Use of a Tissue Microarray To Assess the Accuracy of Lymphoma Classification by Needle Biopsy.
Bibliographic record
Abstract
Abstract Appropriate management of lymphoma patients relies on accurate pathological diagnosis. Pathological criteria, such as those outlined in the WHO classification, are based on relatively large, excisional biopsies. However, pathologists are increasingly asked to render definitive diagnoses based on much smaller, needle biopsy specimens. To the best of our knowledge, the accuracy of lymphoma diagnoses based on needle biopsies relative to “gold standard” excisional biopsies has not been formally determined. In this study, the core samples in a tissue microarray (TMA) were used as needle biopsy surrogates in order to determine the accuracy of diagnoses based on small samples. The array contained randomly arranged pairs of 0.6 mm tissue cores punched from paraffin-embedded, excisional specimens. Slides from the original specimens were reviewed and these diagnoses served as a gold standard. Represented on the TMA were lymphoid hyperplasia (LH, 17 cases), classical Hodgkin lymphoma (cHL, 13 cases), marginal zone lymphoma (MZL, 11 cases), diffuse large B-cell lymphoma (DLBCL, 9 cases), grade 1 or 2 follicular lymphoma (FL, 7 cases), grade 3 follicular lymphoma (FL3, 3 cases), peripheral T-cell lymphoma (PTCL, 7 cases), mantle cell lymphoma (MCL, 6 cases), B-chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL, 6 cases), T-lymphoblastic lymphoma (TLL, 3 cases), nodular lymphocyte predominance Hodgkin lymphoma (NLPHL, 1 case) and anaplastic large cell lymphoma (ALCL, 1 case). TMA sections were stained for morphology and 16 immunomarkers used routinely in lymphoma diagnosis. Three pathologists evaluated the sections independently in a “blinded” manner. They recorded their preferred diagnoses and assigned a “confidence score”, from 0 to 3, with 3 indicating a diagnosis upon which clinical decisions might reliably be based. The results were analyzed based on 252 (i.e., 3 X 84) pathologist-case encounters. A confidence score of 3 was assigned in 118 cases (47%). Not surprisingly, diagnostic accuracy increased with increasing confidence scores: confidence 0, 0% (0/3 cases); confidence 1, 65% (35/54 cases); confidence 2, 79% (61/77 cases); and confidence 3, 92% (108/118 cases). Nine of the 10 errors made at “confidence 3” involved misclassification of lymphomas (DLBCL, ALCL, or MZL misclassified as cHL, 4 cases; MZL misclassified as CLL, 3 cases; FL misclassified as DLBCL, 1 case; and PTCL misclassified as DLBCL, 1 case); the tenth error involved misclassifying MZL as LH. Accurate, “confidence 3” diagnoses were most consistently rendered in TLL, 89% (8/9 cases); MCL, 67% (12/18 cases); cHL, 62% (24/39 cases); and, CLL, 56% (10/18 cases). Conversely, “confidence 3” diagnoses on TMA cores were always accurate in MCL (12 cases), FL, TLL, and PTCL (8 cases each). MZL was diagnostically most problematic. No case of (benign) LH was misdiagnosed as lymphoma at level 3 confidence, although 3 cases were misclassified at level 2 as DLBCL, FL or MZL. Our findings indicate that the pathological diagnosis and classification of lymphomas may be accomplished in a substantial subset of cases using tissue samples of sizes smaller than those typically obtained by needle core biopsy. Accuracy is most reliably achieved in lymphoma types associated with characteristic cytological appearances or immunophenotypic profiles.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.003 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".