Conventional Cytogenetic Analysis of Constitutional Abnormalities: A 20-Year Review of Proficiency Test Results From the College of American Pathologists/American College of Medical Genetics and Genomics Cytogenetics Committee Open Access

All conventional chromosome challenges (CY) from 2003 through 2022 were reviewed to identify those specifically addressing constitutional cytogenetic abnormalities. Three CY program mailings (A–C, 5 challenges per mailing, 15 total challenges per year) were provided annually from 2003 through 2014 (180 total challenges). Two CY program mailings (A and B, 6 challenges per mailing, 12 total challenges per year) were provided annually from 2015 through 2022 (96 total challenges). Of the 276 total CY challenges administered from 2003 through 2022, 161 (58%) were presented in the context of a constitutional study based on the provided clinical history and specimen type (Supplemental Table 1, see supplemental digital content at https://meridian.allenpress.com/aplm in the March 2025 table of contents). In addition, 9 of the 161 constitutional cytogenetic challenges consisted of multiple cases (4 cases each for 2003B-10, 2004B-10, 2006B-10, 2007B-10, 2009B-10, and 2011B-10; 3 cases each for 2008B-10 and 2010B-10; and 2 cases for 2013B-10), resulting in 184 total cases. A designated number of metaphase cells was provided per case (paper and/or electronic) to each enrolled cytogenetic laboratory (participants). Cases were chosen by the CyC to reflect chromosomal findings likely to be encountered during routine conventional chromosome analysis of constitutional cytogenetic studies, including cases with “normal” chromosome results.

From 2003 through 2008, 4 grading components of equal weight were used for each challenge, including modal chromosome number, sex chromosome designation, recognition of abnormalities, and karyotype nomenclature. In 2009, the grading components were consolidated to include M, S, and abnormalities under “recognition of abnormalities,” while maintaining “karyotype nomenclature” as a separate grading component (Supplemental Table 1). Grading was performed using the most current International System for Human Cytogenomic Nomenclature (ISCN) designated in the kit instructions for each challenge. Each of the 4 grading components (modal chromosome number, sex chromosome designation, recognition of abnormalities, and karyotype nomenclature) were assigned a grade of “1” (good performance), “2” (acceptable performance), or “3” (unacceptable performance). While a score of “2” may not be the ideal response, it is technically not incorrect and still considered “acceptable.” The “modal karyotype” is defined as the most common response provided by all participating laboratories; all participant responses were graded against the modal karyotype. If at least 80% of “referees” (15 randomly selected laboratories from a pool of anonymized laboratories with 100% performance on the CY program for the previous 3 mailing periods) responded with the modal karyotype, the CyC formally graded the challenge; challenges that did not meet 80% consensus were not graded. However, as there can be subjective morphologic interpretation of certain chromosomal abnormalities and/or various acceptable forms of ISCN designation, the CyC accepted multiple karyotypes in 14 cases rather than grading based solely on the modal karyotype. All participant data, as well as educational summaries by members of the CyC, were provided to participants in a printed document, the Participant Summary Report.

From 2003 through 2022 a total of 184 cases from 161 challenges (58% of all CY challenges) were presented in the context of constitutional cytogenetic disorders. The number of participants demonstrated an upward trend during this 20-year period and ranged from a low of 212 in 2003 to a high of 350 in 2018. The upward trend from 2011 through 2018 is mainly attributable to increases in international participation (countries outside of the United States and Canada). Fifteen referee laboratories were used per challenge, except for 2004A-5 (14 referees). The number of participants per challenge can be viewed in Figure 1.

For each of the 184 cases, the specimen source of the metaphase images (eg, amniotic fluid), in addition to a brief clinical case history, was provided (Supplemental Table 1). Of the 184 cases, 111 (60%) were peripheral blood specimens, 46 (25%) were amniotic fluid specimens, 21 (11%) were products of conception specimens, 4 (2%) were chorionic villus specimens, and 2 (1%) were skin biopsy specimens. The most common clinical information provided involved phenotypic or developmental abnormalities and/or intellectual delay or disabilities (81 of 184 cases, 44%), a family history of infertility or miscarriages (30 of 184 cases, 16%), fetal demise (21 of 184 cases, 11%), and advanced maternal age (18 of 184 cases, 10%).

Of the 184 total cases, 18 (10%) had a normal karyotype (11 female, 7 male). Sixty-three of 184 cases (34%) had a balanced or unbalanced reciprocal translocation (excluding inversions and whole-arm, Robertsonian, and insertional translocations) (1 case had both unbalanced and balanced translocations [2003A-3], 1 case had 2 balanced translocations [2003A-5], and 1 case had a 3-way translocation [2020B-8]). Thirty-one of 184 cases (17%) had aneusomies including autosomal trisomies (19 of 184 cases, 10%), autosomal monosomies (5 of 184 cases, 3%), monosomy X (11 of 184 cases, 6%), and an extra X chromosome (5 of 184 cases, 3%). Triploidy and tetraploidy were observed in 4 of 184 cases (2%) and 1 of 184 cases (1%), respectively. Thirty of 184 cases (16%) had an interstitial or terminal deletion, and 9 of 184 cases (5%) had a duplication. Twenty-seven of 184 cases (15%) had a pericentric (15 of 184 cases, 8%) or paracentric (12 of 184 cases, 7%) inversion. In addition, 9 of 184 cases (5%) had an isochromosome, 5 of 184 cases (3%) had an isodicentric chromosome, 4 of 184 cases (2%) had a Robertsonian translocation, 3 of 184 cases (2%) had a whole-arm translocation, 2 of 184 cases (1%) had a pseudo-isodicentric chromosome, and 1 of 184 cases (1%) had a dicentric chromosome. Lastly, 2 of 184 cases (1%) had a ring chromosome, and 1 of 184 cases (1%) had an insertional translocation (2010C-14). Importantly, several cases had more than 1 acceptable karyotype (eg, 2006C-11, 2012B-10C). In addition, the abnormalities described may have been observed in isolation (eg, 2003C-11) or in combination with other abnormalities (eg, 2020A-2).

Several recurring chromosomal abnormalities were included in the challenges from 2003 through 2022 (Supplemental Table 1). The 2 most common abnormalities included monosomy X (11 of 184 cases: 2006C-15, 2011B-7, 2011C-13, 2012B-9, 2012C-13, 2013B-9, 2014A-2, 2015A-1, 2015B-9, 2017B-7, 2020B-9) and trisomy 21 (7 of 184 cases: 2003B-6, 2003C-11, 2004C-12, 2007A-5, 2009B-9, 2009B-10A, 2014B-10).

From 2003 through 2008, the modal chromosome number and sex chromosome designation were independently graded for each challenge. None of the modal chromosome number or sex chromosome designation components (referee or participants) failed to meet the greater than or equal to 80% grading threshold from 2003 through 2008 (data not shown).

Recognition of abnormalities and karyotype nomenclature components were graded independently from 2003 through 2022. Of the 184 total cases, 2 failed to meet 80% consensus for abnormality identification by both referees and participants (2003B-10D [Figure 2, A through C], 2017A-3 [Figure 3]). The first case (2003B-10D) had a recombinant chromosome 5 resulting from a pericentric inversion of chromosome 5 from a carrier father. However, the parental history was not provided, thus there was uncertainty about reporting the abnormality as a derivative chromosome or a recombinant chromosome. Furthermore, the most up-to-date version of the ISCN at the time of this challenge did not provide guidance on how to report derivative or recombinant chromosomes. The second case (2017A-3) demonstrated an isochromosome 12p that was also incorrectly reported by some laboratories as isochromosome Xp10, 5p10, or 21q10.

While the remaining 182 cases were graded (referees exceeded the 80% consensus for abnormality identification), 4 cases (2006B-10D, 2017B-7, 2020A-2, 2020B-8) from the participant group were less than 80% for both recognition of abnormalities and nomenclature. For case 2006B-10D, only 75% (170 of 226) of participants correctly identified the interstitial 3q deletion. Many participants incorrectly described the abnormality as either a balanced pericentric or paracentric inversion involving chromosome 3. For case 2017B-7, 79% (265 of 336) of participants correctly identified the isodicentric (or pseudo-isodicentric) X chromosome. Many participants incorrectly reported the structurally abnormal X chromosome as a derivative from an unbalanced translocation with another X chromosome homolog. For case 2020A-2, only 77% (230 of 298) of participants correctly identified both inv(6)(q13q25) and trisomy 22. The majority of participants that failed this challenge did not identify the inv(6). Lastly, only 57% (181 of 318) of participants correctly identified the balanced t(4;5;11)(p15.2;q15;p13). The majority of participants that failed this challenge only reported a t(5;11).

Of the 184 total cases, 2 (1%) failed to meet the 80% consensus for karyotype nomenclature by both referees and participants (2006C-11, 2007C-11). For case 2006C-11, karyotype nomenclature [46,X,der(X;6)(q10;p10),+6 or 46,X,+der(6)t(X;6)(q13;q11) or 46,X,der(X)t(X;6)(p11.1;p11.2)] failed to meet consensus due to the designation of the centromere of the derivative chromosome. Referees and participants designated the derivative as a derivative chromosome 6, a whole-arm translocation, or a derivative X chromosome. For 2007C-11, only 73% (11 of 15) and 57% (140 of 245) of referees and participants correctly described the karyotype nomenclature, respectively. The description “+10,der(10)” was incorrectly employed by several participants and 2 referees. Although this was accepted for “recognition of abnormalities,” it was not acceptable for the karyotype nomenclature. In addition, the absence of brackets and the order of abnormalities was problematic for some laboratories. Lastly, 4 of 184 cases (2%) failed to meet the 80% consensus for karyotype nomenclature by the participants (2006B-6, 2011B-10A, 2013A-2, 2021A-1). Of these 4 challenges, 2 had an isochromosome 12p, 1 had a duplication, and 1 had a pericentric inversion (Supplemental Table 1).

Of the 184 total cases, 170 (92%) had a single acceptable karyotype, while 14 (8%) had multiple ISCN designations that were deemed acceptable by the CyC. Varying breakpoints were accepted for terminal and interstitial deletions (2003B-8, 2004A-3), a duplication (2008B-10A), a pericentric inversion (2011C-15), an unbalanced translocation (2009A-4), and balanced translocations (2013A-3, 2019A-1). The use of “idic” or “psu idic” was accepted for 2 cases (2012B-10C, 2017B-7), the use of “mat” once or twice in the same karyotype was accepted in 1 case (2003A-3) and the presence or absence of “mos” (indicating mosaicism) was accepted in 1 case (2015A-1). Three ISCN designations were accepted for a single case that each accurately described the abnormalities (2006C-11), and both i(12)(p10) and i(21)(q10) were accepted for 1 case (2016A-1). Lastly, the inclusion of a benign heteromorphism was accepted in a single case (2014B-10).

Throughout the 20-year review period, the number of laboratories participating in CyC proficiency surveys increased, peaking in 2018 with 350 laboratories participating (of note, the significant decrease in laboratory participation in early 2020 was presumably due to the COVID-19 pandemic). During this time many new technologies for genetic interrogation were developed and adopted for clinical use. Despite this, there has not been a noticeable decline in the number of CyC participants. This is an indirect indication that G-banded karyotyping remains a valuable diagnostic tool despite the emergence of newer cytogenomic technologies.

Out of 184 total cases, 2 (1%) did not meet the 80% grading threshold in referee and participant groups for both abnormality recognition and, by extension, karyotype nomenclature. The anticipated karyotype in case 2003B-10D was 46,XY,rec(5)dup(5q)inv(5)(p15.3q33.1) (Figure 2, A through C). Since the parental history for this patient was not provided, there was uncertainty on whether to designate the abnormality as a derivative chromosome or a recombinant chromosome. For this challenge, 43% of the participants assumed that there was a carrier parent. At the time, the ISCN 1995 did not provide guidance on how to report these 2 findings (derivative versus recombinant); the difference between them can only be distinguished with clinical information. Subsequent to the time this challenge was issued, ISCN has corrected this oversight to specify that recombinant (rec) should only be used when the clinical history is known to support an inherited abnormality. If this challenge were to be reissued as it was in 2003, based on ISCN 2020 nomenclature guidelines,⁴  describing the abnormality with “rec” would be incorrect. However, if the abnormality were to be described as a derivative, 2 ISCN designations would be acceptable: 46,XY,der(5)t(5;5)(p15.3;q33.1) or 46,XY,der(5)t(5;5)(p15.3;q33.1). The “underlining” convention, which is utilized to distinguish between homologous chromosomes, added further difficulty for participants. Using ISCN 2020 guidelines, 64 laboratories (6 referees, 58 participants) would have correctly identified the derivative chromosome, but only 13 (1 referee, 12 participants) used the correct “underlining” nomenclature to distinguish homologous chromosomes. While the ISCN 2020 guidelines included additional guidance for recombinant (rec) nomenclature, the underlining convention has not changed since ISCN 1995. As per ISCN 2020, the 2003 challenge should currently be appropriately described using “der” and using the “underlining” convention to distinguish homologous chromosomes. Given the complexity of this rearrangement, it would be interesting to see if this case would successfully reach the 80% grading cutoff with current CyC participants using today’s conventions.

The second case that did not reach the 80% grading threshold between both referees and participants, 2017A-3, was a mosaic case consisting of a late-term pregnancy loss with the modal karyotype: 47,XY,+i(12)(p10)[3]/46,XY[2] (Figure 3). Unlike 2003B-10D, the lack of consensus was not related to the question design. Most scenarios set forth by the committee are chosen to ask about a known phenotype or syndrome. In this case, the question was intended to query Pallister-Killian syndrome (OMIM 601803). However, the clinical history included intrauterine growth restriction, which is not usually associated with the condition. Confounding this, the metaphase cells provided to the laboratories had suboptimal morphology, which made reaching consensus difficult. Eighty percent of the referees and participants agreed there was an isochromosome and used correct nomenclature to describe the finding. However, participating laboratories were unable to agree that the isochromosome was derived from the short arm of chromosome 12. Describing Pallister-Killian syndrome adequately for CyC participants proved to be difficult in 2 separate challenges (2006B-6 and 2013A-2), as 80% consensus was not attained in either challenge. While not available for CY challenges, FISH or chromosomal array studies could be performed in the clinical setting to further characterize isochromosomes.

The timeframe examined by this study included 5 different editions of the ISCN. Proficiency testing provides useful but limited data on the impact of each new edition. In addition, previous challenges can be used to identify areas of the ISCN that require additional clarification such as the recombinant versus derivative chromosome question seen in 2003B-10D. From 2003 through 2008, the modal chromosome number and sex chromosome number achieved 100% consensus among participants. As a result, modal chromosome number, sex chromosome number, and recognition of the abnormality were all combined. This did not change the frequency of participant agreement, thus, in 2009 the grading criterion was simplified to 2 primary components: recognizing the abnormality and proper karyotype nomenclature. This simplified both grading and reporting of the results for participants.

The clinical scenario demographics have remained relatively similar for the past 20 years. Most cases feature patients referred for advanced maternal age or pregnancy loss, and there are a few cases with a clinical indication of autism and/or intellectual disability. Interestingly, the challenges were slightly skewed from what might be expected proportionally in clinical practice. Despite G-banded karyotyping being uniquely useful for identifying mosaicism, it made up only 12 of 186 cases (6%). In addition, monosomy X (Turner syndrome) has a lower birth incidence than 47,XXY (Klinefelter syndrome) or 47,XXX,⁵  but was represented twice as often as 47,XXY (11 versus 6 cases, respectively). There were 2 cases of XXX syndrome, one of which was a mosaic 45,X/47,XXX case.

Proficiency testing provides an opportunity to ensure that participants offering clinical testing have the appropriate skill set to provide optimal quality patient care. It also provides a mechanism for delivering continuing education topics to participants by means of the Participant Summary Report, supplemental questions and subsequent discussions, and ungraded CY-99 challenges. Seven total CY-99 ungraded challenges were provided (2008C, 2010C, 2013C, 2014C, 2016B, 2018B, 2021B) that addressed concepts pertinent to constitutional cytogenetics. Most of these challenges were composed of representative karyograms, additional cytogenomic studies (such as FISH), and a clinical history that requires integration of data and a deeper level of understanding compared to graded CY challenges. G-banded karyotyping set the gold standard as a first-tier assay for assessment of products of conception following spontaneous miscarriage, patients with intellectual disability and/or multiple congenital abnormalities, pregnancies with an abnormal screening result, and parents with repeated pregnancy loss. The challenges included in the CyC program aim to include cases that reflect these real-life scenarios.

While newer cytogenomic technologies are continually being adopted by many clinical laboratories, the steady state of CyC participants indicates the ongoing need for a cytogenetics proficiency program. The CyC committee routinely surveys participants for input on numerous practical laboratory issues and performs data analysis on responses as well as on historical data. In some cases, the data are published, as in this study. In other scenarios, the data are synthesized to provide feedback to the committee, which uses the information to improve future challenges and generate a robust cytogenetics proficiency testing program.