How TB Diagnostics Have Evolved Since the Second Century

March 22, 2021

Tuberculosis has been with humans since before recorded history and identifying patients with the disease has been a challenge for just as long. Molecular analysis of the causative organism, Mycobacterium tuberculosis, suggests that the first infections of humans occurred as much as 70,000 years ago when hunter-gatherers were still migrating to the European subcontinent. Population density was low, and tuberculosis was likely a sporadic affliction due to limited person-to-person interaction. With the advent of agriculture, and ultimately, cities, close contact between people created the conditions needed for tuberculosis to be widely transmitted. In this way, tuberculosis became an endemic disease throughout the ancient world.

Ancient Diagnostic Methods

In ancient times, the only way to diagnose the disease was through physical examination. Early in the second century CE, Galen defined classic tuberculosis symptoms that we know today: hemoptysis (coughing up blood), night sweats and fever. Even then, however, it was recognized that diagnosis based on symptoms alone was insufficient. Tuberculosis was certainly not the only chronic disease of ancient people and shares symptoms with viral or bacterial pneumonias, cancer, asthma and allergies.

Soranus of Ephesus attempted to improve on the diagnosis of tuberculosis and described a rudimentary method based on sputum, which was burned on hot coals. Presence of tuberculosis was inferred by a characteristic smell unsurprisingly described as “foul” and suggestive of necrotic flesh. Although burning sputum may initially seem more akin to witchcraft than medicine, this method could be described as prescient in light of modern attempts to use “electronic noses” to diagnose tuberculosis using volatile organic compounds elicited from patients’ sputum or breath.

Animals as Culture Vessels

Today, inhaling fumes from burning sputum would be unconscionable due to the risk of disease transmission. However, the infectious nature of tuberculosis remained a topic of controversy for over 1,000 years. The idea that tuberculosis could be transmitted from person to person was first proposed by Aristotle in the fourth century BCE and again in the 16th century by Girolamo Fracastoro. Unfortunately, neither provided experimental evidence. It was not until 1865 that Jean-Antoine Villemin demonstrated that infectious material from human patients could cause tuberculosis in rabbits.

Nevertheless, rabbits are not easily infected with tuberculosis and therefore not the ideal experimental subjects. Hermann von Tappeiner improved on Villemin’s experiments using guinea pigs, which are inherently more susceptible to the disease. Tappeiner’s work paved the way for these animals to be used as diagnostic tools, despite lack of knowledge of the causative agent of the disease. Patient samples were introduced into guinea pigs and characteristic pathology was observed if tuberculosis was transmitted. Although complex and unwieldy, this methodology persisted until at least the mid-1960s.

Identifying and Visualizing the Cause of Tuberculosis

The next breakthrough in tuberculosis diagnosis came in 1882 when Robert Koch presented his discovery of M. tuberculosis to the Berlin Physiological Society. That date, March 24, is now commemorated annually as World Tuberculosis Day. Soon after, Koch successfully grew the organism in pure culture and in 1905, he was awarded a Nobel Prize. By the 1930s, E. Löwenstein and K. Jensen (their first names appear to have been lost to history) developed selective media containing malachite green, now known as Löwenstein-Jensen medium, which is still in common use for propagation of mycobacteria. 

On the left, Mycobacterium tuberculosis stained using the Kinyoun method. On the right, growth of  M. tuberculosis on solid media.
On the left, Mycobacterium tuberculosis stained using the Kinyoun method. This image shows “cording” where hundreds of individual bacilli cluster together to form filamentous structures. On the right, growth of M. tuberculosis on solid media.
Source: K.P. Smith (left), CDC's Public Health Image Library (right)

In his early studies of M. tuberculosis, Koch noted that this organism was much more difficult to stain than other bacteria that he had previously discovered, including Bacillus anthracis. We now know that M. tuberculosis resists staining because of its exceptionally resilient cell wall structure containing mycolic acids, long chain hydrocarbons akin to paraffin wax.

This observation led to development of a special staining method by Franz Ziehl, which was later modified by Friedrich Neelsen. This method, known as the Ziehl-Neelsen stain, uses heat to allow stain to penetrate the waxy surface of mycobacteria. Use of a strong acid as a decolorizer removes the stain from all organisms without such a waxy coat and gives the method its common name: “acid-fast staining.” The Ziehl-Neelsen method was further improved on by Joseph Kinyoun, who omitted the heating step, therefore limiting the infectious risk inherent with aerosolization of M. tuberculosis.

Immunologic Diagnoses

In vitro culture of M. tuberculosis also opened new opportunities for diagnostic tests based on extracts from the organism itself, known as tuberculin. Koch discovered that animals injected with tuberculin were protected against infection. Contemporaneously referred to as a potential “therapeutic vaccine,” Koch hypothesized that tuberculin could be used to treat and potentially protect humans from tuberculosis. Unfortunately, this hypothesis was incorrect and tuberculin treatments often made patients even sicker. Among the first to recognize the problems with tuberculin treatment was none other than Arthur Conan Doyle, a physician who is also widely known for his creation of the Sherlock Holmes character. In reviewing Koch’s clinical data for a news story, Conan Doyle concluded that tuberculin was a promising diagnostic tool, but not a treatment.

Tuberculin did go on to serve as an important diagnostic tool.  Frederick Roland George Heaf discovered that coating needles in tuberculin results in papule formation in patients previously exposed to tuberculosis. This tuberculin skin test (TST) was further standardized in 1907 by Charles Mantoux, who used intradermal injection of a defined volume of tuberculin. Use of TSTs in apparently healthy individuals also uncovered a population of patients in which the test was positive, but who showed no overt signs of disease. These findings led to the discovery that tuberculosis may lie dormant in the lungs, so called latent disease. Over 100 years later, TSTs remain in use around the world and are recommended by the U.S. Centers of Disease Control and Prevention (CDC).

Chest X-rays for Diagnosis

Chest x-ray showing a cavitary lesion of the lung (red arrow).
Chest x-ray showing a cavitary lesion of the lung (red arrow).
For most of history, clinicians were mainly restricted to observing outward signs and symptoms of disease in their patients. Although specimens such as sputum or tissue could be obtained, there was no way to directly look inside the body for potential damage. This changed in 1895 with the discovery of Röntgen radiation, now known as X-rays, which can pass through a human body, allowing some internal structures to be captured on film. Almost immediately after their discovery, X-rays were used to assist in the diagnosis of tuberculosis and are still used to visualize the cavitary lesions characteristic of active tuberculosis disease.

Modern Diagnostics

Today, the first step in the diagnostic process for latent disease is often the TST, much like it would have been nearly 120 years ago. However, this test may be omitted in lieu of a more specific interferon gamma release assay, which directly assesses T-cell activation when exposed to antigens specific to M. tuberculosis. Following a positive immunologic test, a patient will likely undergo some form of imaging, often still a chest x-ray, to discover undiagnosed active disease.

When active tuberculosis is suspected, a combination of imaging, patient history (travel history, immune status) and detection of the organism itself are used to establish a diagnosis. Methods and media for culture of M. tuberculosis have not changed substantially in nearly 100 years, although some advances, notably automated, liquid-based methods that detect growth without human intervention, have reduced labor requirements while increasing culture sensitivity.

Similarly, acid-fast staining techniques also remain in use, essentially unchanged from their inception. The stain used in newer methods is a fluorescent dye that lights up when exposed to specific wavelengths of light. This makes it much easier to screen specimens for the presence of mycobacteria because the organisms of interest appear as light against a dark background.

Advances in molecular diagnostics have given us a third way to detect M. tuberculosis through specific amplification of nucleic acids from the organism. This reduces time to detection to hours, and even allows for identification of some antimicrobial resistance mechanisms (notably rifampin resistance), which otherwise would take weeks to identify phenotypically. Going forward, it is almost certain that diagnosis of tuberculosis will continue to rely on a mixture of old and new technologies.

Author: Kenneth (K.P.) Smith, Ph.D.

Kenneth (K.P.) Smith, Ph.D.
Dr. Kenneth (K.P.) Smith is assistant director of the Infectious Disease Diagnostics Laboratory at Children's Hospital of Philadelphia.