Diagnostic utility of cerebrospinal fluid flow cytometry in patients with and without prior hematologic malignancy

The spectrum of malignant diagnoses in cerebrospinal fluid cytology from an adult population: a multi-institutional retrospective review

INTRODUCTION

Limited updated literature exists about the prevalence and spectrum of malignancies involving cerebrospinal fluid (CSF). In this multi-institutional study, we review our experience with focus on first time malignancy diagnosis in CSF samples of adults.

MATERIALS AND METHODS

Institutional databases at 4 academic centers were queried retrospectively for CSFs over a 10-year period. The following data elements were collected: total # of CSFs, total # of CSFs with a malignant diagnosis; for each patient with a first time CSF diagnosis of malignancy: age, gender, diagnosis, prior history of malignancy, and ancillary studies.

RESULTS

Twenty-four thousand one hundred forty-two CSFs were collected with a positive for malignancy rate of 2.3% (n = 551). Out of 347 (1.4%) adults with a first-time diagnosis of CSF malignancy 182 (52%) were female (age range: 19-89/mean: 57) and 165 (48%) were male (age range: 20-95/mean: 60). Hematolymphoid malignancies (48%, n = 168) were overall the most common neoplasm. In women, metastatic carcinomas (63%, n = 114) were the leading malignancy, of which the majority were breast primaries. In men, lymphomas/leukemias (64%, n = 106) were the leading malignancy, of which the majority were B-cell lymphomas. Ancillary studies aided the final diagnosis in 110 (32%) cases. For 286 (82%) cases, a prior history of malignancy was available to correlate CSF findings.

CONCLUSIONS

A malignancy diagnosis in the CSF of adults is rare. The most common malignancies in females and males are metastatic breast carcinoma and hematolymphoid malignancies, respectively. Metastatic neoplasms account for the majority, with primary central nervous system neoplasms being quite uncommon. History of malignancy and ancillary tests can be helpful.

Cerebrospinal Fluid Flow Cytometry: Utility in Central Nervous System Lymphoma Diagnosis

BACKGROUND

Flow cytometry of the cerebrospinal fluid (CSF) is used in isolation or as an adjunct to cytology to increase the sensitivity of detecting central nervous system (CNS) lymphoma. We aimed to evaluate the sensitivity of CSF flow cytometry as a diagnostic screening tool for primary CNS lymphoma in patients presenting with undifferentiated neurologic symptoms.

METHODS

We retrospectively reviewed all CSF samples received by the Calgary Laboratory Services Flow Cytometry Laboratory from 2012 to 2015. Clinical data, laboratory investigations, radiologic imaging studies, and pathological data were analyzed. Clinical review extended to 2 years post-CSF flow cytometric testing.

RESULTS

Only 43/763 (5.6%) samples of CSF flow cytometry in 28/573 (4.9%) patients were found to be positive for a hematological malignancy in patients with undifferentiated neurologic symptoms. The overall sensitivity of the test was 13.8% with 25 patients with negative CSF flow cytometry later having a positive biopsy for CNS lymphoma. CSF flow cytometry was negative in all cases when at the time of CSF examination the patient did not have a previous hematological malignancy or findings of abnormal enhancement on MRI (n = 249).

CONCLUSION

CSF flow cytometry has low utility in screening for primary CNS lymphoma in the absence of a previous history of hematologic malignancy or findings of abnormal enhancement on MRI.

Improving Malignancy Detection Rates in Body Fluids Submitted to the Hematology Laboratory for Nucleated Cell Count and Differential: A Quality Improvement Study

CONTEXT.—

Body fluid specimens are regularly submitted to the hematology laboratory for cell count and differential. Unless there is high clinical suspicion for malignancy, most cases lack concurrent cytology review and may not benefit from more focused examination for malignancy.

OBJECTIVE.—

To compare rates of malignancy detection before and after fluid-focused training for hematology technologists as part of a quality improvement initiative.

DESIGN.—

During an 8-week pretraining period, body fluids submitted to the cytology laboratory were correlated with concurrent hematology specimens. After slide review and training sessions for the hematology technologists, the same data were collected for a 4-week period. Discrepant cases were reviewed by hematology laboratory supervisors and pathologists.

RESULTS.—

We collected 465 pretraining and 249 posttraining body fluids with concurrent cytology and hematology evaluation. In the pretraining cohort, 48 cases (10.3%) were diagnosed as malignant by cytology; of those, 33 were detected by hematology. In the posttraining cohort, 30 cases (12.0%) were diagnosed as malignant by cytology of which 27 were detected by hematology. Of the 18 discrepant cases (all carcinomas), hematology slide review showed definite features of malignancy in 15 and no tumor cells in 3. The malignancy detection rate by the hematology laboratory significantly improved after training (68.8% versus 90.0%, P = .01).

CONCLUSIONS.—

We demonstrate the comparatively lower malignancy detection rate for body fluid specimens processed in our hematology laboratory, particularly for carcinomas. Hematology technologist education/training improved the malignancy detection rate, an important quality improvement given the large proportion of body fluids undergoing hematology evaluation without concurrent cytology reviews.

Machine Learning Models Improve the Diagnostic Yield of Peripheral Blood Flow Cytometry

OBJECTIVES

Peripheral blood flow cytometry (PBFC) is useful for evaluating circulating hematologic malignancies (HM) but has limited diagnostic value for screening. We used machine learning to evaluate whether clinical history and CBC/differential parameters could improve PBFC utilization.

METHODS

PBFC cases with concurrent/recent CBC/differential were split into training (n = 626) and test (n = 159) cohorts. We classified PBFC results with abnormal blast/lymphoid populations as positive and used two models to predict results.

RESULTS

Positive PBFC results were seen in 58% and 21% of training cases with and without prior HM (P < .001). % neutrophils, absolute lymphocyte count, and % blasts/other cells differed significantly between positive and negative PBFC groups (areas under the curve [AUC] > 0.7). Among test cases, a decision tree model achieved 98% sensitivity and 65% specificity (AUC = 0.906). A logistic regression model achieved 100% sensitivity and 54% specificity (AUC = 0.919).

CONCLUSIONS

We outline machine learning-based triaging strategies to decrease unnecessary utilization of PBFC by 35% to 40%.

Cerebrospinal fluid from patients with hematopoietic neoplasms: the need for an integrated approach

INTRODUCTION

Cerebrospinal fluid (CSF) evaluation plays an increasing role in the diagnosis and staging of hematopoietic neoplasms. The evaluation is based on the cytomorphologic evaluation (CE) of the specimen and flow cytometry study (FCS). The impact of the increased sensitivity of multicolor FCS and its correlation with the morphological analysis of CSF needs to be evaluated to better guide clinical management.

MATERIALS AND METHODS

CSF specimens sent for CE and FCS obtained over a 9-month period were retrospectively analyzed. Cases were considered completely discordant if one method detected an abnormal hematologic population and the corresponding method was negative and partially discordant if FCS detected an abnormal hematologic population and the CE was atypical or suspicious. Root cause analysis of these discrepancies was performed.

RESULTS

A total of 78 of 361 cases (22%) had discordant results; 72 cases were from patients with hematopoietic neoplasms-22 cases were completely discordant and 50 were partially discordant. FCS had a sensitivity of 95.3% and a specificity of 98.1% for detecting abnormal hematopoietic populations. CE rendered a positive diagnosis in 17.6% of cases with a specificity of 100%.

CONCLUSIONS

Our series demonstrates evaluation of CSF involvement by hematopoietic malignancy with FCS is more sensitive than and equally specific as CE. Nonetheless, CE remains one of the mainstays in the evaluation of CSF specimens. Optimizing cytologic evaluation through process modifications including decreasing screening area and evaluation of multiple preparations increased the detection rate of malignant cells. As such, optimal CSF evaluation and integration FCS is critical in patient management.

Accuracy of flow cytometry and cytomorphology for the diagnosis of meningeal involvement in lymphoid neoplasms: A systematic review

Central nervous system (CNS) involvement by lymphoid neoplasms is a relatively infrequent event that demands accurate identification. The purpose of this article is to review studies comparing diagnostic accuracy of flow cytometry (FCM) and cytomorphology (CM) for meningeal involvement from lymphoid neoplasms. Primary publications from the last 26 years were identified searching MedLine, Scopus, and Web of Science and systematically scanning bibliographies of identified articles. Only studies reporting complete results were included. We assessed study quality using the QUADAS-2 tool. For each study, we extracted informations regarding study population, technical details about sample preparation, data analysis, and results. Twenty-seven studies were included. A great heterogeneity regarding study populations and analytical procedures was observed among studies. Percentages of samples giving a positive result with both FCM and CM range from 0.3% to 42.9% among studies, whereas double negative samples go from 0% to 96.3%. Samples with positive FCM but negative CM are reported by 89% (24/27) of the studies with rates ranging from 3.5% to 61.5% of total specimens. On the contrary, samples with positive CM and negative FCM are found in 48% (13/27) of the studies with percentages ranging from 0.5% to 10%. Despite all the differences observed among studies, almost all of them state that employing flow cytometry along with conventional cytology increases the number of positive CSF samples for lymphoma involvement, although a few cases remain in whom only morphology can correctly identify malignant cells. Diagn. Cytopathol. 2016;44:841-856. © 2016 Wiley Periodicals, Inc.