Low Third-Trimester Serum Levels of Lamivudine/Zidovudine and Lopinavir/Ritonavir in an HIV-Infected Pregnant Woman with Gastric Bypass

Abstract

Pharmacokinetics of lamivudine (3TC)/zidovudine (ZDV) and lopinavir/ritonavir (LPV/r) are described in a gravid 27-year-old HIV-infected woman with gastric bypass. Blood levels were obtained for these medications at time points 0 (predose) and 1, 2, 4, 6, 8, and 12 hours postdose. For these times, the levels (µg/mL) of 3TC were 0.0801, 0.69, 0.339, 0.237, 0.202, 0.108, and 0.0461; the levels of ZDV were 0.0153, 0.433, 0.0717, 0.0481, 0.0107, 0.0214, and 0.00864; the levels of lopinavir (LPV) were 2.45, 2.64, 1.95, 2.78, 3.83, 3.20, and 1.92; and the levels of ritonavir (RTV) were 0.09, 0.10, 0.07, 0.11, 0.15, 0.15, and 0.06. These data suggest that gastric bypass affected these antiretroviral drug levels. A functional, intact small bowel is responsible for absorption of these medications.

Introduction

The United States has seen a significant increase in the number of patients receiving bariatric surgery. A national inpatient sample from 2003 to 2008 reported up to 135 985 cases in 2008. ¹ Among these cases, the Roux-en-Y procedure was the most popular. In the state of California, however, the rate of this procedure has declined in adolescents (from 3.8 to 2.7 per 100 000 population) in favor of the adjustable gastric band (6.9-fold increase). ² Given the high number of individuals who have undergone Roux-en-Y gastric bypass, it is anticipated that the absorption of certain drugs may be impaired. This poses a particular challenge in HIV-infected individuals, especially in HIV-positive women during pregnancy, a time when viral suppression is even more critical to minimize the risk of vertical transmission of HIV.

Medical literature examining the effects of bypass surgery on the pharmacokinetics (PKs) of antiretroviral (ARV) drugs is limited primarily to case reports. Specific cases include a 28-year-old HIV-infected man who underwent gastric bypass (the duodenum was bypassed 20 cm distal to the ligament of Treitz) due to duodenal malignant lymphoma showed successful absorption of lopinavir (LPV) ³ ; a 38-year-old HIV-infected woman who underwent a total gastrectomy with esophagojejunostomy and Roux-en-Y reconstruction due to Hodgkin lymphoma with LPV trough levels which adequately suppressed the viral load to a nearly undetectable level ⁴ ; 2 unrelated patients who underwent Roux-en-Y bypass with no mention of the surgical effect on suppression of ARV PK ⁵ ; and an HIV-infected woman (no age stated) on zidovudine (ZDV), lamivudine (3TC), and nelfinavir (NFV) who underwent Roux-en-Y bypass (150 cm jejunum bypassed), with lower serum concentrations of NFV but no effect on ZDV and 3TC levels. ⁶

Although ARV PK in women who have previously undergone gastric bypass may be problematic, by adding pregnancy as a confounder, PK interpretation becomes even more complex. Several physiologic changes can occur during pregnancy. These include prolonged gastrointestinal transit time; increase in body fat and water; and changes in cardiac, circulatory, hepatic, and renal function. Any of these factors may affect the absorption, distribution, and elimination of drugs. Findings from International Maternal Pediatric Adolescent AIDS Clinical Trial (IMPAACT) Study P1026s showed lower levels of LPV during the third trimester (400/100 mg capsule twice daily [BID]), ⁷ leading to the current recommendation for an increased dose in the second and third trimesters to 600/150 mg tablet BID. ⁸ Our case is the first to describe the PK of specific ARV drugs in an HIV-infected pregnant woman with gastric bypass. This case was submitted to the local institutional review board and deemed exempt from review.

Patient History

The patient was a gravid 27-year-old woman at 33-2/7 weeks of gestation during the time of the evaluation. Her Roux-en-Y bypass with cholecystectomy was performed at 18 years of age. The bypass was 30 cm distal to ligament of Treitz where the biliary limb was plugged into jejunum at 400 cm, leaving 140 cm of small bowel for absorption.

She was diagnosed with HIV at 9 weeks of gestation during her initial obstetric appointment and self-reported a negative HIV test 4 months prior. The viral phenotype testing reported a pan-sensitive HIV-1 with the genotype showing 2 notable protease inhibitor mutations (E35D and M36I). Viral replication capacity was 30% (range 19%-48%). The initial viral load at diagnosis was 5474 copies/mL, which reached an undetectable level within 2 months of the starting antiretroviral therapy (ART). Her viral load remained undetectable throughout her pregnancy. Her ARV regimen included 3TC /ZDV 150/300 mg 1 tablet BID and lopinavir/ritonavir (LPV/r) 200/50 mg 3 tablets BID. She had been fully compliant on this regimen since 12 weeks of gestation with an appropriate increase in the LPV/r dose at the beginning of the third trimester. Her other medications included iron sulfate 325 mg once daily (QD) for iron deficiency and 1 prenatal vitamin QD. She has a history of hypothyroidism but recent thyroid function studies were normal.

She normally eats several meals a day. On the morning of the PK evaluation, she consumed a chicken sandwich, egg roll, and juice. Her height was 162.5 cm (5 ft 4 in) and weight 118.4 kg (260 lbs), for a body mass index (BMI) of 44.8. Her vital signs were normal and examination was notable only for obesity and a gravid pelvis.

Results

On the day of her scheduled evaluation, serial blood draws were performed at hour 0 (predose) followed by 1, 2, 4, 6, 8, and 12 hours post-morning dose of 3TC /ZDV and LPV/r. Samples were collected and immediately spun at 800× gravity for 10 minutes, aliquoted into 1 mL vials, and frozen at −70°C prior to shipping to Consolidated Laboratory Services (Van Nuys, California). This laboratory uses a solid phase extraction procedure followed by high-performance liquid chromatography and detection with a liquid chromatography–tandem mass spectrometry system.

Drug level results (µg/mL) for time points 0, 1, 2, 4, 6, 8, and 12 hours were as follows: 3TC 0.0801, 0.69, 0.339, 0.237, 0.202, 0.108, and 0.0461; ZDV 0.0153, 0.433, 0.0717, 0.0481, 0.0107, 0.0214, and 0.00864; LPV 2.45, 2.64, 1.95, 2.78, 3.83, 3.20, and 1.92; and ritonavir (RTV) 0.09, 0.10, 0.07, 0.11, 0.15, 0.15, and 0.06. These drug levels were compared to previously published data and are shown in Figure 1A and B. ^9,10 Other standard laboratory tests included a CD4 count of 50%/588 cells/mm³, HIV viral load <20 copies/mL, creatinine 0.44 mg/dL, aspartate aminotransferase 18 units/L, and alanine aminotransferase 16 units/L.

Figure 1.

Serum levels for (A) lopinavir (LPV) and (B) ritonavir (RTV) in our patient versus published HIV-infected pregnant controls. ⁹

Figure 2.

Serum levels for (A) lamivudine (3TC) and (B) zidovudine (ZDV) in our patient versus published HIV-infected pregnant controls. ¹⁰

Discussion

The PK of ARV drugs in HIV-infected pregnant women is a topic of ongoing research. Previous studies have demonstrated that pregnancy-induced physiologic changes can affect the metabolism of certain ARV drugs in HIV-infected women. ^7,11,12 Early PK studies of ZDV and 3TC did not demonstrate a difference between pregnant and nonpregnant HIV-infected adults. However, data from IMPAACT Study P1026s did show insufficient LPV/r concentrations during the second and third trimesters in HIV-infected pregnant women when compared to nonpregnant women. ⁹ Consequently, an increased dose of LPV/r capsules is recommended during the second and third trimesters. The reason for decreased levels is likely multifactoral. Certain cytochromes such as P450 34A are more abundant in the proximal small intestine and may be expressed at higher levels which metabolize drugs more quickly. Additionally, intestinal p-glycoproteins are elevated post-gastric bypass surgery, which may decrease gut absorption.

Nonetheless, the addition of gastric bypass to an already altered physiology seen in pregnancy could certainly create a valid concern for virologic breakthrough. In the typical adult, the duodenum is approximately 25 cm long and 5 cm wide. ¹³ As most medications are primarily absorbed in the small intestine, Roux-en-Y bypass alters mechanisms necessary for optimal absorption. Factors specific to the duodenum and jejunum include impairments in bile salt solubilization, enterohepatic recirculation, intestinal adaptation, and decreased mucosal exposure. Furthermore, gastric changes that may also alter absorption of some drugs include smaller stomach volume leading to reduced gastric acid production, higher gastric pH, and delayed gastric emptying. ¹⁴ Finally, the patient described in this report had an elevated BMI which may also be a factor that influences the PKs of her ARV regimen.

Results from our patient’s PK data demonstrate considerably lower serum levels of 3TC, ZDV, and LPV/r when compared to HIV-infected women who have not undergone gastric bypass. We would expect to find similar results in other HIV-infected individuals on ART who have either underwent Roux-en-Y gastric bypass or other conditions that impair intestinal function. Despite low levels, our patient maintained an undetectable viral load throughout her pregnancy, and her newborn infant remains HIV negative. The medical literature does not support the theory that low ARV drug levels as seen in our patient were sufficient to induce viral suppression. ¹⁵ We surmise that she may be an elite controller or her virus may be less virologically “fit,” given her low replicative capacity. Thus, virologic suppression in our patient may be an exception rather than what would otherwise be expected in the typical patient. Despite low ARV drug levels, the persistence of an undetectable viral load allowed us to maintain standard recommended ARV dosing.

We acknowledge several limitations with this analysis. First, the PK analysis was done only once for a single patient. In addition, there may be individual patient variables such as thyroid and gallbladder function, diet, gastric motility, gastric pH, pharmacogenomics as well as other unknown factors that were not accounted for. Finally, ARV levels <0.25 µg/mL are deemed unreliable per the commercial laboratory assay specifications.

Given the increasing number of patients who undergo gastric bypass, further PK studies of ARV medications in these individuals are warranted, especially in pregnant women where the risk of perinatal transmission is of significant concern. We suggest that patients who do not achieve virologic suppression may require an adjustment in the ARV regimen, which may include a dose change, a possible switch, and/or adding additional ARV medications.

Conclusion

There was a notable reduction in the serum levels of 3TC, ZDV, and LPV/r in this pregnant HIV-positive patient with Roux-en-Y gastric bypass. This highlights the effect of a nonintact gastrointestinal system on ARV PKs. It is possible that viral suppression may still be achieved despite low serum levels of ARV medication, however we advise close monitoring of the viral load.