4.1 In-Life Clinical Signs, FOB, Body Weight, Food/Water, Ophthalmology

Why in-life endpoints characterize in vivo toxicity

Handbook task I.B.2 asks you to characterize toxicological effects in vivo: name typical laboratory models and the endpoints that describe what happened while the animal was alive. Independent OpenExamPrep teaching for this section treats the in-life phase as a data-generating system, not as a holding period until histopathology. If cageside notes, body weights, and food records are thin, the pathology report cannot reconstruct onset, severity, reversibility during recovery, or whether a finding was a humane-endpoint event versus a terminal annotation.

Typical mammalian general-toxicity models remain the rat (often Sprague Dawley or Wistar) for oral repeated-dose work such as Organisation for Economic Co-operation and Development (OECD) Test Guideline 407 (28-day oral) and OECD 408 (90-day oral), a mouse strain when the scientific or carcinogenicity question prefers it, and a nonrodent (Beagle dog, minipig, or nonhuman primate) when a pharmaceutical package under International Council for Harmonisation (ICH) M3(R2) needs a second species. OECD 409 covers 90-day oral nonrodent designs. OECD 413 (90-day inhalation) and dermal repeated-dose guidelines add portal-of-entry in-life signs to the same systemic list. Safety pharmacology ICH S7A adds a timed core battery (central nervous system, cardiovascular, respiratory) on top of general-toxicity cageside work. None of those models is useful if the only recorded in-life fact is that animals were found dead on Monday.

The in-life vocabulary you must keep separate:

• Morbidity and mortality (welfare census, often twice daily)
• Cage-side clinical observations (usually at least once daily)
• Detailed physical examinations (hands-on, typically pretest and weekly)
• Irwin screen or functional observational battery (FOB) (structured neurobehavioral stages, timed to pharmacokinetics)
• Body weight and body-weight gain
• Food and water consumption
• Ophthalmology (pretest and termination)
• Targeted physiology: body temperature, respiration, local irritation at administration sites

Morbidity and mortality checks

Mortality is a binary in-life endpoint: the animal is found dead or is euthanized. Morbidity is the clinical state that precedes that outcome—pain, distress, inability to reach food or water, or a constellation of signs that predicts imminent death.

In Good Laboratory Practice (GLP) toxicology facilities, morbidity and mortality checks are typically performed at least twice daily, including weekends and holidays, with enough hours between checks that a moribund animal is not left overnight without observation. An early-morning census and a late-afternoon or evening census are the usual pattern. A once-daily walk-through while technicians are dosing is not a substitute. Found-dead animals need prompt necropsy (plus notes on autolysis and refrigeration) because delayed discovery destroys histology and can hide the last clinical state.

Twice-daily checks answer a welfare question and a scientific one. Time-to-death and the last recorded clinical state support dose-response and help separate acute overdose from cumulative toxicity. Record found dead versus euthanized moribund separately. Combining them as died hides whether humane endpoints were used and whether the high dose remained interpretable.

Satellite toxicokinetic (TK) bleeds, 24-hour urine collection, and open-field sessions all change apparent activity. The mortality check is still a welfare census, not a substitute FOB.

Cage-side clinical observations versus detailed physical examinations

Cage-side (clinical) observations are usually performed at least once daily, often at a consistent clock time and, for bolus dosing, at a time that considers time of maximum plasma concentration (Tmax). The observer scores posture, gait, activity, respiration, skin and fur, eyes, mucous membranes, excretions, and behavior, often without removing every animal unless something is abnormal. Findings use a standardized glossary (hunched, piloerection, ataxia, clonic convulsion, stereotypy) plus severity and time after dose. Inter-observer training matters: two technicians using different words for the same gait will manufacture a false dose-response.

A detailed physical examination is hands-on. The animal is removed and examined systematically: body-condition score, palpation for masses, examination of eyes, nose, oral cavity, teeth (important in dietary studies), skin, and administration sites, plus a coarse neurologic screen that is still less complete than a full FOB. In OECD 407/408, a detailed clinical examination is typically done once pretest and weekly thereafter. Nonrodent studies usually have richer veterinary examinations because each dog or monkey is an individual statistical unit.

Do not treat a weekly physical as if it captured peak-drug neurobehavioral effects. A dog examined at 08:00 before dosing can look normal and still have stereotypy at a 1-hour Tmax. Conversely, a cageside note of slightly reduced activity at Tmax does not replace a weekly palpation that finds a mammary mass.

Dark-phase observations are sometimes specified when the test article is a stimulant or sedative whose peak falls in the animals' active period. Whatever the clock, the protocol must state when relative to dosing the observation occurs.

Irwin screen, FOB, and neurobehavioral batteries

The Irwin screen (a structured rodent neurobehavioral examination used heavily in safety pharmacology) and the FOB used in general toxicity and in OECD 424 neurotoxicity studies share a staged logic:

1. Home-cage: posture, palpebral closure, convulsions, wrapping in bedding, ease of removal, vocalization
2. Handheld: fur condition, lacrimation, salivation, piloerection, pupillary response, palpebral reflex, muscle tone, body temperature, and often landing foot splay
3. Open-field: gait, mobility, rearing, arousal, stereotypy, grooming, fecal boli and urine pools, and sometimes quantitative motor activity in a separate apparatus
4. Stimulus-response: approach, touch, click or snap, tail pinch, righting reflex, grip strength, and sometimes a pain response

ICH S7A expects a central-nervous-system evaluation that can be an Irwin or FOB, timed to Tmax (and often an additional later point such as 24 hours) after dosing, plus a pretest baseline. In a 90-day general-toxicity study, OECD 408 expects sensory reactivity to stimuli, grip strength, and motor activity toward the end of the treatment period. That end-of-study battery does not replace daily cageside notes and does not by itself satisfy an ICH S7A Tmax CNS study if the pharmaceutical program still needs a dedicated safety-pharmacology session.

Timing relative to Tmax is the classic characterization error. If oral Tmax is 0.5–2 hours, an FOB done only 24 hours later may miss sedation, tremor, or increased locomotion. For dietary studies without a bolus Tmax, define a clock time, keep it consistent, and still collect pretest baselines. Repeat sessions should consider handling-stress carry-over. Observers should be unaware of dose group when feasible.

FOB decreases in activity can reflect CNS depression, malaise from systemic illness, or hypothermia. FOB increases can reflect stimulant pharmacology or hyperthermia. Interpret the battery with temperature, body weight, and clinical pathology—not as a stand-alone neurotoxicity yes/no.

Body weight as a sensitive general-toxicity endpoint and MTD driver

Body weight is often the most sensitive routine in-life marker of general toxicity. Animals are typically weighed at least weekly (more often in the first week or in acute work). Plot absolute weight and body-weight gain. A group that gains 10% less than concurrent controls over 90 days is not the same as a group that lost 10% from its own pretest baseline. Growing rats can look “stable” while treated animals fail to gain; adult dogs can lose absolute mass. Fasted versus fed weights must not be mixed across groups.

Maximum tolerated dose (MTD) selection in range-finding and in chronic or carcinogenicity design commonly uses body-weight change as a quantitative brake. Many programs treat about a 10% decrement in body-weight gain relative to concurrent controls as evidence that the high dose is at or above an MTD for a subchronic or chronic study, without deaths, severe suffering, or cachexia that wreck interpretability. That is an in-life characterization decision you will already have seen in dose-selection teaching; here the operational point is that weekly weights are how you see the MTD arriving in real time. Larger decrements, especially with clinical signs, push the dose into overt toxicity that confounds liver enzymes, organ weights, and tumor tables through starvation.

Body weight also drives mg/kg gavage calculations. If you do not update weights, rapidly growing juveniles are overdosed. Dietary parts per million (ppm) do not automatically hold mg/kg constant when food intake and body weight diverge.

Food and water consumption

Food consumption is recorded by cage (rodents are often pair- or group-housed) or individually. Express it as grams per animal per day and often as grams per kilogram body weight per day. Water consumption is required when the test article is in drinking water and whenever polyuria/polydipsia is a plausible renal or endocrine effect.

Three mechanistic stories look similar on a food graph:

Pair-feeding (offering controls only what treated animals ate) can separate calorie restriction from direct organ toxicity. It is a design amendment, not a cageside afterthought. If animals eat less of a dietary mix, achieved mg/kg dose falls. Reporting protocol ppm as if it were consumed is a characterization error.

Drinking-water formulations add a second palatability problem: animals may refuse water, producing dehydration that looks like renal toxicity. Measure both intake and clinical hydration (skin tent, mucous membranes, body weight).

Ophthalmology

Ophthalmology in general toxicity typically includes indirect ophthalmoscopy (fundus: retina, optic disc, choroid, vitreous) and slit-lamp biomicroscopy (adnexa, cornea, anterior chamber, iris, lens). Fluorescein staining is added when corneal injury is in scope (irritants, freeze-dried biologics, solvents).

Perform examinations pretest to exclude animals with spontaneous lesions (corneal dystrophy in some rat strains, embryonic remnants, prior trauma) so a terminal cataract is not blamed on a preexisting finding. Repeat near termination (and sometimes at an interim) on at least high-dose and control animals; many protocols examine all groups if high-dose findings appear. Mydriasis must be standardized; some drugs themselves alter pupil size and confound the examination.

Ocular findings can be primary (test article in the eye, phototoxicity, sugar-alcohol cataract) or secondary (dehydration, infectious sialodacryoadenitis in rats, anesthesia artifacts). Skipping pretest and then discovering strain-background corneal crystals at necropsy is an in-life design failure, not a pathology surprise.

Body temperature, respiration, and local irritation

Body temperature is a core Irwin/FOB measurement and a clinical emergency marker. Hypothermia accompanies severe systemic toxicity, sedation, and moribundity. Hyperthermia can reflect uncoupling of oxidative phosphorylation, seizure activity, or pyrogenicity. A temperature taken only at a weekly physical will miss Tmax-related hypothermia.

Respiration is watched cageside (rate, effort, noises, cyanosis) and, under ICH S7A, with quantitative respiratory rate, tidal volume, and minute volume in a dedicated safety-pharmacology study. OECD 413 inhalation studies add clinical signs of respiratory distress as a portal-of-entry endpoint, not only as a systemic one. Gavage trauma can mimic inhalation injury: bloody oral discharge and dyspnea after a difficult dose are administration-site events until proven otherwise.

Local irritation at the administration site is an in-life endpoint: erythema and edema after dermal dosing, injection-site swelling or necrosis, nasal rubbing after inhalation, and esophageal trauma from gavage. These findings often explain later site histopathology and can justify a humane endpoint (severe necrosis) even when systemic organs are normal. Do not bury injection-site scores only in the pathology narrative if the animals were already in pain in week 2.

Humane endpoints versus findings that merely annotate pathology

Humane endpoints are pre-specified clinical states that trigger euthanasia (or a dose holiday, veterinary treatment, or removal) to prevent unnecessary pain or distress. Typical triggers include moribund condition, inability to eat or drink, status convulsion, severe dyspnea, self-mutilation, and rapid body-weight loss beyond an Institutional Animal Care and Use Committee (IACUC) limit (often on the order of 20% from baseline in adult animals—use the approved protocol number, not folklore). Once triggered, the animal is killed and necropsied; the finding is why the animal left the study.

Annotative clinical signs (mild piloerection, slightly reduced activity at Tmax, a small scab, transient salivation after gavage) are recorded, graded, and later correlated with organ weights and histopathology. They do not by themselves stop the study. Collapsing both categories into clinical signs were noted hides whether the high dose was uninterpretable because half the group was euthanized in week 2.

The Study Director, with the attending veterinarian and the IACUC protocol, owns the humane-endpoint decision. Technicians do not wait for histopathology to decide whether a moribund animal should be killed. Early deaths are characterization: they tell you the dose was not tolerated, and they change statistical n for every later endpoint.

Common in-life findings and likely target organs

Scenario: declining food intake in a dietary 90-day study

A contract laboratory is running an OECD 408 dietary study. High-dose food consumption drops 40% during week 1, body-weight gain flattens, cageside exams are otherwise almost normal, and water intake is unchanged. Two camps form: the diet tastes bad versus we have hit the MTD.

Test palatability as a competing hypothesis, not as a wish. Immediate week-1 rejection concentrated at the highest ppm, without tremor, jaundice, or FOB deficits, is classic palatability. Confirm homogeneity and concentration of the mix (a bitter hot spot in the hopper is still a palatability or formulation problem). Check whether achieved mg/kg dose collapsed because grams eaten collapsed. Consider a short gavage bridge or a flavored or diluted diet if the scientific question is systemic toxicity, not whether rats will eat this chow.

Toxicity-driven anorexia usually accompanies other in-life signals: progressive (not only day-2) food drop, deteriorating clinical scores, hypothermia, and later clinical chemistry or pathology. If food drop persists after animals have had time to accept the diet, or if a gavage arm at the same mg/kg still suppresses intake, you are looking at a systemic effect and possibly an MTD conversation in the sense taught with body-weight gain: some toxicity, not mass starvation.

Humane-endpoint question: if high-dose animals continue to lose weight toward the protocol euthanasia limit, you euthanize individuals. You do not wait for week-13 histopathology to confirm starvation. Those early deaths are the characterization: the dietary high dose was not tolerated, whether the first mechanism was taste or toxicity. The report must say which evidence supported which mechanism, including diet analysis, achieved dose, clinical scores, and whether ophthalmology or FOB ever became abnormal.

High-yield traps

• Treating once-daily technician presence as twice-daily morbidity and mortality checks.
• Running FOB only at 24 hours after a Tmax of 1 hour.
• Calling a 10% lower body-weight gain in growing rats a 10% body-weight loss from baseline.
• Ignoring achieved dose when dietary food intake falls.
• Skipping pretest ophthalmology.
• Recording moribund euthanasia as routine clinical signs with no humane-endpoint flag.
• Assuming unchanged water intake rules out toxicity when the article is in diet, not in water.