Non-Parametric Combined Reference Regions and Prediction of Clinical Risk.

BACKGROUND: Many clinical decisions depend on estimating patient risk of clinical outcomes by interpreting test results relative to reference intervals, but standard application of reference intervals suffers from two major limitations that reduce the accuracy of clinical decisions: (1) each test result is assessed separately relative to a univariate reference interval, ignoring the rich pathophysiologic information in multivariate relationships, and (2) reference intervals are intended to reflect a population's biological characteristics and are not calibrated for outcome prediction. METHODS: We developed a combined reference region (CRR), derived CRRs for some pairs of complete blood count (CBC) indices (RBC, MCH, RDW, WBC, PLT), and assessed whether the CRR could enhance the univariate reference interval's prediction of a general clinical outcome, 5-year mortality risk (MR). RESULTS: The CRR significantly improved MR estimation for 21/21 patient subsets defined by current univariate reference intervals. The CRR identified individuals with >2-fold increase in MR in many cases and uniformly improved the accuracy for all five pairs of tests considered. Overall, the 95% CRR identified individuals with a >7× increase in 5-year MR. CONCLUSIONS: The CRR enhances the accuracy of the prediction of 5-year MR relative to current univariate reference intervals. The CRR generalizes to higher numbers of tests or biomarkers, as well as to clinical outcomes more specific than MR, and may provide a general way to use existing data to enhance the accuracy and precision of clinical decisions.

Mechanistic modeling of hemoglobin glycation and red blood cell kinetics enables personalized diabetes monitoring.

The amount of glycated hemoglobin (HbA1c) in diabetic patients' blood provides the best estimate of the average blood glucose concentration over the preceding 2 to 3 months. It is therefore essential for disease management and is the best predictor of disease complications. Nevertheless, substantial unexplained glucose-independent variation in HbA1c makes its reflection of average glucose inaccurate and limits the precision of medical care for diabetics. The true average glucose concentration of a nondiabetic and a poorly controlled diabetic may differ by less than 15 mg/dl, but patients with identical HbA1c values may have true average glucose concentrations that differ by more than 60 mg/dl. We combined a mechanistic mathematical model of hemoglobin glycation and red blood cell kinetics with large sets of within-patient glucose measurements to derive patient-specific estimates of nonglycemic determinants of HbA1c, including mean red blood cell age. We found that between-patient variation in derived mean red blood cell age explains all glucose-independent variation in HbA1c. We then used our model to personalize prospective estimates of average glucose and reduced errors by more than 50% in four independent groups of greater than 200 patients. The current standard of care provided average glucose estimates with errors >15 mg/dl for one in three patients. Our patient-specific method reduced this error rate to 1 in 10. Our personalized approach should improve medical care for diabetes using existing clinical measurements.

In vivo volume and hemoglobin dynamics of human red blood cells.

Human red blood cells (RBCs) lose ∼30% of their volume and ∼20% of their hemoglobin (Hb) content during their ∼100-day lifespan in the bloodstream. These observations are well-documented, but the mechanisms for these volume and hemoglobin loss events are not clear. RBCs shed hemoglobin-containing vesicles during their life in the circulation, and this process is thought to dominate the changes in the RBC physical characteristics occurring during maturation. We combine theory with single-cell measurements to investigate the impact of vesiculation on the reduction in volume, Hb mass, and membrane. We show that vesicle shedding alone is sufficient to explain membrane losses but not volume or Hb losses. We use dry mass measurements of human RBCs to validate the models and to propose that additional unknown mechanisms control volume and Hb reduction and are responsible for ∼90% of the observed reduction. RBC population characteristics are used in the clinic to monitor and diagnose a wide range of conditions including malnutrition, inflammation, and cancer. Quantitative characterization of cellular maturation processes may help in the early detection of clinical conditions where maturation patterns are altered.

Quantitative absorption cytometry for measuring red blood cell hemoglobin mass and volume.

We present an optical system, called the quantitative absorption cytometer (QAC), to measure the volume and hemoglobin mass of red blood cells flowing through a microfluidic channel. In contrast to clinical hematology analyzers, where cells are sphered in order for both volume and hemoglobin to be measured accurately, the QAC measures cells in their normal physiological shape. Human red blood cells are suspended in a refractive index-matching absorbing buffer, driven through a microfluidic channel, and imaged using a transmission light microscope onto a color camera. A red and a blue LED illuminate cells and images at each color are used to independently retrieve cell volume and hemoglobin mass. This system shows good agreement with red blood cell indices retrieved by a clinical hematology analyzer and in fact measures a smaller coefficient of variation of hemoglobin concentration. In addition to cell indices, the QAC returns height and mass maps of each measured cell. These quantitative images are valuable for analyzing the detailed morphology of individual cells as well as statistical outliers found in the data. We also measured red blood cells in hypertonic and hypotonic buffers to quantify the correlation between volume and hemoglobin mass under osmotic stress. Because this method is invariant to cell shape, even extremely nonspherical cells in hypertonic buffers can be measured accurately.

Developmental plasticity of red blood cell homeostasis.

Most human physiologic set points like body temperature are tightly regulated and show little variation between healthy individuals. Red blood cell (RBC) characteristics such as hematocrit and mean cell volume are stable within individuals but can vary by 20% from one healthy person to the next. The mechanisms for the majority of this inter-individual variation are unknown and do not appear to involve common genetic variation. Here, we show that environmental conditions present during development, namely in utero iron availability, can exert long-term influence on a set point related to the RBC life cycle. In a controlled study of rhesus monkeys and a retrospective study of humans, we use a mathematical model of in vivo RBC population dynamics to show that in utero iron deficiency is associated with a lowered threshold for RBC clearance and turnover. This in utero effect is plastic, persisting at least 2 years after birth and after the cessation of iron deficiency. Our study reports a rare instance of developmental plasticity in the human hematologic system and also shows how mathematical modeling can be used to identify cellular mechanisms involved in the adaptive control of homeostatic set points.

Evidence for bistable bacteria-neutrophil interaction and its clinical implications.

Neutropenia, which may develop as a consequence of chemotherapy, increases the risk of bacterial infection. Similarly, increased risk of bacterial infection appears in disorders of phagocytic functions, such as the genetic disorder chronic granulomatous disease. To elucidate the organizing principles behind these distinct immunodeficiency conditions, we investigated the interaction between in vitro bacteria and human neutrophils by experiments and mathematical modeling. The model and the experiments showed that the in vitro bacterial dynamics exhibit bistability for a certain range of neutrophil concentration and function. Thus, there is a critical bacterial concentration above which infection develops, and below which neutrophils defeat the bacteria. Whereas with normal neutrophil concentration and function, an infection may develop when the initial bacterial concentration is very high, under neutropenic conditions or when there is neutrophil dysfunction, the critical bacterial concentration can be lower, within the clinically relevant range. We conclude that critical bacterial concentration has clinically relevant implications. The individual maximum bearable bacterial concentration depended on neutrophil concentration, phagocytic activity, and patient barrier integrity; thus, the resulting maximal bearable bacterial concentration may vary by orders of magnitude between patients. Understanding the interplay between neutrophils and bacteria may enhance the development of new therapeutic approaches to bacterial infections.

Bacteria-phagocyte dynamics, axiomatic modelling and mass-action kinetics.

Axiomatic modeling is ensued to provide a family of models that describe bacterial growth in the presence of phagocytes, or, more generally, prey dynamics in a large spatially homogenous eco-system. A classification of the possible bifurcation diagrams that arise in such models is presented. It is shown that other commonly used models that do not belong to this class may miss important features that are associated with the limited growth curve of the bacteria (prey) and the saturation associated with the phagocytosis (predator kill) term. Notably, these features appear at relatively low concentrations, much below the saturation range. Finally, combining this model with a model of neutrophil dynamics in the blood after chemotherapy treatments we obtain new insights regarding the development of infections under neutropenic conditions.

Bistability and bacterial infections.

Bacterial infections occur when the natural host defenses are overwhelmed by invading bacteria. The main component of the host defense is impaired when neutrophil count or function is too low, putting the host at great risk of developing an acute infection. In people with intact immune systems, neutrophil count increases during bacterial infection. However, there are two important clinical cases in which they remain constant: a) in patients with neutropenic-associated conditions, such as those undergoing chemotherapy at the nadir (the minimum clinically observable neutrophil level); b) in ex vivo examination of the patient's neutrophil bactericidal activity. Here we study bacterial population dynamics under fixed neutrophil levels by mathematical modelling. We show that under reasonable biological assumptions, there are only two possible scenarios: 1) Bacterial behavior is monostable: it always converges to a stable equilibrium of bacterial concentration which only depends, in a gradual manner, on the neutrophil level (and not on the initial bacterial level). We call such a behavior type I dynamics. 2) The bacterial dynamics is bistable for some range of neutrophil levels. We call such a behavior type II dynamics. In the bistable case (type II), one equilibrium corresponds to a healthy state whereas the other corresponds to a fulminant bacterial infection. We demonstrate that published data of in vitro Staphylococcus epidermidis bactericidal experiments are inconsistent with both the type I dynamics and the commonly used linear model and are consistent with type II dynamics. We argue that type II dynamics is a plausible mechanism for the development of a fulminant infection.