Brief Synopsis: “Instruction Manual for Juvenile Clinical Pathology”

Abstract

This symposium synopsis summarizes key points discussed related to clinical pathology data interpretation for reproduction and juvenile toxicology studies. In pregnant and growing animals, several changes in clinical pathology parameters linked to growth/maturation of organ and physiological functions can occur, and understanding these changes is important to enable accurate interpretation of clinical pathology data. A brief overview of the general approach to clinical pathology data analysis according to contemporary practices is provided, followed by a discussion focused specifically on reproductive and juvenile clinical pathology. In this context, the approach to recognize and differentiate changes that may be related to pregnancy and growth as opposed to those that may be related to test article effects is highlighted.

The Evolution of the Approach to Clinical Pathology Data Interpretation

There is a large amount of clinical pathology data that need to be carefully considered in typical toxicology studies. For example, for a rat study with 4 dose groups comprising 10 rats/group, ∼30 parameters/animal, and 1 timepoint, there are ∼1200 data points to evaluate. The purpose of the evaluation is rather straightforward: (a) to investigate whether there is a relationship between a change in any given parameter to test item administration and/or experimental conditions, (b) to determine the relevance of the change to human risk assessment, and (c) to assess adversity of the changes (if warranted). There is no single valid method for clinical pathology data evaluation given the many variables that need to be considered, and, rather a “weight of evidence” or integrative approach is recommended.

The traditional approach of data evaluation consists primarily of comparison of dosed groups to control groups using mean data to look for statistical relationships or occasionally, comparison against reference intervals. However, this basic approach has needed to evolve as a result of a more diverse array of test items (eg, small and large molecules, lipid nanoparticles, complicated proteins), inclusion of multiple control groups and reference items, addition of functional assays, and expansion of the various routes of test article administration (eg, intrathecally), among many aspects to consider, which complicate data sets and renders the traditional approach limited. An integrative “weight of evidence” approach to data analysis is considered an industry best practice ^{1 -3} and examines the multiple various factors that can impact clinical pathology data including, but not limited to, preanalytical and analytical variables, species differences, the biological significance of a given parameter at a particular magnitude of change, the overall study design, exposure data, and in-life and anatomic pathology findings.

Preanalytical variables that are particularly critical to consider in juvenile studies include for example “age,” which can render an animal more susceptible to particular changes, such as excitement or fear which can, in turn, impact leukocyte parameters or “fasting status,” which can affect carbohydrate metabolism markers since juvenile animals are typically not fasted prior to blood collection. Given their smaller body sizes, one must also consider the impact of blood loss due to the difference in the ratio of required blood sample volume to body weight versus an adult animal. Other preanalytical variables that need to be accounted for are a route of sample collection, type of anesthesia, order of sample collection, study design, and the anatomical site of collection (eg, cardiac puncture). Analytical variables such as sample handling, quality of assay validation, and sample stability can also impact clinical pathology data. ^{4 -6}

Reference intervals are best established by the laboratory performing the analysis of the clinical pathology data and can be helpful to appreciate general differences in adult versus juvenile animals resulting from the stage of growth and maturation. Reference intervals, however, should not typically be used as the primary source of comparison to determine test item-related changes and values within the reference interval should not be automatically considered “normal” or unrelated to test item administration as trends can be present within a “normal range.” ⁷ For a few particular study types (ie, agrochemical or target animal safety studies) the use of reference intervals may be required; however, an integrative approach should still be applied for the interpretation of clinical pathology data and determination of test item–related changes. Statistics can be useful but should not be a substitute for a thorough review and it is important to be mindful of the adage that not all statistically significant changes are test article–related and not all test article–related changes are statistically significant. Because there is no one standard approach for the statistical analysis of clinical pathology data, the analysis should be dictated by the unique type, variability, and distribution of the data set in hand and a consultation with a statistician can be invaluable for analysis of complicated study designs and data sets. Multiple, excellent reviews on this subject have been authored by veterinary clinical pathologists, and the reader is encouraged to consult these. ^7,8

Pregnancy Considerations on Clinical Pathology Data

There are many physiological changes in normal, healthy pregnant animals that will affect hematology, coagulation, and clinical chemistry parameters which should not be confused with pathological states. One of the most striking changes in pregnancy is the expansion of plasma volume (30%-50% from pre-pregnancy values in humans) and increased red blood cell (RBC) volume; however, these increases do not occur simultaneously. This sets up a scenario where there is an increased demand for the production of new RBCs, typically a nonhypoxia induced increase in erythropoietin production, which can sometimes result in decreased cell life span of RBCs due to the accelerated rate of production. Because this plasma expansion outpaces RBC production it can result in hemodilution, a “relative anemia,” whereby RBC count, hemoglobin, and hematocrit (HCT) values are decreased. Platelet counts may also be similarly decreased. There are well-described white blood cell count changes in human pregnancy, whereby neutrophil and lymphocyte counts will increase in late pregnancy and total white blood cell counts increase at delivery due to the contribution of stress. In rats, there are limited data, but there may be increases in neutrophils and lymphocytes seen close to the time of labor.

Coagulation parameters may be impacted, as pregnancy is a hypercoagulable state in many mammals. In women, pregnancy-related increases in fibrinogen, erythrocyte sedimentation rate, von Willebrand factor, and production of clotting factors VII, VII, and X and decreased production of factor XIII have been reported which can lead to prolongations in activated partial thromboplastin time (APTT) and prothrombin time (PT). Similarly, in dogs, there are reports of pregnancy-related increases in factors VII, VIII, IX, and XI production ⁹ and prolongations of APTT and PT times in rats. ¹⁰

There are many metabolic functions that are altered during pregnancy. For example, there are changes in hepatic enzyme activities which can impact data assessment. In women, decreases in alanine and aspartate aminotransferase activities (ALT and AST) and increases in alkaline phosphatase (ALP) from placental origin have been reported, all of which can make the diagnosis of liver abnormalities challenging as pathological increases in transaminase activities may not exceed reference intervals in pregnant women with liver disease. In our preclinical species, there are no clear trends for alterations in enzyme activities, however, in most species, decreases in ALP have been observed. ¹¹ The kidney can also be impacted, with increases seen in renal plasma flow and decreased glomerular filtration rate (GFR) which can subsequently lead to decreases in serum urea nitrogen and creatinine concentrations (women, dogs, rabbits, and cynomolgus monkeys) ^11,12 and rats (personal observation—AP). Sodium and water balance typically remain unchanged. There is an increase in lipid mobilization from maternal adipose tissue which can present as increased triglyceride, cholesterol, phospholipid, and free fatty acid concentrations and varies depending on the level of fetal organ development. Cholesterol and triglycerides are generally in higher demand during early organogenesis in rats, rabbits, and dogs, except for monkeys. ¹¹ As a result of the hemodilution described above, there can be resultant decreased serum albumin with or without concomitantly decreased serum protein concentrations. Globulin concentrations show variable trends. Glucose demand is higher in early pregnancy due to organogenesis and late pregnancy. There can be insulin resistance, with a decrease in the renal glucose threshold and increased GFR resulting in glucosuria with no change in serum glucose concentrations. In our preclinical species, these changes are not as consistently seen as they are in women. ¹¹

Fetal demands for bone mineralization will result in increased maternal parathyroid hormone and vitamin D and intestinal calcium absorption to provide calcium and phosphorus for the growing fetus while maintaining maternal bone mass. As a result of plasma volume expansion, total albumin concentrations may be decreased as a result which will confound interpretation of total calcium concentration (which reflects calcium bound to albumin) however, when ionized calcium levels are measured, they are typically normal.

Considerations for Clinical Pathology Data Interpretation in Young Animals

The growth that occurs in young animals will impact clinical pathology data and it is important to consider the inclusion of an age-matched control group to help parse out test article–related effects. There is decreased functional capacity of many organs and animal species differ widely in the degree of organ development present at the time of birth and the rate of organ development. ¹³ Clinical pathology data for the various organ systems will change with age and these changes should not be confused with pathological states. One of the most striking changes in young animals is the change in RBC mass. During fetal development, RBCs increase in number; HCT increases during the second trimester to ~50% to 63% at term. Upon birth, RBC mass will decline in response to an increase in the availability of oxygen and downregulation of erythropoietin and also as a result of the shorter life span of neonatal RBCs, which leads to physiological anemia of the newborn. Red blood cell size decreases, and fetal hemoglobin is ultimately replaced by adult hemoglobin. The RBC mass in young animals slowly reaches adult values around the time of sexual maturity, which varies widely by species. White blood cell counts also change in young animals with no clear trend in all species. Newborns have increased neutrophil counts yet have decreased neutrophil chemotactic function and immature cellular defense mechanisms and humoral immunity. ¹⁴ Lymphocyte counts increase with growth as lymphoid organs mature. Growing rats show higher lymphocyte and neutrophil counts early in life in comparison to adult animals and then the proportion of neutrophils decreases until maturity is reached. ¹⁵ In rabbits, dogs, and cynomolgus monkeys, reports are conflicting or show no significant change. This highlights the benefit of comparing values to age-matched controls.

Coagulation parameters are impacted in young animals due to immaturity of liver function, and there is sparse information regarding coagulation findings in young animals, particularly rodents, likely due to the assay volume requirements relative to available animal blood volume which often precludes testing.

Many physiologic changes that occur from birth to adulthood will also impact clinical chemistry parameters because of the degree of organ maturation and growth, for example, bone, liver, and kidney. Liver function in dogs typically reaches maturity by 5 to 8 weeks of age and puppies and juvenile animals can have decreased albumin and glycogen stores with a tendency to hypoglycemia. ¹¹ Also, there is a potential impact on coagulation factor production and hepatic microsomal enzyme activity. Hepatobiliary enzyme activity is strongly influenced by age, especially as a result of colostrum absorption. Serum ALP activity can be high in newborns due to maternal contribution, and rapidly growing animals will have higher levels than those reaching skeletal maturity as the bone isoform contribution proportionally decreases in most species. Aspartate aminotransferase is linked to muscle mass development, and both ALT and AST activities increase from birth toward adult levels and γ-glutamyltransferase (GGT) activity decreases generally after weaning. ¹¹ Kidney development is complex and immature renal function at birth will manifest as decreased GFR and renal plasma flow, and neonates are often unable to concentrate or dilute urine. As the animal grows, serum urea nitrogen will increase with growth and creatinine concentrations will vary due to muscle mass, and therefore, until the animals reach maturity, the use of these parameters in combination as indicators of renal health is challenging. Demands for skeletal growth causes increases in calcium (total and unbound), phosphorus, magnesium, and bone ALP levels which normalize with skeletal maturity. ¹¹

Take-Home Points for Data Interpretation in Pregnant, Young, or Adult Animals

• Use the same “weight of evidence” approach to data interpretation irrespective of animal age. Be mindful of the many factors that can affect clinical pathology data including preanalytical variables (eg, stress, nutritional status, maturation stage), species, biological significance of each parameter in the context of reproductive/growth status of the animals (eg, RBC mass); the analytical method; study design; exposure data; in-life and pathology findings.

• Physiological changes can affect clinical pathology parameters during pregnancy and growth and should not be confused with pathological states (eg, anemia of newborn animals).

• Use of age-matched concurrent control group individual and group mean data is important to help parse out the impact of both the test-article administration and stage of maturation on clinical pathology parameters, especially on longer term studies where pretreatment versus posttreatment data and time point data can vary considerably as a result of the stage of maturation. Also, confirm whether expected growth-related changes are seen (eg, effects on calcium and phosphorus concentrations as a result of skeletal maturity).

• The time frame of pregnancy is an important consideration and the need to be mindful of potential rapid changes in certain parameters (eg, RBC mass, protein concentration) when comparing pretreatment versus posttreatment values.

• Individual variability will be more pronounced in data generated from juvenile animals versus adults.

• Recommend the limited and judicious use of statistics and that reference intervals be established at the clinical pathology testing laboratory for specific growth stages.

• Consider the known differences between animal species and humans for extrapolation of the significance of a given finding for a pediatric or juvenile population.