Wound botulism,its neurological manifestations,treatment and outcomes: A case series from the Glasgow outbreak,2015

Abstract

Background and aims

We examined the neurological manifestations, treatment and outcomes of a subset of 25 patients within the largest ever outbreak of wound botulism in Europe.

Methods and results

All 25 cases were intravenous drug users. The most common presenting symptom was dysarthria in 19/25 (76%), followed by dysphagia in 12/25 (48%), blurred vision in 10/25 (40%) and double vision in 8/25 (32%). Microbiological analysis confirmed the diagnosis in nine cases (36%). Duration of admission positively correlated with time to antitoxin, time to wound debridement and female sex.

Conclusion

As the outbreak continued, hospital stays shortened, reflecting growing awareness of the outbreak and quicker treatment initiation.

Keywords

Wound botulism dysarthria muscle-popping

Introduction

Botulism is a rapidly evolving, symmetrical, descending flaccid paralysis caused by the action of botulinum neurotoxin (BoNT) at the neuromuscular junction.¹ This potent neurotoxin is produced by the anaerobic bacterium, Clostridium botulinum. There are seven, antigenically distinct types of BoNT, labelled A to G. Only types A, B, E and F cause illness in humans (most commonly A and B) with toxin A considered to be the most potent.²

Botulism usually occurs through one of three pathways:

Food borne, following ingestion of preformed BoNT in food contaminated with C. botulinum.

Wound botulism, where BoNT is produced in a C. botulinum infected wound.

Intestinal colonisation, whereby neurotoxic clostridia grow and produce BoNT in the gut, usually in infants under 12 months old.²

BoNT binds specifically to receptors in the presynaptic cell membrane of voluntary and involuntary cholinergic neurons, preventing synaptic vesicle fusion and inhibiting release of the neurotransmitter acetylcholine into the synaptic cleft.^3,4 This gives rise to a characteristic descending paralysis with autonomic system involvement.

Owing to its rarity, there are a limited number of studies evaluating the clinical care of patients with wound botulism. In particular, the evidence for the use of anti-toxin in wound botulism is limited to one study.⁵ A recent Cochrane review mainly focused on food-borne botulism, for which there is a greater evidence base, and did not comment on debridement as a treatment for wound botulism.⁶ Our case series contributes to the limited published data on wound botulism, a rare medical emergency which, even if treated aggressively, has a mortality rate of 10%.⁷ The mortality rate associated with wound botulism has fallen significantly from approximately 70% in the first half of the twentieth century, primarily due to advances in mechanical ventilation.⁸

Methods

A total of 40 cases were notified to the Health Protection Scotland (HPS) between December 2014 and June 2015. This is the largest outbreak of wound botulism ever recorded in Europe.^9–11 Electronic case records of the 25 patients treated within NHS Greater Glasgow and Clyde (NHS GGC) were identified and this subset case series registered with NHS GGC audit authority. The patients were then ordered by date of presentation. Data collected included; patient demographics, injection habits, time to presentation, clinical features on presentation and subsequent management. We also reviewed duration of intensive care admission, total inpatient stay, rate of clinical improvement and results of botulism testing. Public health and microbiological data were also available on the remaining 15 cases who were managed outwith NHS GGC.

The data were analysed using SPSS. Descriptive statistics were produced in addition to correlation coefficients. As the data were non-parametric, Spearman Rank correlation and Mann–Whitney U testing were performed. There were two outliers within the data (not omitted from statistical analysis) – one whose time to wound debridement was 13 days and a second whose admission totalled 120 days.

The patient who underwent wound debridement on day 13 initially presented with a 5-day history of ptosis, diplopia and mild generalised weakness in all four limbs. He received antitoxin and intravenous antibiotics at 24 h, at which point, his injection sites were reviewed. Due to ‘lack of systemic inflammatory response’, the wound was aspirated only. Twelve days later, an ultrasound of the wound revealed an abscess, which was then debrided. The patient self-discharged, but survived.

Results

Table 1 describes the distribution by sex and admitting teams, plus mean age. Females had a higher median duration of admission than males, Figure 1(a), p = 0.016. On average, each patient was discussed with three additional medical and surgical specialities, most commonly intensive care in 19/25 (76%). Table 2 describes the clinical features documented at presentation and immediately prior to anti-toxin administration, revealing typical disease progression.

Table 1.

Demographics and breakdown of managing teams.

Patient demographics	Sex	Male 68%
		Female 32%
	Mean age	43 years old
Managing team	General medicine	64%
	Infectious disease	12%
	General surgery	8%
	Stroke	8%
	ENT	4%
	Plastic surgery	4%

Figure 1.

Box and dot plots of variables against admission duration. (a) Comparison of patient sex against admission duration. (b) Comparison of group 1 (first 12 patients in outbreak) and group 2 (subsequent 13 patients in outbreak). (c) Scatter graph comparing time from presentation to anti-toxin against number of nights inpatient stay. (d) Comparison of patients receiving anti-toxin less than 12 h after presentation versus those receiving it more than 12 h. (e) Scatter graph comparing time from presentation to debridement against number of nights inpatient stay. (f) Comparison of patients who underwent wound debridement at less than 24 h versus those after 24 h.

Table 2.

Clinical characteristics of the 25 cases.

		At presentation	Prior to anti-toxin
Symptoms	Dysphagia	48%	28%
	Blurred vision	40%	16%
	Double vision	32%	16%
	Dyspnoea	32%	36%
Signs	Dysarthria	76%	52%
	Ptosis	32%	64%
	Absent/sluggish pupils	20%	20%
	Ophthalmoplegia	8%	32%
	Facial weakness	4%	12%
	Respiratory failure	8%	36%
Intubation requirements		4%	24%

All 25 patients received anti-toxin and antibiotics, the most common combination being benzyl penicillin and metronidazole. The correlation coefficients and significance levels for the interventions and categories are shown in Table 3. The scatter graphs for time from presentation to anti-toxin and wound debridement are shown in Figure 1(c) and (e). Both positively correlated with total inpatient stay – i.e. the longer the delay in initiating treatment, the longer the hospital admission.

Table 3.

Correlation coefficients of the interventions and categories.

Intervention/category	Spearman correlation with number of nights inpatient stay (significance)
Time from presentation to anti-toxin	0.604 (p = 0.001)
Time from presentation to debridement	0.541 (p = 0.021)
Patient number (ordered by date of presentation)	−0.427 (p = 0.033)

Individuals who received anti-toxin within 12 h of presentation had a shorter median admission duration compared to those who received anti-toxin later than 12 h (Mann–Whitney U test), p = 0.029 (Figure 1(d)). Anti-toxin was given prior to wound debridement in 12 of the 17 debrided (71%).

Botulism was considered within 24 h of presentation in 19/25 (76%). Public Health Protection Unit (PHPU) was notified within 48 h in 20/25 (80%). The first 12 patients to present had longer median admissions compared to the second 13 patients (Mann–Whitney U test), p = 0.026 (Figure 1(b)).

Improvement occurred over days in the majority, 14/25 (56%), and over weeks in 8/25 (32%). It was difficult to assess the speed of recovery in three cases due to additional medical problems.

Two of the Glasgow cases died in addition to two cases outwith NHS GGC (National mortality 10%). This is similar to expected mortality in treated patients.⁷

The first patient to die was the second to present during the outbreak. He presented with shortness of breath and was diagnosed as community acquired pneumonia. He was discharged from A&E only to re-present several hours later with ptosis, dysarthria and dysphagia. He suffered a cardiac arrest whilst on the receiving medical ward. Despite recovery of cardiac output, he never regained consciousness and died after 41 days ventilated in ICU.

The second death occurred in the middle of the outbreak, despite the administration of antibiotics and anti-toxin at 8 h, and was thought to be due to ongoing injection of infected heroin whilst an inpatient. This patient in-fact shared heroin with his wife, who was also treated during the outbreak.

All patients were intravenous drug users and had used heroin, however in some, this was not confirmed until intensive care admission. A significant proportion of patients were thought to have ‘skin popped’ (injected subcutaneously or intradermally, as opposed intravenously) either accidentally or on purpose. Individuals who have ‘skin popped’ have increased odds of developing wound botulism by a factor of more than 15.¹²

Microbiological confirmation of a diagnosis of botulism was possible in 9 of the 25 (36%) Glasgow cases and a further 9 nationally (45% of total population), either by detection of the toxin in serum by bioassay, or by isolation of C. botulinum from debrided tissue by culture and PCR detection of neurotoxin genes.^13,14 Amongst the 25 Glasgow cases, botulism subtype was confirmed by detection of neurotoxin type B in one patient and the detection of C. botulinum type B from tissue or pus samples from eight patients. Confirmation of subtype is important if assuming all cases resulted from a single source. Nationally (including the Glasgow cases), botulinum neurotoxin was detected in the serum of 6 patients (able to be confirmed as type B in 3 patients) and C. botulinum type B in the tissue of 12 patients. The organism was isolated from specimens of 10 patients and molecular typing by fluorescent Amplified Fragment length polymorph showed that all 10 isolates were of the same molecular type, which is consistent with a common source exposure.

Discussion

We describe a subset of patients within the largest outbreak of wound botulism ever recorded in Europe. The likely source of the C. botulinum was a contaminated batch of heroin distributed in, or via, Glasgow. Unfortunately, there have been previous reports of such outbreaks and this is unlikely to be the last.

Wound botulism is a rare disease. Presenting features may be vague and the differential diagnosis can include diseases uncommon to non-neurologists, such as Miller-Fisher variant of Guillian–Barre syndrome, myasthenia gravis or even stroke disease. Microbiological confirmation can take up to 3 weeks. Electromyography (EMG), if available, may be helpful, but is rarely diagnostic. Rather it is the evolution of signs, in an at-risk patient, which is the most useful diagnostic pointer, and of course more than one case of wound botulism in an area is pathognomonic for an outbreak.

In our case series, the most common presenting symptom was dysarthria, with evidence of evolving neurological deficit (over hours) resulting in ptosis, blurred or double vision, difficulty swallowing and eventually respiratory distress. Much is made of autonomic disturbance in botulism, particularly abnormal (absent or sluggish) pupillary responses to light; however, these abnormalities were less well identified from these data. Females appear to have longer duration of admission, perhaps reflecting their smaller body size and injecting habits (more likely to skin pop in order to hide injection sites).

The diagnosis of wound botulism, it should be emphasised, is a clinical one.¹ The patient may have no clinical evidence of an abscess (‘cold’ abscess), remain apyrexial and have normal inflammatory markers.¹ In all current or ‘ex’ intravenous drug users, it is therefore important to have a high degree of suspicion that an evolving bulbar paralysis may be botulism.¹² In these cases, ultrasound or CT imaging can identify sites where the bacteria may be producing toxin, and should guide urgent surgical debridement.

By showing that early treatment was associated with shorter inpatient stays, this case series supports established practice of prompt use of anti-toxin and wound debridement in addition to standard antimicrobials in patients with a clinical diagnosis of wound botulism.^15,16 Our results replicate findings of shorter hospital admissions associated with early administration of anti-toxin in food-borne botulism.¹⁷ We also draw parallels with results published in 2001, which showed duration of mechanical ventilation in patients with wound botulism dropped from a median of 54 days to 11 days if antitoxin was given within 12 h of presentation.¹⁸

Botulism is a notifiable disease in the UK. Notification of the early cases led public health to form a multidisciplinary incident management team (IMT), including representation from Police Scotland, to coordinate investigation of the outbreak. PHPU alerted medical practitioners and the at-risk community, and indeed several patients presented because they knew which symptoms to be aware of. The importance of these interventions is supported by the observation that as the outbreak continued patient’s length of stay in hospital shortened.^9,19

Prompt treatment can lead to good outcomes, as can be seen in this series with patients improving over days and the majority making full recoveries.^2,13,20 It should be noted, however, that botulism can reoccur in the same individual, with no protection offered by previous exposure.²¹

Footnotes

Author Contributions

Sarah-Jane Martin designed the study, collected and analysed data on the 25 patients from NHS GGC and drafted the manuscript with help from George Gorrie. Additional information from Public Health Protection Unit, Scotland, was provided by Gillian Penrice who also reviewed the manuscript. Corrine Amar and Kathie Grant provided details on microbiology testing and results, and also reviewed the manuscript.

Acknowledgements

The authors thank Dr Angus MacLeod for his help with statistical analysis of the data.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Sam

Beynon

HLC

. Wound botulism. New Engl J Med 2010; 363: 2444–2444.

Goonetilleke

Harris

. Clostridial neurotoxins. J Neurol Neurosurg Psychiatry 2004; 75(Suppl 3): iii35–39.

Ahnert-Hilger

Münster-Wandowski

Höltje

. Synaptic vesicle proteins: targets and routes for botulinum neurotoxins. Curr Top Microbiol Immunol 2013; 364: 159–177.

Rossetto

Pirazzini

Montecucco

. Botulinum neurotoxins: genetic, structural and mechanistic insights. Nat Rev Microbiol 2014; 12: 535–549.

Burningham

Walter

et al.

Wound botulism. Ann Emerg Med 1994; 24: 1184–1187.

Chalk

Benstead

Keezer

. Medical treatment for botulism. Cochrane Database Syst Rev 2014; 2: CD008123–CD008123.

Anderson

et al.

Wound botulism associated with black tar heroin. Acad Emerg Med 1997; 4: 805–809.

Sobel

. Botulism. Clin Infect Dis 2005; 41: 1167–1173.

MacDonald

Arnesen

et al.

Outbreak of wound botulism in people who inject drugs, Norway, October to November 2013. Eurosurveillance 1996–2016 2013; 18: 20630–20630.

10.

Alpers

van Treeck

et al.

Outbreak of wound botulism in injecting drug users in Germany, October-December 2005. Eurosurveillance 2005; 10: E051215.4–E051215.4.

11.

European Centre for disease control (ECDC) and European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) release joint rapid risk assessment on wound botulism among people who inject heroin. Wound botulism in people who inject heroin in Norway and the United Kingdom. Stockholm/Lisbon 16.02.2015.

12.

Gordon

et al.

Bacterial infections in drug users. New Engl J Med Rev 2005; 353: 1945–1954.

13.

Cherington

. Clinical spectrum of botulism. Muscle Nerve 1998; 21: 701–710.

14.

Lindström

Korkeala

. Laboratory diagnostics of botulism. Clin Microbiol Rev 2006; 19: 298–314.

15.

Offerman

Schaefer

et al.

Wound botulism in injection drug users: time to antitoxin correlates with intensive care unit length of stay. Western J Emerg Med 2009; 10: 251–256.

16.

Chang

Ganguly

. Early antitoxin treatment in wound botulism results in better outcome. Eur Neurol 2003; 49: 151–153.

17.

Tacket

et al.

Equine anti-toxin use and other factors that predict outcome in type A food borne botulism. Am J Med 1984; 76(5): 794–798.

18.

Sandrock

Murin

. Clinical predictors of respiratory failure and long-term outcome in black tar heroin-associated wound botulism. Chest 2001; 120: 562–566.

19.

Schroeter

Alpers

et al.

Outbreak of wound botulism in injecting drug users. Epidemiol Infect 2009; 137: 1602–1608.

20.

Terranova

Palumbo

Breman

. Ocular findings in botulism type B. J Am Med Assoc 1979; 241: 475–477.

21.

Yuan

Inami

et al.

Recurrent wound botulism among injection drug users in California. Clin Infect Dis 2011; 52: 862–866.