Can We Rationalize Oral Drug Exposure Following Bariatric Surgery to Meet the Pharmacotherapeutic Needs of a Growing Patient Population? Commentary on: “Lithium Toxicity Following Roux-en-Y Gastric Bypass”

Despite a dramatic increase in the incidence of bariatric surgery over the last decade, a significant gap still exists in the literature concerning the systemic exposure of drugs after oral administration for many commonly prescribed drugs in a patient population suffering from a variety of obesity-related comorbidities. ¹

The case report by Walsh and Volling ² (published in Bariatric Surgical Practice and Patient Care) provides an important addition to the literature, highlighting the impact of Roux-en-Y gastric bypass (RYGB) on oral drug exposure, and is the second published case to demonstrate lithium toxicity due to adverse lithium levels postoperatively.^2,3 This is a relevant case not only in terms of highlighting altered pharmacokinetics following bariatric surgery but also in terms of pharmacotherapy, where pharmacological treatment of depression and psychiatric disorders remains a relevant long-term indication following surgery.^4,5

Although the mechanism behind the observed lithium toxicity following RYGB is yet to be determined, one cannot overlook the role that altered gastrointestinal (GI) anatomy and physiology may play in altering drug dissolution, absorption, and intestinal first-pass metabolism. ⁶ The direction and extent to which drug exposure may be altered following bariatric surgery varies significantly between surgical procedures and depends on a given drug's pharmacokinetic properties. ⁴

In a recent publication, ⁴ we reported on the utilization of drug-specific properties in order to rationalize the trend in oral drug exposure pre to post bariatric surgery. ⁴ Originally developed to predict oral drug bioavailability, the Biopharmaceutics Classification System (BCS) was considered a strong candidate for predicting the direction of drug exposure postoperatively. The BCS classifies drugs into four classes based on their solubility and absorption properties ⁷ : class I: high solubility, high permeability; class II: low solubility, high permeability; class III: high solubility, low permeability; and class IV: low solubility, low permeability.

The cutoff point for permeability is defined as a 90% fraction of the orally administered dose absorbed. The cutoff point for solubility is defined as a dimensionless dose number (D_o), where a value above or equal to 1 is indicative of solubility limitations. D_o is a function of the maximum therapeutic dose (M_o [mg]), an assumed concomitant fluid intake of 250 mL in healthy volunteers (V_o [mL]), and the solubility of the drug (C_S [mg/mL]) over the physiological pH range (equation 1).^7,8 \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}D_o = \frac { M_o / V_o } { C_S } \tag { 1 } \end{align*} \end{document}

Simple algorithms such as the BCS may provide an indication of whether a drug will be subject to absorption or solubility limitations following bariatric surgery. This was illustrated for a number of reported drugs. Although no statistically significant difference could be determined between BCS classes I–IV with regard to the trend in oral drug exposure pre to post bariatric surgery, there was an indication that solubility-limited drugs (BCS class II and IV, D_o≥1) displayed a reduced exposure following surgery, whereas the outcome for highly soluble drugs (BCS class I and III, D_o <1) was much more variable. ⁴ It should be kept in mind that the BCS cannot quantify the magnitude of potential changes.

Lithium carbonate is considered highly permeable and highly soluble (C_s=13 mg/mL) at a therapeutic dose of 1,200 mg, with a D_o of 0.4. Due to the restrictive nature of RYGB, one can hypothesize that since the drug appears to dissolve completely prior to surgery, there should be no opportunity for more drug molecules to dissolve following surgery. As a consequence of gastric resection following RYGB (10–30 mL gastric pouch), a similar limitation in the volume of concomitant fluid intake can be assumed, resulting in an approximate D_o of around 3.1 post-RYGB.^9,10

With a D_o≥1 following surgery, lithium carbonate may potentially display solubility limitations postoperatively. However, one needs to consider the pH-dependent solubility behavior of lithium, as several studies have indicated a reduction in gastric acid secretion following RYGB.^11,12 At a higher pH, the carbonate salt is more likely to become deprotonated, resulting in more readily dissolved lithium ions available for absorption. This hypothesis is supported by Seaman et al. who employed a comprehensive in vitro dissolution study for numerous psychiatric medications through mimicking the gastric luminal conditions in controls and after RYGB, the results of which indicated a potential increase in lithium solubility following RYGB. ¹³ Further, the gastric emptying of liquids is significantly accelerated following RYGB, which may lead to earlier and higher systemic exposure. ¹⁴ It should also be kept in mind that in some hospitals, it is common practice to crush tablets before oral administration post-RYGB, which may further influence the dissolution of the drug in a positive direction following surgery.^4,13 The net effect on lithium dissolution may therefore be positive thus resulting in higher absorption leading to toxic lithium levels.

The alteration in oral drug exposure is not only a concern for lithium carbonate. Clinical studies have reported altered exposure for a number of drugs, including several antidepressants such as sertraline, citalopram, venlafaxine, and duloxetine. In addition, there is a potential for altered dissolution behavior for a number of drugs used in the treatment of bipolar disorders, such as quetiapine, olanzapine, and risperidone. Contrary to lithium carbonate, all of the above examples display indications of lower solubility and/or exposure following RYGB. As was illustrated using the example of lithium carbonate, the BCS may provide an indication of the absorption behavior following surgery. Assessing the D_o and BCS classification of commonly prescribed medications for bipolar disorder identified several drugs that may potentially be subject to altered dissolution following RYGB (D_o≥1) and which could result in subtherapeutic oral exposure, albeit clinical data to confirm this are missing for most of the drugs (Table 1). Thus, there is reason for great concern for both the safety and the efficacy of pharmacological treatment of depression and bipolar disorders post bariatric surgery.^13,15–17

OPEN IN VIEWER

Table 1.

Biopharmaceutics Classification System Properties for Most Common Medical Therapies for Bipolar Disorder

Drug	Maximum dose strength (mg)	Solubility (mg/mL)	Do,HV	Do,RYGB	BCS class
Mood stabilizers
Carbamazepine	300	0.256	4.7	39.1	II
Divalproex sodium	500	36	0.06	0.5	III
Lamotrigine	200	0.17	4.7	39.2	II
Lithium carbonate	600	13	0.2	1.5	I
Antipsychotics
Aripiprazole	30	0.0001	1,200	10,000	II
Lurasidone hydrochloride	120	0.349	1.4	11.5	IV
Olanzapine	40	0.01	16	133.3	II
Quetiapine fumarate	300	94	0.01	0.1	I
Risperidone	4	0.25	0.06	0.5	I

D_o,HV, assuming a concomitant fluid volume of 250 mL; D_o,RYGB, only considering altered concomitant fluid volume of 30 mL; BCS, Biopharmaceutics Classification System.

References: ^{7,9,10,18–25}

The BCS does, however, not provide a sufficient framework to predict oral drug exposure, thus pointing to the necessity for more complex models to account for the relation between GI physiology and drug-specific properties in the prediction of postoperative exposure. In order to consider the complex interplay between drug-specific factors (e.g., pH-dependent solubility, permeability, metabolism, etc.) and GI physiology (e.g., gastric resection, intestinal bypass, and bile diversion), one should turn to pharmacokinetic computer-based modeling and simulation approaches that can account for such factors, that is, physiologically based pharmacokinetics (PBPK). In the absence of clinical data, the PBPK approach may provide more reliable predictions of trends in oral drug exposure pre to post bariatric surgery and rationale as to why these differences occur.^26,27

Efforts to model oral drug bioavailability following RYGB using PBPK models have proven successful in predicting both the direction and extent of the altered oral drug exposure following bariatric surgery.^26,27 These models have allowed the elucidation of the potential mechanisms governing the change in drug exposure postoperatively, of which solubility, permeability, and gut-wall metabolism were the main implicating factors. The usage of PBPK modeling and simulation has great potential in assisting dosage optimization and establishing pharmacotherapeutic guidance for patients subject to bariatric surgery where national clinical guidelines are absent. ⁴ One important obstacle in implementing PBPK modeling and simulation in clinical practice is the general lack of user-friendly tools, since these tend to be geared more toward drug development than clinical pharmacotherapy.

Interdisciplinary efforts between surgeons, clinicians, and other healthcare professionals, and researchers in pharmaceutical and medical sciences are warranted in order to meet the pharmacotherapeutical needs of a growing post bariatric surgery patient population. To ensure that post-RYGB patients receive appropriate long-term care, it is of great importance that general practitioners be informed of the potential challenges that bariatric surgery may pose in terms pharmacotherapy.