Contributions
Abstract: PB2331
Type: Publication Only
Background
The diagnosis of NHL is based on histological findings from biopsies, but clinicians often perform fine-needle aspirations (FNA). This rapid and cost-effective procedure provides material for both cytology (Cy) and flow cytometry immunophenotyping (FCI) studies.
Aims
To determine the accuracy of the combined use of Cy and FCI in samples obtained by FNA to discriminate between reactive/benign processes (RP) and neoplasia (N). For this purpose, FNA report was correlated with clinical follow-up and histology from patients submitted to excisional biopsies to achieve the definitive diagnosis (Image 1).
Methods
From March 2010 to December 2016, 380 FNA specimens from 357 patients were obtained from lymphoid tissues (n=219), transbronchial aspirations (n=95) and extranodal sites (n=66). Median follow-up from the time of FNA study was 35.5 months (0-95.4). Conventional cytological techniques were applied. For FCI, a stabilisating reagent was added to fresh samples (≥16 hours) before using a standard stain-lyse-wash protocol. For screening, mAbs against CD45, CD3, CD4, CD8, CD19, CD20, CD22, CD56, kappa and lambda were used. Additional reagents were only included if the screening test showed data of malignancy. Results from Cy&FCI were considered as concordant when they matched together with their discrimination between RP and N, further classified as lymphoma, carcinoma, other hematological (OHN), and other neoplasms (ON).
Results
Twenty-two samples were discarded because loss of patient´s follow-up (n=7), or necrosis/insufficient material for study (n=15). In 30 of the remaining 358 samples, material was adequate for either Cy or FCI, and 328 samples (91.6%) were properly evaluated by both methods. A concordant result between Cy&FCI correlated with the definitive diagnosis in 69.3% of samples, mostly RP (80.8%) (Image 1). Within RP, 3 double Cy&FCI false positive results were reported (1.3%); in contrast, 16/68 solid tumors (23.5%) were misdiagnosed by both techniques. Regarding the hematological diseases, an exact identification was done by Cy&FCI in 81.4% of all NHL (n=35) and 71.4% of OHN. Interestingly, 33 out of 35 NHL (94.3%) were new diagnosis, and FCI data were concordant with the definitive histological subclassification in 26/33 NHL (78.8%). In contrast, both Cy&FCI reported a false negative result in 1 Hodgkin lymphoma (HL) and 3 NHL (6.98%). Thirteen samples (3.6%) had discrepant Cy&FCI results: Cy correctly identified 2 carcinoma, 2 NHL and 3 RP; FCI correctly identified 3 low-grade NHL and 3 RP. The sensitivity, specificity, and positive and negative predictive values of concordant Cy&FCI results were 81.2%, 98.4%, 95.4% and 92.9% respectively for all type of malignancies as compared to 90.9%, 98.9%, 93% and 98.5% for hematological neoplasms only.
Conclusion
Adequate material for both Cy&FCI studies was obtained in most samples of our series. Their combined use didn´t preclude discrepant results, but the high correlation between a concordant negative Cy&FCI result and RP may avoid unnecessary biopsies. Regarding neoplasms, there was a much better correlation between the results of Cy&FCI in hematological diseases than in carcinomas. The diagnosis of HL relied on Cy only, but the feasibility of FCI for grading NHL allows a prompt start-up of extension studies until achievement of definitive diagnosis with a second diagnostic procedure.
Session topic: 19. Non-Hodgkin lymphoma Biology & Translational Research
Keyword(s): Diagnosis, flow cytometry, Non-Hodgkin's lymphoma
Abstract: PB2331
Type: Publication Only
Background
The diagnosis of NHL is based on histological findings from biopsies, but clinicians often perform fine-needle aspirations (FNA). This rapid and cost-effective procedure provides material for both cytology (Cy) and flow cytometry immunophenotyping (FCI) studies.
Aims
To determine the accuracy of the combined use of Cy and FCI in samples obtained by FNA to discriminate between reactive/benign processes (RP) and neoplasia (N). For this purpose, FNA report was correlated with clinical follow-up and histology from patients submitted to excisional biopsies to achieve the definitive diagnosis (Image 1).
Methods
From March 2010 to December 2016, 380 FNA specimens from 357 patients were obtained from lymphoid tissues (n=219), transbronchial aspirations (n=95) and extranodal sites (n=66). Median follow-up from the time of FNA study was 35.5 months (0-95.4). Conventional cytological techniques were applied. For FCI, a stabilisating reagent was added to fresh samples (≥16 hours) before using a standard stain-lyse-wash protocol. For screening, mAbs against CD45, CD3, CD4, CD8, CD19, CD20, CD22, CD56, kappa and lambda were used. Additional reagents were only included if the screening test showed data of malignancy. Results from Cy&FCI were considered as concordant when they matched together with their discrimination between RP and N, further classified as lymphoma, carcinoma, other hematological (OHN), and other neoplasms (ON).
Results
Twenty-two samples were discarded because loss of patient´s follow-up (n=7), or necrosis/insufficient material for study (n=15). In 30 of the remaining 358 samples, material was adequate for either Cy or FCI, and 328 samples (91.6%) were properly evaluated by both methods. A concordant result between Cy&FCI correlated with the definitive diagnosis in 69.3% of samples, mostly RP (80.8%) (Image 1). Within RP, 3 double Cy&FCI false positive results were reported (1.3%); in contrast, 16/68 solid tumors (23.5%) were misdiagnosed by both techniques. Regarding the hematological diseases, an exact identification was done by Cy&FCI in 81.4% of all NHL (n=35) and 71.4% of OHN. Interestingly, 33 out of 35 NHL (94.3%) were new diagnosis, and FCI data were concordant with the definitive histological subclassification in 26/33 NHL (78.8%). In contrast, both Cy&FCI reported a false negative result in 1 Hodgkin lymphoma (HL) and 3 NHL (6.98%). Thirteen samples (3.6%) had discrepant Cy&FCI results: Cy correctly identified 2 carcinoma, 2 NHL and 3 RP; FCI correctly identified 3 low-grade NHL and 3 RP. The sensitivity, specificity, and positive and negative predictive values of concordant Cy&FCI results were 81.2%, 98.4%, 95.4% and 92.9% respectively for all type of malignancies as compared to 90.9%, 98.9%, 93% and 98.5% for hematological neoplasms only.
Conclusion
Adequate material for both Cy&FCI studies was obtained in most samples of our series. Their combined use didn´t preclude discrepant results, but the high correlation between a concordant negative Cy&FCI result and RP may avoid unnecessary biopsies. Regarding neoplasms, there was a much better correlation between the results of Cy&FCI in hematological diseases than in carcinomas. The diagnosis of HL relied on Cy only, but the feasibility of FCI for grading NHL allows a prompt start-up of extension studies until achievement of definitive diagnosis with a second diagnostic procedure.
Session topic: 19. Non-Hodgkin lymphoma Biology & Translational Research
Keyword(s): Diagnosis, flow cytometry, Non-Hodgkin's lymphoma