Bone Marrow Genetic Studies for Malignant Lymphoma Staging
Optimizing Laboratory Testing for Hematologic Disorders Series
Click CC box for captions; full transcript is below.
Published: January 2013Print Record of Viewing
Dr. Kurtin presents the third in our series “Optimizing Laboratory Testing for Hematologic Disorders.” The series addresses guidelines for appropriate laboratory test utilization in hematologic disorders. This presentation will discuss the appropriate utilization of genetic testing when evaluating the bone marrow for involvement by malignant lymphomas.
Presenter: Paul J. Kurtin, MD
- Consultant in the Division of Hematopathology, Department of Laboratory Medicine and Pathology at Mayo Clinic in Rochester, Minnesota
Questions and Feedback
TranscriptDownload the PDF
The title of my talk today is “Optimizing Laboratory Testing for Hematologic Disorders Series: Bone Marrow Genetic Studies for Malignant Lymphoma Staging."
I have no conflict of interest disclosures.
Let’s start with a case. The patient is a 64-year-old female with cervical, axillary, and inguinal adenopathy that slowly developed over several months. To evaluate the cause of the adenopathy, a lymph node biopsy was performed. I will show you photographs of the lymph node in the next slides. The lymph node diagnosis was follicular lymphoma. A bone marrow biopsy was then performed for staging. There was no other relevant past medical history.
Here is a composite photograph of the lymph node specimen. On the left, the lymph node architecture is effaced by an abnormal lymphocyte population growing in a nodular pattern. Top right, the nodules contain a monomorphous population of lymphocytes, bottom right, that have elongated, irregular, grooved and cleaved nuclei, partially clumped chromatin and sparse eosinophilic cytoplasm.
This slide illustrates the phenotype of the lymphoma. The tumor cells express the pan B-cell marker CD19, the germinal center B-cell antigens CD10 and bcl-6 and there is aberrant expression of bcl-2. The nodules are associated with CD21-positive follicular dendritic cell meshworks and there are only few intermixed CD3-positive non-neoplastic T cells.
This is a typical case of follicular lymphoma, grade 1 by WHO lymphoma classification criteria.
Here is a photograph of the staging bone marrow specimen from our patient. There intertrabecular and paratrabecular nodules of abnormal lymphocytes that, as seen in the top right photograph, are cytologically similar to those in the lymph node specimen.
Now it is quiz time. This is the menu of ancillary tests that could be performed on the bone marrow specimen: Immunohistochemistry for follicular lymphoma-associated markers, flow cytometry, cytogenetic analysis, FISH for IGH/BCL2 rearrangements which is typically seen in 90% of grade 1 follicular lymphomas as the result of t(14;18)(q32;q21)], and immunoglobulin gene rearrangements for B-cell clonality.
Here is what the clinician ordered to be done on the bone marrow.
What would you do? What tests are appropriate do you think in this context? So keep your answer in mind and we will continue with a discussion of the value of genetic testing in lymphoma staging bone marrow specimens.
For purposes of review, malignant lymphomas are a complex family of cancers of lymphocytes. There are many types and they are diagnosed and classified using the criteria of the World Health Organization Classification of Haematologic and Lymphoid Neoplasms. In most patients, the initial diagnosis of lymphoma is made based on morphologic and phenotypic evaluation of biopsies of lymph nodes or other extranodal sites such as GI tract, skin, lung, liver, or spleen. While the bone marrow is occasionally the first site of evaluation for a suspected lymphoma, bone marrow examination is performed in almost all lymphoma patients to evaluate the extent (or stage) of disease. In addition to lymphoma type, stage is a major determinant of both treatment and prognosis for lymphoma patients. Therefore, accurate assessment of bone marrow for involvement by lymphoma is essential for patient management.
Traditionally, morphologic evaluation of bilateral bone marrow aspirate and biopsy specimens had been the cornerstone for determining bone marrow involvement by lymphoma in staging specimens. In recent years, flow cytometry, conventional cytogenetic analysis, FISH studies, and molecular genetic tests for immunoglobulin and T-cell receptor gene rearrangements have been increasingly used to make this determination. However, there is little evidence about how effective these ancillary tests actually are to provide information over and above morphology when evaluating the bone marrow for involvement by lymphoma. And there are even fewer clinical studies indicating that flow cytometric, cytogenetic, or molecular genetic evidence of bone marrow involvement has the same clinical consequences as morphologic evidence of involvement does.
So we performed a study focusing on genetic tests because it was our informal impression that these added little to the assessment of presence or absence of bone marrow involvement in the setting of lymphoma staging. Our assumptions going into the study were that the diagnosis and classification of the lymphoma had been definitively established based on a biopsy of an anatomic site other than bone marrow and that the bone marrow was performed subsequent to the lymphoma diagnosis for staging, but NOT to make a primary diagnosis.
Our goal was to determine if routine karyotyping (cytogenetic studies) of bone marrow lymphoma staging samples: Added information about the presence or absence of lymphoma in the bone marrow; Added information about lymphoma classification that might alter therapy or; Added information about any other non-lymphoma disorder that might be present in the bone marrow
We started with 574 lymphoma cases that had been diagnosed on a site other than bone marrow at Mayo Clinic over a 3-month period. In 298 of them, a contemporaneous bone marrow was performed. Of those, routine cytogenetic analysis was performed in 112. These 112 cases became the study group. Of the 112 karyotyped cases, 41 were positive for lymphoma by morphology, supplemented by flow cytometry or immunohistochemistry in a small number. Of these 41 lymphoma-involved bone marrows, the karyotype was normal in 32 and abnormal in only 9. All 9 karyotypically abnormal cases had cytogenetic abnormalities that were predictable based on the known lymphoma type that involved the marrow (eg, the t(14;18) was identified in the bone marrows involved by follicular lymphoma, the t(11:14) was detected in the bone marrows involved by mantle cell lymphoma).
There were 71 bone marrow specimens that were negative for lymphoma based on morphology and phenotype. Eight cases had abnormal cytogenetic results. But in none of these was the karyotype typical for lymphoma.
These are the karyotype results from the 8 bone marrows that were negative for lymphoma, but karyotypically abnormal.
In the first 3 cases these patients with follicular lymphoma, diffuse large B-cell lymphoma and classical Hodgkin lymphoma had single metaphases that had a chromosomal abnormality. Since the accepted cytogenetic definition of an abnormal clone requires that the same abnormality be present in 2 or more metaphases, these results are of doubtful significance.
The next 3 patients, 2 with diffuse large B-cell lymphoma and 1 with a MALT lymphoma, were elderly males with loss of the Y chromosome from a fraction of the bone marrow cells. This is a well-recognized phenomenon of aging and is not considered to be a diagnostically significant finding.
There was 1 patient with follicular lymphoma who had a bone marrow negative for lymphoma, but positive for acute myeloid leukemia. The bone marrow in this patient was done for restaging after follicular lymphoma recurrence following a long history of lymphoma treatments. The karyotype with add(5q)(q13) and -7 is consistent with therapy-related AML. The final diffuse large B-cell lymphoma patient with a morphologically normal bone marrow had a chromosome abnormality in all the tested metaphases that was proved to be a constitutional abnormality. So it did not indicate lymphoma involvement or any other hematologic disorder involving the bone marrow.
There were 8 cases, negative for lymphoma, but with morphologic findings that suggested another bone marrow abnormality. In the top 5 specimens with normal karyotypes, the pathologists evaluating the cases were concerned about myelodysplasia so the karyotypes were of potential utility to support MDS diagnoses, but as you can see, they had no cytogenetic abnormalities. Two patients had myeloid abnormalities associated with typical karyotypes for these disorders: a myelodysplastic syndrome patient had [del(7)(q22)] and polycythemia vera sample had [add(12)(q22)]. Finally the AML patient is the one that we discussed previously.
So our study points to the following conclusions:
Cytogenetic analysis does not improve sensitivity for lymphoma detection in staging bone marrow specimens over and above morphology
Cytogenetic analysis does not add useful additional data about lymphoma type or prognosis in lymphoma-positive bone marrows over and above what is already known about the lymphoma based on its diagnosis in the non-bone marrow site.
Normal bone marrow morphology in lymphoma patients is sometimes associated with abnormal cytogenetic results, but the abnormalities are of questionable significance.
If there is a specific myeloid abnormality identified by morphology, genetics can help determine its significance and this is particularly helpful for evaluating treatment-related myelodysplasia.
Finally, in the pre-auto-bone marrow transplant setting, cytogenetics justified to exclude morphologically occult myeloid dysplastic syndromes.
So based on data, our recommendation for cytogenetic analysis for lymphoma staging specimens is the following:
At the time of the bone marrow biopsy procedure, cytogenetic samples should be obtained on all lymphoma staging bone marrows and held.
The bone marrow morphology should be reviewed.
Cytogenetic testing should be performed only in the following circumstances:
If there are unexplained cytopenias, if there are morphologic features suggesting an underlying myeloid disorder.
Or prior to autologous bone marrow transplant because sometimes morphologically normal bone marrows are associated with cytogenetic abnormalities and there is literature that suggests that auto transplants with cytogenetically abnormal bone marrow is associated with poor outcome.
No cytogenetic testing should be done on all other bone marrow staging samples
We have reviewed our practice and the literature for FISH testing in the context of lymphoma staging as well. Our conclusions are similar to those for conventional cytogenetic analysis:
FISH is insensitive compared to morphology in detecting bone marrow involvement by lymphoma
In FISH abnormal cases, the results do not add diagnostic information over and above what is already known about the lymphoma based on the initial biopsy.
FISH should not routinely be ordered on bone marrow specimens performed for lymphoma staging.
Coupled with morphology and phenotyping on blood and bone marrow specimens, FISH may help define a particular type of lymphoproliferative disorder (eg, mantle cell lymphoma, follicular lymphoma, Burkitt lymphoma, or T prolymphocytic leukemia) when the disease in question has a disease-defining chromosomal abnormality and no tissue biopsy is available.
Immune receptor gene rearrangement studies are also commonly ordered in the context of bone marrow evaluations for lymphoma and lymphoid leukemias. The EuroClonality/BIOMED-2 consortium recommended primer sets and interpretive guidelines are used at Mayo and in other high-quality reference laboratories for these analyses. Based on data from the BIOMED-2 projects, the following conclusions can be drawn about the overall sensitivity of these analyses.
First, T-cell receptor gene rearrangements:
A clonal pattern of T-cell receptor rearrangements is identified in 90% of T-cell lineage lymphomas and leukemias. BUT a clonal pattern is also present in a variable percentages of reactive conditions (dermatitis, oligoclonal immune reactions such as the response to Epstein-Barr virus, cytomegalovirus, human immunodeficiency virus, or hepatitis B virus) and a clonal pattern can be associated with the predicted contraction of the T-cell repertory that occurs as part of normal aging (ie, the older you are the more likely it is that clonal T-cell receptor gene rearrangements will be detected in your blood and bone marrow in the absence of a T-cell malignancy. So the problem is T-cell receptor gene rearrangements have low specificity for T-cell lymphomas. Finally, non-clonal patterns are associated with most lymphoid hyperplasias.
Second, immunoglobulin gene rearrangements
A clonal pattern is detected in 90% of B-cell lineage lymphomas and leukemias. Clonal patterns are rare in reactive conditions. Non-clonal patterns are associated with lymphoid hyperplasias. So, immunoglobulin rearrangements have higher specificity for malignancy than T-cell receptor gene rearrangements BUT there are other tools that are cheaper and faster to detect B-cell clones: specifically, immunohistochemistry and flow cytometry readily detect light-chain restricted B-cell and plasma cell populations, making detection of immunoglobulin gene rearrangements unnecessary in most B-cell lymphomas.
To test this we reviewed our practice data on clinician-ordered immunoglobulin gene rearrangements in bone marrow specimens.
Over a 1-year period, 47 cases were evaluated for immunoglobulin gene rearrangements.
In 14 cases the immunoglobulin gene rearrangement indicated clonality. In all 14, though, flow cytometry or immunohistochemistry indicated a light-chain restricted B-cell population, rendering the molecular tests unnecessary.
In 33 cases, the immunoglobulin gene rearrangement result was non-clonal.
Three patients had myeloid disorders and all the rest had morphologically normal bone marrows.
All of the immunoglobulin gene rearrangement tests were considered to be unnecessary.
We then turned our attention to the clinician-ordered T-cell receptor gene rearrangement studies performed on bone marrow specimens.
Again, the results are based on 172 total cases.
In 133 cases, we detected no clonal T-cell population, but in 12 of these, the patient had bone marrow involvement by a T-cell malignancy (8 peripheral T-cell lymphomas, 2 T-cell large granular lymphocytic leukemias, 1 CD4+ T-cell lymphoproliferative disorder that couldn’t otherwise be classified and 1 case of T-acute lymphoblastic leukemia). These 12 were considered to be false-negative results.
The rest of the bone marrow specimens were involved by other myeloid or B-cell malignancies, as shown here, or were normal. And these are considered true negative results.
Again considering clinician-ordered T-cell receptor gene rearrangement studies performed on 172 bone marrow specimens.
In 39 cases, we obtained clonal results.
In 20 cases, the T-cell receptor gene rearrangement results helped to diagnose T-cell malignancies: 11 cases of large granular lymphocytic leukemia, 6 peripheral T-cell lymphomas, 2 T-acute lymphoblastic leukemia and 1 CD4+ T-cell lymphoproliferative disorder.
In all the rest (19 cases) the results were considered to be false positives. They involved B-cell disorders in the bone marrows (2 myelomas, 1 hairy-cell leukemia) myeloid disorders involved in the bone marrows that had hybrid features between myelodysplastic syndromes and myeloproliferative neoplasms, 1 case of chronic myelomonocytic leukemia and 12 normal bone marrows.
So, based on 172 cases, 20 true-positive results were obtained. There were 19 false-positive results and 12 false-negative results. One hundred and twenty-one tests were considered to be unnecessary based on the other studies that had been performed on the bone marrows
Here was the reaction of 1 of our hematologists when we presented the data to them:
Our conclusion was that T-cell receptor and immunoglobulin gene rearrangement tests are poor screening modalities. Their use in bone marrow specimens should be restricted to those cases where they are likely to help resolve differential diagnostic problems posed by clinical history, morphology, and phenotype, and pathologists are in the best position to address these issues and order the rearrangement studies.
Based on practice data that we shared with our clinicians, here are the utilization principles that we have collectively have initiated in Mayo Clinic lymphoma practice and which we recommend to our MML clients:
Cytogenetics requests on staging bone marrows for lymphoma are approved for processing by the hematopathologist only when there is a question of a myeloid disorder or pretransplant.
Lymphoma FISH requests will be approved for processing by the hematopathologist only when they help to resolve a differential diagnosis problem posed by morphology and phenotype. Most of these will be canceled.
Immunoglobulin and T-cell receptor gene rearrangement requests on bone marrow specimens will be approved for analysis by the hematopathologist only when they help to resolve a differential diagnosis problem posed by morphology and phenotype. Again, most of these tests will be canceled.
Now back to our case. To refresh your memories, here is the bone marrow morphology from our patient with follicular lymphoma. Nodules of intertrabecular and paratrabecular abnormal lymphocytes that are cytologically identical to the neoplastic cells in the original follicular lymphoma biopsy.
Here is the list of possible ancillary studies and the highlighted tests ordered by the clinician.
After hematopathology review here’s what we did and what we would recommend. Only bone marrow aspirate and biopsy and evaluation of the morphology.
Our staging approach for malignant lymphoma can be found on the Mayo Medical Laboratories Web site. It is encapsulated in this algorithm. Since the presentation that I just gave summarizes our approach I would refer you to the Web site for further details about the algorithm.
So in conclusion, in an environment where it is our fiduciary responsibility to our patients, our medical institutions, and the health care system in general, we must be engaged in optimizing test utilization so that we get the biggest diagnostic bang for the buck. An ongoing process of practice review helps to generate the data on which test utilization principles can be derived. Implementation of these strategies requires a collaborative process. It is best done where clinicians and pathologists work together to optimize test utilization.
Clinicians evaluate the patients, generate a differential diagnosis based on their findings, order tests, including tissue biopsies.
Pathologists review the clinical history and the morphology of the tissue biopsy, refine the differential diagnosis, consult the mutually agreed upon practice guideline for test utilization and then refine the test order to resolve the remaining differential diagnosis, canceling all unnecessary tests. We have been successful at implementing these principles in our own practice and I hope that you can too.
Finally, I would like to acknowledge those who worked on the utilization projects: From Hematopathology, Dr. Curtis Hanson, Dr. Rebecca King, from cytogenetics, Dr. Rhett Ketterling and Dr. Daniel Van Dyke. And, of course, our hematologists, too numerous to mention, but without whose support we would not have been able to implement a successful test utilization strategy.
Thank you for listening to this presentation.