Impact of expert review on 26 117 referral classified diagnoses according to sampling procedures
| Diagnostic changes . | CNB (n = 8026) . | Surgical excision (n = 18 091) . | P . |
|---|---|---|---|
| Discordant cases (n = 2339) | 752 (9.3%) | 1587 (8.7%) | .12 |
| Major changes | 646 (8.0%) | 1371 (7.6%) | .40 |
| Misclassifications of lymphoma subtypes | 472 (5.9%) | 1130 (6.2%) | .26 |
| B-NHL vs PTCL∗ | 29 (0.3%) | 76 (0.4%) | |
| cHL vs B-NHL† | 39 (0.5%) | 96 (0.5%) | |
| cHL vs PTCL‡ | 34 (0.4%) | 90 (0.5%) | |
| NLPHL vs other lymphoma subtypes§ | 32 (0.4%) | 160 (0.8%) | |
| “High-grade” vs “low-grade” B-NHLǁ | 153 (1.9%) | 344 (1.9%) | |
| Misclassifications in “low-grade” B-NHL with therapeutic impact¶ | 119 (1.5%) | 260 (1.4%) | |
| Misclassifications in “high-grade” B-NHL with therapeutic impact# | 65 (0.7%) | 96 (0.5%) | |
| Misclassifications in classified PCTL with therapeutic impact∗∗ | 1 (0.01%) | 8 (0.04%) | |
| Misclassifications between lymphomas and other neoplasms | 28 (0.3%) | 50 (0.3%) | .32 |
| From classified lymphomas to nonlymphoid neoplasms | 15 (0.2%) | 26 (0.2%) | |
| From nonlymphoid neoplasms to classified lymphomas | 13 (0.1%) | 24 (0.1%) | |
| Misclassifications between malignant and reactive lesions | 146 (1.8%) | 191 (1%) | <.00001 |
| From classified lymphomas or nonlymphoid neoplasms to reactive lesions | 89 (1.1%) | 90 (0.5%) | |
| From reactive lesions to classified lymphomas or nonlymphoid neoplasms | 57 (0.7%) | 101 (0.5%) | |
| Minor changes | 106 (1.3%) | 216 (1.2%) | .39 |
| From DLBCL subtypes to other DLBCL subtypes without therapeutic impact†† | 56 (0.7%) | 57 (0.3%) | |
| DLBCL vs grade 3B FL | 22 (0.3%) | 51 (0.3%) | |
| From PTCL to other PCTL subtypes without therapeutic impacts‡‡ | 28 (0.4%) | 108 (0.6%) | |
| Concordant cases (n = 23 778) | 7274 (90.7%) | 16 504 (91.3%) | .12 |
| Diagnostic changes . | CNB (n = 8026) . | Surgical excision (n = 18 091) . | P . |
|---|---|---|---|
| Discordant cases (n = 2339) | 752 (9.3%) | 1587 (8.7%) | .12 |
| Major changes | 646 (8.0%) | 1371 (7.6%) | .40 |
| Misclassifications of lymphoma subtypes | 472 (5.9%) | 1130 (6.2%) | .26 |
| B-NHL vs PTCL∗ | 29 (0.3%) | 76 (0.4%) | |
| cHL vs B-NHL† | 39 (0.5%) | 96 (0.5%) | |
| cHL vs PTCL‡ | 34 (0.4%) | 90 (0.5%) | |
| NLPHL vs other lymphoma subtypes§ | 32 (0.4%) | 160 (0.8%) | |
| “High-grade” vs “low-grade” B-NHLǁ | 153 (1.9%) | 344 (1.9%) | |
| Misclassifications in “low-grade” B-NHL with therapeutic impact¶ | 119 (1.5%) | 260 (1.4%) | |
| Misclassifications in “high-grade” B-NHL with therapeutic impact# | 65 (0.7%) | 96 (0.5%) | |
| Misclassifications in classified PCTL with therapeutic impact∗∗ | 1 (0.01%) | 8 (0.04%) | |
| Misclassifications between lymphomas and other neoplasms | 28 (0.3%) | 50 (0.3%) | .32 |
| From classified lymphomas to nonlymphoid neoplasms | 15 (0.2%) | 26 (0.2%) | |
| From nonlymphoid neoplasms to classified lymphomas | 13 (0.1%) | 24 (0.1%) | |
| Misclassifications between malignant and reactive lesions | 146 (1.8%) | 191 (1%) | <.00001 |
| From classified lymphomas or nonlymphoid neoplasms to reactive lesions | 89 (1.1%) | 90 (0.5%) | |
| From reactive lesions to classified lymphomas or nonlymphoid neoplasms | 57 (0.7%) | 101 (0.5%) | |
| Minor changes | 106 (1.3%) | 216 (1.2%) | .39 |
| From DLBCL subtypes to other DLBCL subtypes without therapeutic impact†† | 56 (0.7%) | 57 (0.3%) | |
| DLBCL vs grade 3B FL | 22 (0.3%) | 51 (0.3%) | |
| From PTCL to other PCTL subtypes without therapeutic impacts‡‡ | 28 (0.4%) | 108 (0.6%) | |
| Concordant cases (n = 23 778) | 7274 (90.7%) | 16 504 (91.3%) | .12 |
The 5021 cases submitted by referral pathologists with no definitive diagnosis were excluded. cHL, classical HL; FL, follicular lymphoma.
Twelve CNB and 42 surgical excision cases initially referred as classified B-NHL were modified to classified PTCL, and 17 CNB and 34 surgical excision cases referred as classified PTCL were modified to classified B-NHL.
Twenty-five CNB and 76 surgical excision cases initially diagnosed as cHL were modified to classified B-NHL, 14 CNB and 20 surgical excision cases initially referred as classified B-NHL were modified to cHL.
Thirteen CNB and 64 surgical excision cases initially diagnosed as cHL were modified to classified PTCL, 21 CNB and 26 surgical excision cases initially referred as classified PTCL were modified to cHL.
Fourteen CNB and 59 surgical excision cases initially diagnosed as NLPHL were modified to other classified lymphomas, 18 CNB and 101 surgical excision cases classified as lymphomas other than NLPHL were modified to NLPHL.
Eighty CNB cases referred as “high-grade” B-NHL (including 53 DLBCL NOS, 4 T-cell/histiocyte-rich B-cell lymphomas [TCRBCL], and 23 grade 3B FL) were modified to 4 chronic lymphocytic leukemias (CLL), 65 FL, 3 lymphoplasmacytic lymphomas/Waldenström macroglobulinemia (LPL/WD), 2 mantle cell lymphomas (MCL), and 6 NMZLs. A total of 204 surgical excision cases referred as “high-grade” B-NHL were modified to 31 CLL; 111 grades 1, 2, and 3A FL; 1 LPL/WD; 35 MCL; and 26 NMZL. Seventy-three CNB referred as “low-grade” B-NHL (including 3 CLL; 47 grades 1, 2, and 3A FL; 1 LPL/WD; 17 MCL; and 5 NMZL) were modified to classified “high-grade” B-NHL (including 55 DLBCL NOS, 1 TCRBCL, 1 Burkitt lymphoma [BL], 2 high-grade B-cell lymphoma with MYC and BCL2 and/or BCL6 rearrangements [HGBL], and 14 grade 3B FL). A total of 140 surgical excision referred as “low-grade” B-NHL (including 6 CLL; 105 grades 1, 2, and 3A FL; 6 LPL; 20 MCL; and 3 NMZL) were modified to classified “high-grade” B-NHL (107 DLBCL NOS, 2 HGBL, and 31 grade 3B FL with surgical excision).
A total of 119 CNB (including 34 CLL; 45 grades 1, 2, and 3A FL; 6 LPL/WD; 10 MCL; and 24 NMZL) and 260 surgical excision (45 CLL; 92 grades 1, 2, and 3A FL; 21 LPL/WD; 48 MCL; and 54 NMZL) were misclassifications between “low-grade” B-NHLs with a potential therapeutic impact.
Sixty-five CNB and 96 surgical excision cases were misclassifications between DLBCL NOS and HGBL (40 CNB and 66 surgical excision), misclassifications between BL and HGBL (14 CNB and 22 surgical excision), and misclassifications between BL and HGBL (11 CNB and 8 surgical excision cases).
One CNB case and 8 surgical excision cases were misclassifications between anaplastic large cell lymphomas (ALCL) ALK+ or ALK– and other PTCL.
Fifty-six CNB and 57 surgical excision cases were DLBCL subtype misclassifications without potential impact on patients’ care and included changes between DLBCL NOS, TCRBCL, DLBCL Epstein-Barr virus positive, and DLBCL ALK+.
Twenty-eight CNB (6 AITL, 12 PTCL NOS, and 10 ALCL ALK–) and 108 surgical excision (44 AITL, 44 PTCL, and 20 ALCL ALK–) were PTCL subtype misclassifications without potential impact on patients’ care.