Data
- A sample of the MultiClinSum dataset is now available on Zenodo!
Clinical case reports are a type of textual genre in the field of medicine that describe a patient’s medical history, symptoms, diagnosis, and treatment in detail. They are usually written by healthcare providers, such as physicians, nurses, or other medical professionals, and are used to document and share information about a specific patient’s condition. These reports, which are often published in peer-reviewed medical journals, are an important source of information in the field of medicine and are used to contribute to the advancement of medical knowledge and to improve patient care. They are also an important textual source for Natural Language Processing in the clinical domain, as they provide a rich source of medical information in unstructured text format that is similar to real hospital records. In some cases, these reports are published together with a summary of the text provided by the article’s authors.
The MultiClinSum dataset consists of clinical case reports from various specialties extracted from open journals, paired with their human-generated summary. Together with the summarization dataset, a set of background documents will be provided. In addition, participants are allowed to use any other data source available online as long as they report it.
The link to the MultiClinSum dataset, as well as a selected list of additional resources, will be posted here soon.
Dataset Samples
These are some examples of the full-text and summary document pairs in each language contained in the dataset:
MultiClinSum-en
Full Text
A 60-year-old male patient complained of a painless mass under the nipple for more than four months. No pain, itching or nipple discharge were noted. The patient had no history of cancer and no specific family history but felt that the tumor was growing. Further immunohistochemistry investigations were recommended because a diagnosis of an invasive breast carcinoma does not exclude breast polymorphic invasive lobular carcinoma or cancer with syngenetic differentiation. The results of these investigations showed that ER, PR, HER2, E-cadherin, P120, CK5/6, and P40 were all negative, although a high expression of Ki-67 was noted. The results obtained from the local hospital that initially treated the patient showed that these immunohistochemistry results were not specific to the liquid obtained from the biopsy, with histological analysis revealing an aggressive cancer morphology. Following intradepartmental consultation, the patient was diagnosed with a highly aggressive breast cancer consistent with a triple-negative breast lesion. The patient decided to receive a new adjuvant treatment, although his relatives complained and asked for further treatment options as there were no other obvious symptoms due to the patient’s preexisting health conditions. This led to the patient seeking treatment at our hospital. An ultrasound examination carried out on admission showed that the larger gland was located below the nipple and measured 4.4 cm x 4.4 cm x 1.4 cm. Striped blood flow signals were observed in the region, with low echoes seen in the shallow fat layer measuring 1.1 cm x 0.5 cm. The integrated ultrasound diagnosis based on the mammography reports and data systems (BI-RADS) was a type 4C right breast parenchymal occupying lesion.
Only a small number of tumor tissue biopsy samples were obtained for histology. Light microscopy showed a greater number of tumor cells and reduced stroma. The tumor cells were linear or trabecular, with some areas being monoline. The local lesions harbored a mixture of cells, with the tumor cells being large and containing fine chromatin and multiple mitotic images. The tumor cells exhibited deep staining and were disproportionate in size. Necrosis and bleeding were not observed. The results of the immunohistochemistry investigations were as follows: AE1/AE3 (-), EMA (+), S-100 (-), CD68 (-), CD3 (weak+), CD20 (-), Pax5 (-), CD30 (+), Bcl-2 (-), CD2l (-), CD23 (-), Bcl-6 (-), MUM1 (+), CD2 (-), CD4 (+), CD8 (-), CD5 (-), CD7 (-), TIA-1 (-), GrB (+), ALK (+), CD56 (-), EBER (-), and Ki-67 (70%+). Based on the above results the pathologic diagnosis was non-Hodgkin’s anaplastic large cell lymphoma in the right breast, ALK-positive.
Interphase fluorescence in situ hybridization (FISH) analyses were then performed using the two-color fusion probes IRF4 and DUPS22, obtained from Ambipycin Pharmaceuticals Co., Ltd., Guangzhou, China. Where possible, at least 100 nuclei were evaluated for each probe. The FISH analyses were evaluated and recorded using ISIS Digital Image Analysis version 5.0 (MetaSystems, Altusheim, Australia). To analyze the whole genome for imbalance, DNA was extracted from the FFPE material using the QIAmp DNA FFPE tissue kit (Qiagen, Lab Division, America) and processed using the Oncoscan™ FFPE express 3.0 kit (Affymetrix, Santa Clara, CA, USA). Processing involved analysis of copy number aberrations (CNA) and constructing copy TuScan algorithms using Nexus Express for Oncoscan 3 software (Beijing Yuxin Biotechnology Co., China). The human reference genome GRCh37/hg19 was used for gains and deletions smaller than 100 Kb or containing less than 20 probes, and CNN-LOH for those smaller than 5000 Kb or containing deletion regions not considered in the analyses. Molecular cytogenetic analysis using interphase fluorescence in situ hybridization (FISH) analysis showed that the vast majority of cells in the tissue sections had chromosomal breakpoints that affected the fusion of the IRF4 motif and DUPS22, and chromosomal breakpoints that affected the DUPS22 motif. We also detected an additional signal for the nonrecombinant allele, suggesting a gain in the DUPS22 locus. In addition, we observed a gain in the IRF4 locus without breaks. Based on these results and with reference to the current classification of lymphoma of the World Health Organization (WHO), we changed the diagnosis to “non-Hodgkin’s mesenchymal metaplastic large cell lymphoma, ALK-positive”. This malignancy was not associated with an IRF4/DUPS22 rearrangement. Chromosomal imbalance mapping using the Oncoscan™ Chromosomal Imbalance Mapping platform revealed mutations in the following genes: BCOR_p.Q600X, DNMT3A_p.F609fs, NOTCH1_p.P2320fs, and IDH2_p.R140Q. Finally, we attempted to sequence the whole exome of the tumor, although unfortunately, this failed due to technical reasons.
Summary
A biopsy of a breast mass in a 60-year-old Caucasian man showed a morphologic-immunophenotypic profile with features characteristic of an ALK-positive (AKT+), anaplastic large cell lymphoma. Fluorescence in situ hybridization (FISH) analysis of fixed, paraffin-embedded tissue of this lesion was performed at our institution for IRF4/DUSP22 gene rearrangement. No rearrangement was detected. The patient presented with mutations in the following genes; BCOR_p.Q600X, DNMT3A_p.F609fs, NOTCH1_p.P2320fs, and IDH2_p.R140Q. However, the patient's consultation was complicated by the fact that he had been diagnosed with breast cancer at a local hospital and had come to our institution for further consultation. The histology findings were confirmed by immunohistochemistry and FISH. Computed tomography and positron emission tomography did not reveal nodules elsewhere in the body, which allowed the staging of the patient to be completed. However, although the patient had previously received the chemotherapy CCOP regimen (ie, cyclophosphamide, vincristine, prednisolone acetate) he did not go into remission in a timely manner and relapsed after six months, followed by a drastic deterioration in his condition after four months, resulting in his death in less than one month.
MultiClinSum-es
Full Text
Presentamos el caso del manejo de un onfalocele gigante, en un paciente con diagnóstico prenatal en semana 14. Tras el diagnóstico inicial se descartaron anomalías cromosómicas. Desde la consulta de diagnóstico prenatal del servicio de Ginecología y Obstetricia se contactó con el servicio de Cirugía pediátrica para evaluación prenatal y se consideró la infiltración prenatal de toxina botulínica. En semana 37 se programó infiltración de toxina botulínica en el plano transverso del abdomen (TAP) del feto. Este proceso lo realizó un equipo multidisciplinar formado por ginecólogos, cirujanos pediatras, neonatólogos y anestesistas, con la gestante bajo sedación. Se realizó sedación fetal con una inyección intramuscular de atropina, fentanilo y rocuronio. Dada la posición del feto intraútero únicamente se pudo infiltrar el plano transverso del abdomen del lado derecho del feto. Mediante una única infiltración placentaria se infiltró toxina botulínica (dosis 12 UI/kg) a nivel subcostal, 10ª costilla y fosa ilíaca derecha. Durante el proceso y en los días siguientes no se evidenciaron complicaciones. Se programó la cesárea en semana 39. Al nacimiento se colocó silo con bolsa de plástico sobre el defecto y una vez el paciente estuvo relajado e intubado se infiltró toxina botulínica en el lado izquierdo y se realizó el test de relajación, consiguiendo reducción total del contenido eviscerado e inversión del amnios el primer día de vida. La presión intraabdominal (PIA) estuvo monitorizada en todo momento mediante sondaje vesical y mediante control del flujo renal por sistema INVOS. Se colocó un apósito hidrocoloide (Varihesive Gel ControlTM) sobre el defecto, que se cambiaría cada 48 horas. El tercer día de vida se inició alimentación trófica y el cuarto se procedió a la extubación del paciente sin observar incidencias, pudiendo iniciar lactancia materna. El octavo día de vida se programó para cierre definitivo de pared y se realizó técnica de separación de componentes (Técnica de Ramírez)(7), mediante la que se incide la línea semilunar para separar músculo recto del músculo oblícuo externo. Al finalizar la intervención se comprobó presión intraabdominal de 10 y buen flujo renal. El paciente permaneció intubado las primeras 24 horas postoperatorias. Inició tolerancia el primer día postoperatorio, siendo la evolución favorable procediendo al alta domiciliaria el 14 día de vida.
Summary
Presentamos el caso de un hepatoonfalocele con diagnóstico prenatal, sin anomalías asociadas. En semana 37 bajo sedación materna y fetal se realizó inyección de toxina botulínica en el hemiabdomen derecho. Tras la cesárea programada en semana 39 se completó inyección en el hemiabdomen izquierdo y se logró reducción completa del contenido hepático e inversión total del amnios. Se realizó reparación definitiva con separación de componentes el octavo día de vida y alta el decimocuarto día de vida.
MultiClinSum-fr
Full Text
Patiente de 32 ans, cultivatrice, en provenance de la Guinée, admise dans le service le 8 juin 2022 pour fièvre et conscience altérée. Sa symptomatologie serait d'installation progressive depuis dix jours environ, précédée d'une fièvre permanente, de céphalées et de vomissements. Elle est apparue après une révision utérine dans une structure sanitaire guinéenne le 25 mai 2022 pour avortement spontané d'une grossesse de 18 semaines d'aménorrhée. Elle était traitée avant son admission pour méningite bactérienne et infection génitale suspecte par ceftriaxone et métronidazole injectables. Elle n'avait aucun antécédent médico-chirurgical connu, ni de notion d'usage d'immunosuppresseurs. L'examen physique de la patiente a objectivé un indice de masse corporelle (IMC) à 20,9 kg/m2 pour un poids de 61 kg, une fièvre à 39,2°C, un score SOFA (Sequential Organ Failure Assessment) à 3 (1 pour la pression artérielle moyenne à 65 mmHg et 2 pour le score de Glasgow à 11/15), des convulsions tonicocloniques généralisées, une raideur méningée, une hémiplégie droite. Il existait une infection génitale : lochies purulentes, utérus pelvien non retracté.
La tomodensitométrie crânio-cérébrale a mis en évidence une large plage d'hypodensité hémisphérique gauche, rehaussée par le produit de contraste avec important effet de masse. Au bilan biologique, on constatait une hypoglycémie capillaire à 3,2 mmol/l, une clairance de la créatinine (selon la formule CKDEPI) à 155,32 ml/mn pour une créatininémie à 38,5 µmol/l, des transaminases ALAT à 58 Ul/l, et un groupe sanguin O positif. L'examen du liquide cérébro-spinal (LCS) montrait un liquide purulent, une hyperleucocytorachie à 1 100/mm3 à prédominance neutrophile, une hypoglycorachie à 1,2 mmol/l, une hyperprotéinorachie à 1,8 g/l, des bacilles gram négatif E. coli résistants aux pénicillines (amoxicilline, ampicilline, amoxiacide clavulanique), aux carboxypénicillines (ticarcilline), aux céphalosporines (ceftriaxone, cefixime, cefalotine, cefepime), aux quinolones (ciprofloxacine, norfloxacine), aux cyclines (tétracycline) et au cotrimoxazole.
L'hémoculture sur milieu aérobie et l'examen du prélèvement vaginal ont isolé la même souche. L'hémogramme a mis en évidence une anémie sévère (à 4,2 g/dl d'hémoglobine), normocytaire (VGM à 85,2 fl), normochrome (CCMH à 32,7 g/dl), régénérative (réticulocytes à 160 g/l), des leucocytes à 8 000/mm3 (neutrophiles à 6 200/mm3, lymphocytes à 1 000/mm3) et des plaquettes à 14 000/µl. Deux sérologies VIH se sont révélées négatives, et le taux de CD4 était à 850 cellules/mm3 de sang. Le diagnostic de sepsis post abortum, compliqué de méningoencéphalite à E. coli multirésistant et d'anémie sévère a alors été retenu.
Dès l'admission, elle avait été mise sous amoxiacide-clavulanique en raison de 2 g toutes les huit heures et amikacine 1 g/jour en intraveineuse. Au 4e jour, à la suite des résultats de l'antibiogramme, le traitement par méropenem a été instauré à raison de 2 g toutes les huit heures en intraveineuse pendant 21 jours. L'anémie sévère a été corrigée par la transfusion journalière de 450 ml de concentré érythrocytaire isogroupe-isorhésus pendant six jours. L'hypoglycémie a été corrigée par une perfusion de 500 ml de sérum glucosé 10 %, les convulsions par l'administration de diazépam. Par ailleurs, des toilettes génitales biquotidiennes à la chlorhexidine pendant cinq jours et de la kinésithérapie au 8e jour du traitement par méropenem ont été réalisées. L’évolution a été favorable, marquée par l'apyrexie au 4e jour de la bi-antibiothérapie, l'amendement des convulsions, et la reprise de la conscience respectivement au 2e et au 7e jour du traitement par méropenem. L'hémoglobine de contrôle est revenue à 11,9 g/ dl au 7e jour de la transfusion sanguine. La force motrice des membres droits déficitaires a été mesurée à 1/5 et 3/5 respectivement au 7e et au 4e jour de la kinésithérapie. L'exéat a été autorisé le 5 juillet 2022, avec kinésithérapie en ambulatoire et un rendez-vous quatorze jours après. À ce rendez-vous, l’évolution était globalement favorable avec une force motrice des membres initialement déficitaires à 5/5.
Summary
La première, qui avait avorté récemment, était sous traitement par ceftriaxone quand elle a été admise pour syndrome méningé et syndrome pyramidal lié à une lésion cérébrale. Du liquide céphalorachidien (LCR), du sang et d'un prélèvement de sécrétions génitales purulentes a été isolé E. coli résistant aux pénicillines et céphalosporines. Elle a guéri sous traitement par méropénem pendant 21 jours, avec peu de séquelles.
MultiClinSum-pt
Full Text
Mulher de 38 anos com história de neurite óptica do lado esquerdo diagnosticada aos 30 anos, com síndrome clinicamente isolada, obteve recuperação completa da deficiência visual com tratamento com esteroides. Desde então, tomava regularmente cápsulas de B. serrata (200mg/dia, dose recomendada pelo fabricante) para fortalecer seu sistema imunológico. Ela desenvolveu hipersensibilidade à luz, dor ocular, náusea, tontura e fraqueza nos membros inferiores 4 dias após tomar a primeira dose da vacina BNT162b2 em julho de 2021 e aumentou a dose de B. serrata para cinco cápsulas (1.000mg/dia) 1 semana após a vacinação. Após tomar B. serrata na dose de 1.000mg/dia por 3 semanas, a paciente sofreu uma convulsão tônico-clônica generalizada não provocada e foi internada na unidade de terapia intensiva. A investigação revelou hiponatremia sérica (112mmol/L [n, 135 - 150mmol/L]), concentração de sódio urinário de 58mmol/L, osmolaridade sérica de 234mosm/kg (n, 280 - 300mosm/kg), osmolaridade urinária de 739mosm/kg (n, 450 - 600mosm/kg), concentração de ACTH de 85,9pg/mL (n, 7,2 - 63,3pg/mL), concentração normal de cortisol basal, concentração normal de proteína C-reativa, contagem de leucócitos de 11,4 (n, < 10/l), neutrofilia, linfopenia e rabdomiólise (concentração máxima de creatina quinase de 76348U/L (n, 1 - 145U/L). Realizou ressonância magnética do crânio (RMC), que revelou três lesões periventriculares, não realçadas, que não mudaram em número e nem em extensão em comparação com a RMC feita 4 anos antes. A glândula pituitária estava normal. A investigação de malignidade foi pouco esclarecedora. A paciente foi diagnosticada com SIHAD e tratada com levetiracetam, diurese forçada e infusões de cloreto de sódio. Após 3 semanas de tratamento e interrupção do uso de cápsulas de B. serrata, a paciente se recuperou completamente.
Summary
Uma mulher de 38 anos diagnosticada com síndrome clinicamente isolada tomava regularmente cápsulas de B. serrata (200mg/dia) para fortalecer seu sistema imunológico. Ela desenvolveu hipersensibilidade à luz, dor ocular, náusea, tontura e fraqueza nos membros inferiores 4 dias após tomar a primeira dose da vacina BNT162b2 e aumentou a dose de B. serrata para 1.000mg/dia 1 semana após a vacinação. Após tomar B. serrata na dose de 1.000mg/dia por 3 semanas, ela foi internada na unidade de terapia intensiva devido à convulsão tônico-clônica generalizada não provocada. A investigação diagnóstica revelou síndrome da secreção inapropriada de hormônio antidiurético, que se resolveu completamente após tratamento e interrupção do uso de B. serrata.