Post liver transplant tuberculosis

Abstract

We report on a patient diagnosed with disseminated (hepatic and pulmonary) tuberculosis in the context of immunosuppression following liver transplant. During the administration of anti-tuberculosis drugs an abrupt elevation of liver enzymes was detected leading us to suspect drug toxicity rather than graft rejection. Nevertheless, careful surveillance and adjustment of serum levels of immunosuppressant drugs permitted continuance of tuberculosis treatment with no further side effects.

BACKGROUND

Immunosuppression following solid organ transplantation is often associated with different patterns of infection. Managing these infections can be quite challenging depending on the pathogen involved and type of drugs needed for treatment. In this particular case, and because the transplanted organ was affected, things were even more complicated. There are no specific diagnostic procedures to determine whether liver enzyme elevation was due to infection, drug toxicity or, even, graft rejection. This case demonstrates how we were able to manage this problem and still maintain an effective anti-tuberculosis drug regimen. Had we been unable to do so, death would have been the most probable result.

CASE PRESENTATION

A 53-year-old woman was sent to our outpatient clinic in January 2008 after discharge from hospital in December. She had been diagnosed with post liver transplant tuberculosis.

The patient had been submitted to orthotopic liver transplant in June 2007 due to liver involvement in familial amyloidotic polyneuropathy. After the surgical procedure she had begun immunosuppressive drugs (azathioprine 100 mg/day during the first month, ciclosporin 10 mg/kg during the second month and prednisolone tapered to 5 mg/day). A favourable clinical progress was observed and she was stable until December 2007 (approximately 6 months). At that time she started to complain of high fever, predominantly nocturnal, asthenia, anorexia and night sweats. These symptoms evolved for about 15 days, when she finally consulted her attending physician. These complaints, together with elevated liver enzymes (up to seven times the normal value), raised the suspicion of cholangitis or, most importantly, transplant rejection. The patient was admitted to the surgical ward of the local hospital. No other signs and symptoms were found to be present at the time.

There was no history of cough, sputum production, chest pain or haemoptysis. The patient had a previous history of respiratory tuberculosis (diagnosed and treated 20 years before), but no information was available regarding the drugs and doses used. No history of recent exposure to tuberculosis or other infectious diseases was apparent. Results of previous tuberculin skin tests were unclear and unavailable for verification. The chest x ray, prior to the surgical procedure, was normal and BCG vaccine had been administered at birth (suggested by the presence of a typical scar of the left arm).

On the day of admission, an ultrasound exam of the abdomen (fig 1) was performed and revealed a nodular lesion in the posterior region of liver segment VI; cholangiectasis of the left lobe and splenomegaly were also found but endoscopic retrograde cholangiopancreatography (ERCP) failed to show significant anomalies.

Figure 1.

Transverse and longitudinal sonographic images. Solid nodule in segment VI with smooth contours and isoechoic echo structure to the liver parenchyma.

These findings did not corroborate the diagnostic hypothesis put forth and a CT scan of the upper abdomen was performed (fig 2). This exam confirmed the presence of the nodular lesion, which had a low density and did not enhance with contrast (fig 2B–D). Images acquired above the diaphragm revealed multiple diffuse centrilobular ground-glass opacities with a tree-in-bud pattern (fig 3A).

Figure 2.

CT scan. (A) non-enhanced axial thoracic image—multiple diffuse centrilobular ground-glass opacities, assuming in some segments a tree-in-bud pattern. (B) Hypodense lesion without significant enhancement after intravenous contrast administration, both during arterial (C) and portal-venous (D) phase.

Figure 3.

Histological exam of liver biopsy specimen. (A) H&E staining, original magnification ×10, revealing the presence of granulomatous hepatitis with occasional spots of necrosis. (B) H&E staining, original magnification ×40, focusing on a granuloma. (C) Ziehl-Nielsen staining, original magnification ×100, revealing the presence of acid-fast bacilli.

A CT guided biopsy of the nodule was obtained. Histology revealed a granulomatous hepatitis with areas of necrosis. Ziehl-Nielsen staining revealed the presence of acid-fast bacilli (fig 3C).

A fibreoptic bronchoscopic exam was also performed and acid-fast bacilli were identified in the bronchoalveolar lavage fluid.

Microbial culture of both the bronchoalveolar lavage fluid and liver biopsy specimen confirmed the presence of Mycobacterium tuberculosis with in vitro sensitivity to all of group 1 anti-tuberculosis agents as well as streptomycin.

OUTCOME AND FOLLOW-UP

A combination of isoniazid (H) 300 mg/day, rifampicin (R) 600 mg/day, ethambutol (E) 1200mg/day and pyrazinamide (Z) 1000 mg/day was initially administered. Simultaneously doses of prednisone and ciclosporin were reduced to 2.5 mg/day and 80 mg/day, respectively. Monitoring of liver function revealed a sustained decrease in serum liver enzyme values until the second month of anti-tuberculosis treatment. At that time, an unexpected rise in aspartate aminotransferase and alanine aminotransferase values was observed (up to seven times the normal upper limit; fig 4).

Figure 4.

Serum measurements of aspartate aminotransferase (AST; reference interval: 15 – 46 U/L) and alanine transaminase (ALT; reference interval: 10 – 66 U/L).

Liver toxicity due to anti-tuberculosis drugs was suspected and the question now was how to manage this situation. Should anti-tuberculosis treatment be stopped? Should these be replaced by drugs with no liver toxicity?

It was decided that the current drug regimen would be continued since, at that time, levels of immunosuppressant drugs were very low (fig 5) and liver function shift could also be due to transplant failure. Careful surveillance of liver function was performed and, in fact, when the dose of ciclosporin was raised to 300 mg (serum levels: 5.6 mg/Kg) liver function improved.

Figure 5.

Dose administered of prednisolone (mg) and measured serum levels of ciclosporin (mg/Kg).

The patient completed 6 months of treatment (2 months of HRZE followed by 4 months of HR). No other side effects were observed during the treatment period. Liver function and all other laboratory values remained normal. The patient is still on prednisolone (5 mg/day) and has returned to maintenance doses of ciclosporin (150 mg/day). She is clinically stable and no unprogrammed hospital visits have been registered.

DISCUSSION

Immunosuppression following solid organ transplantation is often associated with different types of infection. These can be nosocomial, community acquired or, as is the case, reactivation of latent infection (either of the transplanted organ or host). The interval between organ transplant and diagnosis of an infectious disease has implications as to the most probable pathogen. Hence, in the first month after surgery, infection is most probably due to the donor, complications resulting from the surgical procedure or admission to hospital. In the period 1–6 months after transplantation, the pattern of infection changes. During this time, immunosuppression is high and patients are most at risk for the development of opportunistic infections. Risk for reactivation of latent infection is higher during this time frame. After this period, the patient is usually receiving lower doses of immunosuppressive agents and infection is most likely caused by a community acquired pathogen.¹

In order to exclude possible sources of a latent infection, clinical evaluation of a patient about to receive a transplant must include detailed information about travel history, employment, hobbies, high-risk behaviour, previous infections and admissions to hospital. Current guidelines^1,2 suggest the screening for a number of infectious agents to detect asymptomatic infection in the transplant candidate. These include HIV 1 and 2, human T lymphotropic virus (HTLV) I and HTLV II, viral hepatitis A, B and C, cytomegalovirus, Epstein-Barr virus, herpes simplex virus, varicella zoster virus, syphilis and Toxoplasma gondii.

As it pertains to Mycobacterium tuberculosis, testing for latent infection is currently reserved for those patients with confirmed or suspected exposure within the 2 years prior to transplant. Prophylaxis is indicated whenever there is a recent tuberculin skin test conversion, history of recent contact with a tuberculosis case and recipients of organs from donors with a history of untreated tuberculosis.³ Testing for latent infection using the new interferon gamma release assays promises to add an advantage due to its higher sensitivity in the setting of immunosuppression, but guidelines are still lacking.⁴ In our unit, both tests are performed and decision to treat latent tuberculosis is based on the positivity of either one.

The treatment of post-transplant tuberculosis presents a few challenges, mainly due to possible interaction between the antimicrobial agents and immunosuppressive drugs. Of these, the most troublesome is rifampicin, which can interact with most of the immunosuppressive drugs, in particular corticosteroids and ciclosporin, which need dose adjustment. Simultaneous use of rifampicin leads to increased metabolism of both of these drugs via cytochrome P450 (CYP) isoenzymes.

When the transplanted organ is the liver, problems with the use of anti-tuberculosis drugs are of greater significance since liver toxicity may be seen with most of them, particularly isoniazid and pyrazinamide. Distinguishing the cause of liver dysfunction presents a challenge. Increase in serum values for liver enzymes can be due to infection, drug toxicity or graft rejection, and there is no tangible method to determine which is responsible for the changes observed. The best approach is to carefully assess both liver function and serum levels of immunosuppressive drugs.

When liver enzymes rise in a patient being treated with both anti-tuberculosis and immunosuppressive drugs, careful dosing of the second should be performed in order to prevent graft rejection and allow the body to respond to the infection. All attempts should be made to maintain an effective anti-tuberculosis regimen since death is the most probable result if the patient is left untreated.

LEARNING POINTS

• Immunosuppression following solid organ transplantation is often associated with different types of infection, which can be nosocomial, community acquired or, as is the case, reactivation of latent infection.

• Testing for latent infection with Mycobacterium tuberculosis is currently reserved for patients with confirmed or suspected exposure to tuberculosis within the 2 years prior to transplant.

• Treatment of post-transplant tuberculosis presents a few challenges mainly due to possible interaction between the antimicrobial agents and immunosuppressive drugs.

• Carefully assessment of both liver function and serum levels of immunosuppressive drugs was the key to conduct a successful treatment.