17.8 — Tuberculosis

Around a quarter of the world's population carries Mycobacterium tuberculosis. Around 5 to 10 percent of them will ever develop the disease.

That ratio is the whole of tuberculosis. The organism is necessary and nowhere near sufficient (Chapter 16.1), and what determines who becomes ill is the state of the immune system.

It killed an estimated one billion people between 1800 and 2000 — probably more than any other infectious disease in history — and it remains among the leading infectious causes of death worldwide.

And it is curable.

The organism

Mycobacterium tuberculosis has an unusual cell wall: extremely thick and waxy, rich in complex lipids.

And nearly everything about the disease follows from that wall.

It resists the Gram stain, which is why a different method — the acid-fast or Ziehl–Neelsen stain — is used, and why the organism is called an acid-fast bacillus.

It resists drying and survives in the environment for weeks.

It resists killing inside macrophages, which is exactly where it lives. The macrophage engulfs it and cannot destroy it, so the bacterium survives and multiplies within the cell that was supposed to remove it (Chapter 13.1).

And it grows extraordinarily slowly — dividing every 15 to 20 hours against 20 minutes for E. coli.

Which explains two things: culture takes 2 to 6 weeks, and treatment takes months rather than days, because antibiotics act on dividing organisms.

What happens after exposure

Airborne transmission, from someone with active pulmonary disease coughing.

Four possible outcomes, and the proportions matter.

The immune system clears it entirely — in a substantial proportion.

Latent infection — the commonest outcome. The organism is contained inside granulomas (Chapter 16.3) and remains alive but dormant. No symptoms, not infectious, positive immune test.

Primary progressive disease — proceeding directly to illness, more common in young children and the immunosuppressed.

Reactivation — years or decades later, when immunity falls.

And the lifetime risk of reactivation is around 5 to 10 percent — rising to around 10 percent per year in untreated HIV.

Which is why HIV transformed tuberculosis epidemiology. The two diseases amplify each other, and tuberculosis is the leading cause of death in people with HIV worldwide.

Other risk factors for reactivation: malnutrition, diabetes, smoking, alcohol, kidney failure, silicosis, and immunosuppressive drugs — particularly TNF inhibitors (Chapter 13.6), which is why screening for latent infection is mandatory before starting them.

The disease

Pulmonary tuberculosis — around 80 percent of cases.

Cough for more than 2 to 3 weeks, weight loss, night sweats, fever, fatigue, and later coughing blood.

And the classic combination of a chronic cough with weight loss and drenching night sweats is one of the more reliable symptom patterns in medicine (Chapter 16.4).

Its slowness is what makes it dangerous socially. A person with untreated pulmonary tuberculosis infects an estimated 10 to 15 people a year, because they remain ambulant and coughing for months.

Extrapulmonary tuberculosis — around 20 percent, and higher in HIV.

Lymph nodes — most commonly in the neck, painless and matted. Pleura — effusion. Spine (Pott's disease) — vertebral destruction, and it can cause spinal cord compression. Meninges — tuberculous meningitis, with high mortality and a subacute onset over weeks that distinguishes it from bacterial meningitis (Chapter 11.10). Abdomen, pericardium, kidneys, joints, and skin.

Miliary tuberculosis — widespread dissemination, named for the millet-seed appearance on a chest X-ray.

Diagnosis

And the tools have improved substantially, which is one of the encouraging parts.

Sputum microscopy — fast, cheap, and misses a substantial proportion of cases, particularly in HIV.

Molecular testing — and this is the significant advance. Automated cartridge-based tests detect tuberculosis DNA and rifampicin resistance simultaneously, in under two hours, in a machine that can be run by a non-specialist.

Which replaced a process that took weeks with one that takes an afternoon, and it has been deployed widely in high-burden settings.

Culture remains the reference standard and gives full drug sensitivities.

Chest X-ray — classically upper lobe changes, cavities, and scarring. In HIV the appearance is frequently atypical, which is a common source of missed diagnosis.

Testing for latent infection — the tuberculin skin test, and interferon-gamma release assays, which are blood tests unaffected by prior BCG vaccination.

And neither distinguishes latent from active disease — they detect an immune response, not the presence of disease. Active disease must be excluded clinically and radiologically before treating someone as latent.

Treatment

And it works. Drug-sensitive tuberculosis is cured in around 85 percent of cases with standard treatment.

The standard regimen: four drugs for 2 months, then two drugs for 4 months — six months in total.

Rifampicin, isoniazid, pyrazinamide and ethambutol.

Why four drugs, and why so long:

Resistance emerges readily to any single agent, because the bacterial population is large enough to contain pre-existing resistant mutants (Chapter 3.2). Multiple drugs mean a single organism would need several simultaneous mutations.

And the organism exists in several metabolic states — actively dividing, slowly dividing, and dormant — and different drugs act on different populations. The dormant ones are why treatment cannot be short.

Side effects worth knowing:

Rifampicin turns urine, tears and sweat orange. Harmless, alarming if unexpected, and a useful adherence check. And it powerfully induces liver enzymes (Chapter 1.5), which reduces the effectiveness of many drugs including hormonal contraception — a genuinely important interaction.

Isoniazid causes peripheral neuropathy by interfering with vitamin B6, which is why pyridoxine is given alongside.

Pyrazinamide raises uric acid and can precipitate gout.

Ethambutol causes optic neuritis — loss of colour vision, particularly red-green, is the early sign, and patients are told to report visual change immediately.

All four can cause hepatitis.

Adherence is the central problem, because six months of four drugs in someone who feels better after three weeks is a demanding ask.

Directly observed therapy — watching the person take each dose — improves completion, and video-observed therapy has produced equivalent results with far less burden.

And there is genuine progress on duration. Four-month regimens using newer combinations have been shown non-inferior in trials, which would substantially improve completion rates.

Drug-resistant tuberculosis

Multidrug-resistant (MDR-TB) — resistant to at least rifampicin and isoniazid. Around 400,000 cases a year, and only a minority are diagnosed and treated.

Extensively drug-resistant (XDR-TB) — resistant to further key drugs.

And treatment of MDR-TB used to mean 18 to 24 months of injections and toxic drugs, with cure rates around 50 to 60 percent and side effects including permanent deafness.

That has changed substantially, and it is one of the better recent stories in infectious disease.

New oral regimens using bedaquiline, pretomanid and linezolid — the BPaL regimen and its variants — achieve cure rates around 90 percent in 6 months, entirely by mouth.

Six months of tablets instead of two years of injections, with better outcomes. WHO guidance now recommends these regimens, and rollout is the limiting factor rather than the science.

Prevention

Treating active cases — the most effective control measure, because it removes the source.

Treating latent infection — reduces reactivation risk by 60 to 90 percent. Offered to those at highest risk: HIV, recent contacts, and people starting immunosuppressive treatment.

Shorter regimens — three months of weekly rifapentine and isoniazid, or three to four months of rifampicin — have largely replaced the older nine-month isoniazid course, and completion rates are much better.

BCG vaccination — and its performance needs stating honestly.

It protects children well against the severe forms — miliary tuberculosis and tuberculous meningitis, with efficacy around 70 to 80 percent.

Its protection against pulmonary tuberculosis in adults is inconsistent, ranging from around 80 percent in some trials to essentially zero in others.

The variation appears to relate to prior exposure to environmental mycobacteria, which is commoner nearer the equator — which would explain why the trials in temperate regions performed best.

So BCG is given universally in high-burden countries and selectively elsewhere, and new tuberculosis vaccines are in late-stage trials, with one candidate showing around 50 percent efficacy against progression from latent infection to disease. That would be a substantial advance if confirmed.

Infection control — ventilation, and airborne precautions for infectious cases.

And addressing the social determinants, because tuberculosis tracks poverty, crowding, malnutrition and HIV more closely than it tracks anything medical. Its decline in Western Europe began well before any effective drug existed, driven by better housing, nutrition and working conditions.

Where it stands

Incidence is falling by around 1 to 2 percent a year — too slow to meet elimination targets.

COVID-19 caused the first rise in tuberculosis deaths in over a decade, through disrupted diagnosis and treatment, and recovery has been partial.

And the encouraging half is real: rapid molecular diagnosis, shorter treatment for both drug-sensitive and drug-resistant disease, better latent treatment, and new vaccines in trials.

A disease that killed a billion people is now curable in six months with tablets. The gap is delivery, not knowledge.

What the next page fixes

Chapter 17.9 covers the infection that transformed tuberculosis epidemiology, and which has itself gone from a certain death sentence to a manageable condition within a single generation.