Modied culture system and combination of FICTION could increase the detection rate of abnormalities in multiple myeloma

of


Conclusion
The established modi ed culture system could improve karyotyping quality in MM. Owing to its obvious advantages compared with FISH, FICTION is recommended for detecting genetic abnormalities in MM.
Combining modi ed culture system and FICTION could improve the sensitivity of cytogenetic risk strati cation in MM patients.

Background
Multiple myeloma (MM) is a heterogeneous disease associated with underlying chromosomal abnormalities and has greatly variable prognosis. Recent studies have shown that cytogenetic abnormalities are important factors affecting MM prognosis. However, MM karyotyping has been hampered not only by a low level of bone marrow (BM) in ltration and the low mitotic index of plasma cells (PCs) in vitro, but also the high viscosity of bone marrow uid in MM patients, which often leads to specimen coagulation. In addition, in many MM cases, poor metaphase quality has prevented the identi cation of recurrent aberrations, such as translocations by chromosome banding. As a result, cytogenetic analysis of MM patients often fails, and the detection rate of abnormal karyotypes is low.
Fluorescence in situ hybridisation (FISH) is the method of choice for fast and reliable identi cation of chromosomal alterations, including copy number abnormalities, genomic break points, and translocations or fusion genes. This method can also improve the detection rate of genetic abnormalities in MM. However, ordinary FISH may be unable to distinguish PCs from other nucleated cells when the proportion of PCs is low. As a result, the proportion of abnormal signals may be reduced, potentially resulting in false negative results. Fluorescence immunophenotyping and interphase hybridisation as a tool for the investigation of neoplasms (FICTION) is a technique that combines ordinary FISH and uorescence immunophenotyping. The probe signals of the target cells can therefore be selectively analysed under a microscope. CD138 is usually employed to mark PCs, for it is almost 100% expressed in PCs, but not in BM hematopoietic progenitors and lymphocytes. Owing to its capacity to improve the e cacy of results interpretation, FICTION may be extremely useful in tumour diagnosis, particularly, for the detection of low tumour load, minimal residual disease (MRD), or composite tumour diagnosis [5].
The purpose of the current study was to assess the effect of cytokine stimulation and culture-time extension on the success and detection rates of genetic abnormalities in MM. In addition, the ability of FICTION to increase the detection rate of genetic abnormalities was investigated in this study.

Conventional cytogenetic analysis (CCA) of PCs cultured in improved conditions
The analysis was double-blind. Ordinary culture served as a control. Both the number and quality of karyotypes available for analysis were signi cantly increased when BM cells were cultured in improved conditions by stimulating by cytokines and prolonging culture time (Fig. 1).
In the control group, chromosome analysis achieved a success rate of 77.3% (17/22), and 9.1% (2/22) of abnormal karyotypes were detected. Both of the two abnormal karyotypes were complex (three or more abnormalities present). In the improved group, chromosome analysis achieved an 86.4% (19/22) success rate, and 40.9% (9/22) of abnormal karyotypes were detected. Of these abnormal karyotypes, 77.8% (7/9) were complex karyotypes. Abnormalities involved all chromosomes except chromosomes 16 and 22 (Table 1). All abnormalities detected in the control group were also found in the improved group. The detection rate of abnormal karyotypes in the improved group was signi cantly higher than that in the control group (P = 0.015). Chromosomal abnormalities were detected in 17 (89.5%) of 19 patients by FICTION, and all abnormalities detected by FISH were also found by FICTION (Fig. 2). The highest positive rate was observed for detecting RB1 deletion, with a positive rate of 73.7%. The positive rates of IgH break-apart (signal patterns including 1O1R1G, 1O1R, 2O1R, 1O1R2G, and 1R), 1q21 ampli cation, D13S319 deletion, and P53 deletion were 68.4%, 57.9%, 57.9%, and 42.1%, respectively. Two samples manifested 1q21 deletion (1R) and IgH deletion (1O), and one sample manifested IgH ampli cation (3O).
The abnormal detection rates of each probe are shown in Fig. 3. Signi cant differences in the abnormal rate were found between FISH and FICTION when detecting RB1, D13S319, and P53, with P values of 0.003, 0.013, and 0.006, respectively.
Of the seventeen abnormal specimens detected by FICTION, only one patient (patient No. 8) yielded results consistent with those of FISH. However, the proportion of positive cells was higher than in FISH. 1q21 ampli cation yielded 60.0% and 18.0% positive cells in FICTION and FISH, respectively, and IgH break-apart yielded 38.0% and 15.0%. The other 16 cases with inconsistent results are shown in Table 2.
The abnormal detection rate of FICTION (89.5%) was signi cantly higher than that of FISH (60.0%) (P = 0.030). As shown in Table 3, the median proportion of abnormal positive cells detected by FICTION was higher compared to that by FISH, and the median proportion of PCs was lower in patients with chromosomal abnormalities. In 11 cases, abnormal clones were detected by both FISH and FICTION. Four other cases were determined to be abnormal via FICTION, but the proportion of abnormal clone-positive cells detected by FISH was a critical value. In these 15 samples, a high proportion of abnormal cells were detected by FICTION (Fig. 4). The differences between the positive rates of 1q21, RB1, D13s319, and IgH detection were statistically signi cant, with P values of 0, 0.038, 0.004, and 0.001, respectively.

Discussion
Genetic abnormality is an important factor for risk strati cation in MM [6], as well as for predicting survival and de ning a treatment strategy. It has been reported that almost all MM patients possess cytogenetic abnormalities, sometimes during the course of the disease [7]. However, PCs occur at the nal stage of differentiation, which have a low proliferation rate and long cell cycle. The viscosity of bone marrow uid in MM patients is high, which often causes specimens to coagulate during extraction. As a result, cytogenetic studies of MM after conventional short-term culture methods are hampered by di culty in or even failure to obtain good metaphases for analysis. Furthermore, the abnormal detection rate of karyotypes is also low. In previously reported studies [8][9][10], only 20-40% of abnormal karyotypes could be detected by traditional metaphase cytogenetic analysis.
To improve the detection of chromosomal abnormalities in MM, the application of different cytokines was explored. IL-6 has been recognised as the major growth factor for PCs both in vitro and in vivo. It has been used in combination with haematopoietic factors such as GM-CSF and IL-3, as these growth factors increase the responsiveness of IL-6 for supporting PC growth. However, controversial results have been reported. Lai [4], Facon [2], and Cuneo [1] have suggested that adding different combinations of GM-CSF, IL-6, and IL-3 to the culture medium and prolonging culture time to 6 d offers an advantage over unstimulated overnight cultures. In contrast, Smadja [11] and Brigaudeau [12] found that compared to adding growth factors, the abnormal detection rate of a 3-day unstimulated culture was comparable or higher.
Based on previous studies, IL-6 (10 ng/mL) and GM-CSF (40 ng/mL) were added to the culture medium in this study and prolonged the culture time to 6 d to improve the abnormal detection rate of karyotypes in MM patients. As a result, the number of metaphases and the quality of chromosomes improved signi cantly in most samples compared to ordinary CCA. However, abnormal karyotypes were detected in samples 10 and 17 under improved culture conditions, while the conventional short-term culture method detected normal karyotypes. The number of karyotypes was lower after culture in improved conditions. This may be because the proliferation of non-PCs is faster during short-term culture, and PC proliferation inhibits that of non-PCs after the addition of IL-6 and GM-CSF.
Compared with conventional culture, culturing in improved conditions increased the karyotype and abnormal karyotype detection rates of MM patients. The abnormal detection rate reached 40.9%, which was similar to that reported in the literature [2,4,10].
It is now recognised that there are two broad genetic subtypes of MM, as de ned by chromosome number. Hyperdiploid myeloma (48-74 chromosomes) accounts for about 50% of cases and are mostly trisomies of chromosomes 3, 5, 7, 9, 11, 15, 19, and 21. This subtype rarely involves IgH translocations at the locus 14q32. Non-hyperdiploid myeloma (< 48 or > 74 chromosomes) is associated with the presence of IgH translocations [13]. This study con rms and expands upon prior reports, identifying that 2 cases of hyperdiploidy and 7 cases of non-hyperdiploidy were detected in improved culture conditions. Further, 88.9% (8/9) had both numerical and structural abnormalities. This suggested that the chromosomal abnormalities in MM were often complex. The abnormality involves all chromosomes except chromosomes 16 and 22, as abnormal hyperdiploid MM typically involves chromosomes 2, 4, 5, 7, 9, 10, 11, 12, 13, 17, 19, and 21. Three cases of non-hyperdiploid MM were associated with IgH translocation and 2 cases with structural abnormalities involving 19p13.
A considerable proportion of MM patients have only a small proportion and uneven distribution of PCs in the BM. Therefore, determining a means by which to accurately identify PCs under a microscope without the interference of normal cells is the primary issue to be addressed. The International Myeloma Working Group and the European Myeloma Network recommend that FISH detection in MM should not be conducted without rst concentrating the PCs or employing a means for identifying PCs, ensuring that only these cells are scored [14] To ful l this recommendation, PCs need to be sorted by CD138 immunomagnetic bead-based sorting or by labelling the cytoplasmic immunoglobulin light chain (cIg).
However, enriching PCs via CD138 immunomagnetic bead-based sorting requires at least 20 mL [15] to satisfy the detection requirements, specimen acquisition is di cult, and the sorting process is labour and time-intensive. Cytoplasmic light chain immuno uorescence combined with FISH (cIg-FISH) requires identi cation of the light chain types of clonal PCs before detection, ensuring the correct light chain antibody is selected. This is not easily incorporated into the routine work ow of a cytogenetic laboratory.
Another proposed technique is the use of an initial Giemsa stain with PC mapping by image analysis software, followed by subsequent FISH analysis of the previously mapped PCs [16]. However, this technique relies on expensive image analysis software, which inevitably limits its clinical application.
Unlike the techniques mentioned above, when used in combination with simultaneous immuno uorescence (IF) using CD138 and FISH, FICTION does not require a large number of PCs, nor prior knowledge of the restricted light chain involved. Further, specialised procedures such as cell sorting or image analysis are unnecessary for this method. This makes it possible to recognise PCs under a uorescence microscope and interpret genetic changes in PCs with cytoplasmic staining. FICTION can be implemented in laboratories that routinely perform FISH analysis, which is suitable for both fresh BM specimens and archived BM smears [16].
In the present study, FICTION detected a higher percentage of abnormal cases than FISH. In one CR sample (No. 21), the short-term conventional culture method, CCA with the improved culture system, and FISH did not detect abnormalities, but FICTION detected a low proportion of positive signals. These results suggested that the patient may be in the early stages of relapse. Compared to conventional FISH analysis, FICTION is more sensitive and could detect abnormalities when the proportion of PCs was low. FICTION is also more speci c, detecting a higher proportion of abnormal cells compared with FISH when the proportion of PCs was the same.
Hanamura [17] showed that the 1q21 ampli cation rate of newly diagnosed MM patients is less than 20%, while the detection rate of relapsed patients can increase to 44%. Patient prognosis is worse when increased ampli cation signals are detected. In the current study, 4 1q21 ampli cation signals were detected by FICTION in patient No. 8 54 months after diagnosis. FICTION detected 60% positive cells, while FISH detected only 18%. Two months after the detection of 4 ampli ed signals in 1q21, the patient discontinued treatment due to septic shock, paroxysmal atrial brillation, cardiac insu ciency, decompensated respiratory alkalosis, coagulation abnormalities, and electrolyte disturbance.
A partial deletion of chromosome 13 is the rst genetic abnormality found in MM and is related to drug resistance. In terms of prognosis, 13/13q abnormalities are often attributed to a poor outcome, regardless of whether they receive conventional chemotherapy or autologous stem cell transplantation [18]. The positive rate of RB1 and D13S319 detection by FISH was higher in this study than those reported in the literature [17]. This may be related to different methodologies and was also affected by the small sample size.
This study determined that combining CCA with improved culture conditions and FICTION could increase the abnormality detection rate to 92.3%. Combining CCA with improved culture conditions and FICTION could improve the sensitivity of cytogenetic risk strati cation in MM patients, allow clear interpretation of the results of numerical critical sample detection, and provide more accurate information for personalised medicine. In future, the number of samples will be expanded to con rm the difference in detection quality between the two methods.

Conclusion:
The established modi ed culture system could improve karyotyping quality in MM. Owing to its obvious advantages compared with FISH, FICTION is recommended for detecting genetic abnormalities in MM.
Combining modi ed culture system and FICTION could improve the sensitivity of cytogenetic risk strati cation in MM patients.

Materials And Methods
Patient selection and basic characteristics BM specimens were extracted from 28 MM patients who visited Fujian Medical University Union Hospital from January through December 2018. The criteria of "Hematology Diagnosis and E cacy Standards Third Edition" by Zhinan Zhang [19] were used to establish the diagnosis, and all cases ful lled the criteria for MM. There were 21 males and 7 females, with a median age of 58 years (range, 35-88 years) and median BM PC percentage of 27% (range, 0-78%). The monoclonal component was IgA in 7 patients, IgD in 2 patients, IgG in 15 patients, and Bence-Jones protein in 4 patients. The staging was done according to the international staging system (ISS). Five patients were in stage I, 2 were in stage II, 20 were in stage III, and 1 was in complete remission (CR). Nineteen patients were newly diagnosed, and nine were revisited. A bone aspirate smear of each patient was stained with Wright-Giemsa. At least 200 cells were counted for cell classi cation, and the PC proportion was calculated. Due to the limited number of specimens in some patients, it was impossible to complete one of the cytogenetic analytic techniques, either FISH or FICTION. Therefore, the numbers of cases in each group were slightly different. The study was approved by the ethics committee of Fujian Medical University Union Hospital (Ethics protocol number: 2016KY050), and BM samples were collected after obtaining written informed consent from each patient. CCA CCA was performed on 24-h BM cultures [20]. At least 20 metaphases or all available metaphases were examined. The International System for Cytogenetic Nomenclature (2013) was used to describe abnormal clones. Each specimen was analysed by at least two experienced laboratories.
Effect of stimulating PCs with cytokines and extending culture time on CCA Nucleated cells (2 × 10 6 /mL) from BM were seeded into 5 mL RPMI 1640 (Hyclone, USA) medium containing 10% foetal calf serum, followed by the addition of IL-6 (10 ng/mL, PeproTech, USA) and GM-CSF (40 ng/mL, PeproTech). Cells were cultured at 37 °C under 5% CO 2 conditions for 6 d, followed by harvesting. Cytogenetic analyses were carried out as above.
The percentage of uorescence-positive cells ≥ the mean ± 3 standard deviations for the controls was applied as a positive criterion. The threshold was 6.6% for 1q21 ampli cation, 6.45% for RB1 deletion, 6.95% for D13S319 deletion, and 8.25% for P53 deletion. A break was de ned in the IgH break-apart probe when a red/green or orange fusion signal (O) was split into separate red and green signals. Only red and green signals, which were more than one signal diameter apart from each other, were counted as a break. The threshold for IgH break-apart was 6.85%. Fluorescent hybridisation signals were counted in 200 cells underneath a uorescence microscope (Olympus BX51, Tokyo, Japan) using an oil immersion objective and lter sets for DAPI, FITC, and RHOD. Raw uorescence images were analysed using IMSTAR

FICTION analysis
Red blood cells from uncultured BM were fully lysed with red blood cell lysis buffer (Genview, USA). Prexed cell suspension was mixed with − 20 °C pre-cooled methanol (1:1 cell suspension: methanol). BM cell suspension was dropped onto glass slides, and the slides were air-dried. The slides were then xed with Triton 100-X/4% PFA (1:1, 0.5% Triton X-100: 4% PFA) for 10 min. The following steps were the same as those carried out in ordinary FISH. The glass slides were washed and dehydrated in gradient ethanol solutions (70%, 85%, and 100%) after hybridisation. They were then blocked with 2.0% BSA for 5 min, and mouse anti-human CD138 antibody (primary antibody) and AMCA-labelled goat anti-mouse IgH antibody (secondary antibody) (PeproTech) were added. The slides were incubated at room temperature and protected from light for 1 h. The samples were air-dried in the dark, and 10 µL 4',6-diamidino-2phenylindole (DAPI) counterstain was added to each. Slides were covered with a coverslip and stained for 5 min. Using an oil immersion objective and lter sets for DAPI, AQUA, FITC, and RHOD, FISH signals were analysed only in cells with blue cytoplasm. At least 200 cells were counted. When the percentage of PCs in BM was low, at least 50 cells were counted.

Statistical analysis
Fisher's test was used for comparing the detection rate of karyotypic abnormalities between the two groups. Chi-square test was applied for analysing the differences in the abnormal detection rate of each probe. The number of positive cells detected by FISH and FICTION were compared between groups with a paired t-test. All analyses were performed using SPSS 25.0 Statistics (SPSS Inc., Chicago, IL, USA).  Positive rates of 15 samples detected both by uorescence immunophenotyping and interphase cytogenetics as a tool for the investigation of neoplasms (FICTION) and uorescence in situ hybridisation (FISH). P53 specimens were insu cient for analysis. The differences between 1q21, RB1, D13S319, and IgH were statistically signi cant (P = 0, 0.038, 0.004, and 0.001, respectively). Results of uorescence immunophenotyping and interphase cytogenetics as a tool for the investigation of neoplasms (FICTION). a. Arrows indicate plasma cells. 2R2G denotes D13S319/ P53 normal; 2R1G denotes D13S319 normal and P53 deletion. b. Arrows indicate plasma cells with IgH break-apart (1O1R1G). c. Plasma cells with 1q21 ampli cation and RB1 deletion (3R1G).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. renamedcea64.xls