Clinical Laboratory Reports in Molecular Pathology...est Name: Fragile X genotyping rocedure: Southern blot analysis was performed using EcoR1 and methylation-sensitive Eag1 restriction

Clinical Laboratory Reports in Molecular Pathology

Margaret L. Gulley MD, Raymond R. Tubbs, DO, Rita M. Braziel, MD, Kevin C. Halling, MD, PhD, Eric D. Hsi, MD, Jan A. Nowak, MD, PhD, Andre Oliveira, MD, Herbert F. Polesky, MD, Vivianna M. Van Deerlin, MD, PhD, Gail H. Vance, MD, James Versalovic, MD, PhD, Marina N. Nikiforova, MD, Lawrence Silverman, PhD, for the College of American Pathologists Molecular Pathology Resource Committee Corresponding Author: Margaret L. Gulley, MD Dept of Pathology Brinkhous-Bullitt Building University of North Carolina Chapel Hill, NC 27599-7525 Telephone: 919-966-4676 Fax 919-966-6718 [email protected] Abstract Molecular pathology is a rapidly growing area of laboratory medicine in which DNA and RNA are analyzed. The recent introduction of array technology has added another layer of complexity involving massive parallel analysis of multiple genes, transcripts or proteins. As these technologies are increasingly implemented in clinical settings, it is important to bring uniformity to the way that molecular test results are reported. This article describes elements that are already common to virtually all molecular pathology reports, as set forth in the checklists used by the College of American Pathologists as part of the laboratory accreditation process, as well as new recommendations on which public discussion is welcome in order to gain consensus prior to checklist revisions. Resources to insure proper gene nomenclature are described, including methods of describing mutations, translocations, microsatellite instability, and other genetic alterations related to inherited disease, cancer, HLA typing, identity testing, and microbiology. These guidelines provide a framework for writing patient reports so that the results and their clinical significance are more easily understood by members of the healthcare team. The clinician who orders a molecular test expects the pathology report to state what test was done, the method that was used, and the result of the test. Many clinicians find it helpful if the report further describes what the result means for their patient based on analytic and clinical performance characteristics of the test in correlation with the clinical setting in which the test was done. This is true regardless of whether it is a stand-alone molecular pathology report or part of a

9-6-05, Gulley Draft page-

1

larger surgical pathology or hematopathology report in which molecular data is integrated with morphology and other tests done on the same tissue.1

The College of American Pathologists (CAP) has developed recommendations for composing laboratory reports. Many of these guidelines are outlined in the checklists that are used by inspectors to evaluate compliance as part of the laboratory accrediation process. The Laboratory General checklist summarizes guidelines that are common to every laboratory report, while the subspecialty checklists such as those for Molecular Pathology, Microbiology, Cytogenetics, Anatomic Pathology, and HLA contain additional elements specific to the molecular testing done in these specialty areas of the clinical laboratory. These checklists may be downloaded from the CAP website, www.cap.org. Some of the items are federal requirements, while others are recommendations. In this report, we also make suggestions that extend beyond those in the checklist.

Content of Molecular Pathology Reports

The recommended elements of a molecular pathology report are shown in Figure 1. The report should include fundamental information about the patient, the specimen, the ordering physician, and the laboratory where testing was done. Reports from an outside laboratory must be made available to the clinician, and they may also be transcribed into the local information system either verbatim or as summarized by a laboratory professional who is knowledgeable in the field.

Laboratory information *Name and address of Testing Laboratory (optional: phone, FAX, e-mail,

website) Demographic and identifying information *Patient’s name (first and last with middle initial or middle name) *ID# and/or date of birth Accession number, and Specimen number from referring lab *Date of specimen collection (and time, if appropriate) *Date of receipt in laboratory *Date of report *Sample type (even if only one sample type is accepted) and how tissue was

received (fresh is assumed unless designated as frozen, paraffin-embedded, etc.)

*Ordering physician, when appropriate Clinic/ Inpatient Location; or Name/ address/ phone of outside facility Genetic Counselor Pedigree diagram and/ or family history Indications for testing

9-6-05, Gulley Draft page-

2

Pertinent clinical history (clinical setting, ethnicity/ race, previous molecular / genetic studies and other relevant laboratory findings, etc)

Results: *List by procedure; use standardized nomenclature for genes/ proteins

(www.gene.ucl.ac.uk/nomenclature/) and standard units of measure. *Reference range; or normal versus abnormal. Interpretation: Analytic and/ or clinical interpretation of results: Analytic interpretation involves

synthesizing raw data to produce a reportable result. Clinical interpretation involves synthesizing analytic and clinical information to describe what the result means for the patient.

Comments: Significance of the result in general or in relation to this patient. Recommend additional measures (e.g. further testing, genetic counseling). Unusual circumstances relating to specimen or test conditions (e.g. sample

received thawed, partially degraded DNA). Cautions regarding assay performance characteristics. Residual risk of disease or carrier status by Bayesian analysis. Control test results, if unusual or especially pertinent. Cite peer-reviewed medical literature or reliable web sites on the assay and its

clinical utility (e.g. educational materials on genetests.org). Document to whom preliminary results, verbal results, or critical values were

reported and when. Chain of custody documentation, if needed. *If the report is an amended or addendum report, describe the changes or

updates. *Describe discrepancies between preliminary and final reports. *Name of testing laboratory, if transmitting or summarizing their results. Procedure: *Type of procedure (Southern blot, PCR, rtPCR of RNA, Q-PCR, CISH, FISH,

gene dosage array, expression array, sequencing, protein truncation test, etc.)

*Defined target (i.e. name of analyte tested such as gene, locus, or genetic defect; use standardized gene nomenclature)

Pertinent details of procedure, e.g. analyte-specific reagent or kit version & manufacturer, instrument type

Signature and printed name of reporting physician, for any test having a

physician interpretation. Signature of Lab Director or designee when interpretation is performed.

(Reports may be signed electronically.)

9-6-05, Gulley Draft page-

3

Disclaimer on non-FDA approved tests in which a commercial analyte-specific reagent was used.

Billing information:

ICD9 code (clinical indication for test)

CPT codes (laboratory procedures performed) *At minimum, reports should contain this information. Figure 1: Molecular Pathology Report Content

to be , if

e e accurately tracked. It is useful to cite a

literatu

A), A),

). rse

ll out the words or use the term “Q-PCR” to designate quantitative PCR and “Q-rtPCR of RNA” to designate quantitative

or

that

Methodology It is important to specify the gene or locus tested and the method used

analyze it, particularly since there are so many different methods that can used. Describe which commercial kit or analyte-specific reagent was usedany, since this information imparts some information on the performance characteristics of the assay. Furthermore, when serial testing is done, for example to measure viral load or to monitor tumor burden, it is advisable to usthe same method so that trends can b

re reference or website that provides analytic and clinical performancedata relevant to the test procedure.

It is recommended that abbreviations not be used in clinical reports. However, abbreviations are sometimes required because of limitations in the number of characters allowed in a computerized database. Furthermore, certain abbreviations are universally understood such as deoxyribonucleic acid (DNcomplementary DNA (cDNA), ribonucleic acid (RNA), messenger RNA (mRNfluorescence in situ hybridization (FISH), chromogenic in situ hybridization (CISH), polymerase chain reaction (PCR), base pair (bp) and kilobase (kbSince confusion exists about whether rtPCR stands for real-time PCR or revetranscriptase PCR, either spe

reverse transcriptase PCR.

Results and Interpretation The result is the crux of every report. The result should be stated clearly,

and should be based on objective criteria that are described in the procedure manual. Many molecular tests yield raw data which must be interpreted before the result is evident. For example, bands on a gel must be evaluated visually scanned to yield an interpretable electropherogram. Quantitative amplification assays may require extrapolating from a standard curve and interpretingresult in the context of the amount of control sequence that was amplified fromthe same sample. Interpretation criteria depend on whether the control sequence was an exogenous DNA or RNA that was spiked in known amount

9-6-05, Gulley Draft page-

4

versus an endogenous control sequence that is expected to vary (e.g. when a patient

d ds

ecific knowledge of that patient’s situation. An example of the difference between analytic and clinical interpretation is shown in the report in F

’s white blood cell count is low then the level of DNA or housekeeping RNA in the blood might also be low).

There are two kinds of interpretation, analytic interpretation and clinical interpretation. Analytic interpretation involves examining the raw data anforming a conclusion about the quality or quantity of the analyte, in other worproducing a reportable result. Clinical interpretation involves forming a conclusion about what the result means for the patient, either in general, or based on sp

igure 2. Patient Name: John Doe Date of Birth: 1/2/2003 Testing Laboratory: (Lab Name, phone #, website) Date of Collection: 1/10/05 Date received: 1/10/05 Date of Report: 1/17/05 Ordering physician: John Smith, M.D. Sample type: Blood

est Name: Fragile X genotyping rocedure: Southern blot analysis was performed using EcoR1 and methylation-sensitive Eag1 restriction enzymes followed by hybridization with the StB12.3 probe ta

TP

rgeting the FMR1 gene on chromosome Xq27. This region contains a trinucleotide CGG repeat sequence with a range of 5 to approximately 44 repeats expected in the normal population.

Result: This patient has an abnormally large FMR1 gene segment, the size of which is in the “full mutation” range estimated as more than 200 trinuclotide repeats.

Interpretation: Analytic interpretation: Fragile X genotype. Clinical interpretation: The findings are consistent with a diagnosis of fragile X syndrome of mental retardation.

Comment: Genetic counseling is recommended. For further information see www.geneclinics.org and www.acmg.net. isclaimer: This test was developed and its performance characteristics determined by the Molecular Genetics Laboratory. It has not been cleared by the US Food and Drug

wever, such approval is n

D

Administration. Ho ot required for clinical implementation, and test results have been shown to be clinically useful. This laboratory is CAP accredited and CLIA certified to perform high complexity testing. CD9 code: 315.2 I

CPT codes: 83891, 83896, 83892x2, 83894, 83897, 82397, 83912 Signature: Electronically signed by (Signature of Pathologist or Designee) Figure 2: A sample molecular report is shown. The interpretation section is divided into analytic and clinical portions for purposes of illustration.

The expected result in a normal individual should be stated. This is expressed as a normal range for a quantitative test, along with the units ofmeasurement. Interassay reproducibility is an important parameter that aclinician to determine whether two sequential test results are significantly different fr

llows a

om each other, and information about assay precision can be convey o the ed in the individual report or in a document that is readily accessible t

9-6-05, Gulley Draft page-

5

clinician. Most qualitative assays have a clearly defined normal and abnormal

outcome, although further explanation is sometimes required to interpret whetha base change is a disease-related mutation, a benign polymorphism, or a variant of undetermined significance. Likewise, it is useful to specify whether detection of a microbe is interpreted as pathologic or normal flora. Some tests

er

have n ing le.

ing

t

important

on on e

g

sults should be interpreted with caution since our labora

ual

o normal range, such as parentage testing or HLA-based organ matchwhere the result is interpreted in concert with results on another person’s samp

Most laboratory tests have limitations that make them subject to false positive or false negative results, thus imparting a probabilistic risk of disease rather than an absolute answer. Depending on the performance characteristics of each assay and the clinical impact of an incorrect result, it is worth describthese limitations in the report. Thus, some molecular pathology reports routinely include verbiage on assay sensitivity, specificity, and/ or precision. Analytic sensitivity refers to how good the assay is at detecting a specific molecular targeor defect, while clinical sensitivity refers to how good the assay is at detecting the associated disease. When multiple genotypes can cause the same phenotype, or when the assay is not 100% sensitive for detecting the disease, it isto describe the residual risk of disease in the report. As an example, validationstudies may show that a PCR assay is 100% sensitive for detecting a translocation involving the CCND1 major translocation cluster (MTC) locus on chromosome 11q13 and the immunoglobulin heavy chain gene joining regichromosome 14q32, as long as the tumor clone composes at least .001% of thcells in the specimen. Yet the PCR assay is only about 50% sensitive for detecting mantle cell lymphoma since about half of all mantle cell lymphomas that were tested had breakpoints outside of the MTC region. When reportinnegative molecular results in a marrow with a morphologic differential diagnosis of mantle cell lymphoma, the report might state that “no IGHJ/CCND1 was detected to a sensitivity of approximately 1 in 100,000 marrow cells; although noIGHJ/CCND1 translocation was detected, this does not exclude mantle cell lymphoma since only about half of all mantle cell lymphomas have a detectable translocation”. When testing for minimal residual disease in a patient whose original tumor was never assayed for this translocation, the report of a negative test might add that “these re

tory has no record that this patient’s original tumor was characterized by a detectable translocation”.

Complex genetic test reports should include information about the residrisk of disease. The British Clinical Molecular Genetics Society has published Best Practice Guidelines for performing and reporting molecular genetic test results on their website, www.cmgs.org/new_cmgs. . The American CollegeMedical Genetics has also devised guidelines for clinical genetics laboratorie(www.acmg.net

of s

). The Clinical and Laboratory Standards Institute (formerly NCCLS) has published several guidelines for molecular testing that include recommendations for report content. (www.nccls.org) (ref clsi)

9-6-05, Gulley Draft page-

6

A report of a negative screening test for cystic fibrosis carrier status might ie

T

nclude the wording shown in Figure 3, which relates the molecular findings to thnicity.

est Name: Cystic Fibrosis Carrier screen esult: Negative for 32 CFTR gene mutations.

nterpretation: There is no detectable CFTR gene mRI utation, decreasing the probability that this patient is a carrier of cystic fibrosis. omment: This test is estimated to detect cystic fibrosis carrier status in 88% of Caucasians of Northern European descent, in 94% of people of Ashkenazi Je

Cwish ancestry, in 64% of African

Americans, and in 72% of Hispanic Americans. A negative result does not eliminate the possibility of a CFTR mutation not tested for in this panel. (Watson et al. Genetics in Medicine;2004 6: 387-391; www.geneclinics.org). rocedure: The CFTR gene was analyzed forP the 32 mutations listed below by polymerase chain reaction and allele specific oligonucleotide ligation (Abbott Laboratories CF v3.0). [Each laboratory should list the specific mutations tested for here.]

Figure 3: Partial report of a negative cystic fibrosis screen. See “Ogino et al, submitted” for recommended gene nomenclature for CFTR mutations, and refer to the ACMG guidelines for calculations of residual risk by ethnic group.

Every laboratory must do validation studies to show how well each asperforms in their own hands. Detailed information on assay performanccharacteristics should be made available upon request. Some laboratories routinely display such

say e

information on their web site, as well as information on indicat

14-P,

s lleagues in coagulation laboratories who recently showed that most

orderin

ion-

dical .

ding of

ions for testing, sample requirements, requisition forms, shipping information, and literature references on analytic and clinical utility. Further guidance on new test validation is found in the CAP checklists and in a publication by the Association for Molecular Pathology.2(ref NCCLS mmCDC QA/QC effort)

Because molecular tests are relatively complex and may be done using any variety of methods, interpretation is recommended to insure that the clinical significance of the result is apparent and that the analytic limitations of the assayare disclosed in relation to the findings of that case. Assays should be interpreted by individuals who are competent in the analytic and clinical aspects of the test.3 A lesson regarding interpretation of complex laboratory data comefrom our co

g clinicians perceive that pathologist-generated test interpretation saves them time, assists with differential diagnosis, and helps prevent misdiagnosis.4Clinicians who were surveyed about cystic fibrosis and factor V Leiden reports preferred comprehensive molecular reports that helped guide clinical decismaking.5

Interpretations should combine expert technical knowledge with meexpertise to create a report that synthesizes complex laboratory informationClinical interpretation further offers patient-specific analysis and understanthe laboratory findings. Clinical interpretation of a molecular test result is generally billable by a pathologist (or other physician) using CPT code 83912-26 (interpretation and report) or, when a consult of an abnormal test result is

9-6-05, Gulley Draft page-

7

requested by the patient’s attending physician, using codes 80500 (clinical pathology consultation, limited) or 80502 (comprehensive consultation of comple t

n s

r

e

ort

r reports

a and

ce-based laboratory medicine are summ tory

x diagnostic problem). Analytic interpretation is billable by a pathologisusing code 83912-26 except for certain microbiology tests for which interpretatiois already included in the molecular test code for that organism. Physicians awell as non-physician scientists who interpret and report molecular results can bill the technical component of 83912.

If general comments are automatically inserted into every report, theyshould be brief and educational with regard to test selection, test limitations, oclinical significance. Alternatively, reference a website or publication whergeneric test information can be found. Too much verbiage is burdensome to theclinician who wants to quickly evaluate laboratory findings and their significance. Clinicians seem to prefer a summary of the findings at the top of the repfollowed by a concise explanation of details.(Ref Karen Titus, Clinicians Talk Shop- Here’s what they want. CAP Today, 17(9):1-82, 2003) Too much verbiage may be a liability if actionable findings are hidden in a sea of words. It is important to communicate with clinicians who are reading youto get feedback on their content and utility. The shaded terminology of pathologists often contrasts with the preferred terms of surgeons.(Ref6) But clinicians must keep in mind that we can only report what we know to be true based on objective results and our expert technical and medical knowledge, and then we interpret from there using terms such as “consistent with”. The principles of evidence-based medicine guide our interpretations, thus requiringfundamental knowledge base combined with technical expertise, experience, medical judgement. The principles of eviden

arized at the International Federation of Clinical Chemistry and LaboraMedicine website, www.ifcc.org. This website also contains a database of systematic reviews and meta-analyses to facilitate medical decision-making. Undoubtedly the molecular aspects of this database will continue to grow as additional clinical research is forthcoming.

The difficulty of composing a comprehensible report was evident in a study entitled “Clinicians are from Mars and Pathologists are from Venus” in which surgeons misunderstood anatomic pathologist’s reports 30% of the time.The gap can be bridged by personal communication, education of both the report writer and the reader, and attention to guidelines of report composition. More experienced or subspecialist clinicians may have different perceptions of report content than do generalists or junior colleagues. In general, it is advisable to write each report so that it is

7

urses,

liable ation is available. For inherited disease

sting, consider referencing www.geneclinics.org

easily understood by a non-specialist physician. Keep in mind that the report may be read by non-physicians ranging from nto genetic counselors to epidemiologists or trainees. It is reasonable to include references to published literature, to your own laboratory website, or to reexternal websites where further informte , www.ncbi.nlm.nih.gov/omim, or www.acmg.net. For cancer testing, useful information is often found at www.infobiogen.fr/services/chromcancer, and at http://cgap.nci.nih.gov. And for many types of laboratory tests, technical and clinical information is often

9-6-05, Gulley Draft page-

8

available at www.labtestsonline.org. Reporting Control Test Results Nearly every clinical laboratory test is run alongside external and internal controls. External controls are used to insure that the assay performed as expected, while internal controls are used to assess quality or quantity patient sample. Because controls are required by US federal law and their outcomes are assumed to be within required limits in order for the result to bconsidered valid, the results of controls are not generally reported. Howevesome cases a control result is considered worthy of mention if it is unusual or pertinent to interpretation of the findings. For example, if morphologic examination shows that the proportion of tumor is low but not so low as to beunacceptable for analysis, it is appropriate to comment on marginal sample adequacy. Likewise if a control RNA is shown to be partially degraded by quantitative rtPCR but is still within acceptable limits, then it is appropriatmention that assay sensitivity for detecting a low-level transcript may be reduced. In bone marrow engraftment assays where donor and recipient controls are normally used to assess the degree of engraftment in an allogeneic transplant patient sample, if only one of these two controls is avail

of the

e r, in

e to

able for comparison then cautionary comment is warranted regarding the limitations of the DNA

of a fetal sample is often

ly that each testing laboratory here

sting

required to ely

nsus.

le fits

afingerprinting result. As a final example, genetic testingaccompanied by an identity test to show what proportion of the sample is of fetal versus maternal origin, with additional genetic testing of one or both parents to help interpret the significance of the fetal test results. Units of Measurement and Standardized Reagents For some molecular pathology tests, it is likehas a unique protocol that is different from nearly every other laboratory. Tare few kits, instruments, or standardized reagents available that would help insure that the same result and interpretation is achieved across multiple telaboratories. FDA-approved molecular products are listed on the Association for Molecular Pathology website, www.ampweb.org. The paucity of standardized reagents and procedures comes about because it is expensive to produce and market high quality products, and few comparative studies have been done to show which methods or products achieve better clinical outcomes. More clinical research investment issort out which methods are superior, recognizing that multiple methods are likto rise to the top while a few others may fall by the wayside. Professional societies can help by bringing laboratorians together to encourage conseFor example, HIV viral load testing benefits from a universal standard reagent that is available and agreed upon as a reference for quantitative HIV measurement, and we must likewise encourage the production of affordabstandards for all of the analytes on our menu. HIV viral load testing also benefrom abundant prior research that provided solid data on which to base recommendations regarding which sample type is best (plasma), and which unitsof measurement to use in reporting HIV viral load results (copies per ml). Many

9-6-05, Gulley Draft page-

9

other molecular tests have not yet achieved such consensus. Until more comprehensive research datasets are available on which to base consguidelines, many laboratories continue to use their own unique testing protocolsand reporting schemes that are validated in-house. It is unreasonable to expect each laboratory to compare their in-house protocol to every published variant to insure that theirs is superior, but it is prudent to show that a given method is at least equivalent to any FDA-approved method. The end result of non-standarization is a lack of consensus that often requires that a given patient serially monitored by the same testing laboratory, thus establishing a monopoly of sorts that impedes progress and restricts access to healthcare. It is

ensus

be

essential at more clinical research be done to answer the many outstanding questions

est and to establish affordable standardized reagents that can be ted

lude…”.

Amend on for

r changes in the way that an xisting test is done or reported, should be explained in a memo to clinicians

n your la

irate,

the findings. In

clinician, to

site

thabout each molecular t

used to facilitate interlaboratory consensus. In the area of inheridisease, the CDC is already making strides in identifying reagents and cell lines that can be used for validation, proficiency testing, and controls.(ref) Changes to Reports

An “addendum” report generally adds new information, while an “amended” report changes information that was previously reported. Addendumreports often contain a phrase such as “this report was updated to inc

ed reports should describe the change and, in many cases, the reasthe change. Likewise, discrepancies between a preliminary report and a final report must be documented. It is advisable to personally contact the ordering physician if the change is likely to impact clinical management, and to discuss these and other problematic cases at interdisciplinary conferences.

The introduction of a new test, or majoewhich is subsequently made available by posting in an accessible site (e.g. o

boratory website). It is advisable to get feedback from key clinicians prior to making major changes. In some cases, it is appropriate to emphasize the change within the text of the patient report. Integration of Molecular Tests into Anatomic Pathology Consultations

When molecular testing is performed on a sample that is also undergoing routine anatomic pathology consultation, such as a biopsy or a marrow aspthe option exists to report the molecular findings in a freestanding report or to integrate the molecular findings with the other clinicopathologic data into a single anatomic pathology report. Regardless of whether the molecular result is reported separately or not, it is important that a physician reviewgeneral, it is the responsibility of the physician who selects and orders an ancillary test, whether it be the consulting anatomic pathologist or thesynthesize the test results with other pertinent clinicopathologic information. This synthesis is above and beyond any work performed by the molecular pathologist in the analytic and clinical interpretation of the molecular test. Guidelines for anatomic pathology reports are summarized on the webof the Association of Directors of Anatomic and Surgical Pathology,

9-6-05, Gulley Draft page-

10

www.adasp.org. In the past, there was little consistency in the way that anatompathology reports were written. In fact, the process of standardizing reports can be surprisingly difficult, as was evident when the American Cancer SocCommission on Cancer recently adopted reporting guidelines created by the CAP.(Ref Anne Paxton, Cancer protocols: leaner, later, more lenient, CAP Today, 18: 58-66, 2004) These guidelines, which represent a list of essential elements in pathology reports for cancer-directed surgical resection specimens, were initially perceived by some pathologists to be confusing until the recommendations were refined and the flexibility of reporting

ic

iety

styles was assured. any institutions are trending towards synoptic cancer reports in which checklists

lates are used to standardize reporting of the essential elements and ted,

les

r on

de to er

learning difficulty) but the code may be revised to 759.83 (Fragile syndrome) once the positive test result is correlated with the findings on

cument the analytic procedures that are …

ults should be shared with

ere it

ption to their standard policy or

Mand temptheir sequence. McGowen et al recently showed the superiority of a formatsynoptic anatomic pathology report. (Ref Tom McGowen) The same prinicipand benefits are pertinent to molecular pathology reports. Coding The pathology report is often used by administrators and third-party payers to justify payment for services rendered. The ICD9 code is used to document the reason(s) for testing. It is the responsibility of the ordering physician to use sound judgement in deciding which laboratory tests to ordetheir patient. Laboratory professionals often provide crucial advice on which test(s) to perform in a given situation. Payers, in turn, may use the ICD9 codetermine if a test request is legitimate. It is acceptable for a pathologist to altan ICD9 code once the test result reveals a more specific diagnosis. For example, fragile X genotyping may be ordered because of ICD code 315.2 (developmental Xphysical examination. CPT codes doperformed in the laboratory. A list of molecular pathology test codes is found(Ref, I could no long find this on the Association for Molecular Pathology website,any ideas??). Reporting Research Test Results Research test results should not be included in clinical reports. If an assay is not yet validated for clinical use, then test resa subject’s physician only in the context of an IRB-approved study protocol. Consider reporting unvalidated test results in a spreadsheet or in a letter whis clearly stated that the test was done on a research basis and is not for clinical use. Just because an assay is validated for use on one sample type does not imply that the assay is valid on other sample types. There are some grey areas that warrant the discretion of the professionalsinvolved, analogous to the off-label use of a drug. An individual laboratory director or clinician can choose to make an exceto take risks that are perceived to benefit an individual patient, for example when an orphan disease is involved and there is no option to send the sample out for validated testing. But such exceptions should be handled with caution because

9-6-05, Gulley Draft page-

11

federal law defines minimal standards that must be met by clinical labs, and the public expects that relevant laws are followed. Clinical trials often include laboratory testing in the research protocol, and

patient management based on the results of these tests.

ed,

h s worldwide, sometimes after years of concurrent research, it was

that we would end up with multiple names for the same gene. It is to have multiple names for the same gene, so it was equally inevitable ould need to settle on a single name and develop a uniform format for e gene name. Do you know which of the following gene symbols is

3. bcl2

ith basic and clinical scienti

g

he

t

ek ,

ctuation is permitted within a gene name (except for HLA and an

of the HLA System;

the clinician may alter Any such tests that are meant to be used for patient management, even as part of a clinical trial that is paid for by the trial sponsor, must be validated, performand reported by a CLIA-certified clinical laboratory. Gene Nomenclature Because genes were discovered and named by many different researclaboratorieinevitableconfusingthat we wwriting thcorrect? 1. bcl-2 2. Bcl-2

4. Bcl2 5. BCL2 6. BCL2 7. None of the above

The immunology community survived and thrives two decades after developing the Cluster Designation system for naming surface antigens on hematopoietic cells. Another precedent dates back to the 1930s when it was recognized that the ABO blood group antigens needed consistent nomenclature. We in clinical molecular laboratories are obligated to work towards consensus nomenclature in our field. This requires collaboration w

sts from all areas of medicine to insure consistency across disciplines.Widespread consensus is not always achieved as evidenced by our persistent disagreement on an International System of Units (SI). It is timely and important to patient care that we now attempt to move forward with a system for reportinmolecular pathology results in a systematic fashion.

According to the HUGO Human Gene Nomenclature Committee, the correct answer to the question above is BCL2. Gene symbols are in upper caselatin letters +/- arabic numerals. Gene symbols are italicized (BCL2) whereas tprotein is not (BCL2). To distinguish between mRNA and cDNA the relevanprefix is shown in parentheses: (mRNA)BCL2 or (cDNA)BCL2. No superscripts or subscripts are accepted. No roman numerals (I, II, III, IV, V, etc.) or greletters (α, β, γ, ... κ, λ) are accepted. Oncogenes should have no prefix (MYCnot c-MYC). No pun

tigen receptor genes where hyphens are permitted; HLA alleles are assigned by the WHO Nomenclature Committee for FactorsImmunoglobulin and T cell receptor genes are named by the International

9-6-05, Gulley Draft page-

12

ImMunoGeneTics Information System (IMGT) Nomenclature Committee, http://imgt.cines.fr).

A searchable database of correct gene names and symbols is found othe website of the Human Gene Nomenclature Committee,

n

www.gene.ucl.ac.uk/nomenclature. Selected gene symbols associated with leukemia, lymphoma, and sarcoma are shown in Table 1. The correct gene name is not necessarily the same as the common name used the medical

yelogenous leukemia, and ETV6/RUNX1 (not TEL/AML1) is the common found in childhood acute tic leukemia.

literature. Note that ABL1 (not ABL) is the gene associated with chronic mtranslocation lymphoblas

Table 1: Common translocations in hematopoietic neoplasia and sarcoma Tumor Type Karyotype Genes in karyotpic order Myeloid leukemias CML and pre-B-ALL t(9;22)(q34;q11) ABL1/BCR AML-M2 t(8;21)(q22;q22) RUNX1/CBFA2T1 AML-M3 t(15;17)(q24;q21) PML/RARA AML-M3, atypical t(5;17)(q35;q21) NPM/RARA AML-M3, atypical t(11;17)(q23;q21) PZLF/RARA AML-M3, atypic t(11;17)(q13;q21) NUMA/RARA al AML-M3, atypic t(17;17)(q21;q21) STAT5B/RARal A AML-M4eo inv16(p13;q22) MYH11/CBFB CMML t(5;12)(q33;p13) PDGFRB/E TV6 Eosinophilic leukem t(4;4)(q11;q11) FIP1L1/PDia GFRA B cell leukemias and lymphomas pre-pre-B ALL t(12;21)(p13;q22) ETV6/RUNX 1 pre-pre-B ALL t(1;19)(q23;p13) PBX1/TCF3 Burkitt lymphoma t(8;14)(q24;q32) MYC/IGH Burkitt lymphoma t(2;8)(p12;q24) IGK/MYC Burkitt lymphoma t(8;22)(q24;q11) MYC/IGL mantle cell lymphoma t(11;14)(q13;q32) CCND1/IGH follicular lymphoma t(14;18)(q32;q21) IGH/BCL2 Diffuse large B cell lymphoma t(3;14)(q27;q32) BCL6/IGH Lymphoplasmacytic lymphoma t(9;14)(p13;q32) PAX5/IGH MALT lymphoma t(14;18)(q32;q21) IGH/MALT1 MALT lymphoma t(11;18)(q21;q21) API2/MALT1 MALT lymphoma t(1;14)(p22;q32) BCL10/IGH splenic lymphoma, villous lymphs t(7;14)(q21;q32) CDK6/IGH Plasma cell myeloma myeloma t(4;14)(p16;q32) WHSC1/IGH myeloma t(6;14)(p25;q32) IRF4/IGH myeloma t(14;16)(q32;q22) MAF/IGH myeloma t(16;22)(q22;q11) IGL/MAF T cell leukemias/ lymphomas T-ALL del(1)(p32p32) SIL/TAL1

9-6-05, Gulley Draft page-

13

T-ALL t(7;11)(q35;p13) LMO2/TRB

ALCL t(2;5)(p23;q35) ALK/NPM Alveolar rhabdomyosarco t(2;13)(q35;q14) PAX3/FOXO1A

Sarcomas and tumors of bone and soft tissue

ma Alveolar rhabdomyosarcoma t(1;13)(p36;q14) PAX7/FOXO1A Ewing’s sarcoma/ P t(11;22)(q24;q12) FLI1/EWSR1 NET Ewing’s sarcoma/ PN t(21;22)(q22;q12) ERG/EWSR1ET Ewing’s sarcoma/ PN t(7;22)(p22;q12) ETV1/EWSR1ET Desmoplastic round cell tumor t(11;22)(p13;q12) WT1/EWSR1 Clear cell sarcoma t(12;22)(q13;q12) ATF1/EWSR1 Myxoid liposarcoma t(12;16)(q13;p11) DDIT3/FUS

Myxoid liposarcoma t(12;22)(q13;q12) DDIT3/EWS

Myxoid chondrosarc, extraskeletal t(9;22)(q22;q12) NR4A3/EWS Myxoid chondrosarc, extraskeletal t(9;17)(q22;q11) NR4A3/RBP56 Myxoid chondrosarc, extraskeletal t(9;15)(q22;q21) NR4A3/TCF12 Synovial sarcoma t(X;18)(p11;q11) SSX1 or 2/SYT Dermatofibrosarc. protuberans t(17;22)(q22;q13) COL1A1/PDGFB Alveolar soft part sarcoma t(X.;17)(p11;q25) TFE3/ASPSCR1 Mesoblastic nephroma, congenital t(12;15) )(p13;q25) ETV6/NTRK3 Infantile fibrosarcoma t(12;15)(p13;q25) ETV6/NTRK3 Endometrial stromal sarcoma t(7;17)(p15;q21) JAZF1/JJAZ1 Inflammatory myofibroblastic tumor t(2;19)(p23;p13.1) ALK/TPM4

Now that we have described the rules set forth by the professional grouthat are responsible for naming genes, we must discuss how to maximize their utility in the healthcare system. First, it is important that we use proper gene nomenclature in clinical documents to facilitate communication among heaproviders and those who interface with providers. Recognize that a change in the way we re

ps

lthcare

port gene tests may cause confusion in the short term but will ltimately serve the healtcare system well in the longterm. Second, we must

te specific reagents to use proper gene

o ls

se

man genes because they somatically rearrange during physiologic development of B and T lymphocytes. The proposed symbols for these genes a In the T cell receptor

uencourage manufacturers of analynomenclature. Third, we must encourage researchers and journal editors toadopt the same nomenclature system that we are adopting in clinical settings, sthat all of us can communicate using the same “language”. These are lofty goathat most will agree on, but we will now describe a few areas where gene nomenclature is problematic, and our recommendations for overcoming theproblems. Antigen Receptor Gene Names The immunoglobulin and T cell receptor genes are among the most complex hu

nd selected portions thereof are shown in Table 2.

9-6-05, Gulley Draft page-

14

gamma locus, TRGV1 through TRGV11 are the 11 variable regions, while TRGJ1, TRGJ2, TRGJP, TRGJP1, and TRGJP2 are the five joining regions. Note that some official gene names include the “at” symbol, @. As shown in Table 2, we recommend dropping the “@” from the gene name in clinical reports, in t because of the risk that info ol p Table 2: Nomenclature of Selected Antigen Receptor Genes

ene Symbol

par rmation systems will not process this symbroperly.

Official G Propo sed for Clinical Reports* IGH IGH IGHJ@ (IGH joining group) IGHJ IGK IGK IGKJ@ IGKJ IGKC (kappa constant region) IGKC IGL IGL TRA@ (T cell receptor alpha locus) TRA TRB@ TRB TRG@ TRG TRD@ TRD

*Caution: The HUGO Gene Nomenclature Committee proposes TRD as the official symbol of the tRNA aspartic acid gene, but we recommend using TRD as the symbol for the T cell receptor delta gene in the context of reporting molecular test results for lymphoid neoplasia. [PerIMGT and the HUGO Gene Nomenclature committee will consider renaming these genes TCTCRB, TCRG and TCRD.]

haps the RA,

ion en altered in structure or in transcriptional regulation, thus

he lesion

)(ABL

inical

d or fused genes separated by a slash (PML/RARA). Some have

arguedA) would

Translocations Translocations are common in hematopoietic malignancies and in sarcomas. A translocation can influence disease pathogenesis by fusing two genes into one transcriptome, as with PML/RAR, or by juxtaposing promoter and enhancer elements from one gene with another gene whose transcription is then dysregulated, as with IGH/BCL2. The genes at or near the translocatbreakpoint are oftcontributing to tumor pathogenesis or maintenance, or at least marking tas clonal. Translocation nomenclature is already partly established. The chronic myelogenous leukemia translocation is expressed as t(9;22)(q34;q11) when karyotype is used, or as "nuc ish 9q34(ABLx1,ABL spx2),22q11.2(BCRx2sp con BCRx1)/9q34(ABLx2),22q11.2(BCRx2)" when interphase FISH is used.8 The ISCN nomenclature for FISH has been criticized as inappropriate for cllaboratory reporting,9 and revised FISH nomenclature guidelines are forthcoming.(ref) There is no consensus on nomenclature when a translocation is discovered by amplification methods. When a translocation is found by amplifying across a translocation breakpoint, it seems reasonable to list the twojuxtapose

that a hyphen should be used (PML-RARA) but a hyphen means “deleted” in cytogenetic terminology. Perhaps a semicolon (PML;RARbe more consistent with karyotype nomenclature, as recommended by the Human Genome Variation Society (HGVS). Note that the HUGO Gene

9-6-05, Gulley Draft page-

15

Nomenclature Committee recommends no syntax at all between two regenes (PMLRARA).10 We believe that it is essential to use some form of punctuation so that it is clear when one gene name ends and when the nextstarts. Our recommendation is that a slash be used to separate the two gene names.

As for which gene is listed first, some prefer that the genes be listed in order of their appearance in the karyotype, as is done for cytogenetic nomenclature. In this scheme, genes on chromosome 1 are listed beforeon chromosome 2, p arm before q arm, and centromeric before telomeric. Others argue that karyotypic location is less important than oncogenic mechanism when describing an event at the molecular level, and therefore we should list fused genes starting 5’ and ending at the 3’ terminus of a fusion transcript. Thus, chronic myelogenous leukemia has a BCR/ABL1 translocatbecause the 5’ portion of BCR is fused with the ABL tyrosine kinase domathe 3’ portion of ABL. In fact, the leukemic cells also produce ABL1/BCR fusiproduct, but these reciprocal transcripts are not known to be oncogenic, in contrast to BCR/ABL1 transcripts. Must we thus know the structure and oncogenicity of the genes involved in the translocation before we decide how to report it? In our opinion, we shouldn’t get too worried about which gene is listed first. After all, there are some translocations in which the affected

combined

one

those

ion in of

on

gene may be quite d

ther mantle

ND1 r

L1

ocation

een d, consistent with the

presence of t(11;14)(q13;q32)”. The “consistent with” hedge is necessary since

istant from the translocation breakpoint (as with CCND1 dysregulation in mantle cell lymphoma), or for which the affected genes remain uncertain, or perhaps multiple genes are directly affected. To simplify the process, we propose that clinical reports use “karyotype order” rather than “5’ to 3’” order when reporting two genes that are juxtaposed by translocation. Not that this schema implies that ABL1/BCR is the correct order for the leukemia-associated translocation that is commonly known as “BCR/ABL”, which incites a negative gut reaction among most hematology/oncology experts, and could lead to confusion unless an educational campaign accompanies its introduction.

When the affected genes have not yet been defined, then the locus that is probed should be designated. For example, BCL1 or 11q13 were commonly used to describe the target region before the CCND1 gene was identified as being consistently dysregulated in mantle cell lymphoma. This raises anoquestion: When amplifying across the typical translocation breakpoint of cell lymphomas, is it reasonable to report that PCR for the “major translocation cluster subset of IGHJ/CCND1” was done, even though no probe for the CCgene itself was used in the test? Or would clinicians understand our report betteif we retained the traditional BCL1 terminology by reporting that “IGHJ/BCamplification was done to detect translocations involving the major translocation cluster”? Or should we build on cytogenetic nomenclature, even when karyotypewas not performed, by reporting that PCR for the “IGHJ and major translcluster subset of t(11;14)(q13;q32)” was done? We favor using the term IGHJ/CCND1 to describe the molecular assay that was performed. The corresponding analytic interpretation might we worded as “translocation betwIGHJ and the major translocation cluster was detecte

9-6-05, Gulley Draft page-

16

we can has

level translocations in

athology reports. For now, the decision on how to report procedures and findings remains largely at the discretion of the interpreting laboratory scientist. Bint

t is used to detect a translocation, only one of the gene p nt length polymorphism ire T

not guarantee that this translocation would be seen by karyotype; translocations can be occult, or there might be a three-way translocation thata different karyotypic string. Further work is needed to achieve an internationalconsensus on how to describe various molecular-p

ut we must continue to work towards consensus guidelines so that we can erpret others’ reports and the medical literature.

When Southern bloartners is typically probed, and a restriction fragme

mplies that the other partner is juxtaposed. Report of a positive Southern blot sult might be composed as shown in Figure 4.

est Name: BCR Southern blot rocedure: Southern blot analysis was performed using Bgl2 and Xba1 restriction enzymefollowed by hybridization with a probe t

P s argeting the major breakpoint cluster region of the BCR

gene on chromosome 22. Result: Abnormal bands are seen in both enzyme digests. Interpretation: The BCR gene is rearranged, implying the presence of ABL1/BCR translocation

involving the major breakpoint cluster region. The findings are consistent with a diagnosis of leukemia.

Comment: ABL1/BCR major breakpoint translocation is found in chronic myelogenous leukemia and in about one quarter of cases of Philadelphia-chromosome-positive acute lymphoblastic leukemia. Serial measurement by Q-rtPCR of ABL1/BCR RNA can be used to estimate leukemia burden and risk of relapse.

Figure 4: Partial report of a positive BCR Southern blot result

Mutations, deletions, and other Genetic Variations Nomenclature for allele variation is the purview of the HGVS. Theirdescribes the rules and gives examples, www.genomic.unimelb.edu.au/m

website

di/mutnomen/. Nomenclature rules are complex

he search,

k. It

found

and depend at least in part on the consequence of the nucleotide-level defect on translation of the protein. Of course, it would be unrealistic to determine tconsequence of the defect in every patient sample, so one depends on repreferably published in the medical literature or cataloged in reliable public databases, to infer the biochemical consequence and interpret the clinical significance of a given defect. Progress has been made in establishing databases to catalog disease-causing mutations and other variants.11(Ref Bentley DR, nature 429, 2004) Nomenclature is generally based on sequences deposited in GenBanmust be made clear which reference sequence is used by specifying the accession number from a primary sequence database, and whether this sequence represents cDNA or in genomic DNA, which impacts on whether c. or g. should proceed the nucleotide number.12 A web resource called RefSeq at www.ncbi.nlm.nih.gov/RefSeq seeks to establish a reference sequence for each human gene and for the genomes of selected microorganisms. This is accomplished by synthesizing the data in GenBank to create a standard gene

9-6-05, Gulley Draft page-

17

sequence along with corresponding RNA and protein sequences. Because provisional sequences may change before they are validated and finalized, this database is a moving target. Furthermore, a single gene may have multiple RefSeqs reflecting cDNA splice variants. For example, the HFE gene currently

. The

se

heir

hould describe the genetic defect that is found, i.e. a DNA-ased test should describe the nucleotide-level defect that is found. It is commended that the anticipated amino acid change also be described, using

lly proved. The three letter amino acid nomenclature could be used, e.g. Cys282Tyr rather t oid c r A the a lanine and glycine E .” c.” fo cDNA, or “p.” for protein. Table 3: Nomenclature for Common Inherited Disease Genes

clature or Name

has 11 different cDNA RefSeqs in GenBank. The longest cDNA is 2727bp (NM_000410) whereas the full gene sequence is 269,712bp (NG_001335)best reference sequence is not always evident, but use of the longest cDNA sequence seems reasonable for most clinical applications. Aside from thecomplications, the NCBI database demonstrates progess towards establishmentof a common reference by which all other sequences are compared. In clinicalreports, it remains important to designate the reference sequence number by which your patient’s variations are compared. Examples of nomenclature for common recurrent genetic defects and tassociated amino acid changes and diseases are shown in Table 3. The pathology report sbrebrackets if the protein level change is deduced but not experimenta

han C282Y, in order to av onfusion over whethe and G stand for mino acids a or the nucleotides adenine and guanosine.ven better, use the prefix “g for genomic DNA, “ r coding DNA or

Correct Nomen Common or Pri Associated Disease AF478696.1:g Prothrombin 20210G>A Venous thrombosis

F5 c.1517G>A FV r 1691G>A or Leiden op.Arg506Gln

Venous thrombosis

MTHFR NM_005957:g.716C>T r THFR c.665C>T

or stenemia OM

MTHFR C677C>T p.Ala222Val

Homocy

MTHFR c.1286A>C MTHFR 1298A>C or p.Ala222Val

ia Homocystenem

F2 .21538G>A

HFE c.845G>A HFE C282Y or Hemochromatosis p.Cys282Tyr

HFE c.187C>G HFE H63D or p.His63Asp Hemochromatosis SERPIN1A Alpha-1-antitrypsin deficiency ACADM 985 G>A MCAD MTTL1 3243A>G MELAS MTTK 8344A>G MERRF CFTR.c.1522_1524del CFTR.p.508delF Cystic fibrosis

Table 3 is under construction. The use of divergent reference gene sequences has led to confusion about which defect is being described. For example, MTHFR c.665C>T is thcorrect nomenclature for a mutation that is more commonly referred to in older literature as c.677C>T. Counting nucleotides in cDNA can be problematic if

e

9-6-05, Gulley Draft page-

18

there is alternative splicing, and so it is reasonable to reference the longest transcript variant for any given gene. Unfortunately the longest transcript in GenBank is a moving target since it is subject to change by newly deposited

utation is usually identified at the DNA ther than the cDNA level, it seems reasonable to avoid the splicing problems

o the correct terminology nk

eport, by the

ww.acmg.net/resources/policy-list.asp.

cDNA sequences. Since MTHFR gene mraaltogether and report the result at the gene level, swould be MTHFR NM_005957:g.716C>T, with NM_005957 being the GenBaRefSeq-file accession number. The corresponding cDNA and protein level changes can be inferred and described in the interpretation section of the rif appropriate. Guidelines for interpreting sequence variants are published American College of Medical Genetics, w Me nomena egenconme

R

thylation and other epigenetic pheThere are no established rules on how to report methylation or other

pigenetic events. We recommend a practical approach whereby the target e is identified, the molecular findings are stated, and the report is kept cise but medically informative. Figure 5 is an example of a report of a thylation study of a patient suspected of having Prader-Willi syndrome.

esults: The methylated (maternal) SNRPN allele was detected, but the unmethylated (paternal) SNRPN allele was not detected. terpretation: Only the methylated (maternal) SNRPN allele was detected, consistent wdiagnosis of Prader-Willi syndrome (see comment)

In ith a .

Comment: Methylation studies revealed the presence of only the methylated (maternal) allele of the SNRPN gene. This result suggests that there is either a deletion involving the paternally derived chromosome 15, maternal uniparental disomy (UPD) 15, or an error in imprinting (methylation) of the paternal chromosome 15. Any of these alterations is consistent with the diagnosis of Prader-Willi syndrome. Genetic Counseling is strongly recommended to discuss the implications of this result and to discuss further testing that can be done to determine which genetic mechanism is responsible for the condition in this patient. A test for deletion of the relevant region of chromosome 15 is available in the Cytogenetics Laboratory. (www.geneclinics.org).

Figure 5: Partial report of a positive gene methylation result

Microsatellites and Non-coding Regions Clinical testing often involves analysis of highly polymorphic regionsDNA. Minisatellites, short tandem repeats (STRs, also known as microsatellitesand single nucleotide polymorphisms (SNPs) are com

of ),

monly interrogated.

rofessionals for assessing donor cell engraftment, chimerism, allele loss,

uman source. A list of common STRs has been compiled by e National Institute of Standards and Technology,

www.cstl.nist.gov/biotech/strbase/index.htm

Thousands of these polymorphic regions have been identified in human DNA, and 13 of the STR loci are used by the forensic community for the national convicted offender database, National DNA Index System. (See Table 4.)These or other polymorphic loci are often targeted by the health care pmicrosatellite instability, parentage or other family relations, and matching samples to their hth

. A searchable catalog of SNPs is

9-6-05, Gulley Draft page-

19

available a b bi.nlm ery.fcgi?t the NCBI we site, www.nc .nih.gov/entrez/qu db=snp he Nomencla

guidelture for describing variations in STR found in t

ines of the HGVS at www.genomic.unimelb.edu.au/mdi/mutnomens and SNPs is

Table 4: phic S RepeIdentity T

Locus, Common Name

osom peat Mo of Common at Alle

B

13 Core Loci Used for Identity Te sting CSF1PO 5 TAGA 6-15 D3S1358 [TC A] 3 TG][TCT 12-19 D5S818 5 AGAT 7-16 D7S820 7 GATA 6-14 D8S1170 8 [TCTA][TCTG] 8-18

D16S539 16 GATA 5, 8-15 D18S51 18 AGAA 7, 9-27 D21S11 21 [TCTA][TCTG] 24-38 FGA 4 C 17- .2 TTT 46TH01 11 TCAT 4-13.3 TPOX 2 GAAT

.

Common Polymor hort Tandem at Loci Used for esting

Chrom e Re tif Range Repe les

Gen ank Reference

D13S317 13 TATC 7-15

6-13 vWA 12 [TCTG][TCTA] 11-22

5 Loci Used for Microsatellite Instability Testing (The Bethesda Panel) BAT-26 2 mono U41210 NR-21 ??? di XM033393NR-24 2 di X60152 MONO-27 2 di AC007684 BAT-25 di U63834 Mitochondrial DNA analysis

Mitochondrial DNA (mtDNA) is a circular, double-stranded DNA that ncodes 13 polypeptides required for oxidative phosphorylation, 22 transfer

mal RNAs. Point mutations or deletions in mtDNA usually increas o

eRNAs, and 2 riboso

e exponentially with age and are passed along in the egg from mother tchild. Useful information on the mtDNA is found at www.mitomap.org. The mtDNA sequences in the NCBI database should be used as a reference by which to report sequence variants, analogous to what is recommended above forreporting alterations of chromosomal DNA. Microorganisms

Decades ago, the American Society for Microbiology (ASM) de

veloped

ee standards for bacterial gene names, and these rules are still used today in most clinical and research applications. Typically, bacterial gene names have thrsmall letters representing the function or class of gene and a fourth capital letter

9-6-05, Gulley Draft page-

20

for the specific gene (e.g. dnaG encodes a DNA primase). Homologous geother organisms may have the same name, and therefore it is important to designate the bacterial genus and species. A compilation of bacteria names along with their genomic reference sequences (RefSeqs) is found at www.ncbi.nlm.nih.gov/genomes/lproks

nes in

.cgi. Each RefSeq document specifies the gene n

ting the

previous sentence, and

eqs is l

r

ient

For HIV mutations and genotypes, laboratories should consult the AIDS Society at

d

only called T69D)

tical when al

s, and

at it in defects

whose

e

it

may ystems cannot process such formatting.

hus,

rms

ames, DNA sequences, and amino acid translations. Mutations are designated by the nucleotide number, e.g., [need to insert a clinically relevant example of an organism name, gene name, nucleotide substitution]). The corresponding amino acid substitution can be inferred and reported by liswild type amino acid before, and the mutant amino acid after, the codon position (e.g., N138E [change this to match the example in theuse the three letter aa designation]).

Fungal genes have similar designations, and a list of fungal RefSfound at www.ncbi.nlm.nih.gov/genomes/FUNGI/funtab.html. In the list of viraRefSeqs at www.ncbi.nlm.nih.gov/genomes/static/vis.html, note that Epstein-Barvirus is called by its official name of human herpesvirus 4, while human cytomegalovirus is called human herpesvirus 5. In clinical reports, it is sufficto use the common medical name for an organism. compilation of mutations updated by the International http://iasusa.org/resistance_mutations/index.html. Mutations identified in the HIV-1 reverse transcriptase (RT) or protease (P) genes are commonly reportein conjunction with the inferred amino acid substitution (e.g., HIV-1 NC_001802 RT [insert genotype here] encodes [show three letter aa], comm

. Multiple mutations should be listed individually. Practical Considerations and Problematic Issues The rules set forth by professional groups are not always pracapplied in clinical laboratories. Thus we have already recommended severmodifications, such as using a slash to separate translocated gene namedropping the “@” from the names of antigen receptor genes.

A few other problematic issues are now addressed. We recognize thwill be difficult to adopt new nomenclature for certain genes or certa

common names are entrenched in the literature and in patient records. It may even be dangerous if a patient report is worded in a way that is misunderstood by a clinician. But the longer we wait to change, the morentrenched the "incorrect” name becomes, and the harder it becomes to transition to a standard nomenclature system. During the transition period, seems reasonable to use the correct terminology followed by the common terminology in parentheses, e.g. RUNX1/CBFA2T1 (AML1/ETO).

Although it is proper to display gene names in italics, the use of italicsbe impractical if electronic information sFurthermore, it seems cumbersome to describe transcripts as “(mRNA) PML/RARA” when instead they could be termed “PML/RARA transcripts”. Twe recommend that laboratories balance the rules with the practicalities of implementing them. It is certainly acceptable to omit italics and to use the te

9-6-05, Gulley Draft page-

21

“PML/RARA translocation”, “PML/RARA transcripts”, “PML/RARA cDNA”, and “PML/RARA protein” to describe the various analytes and products that are relevant in clinical laboratory reports. We recommend that a single reference sequence, as defined by the National Center for Biotechnology Information (NCBI), be used as the gold standard by which all gene variants are described. We are working with NCBI to devise a plan for longterm stability of such a reference sequence. Unfortunately the current reference sequences are subject to updates as more information is submitted to GenBank, which makes for a moving target that donot suit our purposes. We recommend that a single DNA-based reference sequence be established for each human gene, including introns, exons, and 3’ and 5’ untranslated regions as well as promoter and enhancer regions, so thvirtually all local variants affecting structure or function or transcriptional regulation are captured. Likewise, a single cDNA sequence could be designatedfor each gene to use as a reference for changes in the coding region. For smnon-human genomes like v

the

es

at

all

iruses, it might be more straightforward to use the whole

e of

at

rting, a product which we coin “ClinSeq”.

other problematic issues in molecular reporting will undoubtedly

ms and

eedback ect

o n the

genome as the reference sequence. Given the diversity of “normal” sequences, arguments may ensue as to which variant is most representativ“wild type”, but in truth it is more important to have a reasonable sequence written in stone than it is to have an accurate or representative sequence thtakes years to devise. We will keep you posted on our progress in this quest todefine “the” reference sequence that is used for clinical repo

Manycome our way as we continue to learn about genomics and disease. Such issues are commonly addressed by medical professionals working in teain conjunction with pertinent professional societies to gain consensus, and always considering what is best for our patients. We hope that the issues and questions raised herein will help guide future discussions.

F We recognize that the guidelines and proposals in this article are subjto discussion and revision, and we welcome input from individuals and professional groups. To provide feedback, please send correspondence [email protected]. Updated versions of our recommendations are posted oCAP website at www.cap.org. We are working towards a deadline for implementation of consensus reporting guidelines on July 1, 2007. Acknowledgements: Guidelines for reporting were developed by the CAP

CAMolec Cytogenetics Resource Committees. We thank Jill Kachin for excellent staff suhelpfu

Molecular Pathology Resource Committee. We would like to thank the other P groups providing input, including the Microbiology, Biochemical and

ular Genetics, Cytogenetics, Histocompatibility/ Identity Testing and

pport. We thank Shuji Ogino MD, PhD and Marie-Paule LeFranc of IGMT for l discussions.

9-6-05, Gulley Draft page-

22

Ite1. Ji rovide a mechanism for new versions of this document to be

2. Tosted.

ar

4. S the g 5T

5. Id aule

e Grody, ractice Committee, Maria

Costanzo (e-mail: [email protected]

ms under active discussion, DUE DATE for response is Dec 1, 2005. ll, please pposted, and an email address to which comments should be sent (currently Jill’s email). ables 1 and 3 will be published in abbreviated form with reference to the CAP website where more complete and updated versions are p

3. We are working with NCBI to establish a ClinSeq database, with a 1.5 yehorizon. huji and Jeff, using the established rules for expressing two changes in same allele, e.g. CFTR [p.Arg117His; 5T], please describe briefly usinin CFTR as an example. Also, please work on expanding Table 3. Will reference the new CF guidelines from ACMG when available. eas for persons/groups to comment on this manuscript—Marie-PLeFranc for IGMT, Jean-Loup Huret, Felix Mitelman, Jacques van Dongen, Elizabeth MacIntyre, Karen Mann, Dan Arber, WaynMark Sobel for AMP and the AMP Clinical P

), Jim Eschelman (F5 study), Domnita Crisan (BCR study), Steve Schichman (fingerprinting study), Ira

d Michele Caggana (CF study), Steve Dumler. HGVS= Lubin [email protected] and [email protected], International Committee of Medical Journal Editors= claine @acponline.org, [email protected]

nces: tions for

n of Directors of

S, Kashalo MS,Dighe . Ph f the

8: 1424-

t reports for cystic fibrosis

parative study of interpretation among

ists are 000.

Refere1. Jaffe ES, Banks PM, Nathwani B, Said J,Swerdlow SH. Recommendathe reporting of lymphoid neoplasms: A report from the AssociatioAnatomic and Surgical Pathology. Mod Pathol 17: 131-135, 2004. 2. Association for Molecular Pathology statement. Recommendations for in-house development and operation of molecular diagnostic tests. Am J Clin Pathol 111: 449-463,1999. 3. Lim EM, Sikaris KA, Gill J, et al. Quality assessment of interpretative commenting in clinical chemistry. Clin Chem 50: 632-637, 2004. 4. Laposata ME, Laposata M, Van Cott EM, Buchner DAS ysician survey of a laboratory medicine interpretive service and evaluation oinfluence of interpretations on laboratory test ordering. Arch Pathol Lab Med 121427, 2004. 5. Krousel-Wood M, Andersson HC, Rice J, Jackson KE, Rosner ER,Lubin IM. Physicians' perceived usefulness of and satisfaction with tes(DeltaF508) and factor V Leiden. Genet Med 5: 166-171, 2003. 6. Attanoos RL, Bull AD, Douglas-Jones AG, Fligelstone LJ,Semararo D. Phraseology in pathology reports. A compathologists and surgeons. J Clin Pathol 49: 79-81, 1996. 7. Powsner SM, Costa J,Homer RJ. Clinicians are from Mars and pathologfrom Venus. Arch Pathol Lab Med 124: 1040-1046, 2

9-6-05, Gulley Draft page-

23

9-6-05, Gulley Draft page-

24

lature. Birth

en

S.

itis O,Cotton RG. The challenge of documenting mutation across the enome: the human genome variation society approach. Hum Mutat 23: 447-452, 2004.

r

Infectious Diseases; MM5-A Nucleic Acid Amplification Assays for olecular Hematopathology; MM6-A Quantitative Molecular Methods for

Infectious Diseases; MM9-A Nucleic Acid Sequencing Methods in Diagnostic Laboratory Medicine; MM7-A Fluorescence In Situ Hybridization (FISH) Methods for Medical Genetics. [ISBN 1-56238-]. Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 19087-1898 USA, 2005.

8. An International System for Human Cytogenetic Nomenclature (1985) ISCN1985. Report of the Standing Committee on Human Cytogenetic NomencDefects Orig Artic Ser 21: 1-117, 1985. 9. Mascarello JT, Cooley LD, Davison K, et al. Problems with ISCN FISH Nom clature make it not practical for use in clinical test reports or cytogenetic databases [corrected]. Genet Med 5: 370-377, 2003. 10. Wain HM, Bruford EA, Lovering RC, Lush MJ, Wright MW,PoveyGuidelines for human gene nomenclature. Genomics 79: 464-470, 2002. 11. Horag12. Antonarakis SE. Recommendations for a nomenclature system for human gene mutations. Nomenclature Working Group. Hum Mutat 11: 1-3, 1998. 13. Forbes BA. Introducing a molecular test into the clinical microbiology laboratory: development, evaluation, and validation. Arch Pathol Lab Med 127: 1106-1111, 2003. Clinical and Laboratory Standards Institute. MM1-A Molecular Diagnostic Methods for Genetic Diseases; MM2-A Immunoglobulin and T-Cell ReceptoGene Rearrangement Assays, second edition; MM3-A Molecular Diagnostic Methods for M