Shadyside Hospital

Abstract Introduction/Objective The BRAF VE1 immunohistochemical (IHC) stain can recognize BRAF V600E-mutant proteins and coincidently highlights cilia due to cross-reactivity of the antibody with epitopes in axonemal dyneins. This phenomenon has not been extensively studied. Our aim was to therefore investigate the utility of BRAF IHC staining in pulmonary tissue. Methods/Case Report Archival cases where BRAF IHC was performed on pulmonary tissue were retrieved. Clinical and pathological findings were recorded and IHC slides reviewed. Results BRAF IHC was available for 49 thoracic specimens including 26 biopsies, 22 surgical resections and 1 cytopathology cell block case. BRAF staining highlighted cilia at the apical surface of bronchial cells in 39 cases (80%). Cytoplasmic or nuclear staining was considered abnormal, with artifactual staining seen in 18 cases (e.g. histiocytes, mucin). Unexpected, non-apical staining in 10 slides occurred with adenocarcinoma (3 of 14 adenocarcinomas in the cohort), thoracic sialadenoma papilliferum, atypical adenomatous hyperplasia, benign cartilage, and vessels. BRAF failed to stain two entirely benign cases (one with scar tissue). Conclusion BRAF IHC is an underutilized ancillary tool in thoracic pathology. By highlighting the presence of cilia, BRAF has potential in most cases to distinguish benign from malignant pulmonary tissue.

Continuous epidural analgesia is desirable for improving infant outcomes after surgeries. However, its contribution to facilitating extubation is not well known. A retrospective chart review was conducted at the UPMC Children's Hospital of Pittsburgh to identify all infants who received an epidural catheter between 2018 and 2024 and required postsurgical admission to the Neonatal Intensive Care Unit (NICU). The study examined the timing of extubation and reintubation, along with associated factors, in 100 infants who underwent major surgeries. In total, 100 infants, 43 females and 57 males, 40 (38.39-42.07) weeks corrected gestational age, 3 (2.52-3.42) kg received epidural catheters. Sixty-two patients had a pulmonary condition. Of 45 infants extubated in the operating room, 32 received fentanyl intraoperatively, and 16 required a morphine infusion in the NICU. Among 55 infants that remained intubated, 24% underwent a thoracic procedure, 46 received intraoperatively fentanyl, and 21 needed an opioid infusion postoperatively. The extubation day was median (IQR) 2 (1-4), and 24% remained intubated beyond day 5. Twelve infants were intubated preoperatively, and six required prolonged ventilation beyond day 5. Of 15 infants that required reintubation, 8 received a morphine infusion. The medians (IQR) of the average of three pain and sedation scores before reintubation were 1.67 (1-3) and 0 (-1.67-0), respectively. Epidural analgesia may facilitate early extubation in some infants undergoing surgeries. Morphine infusion was administered at a similar rate between infants extubated and those who remained intubated, and its role in delaying extubation timing remains unclear.

The Achilles tendon is the strongest and largest tendon in the human body, composed of the distal confluence of the gastrocnemius and soleus muscles and may also include the plantaris longus tendons. Despite its tremendous strength, it is frequently injured. Soft tissue tumors or tumor-like conditions of Achilles tendon are much less common in comparison to traumatic and infectious/inflammatory lesions. These include giant cell tumor of the tendon sheath, pigmented villonodular synovitis, xanthomas, hematomas, and tophaceous gout. We present a case of a myxochondroid mass arising from the Achilles tendon in a 17-year-old female softball player detected after a minor trauma. According to our knowledge, this is the first case reported in the literature detailing the radiological findings and histopathological diagnosis of a myxochondroid mass arising from a tendon.

Superficial anaplastic lymphoma kinase (ALK)-rearranged myxoid spindle cell neoplasm (SAMS) is a recently described entity which coexpresses ALK, CD34, and commonly S100. These neoplasms are characterized morphologically by concentric spindle cell whorls and cords and are commonly set in an abundant myxoid to myxocollagenous stroma, thus mimicking perineurioma or hybrid nerve sheath tumor. EMA immunostain has been reported to be negative in SAMS which helps in excluding the latter entities. Herein, we report the first EMA-positive SAMS of the right leg in a 37-year-old female patient masquerading as perineurioma/hybrid nerve sheath tumor. The tumor morphologically was comprised of spindle cells arranged in loose whorls and short fascicles set in myxoid to collagenous stroma and coexpressed CD34 and EMA, reminiscent of perineurioma. S100 showed focal staining. ALK immunostain was subsequently performed and was positive. ALK gene rearrangement was identified by fluorescence in situ hybridization break-apart assay and was further confirmed by next-generation sequencing-based RNA sequencing demonstrating FLNA::ALK fusion, thus supporting the diagnosis of SAMS. In conclusion, EMA can be expressed in SAMS, thus posing as a diagnostic pitfall. ALK immunostain and molecular studies are essential for confirming the diagnosis of SAMS and excluding potential mimickers, particularly perineurioma or hybrid nerve sheath tumor.

There is an anecdotal impression that teenage patients report exaggerated postoperative pain scores that do not correlate with their actual level of pain. Nurse and parental perception of teenagers’ pain can be complemented by knowledge of patient pain behavior, catastrophizing thoughts about pain, anxiety, and mood level. Two hundred and two patients completed the study—56.4% were female, 89.6% White, 5.4% Black, and 5% were of other races. Patient ages ranged from 11 to 17 years (mean = 13.8; SD = 1.9). The patient, the parent, and the nurse completed multiple questionnaires on day one after laparoscopic surgery to assess patient pain. Teenagers and parents (r = 0.56) have a high level of agreement, and teenagers and nurses (r = 0.47) have a moderate level of agreement on pain scores (p < 0.05). The correlation between patient APBQ (adolescent pain behavior questionnaire) and teenager VAS (visual analog scale) and between nurse APBQ and teenager VAS, while statistically significant (p < 0.05), is weaker (r range = 0.14–0.17). There is a moderate correlation between teenagers’ pain scores and their psychological assessments of anxiety, catastrophic thoughts, and mood (r range = 0.26–0.39; p < 0.05). A multi-modal evaluation of postoperative pain can be more informative than only assessing self-reported pain scores.

Abstract ObjectiveThis study evaluates an AI assistant developed using OpenAI’s GPT-4 for interpreting pharmacogenomic (PGx) testing results, aiming to improve decision-making and knowledge sharing in clinical genetics, and to enhance patient care with equitable access.MethodsThe AI assistant employs Retrieval Augmented Generation (RAG) combining retrieval and generative techniques. It employs a Knowledge Base (KB) comprising Clinical Pharmacogenetics Implementation Consortium (CPIC) data, with context-aware GPT-4 generating tailored responses to user queries from this KB, refined through prompt engineering and guardrails.ResultsEvaluated against a specialized PGx question catalog, the AI assistant showed high efficacy in addressing user queries. Compared with OpenAI’s ChatGPT 3.5, it demonstrated better performance, especially in provider-specific queries requiring specialized data and citations. Key areas for improvement include enhancing accuracy, relevancy, and representative language in responses.DiscussionThe integration of context-aware GPT-4 with RAG significantly enhanced the AI assistant’s utility. RAG’s ability to incorporate domain-specific CPIC data, including recent literature, proved beneficial. Challenges persist, such as the need for specialized genetic/PGx models to improve accuracy and relevancy and addressing ethical, regulatory, and safety concerns.ConclusionThis study underscores generative AI’s potential for transforming healthcare provider support and patient accessibility to complex pharmacogenomic information. While careful implementation of large language models like GPT-4 is necessary, it is clear that they can substantially improve understanding of pharmacogenomic data. With further development, these tools could augment healthcare expertise, provider productivity, and the delivery of equitable, patient-centered healthcare services.