After spinal cord injury (SCI), most men cannot ejaculate without medical assistance. A major advance in the knowledge of the spinal control of ejaculation has been achieved with the discovery of a spinal generator of ejaculation (SGE) in the rat. The aim of this report was to review studies about ejaculation after SCI in order to revisit the spinal control of ejaculation and especially to assess the existence of an SGE in man.METHODS
Studies were identified from Embase, PubMed, EBSCOhost and Cochrane Library. Studies were eligible when they specify the occurrence of antegrade ejaculation as a function of the neurological characterization of SCI. Studies were excluded when ejaculation was elicited by rectal electrical stimulation or when ejaculation could not be discriminated from climax. Meta-analyses were performed to assess the reference ejaculation rates for each procedure used to elicit ejaculation, i.e. masturbation or coïtus, penile vibratory stimulation (PVS) or acetylcholine esterase (AchE) inhibitors prior to masturbation. Subgroup analyses were performed according to the procedure used to elicit ejaculation on (i) the completeness of the SCI and (ii) the upper and lower limits of the SCI. To assess the existence of an SGE, the effect of concurrent lesions of different spinal segments was assessed by means of a stratified bivariate analysis.RESULTS
From 523 studies, 45 were selected (n = 3851). Ejaculation occurred in response to masturbation or coïtus, PVS or AchE inhibitors followed by masturbation in, respectively, 11.8% (n = 1161), 47.4% (n = 597) and 54.7% (n = 309) of patients with complete SCI and in, respectively, 33.2% (n = 343), 52.8% (n = 305) and 78.1% (n = 32) of patients with incomplete SCI. Ejaculation, in the case of complete lesion of the sympathetic centres (T12 to L2), of the parasympathetic and somatic centres (S2–S4) or of all spinal ejaculation centres (T12 to S5) occurred in response to PVS in none of the patients (respectively, n = 5, n = 4 and n = 21) and in response to AchE inhibitors followed by masturbation in 4.9% (n = 61), 30.8% (n = 26) and 0% (n = 16) of the patients, respectively. Ejaculation in response to PVS or AchE inhibitors prior to masturbation was rhythmic forceful in 97.9% (n = 48) of the patients with complete lesion strictly above Onuf's nucleus (segments S2–S4). Complete lesion of the S2–S4 segments precluded the occurrence of rhythmic forceful ejaculation (n = 5). Controlling for the number of the injured segments between T12 and L2, the ejaculation rate sharply decreased when the lesion extended to the L3 segment and below.CONCLUSIONS
The results reinforce the crucial roles of the spinal sympathetic and parasympathetic centres for emission and the somatic centre for expulsion. The spinal segments between L2 and S2 is more than a pathway to connect the ejaculation centres and likely harbours an SGE in man located in the L3, L4 and L5 segments.