Advances in Peritoneal Dialysis, Vol. 27, 2011

Meltem Gursu, 1 Filiz Tayfur, 1 Mine Besler, 2 Ozlem Kaptanogullari,3 Mehmet Kucuk,4 Zeki Aydin,1 Taner Basturk, 5 Sami Uzun, 1 Serhat Karadag, 1 Emel Tatli, 1 Abdullah Sumnu,1 Savas Ozturk,1 Rumeyza Kazancioglu6

Tuberculosis has been paid more attention in recent years because of the increase in the number of patients with immune suppression—such as those with renal failure. In the present study, we analyzed patients on peritoneal dialysis (PD) in our city to determine the prevalence and clinical characteristics of tuberculosis in those patients. Patients who had been on a PD program for more than 3 months were reviewed. Demographic characteristics, primary renal disease, comorbidities, and duration of PD were recorded. With regard to tuberculosis, the timing of the diagnosis, any previous history of antituberculosis treatment, family history, site of presentation, drugs used, drug side effects, and disease outcome were recorded. Among 322 patients from 5 PD units who were reviewed, 4 (1.24%) were found to have tuberculosis. Pulmonary involvement was noted in 2 (50%). The diagnosis was made through microbiology in 1 patient, through pathology in 1, and through clinical and radiologic assessment in the remaining 2. Mild transaminitis was recorded in 2 patients as a side effect of treatment. Of the 4 patients, 2 were cured, 1 died, and 1 was taking ongoing treatment. The prevalence of tuberculosis was significantly higher in the study population than in the general population. In a dialysis population, a diagnosis of tuberculosis is often difficult, and extrapulmonary involvement is more common, as observed in our study. The diagnosis of tuberculosis may be made through non-microbiologic approaches, and temporary transaminase elevations may be seen during therapy. From: 1Nephrology, Haseki Training and Research Hospital; 2Nephrology, Istanbul Training and Research Hospital; 3Nephrology, Bakirkoy Training and Research Hospital; 4Nephrology, Okmeydani Training and Research Hospital; 5Nephrology, Bagcilar Training and Research Hospital, and 6Nephrology, Bezmialem Vakif University Medical Faculty; Istanbul, Turkey.

Tuberculosis in Peritoneal Dialysis Patients in an Endemic Region Key words Tuberculosis, diagnosis, treatment Introduction Tuberculosis has been paid more attention in recent years because of the increase in the number of patients with immune suppression—such as those with renal failure. The clinical presentation of tuberculosis is different in patients with chronic renal failure than in those with a competent immune system. Extrapulmonary tuberculosis is more common in the hemodialysis (HD) population, with rates of 38% and 50% having been reported (1–3). The diagnosis is usually late, sometimes postmortem, because of nonspecific clinical presentation (fever of unknown origin, fatigue, loss of weight), negative tuberculin skin test (anergy), and a low probability of microbiologic evidence (acidresistant bacilli, culture in tuberculosis media) (4). Moreover, because most of the reports in the literature are related to HD patients, data about the prevalence of tuberculosis in peritoneal dialysis (PD) populations are limited. As a result, an effective program for the prevention, screening, early diagnosis, and treatment of tuberculosis is needed in this population. The PD population differs somewhat from the HD population in terms of immune disturbances, which may affect the prevalence of tuberculosis in the former group. Some cases present as culturenegative peritonitis or culture-positive peritonitis unresponsive to appropriate antibiotics. Diagnosis requires a high index of suspicion. In the present study, we analyzed patients on PD in our city to determine the prevalence and clinical characteristics of tuberculosis in those patients. Methods Patients who had been on a chronic PD program for more than 3 months in any of 5 PD centers representing various socio-economic and cultural characteristics of our city’s population were reviewed for the study. Patients who were not residents of the city, but only

Gursu et al. guests; patients who had temporarily been switched to PD from HD for mechanical or metabolic reasons; and patients had been on PD for less than 3 months were excluded. Clinical, demographic, and laboratory parameters were obtained for those among the included patients who had been diagnosed with tuberculosis at least 3 months after the start of PD. Demographic parameters (age, sex, birthplace, education status, occupation), primary renal disease, comorbid diseases, and duration of PD were recorded. All PD patients in Turkey have a chest X-ray annually. If clinical or radiologic evidence raises suspicion of either pulmonary or extrapulmonary tuberculosis, then further radiologic and microbiologic examinations are performed. Our retrospective review of the patient files found those who had been diagnosed with tuberculosis. Using a predefined form, we recorded the timing of the tuberculosis diagnosis; previous history of anti-tuberculosis treatment; family history; site of presentation; names, doses, and durations of drugs; side effects of treatment; and disease outcome. If patients switched between PD centers, they were recorded as attending the center in which they were diagnosed with tuberculosis. The statistical analysis was carried out using the SPSS software application for Windows (version 13.0: SPSS, Chicago, IL, USA). Numeric parameters are expressed as mean ± standard deviation. Results We reviewed 322 patients (156 women, 166 men) from the 5 PD units. Among the patients reviewed, 4 (1.24%) were diagnosed with tuberculosis. All were young adults, and 3 of the 4 were men. None of the patients had history of immunosuppressive medications or any disease other than chronic renal failure that might cause immunosuppression. Table I presents demographic characteristics and PD treatment data for the study patients. Neither the patients nor their families had a history of tuberculosis. Pulmonary involvement was noted in 2 of the 4 patients (50%); of the others, 1 had peritoneal tuberculosis, and 1 had tuberculous lymphadenitis. The diagnosis was made through microbiology in 1 patient, through pathology in 1, and through clinical and radiologic assessment in the remaining 2. The 2 patients with presumed pulmonary tuberculosis presented with low-grade fever, weight

49 loss, nonproductive cough, and pulmonary infiltrates, without cavity formation, on chest radiography. Nonspecific and specific cultures of blood and sputum were negative. With no clinical response to broadspectrum antibiotics, they were accepted as having pulmonary tuberculosis. Appropriate treatment resulted in improved wellness and resolution of symptoms and radiologic findings. Planned treatment for 3 of the patients involved the use of 4 antituberculosis drugs in the first 3 months, followed by therapy with 2 drugs for another 9 months (Table II). One patient was treated with 3 drugs, but the reason for the omission of the 4th drug could not be ascertained from the patient’s chart. In the patient with tuberculous peritonitis, the peritoneal catheter was removed; he was switched to HD, first with a non-cuffed jugular catheter and than with an arteriovenous fistula. No serious side effects from the drug therapy were observed, except for mild transaminitis in 2 patients. Transaminase levels returned to normal levels after withdrawal of the drugs; the drugs were then reinstated with gradually increasing doses. In 2 patients, a complete cure was achieved; 1 patient died from tuberculoma of the brain in the 4th month of treatment; and 1 patient with tuberculous peritonitis was receiving ongoing treatment, with a good clinical response at the 4th month (Table II). Discussion Studies of the incidence and prevalence of tuberculosis in the dialysis population have reported varying results (1,4,5). Data on the PD population are limited. The 322 patients reviewed in the present study make it the largest study of this type performed in Turkey. According to reports from the local health authority, there are 677 PD patients in our city; we therefore reviewed 47.56% of the PD population. Compared with HD patients, PD patients appear to have an increased incidence of peritoneal involvement in tuberculosis—findings that come from case reports (3,6,7). According to registry reports from the Ministry of Health in Turkey, the incidence and prevalence of tuberculosis in the general population in 2007 were 25.2 per 100,000 and 27.9 per 100,000 respectively. The prevalence in our study population (1.24%) is significantly higher. According to the 2009 statistical reports from the Ministry of Health, Istanbul is the

50 table i

Tuberculosis in a PD Population

Demographic and peritoneal dialysis (PD) treatment data for the study subjects Characteristic

Age (years) Sex Primary renal disease PD modality Immunosuppressive medicationa Comorbidities

Patient

1

2

27 Female Crescentic GN CAPD None Juvenile rheumatoid arthritis

23 Male Unknown APD None None

3

4

35 Male Unknown CAPD None None

22 Male Unknown CAPD None None

a

Prior or current. GN = glomerulonephritis; CAPD = continuous ambulatory PD; APD = automated PD. table ii

Tuberculosis (TB) data for the study subjects Characteristic

Time to diagnosis (months)a Prior history of TB Personal Family Site of involvement Diagnostic method Drugs used

Side effects Outcome of TB

a

1

2

5 None None Lung Clinical and radiologic findings Isoniazid, plus rifampicin, plus morphozinamid None Cure

Patient

3

4

9

22

5

None None Cervical and mediastinal lymph nodes Pathology

None None Lung

None None Peritoneum

Clinical and radiologic findings Isoniazid, plus rifampicin, plus ethambutol, plus pyrazinamide Transaminitis Death from brain tuberculoma in the frontal lobe

Microbiology

Isoniazid, plus rifampicin, plus ethambutol plus morphozinamid None Cure

Isoniazid, plus rifampicin, plus ethambutol, plus pyrazinamide Transaminitis Continuing treatment

From PD start.

city with the highest values for both the incidence and the prevalence (50.8 per 100,000 population) of tuberculosis. But Istanbul is also the city with the highest rate of immigration from other cities, both for socio-economic reasons and for health problems, which might lead to the high incidence and prevalence of tuberculosis. To minimize the potential effect of those factors, we excluded PD patients who were followed in another city, but who were only guests in our city.

The presentation of tuberculosis may be different in PD patients than in HD patients. Pulmonary involvement is still in first place (4), but tuberculous peritonitis has a special place in PD. Early diagnosis and treatment is mandatory in peritoneal tuberculosis. Clinical findings are indistinguishable from those in bacterial peritonitis, with fever, abdominal pain, and cloudy effluent being the most frequent (8–10). Cell count cannot differentiate peritoneal tuberculosis from other peritoneal infections. Serositis caused

Gursu et al. by tuberculosis usually results in fluid rich in lymphocytes, but a recent review found that more than 60% of cases showed neutrophil predominance (11). Removal of the peritoneal catheter is not mandatory in tuberculous peritonitis. Our patient was switched to HD because of drainage problems and dialysis inadequacy. In Turkey, the percentage of extrapulmonary tuberculosis in the nonuremic population in 2007 was 30.5% according to a report from the Ministry of Health (12). Extrapulmonary involvement in the dialysis population has been reported to be high both in Turkey (38% – 77%) and in other countries (up to 50%) (1,2,5,13–19), but most studies grouped PD patients and HD patients together. The rate of extrapulmonary involvement was 50% in the present study, which is similar to the result obtained in a Turkish study conducted by Erkoc et al. (5). In the dialysis population, a diagnosis of tuberculosis is often difficult because of the nonspecific symptoms: fever, fatigue, and weight loss. Skin tests are frequently anergic because of the blunted lymphocyte function and cytokine expression in uremia. In the present study, only 1 diagnosis was made through microbiology by culturing the bacilli in peritoneal effluent. Another diagnosis was made after observation of caseating granulomas during a pathology examination of lymph nodes. In 2 patients (50%), treatment was empirical, based on suspicious clinical and radiologic findings. For the treatment of tuberculosis in dialysis patients, it is important to select drugs with regard to their side effects. Streptomycin should be avoided because of increased ototoxicity. Ethambutol is associated with increased optic neuritis. Treatment should therefore use 4 drugs (including rifampicin, isoniazid, pyrazinamide or morphozinamid, and ofloxacin) for the first 3 months, with at least another 9 months of treatment with rifampicin and isoniazid. In 3 of our patients, the 4-drug treatment included ethambutol. Ethambutol was given, with close follow-up for potential visual side effects, because of a Ministry of Health policy and the difficulty of obtaining ofloxacin. Fortunately, the only side effect seen in the present study was elevated transaminase levels, recorded in 2 patients (50%). Both recovered when the drugs were temporarily stopped; the drugs were then reinstated with a gradually increasing dosage.

51 Conclusions The prevalence of tuberculosis is higher in PD patients than in the general population, and the disease is mostly extrapulmonary. Presentation may not be as obvious in dialysis patients as in a nonuremic population, and the diagnosis may be considered using approaches other than microbiology. Disclosures No financial support was received for the present study. References

1 Taskapan H, Utas C, Oymak FS, Gülmez I, Ozesmi M. The outcome of tuberculosis in patients on chronic hemodialysis. Clin Nephrol 2000;54:134–7. 2 Zahiri K, Ramdani B, Hachim K, Benghanem MG, Fatihi E, Zaid D. Tuberculosis in hemodialysis (French). Nephrologie 1997;18:303–6. 3 Kayabasi H, Sit D, Kadiroglu AK, Kara IH, Yilmaz ME. The prevalence and the characteristics of tuberculosis patients undergoing chronic dialysis treatment: experience of a dialysis center in southeast Turkey. Ren Fail 2008;30:513–19. 4 Cengiz K. Increased incidence of tuberculosis in patients undergoing hemodialysis. Nephron 1996;73:421–4. 5 Erkoc R, Dogan E, Sayarlioglu H, et al. Tuberculosis in dialysis patients, single centre experience from an endemic area. Int J Clin Pract 2004;58:1115–17. 6 Aydoğan O, Gürgün A, Başoğlu OK, et al. Tuberculin skin test reactivity in patients with chronic renal failure (Turkish). Tuberk Toraks 2009;57:268–76. 7 Canbakan B, Ergun I, Ekmekci Y, Ates K, Karatan O. Pulmonary and peritoneal tuberculosis in a CAPD patient. Int Urol Nephrol 2007;39:975–8. 8 Keane WF, Bailie GR, Boeschoten E, et al. on behalf of the International Society for Peritoneal Dialysis. Adult peritoneal dialysis–related peritonitis treatment recommendations: 2000 update. Perit Dial Int 2000;20:369–411. 9 Talwani R, Horvath JA. Tuberculous peritonitis in patients undergoing continuous ambulatory peritoneal dialysis: case report and review. Clin Infect Dis 2000;31:70–5. 10 Chau TN, Leung VK, Wong S, et al. Diagnostic challenges of tuberculosis peritonitis in patients with and without end-stage renal failure. Clin Infect Dis 2007;45:e141–6.

52 11 Akpolat T. Tuberculous peritonitis. Perit Dial Int 2009;29(suppl 2):166–9. 12 Ministry of Health of the Turkish Republic. 2007 Report of the fight with tuberculosis in Turkey (Turkish). Ankara, Turkey: Ministry of Health; 2007. [Available online at: www.verem.saglik.gov.tr/content/files/yayinlar/raporlar/turkiyede_verem_savasi_2007_raporu. pdf;cited June 30, 2011]. 13 Woeltje KF, Mathew A, Rothstein M, Seiler S, Fraser VJ. Tuberculosis infection and anergy in hemodialysis patients. Am J Kidney Dis 1998;31:848–52. 14 Smirnoff M, Patt C, Seckler B, Adler JJ. Tuberculin and anergy skin testing of patients receiving longterm hemodialysis. Chest 1998;113: 25–7. 15 Ekim M, Tümer N, Bakkaloğlu S. Tuberculosis in children undergoing continuous ambulatory peritoneal dialysis. Pediatr Nephrol 1999;13:577–9. 16 Sen N, Turunc T, Karatasli M, Sezer S, Demiroglu YZ, Oner Eyuboglu F. Tuberculosis in patients

Tuberculosis in a PD Population with end-stage renal disease undergoing dialysis in an endemic region of Turkey. Transplant Proc 2008;40:81–4. 17 Kürşat S, Ozgür B. Increased incidence of tuberculosis in chronic hemodialysis patients. Am J Nephrol 2001;21:490–3. 1 8 Kaul KL. molecular detection of mycobacterium tuberculosis: impact on patient care. Clin Chem 2001;47:1553–8. 19 Nagesh BS, Sehgal S, Jindal SK, Arora SK. Evaluation of polymerase chain reaction for detection of mycobacterium tuberculosis in pleural fluid. Chest 2001;119:1737–41.

Corresponding author: Meltem Gursu, md, Department of Nephrology, Haseki Training and Research Hospital, Istanbul, Turkey. E-mail: [email protected]