Does diisocyanate exposure result in neurotoxicity?

Figures & data

Fig. 1. Chemical structure of diisocyanates.

Table 1. Evidence summary of reports meeting inclusion criteria.

Reference & Study type	Study size	Diisocyanate exposure	Co-exposure	Subjective symptoms	Objective findings	Diagnostic studies	Author's conclusions	Analysis
Exposure to Diisocyantaes Only
Mastromateo 1965 Case Series	Twenty-four workers cleaning pipes and vessels for TDI manufacturing facility. Twelve not wearing proper PPE developed symptoms over 4-to 5-day period	TDI—Dermal and Respiratory	Only TDI was mentioned. Work was being done because system became contaminated (not specified)	Primary: 12/24 with upper or lower respiratory complaints. Secondary: (1/12)”acute psychosis” 1/12; 4/12 with anxiety neurosis with strong overtones of psychosomatic complaints	6 of 12 hospitalized for various respiratory symptoms, including pneumonia. 1 worker diagnosed with delusional psychosis - 48 hours after given corticosteroid therapy, resolved in 24 hours after medications stopped	Not mentioned in case report	Workers with proper PPE had no complaints. Noted anxiety with respiratory distress is common. “Highly motivated workers were able to return to full time employment. Workers with little or no motivation, in whom no permanent physical impairment could be found, have not yet returned to full employment”	Analysis of diagnosis and treatment is difficult due to sparse details and must be considered in context of medical system of 1965 Canada. Symptoms of anxiety are non-specific. No definition of “acute psychosis”. Likely that exposure included other contaminants of TDI manufacturing process
Singer 1987 Case Series	Three wharf workers	TDI—punctured drum—Dermal and Respiratory	Not specified	Neurasthenia, emotional lability, headaches, depression, irritability, forgetfullness, disorientation, decreased concentration, hypesthesia stocking/glove	One subject had mild sensory peripheral neuropathy (bilateral median nerves), another had mild sural nerve sensory neuropathy. IQ testing showed progressive decrement ranging from 20–26 points between 2-month and 16-month period	Neurobehavioral testing at 2 and 16 months. Different tests used at intervals. WAIS score at 2 mos, WAIS-R at 16 mos. Nerve conduction studies	Progressive decrement of mental function due to “(a) gradual death of brain tissues that was injured by the initial exposure, (b) the release of TDI stored in body fat as the fat was mobilized…(c) the toxicity of TDI metabolites when they are cleared from the blood, liver, or other organs”	Author incorrectly infers toxic effects of TDI are from toluene and cyanates. Author used different testing methods between intervals, WAIS and WAIS-R scales not fully comparable. Litigation effect may have been present. Author used unvalidated malingering assessment to determine litigation effect unlikely
Mixed Exposure to Diisocyanates and Other Chemicals
Thrasher 1989 Case Series	Fifteen workers in two small offices with complaints after office was remodeled	TDI—adhesive source from remodeling	Volatile organic chemicals (VOCs), formaldehyde (F), trimellitic anhydride (TMA) as “reactive chemicals” associated with building related illness	All subjected reported multiple symptoms involving mucous membranes, lungs, musculoskeletal, and CNS (headaches, memory problems), chronic malaise/fatigue, nausea, dizziness, irritability, altered olfaction (increased sensitivity to chemicals, smells). 100% reported headaches, memory difficulties, malaise. 87% reported flu-like illness	No over exposures found on air sampling (conducted 2-years after construction). TDI was below limits of detection. Positive antibody to albumin conjugate of F-IgE: 4/12, F-IgG 11/12; TDI-IgE: 5/12, TDI-IgG 9/12. Correlation of symptoms to antibody titers: F-IgE (r = .33,p > 0.05) F-IgG (r = .24, p > 0.05), TDI-IgE (r = .48, p > 0.05), TDI-IgG (r = .48, P < 0.05)	Building-related illness questionnaire; albumin conjugates of formaldehyde, TDI, trimellitic anhydride (TMA), VOCs; IgE and IgG antibodies to same substances	“Immunologic mechanisms may explain the existence of hypersensitivity pneumonitis, allergic rhinitis, asthma, and skin eruptions that have been reported in tight building syndrome”	Small sample size with no control population. Data suggest poor correlation of immunologic testing with symptoms. No objective physical findings in this population, yet author infers data may explain objective findings in other reports of “tight building syndrome”
Müeller 1989 Case Series	Forty workers in East Germany in polyurethane production	Diisocyanates	Dimethyl formamide, trichloromethane, phosphoric acid esters, trichlorofluormethane, tertiary amines, organic tin-compounds. Most with exposures for > 5 years	Nine workers (22%) with nonspecific neurasthenic symptoms. Nine workers (22%) with symptoms of peripheral neuropathy	EEG—No differences from age adjusted norms. PNF—11 with abnormalities (neurasthenia). MMA 11/40 with abnormalities (27.5%) EMG—11 with mild NCV slowing peripherally. PNF—11 with abnormalities	Physical Exam, ESR, Blood Count, liver enzymes, urine, respiratory evaluation, EMG, EEG, Mean momentary Arrhythmia (MMA) electrocardiogram(MMA and neurospychological testing (PNF)	Findings considered nonspecific, as neuropsychological testing results may have been explained by age dependent decrease rather than toxic encephalopathy. Recommend against use of EEG for screening. Symptoms of neurasathenia, peripheral neuropathy may be explained by the co-exposures. No specific findings with diisocyanates	Workers in the polyurethane plant have multiple potential exposures. No specific exposure data presented on any substance. Data suggest no specific role of diisocyanates and neurotoxic effects
Reidy 1994 Case Series	Five workers allegedly exposed to MDI over 2-year period—facility type or industrial process not described	MDI—No workplace monitoring data, lack of details on task-specific exposure circumstances	Hydrocarbon vapors	All subjects diagnosed with occupational asthma (surrogate for exposure measures). “All… reported high incidence of vague subjective complaints…such as headaches, mood alterations, forgetfulness and decreased concentrations”	No deficits in Neuropsychological testing for: psychomotor, psychosensory, visuographic, language functions, mental efficiency, rate of information processing, learning ability and abstract reasoning	Neuropsychological testing	“[p]resent results do not clearly identify a single pattern of neuropsychological deficits associated with MDI exposure, the data do suggest the presence of compromised cognitive functions characteristic of CNS involvement”	All subjects unemployed, testing obtained by litigating attorneys. Unknown number of co-workers did not seek testing. Exposure suggested by diagnosis of occupational asthma, although no supporting data presented. Overall, data suggest limited. Conclusion of toxic encephalopathy from MDI with no exposure data and co-exposure to hydrocarbon vapors in undescribed process is speculative
Herbert 1995 Case Report, Series	Single subject as sentinel case, 40 union workers in prevalence study	HDI	Multiple solvents including toluene, xylene, and n-hexane	Case report—light-headedness, loss of balance, headache, irritability, fatigue, and symmetrical paraesthesias in feet and hands	Case—Symmetrical distal mixed polyneuropathy demonstrated by NCV. Prevalence—great toe vibrotactile threshold testing 42% abnormal dominant, 36.4% nondominant toe in roofers (p < 0.001)	Tactile Threshold testing	“The results of this series of roofers suggest a previously unreported health hazard for roofing workers. [W]e maintain the results of this series should be considered as hypothesis generating, at best”	Exploratory study to determine prevalence of abnormal tactile response. Possible selection bias of union workers (15% of eligible population). Data suggest higher percentage of abnormal vibrotactile threshold testing. No correlation with exposure data. Data does not suggest causative link with diisocyanates. Abnormalities most likely associated with organic solvents
Nijem 2001 Cross Sectional Survey	One hundred and sixty-seven workers nonrandomly selected from 20 shoe factories in Palestine	Diisocyanates—nonspecified (use in curing of soles)	n-Hexane, toluene, dichloromethane, polyvinyl chloride	Neurasthenia, parethesias, sore eyes, breathing, difficulties	Prevalence Ratio: > 1 month exposure compared to no exposure (from process involving diisocyanate and n-hexane): Sore eyes PR = 1.7 (1.1–2.7), tingling PR = 1.8 (1.2–2.9) No significant association with headache, breathing difficulty, mental irritability	Q16 Swedish neuropsychiatric symptom questionnaire	“Cumulative exposure in plastic molding was associated with sore eyes and tingling of limbs. Workers in this task could be exposed to diisocyanates, a strong irritant compound that produces eye and airway inflammation and to different solvents such as n-hexane that are responsible for toxic neuropathy”	Participants were recruited from factories in Palestine. Factory owners selected participants, introducing large bias. No exposure data. Analysis of symptoms by job task (and thus exposure) suggest no neurologic effects from diisocyanate, but suggests association with n-Hexane and other solvents
Moshe 2002 Case Report	Single Subject—61-y/o artist with 30 years painting in home studio	TDI	Toluene, xylene, benzene, methyl ethyl ketone, acetone, thinner	Intention tremor, parasthesia hands and feet, memory and concentration difficulties	Peripheral neuropathy, central neuropathy, ototoxic hearing loss, brakykinesis, cognitive impairment, attention and memory problems, decreased executive function compatible with solvent intoxication	Scalp somatoseneory evoked potentials, nerve conduction studies, neurobehavioral testing (limited battery of tests)	“The artist…suffered from sensory-motor peripheral polyneuropathy that was not explained by his diabetes (or alcohol, drugs, head trauma, or Alzheimer's). These exclusions supported a diagnosis of solvent-induced neuropathy”	Type of Diabetes or treatment details missing. No data on TDI or extent of exposure. Description of painting method indicates heavy exposure to organic solvents listed. Data suggest in this individual symptoms and objective findings most consistent with organic solvent peripheral neuropathy and encephalopathy
Exposures to Diisocyanates and Thermal Decomposition Products of Polyurethane
McKerrow 1970 Axford 1976 LeQuesne 1976 Three case-series reports of the same firefighter population with 4-year follow-up	36 Firefighters in 1967 mattress manufacturing fire LeQuesne reports comparison with population of 15 fireman McKerrow reports Group 1:36 Fireman Group 2: 25 Factory workers	TDI—leaking vessel from fire, 1000 gallon spill	Pyrolysis products of polyurethane foam mattress factory—over 100 chemicals(Paabo 1987) McKerrow—Group 1: Pyrolysis products of polyurethane foam mattress factory; Group 2: Non-descript industrial process	LeQuesne: 23/36 reported “neurological” symptoms at some time during 4-year-period post fire. 5 men with acute N/V, euphoria (drunk) during fire. 14 men with variety of sx such as headache, difficulty concentrating, poor memory, irritability or depression at 3 weeks, 13 with some symptoms at 4 years McKerrow: Group 1: Acute—6/36 (two lost consciousness), euphoria, ataxia. Delayed/chronic: late memory and concentration difficulties in 14/36	LeQuesne Report Examination: 6/36 slight ataxia of gait, 1/36 with subjective parasthesia (pinprick). EEG normal. At 4-yr follow-up, 1 man with persistent sensory impairment, no other physical findings. Memory testing showed difference in long term recall mean score 33.2 vs. 27.0 in control vs. exposed and still affected (p < 0.02), no difference in control vs. all exposed p > 0.1) McKerrow Report: No objective results provided	LeQuesne Report: Wechsler memory Scale. EEG, Physical Exam. CXR and Spirometry for pulmonary effects McKerrow Report: Not specified	LeQuesne Report: “Although proof that TDI was responsible for the neurological syndrome described is lacking, it appears that the men affected had in common a greater exposure to TDI than those unaffected” McKerrow Report: “Neurological symptoms including euphoria and ataxia occurred acutely in a few men and difficulty in concentration and a memory defect persisted in 14”	Lack of baseline comparison data for controls presented, therefore may not be representative. Significant age difference apparent in control (younger). Data suggest subset of exposed firefighters with persistent subjective symptoms, but no significant support for TDI exposure vs. other fire related exposures. Pyrolysis of polyurethane foam may release HCN, CO, benzonitrile, acryolonitrile, and many other organic solvents that are more consistent with symptoms described, including acute euphoria

Table 2. Mathematical model for probability of causality.

Hill's Criterion	Evidence Summary	Probability (%) of criterion being true	Product of discriminant function^± and probability, (C1)	Product of discriminant function^± and probability, (C2A)
Constant			− 14.7799	− 10.0835
1. Strength	One study (Nijem) presented relative risk ratio of 1.7*	60	3.7338 (0.06223 × 60)	1.1538 (0.01923 × 60)
2. Consistency	Studies varied in symptoms and findings**	50	2.0305 (0.04061 × 50)	0.9015 (0.01803 × 50)
3. Specificity	No findings specific to diisocyanates	40	− 1.1148 (− 0.02787 × 0)	− 1.5508 (− 0.03877 × 0)
4. Temporality	All case reports preceded by diisocyanates exposure	100	7.657 (0.07657 × 100)	8.281 (0.08281 × 100)
5. Biologic gradient	Dose-response data lacking**	50	− 1.764 (− 0.03528 × 50)	− 1.767 (− 0.03534 × 50)
6. Plausibility	No mechanism of toxicity found	0	0.00 (0.23025 × 0)	0.00 (0.21689 × 0)
7. Coherence	No early objective effects or other abnormalities were measured as a result of exposures	0	0.00 (0.009621 × 0)	0.00 (− 0.00334 × 0)
8. Experimental evidence	Animal studies have not demonstrated neurotoxicity from diisocyanate exposure	0	0.00 (0.00843 × 0)	0.00 (− 0.00659 × 0)
9. Analogy	Data to similar class of agents lacking**	50	− 0.6470 (− 0.01294 × 50)	− 0.5055 (− 0.01011 × 50)
		Sum	C1 = − 4.8844	C2A = − 3.5705
Probability of causality	e^C1/(e^C1 + e^C2A)	21.2%

*1 > RR < 2 assumes probability of 60%.

** Assumes 50% probability if data exists but undefined, otherwise assume 0%.

^±Discriminant function values defined by Swaen (from Swaen 09, Table 1).