In 2018 the College of American Pathologists Diagnostic Immunology and Flow Cytometry Committee designed and implemented a new plasma cell neoplasia flow cytometry proficiency testing program—PCNEO—to allow clinical flow cytometry laboratories to monitor and assess their performance compared with a peer group.
To report the results from the first 4 years of the PCNEO program.
Program participants were sent 2 sets of challenges per year, each including 1 wet challenge and 2 dry challenges, with associated clinical and laboratory findings. The wet challenges were composed of myeloma cell line specimens (with or without dilution in preserved whole blood) for flow cytometric analysis. The dry (paper) challenges were composed of clinical case summaries and images of flow cytometric test results from various flow cytometry laboratories of committee members.
A total of 116 to 145 laboratories from 17 countries enrolled in the proficiency testing program. For the wet challenges, almost all participants (97%–100%; cumulative, 98.2%) correctly identified the presence of neoplastic plasma cell populations based on flow cytometric analysis of undiluted myeloma cell lines. Slightly fewer participants (89.0%–97.4%; cumulative, 95.2%) correctly identified the presence of neoplastic plasma cell populations based on flow cytometric analysis of diluted myeloma cell lines (10% or 50% dilutions into peripheral blood) intended to better represent a typical clinical sample. There was generally agreement among 80% or more of participants for positive or negative staining for CD38, CD138, CD19, CD20, and surface and cytoplasmic κ and λ light chains. Similarly, 84% to 100% of participants were able to correctly identify the presence of neoplastic plasma cell populations in paper challenges, including the presence of small, neoplastic plasma cell populations (0.01%–5.0% clonal plasma cells) and the presence of nonneoplastic plasma cell populations (correctly identified by 91%–96% of participants).
Participant performance in the new proficiency testing program was excellent overall, with the vast majority of participants able to perform flow cytometric analysis and identify neoplastic plasma cell populations and to identify small plasma cell clones or expanded populations of reactive plasma cells in dry challenge flow cytometry results. This program will allow laboratories to verify the accuracy of their testing program and test interpretations for the assessment of patients suspected of having a plasma cell neoplasm.
The Diagnostic Immunology and Flow Cytometry Committee (DIFCC) acts as an expert scientific and educational resource for the College of American Pathologists in diagnostic immunology and flow cytometry clinical testing. Volunteer members from academic and private clinical flow cytometry laboratories oversee the proficiency testing for hundreds of flow cytometry laboratories (primarily in North America) enrolled in proficiency testing surveys. Challenges include both “wet” specimens using established neoplastic cell lines, and “dry” specimens, which are paper challenges from various flow cytometry laboratories of committee members. Proficiency testing surveys include assessment of B-cell and T-cell subsets (FL1), DNA ploidy (FL2), leukemias and lymphomas (FL3), CD34+ cells (FL4), hematopoietic neoplasia paper challenges (FL5), postimmunotherapy analysis of lymphoid neoplasia (FL6), expanded T-cell subset analysis (FL7), minimal residual disease (MRD) assessment of mature B-cell neoplasms (FL8), and MRD assessment of plasma cell neoplasms (FL9).
Plasma cell neoplasms are a heterogeneous group of diseases and include plasma cell myeloma, plasmacytoma, monoclonal gammopathy of undetermined significance, monoclonal immunoglobulin deposition diseases, and plasma cell neoplasms with an associated paraneoplastic syndrome.1 Plasma cell myeloma accounts for approximately 1% of malignant tumors and 10% to 15% of hematopoietic neoplasms.1 Diagnosis is based on clinical, laboratory, morphologic, radiologic, and immunophenotypic findings. Flow cytometry is a sensitive and rapid methodology to demonstrate the presence in a sample of neoplastic plasma cells. Neoplastic plasma cells exhibit monotypic cytoplasmic light-chain staining, lack surface immunoglobulin, express CD38 and CD138, and, in contrast to normal, nonneoplastic plasma cells, may be positive for CD20, CD56, CD200, CD28, and CD117, and typically exhibit decreased to negative expression of CD45, CD19, CD27, and CD81.1–3 Because of a lack of proficiency testing for plasma cell neoplasms and related disorders, in 2018 the DIFCC created a new proficiency testing program for plasma cell neoplasia called PCNEO. Program participants were sent 2 sets of challenges per year, A and B. Each set of challenges included a plasma cell myeloma cell line specimen, which, along with provided clinical history and corresponding cytology images, was intended to represent a patient sample. Cell line specimens were diluted in preserved whole blood starting in 2020, with the intention of better representing typical clinical samples. In addition, each set of challenges included 2 dry (paper) challenges from patients with a plasma cell neoplasm or related disease or condition, each of which consisted of a clinical scenario, pertinent laboratory test results, and representative printouts of flow cytometric test results from various flow cytometry laboratories of committee members, to provide participants the opportunity to interpret provided flow cytometry data for cases with suspected plasma cell disorders. In most instances, the flow cytometry data specifically targeted the plasma cell population.
The purpose of this study was to examine proficiency testing performance of the clinical flow cytometry laboratories that participated in the PCNEO proficiency testing program, and to identify diseases and conditions that were challenging to participant laboratories, based on the results of the new proficiency testing program.
MATERIALS AND METHODS
Case types and specifications for each mailing were chosen by the DIFCC in advance. Plasma cell lines MM1-R4 , U266,5 and RPMI 82266 were grown by a commercial laboratory and distributed undiluted or diluted in preserved whole blood to program participants following pretesting by 1 or more DIFCC committee members, typically representing the 2 most common flow cytometer manufacturers for each pretesting cycle. Committee members supplied clinical histories and photomicrographs from clinical cases to accompany the cell line samples. Dry challenges consisted of actual clinical cases from DIFCC committee members that were distributed to program participants and included clinical history, representative photomicrographs, and flow cytograms. Participant consensus on the immunophenotype of the abnormal cell population required agreement of at least 80% of participating laboratories for the individual antigens (positive versus negative), with the additional requirement that at least 20% of participating laboratories had to perform testing for any particular antigen for inclusion in consensus assessment. PCNEO is an educational program and participants are not formally graded. Data were collected and analyzed from the first 8 mailings (2018–2021) of the PCNEO program.
RESULTS
From 2018 to 2021, a total of 116 to 145 flow cytometry laboratories enrolled in the PCNEO proficiency testing program each year (116 laboratories in 2018, 125 laboratories in 2019, 136 laboratories in 2020, and 145 laboratories in 2021). Of the total of 8 surveys (2 surveys per year), 180 laboratories reported results in at least 5 surveys, and 70 laboratories participated in all 8 surveys. For laboratories that reported their geographic location, from 2018 to 2021, most participating laboratories were from the United States, 92 of 102 (90.2%) to 111 of 141 (78.7%), with the remaining laboratories from Canada, 4 of 102 (3.9%) to 7 of 141 (5.0%), and 14 additional countries, 6 of 102 (5.9%) to 23 of 141 (16.3%).
There were a total of 24 challenges in the 8 surveys, which consisted of 8 wet challenges and 16 dry challenges. For the first 4 surveys from 2018 to 2019, the wet challenges consisted of undiluted myeloma cell lines. For the second set of 4 surveys from 2020 to 2021, the wet challenges consisted of myeloma cell lines diluted with preserved whole blood. The dilutions ranged from 10% to 50%. For the wet challenges, from 2018 to 2021 laboratories employed BD FACSDiva (BD Biosciences, San Jose, California), 40 of 97 (41.2%) to 44 of 142 (31.0%); Beckman Coulter Kaluza (Beckman Coulter Life Sciences, Indianapolis, Indiana), 20 of 97 (20.6%) to 40 of 142 (28.2%); FCS Express (De Novo Software, Pasadena, California), 18 of 97 (18.6%) to 26 of 142 (18.3%); or other data analysis software. More than 95% of participants reported using a gating strategy to identify abnormal cells, with most participants employing a gating strategy that included CD38 and/or CD138. Gating strategies included using CD138, CD45, and CD38, CD38 versus CD138, CD38 versus CD45, CD45 versus side scatter, or other approaches.
From 2018 to 2021, the vast majority of participants tested for 7 antigens: CD19 (97 of 102 [95.1%] to 133 of 142 [93.7%]), CD38 (95 of 102 [93.1%] to 132 of 142 [93.0%]), CD45 (96 of 102 [94.1%] to 131 of 142 [92.3%]), CD56 (96 of 102 [94.1%] to 134 of 142 [94.4%]), CD138 (88 of 102 [86.3%] to 124 of 142 [87.3%]), cytoplasmic κ (97 of 102 [95.1%] to 130 of 142 [91.5%]), and cytoplasmic λ (97 of 102 [95.1%] to 130 of 142 [91.5%]). Most participants also tested for CD20 (79 of 102 [77.5%] to 106 of 142 [74.6%]) and CD117 (67 of 102 [65.7%] to 97 of 142 [68.3%]).
For the undiluted plasma cell myeloma cell line samples, there was generally participant consensus for positive expression of CD38, CD45, CD56, CD138, and cytoplasmic κ or λ light chain, and for absence of expression of CD19, CD20, CD117, and cytoplasmic λ or κ light chain. Two challenges had nonconsensus for CD45, 2 challenges had nonconsensus for CD56, and 1 challenge had nonconsensus for CD117. For these undiluted samples, virtually all participating laboratories agreed on the diagnosis of plasma cell myeloma (102 of 105 [97.1%] to 116 of 116 [100%]; cumulative, 431 of 439 [98.2%]). A representative example that employed an undiluted plasma cell myeloma cell line is shown in Figures 1 and 2.
Representative wet sample example that employed an undiluted plasma cell myeloma cell line. Representative bone marrow aspirate demonstrating a predominant plasma cell population (Wright-Giemsa, original magnification ×1000).
Representative wet sample example that employed an undiluted plasma cell myeloma cell line. Representative bone marrow aspirate demonstrating a predominant plasma cell population (Wright-Giemsa, original magnification ×1000).
Representative flow cytometry cytograms from the wet sample example, which demonstrate the presence of a large abnormal population that is positive for CD38 and CD138, and negative for CD19, CD20, and CD56. The intracytoplasmic κ versus λ plot demonstrates monotypic λ light chain staining.
Representative flow cytometry cytograms from the wet sample example, which demonstrate the presence of a large abnormal population that is positive for CD38 and CD138, and negative for CD19, CD20, and CD56. The intracytoplasmic κ versus λ plot demonstrates monotypic λ light chain staining.
For the diluted plasma cell myeloma cell line samples, there was generally participant consensus for positive expression of CD38, CD45, CD138, and cytoplasmic κ or λ light chain and for absence of expression of CD19, CD20, CD117, and cytoplasmic λ or κ light chain. Two challenges, consisting of a plasma cell myeloma cell line (50%) diluted with 50% preserved whole blood leukocytes, had nonconsensus for CD27, CD56, and CD81. One challenge, which consisted of a plasma cell myeloma cell line (10%) diluted with 90% preserved whole blood leukocytes, had nonconsensus for CD45, CD20, CD117, CD27, and cytoplasmic λ light chain. Participants correctly identified the presence of neoplastic plasma cell populations based on flow cytometric analysis of diluted myeloma cell lines (113 of 127 [89.0%] to 131 of 132 [99.2%]; cumulative, 355 of 373 [95.2%]). Virtually all participating laboratories agreed on the diagnosis of plasma cell myeloma when the plasma cell myeloma cell line was diluted 50:50 with preserved whole blood leukocytes (111 of 114 [97.4%] and 131 of 132 [99.2%]), and most participating laboratories agreed on the diagnosis of plasma cell myeloma when the plasma cell myeloma cell line was diluted 10:90 with preserved whole blood leukocytes (113 of 127 [86.9%]).
One of the diluted cell line challenges was intended to represent a patient with abnormal/clonal B cells without abnormal plasma cells and employed Mino cells, a mantle cell lymphoma cell line,7 diluted 50:50 with preserved whole blood leukocytes. The consensus immunophenotype of the neoplastic cells was reached for the following markers: positive for CD19, CD20, CD38, CD45, and both surface and cytoplasmic λ light chains. Negative markers included CD117, CD200, and both surface and cytoplasmic κ light chains. Consensus was not reached for CD56 and CD138; however, 102 of 129 (79.1%) and 70 of 118 (59.3%) of participants, respectively, reported absence of expression. Based on this immunophenotype and corresponding morphologic findings, 61 of 134 participants (45.5%) correctly interpreted this case as abnormal/clonal B cells present. A total of 57 of 134 participants (42.5%) chose abnormal/clonal B cells with plasmocytic differentiation; however, there was no abnormal plasma cell population in this case.
In addition to an undiluted or diluted cell line challenge, each of the 8 surveys included 2 dry (paper) challenges that were meant to represent clinical cases likely to be encountered by laboratories that perform flow cytometric analysis for plasma cell neoplasia; the cases are summarized in the Table. In general, there was excellent participant performance in the interpretation of these cases, with 119 of 141 (84.4%) to 114 of 114 (100%) participants correctly identifying the presence of a clonal plasma cell population (12 cases), including 1 case with 0.01% neoplastic plasma cells in a polyclonal background from a patient with a history of plasma cell myeloma. Three cases contained nonneoplastic plasma cell populations that were correctly identified by 93 of 102 (91.1%) to 111 of 116 (95.7%) participants. One case of lymphoplasmacytic lymphoma was correctly identified by 117 of 118 participants (99.1%).
The most challenging dry (paper) challenge was an example of a nonexpressor myeloma, shown in Figures 3 and 4. Although serum-free light chains were elevated for both κ and λ, the ratio was normal. Urine protein electrophoresis and IFIX were negative. Although the lytic bone lesions and patient demographics suggested plasma cell myeloma, the laboratory investigations failed to identify a monoclonal protein. A provided Wright-Giemsa–stained bone marrow aspirate (Figure 3) showed increased plasma cells with atypical morphology (such as binucleation and nuclear pleomorphism). The provided flow cytometric plots (Figure 4) demonstrate a relatively large abnormal cell population, which is color gated in black. This population demonstrates slightly dim CD45, positive CD38, low side scatter, heterogeneous CD20, negative-to-dim CD138, negative CD19, and heterogeneous CD56. The intracytoplasmic κ versus intracytoplasmic λ plot reveals a polytypic background plasma cell/B-cell population (in blue and pink); the population of interest had no cytoplasmic light chain expression. Additional provided information indicated that flow cytometric analysis of the plasma cell population was repeated 3 times—once with the same set of antibodies and twice with different antibody combinations (including polyclonal and monoclonal and different vendors/fluorochrome combinations), with similar results. Most participants (84 of 116 [72.4%]) reached the intended response of “Plasma cell myeloma, non-producer”; however, a substantial proportion (26 of 116 [22.4%]) indicated “Plasma cell myeloma, non-secretor.” Other submitted answers included “Reactive plasma cell hyperplasia” (3 of 116 [2.6%]), “Advise laboratory to regate primary data” (1 of 116 [0.9%]), “Normal, no abnormal B or plasma cell population” (1 of 116 [0.9%]), and “Plasma cell myeloma, not otherwise specified” (1 of 116 [0.9%]).
Dry challenge example of a nonexpressor myeloma. A bone marrow aspirate smear shows increased plasma cells with atypical morphology (such as binucleation and nuclear pleomorphism) (Wright-Giemsa, original magnification ×400).
Dry challenge example of a nonexpressor myeloma. A bone marrow aspirate smear shows increased plasma cells with atypical morphology (such as binucleation and nuclear pleomorphism) (Wright-Giemsa, original magnification ×400).
Provided flow cytometric plots demonstrate a relatively large abnormal cell population (black) that demonstrates slightly dim CD45, positive CD38, low side scatter, heterogeneous CD20, negative-to-dim CD138, negative CD19, and heterogeneous CD56. The intracytoplasmic κ versus intracytoplasmic λ plot demonstrates a polytypic background plasma cell/B-cell population (in blue and pink); the population of interest has no cytoplasmic light chain expression. Additional provided information indicated that flow cytometric analysis of the plasma cell tube was repeated 3 times with similar results—once with the same set of antibodies and twice with different antibody combinations (including polyclonal and monoclonal combinations, as well as combinations of antibodies from different vendors).
Provided flow cytometric plots demonstrate a relatively large abnormal cell population (black) that demonstrates slightly dim CD45, positive CD38, low side scatter, heterogeneous CD20, negative-to-dim CD138, negative CD19, and heterogeneous CD56. The intracytoplasmic κ versus intracytoplasmic λ plot demonstrates a polytypic background plasma cell/B-cell population (in blue and pink); the population of interest has no cytoplasmic light chain expression. Additional provided information indicated that flow cytometric analysis of the plasma cell tube was repeated 3 times with similar results—once with the same set of antibodies and twice with different antibody combinations (including polyclonal and monoclonal combinations, as well as combinations of antibodies from different vendors).
DISCUSSION
Here we report the first 4 years (2018–2021) of results from a new proficiency testing program for laboratories performing flow cytometric immunophenotypic analysis for the assessment of plasma cell neoplasia. The number of enrolled laboratories grew from 116 to 145 during the 4-year period, an overall increase of 25%. Most participants employed 1 of 3 commercially available flow cytometry data analysis software products (77.5%–80.4%) and 1 of several gating strategies to identify abnormal cell populations (98.0%–98.6%).
The vast majority of participating laboratories were able to identify the presence of a clonal plasma cell population based on analysis of undiluted plasma cell myeloma cell lines (97%–100%; cumulative, 98.2%), with slightly fewer participating laboratories able to do so when diluted cell lines were analyzed (87%–97%; cumulative, 95.2%). There was poorer overall performance when relatively smaller numbers of neoplastic plasma cells (10% dilution versus 50% dilution) were present in the specimen to be analyzed. In addition, there was decreased participant consensus for the various antigens employed for flow cytometric analysis in diluted versus undiluted samples; however, this conclusion is based on a small number of samples employed for proficiency testing in the first 4 years of the PCNEO program. In a subsequent survey in 2022, in which the same myeloma cell line was diluted (10:90) with preserved whole blood, 94% of participants correctly interpreted the case as a plasma cell neoplasm, with lack of consensus for cytoplasmic λ light chain, CD20, CD27, and CD45 (data not shown), similar to participant performance analyzing the prior cell line sample diluted 10:90 with preserved whole blood (see results above). Additional plasma cell myeloma cell line samples will be employed for future PCNEO surveys, with various dilutions in preserved whole blood, to further evaluate participant performance. There was relatively poorer participant performance in the analysis of a B-cell lymphoma cell line diluted with preserved whole blood, likely due to participant confusion about the inclusion of a B-cell lymphoma case in a plasma cell neoplasia proficiency testing program.
Participant performance in the analysis of dry (paper) challenges from various specimens was excellent (12 cases with neoplastic plasma cells and 3 cases with nonneoplastic plasma cells correctly identified by 84%–100% and 91%–96% of participants, respectively). Neoplastic challenges included cases with as few as 0.01% neoplastic plasma cells in a polytypic background (correctly identified by 97% of participants), and lymphoplasmacytic lymphoma (correctly identified by 99% of participants). The 1 exception was a case of a nonexpressor myeloma, which was correctly identified by 72% of participants, with 22% of participants identifying it as a case of nonsecretor myeloma. This is likely due to the lack of participant familiarity with these 2 rare subtypes of plasma cell myeloma, which together account for approximately 1% of plasma cell myeloma cases.1
In conclusion, participant performance in the new plasma cell neoplasia proficiency testing program was excellent, with the vast majority of participants able to perform flow cytometric analysis and identify neoplastic plasma cell populations, even when the neoplastic cell population was diluted with preserved whole blood so that it represented only 10% to 50% of total cells. In addition, participants were able to identify neoplastic and nonneoplastic plasma cell populations in dry (paper) challenges containing clinical histories, photomicrographs, and representative flow cytograms. The new program will allow laboratories to verify the accuracy of their testing program and test interpretations for the assessment of patients suspected of having a plasma cell neoplasm and to compare flow cytometry test results and overall performance with other clinical flow cytometry laboratories.
The authors acknowledge Rhona Souers, MS, College of American Pathologists, Northfield, Ill, for assistance with statistical analysis.