A proof of concept study establishing Necator americanus in Crohn's patients and reservoir donors

The emergence of autoimmunity, including Crohn's disease (CD) where the immune relationship with commensal bacteria is corrupted, has been linked to hygiene.¹,² A gradual decline in endoparasites is but one argument that might explain this phenomenon.³ Weinstock and colleagues have successfully tested the pig whipworm, Trichuris suis, in patients with inflammatory bowel disease (IBD).⁴,⁵ However, repeated inoculation was required and concern has been raised that aberrant migration could occur.⁶ The haematophagous hookworm, Necator americanus (NA), is proposed as an alternative. We have tested if CD patients tolerate hookworm infection, and the practical issues associated with establishing reservoir donors (RDs).

Over 700 million people remain infected with hookworms. Infective larvae (L3i) are acquired through skin contact with contaminated soil.⁷ Auto‐reinfection, direct person to person infection, aberrant migration, and hypobiosis do not occur. Adult worms live in the host small intestine for an average of five years. Infection can be easily terminated with an anthelminthic. Anaemia is the only disease of consequence but is an unusual outcome in properly nourished individuals. Using L3i originally obtained from Madang, Papua New Guinea, but maintained in a healthy researcher in the UK, five CD subjects with longstanding but mostly inactive disease and three RDs each received a carefully measured inoculum (table 1). Subsequently, four additional CD subjects with chronic and mostly active disease were inoculated with L3i cultured from faeces provided by an RD, and the original CD cohort were reinoculated from week 27 to week 30. Ethics approval was granted by the Townsville Health Service District Institutional Ethics Committee. Haematological and clinical measurements are expressed as mean (95% confidence interval).

Table 1 Crohn's disease activity index (CDAI) in CD subjects inoculated with infective larvae (L3i). Subsequently, the five CD subjects first inoculated were reinoculated from week 27 to week 30 and four CD subjects with chronic and mostly active disease were inoculated with larvae sourced from one of the authors.

ID, age (y), sex	Initial inoculation trial					Reinoculation trial
	Time (weeks)					Time (weeks)
	0–4	5–8	9–12	13–16	17–20	27	30	35	39–41	45
CD1 55 M
Inoculum therapy	25 L3i					25 L3i
CDAI	79	60	89	77	68		96		93	62
CD2 46 M
Inoculum therapy	25 L3i					25 L3i
CDAI	38	114	20	68	48		30			36
CD3 41 F
Inoculum therapy	25 L3i P5 M15	P5 M15	P5 M7.5			25 L3i P5 M10	P5 M10	P5 M10	M10	M10
CDAI	46	71	85	83	90		36	30	92	95
CD4 34 M
Inoculum therapy	50 L3i P38 M30	P38 M30	P50 M30	P25 M30	P5 M30	50 L3i P5 M20	P2.5 M20	M20	P25 M30	P20 M30
CDAI	260	230	232	264	118		60	442	60	122
CD5 21 F
Inoculum therapy	50 L3i P10 M20	P10 M20	P13 M20	P10	P7.5	P5	25 L3i P5	P25 M20	P15 M20	P10 M20
CDAI	144	151	103	79	73		410	180	44	61
CD6 33 F
Inoculum therapy	50 L3i P15 M20	P10 M20	P5 M20	P5 M20	P7.5 M20
CDAI	49	32	6	29	100
CD7 33 M
Inoculum therapy	50 L3i A150	P25 A150	P5 A150	P5 A150	P8 A150
CDAI	260	114	96	125	118
CD8 46 M
Inoculum therapy	50 L3i M20	M20	M20	M20	M20
CDAI	145	159	171	152	186
CD9 44 F
Inoculum therapy	100 L3i P5 M20	P10 M20	P5 M20	P5 M20	P10 M20
CDAI	173	127	106	76	125

L3i, n 3rd stage N americanus larvae inoculated percutaneously; P, prednisone n mg/day; A, azathioprine n mg/day; M, methotrexate n mg/week.

The inoculation caused a mild itch within five minutes that disappeared after a few days in eight CD subjects and a pruritic rash that lasted two weeks in the RDs, who also developed a painful transient enteropathy. Neither respiratory symptoms nor detectable aberrant migration occurred. In the CD cohort, blood eosinophilia developed from week 5 (mean 2.60×10⁹/l (1.89) v week 1 0.18×10⁹/l (0.10) v week 20 0.59 (0.20)). Patent infection had established by week 20 in all cases. CD activity index (CDAI) remained unchanged until week 17, possibly in part due to a hookworm related enteropathy recognisable because of blood eosinophilia and faecal Charcot‐Leydon crystals.⁸ After 20 weeks, the IBD questionnaire was improved (mean 151 (14) v 179 (20)) and the four week cumulated CDAI scores was decreased (mean 141 (31) v 87 (15)).⁹ Haemoglobin fell marginally (week 1 mean 135.6 (7.8) g/l v week 20 129.3 (4.1) g/l). Reinoculation of the five CD subjects first exposed caused no apparent adverse effect. Disease reactivation, as defined by a CDAI >150, occurred in two (CD4, CD5; table 1) after the doses of long term immune suppressive drugs had been reduced. The subject (CD3–7) driven trend was to reduce immune suppression as health improved, a strategy often associated with worsening of symptoms. The five CD subjects first inoculated were in remission at week 45 (fig 1).

Figure 1 Initial Crohn's disease activity index (CDAI) score for each CD patient versus score at week 20 and at week 45 for the first five inoculated cases (mean 165 (95% confidence interval 145) v 64 (25), p = 0.132; mean 165 v 75 (29), p = 0.246).

Our pilot study has established a potential for NA, already a fact of life for many millions, as a candidate parasite to inoculate those with autoimmune disease. The natural advantages are lifecycle and migration predictability, ability to control the size of and eliminate a colony, and the parasite's longevity. Inoculation proved safe, even in immune suppressed patients. Our hope that NA would suppress autoreactivity sufficiently to allow immune suppressive therapy to be stopped was unrealistic. Recent and compelling evidence has shown that IBD is self sustaining.¹⁰ It may be that after remission is achieved, endoparasites will offer an alternative or adjunct to immune suppressive therapy, a priority for some people with CD.