OUP user menu

Necessity of Bilateral Bone Marrow Biopsies for Ancillary Cytogenetic Studies in the Pediatric Population

Kate E. Grimm MD, Connie Chen, Lawrence M. Weiss MD
DOI: http://dx.doi.org/10.1309/AJCPHR1M1EERGEOK 982-986 First published online: 1 December 2010

Abstract

The need for bilateral pediatric bone marrow biopsies for cytogenetic and fluorescent in situ hybridization (FISH) studies has not been clearly delineated. We retrospectively identified 166 pediatric bilateral bone marrow biopsy specimens obtained from patients with a variety of clinical diagnoses, including solid tumors, lymphoma, leukemia, and other hematologic conditions. The cases included all pediatric bilateral bone marrow biopsies performed at our hospital spanning the years of 1992 to 2008. Agreement of FISH and classical cytogenetic results between the 2 sides was assessed. Of a total of 166 bilateral cases, 2 cases showed disagreement (1.2%), both from patients with solid tumors. One case was a rhabdomyosarcoma, in which FISH only was performed; the second was a neuroblastoma in which FISH and cytogenetics were performed (both FISH and classical cytogenetic results disagreed). The remainder of the cases showed complete agreement between the 2 sides (total 98.8%). We conclude that it is usually not necessary to perform bilateral bone marrow biopsies for FISH and cytogenetics in the pediatric population outside of the setting of solid tumor staging.

Key Words:
  • Bilateral bone marrow biopsy
  • Pediatric patients
  • Ancillary studies
  • Fluorescence in situ hybridization
  • Classical cytogenetics

Bone marrow biopsies, although a low-risk procedure, are not entirely without risks.1,2 In the pediatric population, the risks and benefits of an invasive procedure must be balanced with the benefit of the information provided by the study. Histologic evaluation of bilateral bone marrow biopsies is often necessary in lymphoma or solid tumor staging.36 Ancillary studies are important tools in monitoring minimal residual disease or assessing current disease status.3,7 Molecular studies have become more important in determining the diagnosis and prognosis of pediatric leukemia.8,9 The role of bilateral vs unilateral biopsies for ancillary studies such as fluorescence in situ hybridization (FISH) and classical cytogenetics is less well defined. Such studies are expensive and should not be performed unless they contribute to the diagnosis or prognosis in a meaningful way.

The current retrospective study examined the agreement of classical cytogenetics and FISH studies between the 2 sides of bilateral bone marrow specimens.

Materials and Methods

Cases were identified by a computer search for bilateral pediatric specimens in the pathology files of the City of Hope National Medical Center, Duarte, CA. The cases spanned the years 1992 to 2008. The mean patient age was 12 years; the ages ranged from 1 year to 21 years. All cases that were found were used, with the exceptions described in the following text.

Cytogenetic Studies

Cases in which one or both sides had insufficient material for cytogenetic testing were discarded. Three cases were excluded by this criterion; they were all aplastic anemia cases. Some cases had requested cytogenetics on both sides; however, FISH was requested only on one side. These cases were retained and were considered as “cytogenetics only” studies because the FISH results were not bilateral; 6 cases fell into this category. Several cases had one or both sides that were considered to be “limited studies”; these cases were retained because they were still considered diagnostic and reportable. Two cases in this category were aplastic anemia cases; one case was limited on one side with only 9 cells instead of the normal 20 cells for analysis, and the other case showed 3 cells on one side and 2 cells on the other side. One case of Hodgkin lymphoma had only 6 cells for analysis on one side and the normal 20 cells on the other.

Cases were considered to be in agreement if the same karyotype was seen on both sides. Cases with a normal karyotype were considered negative, while cases with an abnormal karyotype were considered positive. Banded chromosomes were analyzed using standard methods and were identified using GTG banding at a banding resolution of 200 to 400. Karyotyping and clonal definitions followed the current International System for Human Cytogenetic Nomenclature criteria.10 Normal male or female karyotypes were considered negative. Nonclonal aberrations were disregarded.

FISH Studies

FISH studies were performed using Abbott probes (Abbott Park, IL) specific for a given clinical history Table 1. FISH results were considered positive if the given abnormality was identified above background limits and considered negative if it was not. Cases were considered in agreement if the rearrangement in question was seen on both sides. In cases in which a given abnormality was reported as detected in a certain percentage of cells, the difference between percentages was not taken into account as long as the abnormality was considered present at a high enough threshold to be reported. We followed the current International System for Human Cytogenetic Nomenclature.10

View this table:
Table 1

Clinical History

The bone marrow biopsy specimens included in the present study consisted of a total of 332 separate specimens (166 bilateral biopsies). Of the bilateral biopsies, 77 were performed to evaluate for involvement by solid tumor, 44 were performed to evaluate for involvement by lymphoma, 18 were performed in cases of leukemia, 26 were performed to evaluate other hematologic conditions, and 1 set was performed in a case in which the underlying disorder could not be ascertained.

Clinical histories were taken from the patients’ requisition, or, if not listed, were taken from the pathology report of the corresponding bone marrow biopsy and are listed in Table 2. Bilateral cases in the solid tumor group included 22 cases of Ewing/peripheral neuroectodermal tumor, 20 neuroblastomas, 18 rhabdomyosarcomas, 2 Wilms tumors, 1 renal cell carcinoma, 1 medulloblastoma, 1 chondrosarcoma, 1 aneurysmal bone cyst, 1 desmoid tumor, 1 atypical teratoid rhabdoid tumor, 1 malignant rhabdoid tumor, 1 ganglioneuroblastoma, and 7 sarcoma cases, not otherwise specified. Bilateral cases in the lymphoma group included 22 cases of Hodgkin lymphoma; 10 diffuse large B-cell lymphomas; 5 anaplastic large cell lymphomas; 5 Burkitt lymphomas; 1 T-cell lymphoma, not otherwise specified; and 1 “malignant hematolymphoid tumor.” Bilateral cases in the leukemia group included 13 cases of precursor B- or precursor T-cell leukemia lymphoma; 3 acute myeloid leukemias, not otherwise specified; and 2 myelodysplastic syndromes, not otherwise specified. Bilateral cases in the “other hematologic conditions” group included 11 cases of aplastic anemia, 2 cases of Langerhans cell histiocytosis, 1 Kikuchi lymphadenitis, 1 pancytopenia, 1 neutropenia, 1 autoimmune hemolytic anemia, 1 idiopathic thrombocytopenic purpura, 1 Shwachman-Diamond syndrome, 1 familial thrombocytopenia, 1 polycythemia, 1 Evans syndrome/common variable immunodeficiency, and 4 cases of reactive inflammatory conditions, not otherwise specified.

The specimens were also grouped according to which combinations of ancillary studies were ordered. Some bilateral cases were submitted for classical cytogenetic analysis only, some cases were submitted for FISH only, and some cases were submitted for classical cytogenetics and FISH.

Results

Results are summarized in Table 3.

Results by Clinical History

In the solid tumor category, 75 (97%) of 77 cases were in agreement. There were 27 cases in the solid tumor category in which cytogenetic analysis only was performed; all were in agreement (100%). Of these 27 cases, 26 showed a normal karyotype (96%), and 1 case showed an abnormal karyotype (4%). There were 21 cases in the solid tumor category in which FISH only was performed; 1 case showed disagreement (95% agreement). Of the 20 cases that were in agreement, 18 were negative (90%) and 2 were positive (10%). There were 29 cases in the solid tumor category in which both cytogenetics and FISH were performed; 1 case showed disagreement (97% agreement). Of the 28 cases that were in agreement, 23 cases had a normal karyotype with negative FISH results (–/–; 82%), 2 cases had an abnormal karyotype with negative FISH results (+/–; 7%), 1 case had a normal karyotype with positive FISH results (–/+; 4%), and 2 cases had an abnormal karyotype with positive FISH results (+/+; 7%). One case disagreed in cytogenetic and FISH findings (4%).

In the lymphoma category, all 44 cases were in agreement (100%). There were 32 cases in which cytogenetic analysis only was performed; all were in agreement (100%). The karyotypic findings in these cases were normal (100%). FISH alone was not requested on any lymphoma biopsy specimens. There were 12 cases in the lymphoma category in which both cytogenetics and FISH were performed; all showed agreement (100%). In 9 of these 12 cases, there was a normal karyotype with negative FISH findings (–/–; 75%), while 3 cases showed an abnormal karyotype and positive FISH findings (+/+; 25%).

View this table:
Table 2
View this table:
Table 3

In the leukemia category, all 18 cases were in agreement (100%). There were 15 cases in which cytogenetic analysis only was performed; all were in agreement (100%). Nine of these 15 cases showed normal karyotypes (60%), and 6 cases showed karyotypic abnormalities (40%). FISH alone was not requested in any leukemia biopsy specimens. There were 3 cases in the leukemia category in which both cytogenetics and FISH were performed; all were in agreement (100%). All 3 cases showed a normal karyotype and negative FISH findings.

In the other hematologic conditions category, all 26 cases were in agreement (100%). There were 24 cases in which cytogenetics only was performed; all were in agreement (100%). All cases showed a normal karyotype (100%). FISH alone was not requested on any other biopsy specimens from this category. There were 2 cases in this category in which both cytogenetics and FISH were performed; both were in agreement (100%). Both cases showed a normal karyotype with negative FISH findings (–/–).

The case for which the underlying disorder was unknown had agreement between the bilateral biopsies. This case had cytogenetics only performed, which showed an abnormal karyotype.

Results by Ancillary Studies

Cytogenetics was performed in a total of 99 bilateral bone marrow biopsy cases, with 100% agreement between the 2 sides. FISH-only analysis was performed on a total of 21 bilateral bone marrow biopsy cases, with 1 case showing disagreement (95% agreement). The case that disagreed was a rhabdomyosarcoma from the solid tumor category. Both cytogenetics and FISH were performed in a total of 46 bilateral bone marrow biopsy cases, with 1 case showing disagreement (98% agreement). The case that disagreed was a neuroblastoma from the solid tumor category. Both cytogenetics and FISH results on this case disagreed.

Discussion

The value of bilateral bone marrow biopsy specimens for histologic examination and staging in certain clinical situations has been well established.35 The role of bilateral biopsies for ancillary studies such as classical cytogenetics or FISH has not been as clearly defined.

Bone marrow biopsy is considered a low-risk procedure. Bain1 and Feeney and Barry2 in the United Kingdom reported a complication rate of 0.12%; however, no procedure is completely risk-free. In addition to the procedural adverse events, pediatric patients may be apprehensive about undergoing additional procedures. The benefit of the bilateral biopsy must be weighed against the potential risks to the patient, the cumulative emotional toll of additional procedures, and the cost of any ancillary studies obtained. In the current study, we asked if analysis of bilateral bone marrow biopsy specimens is superior to analysis of one unilateral biopsy specimen for use in ancillary genetic studies.

In this retrospective study, we reviewed the agreement in the genetic findings of 166 bilateral pediatric bone marrow biopsy specimens. The discordance between the 2 sides was low, at 1.2%. For the purpose of analysis, the specimens were grouped into categories of clinical history and also grouped into cases that were sent for cytogenetics only, FISH only, and both cytogenetics and FISH. We did not group the biopsies according to the reasons for the procedures, such as FISH for specific translocations to assess minimal residual disease following stem cell transplantation, diagnostic bone marrow samples to determine if the marrow was involved, or specimens obtained to assess the cytogenetic response criteria for a given disease. Presumably these are some of the reasons the biopsies were performed, but we did not have enough history about each individual case to definitively group them in this way.

The solid tumor category was the largest, with 77 bilateral cases, possibly reflecting the heavy referral base for pediatric sarcomas in our institution. The lymphoma category was the second largest category, with 44 bilateral cases. The nonneoplastic hematologic category included 26 bilateral cases, while the leukemia category contained 18 bilateral cases. Overall, there was a high level of agreement between the bilateral biopsies, even given the diverse clinical histories. Both cases that showed disagreement were in the solid tumor category.

One explanation for this may be that by its nature, bone marrow involvement by solid tumors may be patchy. The lymphoma group was made up of fewer specimens but showed complete agreement between the sides. The histologic evidence of marrow involvement by lymphoma has been known to vary among different subtypes. Juneja and colleagues11 reported that bilateral biopsies in the setting of small lymphocytic lymphoma were found to have a high rate of concordance (100%) and that diffuse large B-cell lymphoma had a lower level of agreement between the 2 sides (50%). One reason for the high rate of concordance among the disparate lymphoma cases in this study may be that the molecular assays are detecting small amounts of occult disease that may not be histologically apparent. The leukemia category also showed complete agreement, perhaps reflecting the biology of leukemia as a systemic disease of the bone marrow. In addition, the nonneoplastic hematologic group showed complete agreement. One may argue that this would be because many of the diseases in this group may result from diseases intrinsic to the marrow; however, the cytogenetics for all of these cases were normal, and the FISH studies were also all negative. No evidence of intrinsic marrow disease was found using these ancillary techniques.

Many of the specimens in this study were negative or had normal findings. A negative study may be truly negative or could represent a false-negative owing to sampling bias. In this context, there could be false agreement between the sides because one side may be truly negative with no disease in the region, while the other side may be falsely negative via sampling bias. The 2 sides would be erroneously in agreement. Positive studies would carry more weight in this regard. Future studies using a subset of positive bone marrow samples may help decrease this bias.

Clinical decisions such as bilateral vs unilateral biopsies are made in the context of the patient’s individual disease course and overall treatment plan. In the setting of solid tumors, only 2 of 77 cases showed discordance. It remains to be seen if the high concordance will influence clinicians to move to unilateral biopsies in this group. Perhaps a stronger case is made for the other categories: lymphoma, leukemia, and other hematologic conditions, because 100% concordance was seen. Future prospective studies are warranted to test the reproducibility of these findings.

One question that follows from these findings is whether bilateral biopsies might be safely pooled without dilution. Currently each side is analyzed and processed as a separate sample. This level of separation may be redundant, using increased time and reagents in the laboratory. Previous researchers in this area questioned whether pooling would result in dilution of the specimen, making it less likely to detect occult disease. Murphy and colleagues12 found a very low discordance rate (2.9%) in bilateral specimens performed for immunoglobulin and T-cell receptor gene rearrangements. They concluded that bilateral specimens may be pooled without unacceptable risk of false-negatives owing to dilution. From the standpoint of the laboratory, pooling the biopsies is cost-effective and would decrease the workload. These findings would seem to support that pooling specimens may be one possible solution if a clinician is not comfortable sampling only 1 side. The results of our study analysis raise the question of pooling; however, this is a question that cannot be answered by a retrospective review. Again, further study in this area is warranted.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
View Abstract