13.4 — Inflammation and Fever

Aulus Cornelius Celsus, writing in Rome around the year 30, described inflammation in four words: rubor, calor, tumor, dolor — redness, heat, swelling, pain. Two thousand years later the description has not needed correcting, only extending: Galen added functio laesa, loss of function.

Every one of those five signs is a direct consequence of a specific mechanical change, and understanding which is which explains both why inflammation is useful and when it becomes the disease.

The five signs, explained

Redness and heat — blood vessels dilate, so more warm blood flows through the area. Deliberate: more blood means more cells and more plasma proteins delivered.

Swelling — the vessels also become leaky. Fluid and protein escape into the tissue (Chapter 7.5). This delivers antibodies, complement and clotting factors to the site, and it dilutes toxins.

Pain — from pressure of the swelling and, more importantly, from chemical mediators — particularly prostaglandins and bradykinin — that lower the firing threshold of pain nerve endings. The area becomes tender to touch that would normally be painless.

And the pain is functional. It enforces rest and protects the area, which is why congenital insensitivity to pain destroys joints (Chapter 11.13).

Loss of function — partly from pain and swelling, partly deliberate immobilisation.

Note that four of the five follow from just two vascular changes: dilation and increased permeability.

The sequence

Minutes — mast cells release histamine, causing immediate vasodilation and leakiness. The wheal and flare of a nettle sting or a mosquito bite is this alone.

Complement fragments and clotting products add to it.

Hours — neutrophils arrive, and the process by which they do is worth following because it is exploited by several drugs.

Neutrophils in flowing blood must stop, cross the vessel wall and reach the tissue. Four steps.

Rolling — the endothelium near an inflamed site displays adhesion molecules that make neutrophils tumble slowly along the wall rather than rushing past.

Activation — chemical signals from the tissue activate them.

Firm adhesion — stronger adhesion molecules lock them in place.

Migration — they squeeze between endothelial cells and follow the chemical gradient to the source.

Deficiency of the adhesion molecules causes leukocyte adhesion deficiency: the neutrophil count in blood is high because they cannot leave the circulation, and the person has severe recurrent infections with no pus at all — an infection with no visible inflammation, which is a striking demonstration of what the fourth step is for.

Days — macrophages become dominant. They clear debris and dying neutrophils, and switch the process from destruction to repair.

And the switch is active, not passive. Macrophages change phenotype from an inflammatory type to a repair type, releasing growth factors that drive new blood vessels and new collagen. Failure of that switch is what produces chronic inflammation.

Chronic inflammation

When the cause is not removed, the process persists — for months or years.

The cell population changes: macrophages, lymphocytes and plasma cells rather than neutrophils.

And tissue destruction and repair happen simultaneously, which is why chronic inflammation causes fibrosis — functional tissue progressively replaced by scar.

A granuloma is the specific structure formed when macrophages cannot destroy something: they wall it off, fusing into giant cells and surrounded by lymphocytes. Tuberculosis is the classic example (Chapter 17.8), along with sarcoidosis and Crohn's disease.

And chronic low-grade inflammation is now recognised as a contributor to a striking list of conditions — atherosclerosis, type 2 diabetes, some cancers, and neurodegeneration. The obesity link runs through fat as an endocrine organ releasing inflammatory signals (Chapter 12.7).

Where the evidence is strong, the drugs work. Trials of anti-inflammatory drugs in cardiovascular disease have shown reductions in events, which supports the causal role rather than merely the association. Where the evidence is weaker is in the enormous popular literature about "inflammation" as a general explanation for unwellness, most of which runs far ahead of what is established.

Fever

Fever is not the infection overwhelming your temperature control. It is your temperature control doing exactly what it was told (Chapter 4.7).

The mechanism:

Pyrogens — from microbes, or cytokines released by your own immune cells — reach the hypothalamus.

They trigger prostaglandin production, which raises the temperature set point.

The body then behaves as though it is too cold, because relative to the new set point it is. Vasoconstriction, shivering, and seeking warmth — all heat-generating and heat-conserving.

Which explains the experience precisely. At the start of a fever you feel cold and shiver, while your temperature is objectively rising. When the fever breaks, the set point drops, you are now above it, and you sweat profusely.

The shivering and sweating are not the illness. They are the transitions between set points.

Is fever useful?

The evidence says yes, within limits, and this changes what you should do about it.

Fever inhibits many pathogens. Bacterial and viral replication rates fall at higher temperatures, and several organisms have narrow optimal ranges.

Immune function improves. Neutrophil migration, lymphocyte proliferation and interferon activity all increase with modest temperature elevation.

Iron availability falls, and bacteria need iron.

And it is evolutionarily conserved. Even cold-blooded animals seek warmth when infected — a lizard with an infection moves to a warmer rock — and preventing them from doing so measurably increases mortality. A costly behaviour conserved across vertebrates is unlikely to be useless.

Studies in humans are less clear-cut, and some have found that routine antipyretics prolong illness modestly. The honest position: fever is probably beneficial at moderate levels, and the effect is not large enough to justify enduring severe discomfort.

So the practical guidance:

Treat fever for comfort, not for the number. A child who is drinking, alert and playing does not need paracetamol because the thermometer reads 39 °C. A child who is miserable does, and it will help them drink and sleep.

The height of a fever correlates poorly with the seriousness of the illness. A well-looking child with 39.5 °C is usually less concerning than a listless child with 38 °C. How the person looks and behaves is far more informative than the reading.

Above about 41 to 42 °C, treat. Protein denaturation becomes a real threat (Chapter 1.3).

And there is one situation where fever must be treated as an emergency regardless of how the person looks: fever in a person with neutropenia (Chapter 7.1), or with no spleen (Chapter 7.8). Antibiotics within the hour, no exceptions.

Hyperthermia is not fever

A crucial distinction that changes the treatment entirely.

In fever, the set point is raised and the body is achieving it.

In hyperthermia, the set point is normal and the body cannot hold it — heat stroke, malignant hyperthermia (Chapter 6.1), serotonin syndrome, thyroid storm.

So antipyretics do not work in hyperthermia. Paracetamol lowers the set point, and the set point is not the problem. Physical cooling is the treatment, and it must be aggressive (Chapter 23.10).

Getting this wrong — giving paracetamol to someone with heat stroke and waiting — costs lives.

Febrile convulsions occur in about 2 to 5 percent of children between 6 months and 5 years. They are frightening and are usually harmless, and the important reassurance is specific: simple febrile convulsions do not cause brain damage and do not mean the child has epilepsy — the risk of later epilepsy is only slightly above background.

They are related to the rate of temperature rise more than the absolute level, which is why they often occur at the very start of an illness, and why giving antipyretics does not reliably prevent them.

Anti-inflammatory drugs

NSAIDs — aspirin, ibuprofen, naproxen, diclofenac. They block cyclooxygenase, the enzyme making prostaglandins (Chapter 1.3).

So they reduce pain, fever and inflammation, all through one mechanism.

And their side effects come from the same mechanism in other places. Prostaglandins protect the stomach lining (Chapter 9.2), maintain kidney blood flow (Chapter 10.6), and are involved in platelet function. Blocking them therefore causes ulcers, kidney injury and bleeding risk.

There are two forms of the enzyme. COX-1 is present constantly and does the housekeeping — stomach protection, platelet function. COX-2 is induced at sites of inflammation.

Selective COX-2 inhibitors were designed to get the anti-inflammatory effect without the stomach damage, and they largely succeeded at that — and turned out to increase cardiovascular events, because COX-2 also produces a protective substance in blood vessels. Rofecoxib was withdrawn in 2004 after this became clear.

The episode is worth remembering as a general lesson: a drug designed rationally from a mechanism can still have consequences the mechanism did not predict, because the same enzyme does different things in different tissues.

Paracetamol (acetaminophen) reduces pain and fever and has almost no anti-inflammatory effect, which is why it does not help inflammatory conditions much. Its mechanism is still not fully established after more than a century of use, which is worth admitting.

It is safe at recommended doses and dangerous above them (Chapter 23.6), and the margin is narrower than people assume.

Steroids suppress inflammation broadly and powerfully (Chapter 12.4).

And there is a genuine question about when suppressing inflammation is unhelpful. Anti-inflammatory drugs after soft tissue injury may delay healing, since inflammation is the first stage of repair. They are now used for pain rather than prescribed routinely for injury (Chapter 5.8).

Sepsis

The extreme failure of this system, and the most important condition in this chapter.

Sepsis is life-threatening organ dysfunction caused by a dysregulated host response to infection.

Read that definition carefully: the damage is done by the response, not by the organism.

What happens: an infection triggers a massive systemic release of inflammatory mediators. The changes that are useful locally become catastrophic when they occur everywhere at once.

Vasodilation everywhere — blood pressure collapses. Increased permeability everywhere — fluid leaves the circulation into the tissues, so the person is simultaneously fluid-overloaded and intravascularly depleted. Widespread clotting activation — microclots block small vessels, consuming platelets and clotting factors, so the person clots and bleeds at once (Chapter 7.1). Cells cannot use oxygen properly even when it is delivered. And organs fail — kidneys, lungs, liver, brain, heart.

Sepsis causes around 11 million deaths a year worldwide, roughly a fifth of all deaths.

And mortality rises by several percent for every hour of delay in treatment, which is why recognition is everything.

The signs — and it is worth learning these, because sepsis presents to families before it presents to hospitals:

Slurred speech or confusion. Extreme shivering or muscle pain. Passing no urine in a day. Severe breathlessness. A feeling that something is seriously wrong — patients and families often describe a sense of impending doom. And skin that is mottled, blue or unusually pale.

In children, add: fast breathing, a rash that does not fade under pressure, a fit, unusual coldness, and being abnormally sleepy or floppy.

The temperature can be high or low, and a low temperature in sepsis is more ominous than a high one.

The immediate treatment is a defined bundle: blood cultures, antibiotics, intravenous fluids, lactate measurement, oxygen and urine output monitoring — all within the first hour.

Antibiotics within the first hour save lives, and this has been demonstrated repeatedly.

And the single most useful thing a member of the public can do is to say the word. Campaigns encouraging patients and families to ask "could this be sepsis?" have measurably improved recognition, because it prompts a specific set of actions.

Wound healing

The end of inflammation and the beginning of repair, in four overlapping phases.

Haemostasis (minutes) — the clot (Chapter 7.1).

Inflammation (days 1 to 3) — clearing debris and organisms.

Proliferation (days 3 to 21) — new blood vessels grow in, fibroblasts lay down collagen, and epithelium migrates across the surface. Granulation tissue — red and bumpy — is new vessels and matrix, and it is a healthy sign.

Remodelling (weeks to 2 years) — collagen is reorganised along lines of stress and the scar gradually strengthens and fades.

And a scar never regains full strength. About 80 percent of original tensile strength at best, and it reaches only 20 percent at three weeks — which is exactly why wounds are supported for longer than they appear to need it.

What impairs healing: poor blood supply, infection, diabetes, smoking, steroids, malnutrition — particularly protein, vitamin C for collagen (Chapter 4.2), and zinc — old age, and repeated movement of the wound.

And the encouraging counterpart is that most of that list is modifiable. Stopping smoking before elective surgery measurably reduces wound complications, and correcting nutrition before an operation is one of the more effective and least glamorous interventions available.

Abnormal scars: hypertrophic scars stay within the original wound boundaries and often improve with time. Keloid scars grow beyond them, do not regress, and are commoner in darker skin and on the chest, shoulders and earlobes.

What the next page fixes

The whole point of the adaptive system is that the second encounter goes better than the first. Chapter 13.5 covers immunological memory and the technology built on it — how vaccines work, which kinds exist, what they have achieved, and how to think clearly about the arguments made against them.