The Vanguard Has Arrived in the Clinical Laboratory

Abstract

This Commentary provides a state of the art for array-based karyotyping in cancer diagnostics.

Nuclear and mitotic abnormalities have been a hallmark of human cancer since the late 19th century when they were first described by German pathologists. However, the ability to visualize chromosome structures within these abnormal nuclei has only been possible since the middle of the 20th century when “squash and smear” cytogenetic techniques were adopted for analysis of malignant tissues.¹ Despite the low resolution of these crude methods, comparison of the chromosomes in malignant cells with those of healthy cells revealed them to be both morphologically and numerically abnormal even before the correct homo sapiens chromosome number was determined in 1956.² With the advent of chromosome banding techniques in the early 1970s, recurrent and tumor-specific chromosome abnormalities could be accurately identified in malignant cells and these recurrent aberrations have provided insight into the mechanisms of tumorigenesis as well as clinically applicable prognostic disease markers.³ Despite these decades of progress, the universal application of chromosome analysis in clinical oncology has been hampered by the difficulty involved in obtaining metaphases from most types of tumor cells in the laboratory and the relatively low resolution of chromosome banding techniques even when metaphases are successfully obtained.

With the recent clinical availability of fluorescence in situ hybridization (FISH) for the evaluation of structural and numerical chromosomal alterations in oncology cases, molecular cytogenetic analysis has become an integral part of the diagnostic and prognostic workup of many hematological malignancies as well as some solid tumors.⁴,⁵,⁶ Interrogation of tumor chromosomes using DNA-based FISH probes allows higher resolution analysis than conventional karyotyping, but provides only a very limited glimpse of the tumor genome. In addition, this analysis may not yield any clinically useful information when the selected probes interrogate regions outside of the genomic aberrations of the tumor.⁷

Array-based karyotyping platforms, also known as copy number arrays, have thus been developed over the past two decades as a method to circumvent the necessity of obtaining metaphases from tumor cells for subsequent analysis. Array-based testing methods merge the high-resolution locus-specific DNA-based analysis of FISH probes with a global genome view similar to that obtained by conventional karyotyping.⁸ Harkening back to the earliest days of tumor genome analysis when metaphases from malignant cells were compared with those from normal cells, the various array platforms currently used for clinical oncology analyses derive results by comparing hybridization intensity values of malignant (test) and normal (reference) DNA samples. In this issue of The Journal of Molecular Diagnostics, Hagenkord et al⁹ herald the continued advancement of array-based karyotyping into the clinical laboratory with the validation of a single nucleotide polymorphism (SNP) array platform for routine clinical use in the assessment of prognostic markers for newly diagnosed chronic lymphocytic leukemia (CLL). They describe the advantages and specific characteristics of this platform, which joins a growing list of validated arrays for clinical oncology applications, all but one of which were designed for prognostic assessment of CLL.¹⁰,¹¹,¹²,¹³,¹⁴

What has made CLL prognostic marker assessment such an inviting target for the application of array-based karyotyping in the clinical laboratory? In their article, Hagenkord et al⁹ make the point that CLL is actually the ideal neoplasm to analyze using copy number arrays because the clinically relevant genomic lesions in CLL are chromosomal gains and losses, instead of balanced translocations, which would be undetectable using current array technology. In addition, fresh CLL samples with verifiable high tumor burden can readily be obtained from peripheral blood and bone marrow specimens.⁹ In addition to these practicalities, perhaps more important is the long-standing practice of assessing prognostic markers in CLL because of its presentation as a microscopically homogeneous and clinically heterogeneous disease in which patients with morphologically similar abnormal lymphocytes can have dramatically different disease courses.¹⁵ Therefore, evaluation of recurrent chromosomal abnormalities with prognostic significance in newly diagnosed CLL has become one of the major criteria for patient risk assessment, making it imperative that tumor genome markers be evaluated completely and accurately. Although recent studies have revealed the inadequacies of conventional cytogenetics and FISH panels for evaluation of the CLL tumor genome,⁷ the vast majority of newly diagnosed cases of CLL are still evaluated using these methods.

Even though array-based karyotyping methods have been available for almost a decade, oncology applications for this technology in the clinical laboratory have lagged behind applications for identification of congenital chromosome abnormalities. This is probably because of the unavoidable genomic heterogeneity of hematological malignancies and solid tumor samples. In the analysis of constitutional samples, generally every cell in the sample contains the genome of interest, whereas tumor samples may contain multiple clones with slightly different genomic changes and also may include contaminating cells from surrounding normal tissue. To circumvent the problem of genomic heterogeneity in the research laboratory, the vast majority of early array-based karyotyping studies of tumors were performed using DNA extracted from homogeneous tumor cell lines. Although this approach has added to our collective knowledge of tumor biology, it has provided few solutions for the challenges of handling heterogenous tumor samples in the clinical laboratory. Fortunately, the challenges associated with performing tumor-targeted DNA extractions from oncology samples are far from insurmountable obstacles. For example, many hematological malignancy samples are routinely evaluated by flow cytometry and, if necessary, enrichment strategies can be used to increase the percentage of abnormal cells as described by Hagenkord et al.⁹ Solid tumors can be enriched for tumor cell populations by marking tumor cells on H&E slides for microdissection, macrodissection, or laser-capture dissection.¹⁴

Another probable reason for the slow adaptation of array-based karyotyping for oncology samples is the ubiquitous use of formalin-fixed paraffin-embedded methods for preservation of solid tumor tissue. Array-based karyotyping results obtained using DNA extracted from formalin-fixed paraffin-embedded tissue are generally are less satisfactory than those obtained from fresh or frozen samples. However, clinically meaningful results can also be obtained from formalin-fixed paraffin-embedded tumor samples, and given that the use of this preservation method in the pathology laboratory is unlikely to change anytime soon, it is fortunate that acceptable methods for clean extraction of DNA from formalin-fixed paraffin-embedded samples are becoming more widely available.¹⁶ In the meantime, hematological malignancy samples are readily available for direct DNA extractions in the clinical laboratory, and this is most likely a major reason why these samples are among the first to be analyzed by array-based karyotyping.

New platforms for whole-genome scanning in hematological malignancies are becoming increasingly available, and this availability will most likely lead to the continued implementation of this technology into the clinical laboratory.¹⁷ Among these platforms are comparative genomic hybridization arrays using either bacterial artificial chromosome or oligonucleotide probes¹⁸ and SNP arrays. Hagenkord et al⁹ chose SNP arrays for their validation study because this platform is able to simultaneously determine ploidy status, locus-specific copy number changes, and copy number neutral events such as loss of heterozygosity. Copy number neutral loss of heterozygosity is of particular clinical interest in oncology owing to the growing realization that “second hit” deactivation of tumor suppressor genes such as TP53 and RB may be caused by acquired uniparental disomy, an alteration that would be not be detectable by comparative genomic hybridization arrays.⁹ SNP arrays are thus an excellent and extremely precise method for analysis of hematological malignancies for which acquired uniparental disomy of specific genes such as TP53 could have a significant impact on treatment decisions. It should be noted, however, that even when optimal DNA extraction methods are used, SNP platforms have a lower signal/noise ratio than array comparative genomic hybridization platforms and may be most clinically applicable when fresh tissue samples are available.

The argument has been made by some that the new array-based tests may represent a scenario in which “diagnostics have outstripped therapeutics” and that detailed information about a given patient's tumor genome is not clinically useful or even desirable. However, the counter argument is that a more complete understanding of each patient's tumor genome can assist clinicians with diagnostic issues and prognosis for the individual patient and at the same time add to the collective understanding of human cancer. In an age when oncology patients are expecting a more personalized approach to their individual disease, array-based karyotyping will become an indispensable tool in the clinical laboratory for identification of prognostic and predictive markers in hematological malignancies as well as solid tumors.