Chromosome analysis on bone marrow or peripheral blood samples fails in a small proportion of attempts. A method that is more reliable, with similar or better resolution, would be a welcome addition to the armamentarium of the cytogenetics laboratory.
To develop a method similar to banded metaphase chromosome analysis that relies only on interphase nuclei.
To label multiple targets in an equidistant fashion along the entire length of each chromosome, including landmark subtelomere and centromere regions. Each label so generated by using cloned bacterial artificial chromosome probes is molecularly distinct with unique spectral characteristics, so the number and position of the labels can be tracked to identify chromosome abnormalities.
Interphase chromosome profiling (ICP) demonstrated results similar to conventional chromosome analysis and fluorescence in situ hybridization in 55 previously studied cases and obtained useful ICP chromosome analysis results on another 29 cases in which conventional methods failed.
ICP is a new and powerful method to karyotype peripheral blood and bone marrow aspirate preparations without reliance on metaphase chromosome preparations. It will be of particular value for cases with a failed conventional analysis or when a fast turnaround time is required.
Conventional chromosome analysis plays an important role in establishing the diagnosis, prognosis, and treatment planning for hematologic malignancies, in determining the genetic basis of pregnancy loss, and other purposes. However, current karyotypic methods are limited to the availability of adequate mitotically active cells.
These limitations can lead to false-negative results if abnormal mitotic cells are few, the relevant cells are not mitotically active in sufficient numbers in cell culture, or the cells do not proliferate in the media or mitogen provided. A more reliable and less costly high-resolution cytogenetic method would be a welcome addition to the cytogenetics armamentarium. Here we describe the development and validation of interphase chromosome profiling (ICP), a novel cytogenetic technology to assess the karyotype of any hematologic neoplasia by using interphase cells. This new approach can detect all numerical abnormalities and most balanced and unbalanced structural aberrations, including characterization of “add” and “marker” chromosomes.
MATERIALS AND METHODS
This study was determined to be exempt from institutional review board review by the Mayo Clinic Institutional Review Board, Rochester, Minnesota.
Conventional chromosome analysis routinely uses metaphase cell preparations, since the chromatin is condensed at this stage and individual chromosomes can be readily visualized. However, there is good evidence that chromosomes behave similarly in both interphase and metaphase, and their length in interphase is similar to that in a 600-band karyotype.1 With whole chromosome multicolor karyotyping methods (eg, multiplex fluorescence in situ hybridization [M-FISH]), spectral karyotyping, combined binary ratio fluorescence in situ hybridization (COBRA-FISH), some chromosome rearrangements such as translocations can be visualized,2–5 and intrachromosomal changes such as inversions, deletions, and duplications can be identified with a multicolor metaphase banding approach.6,7 However, interphase visualization of specific chromosome rearrangements (eg, fiber-FISH) has presented greater challenges,8,9 and until now, no method has been successfully designed for banded karyotype analysis in interphase nuclei.
Building on the observations of Lemke et al,1 and by assigning FISH probes to target sequences along the length of each chromosome in an equidistant fashion, entire chromosomes can be evaluated in interphase nuclei in normal and disease states (US patents 8,574,836 and 7,943,304). Each chromosome arm includes at least 1 (18p and Yp) and up to 6 (2q, 4q, and 5q) hybridization sites, each assigned to a specific chromosome band. Subtelomeric and paracentromeric sequences are assigned a pure color (aqua, yellow, and red for pter, centromere, and qter, respectively), and interstitial bands are assigned either a pure (far red or green) or a hybrid (fusion) color. Each chromosome is studied individually, with the results compiled into a composite karyotype. This configuration provides the equivalent of a 600-band resolution karyotype10 and facilitates identification of copy number changes of whole chromosomes and balanced and unbalanced chromosome rearrangements.
Cell Culture and Microscope Slide Preparation
Microscope slides were prepared after a standard 24- to 48-hour cell culture and harvested by using a 0.075M KCl hypotonic solution for 20 minutes at 37°C without colcemid followed by 3 changes, at room temperature, of freshly prepared 3:1 methanol–glacial acetic acid fixative. A direct preparation without cell culture can be used for STAT cases.
Keeping with the equidistant concept, each chromosome arm was assigned a certain number of bands. The overall length of the chromosome arm was the primary deciding factor for the ultimate number of bands on that arm. In general, longer chromosome arms are assigned a greater number of bands (targets). For example, the short arms of chromosomes 6 and 11 have more (3) bands than the short arm of chromosomes 5 and 10. The long arm of chromosome 11 has 5 bands, whereas the long arm of chromosome 9 has 4 bands because although the chromosomes are similar in length, 11q is longer than 9q when the 9q heterochromatic region is excluded.
Individual bacterial artificial chromosome (BAC) clones were selected from a Children's Hospital Oakland Research Institute library for each target on the chromosome.11 Two BACs were chosen for each target location. Figure 1 provides a cytogenetic ideogram with the targets and the corresponding probes with their fluorescent labels. Table 1 lists the number of bands in each chromosome arm, and Figure 2, A and B, depicts the color scheme of short and long arms. With the exception of the green color, which was assigned to a short arm as well as a long arm band on some chromosomes, bands in each arm received distinct colors. For example, aqua and far red were restricted to the short arm, and yellow and red were restricted to the long arm. In a few instances when a proper BAC was not available in the long arm, the short arm pericentromeric area was chosen. However, the total number of pure colors in any arm was limited to 3. The interstitial hybrid colors were simply distinct combinations of the pure colors belonging to that arm. The following scheme was used in assigning the number of hybrid bands in any arm. First, 2 landmarks were chosen for each arm. The telomere (aqua) and the most proximal band (green) were the landmarks for the short arm. The centromere (yellow) and the telomere (red) were considered the landmarks for the long arm. In order for a chromosome arm to receive a hybrid band, it needs to have more than the 2 landmark bands. Therefore, chromosome arms large enough to require 4 bands received 1 hybrid distal to the green landmark, and the proximal band to the telomere landmark received the remaining pure color (far red) in the short arm. Similarly, for the long arm, the proximal band to the red landmark received hybrid color, and the distal band to the centromere received the remaining pure color (green). For an arm with 5 bands, the 3 interstitial bands were made hybrids. For an arm with 6 bands, the distal band to the telomere was given the remaining pure color (green), and the 3 remaining interstitial bands became hybrids. Finally, when any arm has only the landmarks, the third pure color was eliminated. Since the short arms of chromosomes 18 and Y were small, only a telomere landmark band was assigned. For each probe, we used a 3-step verification process: (1) checking for known genomic variants,12 (2) end sequencing to match the intended chromosomal location,11 and (3) confirmation of the expected metaphase band location.13,14 Only probes that passed these selection criteria were used for further validation.
FISH probes were generated by using standard nick translation protocols from BAC clone DNA.15 Briefly, 1 to 2 μg nick translation reactions were run for each BAC clone. The average size of the selected BAC clones was approximately 200 kb, and the individual fragments that collectively make up a single probe for each selected target were adjusted to be around 200 base pairs in length to facilitate efficient hybridization. The ratio of dTTP to fluorophore dUTPs was optimized. Individual chromosome hybridizations were done on 4 slides with 6 areas of hybridization on each slide, following standard FISH protocols16 with minor adjustments. About 20 ng of probe was used for each chromosomal target and the overnight hybridization was done in a 10-mm area under a round coverslip. Posthybridization washing conditions included 2 minutes in 0.4× saline-sodium citrate (SSC)/0.3% NP-40 at 69°C, followed by 1 minute in 2× SSC/0.1% NP-40 at room temperature; 4′,6-diamidino-2-phenylindole (DAPI) counterstaining was omitted. Appropriate filter sets from Semrock (Rochester, NY) were used to detect fluorophores DEAC (aqua), Fluorescein-12 (green), Cyanine555 (yellow), Cyanine647 (far red), and CF594 (red).
Initial scanning to place the cells in the correct plane was done with the filter for Cyanine555. A minimum of 20 interphase cells were analyzed for each chromosome. The usual guidelines of metaphase analysis were followed with minor adjustments to identify an abnormal clone: at least 4 of 20 cells for both structural and numerical abnormalities.
To get familiarized with the banding pattern generated for each chromosome, known normal metaphases were studied. The individual color patterns were studied in metaphase spreads and interphase nuclei, and the combined results of all color combinations were verified to assure concordance with the expected result (Figure 3 and Figure 4, A and B).
In contrast to usual single-target interphase FISH assays, the usual guideline of 20 metaphases is considered adequate and each chromosome is analyzed in 20 nuclei, since the entire chromosome is profiled. To accommodate the maximum number of filter cubes on any fluorescent microscope, DAPI counterstaining was eliminated. This did not create any problem in clearly identifying nuclei, as there was enough autofluorescence of the nuclei. To analyze each chromosome, the following protocol was observed: the guidance provided in Table 1 and in Figure 2, A and B, was followed to make sure all of the bands were present on any chromosome. In interphase nuclei, each chromosome typically occupies a “domain,” and one would count the signals within this domain. For example, chromosome 8 has 3 red, green, and yellow signals and 1 aqua signal. Even though there are 3 hybrid bands on this chromosome, by simply counting the total number of bands (both homologs, 2 domains) in each color channel (filter cube), one can determine if the chromosome has the expected number of bands. Any deviation from this signal pattern would indicate a numerical abnormality. A trisomy would produce 9 red, 9 green, 9 yellow, and 3 aqua signals, whereas a monosomy would produce a 3 red, 3 green, 3 yellow, and 1 aqua signal pattern. The number of domains then varies: 2 for normal, 3 for trisomy, and 1 for monosomy.
Signal patterns for a normal chromosome in interphase, a normal chromosome in metaphase, structurally abnormal chromosomes, and a numerical abnormality are all shown in Figure 5, A, B, C through H, and I, respectively. When there is a structural abnormality such as a translocation, whether it be balanced or unbalanced, 3 domains are created (Figure 5, C and D). However, only 1 domain exhibits the expected “normal size and signal pattern” and 2 abnormally patterned domains represent the 2 broken chromosomal segments. The size of the abnormal domains then depends on the translocation breakpoint. Deletions are recognized by the count of signals as well as the smaller size of the domain (Figure 4, A and B; and Figure 5, D, G, and H). Similarly, when there are duplications, depending on the extent of the abnormality, a relatively larger domain would be apparent. A characteristic difference exists between similarly sized duplications depending on whether they represent tandem or not. For tandem duplications, there is no change in the number of domains (Figure 5, E and F), whereas a duplication resulting from an unbalanced translocation will exhibit 3 domains (Figure 5, D).
This form of chromosome rearrangement is easily recognized in ICP as there will be duplication of 1 arm and deletion of the other. So in the abnormal domain for an isochromosome of the long arm, there will be clear absence of the characteristic color band(s) from the short arm, that is, aqua (Figure 5, H). Similarly, for an isochromosome of the short arm, absence of 1 or all red bands would be apparent.
When there is an overlap of 2 signals representing the same band/chromosome location, traditional FISH assays would create an erroneous, deletion signal pattern. In ICP, since the entire chromosome is profiled even when there is an overlap, it is easily recognized by taking into consideration the total number of bands of that color as well as the adjoining signals on that chromosome. This is accomplished by switching the filter cubes during analysis. When 2 telomeres are situated very close to each other, it is possible to assign either of the signals to any of the homologs, and this should not cause misinterpretation as long as the total count for those color sign patterns remained as expected (Figure 5, C). As with traditional FISH, analysis of multiple nuclei will also clarify a misleading pattern in 1 nucleus. Generally, any uncertainty of signal pattern in 1 cell is easily clarified in the analysis of the other 19 cells.
Hybrid/Juxta Color Bands
Depending on the total number of bands, each chromosome arm can have zero, 1, or 3 hybrid bands. When the count matches with the expected number for each color, it is reliable to predict that all expected hybrid patterns exist. Confirmation of the hybrid band is done by simply isolating/focusing a single-color signal under the microscope and switching the filter to observe the other color at the same location. A pure color band, for example a telomere band, will have no other color signal at that same location. Superimposing the pictures taken during or after analysis will clearly identify the hybrid/juxta bands.
Interstitial Deletions and Duplications
While large interstitial deletions and duplications are easily recognized in ICP, small changes either between the adjacent bands or within the band are difficult to identify with the current scheme. This is not dissimilar to the well-known 2-band uncertainty principle in metaphase banding analysis.
Each color band in ICP was given a specific high-resolution G-banded location. When there is a structural abnormality, generally the breakpoint is assigned to the closest distal band to the original remaining band on the chromosome. For example, chromosome 9 has 2 distal bands on the long arm: one at 9q32 and the other at 9q34.3. When the red band at 9q34.3 is separated as a result of a translocation, the breakpoint is assigned to 9q33, the band distal to 9q32.
It is well recognized that 1 of the X chromosomes in females is inactivated, and there could be differences in the physical length of the 2 chromosomes, based on their inactivation status. The active X chromosome could be more diffuse and longer than the inactive chromosome, which is very tight and short and also exhibits a bend during the metaphase stage at the inactivation center.17 In ICP, a similar phenomenon is frequently observed for almost every chromosome. The mechanism for this is not well understood at this point, but in interphase nuclei, it is frequently observed that the homologs differ in overall length. Therefore, while there may be an obvious size difference between the homologs, neither chromosome should be interpreted as abnormal, since the overall difference between the adjacent bands, either due to stretching or condensation, seems to be constant. Chromosome condensation differences are likewise common in metaphase spreads, wherein the homolog positioned at the periphery of the spread is often measurably longer than its partner positioned in the center of the spread.
Since only 1 chromosome pair is analyzed at a time, it is uncommon for a cell to be unanalyzable when using the ICP scoring criteria. As detailed above, even when there is signal overlap, individual chromosomes can be traced following the expected pattern from the short arm telomere through the centromere to the long arm telomere. The only time a cell may have to be considered unanalyzable and therefore rejected would be when most signals are so diffuse that an accurate count is not possible. Diffuse and weak signals could exist in some cells when hybridization conditions are suboptimal, and in these scenarios, these cells could be rejected.
Clinical validation of the ICP method was done in 2 steps. Once the normal chromosome banding pattern was verified and familiarized, 20 bone marrow samples with known cytogenetic results were chosen for blinded analysis (Table 2). Three institutions participated in this study. The institution that developed the methodology received 15 samples from the second institution, and the third institution studied 5 samples in its laboratory.
Next, 2 principal attributes of ICP were tested: improved reliability and sensitivity as compared to conventional chromosome analysis. Interphase nuclei from 39 bone marrow samples were provided by 4 institutions for a blinded ICP analysis (Table 3).
These included 29 samples with no cytogenetic results (cell culture failure) and 10 samples with known cytogenetic results. One participating institution studied 25 samples blindly. To test interlaboratory reproducibility, 5 of these 25 samples were shared with another laboratory.
The first step in the clinical validation used 20 blinded samples with known cytogenetic results, identified by conventional cytogenetics, FISH, or both, and included trisomy, monosomy, deletions, duplications, and balanced and unbalanced rearrangements (Table 2). In almost all cases, there was complete concordance with the conventional cytogenetic or FISH results. In case 5, ICP did not identify the t(1;3) clone.
In case 6, ICP also identified cells with monosomy 7; the originating laboratory was unable to revisit this case. In case 9, ICP did not identify evidence of an abnormality on the short arm of the “add(X).” In case 15, the original testing laboratory was unable to distinguish the small terminal 8p deletion. For several cases with unidentified material in “add” rearrangements, ICP characterized the origin of the additional chromosome material.
The second step in the clinical validation tested the ability of ICP to identify the abnormalities in another set of 35 blinded cases with known cytogenetic results, plus a series of 29 cases for which conventional cytogenetics failed to produce a result (Table 3). Five of the 35 blinded cases were evaluated by using ICP in 2 laboratories, with similar results, which served to demonstrate the interlaboratory reproducibility of the assay.
As with the first series of cases, again there was nearly complete concordance with the conventional cytogenetic or FISH results. In 6 cytogenetically normal cases (cases 35, 42–45, and 50) and 1 normal FISH case (case 46), ICP detected clonal abnormalities diagnostic and/or prognostic of various diseases (Tables 2 and 3). In 4 cases with cytogenetic abnormality, ICP failed to detect the abnormal clone. These included 2 cases with inversions (cases 25, 30) and 2 with a very low level (2 cells) abnormality (cases 15, 26). In 1 case (case 5), ICP failed to detect the t(1;3); however, the same (or a similar) translocation, even when present in low level in the cytogenetic preparations in a different case (case 6), was observed by ICP.
In 3 cases, ICP identified the origin of a marker chromosome (cases 53, 54, 55; Figure 6, A through D). One of these cases had 2 identical copies of a marker chromosome derived from chromosome 9, with a potential neocentromere, since these “acentric” markers were very stable (Figure 6, A through D; Table 3). In 1 case (case 57), ICP refined the breakpoints, which were diagnostic for an IGK/CDK6 fusion (Table 3). In 6 cytogenetically abnormal cases (23, 32, 37, 38, 40, 41), ICP identified additional clonal abnormalities (Tables 2 and 3). Example ICP results are depicted in Figure 5, A through I, and Figure 6, A through D.
The ICP method yielded analyzable results in all 29 bone marrow samples for which conventional cytogenetics was attempted but unsuccessful. Ten had an abnormal ICP karyotype (cases 56–65), and 19 had a normal ICP karyotype (cases 66–84). In a separate preliminary study (R.B., unpublished data, 2015), ICP confirmed a normal conventional chromosome analysis result in 9 cases, confirmed a normal myelodysplastic syndrome or myeloma FISH panel result in 4 cases, and combined normal chromosome analysis and myeloma FISH panel results in 7 cases.
The whole study involved sample types of peripheral blood and bone marrow aspirates (Tables 2 and 3) after cell culture without mitogens. There was no difference in the quality of ICP results between the blood and bone marrow cells. The only time a significant difference in the signal quality was observed was when ICP was done on direct preparations, which tended to yield less satisfactory signal patterns (Figure 7, A and B).
Since the clinical management of patients with hematologic malignancies is often influenced by karyotype findings, obtaining this information in a fast, accurate, and failure-proof manner is critical. Even though other focused molecular technologies can be useful, when the working clinical diagnosis is questionable or there is a differential diagnosis, which often is the case in hematologic malignancies, karyotype analysis is the only approach that identifies genome-wide gross chromosome abnormalities. In this study, we have developed a novel molecular method, which we termed interphase chromosome profiling. Using this process, we showed that a molecular karyotype can be obtained from interphase nuclei without any prior clinical information, mitogen stimulation, or other cell culture modifications. The experiments described here indicate that this technique is robust and highly reproducible.
The resolution obtained from the ICP design yields a karyotype roughly equivalent to a 600-band level, in contrast to the usual karyotypes obtained at a 400-or-fewer band level from hematologic samples. This higher resolution allows for confident breakpoint localization and identification of chromosome material residing in marker chromosomes and unbalanced rearrangements that are difficult to characterize.
As demonstrated, results can be obtained in 48 hours by using a brief cell culture, without the use of mitogens or mitotic arresting agents. In the case of STAT situations for acute promyelocytic leukemia with promyelocytic leukemia/retinoic acid receptor alpha (PML-RARA), a direct interphase cell preparation can be used (Figure 7, A and B), which allows for a next-day ICP karyotype report.
This study demonstrates that ICP is extremely reliable. All 84 samples studied using ICP—including 29 cytogenetically failed cases—had a clinically interpretable karyotype result. This is critical in the workup of hematologic samples. Since the costs and risks associated with a repeated bone marrow aspirate are not trivial, ICP as a first-line or backup method will find utility in patient management from both diagnostic and follow-up treatment samples.
Improved Analytic Sensitivity
Incorrect breakpoint assignment of structural abnormalities has obvious clinical implications. Obtaining the diagnostic chromosome abnormality at the initial workup of patients with hematologic disorders is crucial for disease classification and management. In 7 of the cases in this study, only normal karyotypes were obtained, but 1 or more clonal abnormalities were detected in these by ICP, including 1 with a variant t(15;17;11) characteristic of acute promyelocytic leukemia (case 35). This 3-way translocation was confirmed by using probes flanking the chromosomal breakpoints (data not shown), which neither conventional cytogenetics nor FISH analysis revealed.
Accurate identification of rearrangement breakpoints, and more specifically, determination of the extent of the rearrangement, that is, deletions and duplications, has a prognostic implication. In 1 case with a limited number of metaphase cells and poor morphology, the initial breakpoint assignment of the t(2;7)(p21;q22) failed to recognize a potential CDK6 gene rearrangement at band 7q21, which was identified by ICP.
In general, a more complex karyotype is associated with a less favorable outcome.18 Historically, conventional cytogenetics can only indicate the presence of a marker but cannot identify the nature or origin of it. In several cases, ICP was able to identify the chromosomal origin of marker chromosomes and “add” material, and it identified other abnormalities. In case 55, ICP established the mechanism of marker formation as a stable acentric chromosome with a neocentromere.
ICP on Metaphase Chromosomes
Even though ICP is designed to analyze interphase chromosomes, it certainly can be used as an adjunct, as demonstrated in this study, in characterizing the marker chromosomes and identifying the material in “add” chromosomes. An additional advantage, provided by the identification of material both in the marker chromosomes as well as the material of unknown origin as usually described in the standard karyotypes as “add,” is the possibility of discovering novel fusion genes. The detection of such novel genes was hampered until now by the lack of precise characterization of “material of unknown origin.” ICP clears the path for discovering potential fusion genes by identifying targets for further molecular elucidation with methods such as mate-pair sequencing.19,20 This has clear implications as the pharmaceutical industry develops treatments specifically geared toward fusion gene products.
Some of the small rearrangements involving breakpoints close to telomeres, such as balanced reciprocal translocations exchanging similarly banded material, escape detection by conventional methods.21 One example of that is t(12;21) in pediatric acute lymphocytic leukemia cases (Figure 5, A through I). As illustrated by our results in case 37, significant rearrangements involving telomere regions can be missed in a complex karyotype. ICP identified the balanced translocation between chromosomes 8 and 22, and this was later confirmed by standard interphase FISH with probes for MYC and IGL (Table 3). Likewise, a small terminal deletion involving the chromosome 8 short arm was identified only by the ICP method (case 15).
Generally, ICP is used when traditional cytogenetics fails to produce any result. However, ICP is a very flexible technique, and based on the clinical situation on hand, it can be used by careful selection of only a few chromosomes, that is, partial ICP. For example, in a follow-up failed cytogenetics on a patient with a proven diagnosis of chronic myelogenous leukemia, chromosomes 8, 17, and 22 can be used for partial ICP analysis to rule out blast crisis. With just these 3 chromosomes, the most common blast crisis–related changes—that is, trisomy 8, isochromosome for the long arm of 17, and additional copies of the Philadelphia chromosome—can be assessed. Similarly, in any failed cytogenetic case with established chromosome changes in the original diagnostic or previous studies, for which no FISH probes are currently available, partial ICP for the selected chromosomes will be very useful in providing the current status.
Confirmation of a Translocation
As opposed to conventional karyotyping, whereby all the chromosomes in a cell are analyzed at the same time, in ICP each chromosome pair is analyzed separately. As a result, in a case where a patient has a clonal abnormality, for example, a balanced translocation between chromosomes 12 and 21, when chromosome 12 is analyzed it will show a displacement of the chromosome segment distal to the breakpoint of the translocation (Figure 5, C). Similarly, there will be a displacement on chromosome 21. This is a simple scenario where there is only 1 balanced translocation and only 2 chromosomes are involved. More complex karyotypes may exhibit more than 1 balanced translocation or a translocation involving multiple chromosomes. Therefore, it may be necessary to confirm that what was presumed from the initial ICP analysis is indeed the case with respect to the partners involved in a translocation. We have confirmed the involvement of the presumed partners in a given translocation by designing probes flanking the breakpoints on the involved chromosomes and analyzing both hybrid chromosomes in the same cell. We performed the confirmatory analyses for all common translocations observed in hematologic malignancies, and a few examples are shown in Figure 8, A through D. In clinical practice, each laboratory would have to do these confirmatory tests only once during the initial validation studies. This can be done by using the existing targeted FISH probes for translocations. Afterwards, in a simple scenario involving only 2 chromosomes, it can be reliably presumed that the 2 chromosomes with displaced chromosome segments actually formed the balanced translocation. Additional confirmatory testing is only needed when there are more than 2 chromosomes involved.
Guidelines for Calling a Clonal Abnormality
Historically, the definition for a clone in the cytogenetics literature requires 2 cells with a structural abnormality and 3 cells with a numerical abnormality, for metaphase analysis.10 However, for certain neoplasms, the cell with an abnormality may have a selective proliferative advantage in culture, and as such if only 2 cells are detected with an abnormality in a standard 20 metaphase cell analysis, ICP may have a disadvantage in identifying the abnormal clone in a 20 interphase cell analysis. Increasing the number of cells analyzed could alleviate this concern. There were 2 such discordant cases in this study. Based on 5 cases with normal karyotype by standard metaphase and FISH analyses, it appears that when an isolated abnormality is present in 3 or fewer cells on ICP preparations, it may represent a nonclonal change. Therefore, for ICP analysis, a minimum of 4 cells is required to consider a clonal abnormality, either structural or numerical.
Any time an abnormality of any kind was present in 4 or more ICP cells, it was also present in the conventional chromosome analysis. Therefore, it is considered a “major” abnormality for the purpose of ICP analysis. Thus, we consider an abnormality present in 3 or fewer cells to be a “minor” abnormality for the purpose of ICP analysis. As discussed already, a minor abnormality was not always confirmed in the standard cytogenetic analysis. Therefore, at the present time, the clinical significance of the minor abnormalities detected in the ICP analyses is unclear. Of course, in clinical practice additional analysis may be useful to confirm the presence of a clone. Large prospective studies are needed to clarify the importance of minor ICP abnormalities, and such studies are underway.
Until enough experience is gained from the studies using this technology and the International System for Human Cytogenetic Nomenclature committee issues guidance, we propose to add a prefix “icp” to describe the results generated by using the technology described here. We also use “cp” routinely, because the karyotype interpretation is assembled from multiple interphase cells. For example, a sample from a male with a balanced translocation between chromosomes 9 and 22 with a break in band 9q34 and 22q11.2 could be described as icp.46,XY,t(9;22)(q34;q11.2)[cp20].
The single case where ICP could not detect the t(1;3) identified in the cytogenetic study, even after extending the analysis to a large number of cells, may represent a laboratory error. Since no specimen remained to reexamine the case, the possibility of specimen mix cannot be excluded. Additionally, ICP clearly detected the same t(1;3) in a different case from the same collaborating laboratory even though the abnormality was present in only 4 cells in the cytogenetics study. Both inversion 11q and inversion 3q were missed by ICP. This is an inherent limitation of the current design of ICP, and therefore alternative methods including standard FISH should be used to rule out inversions. Thus for all practical purposes, inversions are the only cytogenetic abnormality that ICP cannot reliably identify.
ICP is virtually failure-proof and can detect both numerical and structural aberrations including characterization of marker chromosomes and add material. For the workup of hematologic malignancies with failed cytogenetics and with “normal” results in plasma cell myeloma cases, ICP for all 24 chromosomes can be a preferred reflex test, since standard FISH panels do not detect all clinically relevant abnormalities. This technology may prove useful for other areas of cytogenetic investigations such as products of conception. Such studies are underway, and the results will be published elsewhere. ICP offers the additional potential to examine hundreds of interphase nuclei for a specific abnormality, similar to conventional FISH studies, as well as evaluation of individual nuclei in prenatal diagnostic studies to evaluate mosaicism. ICP is a very fast method, with a 24-hour reporting time (for STAT cases), and provides very reliable analysis on interphase nuclei. It is faster and less costly than chromosomal microarray or DNA sequencing, while also providing chromosomal structural information in addition to copy number.
The authors would like to thank Srikanthi Kopuri, MS, and Yvonne Banol, HSD, for their technical aid; Lisa Plumley, BS, for whole data organization and specimen mailing; and Anna Chockalingam, PhD, for her help with blind studies.
From the Department of Research and Development, InteGen LLC, Orlando, Florida (Dr Babu, Messrs E. Fuentes and Papa, and Ms S. Fuentes); the Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, Minnesota (Dr Van Dyke); the Department of Clinical Cytogenetics, Genetics Associates Inc, Nashville, Tennessee (Dr Dev and Ms Liu); the Department of Pathology, UT Southwestern Medical Center, Dallas, Texas (Dr Koduru); the Department of Pathology and Laboratory Medicine, David Geffen UCLA School of Medicine, Los Angeles, California (Dr Rao); and the Department of Clinical Cytogenetics, Dianon Pathology (LabCorp), Shelton, Connecticut (Dr Mitter).
Dr Babu is the founder and chief executive officer of InteGen LLC, the company that developed this technology. Mr E. Fuentes, Ms S. Fuentes, and Mr Papa are also employees of InteGen LLC. InteGen LLC has financial interests in this technology. The other authors have no relevant financial interest in the products or companies described in this article.