Short note: Infliximab recovery in a simulated intestinal fluid of the upper intestine tract

Abstract

BACKGROUND:

The oral administration of Infliximab (IFX) antibody would ensure a direct action on inflamed intestinal tissues without side effects. Thus, investigations about its resilience within the intestinal environment are required.

OBJECTIVE:

Quantify the IFX recovery in a simulated upper intestinal environment.

METHODS:

IFX was incubated for different times until 120 min in simulated intestinal fluid (SIF) which differed (i) for pH (7.2 vs 6.8, Exp 1), (ii) for addition or not with pancreatin (Exp 2) and (iii) for addition or not with bovine serum albumin in presence of pancreatin (BSA, Exp 3).

RESULTS:

In Exp 1 the IFX incubated without pancreatin was degraded by about 15% by SIF pH change from 7.2 to 6.8 and after 120 min it was reduced by about 20%. In Exp 2 the presence of pancreatin determined an intense and rapid IFX degradation (recovery $<$ 33%, within 30 min), but when BSA was added to simulate the presence of food protein (Exp 3) the IFX half-life ranged between 59 and 70 min.

CONCLUSIONS:

A discrete in vitro stability of IFX in the upper intestine environment was demonstrated, if food protein is available and competes with pancreatin proteases.

Keywords

Infliximab intestine digestion

1. Introduction

The pro-inflammatory cytokine tumor necrosis factor alpha (TNF $\alpha$ ) plays a relevant role in inflammatory bowel disease (IBD). Antibodies with specific TNF $\alpha$ binding/neutralizing activities, such as the chimeric human-mouse monoclonal Infliximab (IFX), are proposed as therapeutics and currently administered intravenously in order to be delivered systemically to the gastrointestinal mucosa of individuals affected by Chron’s disease [17, 3]. The parenteral therapy has relevant side effects [8, 16], whereas the oral administration of an anti-TNF $\alpha$ antibody would ensure a direct action on the target intestinal tissues [6, 4]. Obviously, the antibodies should be resistant to the digestive environment, or protected, in order to reach their target areas with intact biological properties. Some antibodies can resist in the digestive environment, such as colostrum immunoglobulins [6, 19], or novel engineered antibodies [14], but IFX is intensively degraded in the stomach [18]. Based on this evidence, its oral administration to patients might include the protection within gastro resistant capsules or by means of novel delivery drug systems [9] allowing the release in the upper intestinal tract. It seems therefore crucial to investigate further the resilience of the antibody within this specific environment.

Pigs are often used as animal models to study the physiology of the human digestive tract [15, 10, 7]. In the present study we applied an in vitro digestion approach developed for pigs and already validated in vivo [12], aiming at the quantification of IFX recovery in the upper intestine tract. The IFX resistance in a biologically active form was assessed at progressive incubation times in simulated upper intestinal fluid (SIF) under different conditions (e.g. different pH, absence or presence of pancreatin, different IFX concentration, absence or presence of a food protein).

2. Material and methods

2.1 Therapeutic antibody and experimental design

The IFX (Inflectra ${}^{\rm TM}$ ) was provided by the University Hospital of Udine (Ospedale Santa Maria della Misericordia, Italy) in single vials containing 100 mg of lyophilized powder. The IFX was dissolved in buffer solution [12] at the concentrations of 10, 20 and 40 $\mu$ g/mL.

The research consisted of three experiments: IFX was incubated up to 120 min in SIF at different pH (7.2 vs 6.8, Exp 1); in SIF added or not with pancreatin (Exp 2); in SIF added with pancreatin and in presence or absence of BSA (Exp 3). In Exp 1 the 20 $\mu$ g/mL IFX solution was used, while in Exps 2 and 3 all three concentrations were tested.

2.2 In vitro simulation of the upper intestinal environment

The in vitro digestion was carried accordingly with the method of Boisen and Fernández [12], with some modifications. The simulated gastric fluid was prepared into 125-mL conical flasks where 25 mL of 0.1 M sodium phosphate buffer solution and 10 mL of 0.2 M HCl (pH adjusted to 2.0 by using 1 M HCl or NaOH) were dispensed, without the inclusion of pepsin solution. Flasks were closed with a silicon stopper and were kept into a thermostatically-controlled heating bath to maintain the temperature of 39 ${}^{\circ}$ C for 1 h. Then 10 mL of 0.2 M sodium phosphate buffer and 5 mL of 0.6 M NaOH were added to the flasks to turn the gastric fluid into the SIF. In Exp 1 the pH was adjusted to 6.8 or to 7.2 using 1 M HCl or NaOH and 1 mL of IFX solution (20 $\mu$ g/mL) was added.

In Exps 2 and 3 the pH was adjusted to 6.8 using 1 M HCl or NaOH and 1 mL of freshly prepared pancreatin solution (50 mg/mL, derived from porcine pancreas, 4 $\times$ USP, P1750, Sigma-Aldrich, St. Louis, MO, USA) was added together with 1 mL of IFX solution (10, 20 or 40 $\mu$ g/mL). In Exp 3 100 mg of bovine serum albumin (BSA, A-9647, Sigma-Aldrich, St. Louis, MO, USA) were also added. Flasks were closed with a silicon stopper and incubated for 120 min in a heating bath at 39 ${}^{\circ}$ C. In all the experiments, three flasks were used for each treatment.

To measure the recovery of biologically active IFX by E.L.I.S.A., samples of the SIF (0.35 mL) were collected from each flask immediately before closing and at 30, 60, 90 and 120 minutes of incubation.

Figure 1.

Example of standard curve obtained by E.L.I.S.A. of serial dilutions of IFX. The value of $R^{2}$ was equal to or greater than 0.998.

2.3 E.L.I.S.A. based titration of IFX in the simulated intestinal fluids

The IFX concentration of SIF samples was evaluated by indirect E.L.I.S.A accordingly with the method of Boscolo et al. [13], partially modified by the authors. Ninety-six well plates (Costar 3590) were coated with 100 ng/100 ul per well tumor necrosis factor alpha (TNF $\alpha$ , PHC3011, Life Technologies Italia) and incubated overnight at 4 ${}^{\circ}$ C. Then the aspecific binding sites were blocked with 2% non-fat milk in Phosphate Buffer Saline pH 7.2 (PBS) (200 ul/well) for 1 hour at 37 ${}^{\circ}$ C. The plates were washed three times with PBS plus 0.1% Tween-20 and 100 ul/well of the test samples were added to the plates, then incubated for 1 h at 37 ${}^{\circ}$ C. The plates were washed three times with PBS pH 7.2. A peroxidase conjugated goat antibody specific for human IgG (A070, Sigma-Aldrich) diluted 1: 5000 was added (100 ul/well) and incubated for 1 h at 37 ${}^{\circ}$ C. The plates were washed with PBS and the reaction revealed with 3’3’5’5’-tetramethyl-benzidine dihydrochloride (TMB) (100 ul/well). The color development in wells was stopped after 3 minutes by the addition of 50 ul/well of 2 M H ${}_{2}$ SO ${}_{4}$ and was read by using a Tecan multi spectrophotometer at 450 nm. Each microplate included blank wells in which SIF samples were replaced by Phosphate Buffer Saline. Infliximab concentrations were calculated by referring to repeatable calibration curves included in each plate and obtained using a serial two-fold dilution of IFX in Phosphate Buffer Saline from 1.25 ug/ml to 0.039 ug/ml as that shown in Fig. 1.

2.4 Statistical analysis

The IFX amount in the SIF were statistically analysed as factorial designs with repeated measures according to the following model:

$\displaystyle Y_{ijk}=\mu+\alpha_{i}+\beta_{(i)j}+\gamma_{k}+\alpha\gamma_{ik}% +\varepsilon_{ijk,}$

where $y$ is the measure of the experimental unit (single analytical IFX determination from samples collected in each flask), $\mu$ is the overall mean; in Exp. 1 $\alpha_{i}$ is the fixed effect of the SIF pH ( $i=$ 1, 2), while in Exp 2 and 3 it is the fixed effect of the IFX dose ( $i=$ 1, 3); $\beta_{j}$ is the random effect of flasks ( $j=$ 1, 3) for each SIF pH in Exp 1 and for each IFX dose in Exp 2 and 3; $\gamma_{k}$ is the fixed effect of sampling time (0, 30, 60, 90 and 120 min, $k=$ 1, 5) and $\varepsilon$ is the residual error.

In Exp 3, the IFX amounts of each flask within each IFX dose were regressed on sampling time according to the following linear model:

$\displaystyle Y_{i}=\alpha+\beta x_{i}+e_{i}$

where $\alpha$ is the intercept, $\beta$ is the slope and $x_{i}$ is the $i^{\rm th}$ sampling time. The half-life of IFX was calculated from regression parameters as follows: $(\alpha-(\alpha/2))/-\beta$ .

The regression parameters and half-life values were were statistically analysed with the following factorial model:

$\displaystyle Y_{ij}=\mu+\alpha_{i}+\varepsilon_{ij}$

where $\mu$ is the overall mean and $\alpha_{i}$ is the fixed effect of the IFX dose ( $i=$ 1, 3).

3. Results and discussion

Table 1

Recovery (ug) of IFX (20 ug added) throughout the incubation in the simulated intestinal fluid calibrated at two different pH around neutrality (6.8 and 7.2) without pancreatin (Exp 1)

	pH of the intestinal fluid			RMSE ${}^{1}$
	6.8	7.2	means
Incubation time (in min)
0	18.0	20.1	19.1a	2.19
30	18.8	20.2	19.0a
60	13.3	17.7	14.6b
90	15.2	17.4	15.6bc
120	17.0	15.6	16.9c
means	15.9b	18.2a

Means with different letters are statistically different ( $P<$ 0.05). ${}^{1}$ Residual mean square error.

A rapid and complete degradation of IFX in the stomach environment has been already demonstrated in vitro [18], whereas the resistance of the antibody in the upper intestine is still not clearly stated. In the work of Yadav et al. [18], the recovery in the intestinal environment was lower than that in the cecum-colon tract (approx. 19 vs more than 90 min of half-life), but the in vitro study adopted a fixed ratio between intestinal enzymes/IFX antibody and did not considered the addition of a protein substrate. Also, another study [11] measured a low in vitro stability of IFX (around 10%), but at high concentration of porcine pancreatin (5–10 mg/mL), after a long incubation time (4 h) and without the addition to food proteins. Therefore, we focused our interest in the upper intestinal tract with an experiment aimed at testing increasing levels of IFX and at evaluating the effect of the presence of food protein.

The IFX dosages adopted in the experiments of the present study (10–40 ug/incubation flask) corresponded to an oral administration ranging from 4–5 to 16–20 mg for a person having an average intake of 400–450 g/d of dry matter, because the in vitro technique is established for an amount of 1000 mg of substrate. These dosages are consistent with those suggested by Maurer et al. [5], which range between 10 and 20 mg/d, in the light of an evaluation considering doses adopted for local injections.

Table 2

Recovery (ug) and half-life of IFX added at different doses in the simulated intestinal fluid with pancreatin, not added (Exp 2) or added (Exp 3) with BSA

	10	20	40	D	T	RMSE ${}^{2}$
	IFX doses (ug)			$P^{1}$
BSA not added (Exp 2)
Incubation time (in min)
0	11.5	16.7	33.4	**	**	0.38
30	2.9	4.3	10.0
60	0.6	0.2	5.1
90	0	0	4.5
120	0	0	3.6
BSA added (Exp 3)
Incubation time (in min)
0	11.9	20.4	36.4	**	**	0.56
30	7.2	15.8	29.7
60	3.8	10.2	14.9
90	2.1	6.1	13.5
120	1.0	3.9	5.0
IFX linear degradation (Exp 3)
– Intercept (ug)	10.58	19.83	35.70	**	–	0.7588
– Rate of degradation (ug/min)	$-$ 0.090	$-$ 0.142	$-$ 0.263	**	–	0.0101
$R^{2}$	0.892	0.975	0.940	–	–	–
Calculated half-life ${}^{3}$ (min)	59.2	69.7	67.9	NS	–	4.74

${}^{1}$ Statistical significance (** $P<$ 0.01; NS: not significant) of factors in the model (D: IFX dose; T: sampling time). ${}^{2}$ Residual mean square error. ${}^{3}$ Calculated from linear equation parameters.

In Exp 1, the variation of pH of the SIF without pancreatin from 7.2 to 6.8 determined a reduction in the IFX recovery from 18.2 to 15.9 ug ( $P<$ 0.05), while the permanence of IFX in the SIF for 120 min caused a reduction of about 20% (from 19.1 to 15.6 ug, $P<$ 0.01, Table 1). The results of the first experiment highlight a moderate in vitro degradation (15–20%) of the antibody due to factors independent by the presence of pancreatin (permanence in SIF and pH of the SIF). Firstly, the antibody integrity and/or ability to bind TNF $\alpha$ seems to be affected by the simple permanence in the SIF at 39 ${}^{\circ}$ C without enzymes. In addition, the sub acid conditions of SIF interfere with the IFX molecule and a pH value of 7.2 is more favourable for the preservation of the IFX biological activity. In vitro digestion methods for humans [7] utilise a neutral condition (pH $=$ 7.0) for the intestinal step, while it is assumed that the upper intestinal pH can reach a value of 7.5 at the distal ileum. Therefore, we decided to perform the in vitro tests following the pH suggested by the original method [12] but then we applied a correction ( $+$ 15%) to the recovered amount of antibody in order to consider the actual SIF pH. The neutro-alkaline conditions favourable to the maintenance of the IFX stability in the upper intestine would suggest associating also buffer substances to oral IFX based treatments.

The effect of three dosages of IFX incubated for progressive times in the SIF containing pancreatin or pancreatin plus BSA is reported in Table 2. In both Exps 2 and 3 the interaction “time $\times$ dose” was not significant, whereas there a was a significant effect of factors “time” and “dose” ( $P<$ 0.01).

Figure 2.

Percentage of IFX recovery throughout the incubation in SIF in absence or in presence of bovine serum albumin (BSA, (a) without BSA; (b) with 100 mg of BSA/flask).

In the last experiment we added to the incubation flasks of the in vitro system also BSA to simulate the food protein. In fact, it might represent a suitable target substrate for the pancreatic enzymes, alternative to the administered IFX. The amount of BSA added was calculated on the base of the human food requirements: assuming a protein daily intake of 50 g for an adult person [1] it could correspond (at maintenance) approximately to the 10% of dry matter intake. We added an amount of 100 mg of BSA, which represents the 10% of the amount of protein, contained in 1000 g of the substrate in the [12] method. The results of the present research indicate a relevant role ascribable to food proteins on the IFX recovery because in absence of BSA the IFX recovery was very low within 30 min of incubation (less or equal than 1/3) while the IFX remained biologically active for longer time in the case of BSA addition (Table 2 and Fig. 2a and b). This is an expected consequence of the competition between added protein and IFX as target substrates for pancreatic enzymes and suggests that a possible strategy to increase the IFX stability is to include sacrificial proteins to the drug [2] aiming at neutralising in part the pancreatic enzymes. The coefficients of the linear regressions within each IFX dose in the Exp 3 (Table 2) differed according to the antibody dose ( $P<$ 0.01), while the calculated IFX half-life ranged from 59 to 70 min, but no statistically significant difference was found between antibody doses. A graphical representation of the IFX recovery during the 120 min of incubation is reported in Fig. 2a and b.

In the conditions of presence of food protein and considering the loss of accuracy of E.L.I.S.A. when used to assess the IFX recovery at a 6.8 pH we observed a regular decline of IFX concentration, and the half-life was reached at about 59–70 min. This degree of resilience is set in between the low resistance at the intestinal level (approx. 19 min) and the value considered of great stability of 90 min in the colon, found by [18]. However, caution is requested in such comparisons because the two studies used different analytical methods for IFX recovery titration, which were based respectively on chromatographic determination (e.g. HPLC in Yadav et al. [18]) and on immune enzymatic detection (e.g. TNF binding E.L.I.S.A.).

Finally, our results did not demonstrate any effect of the IFX dose on the degree of recovery, because the antibody half-life, calculated from the linear regressions within each IFX dose, was very similar (ranging between 59 and 70 min) and not statistically significant between doses. This is probably the consequence of an excess of proteolytic activity in the SIF proposed in this study, irrespective to IFX concentrations. In such conditions, the IFX resilience is not dependent by the pancreatin concentration, but rather from the intrinsic structural resistance of the antibody.

In conclusion, we demonstrated a discrete stability of IFX in the upper intestine environment (half-life range 59–70 min) and therefore, IFX could be considered for the oral treatment by delivering it at the duodenum. A dietetic strategy to further increase IFX stability should combine its administration with buffer substances (to improve the intestinal alkaline conditions) and food proteins to enhance the competition with pancreatic enzymes for the degradation. Therefore, the oral treatment should be associated with suitable dietary prescriptions.

Footnotes

Acknowledgments

This work was supported by the Region Friuli Venezia Giulia (POS-FESR 2014–2020, attività 1.3.b.).

Conflict of interest

All authors have nothing to disclose.

References

BNF, British Nutrition Foundation, Nutrient requirements. www.nutrition.org.uk, 2017.

Moroz

Matooris

and Leroux

J.C.

, Oral delivery of macromoleculars drugs: Where we are after almost 100 years of attempts, Advanced Drug Delivery Reviews 101 (2016), 108–121.

Ordás

Feagan

B.G.

and Sandborn

W.J.

, Early use of immunosuppressives or TNF antagonists for the treatment of Crohn’s disease: Time for a change, Gut 60 (2011), 1754–1763.

Crowe

J.S.

Roberts

K.J.

Carlton

T.M.

Maggiore

Cubitt

M.F.

Clare

Harcourt

Reckless

MacDonald

T.T.

Ray

K.P.

Vossenkämper

and West

M.R.

, Preclinical development of a novel orally-administered anti-tumour necrosis factor domain antibody for the treatment of inflammatory bowel disease, Scientific Reports 8 (2018), 4941.

Maurer

J.M.

Hofman

Schellekens

R.C.A.

Tonnis

W.F.

Dubois

A.O.T.

Woerdenbag

H.J.

Hinrichsb

W.L.J.

Kosterink

J.G.W.

and Frijlink

H.W.

, Development and potential application of an oral ColoPulse infliximab tablet with colon specific release: A feasibility study, International Journal of Pharmaceutics 505 (2016), 175–186.

Bhol

K.C.

Tracey

D.E.

Lemos

B.R.

Lyng

G.D.

Erlich

E.C.

Keane

D.M.

Quesenberry

M.S.

Holdorf

A.D.

Schlehuber

L.D.

Clark

S.A.

and Fox

B.S.

, AVX-470: A novel oral anti-TNF antibody with therapeutic potential in inflammatory bowel disease, Inflammatory Bowel Disease 19 (2013), 2273–2281.

Egger

Schlegel

Baumann

Stoffers

Guggisberg

Brügger

Dürr

Stoll

Vergères

and Portmann

, Physiological comparability of the harmonized INFOGEST in vitro digestion method to in vivo pig digestion, Food Research International 102 (2017), 567–574.

Lichtenstein

Ron

Kivit

Y.S.

Horin

S.H.

Israel

I.E.

Fraser

Dotan

Chowers

Cohen

R.C.

and Weiss

, Infliximab-related infusion reactions: Systematic review, Journal of Crohn’s and Colitis 9 (2015), 806–815.

Zeeshana

Ali

Khana

Khan

S.A.

and Weigmann

, Review advances in orally-delivered pH-sensitive nanocarrier systems; an optimistic approach for the treatment of inflammatory bowel disease, International Journal of Pharmaceutics 558 (2019), 201–214.

10.

Swindle

M.M.

Makin

Herron

A.J.

, Jr. Clubb

F.J.

and Frazier

K.S.

, Swine as models in biomedical research and toxicology testing, Veterinary Patholology 49 (2012), 344–356.

11.

Burton

Patel

R.R.

Lemos

B.R.

Tracey

D.E.

and Hartman

, Proteolytic stability of AVX-470, a bovine colostral anti-TNF antibody, determined in vitro and in clinical samples after oral administration to patients with ulcerative colitis, Gastroenterology 148 (2015), 847.

12.

Boisen

and Fernández

J.A.

, Prediction of the apparent ileal digestibility of protein and amino acids in feedstuffs and feed mixtures for pigs by in vitro analyses, Animal Feed Science and Technology 51 (1995), 29–43.

13.

Boscolo

Mion

Licciulli

Macor

De Maso

Brce

Antoniou

M.N.

Marzari

Santoro

and Sblattero

, Simple scale-up of recombinant antibody production using an UCOE containing vector, New Biotechnology 29 (2012), 478–484.

14.

Nurbhai

West

MacDonald

Irving

G.P.

Vossenkaemper

and Crowe

, Measured and modelled data suggest that oral administration of V565, a novel domain antibody to TNF-alpha, could be beneficial in the treatment of IBD, Journal of Crohn’s and Colitis 12 (2018), 448–449.

15.

Hur

S.J.

Lim

B.O.

Decker

E.A.

and McClements

D.J.

, In vitro human digestion models for food applications, Food Chemistry 125 (2011), 1–12.

16.

Quezada

S.M.

McLean

L.P.

and Cross

R.K.

, Adverse events in IBD therapy: The 2018 update, Expert Review of Gastroenterology & Hepatology 12 (2018), 1183–1191.

17.

Klotz

Temland

and Schwab

, Clinical pharmacokinetics and use of infliximab, Clinical Pharmacokinetics 46 (2007), 645–660.

18.

Yadav

Varum

Bravo

Furrer

and Basit

A.W.

, Gastrointestinal stability of therapeutic anti-TNFα IgG1 monoclonal antibodies, International Journal Pharmaceutics 502 (2016), 181–187.

19.

Jasion

V.S.

and Burnett

B.P.

, Survival and digestibility of orally-administered immunoglobulin preparations containing IgG through the gastrointestinal tract in humans, Nutritional Journal 14 (2015), 22.