Tuberculosis is caused by Mycobacterium tuberculosis (MTB), a rod-shaped bacterium that requires special media for culture and that has the peculiarity of “acid-fastness” (resistance to decolorization by strong acidic decolorizing chemicals such as acid alcohol after being stained by certain stains such as carbol fuchsin). Tuberculosis is still very important and common despite advances in drug therapy. It has been reported that about 25% of persons exposed to MTB will become infected; and of those infected, about 10% will develop clinical disease (range, 5%-40%). The disease usually begins in the chest due to inhalation of airborne infectious material. This material is carried to some localized area of the lung alveoli and provokes a host response of granulomatous inflammation around the material (the “Ghon complex”). It is thought that in many cases there is also a silent hematogenous spread of the organisms. In most cases the host is able to contain and eventually destroy the organisms in the chest and those reaching other locations. Those in the lungs seem better able to survive than those deposited elsewhere. In some cases the organisms remain dormant, and the infection can be reactivated at a later date; in some cases the initial infection spreads; and in some cases reinfection takes place. If infection in the lungs progresses to clinical disease, the most important symptoms are cough, fever, and hemoptysis. (The most important diseases to rule out are lung carcinoma and bronchiectasis.) The kidney is involved in a small percentage of advanced cases, with the main symptom being hematuria (Chapter 12). A small number of patients develop widespread extrapulmonary disease, known as miliary tuberculosis. The laboratory findings in tuberculosis depend to some extent on the stage and severity of the disease.

Chest x-ray films

Chest x-ray films often provide the first suggestion of tuberculosis and are a valuable parameter of severity, activity, and response to therapy. Depending on the situation, there are a variety of possible roentgenographic findings. These may include one or more of the following:

1. Enlargement of hilar lymph nodes.
2. Localized pulmonary infiltrates. These occur characteristically in an upper apical location or, less commonly, in the superior segment of the lower lobes. Cavitation of lesions may occur.
3. Miliary spread (small punctate lesions widely distributed). This pattern is not common and may be missed on routine chest x-ray films.
4. Unilateral pleural effusion. The most common causes are tuberculosis, carcinoma, and congestive heart failure. Tuberculosis has been reported to cause 60%-80% of so-called idiopathic pleural effusions, although this percentage varies greatly depending on the patient’s geographic location and other factors.

Sputum smear

Sputum smears provide a rapid presumptive diagnosis in pulmonary tuberculosis. The smear is usually stained by one of the acid-fast (acid-fast bacillus, or AFB) procedures (usually the Ziehleelsen or Kinyoun methods). Fluorescent Auramine-o staining methods are available, faster, and somewhat more sensitive. Smears require about 5 Ч 103 organisms/ml of specimen for microscopic detection. The more advanced the infection, the more likely it is to yield a positive smear. Therefore, the rate of positive findings is low in early, minimal, or healing tuberculosis. Also, the smear may be normal in a substantial minority of advanced cases. Culture is more reliable for detection of tuberculosis and also is necessary for confirmation of the diagnosis, for differentiation of MTB from the “atypical” mycobacteria, and for sensitivity studies of antituberculous drugs. According to the literature, false negative smears (smear negative but culture positive) have been reported in an average of 50% of cases (literature range, 16%-70%). Some of these false negative results may be due to laboratory technique problems and differences in smear staining methods. A high centrifugation speed when concentrating the specimen is said to increase the yield of positive smears. False positive smears (positive smear but negative culture) have been reported, averaging about 1%-5% of positive smears (literature range, 0.5%-55%). Some of these were apparently due to contamination of water used in the smear-staining procedure by saprophytic mycobacteria. Control slides are necessary to prevent this. Some authorities believe that only 1-2 acid-fast organisms/300 oil immersion fields should be considered negative (although indicative of need for further specimens). Smears may sometimes be positive for up to 2 months when cultures are negative if the patient is on antituberculous drug therapy (this would not be considered a genuine false positive, since the drugs inhibit mycobacterial growth or the organisms may be nonviable). After 2 months, persistence of positive smears raises the question of treatment failure. Temporary persistence of positive smears with negative cultures is more likely to occur in severe cavitary disease (in one series, this occurred in 20% of cases). Sputum specimens should be collected (for culture and smear of the concentrated specimen) once daily for at least 3 days. If the smear is definitively positive, further smears are not necessary. Also, a definitively positive smear means high probability that culture of the specimens already collected will obtain positive results, and it is not necessary to collect more than three specimens or to proceed to more complicated diagnostic procedures. If smears are negative, one must consider the possibility that the culture may also be negative, and conventional cultures on standard solid media average 20 days to produce growth from MTB smear-positive specimens and about 27 days from MTB smear-negative specimens (overall range, 2–8 weeks).

Culture

Sputum culture is preferred for pulmonary tuberculosis (gastric aspiration may be done if adequate sputum specimens cannot be obtained); urine culture is preferred for renal involvement; and bone marrow culture is preferred in miliary tuberculosis. Reports indicate that an early morning specimen, either of sputum or urine, produces almost as many positive results as a 24-hour specimen and has much less problem with contamination. Special mycobacteria culture media are needed. The necessity for adequate sputum culture specimens, regardless of the concentrated smear findings, should be reemphasized. Several reports indicate that aerosol techniques produce a significantly greater yield of positive cultures than ordinary sputum collection. The aerosol mixture irritates the bronchial tree and stimulates sputum production. At any rate, it is necessary to get a “deep cough” specimen; saliva alone, although not completely useless, is much less likely to reveal infection and is much more likely to be contaminated. If sputum cultures are negative or if the patient is unable to produce an adequate sputum sample, gastric aspiration may be used. Gastric contents are suitable only for culture; nontuberculous acid-fast organisms may be found normally in the stomach and cannot be distinguished from M. tuberculosis on AFB smear. If renal tuberculosis is suspected, urine culture should be done (Chapter 12). However, renal tuberculosis is uncommon; and even with urine specimens obtained on 3 consecutive days, only about 30% of cases are positive.

Cultures should be grown in high carbon dioxide atmosphere, since this is reported to increase the number of positive cultures by at least 10%. Inoculation on several varieties of media increases the number of positive results by 5%-10%. The 4% sodium hydroxide traditionally used to digest and decontaminate sputum before concentration also kills some mycobacteria. Use of weaker digestion agents increases culture yield, but troublesome overgrowth by other bacteria may also increase.

Culture should be done on all tissue specimens when tuberculosis is suspected. Acid-fast stains on tissue slides reveal tuberculosis organisms in only 30%-40% of cases that are positive by culture. Several newer methods such as BACTEC (which uses liquid media and a machine that detects metabolic products of bacterial growth) have been able to decrease detection time for MTB smear-positive specimens to 8 days and time for MTB smearegative specimens to 14 days (overall range, 1-3 weeks). The system is about 93% sensitive compared to conventional multimedia culture. Once culture growth occurs, the organism must be identified. Conventional methods require biochemical and growth tests to be performed that may take 3-6 weeks to complete. The BACTEC system has a nucleic acid phosphate method that can identify MTB (only) in 3-5 days. Commercial DNA probes are available that can identify MTB and certain non-MTB mycobacteria in 1 day. Gas-liquid chromatography and high-performance liquid chromatography have also been used. Antibiotic sensitivity studies are recommended when a mycobacterial organism is isolated, since multiresistant MTB is increasing in various areas and non-MTB mycobacteria have often been multiresistant. Conventional culture methods take 21 days; BACTEC takes about 5 days.

Data on sputum culture sensitivity using conventional AFB media is difficult to find since culture is usually considered the gold standard of AFB detection. Sputum culture appear to average about 75% sensitivity (range, 69%-82%). Sensitivity using BACTEC averages about 85% (range, 72%-95%).

Nucleic acid probe

Nucleic acid (DNA) probe methods are now becoming available that permit direct nonsmear detection of mycobacteria in clinical specimens. The first such test available (Gen-Probe, Inc.) is reported in two studies to be 83%-85% sensitive compared with culture using sputum specimens. However, it has been reported that antituberculous drugs can interfere with the probe; this study found sensitivity in nontreated patients to be over 90% (when comparing probe to culture, culture only detects about 75%-80% of cases). Ten percent of specimens were positive by probe but negative by culture, which may represent additional true positives that could not be confirmed. A DNA probe is also available specifically for M. tuberculosis. Same-day results can be obtained. Disadvantages of this first-generation method are need for considerable technologist time and certain special equipment. Compared with culture, the general Mycobacterium screen probe will not differentiate M. tuberculosis from other mycobacteria, whereas the specific M. tuberculosis probe will not detect the other mycobacteria. Neither probe would provide therapeutic drug sensitivity information. The major drawback regarding its use as replacement for the acid-fast smear is the relatively high cost of the probe method, very high if only one specimen at a time is processed. DNA probes with PCR amplification have been reported (e.g., Roche Diagnostics) that are said to have a sensitivity of 3-30 organisms/ml (compared to at least 5 Ч 103 organisms/ml required for a positive acid-fast smear). Nevertheless, one study involving 7 prestigious worldwide reference laboratories who were sent sputum or saliva specimens to which various quantities of BCG (M. Bovis) mycobacteria were added, showed false positive PCR rates of 3%-77%. In specimens containing 102 organisms, sensitivity ranged from 0%-55%; in specimens containing 103 organisms, 2%-90%; and in specimens containing 101 organisms, 20%-98%.

Skin test (Mantoux test)

This test is performed with an intradermal injection of purified protein derivative (PPD) or old tuberculin (Table 14-1). A positive result is represented by an area of induration having a specified diameter by 48 hours. The diameter used to be 10 mm but was redefined in 1990 to require different diameters depending on the person’s risk group (see box). In addition, a distinction was made between “reaction” (diameter or width of induration without record of previous test result) and “conversion” (increase in reaction width within 2 years from last previous reaction width). For all persons younger than 35 years of age whose previous reaction was negative, an increase in PPD induration of 10 mm or more in diameter within a period of 2 years would be considered a conversion and presumptive suspicion for occult tuberculosis (TB), whereas the change would have to be at least 15 mm for persons 35 years of age or more (that is, for nonrisk persons above or below age 35 who have had a PPD within 2 years, conversion criteria would replace reaction size criteria).

Table 14-1 Comparison of tuberculosis skin tests*
Comparison of tuberculosis skin tests


A positive skin test is a manifestation of hypersensitivity to the tubercle bacillus. This reaction usually develops about 6 weeks after infection, although it may take several months. A positive reaction means previous contact and infection with TB; the positive reaction does not itself indicate whether the disease is currently active or inactive. However, in children under 3 years of age it usually means active TB infection. Apparently, once positive, the reaction persists for many years or for life, although there is evidence that a significant number of persons revert to negative reactions if the infection is completely cured early enough. In a few cases of infection the test never becomes positive. The Mantoux test may revert to negative or fail to become positive in the following circumstances:

1. In about 20% of seriously ill patients, due to malnutrition (severe protein deficiency).
2. In newborns and occasionally in old age.
3. In some persons with viral infections, or within 1 month after receiving live virus vaccination.
4. In 50% or more of patients with miliary tuberculosis.
5. In a high percentage of patients with overwhelming pulmonary tuberculosis.
6. In a considerable number of patients who are on steroid therapy or immunosuppressive therapy.
7. In many persons who also have sarcoidosis or Hodgkin’s disease.
8. In some persons with chronic lymphocytic leukemia or malignant lymphoma.
9. In some patients with chronic renal failure or severe illness of various types.
10. In some persons with old infection (“waning” of reactivity).
11. When there is artifact due to improper skin test technique (e.g, subcutaneous rather than intradermal injection).

In cachectic patients and those with protein malnutrition, treatment with an adequate protein diet can restore Mantoux test reactivity to most patients after about 2 weeks. In patients after age 50 with M. tuberculosis infection, especially those with old previous infection, the PPD skin test sometimes may slowly decrease in reactivity and eventually become negative. (How often this occurs is controversial; the best estimate seems to be 8%-10%, but studies range from 0.1%-21%, possibly influenced by the elapsed time period since infection and time intervals between testing.) If another skin test is performed, the new skin test itself stimulates body reaction and may restore reactivity (“booster phenomenon”). This phenomenon could simulate a new infection if previous infection were not known, since the next time a skin test is performed the physician would see only conversion of a negative to a positive reaction. Restoration of skin reactivity can take place in only 1 week, so that retesting 1 week after the first negative reaction can usually show whether or not there is potential for the booster reaction. (The 1-week interval would in most cases not be long enough for true conversion in persons with their first infection.) Repeated skin tests will not cause a nonexposed person to develop a positive reaction. Some investigators recommend skin testing with antigens used to demonstrate the presence of skin test anergy (e.g., Candida or Trichophyton antigen) if the Mantoux test is repeatedly negative in a person with substantial suspicion of mycobacterial infection.

The standard procedure for skin testing is to begin with an intermediate strength PPD (or the equivalent). If the person has serious infection, some clinics recommend starting with a first-strength dose to avoid necrosis at the injection site. A significant minority of patients with tuberculosis (9%-17%) fail to react to intermediate strength PPD; a second-strength dose is then indicated.

Miliary tuberculosis

Miliary TB is clinically active TB that is widely disseminated in the body by hematogenous spread. Clinical symptoms are often nonspecific, such as fever, weakness, and malaise. There frequently is an associated condition, such as alcoholism, intravenous (IV) drug abuse, or malignancy, that decreases immunologic defenses. About 20% have a negative tuberculin skin test reaction. About 35% do not show a miliary pattern on chest x-ray film. If routine clinical and culture methods fail, biopsy of bone marrow or liver may be useful. Liver biopsy has a fairly good positive yield (up to 80%), considering that a needle biopsy specimen is such a tiny random sample of a huge organ. However, it is usually difficult to demonstrate acid-fast organisms on liver biopsy even when tubercles are found, and without organisms the diagnosis is not absolutely certain. Bone marrow aspiration is probably the best procedure in such cases. Bone marrow yields much better results for mycobacterial culture than for demonstration of tubercles. Routine marrow (Wright-stained) smears are worthless for histologic diagnosis in TB. Aspirated material may be allowed to clot in the syringe, then formalin-fixed and processed as a regular biopsy specimen for histologic study. Before clotting, some of the aspirate is inoculated into a suitable TB culture medium. It should be emphasized that bone marrow aspiration or liver biopsy is not indicated in pulmonary tuberculosis (since this disease is relatively localized), only in miliary TB.

PPD Reaction Size Considered “Positive” (Intracutaneous 5 TU* Mantoux Test at 48 Hours)
5 MM OR MORE
Human immunodeficiency virus
(HIV) infection or risk factors for HIV
Close recent contact with active TB case
Persons with chest x-ray consistent with healed TB
10 MM OR MORE
Foreign-born persons from countries with high TB prevalence in Asia, Africa, and Latin America
Intravenous (IV) drug users
Medically underserved low-income population groups (including Native Americans, Hispanics, and African Americans)
Residents of long-term care facilities (nursing homes, mental institutions)
Medical conditions that increase risk for TB (silicosis, gastrectomy, undernourished, diabetes mellitus, high-dose corticosteroids or immunosuppression RX, leukemia or lymphoma, other malignancies
Employees of long-term care facilities, schools, child-care facilities, health care facilities
15 MM OR MORE
All others not listed above
* TU, tuberculin units.

Renal tuberculosis

Renal TB is almost always bilateral and presumably results from nonhealing infection produced during the transient bacteremia of the lung primary stage. There is usually a latent period of many years before clinical infection becomes evident. The estimated incidence of eventual active renal infection is about 2%-5%, but this probably represents incidence in high-risk groups. About 25% (range, 20%-75%) of patients are said to have a normal chest x-ray film. About 14% (range, 12%-15%) of patients are reported to have a negative PPD skin test result. Even the intravenous pyelogram results (IVP) are normal in about 25% (range, 14%-39%) of cases. Most patients do not have systemic symptoms, such as fever. The erythrocyte sedimentation rate was elevated in 23% of patients in one report. Only 20%-56% have urinary tract symptoms. Gross hematuria is the classic finding in renal TB but is present in only about 20% of patients. Pyuria (with negative urine culture) and microscopic hematuria are more frequent, occurring in about 65%-85% of cases. Some patients have a positive urine culture with some ordinary pathogen in addition to renal TB. Urine culture for TB was mentioned previously; 30%-80% of patients have positive cultures when three 24-hour or early morning specimens are collected (true culture sensitivity is probably less than 80%, since the diagnosis could easily be missed with negative cultures).