Pharmacological Interaction of Lopinavir/Ritonavir 800/200 mg BID and Rifampicin in Subjects Presenting Tuberculosis with Contraindication for an Efavirenz containing Antiretroviral Regimen

Despite the use of potent antiretroviral (ARV) regimens, the incidence of tuberculosis (TB) in HIV-infected people world wild remains high [1]. Because of poorer treatment outcomes when not used, rifamycin based regimens are the preferred choice for the treatment of TB, based on their proven efficacy, tolerability, and lower costs [2]. Boosted protease inhibitors are recommended if non-nucleoside reverse transcriptase inhibitors cannot be used. However, rifampicin – a potent cytochrome P450 3A4 isoform inducer – yields sub-therapeutic blood concentrations of boosted protease inhibitors (PI),with standard PI doses [3]. An alternative is to use an integrase inhibitor as part of the antiretroviral regimen while treating TB with rifamycin. A recent trial showed similar efficacy of raltegravir and efavirenz used concomitantly with rifampicin based regimens for TB [4]. However, in high burden TB countries the use of raltegravir is limited by its high cost, being unavailable in many TB programs up to now.


Introduction
Despite the use of potent antiretroviral (ARV) regimens, the incidence of tuberculosis (TB) in HIV-infected people world wild remains high [1]. Because of poorer treatment outcomes when not used, rifamycin based regimens are the preferred choice for the treatment of TB, based on their proven efficacy, tolerability, and lower costs [2]. Boosted protease inhibitors are recommended if non-nucleoside reverse transcriptase inhibitors cannot be used. However, rifampicin -a potent cytochrome P450 3A4 isoform inducer -yields sub-therapeutic blood concentrations of boosted protease inhibitors (PI),with standard PI doses [3]. An alternative is to use an integrase inhibitor as part of the antiretroviral regimen while treating TB with rifamycin. A recent trial showed similar efficacy of raltegravir and efavirenz used concomitantly with rifampicin based regimens for TB [4]. However, in high burden TB countries the use of raltegravir is limited by its high cost, being unavailable in many TB programs up to now.
The use of rifabutin has been recommended with PIs, because rifabutin is a less potent cytochrome inducer than rifampicin. However, it is metabolized by cytochrome P450 3A4 isoform, of which ritonavir (RTV) is a powerful inhibitor which can result in toxicities caused by increased rifabutin concentrations. The result of these interactions is an increase in rifabutin blood concentration. In order to avoid the potential rifabutin toxicity, the reduction of rifabutin dose is recommended when co-administered with boosted PIs. However, reducing the rifabutin dose poses the risk of acquired rifamycin resistance if ARV therapy is not adequately taken. Additionally, there are no fixed-dose combinations of TB medications incorporating rifabutin, potentiating thus the risk of drug resistance if taken separately, due to the pill burden. All the issues described above make rifampicin of greater interest in terms of public health. Also, rifabutin is not readily available in resource poor areas due to cost.
Lopinavir/ritonavir (LPV/r) are drugs with increased genetic barrier to resistance [5], which may be an excellent therapeutic option for TB-HIV patients who cannot be treated with nonnucleoside reverse transcriptase inhibitors. To overcome the effects of rifampicin hepatic induction, the standard dose of LPV/r could be doubled or extra ritonavir (super boosting) used. Some studies using these strategies, either LPV/R 400/400 mg or LPV/R 800/200 mg, have shown that there is achievement of adequate pharmacokinetic parameters. The first study, conducted by LaPorte in healthy volunteers, had comparable C min , C max , and AUC 12 of lopinavir for adjusted doses of LPV/R with rifampicin and standard doses of LPV/R without rifampicin; however, a significant number of cases of hepatotoxicity was described [6]. Later, HIV-infected people have shown better tolerance of these drugs combinations than HIV-uninfected individuals. Two studies conducted by Decloedt et al. one in HIV-infected [7], and the other in TB-HIV co-infected individuals [8], showed adequate pre-dose concentrations with much lower rates of hepatotoxicity than LaPorte. In the first study of Decloedt et al., LPV/r doses were escalated to twice the standard dose (800 mg/200 mg BID) and in their second study, as in LaPorte study, a group of patients used LPV/r 800 mg/200 mg BID and another group used LPV/r 400 mg/400 mg BID, both after escalation. However, these studies did not show the tolerance of HIV-infected patients to TB regimens before ARV therapy introduction. We conducted a prospective study to assess the pharmacokinetics of LPV/R800/200 mg in association with rifampicin-containing anti-tuberculosis regimens, in patients presenting tuberculosis that initiated TB treatment first, and later were started on antiretroviral therapy as recommended in Brazil. We also aimed to describe the adverse events observed during the tuberculosis treatment period with rifampicin, and the clinical, immunological and virological endpoints.

Method
This was a pharmacokinetic, descriptive, open-label, prospective, study, conducted at Tuberculosis Clinics of Instituto de Pesquisa Clínica Evandro Chagas, Rio de Janeiro. We enrolled HIV positive patients, 18 years or older, with tuberculosis, with any contraindications to use efavirenz or no genotypic resistance to LPV/ron the screening sample, who signed a written informed consent. Patients were excluded if they had resistance to rifampicin, hepatic enzymes higher than (ACTG) grade 2 (G2), hepatitis B and C or pregnancy. Tuberculosis diagnosis was based on signs and symptoms, consistent radiological abnormalities, and exclusion of other opportunistic diseases, even if the acid fast sputum smears were negative. TB was confirmed if Mycobacterium tuberculosis was identified in culture, or if after two months there was a favorable clinical response to tuberculosis treatment in the case of either negative cultures or contamination. Patients were followed up until 210 days after TB therapy initiation. A clinical resolution was considered a successful outcome o TB therapy, Study endpoints were defined as adequate plasma levels of rifampicin and LPV/R, a drop higher than 1 log10 in viral load (VL) after 90 days compared to baseline and viral suppression after 180 days, an increase in CD4 cell counts from baseline, treatment compliance measured by pill count at each visit and grade 3 and 4 adverse events. Genotyping analysis was done at D180 or at last visit if viral load was above 1,000 copies/mL. The study was approved by lPEC ethics review board. Antiretroviral treatment consisted of two nucleoside reverse transcriptase inhibitorsin combination or associated with a nucleotide analogue (tenofovir) and a combination of lopinavir-800 mg and ritonavir-200 mg (4 tablets of Kaletra™) BID, orally. Anti-TB medications were given in accordance with the current recommendations of the Brazilian Ministry of Health. This consisted of a 6 month regimen of rifampicin 600 mg and isoniazid 400 mg daily given in a fasting state with the addition of pyrazinamide for the first 2 months. Doses were adjusted for subjects who weighed less than 45 and 35 kg.
Patients started anti-TB treatment at D1, when all ARV drugs were discontinued. After a month, they were prescribed a LPV/r based ARV. In the first 3 days, they used two LPV/r gel tablets BID (400/100 mg); then it was escalated to three tablets BID (600/150 mg) and after three more days, to four tablets BID (800/200 mg). Escalation was done as a way to improve tolerance, and was already done before in other studies [6][7][8].
All formulations of LPV/r were supplied by Abbott Laboratories and rifampicin formulations were provided by the Brazilian Ministry of Health network.
The study procedures done on each visit are showed on Figure 1.

Screening
Questionnaire: sex, age, alcohol abuse, use of illicit drugs, weight loss, concomitant medication, date of first positive HIV serology, history of opportunistic diseases and comorbidities, current and previous antiretroviral (ARV) therapy regimens.
Medical history and physical exam.
Laboratory tests: microscopic examination of sputum; culture of sputum, blood and other clinical specimens for mycobacteria; serologic tests for viral hepatitis (B and C), syphilis and toxoplasmosis; blood cell counts; serum levels of urea, creatinine, liver enzymes, uric acid, and albumin; chest x-rays or CT scans; pregnancy test (female patients in reproductive age). Hematologic test and blood chemistry; pregnancy test in urine (in women of childbearing potential)

Day 1
Updated medical history and comorbidities, concomitant medication, and physical exam Antiretroviral drugs in use were discontinued, TB treatment was started.
Laboratory tests: blood cell counts; serum levels of urea, creatinine, uric acid, liver enzymes, and albumin; lipid profile; CD4 cell count and HIV viral load; genotyping; urine pregnancy test in women of childbearing potential).

D180 or last study visit
HIV genotype in case of VL>1,000.

D210
Lopinavir and Ritonavir PK Follow-up visits were scheduled every 15 days for the first two months (D60) after enrollment and then monthly until D210. TB medications were suspended on D180. Adverse events were graduated according to AIDS table for grading severity and adult adverse experience [9]. Treatment compliance was evaluated by tablet accountability and patient compliance history at every clinical evaluation.
Lopinavir and rifampicin were assayed in plasma using a LC-MS/ MS validated method. Chromatographic analysis was carried out on a Varian 1200 L quadrupole LC-MS/MS system equipped with an electrospray ion source, operated in the positive mode. The ion spray voltage and source temperature were 5850 V and 400°C, respectively. Chromatographic separation was achieved on a Pursuit C18 (Varian ® 5 μm, 100×2.0 mm i.d.) column at ambient temperature, using acetonitrile-water containing formic acid 0.1% (70/30, v/v) as mobile phase at a flow rate of 0.3 mL/min.
Prior to the chromatographic analysis, 50 μL of plasma samples containing IS solution (carbamazepine and clozapine 0.6 μg/mL) were deproteinized by the addition of 400 μL acetonitrile, vortexed for 1 min and centrifuged at 14400 rpm for 5 min. The supernatant was diluted (1:4) in acetonitrile/water containing formic acid 0.048% (70/30 -v/v) and 10 μLwas injected into the LC-MS/MS system. For the lopinavir analysis, after deproteinization, 100 μL methanol 50% (in water) was added to the supernatant and then was performed the dilution in acetonitrile/water containing formic acid 0.048% and LC-MS/MS analysis.
Quantification was performed by monitoring the decay of the massto-charge (m/z) ratio 629.70→447.30 for lopinavir, m/z 823.8→791.8 for rifampicin, m/z 237.0→193.7 for carbamazepine and m/z 327.0→270.0 for clozapine. Data acquisition and analysis were achieved using the Varian MS Workstation software (Version 6.6). The assay range for lopinavir and rifampicin was 0.5-15 µg/mL. Inter and intra-day coefficients of variation were below 15% for both drugs.
The pharmacokinetic parameters of lopinavir and rifampicin were determined by non-compartmental pharmacokinetic approach (Excel 2007, Microsoft ® ). The area under the plasma concentration-time curve until the last measurable concentration (AUC 0-12 ) was calculated using the linear trapezoidal rule. The values of peak concentration (C max ) and low concentration (C min ) of drugs in plasma were determined directly from the individual concentration-time data. Statistical analysis was performed using software R, version 2.14 (α=0.05).

Results
Fifteen HIV patients, already HAART experimented; consented to participate in the study which was conducted from September 9 th 2008 to April 19 th 2010. The group had a low median CD4 cell count (121 cells/mm 3 ) and median VL of 4.6 log10 copies/mL). Nine of the fifteen patients were using lopinavir based ARV regimens when TB diagnosis was made, of which three had undetectable viral load, and none of patients who had a VL>1000 copies/mm 3 had virologic resistance to lopinavir/r. TB diagnosis was made by positive culture in 7/15, and pulmonary TB was the most frequent diagnosed form. Alcohol abuse was reported by 5/15 and illicit drug use by 3/15. BMI was <18.5 in 7/15 patients at baseline ( Table 1).
The time line for the study procedures and dropouts during the study are shown in Figure 2. Five patients dropped out of the study during exclusively TB therapy. Five patients dropped out of the study during lopinavir therapy, 3 of them because of adverse events. Patients eligible to participate and the reasons for dropouts are shown in Figure  3. Adverse events were frequent during exclusively TB therapy (3 cases). Among them, hepatoxicity and flu like syndrome were recorded. Additionally, hepatitis C co-infection was diagnosed in one patient who was subsequently excluded from the study. G3 hepatotoxicity was a special concern during TB-HIV therapy (2 cases) and one case of    Table 2). Pharmacokinetics of LPV/r alone was performed in 4/5 patients at D210. Table 2 shows the CD4 counts and viral load for each study visit for those who completed 180 days of TB therapy. An improvement of CD4 cell counts from baseline to D60 was observed. However, from D60 to D180 all CD4 counts declined. Observing the favorable effect of lopinavir therapy on viral load, it was unclear why an increase in CD4 counts was not observed from baseline to D180 as expected. Only two out of five patients achieved undetectable viral load at D180.
Lopinavir genotypic resistance was detected in one patient at D60, when he dropped out from the study. This patient had baseline undetectable VL and had used 4 different ARV regimens before a LPV based regimen. No resistance to LPV/r was detected in the other patients who finished the study with viral load>1000. Resistance was not detected in low adherence patients in our study.
During the follow-up, no deaths related to drugs were observed. However, one patient, who dropped out because of a G3 hepatotoxicity, died later, due to an anaphylactic shock caused by the contrast injection for computed tomography scan. Hepatotoxicity was already controlled at that time.
The median steady-state plasma profiles of rifampicin determined during the study are presented in Figure 4, and the median pharmacokinetic parameters are showed in Table 3. Low concentrations of rifampicin were observed through all the samples. C max values of all patients, except two at D15, and four during ARV treatment, were below the reference range (8-24 µg/mL). Rifampicin AUC were low in the most of patients, but presented wide variations (4.57-110.80 μg.h/ mL) when administered alone or combined to ARV drugs. Although the lower and variable drug concentrations observed for rifampicin, the TB treatment was successful in all patients that completed TB therapy. No significant differences in pharmacokinetic parameters of rifampicin were observed along the study.
Lopinavir plasma median concentrations were stable during the study (Figure 4), as well as pharmacokinetic parameters (Table 3), and comparable to standard dose (400/100 mg BID) when used without rifampicin. Further at D60, three patients presented AUC above125 μg.h/mL -one of them had lower weight than 50 kg, and their LPV/r dose was reduced to 600/150 mg BID for safety reasons.
During the study, 4/10 patients had C min of LPV<1 μg/mL (C min for wild virus), but only one of them had another LPV low C min concentration. This patient received a prescription for LPV/r dose reduction to 600/150 mg BID on D45 due to AUC above the threshold limit (125 μg.h/mL). Nevertheless, 2/4 patients with low C min <1 mg/mL were considered adherent to treatment and had an undetectable viral load at the end of the study. The other two patients that presented C min of LPV insufficient to kill wild virus were discontinued from the study: one due toan adverse event at D45 and the other at D180 due to low adherence.
Lopinavir minimum concentrations were >4 µg/mL (target to HIV with mutations that confer resistance) in 6/10 patients, 4 of them in more than one point and 2 of them in all points of pharmacokinetic evaluation. One of these patients presented severe pancreatitis probably related to LPV and was prematurely discontinued from the study at D60; another patient, discontinued due to low adherence, had lipase grade 2 at D150. Neutropenia grade 1, nausea grade 2 and diarrhea grade 1 were detected in the patients that presented minimum concentration above 4 µg/mL.
Adherence was planned to be evaluated by pill count, however, most of the low adherent patients included in the study did not bring the drugs for counting. One study patient had to be hospitalized due to signs and symptoms of TB and AIDS. During hospitalization, the    patient received the study drugs and recovered. Our conclusion was that the patient did not take the medicines at home. Two other patients were discontinued during TB therapy because they did not come to the scheduled visits after many calls.

Discussion
Concomitant therapy for TB and HIV is still a challenge because of pill burden, drug-drug interaction and a consequent low adherence to both treatments. Few studies have been done to explore these issues. In countries where rifabutin is not available it is not clear how to introduce anti-TB drugs to HIV patients using PI with undetectable viral load. LPV/R is one of the few PI combination options to use with rifampicin and still needs a dose adjustment. Start a TB treatment and maintain antiretrovirals being used with a LPV/R double standard dose could lead to a higher frequency of adverse events and a need of both treatments interruption. In 2008 it was not known which the best moment to initiate antiretroviral therapy in HAART naïve patients, after TB treatment start, was. Then, antiretroviral interruption to start TB treatment and subsequent HAART reintroduction 30 days later with PI doses adjustment was the strategy considered to be used and adopted in this study. Until nowadays, there is no answer on how to proceed with HIV patients previously HAART experimented, especially those with efavirenz resistance, who start TB therapy with rifampicin as part of it. Detectable viral load in patients using LPV/R could occur because of no adherence, what was corroborated in the study since we did not found resistance in baseline genotyping, a reason for reintroducing LPV/R to these patients after TB treatment start.
Recently, a study was published by Decloedt and colleagues to evaluate concomitant therapy for TB-HIV patients with lopinavir 800/200 mg BID in South Africa [8]. The authors showed an overall good virologic control and compliance contrasting with our results, however, in their study; patients were enrolled after being established on tuberculosis treatment which excluded those who had early adverse events related to TB drugs. In our study, patients had a high incidence of adverse events to TB therapy, prior to the initiation of HAART. Compliance was also a limitation for both therapies as well as adverse events, even though the majority of the patients had already used LPV/r before TB diagnosis. In some cases adverse events, such as pancreatitis and hepatotoxicity related to study drugs, were barriers to adherence. Adverse events were controlled by the interruption of both therapies resulting in improvement in all cases. Unfortunately, one patient died after being discontinued due to hepatoxicity, during a Scan with contrast when the adverse event was almost resolved. We did not observe a correlation between adverse events and a higher LPV/r blood concentration.
The low rifampicin plasma levels observed in our study are consistent with results previously reported for HIV-infected patients in literature [10][11][12][13] and could be explained by the malabsorption observed in advanced HIV infection [10,14]. Although the low exposure to rifampicin was detected, it did not compromise patients' outcomes. One explanation would be post antibiotic effect of rifampicin [15]. TB treatment resulted in therapeutic success for all adherent patients.
Lopinavir plasma levels showed adequate concentrations of drugs for the majority of patients. Although lopinavir concentrations were below 1 µg/mL in 4/10 patients during the study, two of them, finished the study with undetectable viral load. Considering the target to resistant virus, 6/10 patients had minimum concentrations above 4 µg/mL during the study (5 of them were previously exposed to PI), but none of them showed LPV resistance. These results indicate that the increased dose of LPV/r (800/200 mg BID) when administered concomitantly to rifampicin was enough to achieve LPV therapeutic levels.
A study conducted at South Africa discussed the importance of adjusting lopinavir dose in patients with low weight (<50 Kg) to achieve an adequate serum concentration and to prevent overdose [16]. In our study, only one out of three patients who dropped out due to adverse events weighed<50 kg. Moreover, 3/10 patients taking lopinavir full dose (800 mg) showed an AUC over the target limit, without any signs of overdose. Therefore, it seems that low weight was not a parameter to adjust lopinavir dose in our study population.
The main limitation of the study was the small number of participants, with an elevated number of drop outs what had an impact in the analysis.
In conclusion, LPV/r 800/200 mg dose showed adequate concentration to treat a wild type virus, although low adherence was observed in several patients during TB and HIV therapies, concomitant or not. Adverse events and no compliance with the elevated number of pills for both treatments were limitations to finish the study. CD4 counts and viral load were improving just at the beginning of the therapy, not sustained through the end of the study. Although low adherence was remarkable, only one case of resistance to LPV/r occurred in the study. Other options for concomitant treatment of TB and HIV patients with resistance to efavirenz could facilitate adherence. Unfortunately, drugs