Efficient assessment of peripheral blood lymphocytosis in adults: developing new thresholds for blood smear review by pathologists.

BACKGROUND

Peripheral smear review is a critical, but labor intensive adjunct for evaluation of lymphocytosis. Standard practice based on consensus guidelines is to review cases with absolute lymphocyte count (ALC) >5×109/L. We hypothesize that identifying cases for review by applying appropriately adjusted ALC and age discriminators will decrease laboratory workload without compromising patient care.

METHODS

1170 complete blood counts with ALCs >5×109/L analyzed in the core laboratory during a 2-year period were included. Patients were categorized into diagnostic groups based on follow-up criteria. A total of 402 patients with new onset lymphocytosis who met criteria for reactive lymphocytosis (82%) or lymphoproliferative disorder (18%) were used to establish optimal ALC and age thresholds from receiver operating characteristic (ROC) curve analysis.

RESULTS

ALC as a discriminator for neoplastic lymphocytosis had an ROC area under the curve (AUC) of 0.732. Selecting cases with ALC >10×109/L enriched the proportion of neoplastic cases in the review pool (90% specificity); however, many cases with ALC below this threshold were also neoplastic (52% sensitivity). For cases with ALC between 5 and 10×109/L, age as a discriminator had an ROC AUC of 0.886. Selecting patients >50 years old in this group for review captured the neoplastic cases while excluding the reactive cases (93% sensitivity, 62% specificity). When applied to a validation cohort, the predictive performance of the thresholds was maintained while reducing smears reviewed by 50%.

CONCLUSIONS

We show that modifying the standard 5×109/L ALC smear review threshold through retrospective analysis of institutional data can reduce laboratory workload without compromising quality.