Electroejaculation: its technique, neurological ... · JournalofNeurology, Neurosurgery, andPsychiatry, 1981, 44, 9-18 Electroejaculation: its technique, neurological implications

Journal of Neurology, Neurosurgery, and Psychiatry, 1981, 44, 9-18

Electroejaculation: its technique, neurologicalimplications and usesG S BRINDLEY

From the Department of Physiology, Institute of Psychiatry, London

SUMMARY An improved technique for electroejaculation is described, with the results ofapplying it to 84 men with spinal injuries and five men with ejaculatory failure from othercauses. Semen was obtained from most patients, but good semen from very few. Only onepregnancy has yet been achieved. The technique has diagnostic applications.

"Electroejaculation" is the word generally usedto denote the obtaining of semen by electricalstimulation with electrodes in the rectum. Thename is a little misleading, because the semen israrely or never ejaculated in the strict sense; ittrickles from the external urinary meatus withoutthe help of contractions of striated muscles. Thetechnique has been used in domestic animals since1936. Most veterinary users have employed sinu-soidal alternating current of frequencies from 15to 100 Hz, delivered through rigid rectal probes,a substantial fraction of whose surface was coveredwith electrodes.'

Published accounts of its application to man2-7consider only its use for rendering paraplegic menfertile. Two pregnancies thus achieved have beenreported,4 7 one of them7 yielding a live baby.Stimuli have been sinusoidal at frequencies fromtwo3 to 906 Hz or have consisted of trains of uni-directional pulses of duration one" to 156 ms andfrequency five6 to 1405 Hz, and they have beenapplied through probes similar to those used inveterinary practice. No writer, I believe, has dis-cussed the rational choice of stimulus parametersor electrode sizes or positions, and I can findno substantial body of empirical observation onthese matters in either the abundant veterinaryliterature or the scanty medical literature.

Preliminary investigations

Theory What is needed is to stimulate the right nervefibres, and as few as possible of the wrong ones, with

Address for reprint requests: Professor GS Brindley, Institute ofPsychiatry, De Crespigny Park, Denmark Hill London SE8 8AF.

Accepted 18 September 1980

9

the least possible risk of thermal and electrolyticdamage to the rectal mucosa. To select a particularnerve or compact plexus it is advantageous to havea single circular cathode and a larger anode or groupof anodes. The diameter of the cathode should beroughly equal to its least distance from the nerve orplexus; a greater diameter worsens the discrimination,and a smaller diameter causes the current densityclose to the electrode to be unnecessarily and perhapsharmfully high. These desiderata are roughly obviousfrom theory, and can be confirmed from experiencein stimulating motor points and cutaneous nervesthrough the skin.

In attempting to stimulate from the rectum ana-tomically defined structures in the pelvis one wouldexpect discrimination to be better with an electrodemounted on a glove-finger than with one mounted ona rigid probe, and probably the risk of doing mech-anical damage is less with a glove-finger. To minimiseelectrolytic damage, one should ensure that no netdirect current is passed (which some earlier workershave done), and also minimise the electrical chargetransferred per cycle (which none of them have done).To minimise thermal damage, one should, amongstother things, minimise the electrical power consumed,and this has never before been attempted. For boththese purposes, if the fibres that one is intending tostimulate are myelinated (and it will be argued belowthat they are), the stimulating pulses should be short.Charge per pulse at the threshold for a myelinatedfibre falls with duration down to roughly 80 /is andremains nearly constant below this. Energy per pulseat such a threshold falls with duration down toroughly 150 ,us and rises below this. Thus pulses of100 pis duration will be not very far from minimisingboth charge and energy; at least they will be muchbetter in both respects than the sinusoidal currents or1 to 15 ms pulses used in previous published work.A nimal Experiments All experiments and surgicalprocedures on animals were done under deep pento-barbitone anaesthesia. In ten male baboons and four

Protected by copyright.

on March 30, 2021 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.44.1.9 on 1 January 1981. Dow

nloaded from

http://jnnp.bmj.com/

10

male rhesus monkeys, I explored what happened whensites accessible to a cathode of 4 mm diametermounted on a glove-finger were stimulated electricallythrough the rectum. I was able to produce separately,on either side, the movements appropriate to thepudendal, sciatic, inferior gluteal, superior gluteal,obturator, or genitofemoral nerve. The femoral nervecould be stimulated, but usually not without alsostimulating the genito-femoral. In all 14 animals,stimulation at sites giving obturator or genitofemoraleffects gave full or nearly full erection, with thresholdroughly two or three times that for somatic effects.

In two baboons a urethral catheter was passed andthe bladder pressure measured. The region from whicha rise in bladder pressure was obtained with lowestthreshold was centred on the best site for stimulatingthe obturator nerve. All stimulation which increasedthe bladder pressure also caused some erection, but inthe more caudal part of the effective region the erec-tion was slight though the rise in bladder pressure waslarge (40 mmHg or more). Just rostral to the effectiveregion was one from which erection could be pro-voked without any rise in bladder pressure.

In seven of the 10 baboons and three of the fourrhesus monkeys semen was obtained as a result ofthe procedure, but it was often not clear which ofseveral sites of stimulation had caused the emission. Itherefore repeated the procedure in one of thebaboons and one of the rhesus monkeys with theabdomen opened and the vas deferens of both sides sur-gically exposed in the spermatic cord and upper scro-tum. In both animals, the site that gave contractionof the seminal vesicle and vas deferens with lowestthreshold was just rostral to the ipsilateral obturatornerve point. The response fatigued if stimulation wasrepeated at 15 second intervals, but two minutes' restsufficed for its recovery. The contraction producedfrom either obturator point was restricted to the ipsi-lateral seminal vesicle and vas deferens. Stimulationin the midline just rostral to the prostate caused con-traction on both sides, but with threshold higher by afactor of 1-6 to 1-8. The contractions involved all thevisible part of the vas deferens, that is from the upperborder of the testis to about 2 cm below the pubiccrest and from the lateral inguinal fossa to thebladder. The most easily visible manifestation of themwas a shortening, so that undulations of the vas dis-appeared and it took as straight a course as possiblebetween its points of attachment. No peristaltic waveswere seen. The strength-duration relation for produc-ing contraction of the left vas deferens by stimulatingthe left obturator point was measured in the baboon.The chronaxie was 0-6 ms.

In four male baboons and three male rhesus mon-keys I implanted electrodes on the hypogastric plexusimmediately in front of the bifurcation of the aorta.In all animals, stimulation through the implantedelectrodes caused shrinkage of the penis together witha contraction of the seminal vesicles and vasa defer-entia very similar to that seen during electroejacula-tion. In one of the rhesus monkeys the strength-

G S Brindley

duration relation for producing such contraction wasmeasured. The chronaxie was 0 4 ms.Experiments on myself In me, trains of 100 lAs pulsespassed through a 6 mm circular cathode mountedeither on a glove-finger or on a rigid probe will stimu-late both motor and sensory fibres of the pudendalnerves from the upper part of the anal canal withoutcausing severe pain, though always with some pain.It is even possible to produce a maximal contractionof the muscles innervated by the pudendal nerve.8From the posterolateral rectal wall, pressing back-wards towards the sacral plexus of one side, I canproduce ipsilateral toe flexion or plantarfiexion ofthe foot, and tingling in various parts of the ipsilateralS2 and S3 dermatomes. These are always accompaniedby diffuse deep pelvic pain, but the pain is not pro-hibitively unpleasant at threshold for the muscularand "cutaneous" effects. I cannot produce contractionof the gluteal or adductor muscles, or erection, ejacu-lation, micturition or the sensation of impending mic-turition. The strongest stimuli that pain allows me totolerate are roughly 1/4 (in voltage or current) ofthose needed for electroejaculation in patients.

Methods

EQUIPMENT FOR ELECTROEJACULATION IN MANElectrode mounts My first glove-finger electrodemounts were made from dental acrylic. The rigidityof this material hindered accurate palpation, but notas severely as one might expect. After about a yearof using acrylic electrode mounts I changed to siliconerubber. Whatever the material of the electrodemount, it is made to extend beyond the tip of thefinger by about 20 mm, and the cathode, whose di-ameter is about 8 mm, lies on the palmar surface ofthis extension. There are two anodes in parallel, eachof area equal to that of the cathode or a little greater,on the dorsal surface centred 18 and 50 mm proximalto the cathode. Such electrode mounts are made byMr C M Andrew of 43 Landcroft Road, LondonSE22, and can be bought from him.Stimulator I assume that in man, as in the baboonand rhesus monkey, electroejaculation depends on thestimulation of myelinated fibres. I therefore adhererigidly to 100 /As as the duration of the stimulatingpulses. Until May 1980 I always used 30 pulses persecond, because I had found in 1977 that this waseffective, and there seemed to be no reason to changeit. Since May 1980 I have had reason to use lowerfrequencies, and have often done so. At first I useda general purpose electrophysiological stimulator, put-ting a capacitor in series with the output to preventthe passage of direct current. Now I use a compact(19X 11 X8 cm) battery-driven stimulator designed by(and purchasable from) Mr C M Andrew, giving thefollowing stimulus parameters: pulse duration 100 its;peak voltage of pulse up to 108 V (9 equally-spacedsteps) into 1 MQ2 load, up to 80 V into the usual loadof 250Q2; time-constant of sag of pulse 800 us; peakcurrent into 250Q up to 316 mA; nett current less



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Electroejaculation: its technique, neurological implications and uses

than 0-1 AsA; output impedance 60-902 (highest athighest voltage setting); time-constant of decay ofreturn current between pulses 10 ms; frequency ofpulses 30, 15 or 10 per second (switchable).

PROCEDUREMy usual practice is to have the patient supine withknees bent (though it is possible to use the left lateral,or even the prone position). It is useful to have ahelper standing on the patient's left to hold the leftleg. I stand on the right and put the electrode mounton the right index finger over a plastic or rubberglove. I put some aqueous jelly (KY jelly is suitable)on the tip of the electrode mount and over the elect-rodes, and insert the electrode mount into the analcanal until the cathode and distal anode and half ofthe proximal anode are within the canal. Then stimu-lation at 9 volts should cause contraction of the analsphincter, and, with a little adjustment of the rotationand depth of the electrode mount, the left or rightischiocavernosus muscle. I next turn my hand so thatits palm faces down and, pressing gently downwardwith the finger-tip, advance the electrode mount asfar in as it goes easily with very gentle pressure. Thecathode is now near the best places for stimulating thesacral plexus and its somatic branches. Using 36 to54 volt pulses, I explore the posterior wall of thepelvis for sites giving flexion of the toes, plantar-flexion at the ankles, visible contraction of the ham-strings, abduction of the thighs, visible contraction ofthe buttocks, external rotation at the hips, or palpablecontraction of the obturator internus muscles. It isusually possible to obtain all of these, separately orin various combinations, on either side, and this mayprovide useful diagnostic information. But when theprocedure is being used solely to obtain semen, itsuffices to obtain any one of the somatic motor effects.If I get none at 54 V I try at 80 V. If still none canbe obtained, then either no anterior horn cells sur-vive from the 5th lumbar to the 2nd sacral segment,or the stimulator is not working, or there is muchgas in the rectum, preventing the anodes from makinggood contact. In the last case the gas can be expelledby pressing on the abdomen, pulling laterally with thefinger to assist its release.

I next turn the palmar surface of the finger so thatit faces directly to the patient's right, and advance thefinger as far as possible, pushing hard. I then explorethe right lateral wall of the pelvis for the obturatorpoint, that is the site where 80-volt stimulation causespowerful adduction of the right thigh. Stimulation hereusually yields semen (if semen can be obtained at all)after from 5 to 20 seconds. If there is none in 40 sec-onds I try the left obturator point. To reach the leftobturator point it is sometimes necessary to transferthe electrode mount to the left hand and stand on thepatient's left.

Results

In 256 attempts at electroejaculation on 89 men

(84 with spinal injuries and five others), I haveobtained semen externally in 163 (64'1%), retro-gradely in 44 (17'2%) and not at all in 49 (19'1%).But these figures are overweighted with men onwhom I have had many (in one case 31) successes.Table 1 classifies the results on men with spinalinjuries by patients rather than by attempts. "Ex-ternal success" means that on at least one occasionliquid containing spermatozoa (not necessarilymotile) trickled from the meatus. On 11 of these36 men only one attempt was made. On four ofthem two attempts were made, and both wereexternally successful. On 21 of them from threeto thirty-three attempts were made, of which some(usually most) were externally successful, the re-mainder being either retrograde or unsuccessful.

"Retrograde success" means a patient fromwhom I have never obtained semen externally, butwhose next urine passed after an attempt atelectroejaculation has contained at least 5X 106spermatozoa. On six of these 14 men only oneattempt was made. On five men two or moreattempts were made, all retrogradely successful.On three men two attempts were made, oneretrogradely successful and the other unsuccessful.

"Definite failure" means a patient in whom onone occasion (17 cases) or on three occasions(three cases), attempted electroejaculation hasyielded no liquid at the meatus (15 cases) orliquid containing no spermatozoa (five cases), andthe next urine passed has contained no sperma-tozoa (13 cases) or fewer than 5 X 106 spermatozoa(seven cases).

"External failure" means a patient from whomno semen was obtained externally, and who failedto provide a specimen of urine. All such patientswere seen once only.

"Pain prevented" means a patient in whom, atthe only attempt at electroejaculation, stimulationat less than the strength needed for success inother patients was intolerably painful.Partly retrograde ejaculation Ten of the 36 menlisted in table 1 as external successes also onoccasion gave purely retrograde ejaculation. Itwould thus not be surprising if ejaculation wassometimes partly external but partly retrograde.Merely finding spermatozoa in the next urinepassed after an externally successful electroejacu-lation is insufficient to prove such an occurrence;it is necessary to collect the urine in two lots, thefirst (of at least 20 ml) to wash out the urethra,and the second as a sample of real bladder con-tents. I have done this on three occasions (differ-ent patients) when I already suspected that ejacu-lation was partly retrograde, and on all three

B

I1I



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


12

occasions found more spermatozoa in the bladderurine than in the external ejaculate.Prevention by pain Every patient who couldrecognise pinprick in no lumbar or sacral derma-tome could tolerate electroejaculation withoutanaesthesia. Every patient who could recognise pin-prick in a sacral or L5 or L4 dermatome was un-able to tolerate it. Patients who could recognisepinprick in an LI to L3 dermatome but not belowwere unpredictable.General anaesthesia and neuromuscular blockTable 2 summarises the ten patients on whom Ihave attempted electroejaculation under generalanaesthesia, usually with neuromuscular block(succinylcholine). Six had spinal injuries or spinabifida, and in four of these the reason for usinggeneral anaesthesia was that an attempt at electro-ejaculation without anaesthesia had been preventedby pain. In the other two, the reason was thatattempts without anaesthesia had been retro-gradely successful. I then tried under generalanaesthesia with neuromuscular block, and fromone of the two patients obtained semen externally.A similar conversion from retrograde to externalejaculation was achieved in one of the two other-wise healthy men who had never ejaculated in thewaking state. My first electroejaculation of him,

G S Brindley

done under general anaesthesia without musclerelaxant, was retrogradely successful. Five sub-sequent electroejaculations, all done with neuro-muscular block, have yielded semen externally.

SIDE-EFECTS OF ELECTROEJACULATIONContractions of striated muscles These are a valu-able guide to the position of the stimulatingcathode. Under general anaesthesia they agreewith what would be expected from the stimulationof nearby a motor fibres. In unanaesthetisedpatients these direct motor effects are accom-panied by reflex effects. Often the reflex effects arevariable from time to time in the same patient,and they may occur after rather than duringstimulation; but sometimes they are so repeatableand so immediately related to the stimulus thatonly their anatomical inappropriateness shows thatthey are not direct motor effects. The commonestsuch pseudo-direct response of striated muscle iscontraction of the abdominal muscles on stimu-lation of branches of the sacral plexus; I havenotes of its occurring in seven patients and thinkit occurred in a few others. Almost simultaneouscontraction of the adductors of both sides in re-sponse to stimulation of the obturator nerve onone side has occurred in at least four patients. I

Table 1 Success and failure of electroejaculation in patients with spinal injuries

Highest clinically External success Retrograde success Definite failure Externalfailure Pain prevented Totaldamaged cordsegment

C6 to Tl 13 1 5 3 0 22(4 incomplete) (incomplete) (2 incomplete)

T2 to T12 21 11 13 3 2 50(3 incomplete) (1 incomplete) (6 flaccid) (2 flaccid) (both incomplete)

(1 flaccid) (1 incomplete)Ll orbelow 2 2 2 1 5 12

(1 incomplete) (flaccid) (1 flaccid) (flaccid) (2 incomplete)(1 flaccid)

Total 36 14 20 7 7 84

Table 2 Electroejaculation under general anaesthesia

External success Retrograde success Definite failure1 complete L2 lesion I complete T5 lesion (patient with 1 T12/L2 lesion (flaccid patient)

transurethral resection of bladder neck)1 incomplete T12 lesion 1 abdomino-perineal excision of rectum for

carcinoma (electrodes inserted throughcolostomy)

I complete T4 lesion I diabetic non-ejaculator without otherevidence of neuropathy

I lumbar spina bifida2 otherwise healthy lifelong non-ejaculators(5 and 6 external successes respectively)The fi e patients in table 2 who had spinal injuries are listed also in table 1.



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



have never seen pseudo-direct contraction of thequadriceps, adductors or ilio-psoas in response tostimulation of the sacral plexus, or pseudo-directcontraction of the gluteal muscles, or of anymuscle below the knee, to stimulation on the oppo-site side.

In six of the 146 externally successful electro-ejaculations without general anaesthesia, emissionof semen has been accompanied by rhythmic con-tractions of the abdominal muscles. The temporalpattern of these contractions roughly resembledthat of orgasmic pelvic floor contractions,9 butthey lacked the progressive increase in intervalthat occurs at orgasm. On two of these six oc-casions the rhythmic contractions included theanal sphincter.Rhythmic abdominal contractions (not including

the anal sphincter) were seen in one of 47 unsuc-cessful and in one of 44 retrogradely successfulelectroejaculations.After-effects on spasm Several patients havementioned that for a few hours after electroejacu-lation their legs are less spastic than usual. Others,when directly questioned, have denied any sucheffect. No patient has reported increase in spasm.Micturition Stimulation that is intended to yieldsemen sometimes yields urine instead, or a mixtureof urine and semen. Of the 15 men whom I haveelectroejaculated three or more times with exter-nal success (most of them six or more times), eighthave never given semen mixed with urine, threehave done so on a minority of occasions, and fouron the majority of occasions. Of the 22 men whomI have electroejaculated once or twice with ex-ternal success, three have given semen mixed withurine.There are two means by which the release of

urine with semen can sometimes be prevented. Thefirst uses the anatomical separation of sympatheticand parasympathetic fibres that might be expectedfrom anatomical textbooks, and was seen in thebaboon experiments. Stimulation near the bestsite for stimulating the superior gluteal nerveoften gives urine without semen. When the bladderhas by this means been made emptier than thepatient can get it by his own efforts, stimulation ator a little cranial to the best site for the obturatornerve may give pure semen. The second methoduses the effect of pulse frequency. In the baboon'0and in human patients (Brindley, Polkey & Rush-ton, unpublished), stimulation of preganglionicsacral parasympathetic fibres at 30 pulses/s givesvery much stronger detrusor contractions than at15 or fewer pulses/s. For electroejaculation, how-ever, 15 pulses/s are almost as good as 30/s in the

baboon. In man they are at least sometimes ade-quate, and yield pure semen where 30/s hadyielded mixed semen and urine from the samepatient.Erection I have done some externally successfulelectroejaculations without causing any penileerection, and very many with only a slight erec-tion. But there are men in whom seminal emissioncannot, with my technique, be achieved withoutan accompanying full erection, and others inwhom erection sometimes but not always occurs.Contraction of the dartos muscle A train of largepulses delivered to almost any site in the pelviscauses in nearly all patients a conspicuous con-traction of the dartos after a delay of 3-5 seconds.This response is absent or very feeble under gen-eral anaesthesia, but I have seen it clearly andreproducibly present in a patient with a T8 lesionin whom there was no evidence (unless this dartosresponse be such) of any functioning cord orsurviving anterior horn cells below the level of thelesion. In two otherwise similar patients with T6and T7 lesions the dartos response was absent.Rise in blood pressure In patients with lesionsat T5 or below I do not usually record the bloodpressure. No such patient has reported headacheduring the procedure. In patients with lesionsabove T5, electroejaculation nearly always raisesthe blood pressure. I formerly used an automaticsphygmomanometer (Dynamap 845). Now I usean ordinary sphygmomanometer repeatedly, or feelthe pulse, inflate a sphygmomanometer cuff to200 mmHg, and not whether the pulse reappears.If the systolic pressure reaches 200 mmHg or ifthe patient reports headache I stop stimulating.The blood pressure then always falls within afew tens of seconds, and I often resume stimula-tion (if semen has not yet been obtained) within aminute.

CORRELATES OF FAILUREIn five of the 20 patients classified as definitefailures, electroejaculation yielded liquid at theexternal urinary meatus that looked like semenbut contained no spermatozoa. The volumes were3.5, 2-0, 0 3, 0-2 and 0 1 ml. In another threepatients, no liquid appeared at the meatus, butacid phosphatase and fructose were substantiallymore abundant in the next urine passed afterelectroejaculation than in urine passed before it.Another patient was of eunuchoid appearanceand had very small testes. It seems likely that inall or most of these nine patients the electricalstimulation had its proper effect on the prostate,seminal vesicles, and vasa deferentia, but the

13



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


14

patients were azoospermic from disease or injuryof the upper genital tract.

Six patients with whom I definitely failed had noevidence of any surviving anterior horn cellsbelow T6, T7, T8, T9, LI and L2. In the last twoof these, damage extended on one side up to T12.Two patients, both with complete Tl 1 lesions,had intact muscles of the L5 and sacral myotomes,but severe wasting of the quadriceps, adductors,and iliopsoas. It seems likely that in all or mostof these patients the sympathetic fibres that theprocedure should stimulate were lost.There remain two patients where definite failure

is entirely unexplained. Both had cervical lesions,and only one attempt at electroejaculation wasmade. Their urines were not examined for fruc-tose, acid phosphatase or other genital-tractmarkers (eg arginine or y-glutamyl transferase).

THE INFLUENCE OF SURGICAL OPERATIONS ON THEBLADDER NECK AND URETHRA

Twelve patients have had transurethral resectionof the bladder neck; of these, three have also hadtransurethral external sphincterotomy. Five ofthese patients were external successes, and fiveretrograde successes. Two were "definite failures",in both cases with evidence that retrogradeemission of azoospermic semen had occurred.Thus transurethral resection of the bladder neckcertainly does not make external emission imposs-ible, though it probably increases the proportionof retrograde to external emissions. Only onepatient has had external sphincterotomy and noresection of the bladder neck. He is the one patientwho has become a father. Of 33 attempts atelectroejaculation on him, 31 have been externallysuccessful. The other two (both early in the series)were definite failures.

ELECTROEJACULATION IMMEDIATELY AFTERREMOVING A FOLEY CATHETERThis lhas been done 34 times on 13 patients.Twenty-one attempts (seven different patients)were externally successful, seven attempts (on fourof the preceding patients) were retrogradely suc-cessful, and six attempts (six different patients)were definite failures.

QUALITY OF SEMEN IN MEN WITH SPINAL INJURIESI nearly always measure the specimen, and docounts of motile spermatozoa and all spermatozoa,within 40 minutes. Paraplegic semen is usuallyliquid; only for about one specimen in 10 is thereneed to wait for liquefaction.

G S Brindley

The volume can be from 0-2 ml to 10-5 ml(the latter without evident contamination byurine). The average is about 2 ml. The number ofspermatozoa per unit volume can be normal, butis more often low (less than 40 million/ml). Twopatients have repeatedly given specimens withcounts exceeding 500 million/ml, and six otherpatients have at least once had counts exceeding200 million/ml. The fraction of spermatozoa thatswim is almost always low. Among 166 externalejaculates examined, the highest motilities were78% (T10 complete lesion of one year's duration),48% and 33% (T5 complete lesion of 17 years'duration, injured at age 19 and first electroejacu-lated at age 36), 41% (T5 complete lesion of 1Iyears' duration), 34% (T7 complete lesion ofseven years' duration), and 26% (C7 completelesion of two years' duration). All the other 161specimens had motilities under 26%. In the 44retrograde ejaculates (which by definition con-tained at least five million spermatozoa), motilitieswere zero in 32, and between 1% and 20% in 12.Fourteen wives of men with spinal injuries have

been inseminated with the husband's semen, all ofthem more than once. Only one pregnancy hasbeen achieved. Paternity was verified by exam-ination (by Prof B E Dodd of the London1Hospital) of 17 blood antigens, not includingHLA. The child is healthy, walked at 11 months,and could say six distinguishable words at 14months.

DOMESTIC ELECTROEJACULATIONThe wives of 12 paraplegic men were askedwhether they wished to learn to electroejaculatetheir husbands. Ten wished to and two did not.All the 10 who wished to learn have now learned,and are using the technique successfully at home.One of these wives is a physician, one a nurse, andone a physiotherapist. The other seven had norelevant previous knowledge or experience.

Discussion

SAFETYLocal heating The calculated power dissipationat the cathode during stimulation at 30/s at thehighest voltage setting is at most 0- 11 watt cm2.In two recent electroejaculations done at thesesettings the temperature rise was monitored by abead thermistor mounted on the centre of thecathode. The temperature rises in 30 seconds ofstimulation were 0 25 and 0 300C. There is nokind of single or double failure in the stimulatorthat could increase the power dissipation by more



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



than a factor of five; such an increase could occuronly if there were three simultaneous faults, oneof them being the insertion of the wrong kind offuse in the fuse-holder.Electrolysis The net direct current passed by theblocking capacitors during stimulation at 30/s atthe highest voltage-setting is less than 10-7A, thatis at least 100 times too low to cause electro-lytic damage. In five patients I examined therectal mucosa with a sigmoidoscope a few minutesafter electroejaculation. The sites of stimulationwere indistinguishable from the surroundingmucosa. Blood-staining of the next faeces passedhas been reported to me after only two of 256electroejaculations on 89 patients. There are twoblocking capacitors in series, each rated at 160 V,and the potential across them cannnot exceed 96 V.It is thus very unlikely that either will fail, andmuch more unlikely that both will.Ventricular fibrillation The current density atthe heart produced by the pelvic stimulation usedfor electroejaculation can be calculated if weassume the human body to be a uniform con-ductor. Though this assumption is inaccurate, itis unlikely to overestimate or underestimate thecurrent density by more than a factor of 2 or3. The current density in a uniform conductor atdistance D from a dipole of length d carryingcurrent I, where D is substantially greater than d,is dI/2srD3 along the axis of the dipole and lessat any other orientation. In the present case thelength of the dipole is at most 3 cm, the distancefrom the dipole to the heart at least 20 cm, andthe current during a pulse at most 316 mA, givinga peak current density at the nearest part of theheart of 19 MtAcm-2, and an rms current densityof 1[0 MAcm-2 at 30 pulses per second and 05[Acm'2 at 15 pulses per second. The official safelimit for 50 Hz alternating current is 50 uAcm-2rms. In pigs of weight 15-25 Kg, 300 mA peakcurrent from head to knee in trains of 3-7 mspulses was required to produce ventricular fibrilla-tion." Assuming 600 cm2 (probably a substantialoverestimate) for the non-lung non-fat cross-sectional area of the chests of such pigs, the peakcurrent density at the heart would be 500 ltAcm .Even on these criteria the currents used forelectroejaculation should be safe by a factor of atleast 25, and the 100 ,us pulses used should be less(and perhaps very much less) effective in causingventricular fibrillation than 3-7 ms pulses of thesame peak amplitude. To further confirm that thestimulus parameters used have a large margin ofsafety for the heart, in one anaesthetised baboonthe 100 us pulse stimulator was connected between

electrodes of area 5 cm-2 in the mouth and in therectum and switched on for a minute at maxi-mum amplitude (EMF 108 V) and 30 pulses persecond. This caused very powerful tonic contrac-tion of the trunk muscles which probably involvedevery motor unit in the whole trunk. Breathingceased, but the heart continued to beat in regularsinus rhythm throughout the minute of stimula-tion. Breathing returned within a few seconds ofthe end of stimulation. This baboon weighed 12 Kgand the non-lung cross-sectional area of its chestwas about 350 cm2. The peak current in eachpulse was 520 mA, so the peak current densitythrough the heart was about 1'5 mA cm-.Elevation of blood pressure When electroejacu-lation is being done on patients with high lesionsthe blood pressure must be measured, and stimu-lation stopped if it rises too high or if the patientreports severe headache. What pressure should beregarded as too high, and whether mild headacheshould be taken as a ground for stopping, arematters on which I can give no authoritativeopinion. But it seems reasonable to suggest thatif the systolic pressure does not rise above 200mmHg and the headache is neither distressing tothe patient nor more severe than he has alreadyexperienced in previous episodes of autonomicdysreflexia, no excessive risk has been taken.

WHAT IS BEING STIMULATED TO CAUSE EMISSIONOF SEMEN

Electroejaculation works well under generalanaesthesia, and is then very unlikely to dependbon a reflex. In unanaesthetised patients, theemission of semen was in all but 2 of 146 exter-nally successful trials unaccompained by rhythmicpelvic floor contractions such as occur in normalorgasm. It is thus unlikely that success in the un-anaesthetised ordinarily depends on triggering areflex response of the cord; if it did, this reflexwould have to be very different from the orgasmicreflex of the intact cord. What of the two oc-casions when rhythmic pelvic floor contractionsdid accompany seminal emission? One of thesewas the only electroejaculation of a patient, butthe other was one of five electroejaculations of apatient who emitted semen externally on all fiveoccasions, with rhythmic abdominal contractionsincluding the anal sphincter once, rhythmic ab-dominal contractions without anal sphincter con-tractions twice, irregular abdominal contractionsonce, and no obvious skeletal muscular activityonce. It seems likely that even here the seminalemission depends on ele_trical stimulation ofefferent fibres; rhythmic reflex activity can occur

15



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


16

simultaneously with it, but is probably a separateand independent phenomenon.The sites that are best for causing seminal

emission agree fairly well with those that onewould expect, from known anatomy, to containsubstantial numbers of sympathetic fibres to thegenital tract, and they are remote from the sensoryfibres that serve the known receptive field for re-flex ejaculation. In me, and in the patients withincomplete or low lesions, the sensations producedby stimulating at these sites are nongenital, whollyunpleasant, and not conducive to sexual arousal.These facts provide further reasons for doubtingwhether the seminal emission produced by mytechnique is ever reflex.The chronaxies found for contraction of the vas

deferens in the rhesus monkey and baboon arevery much lower than those found for unmyelin-ated fibres,12 and are typical of myelinated fibres.This leads immediately to the conclusion that therelevant sympathetic fibres are myelinated (andhence presumably preganglionic) at the sites stimu-lated, that is in front of the bifurcation of theaorta and between the rectum and the obturatornerve.

It would be very difficult to measure the chro-naxie for electroejaculation in men, and I havenot attempted it. But the relatively low thresholdsfound for 100 ,us pulses make it probable thatmyelinated fibres are being stimulated. The largestunmyelinated fibres have electrical thresholdsonly twice those of small myelinated fibres forlong pulses, but at least 20 times for short pulses,12so that it is difficult to believe that the present

G S Brindley

stimuli can have excited more than a very fewfavourably placed unmyelinated fibres.

THE SEGMENTAL ORIGIN OF THE NERVE FIBRESINVOLVED IN SEMINAL EMISSIONTable 3 shows all patients with complete traumaticlesions at T6 or below who gave external success,good retrograde success (>20X 106 spermatozoa)or definite failure without evidence for azoosper-mic emission. The more stringent criterion ofretrograde success is to reduce the risk of includ-ing cases where spermatozoa were released spon-taneously into the urine of expressed mechanicallyby the electroejaculation procedure. Two patientsare thus excluded; they have Tl and T12 lesions.Four "definite failures" in the sense of table 1 areexcluded, two (both T6 lesions) because they gaveazoospermic external ejaculates and two (T 11 andT12) because of high fructose and acid phospha-tase in the urine collected after electroejaculation.From cases 1-18 it seems that no single seg-

ment (unless just possibly T8 or Tl 1) is the sourceof all the fibres involved in seminal emission. Sincecord lesions are rarely if ever as short as onesegment, we may even tentatively infer that nopair of consecutive segments is the source of allthe fibres. Case 17 is very informative. He twicegave good retrograde emission, despite completeabsence of responses to electrical stimulation ofmotor nerves in all segments from T12 downwardon the left and from LI downward on the right.He had no trace of erection, reflexly or in theelectroejaculation procedure or (he says) psychic-ally. For him we can say almost certainly that the

Table 3 Patients with complete lesions at T6 or below who gave external success, good retrograde success,or definite failure not attributable to azoospermia

External success Good retrograde success Definitefailure

I T6 14 T7 (resection of bladder neck) 19 T6, flaccid2 T6 15 T12 20 T7, flaccid3 T6 16 T12 21 T8, flaccid4 T6 left, T7 right 17 T12 left, Li right, flaccid 22 T9, incomplete, flaccid5 T7 18 T12 left, L2 right 23 T1 1, partly wasted muscles down to L46 T7 right, T8 left 24 Ti 1, completely wasted muscles down to

L47 T9 25 T12 right, Li left, flaccid8 T9 26 T12 right, L2 left, flaccid (general

anaesthetic)9 T9, partly wasted muscles down to L310 T91 1 TIO12 T1213 L2 (general anaesthetic)



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



lumbar and sacral segments of the cord are un-necessary for seminal emission. The only slightreservation is that loss of all a anterior horn cellsdoes not absolutely prove loss of all intermediatehorn cells in the same segment.

Cases 25 and 26 seem to conflict with case 17,but both were failures on a single occasion withoutchemical analysis of the urine for genital-tractmarkers.

Cases 23 and 24 suggest that loss of all thesympathetic outflow from Tll downwards prob-ably prevents success in electroejaculation. Case 9(in whom I have had many external successes) isnot in sharp conflict, because of the incomplete-ness of his wasting.

Cases 19-22 all showed no trace of somaticmotor responses in the electroejaculation pro-cedure. Even the dartos response was absent incases 19, 20 and 22, though it was present (to mysurprise) in case 21. Case 22 was incomplete onlyin having slightly preserved tactile sensation.The predominance of low lesions among the

retrograde successes suggests that a low lesion mayinterrupt the sympathetic fibres to the bladderneck without destroying all those to the vasa de-ferentia and seminal vesicles.

IMPROVING THE QUALITY OF SEMENThe figure of merit that I use is the estimatednumber of motile spermatozoa in the wholeejaculate, that is the product of volume, count perml, and fraction motile. On this index the bestparaplegic ejaculate that I have seen does notmatch average normal semen, and most paraplegicejaculates are very bad. They are also usually ab-normal in being liquid when emitted.

Possible reasons for the poor quality includenon-drainage, chronic infection, and raised scrotaltemperature. I have some evidence that all thesethree are contributory, and that remedying themis practicable and beneficial; but this work has along way to go before it will be fit to publish,except for the firm finding that deep scrotal tem-peratures are on the average higher in paraplegicsin wheelchairs than in age-matched and similarlyclothed seated normal men.13

THE TREATMENT OF MEN WHO EJACULATERETROGRADELYSpermatozoa may remain motile for three hoursor more in urine provided that it is not too acid(pH less than about 6-2) or too dilute (total solutesless than about 150 mosm/Kg). It thus seemspossible that retrograde ejaculates might be used.after centrifugation, for inseminating. Sodium bi-

carbonate 5 g, given by mouth one or two hoursbefore, ensures appropriate pH and concentration.However, centrifugation may not be harmless,and even the best urine is unlikely to be a goodenvironment for spermatozoa.

In two men who previously ejaculated retro-gradely I have obtained semen externally byelectroejaculating under general anaesthesia withneuromuscular block. It seems likely that thistechnique will be widely applicable. The same end,that is relaxation of the striated muscles that closethe urethra, could doubtless be achieved bypudendal block or sacral epidural block.

RANGE OF APPLICATION

Diagnostic use The electroejaculation procedureprovides a quick and simple means of examining,in a patient of either sex with a complete transec-tion of the spinal cord, which muscles of the lowerlimbs retain a motor innervation. It is applicableduring the period of spinal shock, provided thatthree or four days have elapsed since injury, sothat motor fibres whose perikarya have beendestroyed have lost their excitability. The sameend can be achieved by stimulation of motor pointsthrough the skin, but the electroejaculation pro-cedure is much quicker.

In theory, the electroejaculation procedureshould distinguish between a deafferented and atotally denervated bladder. These behave alikein ordinary cystometry. At present the distinctionmakes little or no difference to treatment, but thismay not remain true in the future.Fertility of men with injury or disease affectingthe spinal cord or cauda equina This is the"established" (though as yet only slightly success-ful) field of application. It will probably be themost important, if means can be found for improv-ing the quality of the semen.Fertility of otherwise healthy men with ejaculatoryfailure Only two such patients have been re-ferred to me. With both I have been regularlysuccessful, and the counts and motilities are withinnormal limits. Though no pregnancy has yet beenachieved, there is ground for optimism.Injuries and disease affecting peripheral nervefibres in the pelvis It seems likely that when theseconditions cause ejaculatory failure they usuallydo so by destroying the efferent pathway, soelectroejaculation will fail. However, there maybe a few cases where it will succeed.

References

1 Ball L. Electroejaculation. In: Klemm WR, ed.Applied Electronics for Veterinary Medicine and

17



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


18

Animal Physiology. Springfield, CC Thomas, 1976;394-441.

2 Horne HW, Paull DP, Munro D. Fertility studiesin human male with traumatic injuries of thespinal cord and cauda equina. New Engl J Med1948; 239:959-61.

3 Bensman A, Kottke FJ. Induced emission ofsperm utilizing electrical stimulation of theseminal vesicles and vas deferens. Arch Phys MedRehab 1966; 47:436-43.

4 Thomas RJS, McLeish G, McDonald IA. Eectro-ejaculation of the paraplegic male followed bypregnancy. M J Aust 1975; 2:798-9.

5 David A, Ohry A, Rozin R. Spinal cord injuries:male infertility aspects. Paraplegia 1977; 15:11-14.

6 Res P, Plevnik S, Suhel P. Electroejaculation inspinal cord injured patients. International Con-tinence Society, 7th Annual Meeting 1977; Paper6:21-22.

7 Francois N, Maury M, Jouannet D, David G,Vacant J. Electroejaculation of a complete para-

G S Brindley

plegic followed by pregnancy. Paraplegia 1978;16:248-51.

8 Brindley GS, Rushton DN, Craggs MD. Thepressure exerted by the external sphincter of theurethra when its motor nerve fibres are stimulatedelectrically. Br J Urol 1974; 46:453-62.

9 Masters WH, Johnson, VE. Human sexual re-sponse. Boston: Little, Brown & Co., 1966:185.

10 Brindley GS. An implant to empty the bladderor close the urethra. J Neurol, Neurosurg Psy-chiatry 1977; 40:358-69.

1 1 Jacobsen J. Die Gefahrdung durch phasenange-schnittene und gleichgerichtete elektrische Strome.Hannover: Technische Universitat (Dr. Ing.thesis) 1973.

12 Blair EA, Erlanger, J. A comparison of thecharacteristics of axons through their individualelectrical responses. Am J Physiol 1933; 106:524-64.

13 Brindley GS. Deep scrotal temperature and theeffect on it of clothing, air temperature, activity,posture, and paraplegia. Br J Urol (in Press).



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from

Documents

Electroejaculation: its technique, neurological ... · JournalofNeurology, Neurosurgery, andPsychiatry, 1981, 44, 9-18 Electroejaculation: its technique, neurological implications