Evidence-Based, Patient-Specific Guidelines Provide Efficient and Cost-Effective Molecular and Cytogenetic Testing in Hematologic Malignancy

Molecular and cytogenetic testing is critical in diagnosis and management of hematologic malignancies. These tests are expensive, complex, and frequently inappropriately ordered. There are no comprehensive guidelines to assist in the disease- and patient-specific selection of these tests. Traditionally, most tests are ordered before morphologic evaluation of the specimen, which could help guide test selection. We hypothesize that this leads to significant over-ordering of tests unnecessary for diagnosis or monitoring, as well as under-ordering of necessary tests. As part of the hematopathology diagnostic management team (DMT) effort, we evaluated this hypothesis by establishing evidence-based rules for molecular and cytogenetic test ordering that are applied by the pathologist after review of clinical history and bone marrow aspirate smears.

Using published evidence and best clinical practices, a team of clinical hematologists and hematopathologists developed rules, termed standard operating procedures (SOPs), for selecting cytogenetic and molecular tests on bone marrow biopsies from adult patients performed for myelodysplastic syndrome/acute myeloid leukemia (MDS/AML), plasma cell myeloma, lymphoma, myeloproliferative neoplasms, lymphoblastic leukemia, and bone marrow failure. The rules specify testing in five clinical scenarios: diagnosis or follow-up with morphologically overt disease, bone marrow staging (for lymphoma), routine follow-up without overt disease, evaluation prior to stem cell transplant (pre-SCT), and follow-up after stem cell transplant (post-SCT). To determine the effect that application of these SOPs would have had on test ordering and results, bone marrow biopsies from adults (>18 years) were evaluated retrospectively over a six-month time period. Each test was ruled concordant or discordant with the corresponding SOP. Omitted tests, those recommended by the SOPs, but not performed, were also identified. Statistical analysis of concordant and discordant tests was performed using the Chi-square test.

A total of 3,007 tests (769 karyotypes, 1,790 FISH, and 448 molecular tests) were performed on 804 adult bone marrows. Of these, 1,080 (36%) were discordant with the SOPs. Discordant tests were most frequent in myeloma (471), MDS/AML (231), and lymphoma (215). The most common clinical settings for discordant tests were routine follow-up (403) and post-SCT (417). To determine if these tests were necessary, we determined the positive rate of both concordant and discordant tests. The positive rate was significantly higher in the concordant (490/1,929; 25%) than the discordant tests (38/1,078; 4%) (P<0.001). Of the 38 positive discordant tests, 15 were redundant, i.e., a more sensitive test for the abnormality was positive on the same sample. An additional 13 discordant tests were only weakly positive. Of these, most were negative at diagnosis or at subsequent bone marrow testing, suggesting that they may represent false positive results. Thus, only 10/1,080 (<1%) discordant tests had potential clinical significance. In 145 of the marrows (18%), there were 308 omitted tests. By eliminating the discordant tests and including the omitted tests, we estimate that institution of these SOPs would reduce the total number of cytogenetic and molecular tests by 26%, from 3,007 to 2,235.

Retrospective analysis of cytogenetic and molecular test ordering demonstrates that more than one-third of cytogenetic or molecular tests ordered on bone marrow specimens at a large tertiary care institution were discordant with rules (SOPs) based on published literature and best clinical practices. Furthermore, the vast majority (>99%) of these discordant tests were negative for any abnormality, redundant, or potential false positives, suggesting that the exclusion of these tests would have little negative clinical impact. In addition, there were a large number of omitted tests, seen in 18% of all marrows. This analysis predicts that institution of rules-based morphology-influenced test ordering would decrease uninformative testing, while increasing appropriate informative tests, leading to more cost-effective, accurate, and complete diagnosis and monitoring of hematolymphoid malignancies. We have therefore instituted a rules-based, morphology-directed approach at our institution.

No relevant conflicts of interest to declare.