC 1.21).2 Introduction of ART has decreased the burden caused by ADC, but has increased the burden caused by NADC.3,4 Specifically, Burkitt lymphoma and Hodgkin lymphoma have been noted in the literature to occur paradoxically in patients with higher CD4 counts (e.g. >250 cell/mcL).3,5 Our patient presented with an undetectable viral load, CD4 count of 776 four days prior, and an incomplete earlier infectious workup. His ultimate diagnosis of EBV-negative BL highlights the importance of investigating cancer etiologies in the absence of positive serology. Patients with HIV are at increased risk of malignancy due to alterations in their immune system. Evaluation of atypical illness should include early imaging and biopsy. Preventive actions include thorough routine history and physical (including skin and mucosal exams), optimization of HIV treatment, and comprehensive vaccination (HBV, HPV, VZV).
References

1. Engels EA, Shiels MS, Barnabas RV, et al. State of the science and future directions for research on HIV and cancer: Summary of a joint workshop sponsored by IARC and NCI. Int J Cancer. 2024;154(4):596-606. doi:10.1002/ijc.34727
2. Hernández-Ramírez RU, Shiels MS, Dubrow R, Engels EA. Cancer risk in HIV-infected people in the USA from 1996 to 2012: a population-based, registry-linkage study. Lancet HIV. 2017;4(11):e495-e504. doi:10.1016/S2352-3018(17)30125-X
3. Rubinstein PG, Aboulafia DM, Zloza A. Malignancies in HIV/AIDS: from epidemiology to therapeutic challenges. AIDS. 2014;28(4):453-465. doi:10.1097/QAD.0000000000000071
4. Ceccarelli M, Venanzi Rullo E, Marino MA, et al. Non-AIDS defining cancers: a comprehensive update on diagnosis and management. Eur Rev Med Pharmacol Sci. 2020;24(7):3849-3875. doi:10.26355/eurrev_202004_20852
5. Guech-Ongey M, Simard EP, Anderson WF, et al. AIDS-related Burkitt lymphoma in the United States: what do age and CD4 lymphocyte patterns tell us about etiology and/or biology?. Blood. 2010;116(25):5600-5604. doi:10.1182/blood-2010-03-275917

Premium content

Accuracy Benchmark Highlights AI's Potential to Transform Medical Education and Curricula Development

Next from AMA Research Challenge 2024

Traumatic Brain Injuries Amongst Rugby Players in the US

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES