Inflammatory Reflex: 신경-면역 상호작용의 기초 메커니즘. 크론병 사례로

[문형철]

Inflammatory Reflex

신경-면역 상호작용의 기초 메커니즘

주어진 논문(Keeler et al., 2023, PMC10798868)을 기반으로

inflammatory reflex(염증 반사)를 설명하면,

이는 미주신경(vagus nerve)을 매개로 한

선천적 신경-면역 메커니즘으로 정의된다.

이 반사 Inflammatory Reflex는

장 및 기타 장기의 염증에 대한 감지와 억제를 담당하며,

중추신경계(CNS)와 내장 사이의 양방향 소통 경로로 작동한다.

미주신경의 약 80%가 감각(afferent) 뉴런으로 구성되어

염증 신호(예: 사이토카인 방출)를 brainstem으로 전달하고,

나머지 20%의 운동(efferent) 뉴런이

이를 바탕으로 항염증 신호를 장기로 되돌려 보낸다.

이 메커니즘은

특히 크론병(Crohn's disease, CD)과 같은 만성 염증성 장 질환에서

미주신경 자극(vagus nerve stimulation, VNS)을 통해

TNF-α와 같은 프로염증 사이토카인을 억제하는 것으로 입증되었다.

논문에서 VNS를 통해 CD 환자에서

혈청 TNF 수준이 46% 감소하고,

내시경적 개선(SES-CD 점수 하락)이 관찰된 것은

이 반사의 임상적 타당성을 뒷받침한다.

대학원 수준에서

inflammatory reflex의 메커니즘을 더 깊이 파헤쳐보자.

염증반사 네이처.pdf

277.29KB

https://pmc.ncbi.nlm.nih.gov/articles/PMC4536565/

https://bioelecmed.biomedcentral.com/articles/10.1186/s42234-019-0029-8

inflammatory reflex는 sensory-motor circuit로 구성되며,

afferent arm은 장벽 내 사이토카인(예: TNF-α, IL-1β, IL6) 방출에 의해 자극되어

cytokine-specific action potentials를 생성한다.

이러한 신호는 nodose ganglion을 거쳐

nucleus tractus solitarius(NTS, brainstem 내)로 전달되어,

dorsal motor nucleus of the vagus(DMV)와 같은 motor nuclei를 활성화한다.

Efferent arm은

cervical vagus nerve를 통해 celiac plexus로 이어지며,

splenic nerve를 매개로 비장(spleen)으로 분기되어

ChAT(choline acetyltransferase)+ T 세포에서 아세틸콜린(ACh)을 방출한다.

이 ACh는

대식세포(macrophages)와 면역세포의 α7 nicotinic acetylcholine receptor(α7nAChR)에 결합하여

downstream signaling을 억제한다.

구체적으로,

α7nAChR 활성화는 NF-κB의 핵 전위를 차단하고,

JAK/STAT 경로 및 inflammasome(예: NLRP3)을 억제하며,

CD11b 발현을 줄여 leukocyte trafficking을 감소시킨다.

또한,

T regulatory cell(Treg) 증가와 Th1 세포 감소를 유도하여 면역 균형을 회복하며,

specialized pro-resolving mediators(SPMs)의 방출을 촉진한다.

장 특이적으로는 myenteric plexus를 통해

enteric nervous system과 직접 인터페이스하며,

spleen-independent pathway(예: intestinal muscularis resident macrophages 직접 조절)가 관여하여

TNBS나 DSS 유발 대장염 모델에서 TNF 억제를 보인다.

이러한 분자-세포 수준의 억제는

VNS 후 24-48시간 지속되는 내구성을 가지며,

이는 α7nAChR agonist의 in vitro/in vivo 연구에서 확인된다.

이 개념의 기념비적인 발전을 위해,

2020-2025 기간 중 인용지수가 높은(74회 이상 인용) seminal 논문으로

Kelly et al.(2022)의 "Manipulation of the inflammatory reflex as a therapeutic strategy"

(Cell Reports Medicine)를 선정하였다.

https://pmc.ncbi.nlm.nih.gov/articles/PMC9381415/

Manipulation of the inflammatory reflex as a therapeutic strategy - PMC

이 리뷰는

Kevin Tracey의 원래 발견(2002, Nature)을 바탕으로

inflammatory reflex를

CNS-주도 peripheral immune modulation의 neural circuit로 재구성하며,

cholinergic anti-inflammatory pathway를

efferent arm의 핵심으로 강조한다.

논문은 transgenic 모델(예: α7nAChR knockout)과 vagotomy 실험을 통해

vagus-splenic nerve-ACh-ChAT+ T cell-α7nAChR-macrophage cascade를 검증하며,

이는 주어진 논문의 CD VNS 연구와 일치한다.

특히,

2020년대 업데이트로 bioelectronic medicine의 잠재력을 강조하는데,

non-invasive VNS(예: transcutaneous auricular stimulation)나

targeted neural interfaces를 통해

염증성 질환(류마티스 관절염, sepsis, IBD)의 임상 시험을 가속화할 수 있음을 제안한다.

이 접근은

기존 생물학적 제제(biologics)의 면역억제 부작용(감염, 악성종양)을 피하며,

preclinical 모델에서 endotoxin-induced TNF 억제와 sepsis 생존율 향상을 입증한다.

Tracey 그룹의 후속 작업처럼,

이 논문은 inflammatory reflex를

'hacking' 가능한 therapeutic target으로 재정의하여,

2022년 이후 VNS 임상 적용(예: biologic-naive CD에서 71% remission rate)을 촉진한 바 있다.

임상적으로, inflammatory reflex는

vagotomy 후 CD 위험 증가와 vagal tone 저하가

TNF 상승과 상관하는 역학 데이터로 지지되며,

VNS의 안전성(125,000+ epilepsy/depression 임플란트 사례에서 장기 합병증 없음)을 고려할 때

refractory IBD의 adjunct therapy로 유망하다.

그러나

장 colonic innervation의 불확실성과 ChAT+ T cell의 gut-specific 역할 규명 등

미해결 과제가 남아 있으며,

대규모 RCT를 통해 spleen-dependent vs. independent pathway의 차이를 밝혀야 한다.

이 반사는

신경과학-면역학의 융합으로,

미래 bioelectronic intervention의 패러다임을 제시한다.

J Crohns Colitis

. 2023 Sep 21;17(12):1897–1909. doi: 10.1093/ecco-jcc/jjad151

Neuroimmune Modulation Through Vagus Nerve Stimulation Reduces Inflammatory Activity in Crohn’s Disease Patients: A Prospective Open-label Study

Geert D’Haens 1,#,✉, Michael Eberhardson 2,3,#, Zeljko Cabrijan 4,5,6, Silvio Danese 7,8, Remco van den Berg 9, Mark Löwenberg 10, Gionata Fiorino 11,12, P Richard Schuurman 13, Göran Lind 14,15, Per Almqvist 16,17,18, Peder S Olofsson 19,20, Kevin J Tracey 21,22,23, Stephen B Hanauer 24, Ralph Zitnik 25,26, David Chernoff 27, Yaakov A Levine 28,29,30

• Author information
• Article notes
• Copyright and License information

PMCID: PMC10798868  PMID: 37738465

See "Targeting the Vagus Nerve to Treat Inflammatory Bowel Disease?" with doi: 10.1093/ecco-jcc/jjad149.

AbstractBackground and Aims

Crohn’s disease [CD] is a debilitating, inflammatory condition affecting the gastrointestinal tract. There is no cure and sustained clinical and endoscopic remission is achieved by fewer than half of patients with current therapies. The immunoregulatory function of the vagus nerve, the ‘inflammatory reflex’, has been established in patients with rheumatoid arthritis and biologic-naive CD. The aim of this study was to explore the safety and efficacy of vagus nerve stimulation in patients with treatment-refractory CD, in a 16-week, open-label, multicentre, clinical trial.

Methods

A vagus nerve stimulator was implanted in 17 biologic drug-refractory patients with moderately to severely active CD. One patient exited the study pre-treatment, and 16 patients were treated with vagus nerve stimulation [4/16 receiving concomitant biologics] during 16 weeks of induction and 24 months of maintenance treatment. Endpoints included clinical improvement, patient-reported outcomes, objective measures of inflammation [endoscopic/molecular], and safety.

Results

There was a statistically significant and clinically meaningful decrease in CD Activity Index at Week 16 [mean ± SD: -86.2 ± 92.8, p = 0.003], a significant decrease in faecal calprotectin [-2923 ± 4104, p = 0.015], a decrease in mucosal inflammation in 11/15 patients with paired endoscopies [-2.1 ± 1.7, p = 0.23], and a decrease in serum tumour necrosis factor and interferon-γ [46–52%]. Two quality-of-life indices improved in 7/11 patients treated without biologics. There was one study-related severe adverse event: a postoperative infection requiring device explantation.

Conclusions

Neuroimmune modulation via vagus nerve stimulation was generally safe and well tolerated, with a clinically meaningful reduction in clinical disease activity associated with endoscopic improvement, reduced levels of faecal calprotectin and serum cytokines, and improved quality of life.

초록

배경 및 목적

크론병[CD]은 위장관을 침범하는 쇠약성 염증성 질환이다. 현재 치료법으로는 완치가 불가능하며, 환자의 절반 미만이 지속적인 임상적·내시경적 관해를 달성한다. 미주신경의 면역조절 기능인 '염증 반사'는 류마티스 관절염 및 생물학적제제 미경험 CD 환자에서 확인된 바 있다. 본 연구의 목적은 치료 불응성 크론병 환자를 대상으로 16주간의 공개 다기관 임상시험을 통해 미주신경 자극의 안전성과 유효성을 탐구하는 것이었다.

방법

중등도에서 중증의 활동성 크론병을 가진 생물학적 제제 불응성 환자 17명에게 미주신경 자극기를 이식하였다. 1명의 환자는 치료 전 연구를 중도 탈락하였으며, 16명의 환자는 16주간의 유도 치료 및 24개월간의 유지 치료 기간 동안 미주신경 자극 치료를 받았다[동반 생물학적 제제 투여 환자 4/16명]. 평가 변수에는 임상적 개선, 환자 보고 결과, 염증의 객관적 측정[내시경/분자학적], 안전성이 포함되었다.

결과

16주차에 CD 활동 지수에서 통계적으로 유의하고 임상적으로 의미 있는 감소가 관찰되었다[평균±표준편차: -86.2±92.8, p=0.003]. 대변 칼프로텍틴 수치도 유의하게 감소하였다[-2923±4104, p = 0.015], 쌍을 이루는 내시경 검사를 시행한 15명 중 11명의 환자에서 점막 염증 감소 [-2.1 ± 1.7, p = 0.23], 혈청 종양괴사인자 및 인터페론-γ 감소 [46–52%]가 관찰되었다. 생물학적 제제 없이 치료받은 11명 중 7명의 환자에서 삶의 질 지표 2개가 개선되었습니다. 연구 관련 중대한 이상반응은 1건 발생했으며, 이는 기기 제거가 필요한 수술 후 감염이었습니다.

결론

미주신경 자극을 통한 신경면역 조절은

전반적으로 안전하고 내약성이 우수했으며,

내시경적 개선, 분변 칼프로텍틴 및 혈청 사이토카인 수치 감소, 삶의 질 향상과 연관된

임상적으로 의미 있는 질환 활동성 감소를 보였습니다.

Keywords: Crohn’s disease, vagus nerve, neuroimmune modulation

Graphical AbstractGraphical Abstract.

Open in a new tab

1. Introduction

Crohn’s disease [CD] is a chronic inflammatory condition that can affect any portion of the gastrointestinal tract. The incidences of CD in the USA and in Europe are approximately 10.7 and 6.3 per 100 000 person-years, respectively, with a trending increase in incidence as well as preval‎ence.1–3 Current treatments include corticosteroids, immunomodulators, and biologic drugs including anti-tumour necrosis factor [TNF] antibodies such as infliximab and adalimumab. Other biologics include anti-interleukin [IL]-12/IL-23 antibodies and the integrin-inhibiting antibody vedolizumab, as second-line therapeutics. However, many patients do not have an adequate response or develop a loss of response. On average, 65% of patients with severe CD fail to attain steroid-free clinical remission.4–7 Furthermore, biologics and conventional therapies for moderately to severely active CD have potential significant side effects, and some include black box warnings for infections or malignancies.

The vagus nerve, a bilateral cranial nerve which arises in the brain stem and innervates the body’s organs, is an established therapeutic target for patients with drug-refractory epilepsy and depression.8 It communicates bidirectional information between the viscera and central nervous system through afferent and efferent neurons [approximately 80% and 20% of neurons, respectively]9 [Figure 1]. For the past three decades, electrical devices to stimulate the vagus nerve have been implanted in more than 125 000 patients to treat drug-refractory epilepsy and depression.8 This procedure is established to be safe, well tolerated, and devoid of significant long-term complications.10

1. 서론

크론병[CD]은 위장관의 어느 부위에도 영향을 미칠 수 있는 만성 염증성 질환이다. 미국과 유럽의 크론병 발생률은 각각 인구 10만 명당 약 10.7명과 6.3명으로, 발생률과 유병률이 증가하는 추세입니다.1–3 현재 치료법으로는 코르티코스테로이드, 면역조절제, 인플릭시맙 및 아달리무맙과 같은 항종양괴사인자[TNF] 항체를 포함한 생물학적 제제가 있습니다. 기타 생물학적 제제로는 2차 치료제로 사용되는 인터루킨[IL]-12/IL-23 항체 및 인테그린 억제 항체인 베돌리주맙이 있다. 그러나 많은 환자들이 충분한 반응을 보이지 않거나 반응이 소실되는 경우가 있다. 중증 크론병 환자의 평균 65%는 스테로이드를 사용하지 않는 임상적 관해에 도달하지 못합니다.4–7 또한 중등도에서 중증의 활동성 크론병에 대한 생물학적 제제 및 기존 치료법은 잠재적으로 심각한 부작용이 있으며, 일부는 감염 또는 악성 종양에 대한 블랙박스 경고를 포함합니다.

뇌간에서 기원한 양측성 뇌신경으로 신체 장기를 신경 분포하는 미주신경은

약물 불응성 간질 및 우울증 환자를 위한 확립된 치료 표적이다.8

https://www.frontiersin.org/journals/medical-technology/articles/10.3389/fmedt.2021.696543/full

미주신경자극(VNS) 치료® 시스템은
약물 내성 간질 치료를 위한
최초의 FDA 승인 의료기기 치료법입니다.

지난 20년간 이 기술은
여러 차례의 개선을 거쳐
소프트웨어 관련 업데이트와 이식형 리드 및 발전기 하드웨어의 발전을 이루었습니다.

현재 의료진은
단일 및 이중 핀 제너레이터(모델 100, 101, 102, 102R, 103, 104, 105, 106, 1000)와
관련 프로그래밍 시스템(모델 250, 3000)을 흔히 접하게 되며,
이들 각각은 미묘하지만 실질적인 차이점을 지닙니다.

따라서 비교를 위해 이식 모델들의 매뉴얼을 검토하는 것은 어려운 작업일 수 있으며, 일부 매뉴얼은 쉽게 구할 수 없습니다. 본 리뷰에서는 장치 승인 이정표와 관련하여 VNS 치료 시스템의 기술적 진화를 강조하고, 기존의 개방형 루프와 최신 폐쇄형 루프 제너레이터 모델을 비교합니다. 또한 배터리 수명 예측과 자극 모드 상호작용에 대한 심층 분석을 제시하여 제너레이터 모델 간 차이를 더욱 명확히 구분합니다.

서론

미주신경자극(VNS) 치료® 시스템은 약물저항성 간질(DRE)의 보조 치료를 위한 최초의 FDA 승인 의료기기 치료법으로, 입증된 안전성과 내약성 프로파일을 보유하고 있습니다(1, 2).

이 시스템은 이식형 펄스 발생기 및 리드와 자극 설정을 변경하는 데 사용되는 외부 프로그래밍 시스템으로 구성됩니다.
펄스 발생기는 다양한 자극 모드를 통해 리드를 경유하여 미주 신경에 전기 신호를 전달하는 다중 프로그래밍 가능 의료 기기입니다.

외부 프로그래밍 시스템을 통해 의료진은 발생기 설정을 변경할 수 있을 뿐만 아니라 기기가 시간 경과에 따라 수집한 데이터를 시각화하고 다운로드할 수 있습니다(3). 수술적 이식 절차, 자기공명영상(MRI) 안전성 및 호환성, 프로그래밍 원리, 그리고 치료 사용을 뒷받침하는 실제 임상 증거에 대한 상세한 검토는 다른 문헌에서 확인할 수 있습니다(4–9).

VNS 치료의 작용 기전은 여러 경로를 통해 이루어집니다(10–19).

회로 관점에서
VNS 치료의 항경련 효과는
뇌간, 피질하 및 피질 구조로 구성된
“미주 신경 구심성 네트워크”의 교차점을 조절함으로써 발생하는 것으로 여겨집니다(20).

간질 환자에서 이 네트워크의 조절은
발작 확산 및 전기피질도 공간적 동기화를 감소시키는 것으로 추정됩니다(13, 21–23).

신경영상 기법은
시상 및 피질 노드에서 VNS의 급성 및 지속적 효과를 추가로 밝혀냈다(15, 24–27).

최근 연결체 연구는
편도체, 시상-피질, 반구 간 연결 섬유 내 좌측 편향 미세구조 및 연결성의 견고성이
VNS 치료 반응성을 신뢰성 있게 예측함을 입증함으로써 이러한 관찰을 뒷받침한다(28–31).

VNS 치료는 도입 이후 여러 차례의 개선을 거쳐 소프트웨어 관련 업그레이드, 하드웨어 개선, 심지어 명칭까지 변화해 왔습니다: 신경사이버네틱 보철(NCP)에서 VNS 치료 시스템(3)으로. 본 리뷰 시점까지 전 세계적으로 125,000명 이상의 환자에게 이 치료가 이식되었습니다(6). 새로운 항간질 신경조절 장치 기술이 치료 영역에 진입함에 따라, 환자에게 가장 적합한 치료법을 선택하기 위해서는 발작 부담 감소 외에도 프로그래밍 기능과 자극 능력의 임상적 유용성을 고려해야 합니다(7–9, 32–37).

현재 의료진은 단일 및 이중 핀 VNS 치료 발생기와 관련 프로그래밍 시스템을 흔히 접하게 되며, 이들 모두 미묘하지만 실용적인 차이점을 지닙니다. 본 리뷰에서는 역사적·진화적 맥락에서 VNS 치료 기술의 발전적 기원을 조명한다. 제조사 의사용 매뉴얼 및 기타 자료의 관련 정보를 통합하여 의료진에게 시스템 구성 요소, 배터리 수명 예측, 자극 모드를 비교하는 전용 참고 자료를 제공한다(3, 5, 38–43).

미주신경은

구심성 및 원심성 신경세포(각각 신경세포의 약 80%와 20%를 차지함)를 통해

내장 기관과 중추신경계 간 양방향 정보를 전달한다.9 [그림 1].

지난 30년간 약물 불응성 간질 및 우울증 치료를 위해

미주신경 자극 전기 장치가 125,000명 이상의 환자에게 이식되었습니다.8

이 시술은 안전성이 입증되었으며

내약성이 우수하고 중대한 장기적 합병증이 없는 것으로 알려져 있습니다.10

Figure 1.

Open in a new tab

The inflammatory reflex. The vagus nerve functionally projects into the coeliac plexus which relays signals to the sympathetic splanchnic nerve. Activation promotes release of acetylcholine [ACh] from choline acetyltransferase [ChAT]-expressing T cells which inhibits the release of tumour necrosis factor [TNF] from splenic macrophages. Vagus nerve endings also project into the gut wall and interface with the enteric nervous system through the myenteric plexus, but the role of ChAT+ T cells in the gut has not yet been fully elucidated. NE, norepinephrine/noradrenaline. Reprinted from International Immunology 2021;33:349–56].

염증 반사.

미주신경은 기능적으로 복강신경총으로 투사되어 교감 내장신경으로 신호를 전달한다. 활성화는 콜린 아세틸트랜스퍼라제[ChAT] 발현 T 세포로부터 아세틸콜린[ACh]의 방출을 촉진하며, 이는 비장 대식세포로부터 종양괴사인자[TNF]의 방출을 억제한다. 미주 신경 말단은 또한 장벽으로 투사되어 장신경총을 통해 장신경계와 연결되지만, 장 내 ChAT+ T 세포의 역할은 아직 완전히 규명되지 않았다. NE, 노르에피네프린/노르아드레날린. International Immunology 2021;33:349–56에서 재인용].

Vagus nerve fibres also mediate the ‘inflammatory reflex’, an innate neuroimmune mechanism that is responsive to and inhibits inflammation in the intestines and other organs.11 Cytokine release in tissues stimulates the sensory arm of the inflammatory reflex, which relays cytokine-specific action potentials to the brainstem.12,13 Arrival of these signals activates brain stem motor neurons that reflexively transmit signals via the vagus nerve back to the organs. These vagus motor signals inhibit cytokine release in tissues via the myenteric plexus to the gut and the coeliac plexus to the spleen.14–19 The downstream mechanism of the inflammatory reflex is mediated by signal transduction via the α7 nicotinic acetylcholine [ACh] receptors on immunocytes which inhibit nuclear factor kappa B [NF-κB], Janus kinase [JAK]/signal transducer and activator of transcription proteins [STAT], and inflammasome activation.20–22 Other anti-inflammatory mechanisms of vagus nerve stimulation include suppressing CD11b expression‎, increasing T regulatory cells, decreasing Th1 cells, reducing auto-antibody production, and increasing specialised, proresolving mediator release23–29].

미주 신경 섬유는

또한 장 및 기타 기관의 염증에 반응하고

이를 억제하는 선천적 신경면역 기전인 '염증 반사'를 매개한다. 11

Vagus nerve fibres also mediate the ‘inflammatory reflex’, a

n innate neuroimmune mechanism that is responsive to and inhibits inflammation

in the intestines and other organs

조직 내 사이토카인 방출은

염증 반사의 감각 경로를 자극하며,

이는 사이토카인 특이적 활동 전위를 뇌간으로 전달한다.12,13

이러한 신호의 도달은

뇌간 운동 뉴런을 활성화시켜

미주신경을 통해 반사적으로 신호를 장기로 되돌려 보낸다.

이러한 미주 신경 운동 신호는

장의 장신경총과 비장의 복강신경총을 통해 조직 내 사이토카인 방출을 억제한다.14–19

염증 반사의 하류 기전은

면역세포 상의 α7 니코틴성 아세틸콜린[ACh] 수용체를 통한 신호 전달로 매개되며,

이는 핵인자 카파 B[NF-κB],

야누스 키나제[JAK]/신호전달 및 전사활성제 단백질[STAT], 그리고

인플라마좀 활성화를 억제합니다.20–22

미주신경 자극의 다른 항염증 기전으로는

CD11b 발현 억제, 조절 T세포 증가, Th1 세포 감소, 자가항체 생성 감소,

그리고 특화된 분해 촉진 매개체 방출 증가 등이 포함됩니다23–29].

Recent preclinical and clinical studies of electrical stimulation of the vagus nerve indicate significant attenuation of experimental colitis and CD signs and symptoms.18,30–36 A pilot, single-centre, clinical study of seven biologic-naïve CD patients used an implanted vagus nerve stimulator programmed with parameters used to treat epilepsy. CD activity was reduced in five [71%] patients and four patients achieved clinical and endoscopic remission within 6 months.36 In a 1-year follow up, and including two additional subjects, both clinical and endoscopic remission were observed in five patients at 12 months.37 The stimulation parameters [500 µs pulse width and 10 Hz] were adjusted during the trial to reach maximum tolerated intensities ranging between 0.25 mA and 1.25 mA, in cycles which ran 24 h per day for 30 s on followed by 5 min off. Recently, we observed that delivering much lower charge [250 µs pulse width, 10 Hz] for only 1–4 minutes per day was sufficient to significantly reduce inflammation and disease severity in preclinical animal models and in patients with rheumatoid arthritis.34,38,39

미주 신경 전기 자극에 대한 최근의 전임상 및 임상 연구는

실험적 대장염 및 크론병 징후와 증상의 현저한 감소를 나타낸다.18,30–36 

생물학적 제제 미경험 크론병 환자 7명을 대상으로 한

단일 센터 파일럿 임상 연구에서는

간질 치료에 사용되는 매개변수로 프로그래밍된 이식형 미주 신경 자극기를 사용했다.

5명[71%]의 환자에서 CD 활동이 감소했으며,

4명의 환자가 6개월 이내에 임상적 및 내시경적 관해를 달성했다.36 

1년간의 추적 관찰(추가 2명 포함)에서 12개월 시점에

5명의 환자에서 임상적 및 내시경적 관해가 관찰되었다. 37 

자극 매개변수[500 µs 펄스 폭 및 10 Hz]는

시험 중에 최대 허용 강도(0.25 mA ~ 1.25 mA 범위)에 도달하도록 조정되었으며,

30초 켜짐 후 5분 꺼짐의 주기로 하루 24시간 작동했습니다.

최근 우리는 훨씬 낮은 전하량[펄스 폭 250µs, 10Hz]을

하루에 단 1~4분만 전달해도 전임상 동물 모델과 류마티스 관절염 환자에서

염증과 질환 중증도를 현저히 감소시키기에 충분하다는 사실을 관찰했습니다.34,38,39

Despite this evidence, it was previously unknown whether stimulation of the vagus nerve with lesser electrical charge and reduced daily frequency [only once to four times daily] would decrease CD severity in biologic-refractory patients. Herein, we provide the results from a 16-week, open-label, multicentre, clinical trial with clinical and objective endpoints investigating the safety and efficacy of vagus nerve stimulation by an implanted electrical pulse generator in patients with moderately to severely active CD and insufficient or absent response to biologic drugs.

이러한 증거에도 불구하고,

더 적은 전기적 충전량과 감소된 일일 빈도(하루에 단 1~4회)로

미주신경을 자극하는 것이 생물학적 치료에 반응하지 않는 환자에서

크론병 중증도를 감소시킬 수 있는지 여부는 이전에 알려지지 않았습니다.

본 연구에서는

중등도에서 중증의 활동성 크론병을 앓고 있으며

생물학적 제제에 대한 반응이 불충분하거나 없는 환자들을 대상으로,

이식형 전기 펄스 발생기를 이용한 미주신경자극의 안전성과 유효성을 평가하기 위한

임상적 및 객관적 평가 지표를 포함한 16주간의 공개 다기관 임상시험 결과를 제시한다.

2. Materials and Methods

2.1. Study design and participants

We performed a 16-week, multicentre, open-label trial in four EU countries, examining the safety and efficacy of electrical stimulation of the vagus nerve as an innovative treatment for CD. Patients were 18–75 years of age, with moderately to severely active refractory CD >4 months after diagnosis and with a Crohn’s Disease Activity Index [CDAI] 220 to 450, endoscopic evidence of ulceration by Simple Endoscopic Score for Crohn Disease [SES-CD ulcer size score of at least 2 in at least one segment], and faecal calprotectin concentrations greater or equal to 200 μg/g.

Patients were eligible for enrolment if they had an insufficient response or were intolerant to at least one TNF inhibitor [ie, infliximab or adalimumab] or vedolizumab. Patients in Sweden were required to have failed both an anti-TNF agent and vedolizumab prior to enrolment. Azathioprine, 6-mercaptopurine, and methotrexate could be continued throughout the study but had to be stable for >12 weeks prior to enrolment. Prohibited medications within the pre-enrolment washout window included any TNF inhibitor, natalizumab, vedolizumab, oral glucocorticoids at doses greater than 10 mg prednisone orally daily, or an equivalent dose of other oral or parenteral glucocorticoids within 4 weeks, ciclosporin, tacrolimus, sirolimus, or mycophenolate mofetil within 4 weeks, intravenous antibiotics for CD within 4 weeks, parenteral or enteral feeding, or elemental diet within 2 weeks, and rectal use of 5-aminosalicylates or corticosteroid enemas or suppositories within 2 weeks. During the study [after the first nine patients were enrolled], a protocol modification was made that allowed for continued use of a stable [for at least 6 months] dose of a TNF inhibitor or vedolizumab throughout the study, to limit the amount of time the patients were untreated prior to Day 0. Exclusion criteria included coeliac disease, ulcerative or indeterminate colitis, enterocutaneous fistulae with abscesses, extensive colonic resection, bowel-related surgery within 12 weeks prior to enrolment, prior vagotomy, history of vasovagal syncope, pharyngeal dysfunction, preexisting vocal cord damage or dysfunction, uncontrolled asthma or obstructive lung disease, peptic ulcer disease, significant cardiac rhythm disturbances, sleep apnoea, or the use of other electrically active medical devices. A full list of inclusion and exclusion criteria can be found in Supplementary Table 1.

The safety-eval‎uable population included all patients who were screened, and in this group all adverse events were reported beginning at the time of signed informed consent. The efficacy-eval‎uable population included all patients who were implanted with the vagus nerve stimulation device and for whom at least one post-implantation documentation of primary efficacy data was available.

2.2. Intervention

The investigational study device was a standard Cyberonics VNS Therapy® System, including an implanted pulse generator (Demipulse Model 103; Cyberonics [now Livanova], London, UK) that was placed in a subcutaneous pocket in the chest wall, and lead [PerenniaFLEX Lead Model 304] that was secured around the left cervical vagus nerve and tunnelled to the pulse generator [Figure 2A–C].40 The external programming system includes the programming wand, the programming software, a compatible computer, and an actuating magnet. This software allows a physician to place the programming wand over the pulse generator to read and change device parameters. The automatic stimulation [optimised for epilepsy] was disabled on the device and the magnet was provided to the patients, with instructions for use for actuating the device daily according to protocol. The system components were treated as investigational study devices due to their off-label use in patients with CD.

Figure 2.

Open in a new tab

Study device and design.

The device settings chosen for this study were based on extensive preclinical studies, the proof-of-concept clinical study that was performed with the same device in two rheumatoid arthritis cohorts, as well as clinical vagus nerve stimulation experience with these devices used in epilepsy and depression.39,41,42 The study used a pulse frequency and pulse duration of 10 Hz and 250 µs, respectively, well below the specified upper limits allowed in the currently approved vagus nerve stimulation device product labelling in approved indications of drug-refractory epilepsy and depression.43

The study flow chart is shown in Figure 2D. Patients had screening assessments and baseline clinical and biomarker assessments at the Week -4 visit and the Week 2 visit, after which the device was implanted under general anaesthesia at the Week 2 Visit. During the implantation procedure prior to wound closure, the patients received a single stimulus as part of the standard intraoperative diagnostic testing to check system function and lead integrity and impedance. The device was then inactivated and the patient allowed recovery from surgery for at least 14 days. On the Week 0 visit, patients had postoperative clinical assessments and the stimulation was titrated to an output current as maximally tolerated between a minimum of 0.25 mA and a maximum of 2.0 mA [in 0.25-mA increments]. The patients self-administered for 1 min once per day active stimulation at a pulse frequency of 10Hz, a 250-µs pulse duration daily by passing the actuating magnet across their chest. On each visit between Weeks 1 and 4, the stimulation output current was further incremented as tolerated. At the Week 4 visit, daily stimulation time was increased to 2 min and at the Week 6 visit to 5 min. At Week 8 clinical scores were assessed, and if the patient had not achieved remission [CDAI <150], daily stimulation time was increased to 5 min four times per day. Early termination visits were analysed as Week 16 visits.

At the conclusion of the study, patients were offered the option to have the device surgically removed, left in place and inactivated, or to continue treatment in a long-term extension study that ends when the last patient enrolled reaches the 24-month study visit. Eleven of 16 patients in the efficacy group opted to continue in the extension study, which will be reported separately.

2.3. Assessments and measurements

2.3.1. Efficacy measures

The primary efficacy endpoint was mean change in CDAI between the preimplantation baseline and the Week 16 visit. Other endpoints included the proportion of patients that achieved CDAI remission [CDAI <150] and response [CDAI -70; drop in CDAI by at least 70 points] or enhanced response [CDAI -100; drop in CDAI by at least 100 points]. CDAI with partially missing CDAI subscores were calculated with the missing subscore imputed by last value carried forward. No CDAI imputations were allowed for study visits missing all CDAI subscore values or carried forward from a pre- to a post-therapy initiated time point.

Endoscopy was performed at screening and at the Week 16 visit and biopsies were collected. Recorded endoscopic videos were scored by two expert central readers blinded to the timing of the recordings. The averaged value of the independent scores was used for the primary analysis of SES-CD. SES-CD remission was defined as all observed segments having an ulcer score of </= 1. Standard forceps biopsies were collected from involved regions of ileum and colon and formalin fixed for standard histology. The most affected regions were read at a central reading facility by a pathologist blinded to the patient and treatment sequence information, using the semi-quantitative Geboes Score.44,45

Biomarker endpoints included change from baseline in faecal calprotectin, high-sensitivity C-reactive protein [hsCRP], and serum cytokines, and the heart rate variability-derived autonomic balance. Faecal calprotectin and hsCRP were measured centrally and serum cytokines were measured by MSD electrochemiluminescence [MSD Chemokine Panel 1 and Proinflammatory Panel 1, Meso Scale Discovery, Rockville, MD, USA].

Two quality-of-life scales were used to assess health-related quality of life: the Inflammatory Bowel Disease Questionnaire [IBDQ; MID of 16] and the Simple Health Score [SHS] instruments.46 The IBDQ scale increases and the SHS decreases as patients improve.

2.3.2. Safety measures

The overall safety and tolerability of the implantation, device, and treatment were assessed for the safety-eval‎uable population. Safety endpoints included were incidence, causality, and severity of serious adverse events [SAEs], adverse events, and clinical laboratory results. These were assessed throughout the study, coded using the Medical Dictionary for Regulatory Activities [MedDRA], and are presented by MedDRA term as incidence rates.

2.3.3. Statistical analyses

Descriptive statistics and the 95% confidence intervals [CIs] of the mean difference from baseline were calculated for CDAI, SES-CD, faecal calprotectin, hsCRP, IBDQ, and SHS. Changes from baseline to the Week 16 primary endpoint were compared by paired t test. As this was a pilot study, no adjustment for multiple comparisons were prespecified. In post hoc analyses, changes in CDAI, faecal calprotectin, and hsCRP from baseline to each study visit were further tested with a paired, mixed-effects analysis of variance [ANOVA] model [Restricted Maximum Likeliness; REML], adjusted with Bonferroni’s multiple comparisons test [Prism V.9, GraphPad, San Diego, CA, USA], and included in the figures. Changes in serum cytokine levels were assessed by Wilcoxon matched pairs signed rank test. Endoscopy across the SES-CD subscores and histopathology across the Geboes subscores were calculated by REML adjusted with Bonferroni’s multiple comparisons test. Correlations between continuous or discrete clinical, molecular, endoscopic, and quality-of-life parameters were quantified using Spearman’s rank correlation coefficients [Prism V.9, GraphPad]. In further post hoc analyses of the patients achieving a clinical response to therapy, median changes in clinical and patient-reported outcomes [CDAI, IBDQ, SHS] and in objective outcomes [faecal calprotectin, hsCRP, SES-CD, and the SES-CD colon subcomponent] were assessed by Wilcoxon column test vs a hypothetical change of 0.

2.4. Ethics statement

This study was done in accordance with the International Conference on Harmonisation guidelines for good clinical practice and the ethical principles of the Declaration of Helsinki. All patients gave written informed consent, which was reviewed and approved by an independent ethics committee or institutional review board. The study was registered with clinicalTrials.gov [NCT02311660].

3. Results

3.1. Patient disposition and baseline characteristics

A total of 31 patients were screened and 17 patients with active CD were enrolled at five sites and received a vagus nerve stimulator system implanted on the left cervical vagus nerve [Figure 2A–C]. Sixteen [94%] were included in the efficacy population because one patient suffered a postoperative wound infection, and the device was removed before stimulation had commenced [‘All Patients group’]. Twelve of the 16 patients [designated as ‘Stimulation Monotherapy group’] were analysed separately because the only nonconventional treatment they received during the study period was vagus nerve stimulation. Four patients continued biologic treatment during the 16-week study, at the treating physician’s discretion [2/4 treated on a stable dose]. Eleven of 17 subjects had any exposure to a corticosteroid during the trial, and 5 of 17 received >10 mg at any point after treatment was initiated on Day 0. Two patients in the efficacy population withdrew from the study prior to Week 16, one due to a CD flare and one to undergo a magnetic resonance imaging [MRI] scan that was incompatible with the specific device implanted [Figure 3].

Figure 3.

Open in a new tab

Patient disposition.

The mean [range] age of the 17 patients at baseline was 35.4 [21–62] years, mean [range] body mass index [BMI] of patients was 22 [16.8–29.0], 76.5% of patients were male, and 82.4% of patients were White. The mean [range] number of biologic drugs previously experienced was 2.2 [1–6]. Patient characteristics are summarised in Table 1.

Table 1.

Baseline characteristics.

Total Safety Population, n	17
Enrolment by country
The Netherlands	6
Croatia	7 [6 in Efficacy Population]
Sweden	2
Italy	2
Sex, n [% male]	13 [76.5]
Ethnicity, n [% Caucasian]	14 [82.4]
Mean age, years [range]	35.4 [21.0–62.0]
Mean age at CD diagnosis, years [range]	23.5 [14.2–57.0]
Mean CD duration, years [range]	12.2 [4.7–25.2]
Mean height, cm [SD; range]	174.7 [5.8; 162.5–186.0]
Mean weight, kg [SD; range]	67.1 [11.7; 50.0–88.7]
Mean BMI, kg/m2 [SD; range]	22.0 [4.0; 16.8–29.0]
Number of prior biologics [SD; range]	2.2 [1.4; 1–6]
Number of patients on biologics at baseline	4
Number of patients with prior bowel resection	7
Number of patients with prior or current perianal fistula	5
Number of patients with confirmed colonic involvement	17
Number of patients with confirmed ileal involvement	11
Total Efficacy Population, n	16
CDAI [SD]	306.4 [59.4]
SES-CD [SD]	20.2 [6.8]
Faecal calprotectin [SD]	5054.4 [5062.54]
High-sensitivity CRP, mg/L [SD]	4.4 [3.4]

Open in a new tab

Abbreviations: BMI: Body Mass Index, CD = Crohn’s Disease, CDAI = Crohn’s Disease Activity Index, CRP = C-Reactive Protein.

3.2. Efficacy endpoints

Mean clinical disease activity was stable from the screening visit [Week -4] to Week 0 until the onset of daily stimulation of the vagus nerve. After the onset of vagus nerve stimulation, the mean decrease in CDAI from baseline to Week 16 was (mean ± standard error of the mean [SEM]: -86.2 ± 24.3, p = 0.003] in the full cohort, and -114.5 ± 23.9, [p = 0.0002] in the Stimulation Monotherapy group [Table 2, Figure 4A]. The individual patient change in CDAI during the study is shown in Supplementary Figure 1A and B. During active treatment, mean CDAI decreased from baseline ([mean ± SEM] in the All Patients group: baseline: 306 ± 15; Week 8: 218 ± 29, p = 0.001; Week 16: 221 ± 27, p = 0.003) [Figure 4B]. Clinical remission [CDAI <150] was achieved in 27% of the All Patients group and 36% of the Stimulation Monotherapy group, respectively, at Week 16. By Week 8, 47% and 64% of patients met the CDAI-100 criterion and more than half of those treated [53% and 64% of the All Patients group and Stimulation Monotherapy group, respectively] met the CDAI-100 threshold at Week 16. The individual patient CDAI scores are plotted in Supplementary Figure 2A and B. The proportions of patients who improved sufficiently to achieve the definition of clinical remission, and the CDAI-70 and CDAI-100 clinical responses following 8 and 16 weeks of vagus nerve stimulation treatment, are shown in Figure 4C and D.

Table 2.

Efficacy statistics.

Week 16 change from BaselineNMeanStdMinQ1MedianQ3MaxLower 95% CIUpper 95% CIT-test p-value

CDAI	151	-86.2	92.79	-249	-160	-121	-7	101	-137.6	-34.8	0.003
SES-CD	15	-2.1	6.43	-11	-6.5	-3	0.5	15	-5.6	1.5	0.23
IBDQ	15	9.4	34.14	-41	-23	7	46	60	-9.5	28.3	0.3
SHS	15	-33.7	107.74	-240	-121	-39	68	128	-93.3	26	0.25
Faecal calprotectin [µg/g]	15	-2923	4104	-10871	-5357	-1623	-4622	4395	-5196	-650	0.015
hsCRP [mg/dL]	16	-0.5	2.71	-4.7	-2.3	-0.4	0.4	5.5	-2	0.9	0.46

Open in a new tab

1Week 16 or Early Termination samples not available for every patient.

Abbreviations: CDAI = Crohn’s Disease Activity Index, hsCRP = High-Sensitivity C-Reactive Protein, IBDQ = Inflammatory Bowel Disease Questionnaire, SES-CD = Simple Endoscopic Score-Crohn’s Disease, SHS = Simple Health Score.

Figure 4.

Open in a new tab

Clinical efficacy. [A] Change from baseline in CDAI [mean ± SEM] and [B] CDAI [mean ± SEM] over time in the All Patients Efficacy population [n = 16] and in the 12-patient Stimulation Monotherapy subpopulation. [C] Percent of All Patients or [D] Stimulation Monotherapy patients who achieved clinical remission [CDAI <150], CDAI-70 [CDAI decrease from baseline ≥70], and CDAI-100 [CDAI decrease from baseline ≥100]. CDAI and its change from baseline were analysed with a paired mixed-effects model [restricted maximum likeliness; REML] and adjusted with Bonferroni’s multiple comparisons test. **p <0.01, ***p <0.001. CDAI, Crohn’s Disease Activity Index; SEM, standard error of the mean.

A decrease in SES-CD was observed from (median [interquartile range [IQR]) 24 [13.5–25.5] at Baseline to 17.5 [12–21.5] at Week 16, and numerically in 11/15 [73%] of patients with paired endoscopies [Figure 5A]. Two patients [13.3%] had ileal-colonic sections that were not scored at baseline yet scored at Week 16, artificially reducing improvement in SES-CD. One patient [7%] achieved endoscopic remission [all observed segments had an ulcer score of </= 1] at Week 16. Six patients [40%] and one patient [7%] had a decrease in SES-CD of >25% and >50%, respectively [Supplementary Figure 3A]. There were modest numerical improvements between Baseline and Week 16 in the mean SES-CD subscores in the sigmoid and left colon and in the terminal ileum [Supplementary Figure 3B]. In patients who achieved a clinical response at Week 16, there was a significant improvement in SES-CD subscores across the entire colon [p = 0.038] [Supplementary Figure 3C]. Ileal biopsies showed significant improvement from baseline at Week 16 across the eight histopathological subcategories [p <0.01] [Figure 5B]. However, histopathology of biopsies from the colon and the rectum did not show significant improvement from baseline [p = 0.14 and 0.93, respectively] [Supplementary Table 2].

Figure 5.

Open in a new tab

Bowel inflammation. [A] SES-CD for each patient at screening visit and Week 16/Early Termination visit. The hollow point denotes patient in SES-CD remission. [B] Histopathology subcomponent score in biopsies of the most affected region of the ileum [samples from six subjects; mean ± SEM]. Ileal histopathology change from baseline was analysed by restricted maximum likeliness [REML]. * p <0.01. SES-CD, Simple Endoscopic Score Crohn’s Disease; ET, early termination; SEM, standard error of the mean.

A significant mean reduction from baseline in faecal calprotectin at Week 16 was observed in the All Patients group ([mean ± SEM] Baseline: 5054 ± 1266, Week 16: 1969 ± 625.5, p = 0.02) and in the Monotherapy group ([mean ± SEM] Baseline: 4705 ± 1295, Week 16: 1496 ± 579, p = 0.004) [Table 2, Figure 6A, Supplementary Figure 4A and B]. The mean level of faecal calprotectin over time is plotted in Figure 6A and shows stable levels from the pre-implantation biomarker baseline [the average of Week -4 and Week -2] to Week 0 when daily stimulation of the vagus nerve commenced. The mean faecal calprotectin level was significantly lower than baseline by 12 weeks of stimulation. Mean levels of CRP were numerically lower from Week 12 onwards compared with baseline [Figure 6B]. The reduction in faecal calprotectin was significant [p = 0.02] and the reduction in CRP trended lower [p = 0.20] in those patients who achieved a clinical response [Supplementary Figure 7].

Figure 6.

Open in a new tab

Disease-related biomarkers. [A] Faecal calprotectin [mean ± SEM] and [B] serum hsCRP [mean ± SEM] over time in the All Patients Efficacy population and in the 12-patient Stimulation Monotherapy subpopulation. Change from baseline was analysed by paired restricted maximum likeliness [REML] and adjusted with Bonferroni’s multiple comparisons test. *p <0.05, **p <0.01. hsCRP, high-sensitivity C-reactive protein; SEM, standard error of the mean.

The majority of patients reported improvement in their IBDQ [change in IBDQ >0] at Week 16, with 6 of 11 Stimulation Monotherapy patients exceeding the ‘minimal important difference’ of 16 [Table 2, Supplementary Figure 5A]. In contrast to the IBDQ, the SHS decreases as patients improve. By Week 16, a majority of patients reported improvements in their SHS compared with baseline [change in SHS <0] [Table 2, Supplementary Figure 5B]. In those patients who achieved a clinical response, the improvements in both IBDQ and SHS were significant [p <0.05] [Supplementary Figure 7].

Serum cytokine concentrations were also measured to assess the inflammatory biomarker response to vagus nerve stimulation. We observed a 46% and 52% decrease from baseline in mean levels of TNF and IFN-γ [Figure 7, Supplementary Figure 6]. Mean total Il-17 levels were 54% higher at Week 16 than at baseline. The full cytokine panel is presented in Supplementary Table 3. Correlations of the change from Baseline to Week 16 in clinical [CDAI], molecular [faecal calprotectin, hsCRP, TNF, IFN-γ, IL-17], endoscopic [SES-CD], and quality of life outcomes [IBDQ, SHS] were analysed to investigate the interactions between the varied endpoints in the context of this therapy [Supplementary Figure 8A and B]. Primarily, the change in CDAI was significantly correlated to changes in the quality-of-life assessments, and it positively correlated to the change in SHS and negatively correlated to the change in IBDQ [Spearman r >|0.5|, p <0.05] [Supplementary Figure 6A and B]. Additional interactions are described in Supplementary Figure 6. In those patients who achieved a clinical response, changes in serum levels of TNF, IL-17, and IFN-γ move together with a significant correlation between changes in IL-17 and IFN-γ levels [Spearman r = 0.8, p = 0.02] [Supplementary Figure 8 C and D].

Figure 7.

Open in a new tab

Serum cytokines: TNF, IFN-γ, and IL-17 [mean ± SEM]. Change from baseline was analysed by paired Wilcoxon test. *p <0.05. SEM, standard error of the mean; TNF, tumour necrosis factor; IL, interleukin.

3.3. Safety

All 17 patients reported at least one treatment-emergent adverse event during the study and most were mild or moderate [Table 3]. Eight patients experienced at least one SAE. All but one event was CD related and 50% of these SAEs [6/12] occurred prior to initiation of stimulation on Day 0. None were deemed to be treatment or device related, and one was related to device implantation/explantation [Supplementary Table 4]. Three patients discontinued the study prematurely, including one patient with a postoperative wound infection following device implantation [the implantation-related SAE]. One patient withdrew; their device was removed to enable diagnostic MRI. Another patient developed a relapse of Crohn’s disease with a prolonged, increased, inflammatory response. No significant adverse vital signs, physical examination or other observational findings were noted.

Table 3.

Treatment-emergent adverse events occurring in > 10% of patients enrolled in the study

Adverse eventSafety Population, n = 17, n [%]

Crohn’s disease exacerbation	7 [41.2]
Abdominal pain	3 [17.6]
Anaemia	3 [17.6]
Pyrexia	3 [17.6]
Cachexia	2 [11.8]
Hypokalaemia	2 [11.8]
Pallor	2 [11.8]
Dysphonia	2 [11.8]
Oropharyngeal pain	2 [11.8]
Alopecia	2 [11.8]
Back pain	2 [11.8]
Joint swelling	2 [11.8]
Pain in jaw	2 [11.8]
Fatigue	2 [11.8]
Any serious adverse event	8 [47]
Any serious infection	2 [11.8]
Any cancer	0 [0]

Open in a new tab

4. Discussion

In this 16-week, open-label, clinical trial of neuroimmune modulation therapy in 17 patients with moderately to severely active CD, we observed a significant reduction in CDAI and faecal calprotectin levels. Clinical response [decrease in CDAI of at least 70 or 100 points], and clinical remission were achieved in a substantial percentage of patients. Together, these data indicate that vagus nerve stimulation improves CD clinical activity within 8 to 16 weeks. Moreover, some improvements in endoscopic severity were observed, even though this did not reach statistical significance. Significant endoscopic improvements were observed in the colons of those patients who achieved a clinical response. Significantly improved histopathological outcomes at Week 16 were also observed in the terminal ileum, but not in colonic segments, based on the Geboes score of paired biopsies with histopathological inflammation at baseline. There were no treatment-related serious adverse events. One implantation-related infection was reported.

The vagus nerve has immunoregulatory functions, and one important mechanism is the inflammatory reflex. Vagus nerve stimulation has been demonstrated to reduce inflammation in various colitis and intestinal inflammation models, including DSS-, oxazolone-, and TNBS-colitis, indomethacin enteropathy, and post-operative ileus.18,30–35 Electrical stimulation of the cervical vagus nerve reduces serum TNF and attenuates the severity of sepsis through a mechanism requiring cholinergic inhibition of pro-inflammatory immune cells.47–49 The vagus nerve innervation of the gut is still not completely mapped, and although vagus nerve innervation of the small bowel is well established, there are conflicting reports as to whether the colon is directly innervated.50–52 Cholinergic nerve endings have been localised adjacent to the myenteric plexus of the intestinal wall, but details of the interaction between intestinal immune cells and the vagus nerve are incompletely understood53–55 [Figure 1].

A number of studies have focused on nicotinic ACh receptor-mediated regulation of gut inflammation and vagus nerve stimulation in murine colitis models. These results indicate that the vagus nerve regulates colonic inflammation and that stimulating the vagus nerve attenuates gut-specific disease activity.30,32,56,57 A recent published study also reported reduction of small bowel inflammation by stimulating the vagus nerve of rats in a model of Crohn’s-like disease, through a spleen-independent mechanism.34 The immunoregulatory role of the vagus nerve in the gut has been studied in humans, and one epidemiological study looked at 15 637 vagotomised patients, finding a significant association between vagotomy and later development of CD with an incidence of 0.38 per 1000 person-years compared with 0.26 in non-vagotomised controls.58 Vagus nerve stimulation has been used as a therapy for refractory epilepsy for 25 years, and vagus nerve stimulation devices have been implanted in more than 125 000 patients, with high patient tolerability.8,10 The independent pilot vagus nerve stimulation study expanded to nine patients with moderately active CD naïve to biologic drugs, and report at 12 months has shown encouraging data on the reduction of symptoms and inflammatory biomarkers as well as endoscopic improvement, also reporting no serious adverse events.37

4. 논의

중등도에서 중증의 활동성 크론병 환자 17명을 대상으로 한 이 16주간의 공개 라벨 신경면역 조절 요법 임상 시험에서, 우리는 CDAI 및 분변 칼프로텍틴 수치의 유의미한 감소를 관찰했습니다. 상당수 환자에서 임상 반응(CDAI 70점 또는 100점 이상 감소) 및 임상적 관해가 달성되었다. 종합적으로 본 데이터는 미주신경 자극이 8~16주 내에 CD의 임상적 활성도를 개선함을 시사한다. 또한 내시경적 중증도에서 통계적 유의성은 달성하지 못했으나 일부 개선이 관찰되었다. 임상 반응을 달성한 환자군에서는 대장에서 유의한 내시경적 개선이 관찰되었다. 기저선에서 조직병리학적 염증이 확인된 대조군 생검 샘플의 Geboes 점수를 기준으로, 16주차에 말단 회장부에서는 유의미한 조직병리학적 개선이 관찰되었으나 대장 부위에서는 그렇지 않았다. 치료 관련 중대한 이상반응은 발생하지 않았다. 이식 관련 감염 1건이 보고되었다.

미주신경은 면역조절 기능을 가지며, 그중 중요한 기전 중 하나는 염증 반사이다. 미주신경 자극은 DSS, 옥사졸론, TNBS 대장염, 인도메타신 장병증, 수술 후 장폐색 등 다양한 대장염 및 장 염증 모델에서 염증을 감소시키는 것으로 입증되었다.18,30–35 경부 미주신경의 전기 자극은 혈청 TNF를 감소시키고, 전염성 염증성 면역 세포에 대한 콜린성 억제를 필요로 하는 기전을 통해 패혈증의 중증도를 완화시킨다. 47–49 장의 미주 신경 분포는 아직 완전히 파악되지 않았으며, 소장의 미주 신경 분포는 잘 알려져 있지만 대장이 직접 신경 분포를 받는지에 대해서는 상반된 보고가 있다. 50–52 콜린성 신경 말단은 장벽의 장신경총에 인접한 위치에서 확인되었으나, 장 면역 세포와 미주 신경 간의 상호작용에 대한 세부 사항은 완전히 이해되지 않았다53–55 [그림 1].

다수의 연구는 쥐 대장염 모델에서 니코틴성 아세틸콜린 수용체 매개 장 염증 조절 및 미주 신경 자극에 초점을 맞추었다. 이러한 결과는 미주 신경이 대장 염증을 조절하며, 미주 신경 자극이 장 특이적 질환 활성을 약화시킨다는 점을 시사한다.30,32,56,57 최근 발표된 연구에서도 크론병 유사 질환 모델에서 쥐의 미주 신경을 자극함으로써 비장 독립적 기전을 통해 소장 염증이 감소하는 것으로 보고되었다.34 장 내 미주 신경의 면역 조절 역할은 인간을 대상으로 연구되었으며, 한 역학 연구에서는 15,637명의 미주신경절제술을 받은 환자를 대상으로 조사한 결과, 미주신경절제술과 이후 크론병 발병 사이에 유의미한 연관성이 발견되었으며, 발병률은 1000인년당 0.38건으로 미주신경절제술을 받지 않은 대조군의 0.26건보다 높았다.58 미주신경 자극은 25년간 난치성 간질 치료법으로 사용되어 왔으며, 미주신경자극장치는 125,000명 이상의 환자에게 이식되었으며, 높은 내약성을 보였다.8,10 독립적인 미주신경자극 파일럿 연구는 생물학적 제제 미경험 중등도 활동성 크론병 환자 9명으로 확대되었으며, 12개월 보고에서 증상 및 염증 생체표지자 감소와 내시경적 개선에 대한 고무적인 데이터를 보여주었으며, 중대한 이상반응도 보고되지 않았다.37

Observational studies in humans support the role of the vagus nerve and the parasympathetic tone in gut inflammation. One study reported lower vagal tone, as measured by heart rate variability, in CD patients compared with healthy controls, and another report showed that patients with high resting vagal tone had lower circulating TNF.59,60 Furthermore, in patients with ulcerative colitis, an association was observed between higher parasympathetic activity during a flare of the disease and lower systemic inflammation during a 3-year follow-up.61 An association has also been reported between a history of vagotomy and the development of CD.58

The first human study of an implanted vagus nerve stimulation device to report inhibition of cytokine biomarkers was performed in seven patients under full anaesthesia during implantation of a vagus nerve stimulator for the treatment of epilepsy. The data showed that endotoxin-induced TNF production in blood drawn from the patients was significantly reduced after a single intraoperative stimulation.39 Several small human trials of vagus nerve stimulation for inflammatory diseases, with promising efficacy outcomes, were reported, including: an open-label multicentre trial of 17 patients with active rheumatoid arthritis refractory to methotrexate and/or multiple biologic agents; a double-blinded multicentre trial of 14 patients with active rheumatoid arthritis refractory to multiple biologic and/or targeted synthetic agents; and an open-label, single-centre trial of nine patients with CD.36–39 The vagus nerve stimulation dosing in the rheumatoid arthritis trials was similar to what was used in the current study. The electrical stimulation significantly decreased TNF concentrations in circulating blood and resulted in significant improvement of clinical signs and symptoms according to the standard disease activity and patient disability indices [DAS28-CRP and HAQ-DI, respectively].38,39 In inflammatory bowel disease [IBD], Sinniger et al. reported a 12-month study in nine biologic-naive CD patients with moderately active disease at entry. The same device as used in the current study was implanted on the left vagus nerve and the same vagus nerve stimulation protocol used clinically to treat drug-refractory epilepsy was applied, i.e. 30 s of stimulation every 5 min. The electrical amplitude was titrated according to individual tolerability, and seven patients reached the 52-week visit [two early terminations]. The cytokine levels normalised toward those seen in healthy controls, especially IL-6, IL-12/23, transforming growth factor β1 [TGF-β1], and TNF, and the patients displayed a reduction in clinical activity and endoscopic scores.37 Even with dosing repeated 12 times every hour, vagus nerve stimulation was well tolerated and without serious complications.

본 연구에서 미주신경 자극은 상대적으로 안전한 중재로 나타났다. 16주간의 시험 기간 동안 주요 안전성 신호는 검출되지 않았으며 모든 이상반응은 경증에서 중등도 수준이었다. 한 환자에게서 수술 후 감염이 발생했으나 장치 제거 후 해결되었다.

본 연구의 강점은 대조군이 없었음에도 임상 활동 지표와 삶의 질 지표, 내시경 및 조직학적 평가, 생화학적 및 사이토카인 분석을 결합한 종점의 엄격한 객관적 평가이다. 유효성 신호에 대한 증거는 이러한 다양한 유형의 기기 전반에 걸쳐 대체로 일관성을 보였다. 연구의 한계는 상대적으로 적은 환자 수, 난치성 환자 집단, 그리고 공개 라벨 설계이다. 16주차에 나타난 내시경적 치유의 미미한 정도는 임상적 결과의 확고한 개선 및 대변 칼프로텍틴 감소와 부합하지 않았다. 6개월 후 측정 시점에 진행된 평가였다면 내시경적 치유가 확실히 이루어졌다는 더 강력한 증거를 제공했을 수 있다. 대장 조직학에서 개선이 관찰되지 않은 점은 내시경적 반응과 일관되지 않았으며, 이는 대장 전체 길이에 걸쳐 분산 채취한 생검과 비교하여 말단 회장부의 상대적으로 짧고 명확히 정의된 부위를 대상으로 한 생검 채취 전략 때문일 수 있다.

종합적으로 본 연구는 중등도에서 중증의 활동성 크론병을 가진 생물학적 요법 불응성 환자에서 미주신경 장치 이식 및 일일 수분간 전기 자극이 일반적으로 안전하고 내약성이 양호함을 입증하였다. 미주신경 자극 치료는 임상적으로 의미 있는 질환 활동도 감소를 초래하였으며, 이에 따른 삶의 질, 대장 염증, 분변 칼프로텍틴 및 혈청 사이토카인 수치의 개선이 동반되었다. 이 연구는 소규모 환자군을 대상으로 수행되었으므로 결과 해석에 신중을 기해야 하며, 더 큰 규모의 이중맹검 대조 임상 연구가 필요함을 유의해야 한다.

In contrast to the high number of daily doses delivered to the vagus nerve in the treatment of epilepsy and in the prior CD study, we restricted the electrical stimulation in the current study to just one to four times daily in sessions lasting 1–5 min. The anti-inflammatory effect of the limited periods of electrical stimulation in our study is supported by previous reports on durable resolution of inflammation from a short electrical stimulation of the vagus nerve. Translational evidence for an anti-inflammatory effect lasting 24–48 h comes from studies with α7 nicotinic ACh receptor agonism of primary human macrophages exposed to endotoxin in culture, with vagus nerve stimulation in mouse endotoxaemia, and with human rheumatoid arthritis.38,39,62,63 Importantly, rats injected with indomethacin, which causes small bowel mucosal inflammation, were protected for up to 48 h after a 60-s electrical stimulation of the vagus nerve.34 Limiting the frequency, duration, and strength of electrical charges will mitigate potential off-target effects [such as hoarseness and discomfort] caused by contraction of laryngeal muscles during stimulation. Less frequent stimulation also reduces the energy use, which can allow for smaller batteries and devices, thereby facilitating implantation and clinical usability.64,65

인간을 대상으로 한 관찰 연구는 장 염증에서 미주신경과 부교감신경 긴장의 역할을 뒷받침한다. 한 연구에서는 건강한 대조군에 비해 크론병 환자에서 심박 변이도로 측정된 미주신경 긴장도가 낮게 나타났으며, 다른 연구에서는 휴식 시 미주신경 긴장도가 높은 환자가 순환 TNF 수치가 낮음을 보여주었다.59,60 또한 궤양성 대장염 환자에서 질병 악화기 동안 높은 부교감신경 활동과 3년간 추적 관찰 기간 동안 낮은 전신 염증 사이의 연관성이 관찰되었다. 61 또한 미주신경절제술 병력과 크론병 발병 사이의 연관성도 보고된 바 있다.58

사이토카인 바이오마커 억제를 보고한 최초의 이식형 미주신경 자극 장치에 대한 인간 연구는 간질 치료를 위한 미주신경 자극기 이식 시 전신 마취 상태의 7명의 환자를 대상으로 수행되었다. 데이터에 따르면, 수술 중 단일 자극 후 환자로부터 채취한 혈액에서 내독소 유발 TNF 생산이 현저히 감소한 것으로 나타났다.39 염증성 질환에 대한 미주신경 자극의 유망한 효능 결과를 보여주는 여러 소규모 인간 시험이 보고되었는데, 여기에는 다음이 포함된다: 메토트렉세이트 및/또는 다중 생물학적 제제에 내성이 있는 활동성 류마티스 관절염 환자 17명을 대상으로 한 공개 다기관 시험; 다중 생물학적 제제 및/또는 표적 합성제제에 내성을 보이는 활동성 류마티스 관절염 환자 14명을 대상으로 한 이중맹검 다기관 임상시험; 그리고 크론병 환자 9명을 대상으로 한 공개단일기관 임상시험.36–39 류마티스 관절염 임상시험에서 사용된 미주신경 자극 용량은 본 연구에서 사용된 것과 유사했다. 전기 자극은 순환 혈액 내 TNF 농도를 현저히 감소시켰으며, 표준 질환 활동도 및 환자 장애 지수(각각 DAS28-CRP 및 HAQ-DI)에 따라 임상 징후 및 증상이 현저히 개선되었습니다.38,39 염증성 장 질환(IBD)에서 Sinniger 등은 생물학적 제제 미경험 중등도 활동성 CD 환자 9명을 대상으로 한 12개월 연구를 보고했습니다. 본 연구와 동일한 장치를 좌측 미주신경에 이식하고, 약물 불응성 간질 치료에 임상적으로 사용되는 것과 동일한 미주신경 자극 프로토콜(즉, 5분마다 30초간 자극)을 적용했습니다. 전기 진폭은 개인별 내약성에 따라 조절되었으며, 7명의 환자가 52주 방문까지 도달했습니다(2명 조기 종료). 사이토카인 수치는 건강한 대조군 수준으로 정상화되었으며, 특히 IL-6, IL-12/23, 변형성장인자 β1(TGF-β1), TNF에서 두드러졌다. 환자들은 임상적 활동도와 내시경 점수에서 감소를 보였다.37 1시간에 12회씩 반복 투여하더라도 미주신경자극은 잘 견디며 심각한 합병증 없이 진행되었다.

간질 치료 및 이전 CD 연구에서 미주신경에 전달된 높은 일일 투여량과 달리, 본 연구에서는 전기 자극을 하루 1~4회, 세션당 1~5분으로 제한하였다. 제한된 기간의 전기 자극이 보여준 항염증 효과는 미주신경에 대한 짧은 전기 자극으로 인한 염증의 지속적인 해결에 관한 기존 보고들에 의해 뒷받침된다. 24~48시간 지속되는 항염증 효과에 대한 전환적 증거는 배양된 인간 1차 대식세포에 내독소를 노출시킨 α7 니코틴성 아세틸콜린 수용체 작용제 연구, 생쥐 내독소혈증에서의 미주신경 자극 연구, 그리고 인간 류마티스 관절염 연구에서 비롯된다.38,39,62, 63 특히, 소장 점막 염증을 유발하는 인도메타신을 주사한 쥐에서 미주신경 60초 전기 자극 후 최대 48시간 동안 보호 효과가 관찰되었다.34 전기 자극의 빈도, 지속 시간 및 강도를 제한하면 자극 중 후두근 수축으로 인한 잠재적 비표적 효과[예: 쉰 목소리 및 불편감]를 완화할 수 있다. 자극 빈도를 낮추면 에너지 소비도 감소하여 더 작은 배터리와 장치를 사용할 수 있게 되어 이식 및 임상적 활용성을 높일 수 있다.64,65

Neuroimmune Modulation Through Vagus Nerve Stimulation Reduces Inflammatory Activity in Crohn’s Disease Patients: A Prospectiv

In this study, stimulating the vagus nerve appeared to be a relatively safe intervention. No major safety signals were detected during the 16-week trial and all adverse events were mild to moderate. One patient developed postoperative infection that was resolved after explantation of the device.

The strengths of this study are the rigorous objective assessment of the endpoints despite lack of a control group with a combination of clinical activity and quality of life indices, endoscopic and histological eval‎uation, and biochemical and cytokine analyses. The evidence for an efficacy signal was generally consistent across these different types of instruments. The limitations of the study are the relatively small number of patients, the refractoriness of the populations, and the open-label design. The modest extent of endoscopic healing at Week 16 did not match the robust improvement in clinical outcome and decrease in faecal calprotectin. A later measurement, perhaps at 6 months, might have provided stronger evidence of robust endoscopic healing. The lack of improvement in colonic histology was inconsistent with the endoscopic response, and may be due to the biopsy-sampling strategy employed in combination with a relatively short and clearly defined segment of terminal ileum, compared with the dispersed biopsies from the total length of the colon.

Overall, this study has demonstrated that device implantation and electrical stimulation of the vagus nerve for several minutes per day was generally safe and well tolerated in biologic-refractory patients with moderately to severely active CD. Vagus nerve stimulation treatment resulted in a clinically meaningful reduction of clinical disease activity, with associated improvements in quality of life, colonic inflammation, and levels of faecal calprotectin and serum cytokines. It should be noted that with this small study population, these results should be interpreted with caution and a larger, double-blinded, controlled clinical study is warranted.

Supplementary Material

jjad151_suppl_Supplementary_Material

Click here for additional data file. (653.4KB, pdf)

Acknowledgments

We thank the additional investigators, the gastroenterology and neurosurgical site staff, and the patients for participation in this clinical trial.

Contributor Information

Geert D’Haens, Department of Gastroenterology and Hepatology, Amsterdam UMC, Amsterdam, The Netherlands.

Michael Eberhardson, Department of Medicine, Karolinska Institutet, Solna, Sweden; Department of Health, Medicine and Caring Sciences, Linköping University, Linköping, Sweden.

Zeljko Cabrijan, Division of Gastroenterology, Hepatology and Clinical Nutrition, University Hospital Dubrava, Zagreb, Croatia; Division of Gastroenterology, University of Applied Health Sciences, Zagreb, Croatia; Josip Juraj Strossmayer University of Osijek School of Medicine, Osijek, Croatia.

Silvio Danese, Department of Gastroenterology and Endoscopy, IRCCS Ospedale San Raffaele, Italy; Department of Gastroenterology and Endoscopy, University Vita-Salute San Raffaele, Milano, Italy.

Remco van den Berg, Department of Gastroenterology and Hepatology, Amsterdam UMC, Amsterdam, The Netherlands.

Mark Löwenberg, Department of Gastroenterology and Hepatology, Amsterdam UMC, Amsterdam, The Netherlands.

Gionata Fiorino, Department of Gastroenterology and Digestive Endoscopy, VIta-Salute San Raffaele Hospital, Milan, Italy; IBD Unit, Department of Gastroenterology and Digestive Endoscopy, San Camillo-Forlanini Hospital, Rome, Italy.

P Richard Schuurman, Department of Neurosurgery, Amsterdam UMC, Amsterdam, The Netherlands.

Göran Lind, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden; Department of Neurosurgery, Karolinska University Hospital, Stockholm, Sweden.

Per Almqvist, Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden; Department of Neurosurgery, Karolinska University Hospital, Stockholm, Sweden; Neurosurgery Stockholm AB, Stockholm, Sweden.

Peder S Olofsson, Department of Medicine, Solna, Karolinska Institutet, Karolinska University Hospital, Stockholm, Sweden; Feinstein Institutes for Medical Research, Manhasset, New York.

Kevin J Tracey, Feinstein Institutes for Medical Research, Manhasset, New York; Department of Neurosurgery, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, USA; Department of Molecular Medicine, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, USA.

Stephen B Hanauer, Division of Gastroenterology and Hepatology, Northwestern University–Feinberg School of Medicine, Chicago, Illinois, USA.

Ralph Zitnik, SetPoint Medical, Valencia, California, USA; Valerio Consulting, Santa Barbara, California, USA.

David Chernoff, SetPoint Medical, Valencia, California, USA.

Yaakov A Levine, Department of Medicine, Karolinska Institutet, Solna, Sweden; Department of Molecular Medicine, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Hempstead, New York, USA; SetPoint Medical, Valencia, California, USA.

Funding

This work was supported by SetPoint Medical, Inc.

Conflict of Interest

GDH: consulting and/or lecture fees from AbbVie, Alimentiv, Boehringer Ingelheim GmbH, BMS, Eili Lilly, Ferring, Galapagos, GlaxoSmithKline, Johnson and Johnson, Pfizer, Takeda, Tillotts Pharma,, Versant; research grants from Pfizer, Takeda, Eli Lilly; and speaking honoraria from AbbVie, Tillotts, Pfizer and Johnson and Johnson. ME: Honoraria and consultancy fees from AbbVie, Merck [MSD], Takeda, Ferring, Orion Pharma, Otsuka, Tillotts, Novartis, Pfizer, Galapagos, Bristol Myers Squibb, and Janssen; research grants from AbbVie and MSD. ZC: nothing to report. SD: speaker, consultant, and advisory board member for Schering-Plough, AbbVie, Actelion, Alphawasserman, AstraZeneca, Cellerix, Cosmo Pharmaceuticals, Ferring, Genentech, Grunenthal, Johnson and Johnson, Millenium Takeda, MSD, Nikkiso Europe GmbH, Novo Nordisk, Nycomed, Pfizer, Pharmacosmos, UCB Pharma, and Vifor. RvdB: nothing to report. ML: speaker and/or principal investigator for: Abbvie, Alimentiv, Bristol Myers Squibb, Celgene, Covidien, Dr Falk, Ferring Pharmaceuticals, Galapagos, Gilead, GlaxoSmithKline, Janssen-Cilag, Medtronic, Merck Sharp & Dohme, Pfizer, Protagonist therapeutics, Receptos, Takeda, Tillotts, Tramedico; research grants from AbbVie, Merck Sharp & Dohme, Dr Falk, Achmea Healthcare, Galapagos, and ZonMW. GF: personal fees: Takeda, Abbvie, Janssen, Pfizer, Celltrion, Sandoz, AlfaSigma, Samsung Bioepis, Amgen, Roche, Ferring, Mylan, Gilead, Galapagos. PRS: nothing to report. GL: nothing to report. PA: nothing to report. PSO: a shareholder of Emune AB. KJT: shareholder of SetPoint Medical. SH: consultant for AbbVie, Allergan, Amgen, Arena, Bristol-Myers Squibb, Celgene, Celltrion, Genentech, Gilead, GSK, Janssen, Lilly, Merck, Nestle, Novartis, Pfizer, Progenity, Prometheus, Receptos, Salix, Samsung Bioepis, Seres Therapeutics, Takeda, TiGenix, UCB Pharma, and Vhsquared; speaker for AbbVie, Janssen, and Takeda. RZ: shareholder of SetPoint Medical. DC: employee and shareholder of SetPoint Medical. YAL: employee and shareholder of SetPoint Medical.

Author Contributions

Concept or design of the work: GDH, ME, ZC, SD, KJT, SBH, RZ, YAL. Data collection: GDH, ME, ZC, SD, RvdB, ML, GF, PRS, GL, PA, SBH, RZ, YAL. Data analysis and interpretation: GDH, ME, ZC, SD, PSO, KJT, SBH, RZ, DC, YAL. Drafting the article: GDH, ME, PSO, KJT, RZ, DC, YAL. Critical revision of the article: GDH, ME, ZC, SD, RvdB, ML, GF, PRS, GL, PA, PSO, KJT, SBH, RZ, DC, YAL. All authors had full access to all the data and had final responsibility for the decision to submit for publication.