Rapid Diagnosis of Disseminated Tuberculosis Using Cell-free DNA Sequencing in a Kidney Transplant Recipient.

Abstract

Here, we report a case of smear-negative disseminated tuberculosis (TB) in a kidney transplant recipient diagnosed using plasma cell-free DNA (cfDNA). A 68-y-old woman presented with 2 mo of fevers and a 15-pound weight loss. Six years earlier, she received a living donor kidney transplant from her husband for adult polycystic kidney disease. She received standard induction with antithymocyte globulin and maintenance immunosuppression with tacrolimus, mycophenolate mofetil, and low-dose prednisone. Her posttransplant course was uncomplicated. She was born and lived in Massachusetts and had not traveled internationally in many years. She worked in finance until she retired a few years prior. She had no history of incarceration or homelessness. Her hobbies included volunteering at a local food bank. On presentation, she was febrile to 38.8 °C and diaphoretic but otherwise appeared well. Her laboratory values were notable for a leukocytosis of 16 000 K/µL and a urinalysis with >100 white blood cells per high power field. There was no growth in the urine or blood cultures. Her nasopharyngeal multiplex polymerase chain reaction (PCR) was negative for common respiratory viruses. She had negative test results for Epstein-Barr virus DNA, cytomegalovirus DNA, HIV 1/2 antigen/antibody, cryptococcal and Histoplasma antigens. Chest computed tomography showed innumerable 1- to 3-mm bilateral nodules (Figure 1A) and lytic lesions of the T6 and T7 vertebral bodies. Spine magnetic resonance imaging demonstrated enhancing lesions involving the entire T6 and T7 bodies (Figure 1B).FIGURE 1.: Abnormal imaging findings. A, Chest CT scan with diffuse bilateral pulmonary nodules in a miliary pattern. B, MRI of the thoracic spine with enhancing lesions of the T6 and T7 vertebral bodies. CT, computed tomography.Further workup included bronchoalveolar lavage with multiple negative studies: bacterial cultures, Pneumocystis direct fluorescent antibody, galactomannan antigen, acid-fast bacilli (AFB) and modified AFB smear, and Mycobacterium tuberculosis PCR. Transbronchial biopsy was nondiagnostic, with negative AFB and fungal stains. Mycobacterial blood and urine cultures were also obtained. A biopsy of her T7 vertebral body noted chronic inflammation and 1 nonnecrotizing granuloma. On hospital day (HD) 13, while she remained febrile and without a definitive diagnosis, we sent a quantitative cfDNA test (Karius, Inc, Redwood City, CA). On HD 15, the Karius cfDNA test returned positive for M tuberculosis. Subsequently, the mycobacterial blood, urine, and bronchoalveolar lavage cultures grew M tuberculosis on HD 17, 17, and 21, respectively. Bone cultures grew M tuberculosis at 34 d after biopsy after the patient had been discharged from the hospital. Cell-free DNA tests have become increasingly prevalent in recent years, including donor-derived cfDNA, which is routinely used to detect allograft rejection. The Karius cfDNA test uses metagenomic next-generation sequencing to specifically detect microbial nucleic acid from over a thousand clinically relevant species in plasma.1 This was the first case of TB diagnosed by plasma cfDNA sequencing in our institution. The turnaround time of 48 h is a promising characteristic of this test, especially when slow-growing pathogens are suspected. Had it been obtained earlier in the course, this test would have not only expedited diagnosis and treatment initiation but also eliminated the need for invasive diagnostics. Nevertheless, like other molecular tests, cfDNA sequencing does not provide antimicrobial susceptibilities to guide antimicrobial therapy and thus does not eliminate the need for cultures. In a recent meta-analysis, cfDNA sequencing showed excellent specificity but poor pooled sensitivity for the diagnosis of TB.2 The very low-risk epidemiologic history for TB and disease onset in the late posttransplant period were other interesting features of this case. Her pretransplant T spot TB was negative. We considered donor-derived TB, but this was deemed unlikely because her donor was her husband, who was healthy and had a negative TB screening. She had no known exposures to TB, and her exposure to higher-risk populations via volunteerism was her only epidemiologic risk factor.