Review Br. J. Surg. 1990, Vol. 77, February, 12S132
1. G . Allen-Mersh Westrninstef Hospital, Dean R yle Street, London W l , UK Correspondence to: Mr T. G . Allen-Mersh
Pilonidal sinus: finding the right track for treatment Management of pilonidal sinus is frequently unsatisfactory. N o method satisfies all requirements for the ideal treatment - quick healing, no hospital admission, minimal patient inconvenience, and low recurrence but greater awareness of the strengths and weaknesses of existing methods would lead to improved management. Early excision of the pilonidal pit at the time of treatment of pilonidal abscess reduces the high (40per cent) risk of subsequent sinus. Treatments for pilonidal sinus that flatten the natal cleft halve the risk of recurrence. En bloc excision of pilonidal sinus with secondary healing should be abandoned and emphasis given to development of treatments, such as primary asymmetric closure, which have more potential. Some treatments are operator-dependent and, to achieve the best results, junior surgeons must be correctly trained and supervised. Future treatment studies must be prospective and randomized, and should compare healing time, recurrence rates beyond 3 years, nurse and hospital visits, patient inconvenience and loss of income. Keywords:
Pilonidal sinus, pathogenesis, treatment
The ideal treatment of pilonidal sinus should provide a high chance of cure with a low recurrence rate, and should avoid hospital admission and general anaesthetic while involving minimalinconvenience and time off work for the patient. In 1985 over 7000 patients were admitted, for an average stay of 5 days, to hospitals in England for treatment of sacrococcygeal pilonidal disease'. Most pilonidal sinuses resolve with treatment by the age of 40years irrespective of the method of treatmentzs3.This encourages surgeons to believe that their method of treatment is adequate and that there is no compelling need to change. However, the average duration of symptoms from pilonidal sinus exceeds 3 years2 and many patients endure discomfort, inconvenience, and loss of income during prolonged or multiple courses of treatment which are often left to relatively inexperienced surgeons. Although most patients are eventually cured, treatment falls short of the ideal. Over the last 25 years many treatments have been advocated for pilonidal sinus but no consensus has emerged, and there have been few studies comparing relative treatment costs, discomfort and time off work. In this paper the rationale for, and results of, various treatments are reviewed in order to determine those that are promising and require further evaluation and those that should be discarded.
Incidence Sacrococcygeal pilonidal disease affiicts young adults after puberty4. In a population study of Minnesota college students, pilonidal sinus was noted at routine physical examination in 365 (1.1 per cent) of 31497 males and 24 (0.11 per cent) of 21 367 females', but proportionately more females undergo treatment and so the ratio of patients treated is closer - roughly 4:14. The ratio of ma1es:female.s admitted to English hospitals in 1985' for treatment of pilonidal disease was 15:l. The abscess:sinus ratio (0.7:l) was lower for males (05:l) than for females 1.1:l). This may be because pilonidal sinus is more likely to be complicated, requiring multiple hospital admissions, in males. The observation that the onset of pilonidal disease coincides
0007-1323/90/0201234953.00 Q 1990 Butterworth & Co (Publishers) Ltd
with puberty and that de nouo pilonidal disease is rare after age 40 is compatible with an association with sex hormones which are known to affect pilosebaceous glands6. This may explain the earlier onset of the condition in women7 since puberty occurs earlier in females than in males. Differences in incidence between races are also found. The incidence is highest among Caucasians and less among Africans and Asians'. This probably results from differences in hair characteristics' such as kinking, medullation, cuticular scale count and average curvature, and from different daily rates of hair growth' between races. The incidence is also affectedby factors that are not related to hair characteristics. During the 193945 war the increase in sweaty activity associated with buttock friction" - particularly jeep and lorry driving" -and reduced opportunity for personal hygiene produced an epidemic of pilonidal disease. Seventyseven thousand American servicemen were admitted to US Army hospitals'2 where average inpatient stay for treatment was 55 days. Karydakis has noted" that between 1960and 1971 the prevalence of pilonidal sinus at the time of enrollment into the Greek army rose from 4.9 per cent to 14.8 per cent. He suggests that this increase may be due to a 3.2 kg rise in average body weight among Greek army recruits. An association between incidence and obesity has also been observed among American college students'.
Pathogenesis There are two conflicting theories of pathogenesis. Both have implications for the extent of surgical resection thought necessary to achieve a cure. Acquired theory It is most likely that sacrococcygeal pilonidal infection originates within a natal cleft hair follicle which becomes distended with keratinI4. Eccrine and apocrine sweat glands and their ducts are normally sterile, but anaerobic organisms are found in pilosebaceous follicles' '. The distended follicle becomes inflamed and the resulting folliculitis produces oedema which occludes the follicle opening. The obstructed follicle then
123
Pilonidal sinus: T. G. Allen-Mersh
expands and eventually ruptures into the underlying subcutaneous fat to produce a pilonidal abscess. I n 93 per cent of cases the direction of this abscess and subsequent secondary tracks is cephalad’6. MillarI7, who studied the direction of hair growth in the postanal region of 112 patients, observed that where there was cephalad tracking the inflamed follicle was situated at a mean of 8 cm from the anus, but where there was caudal tracking the follicle was a mean of 4.5 cm from the anus. The direction of the sinus track always coincided with the orientation of hair follicles at the site of the inflamed follicle. Normal perianal hair distribution is in a ring around the anus pointing away from the anus towards the natal cleft. Cephalad to this there is a change in the direction of hair which then points caudally from above into the natal cleft. There were no cases in which the direction of the pilonidal track ran in the opposite direction to the orientation of hair follicles in that area. The subcutaneous pilonidal abscess then drains on to the skin surface - usually via a laterally situated track overlying the sacrum18. The natal cleft follicle remnant now communicates via a track extending from the hair follicle through the abscess cavity to a laterally situated granulating sinus track opening. An epithelialized tube, extending into the sinus track from the natal cleft opening, is then formed by the residue of epithelial cells ~. present at the site of rupture of the original f ~ l l i c l e ’ Movement of the buttocks exerts both a suctionlg and a drilling or ‘cigarette rolling’20 effect which encourages loose natal cleft hair and debris to enter the track whenever the patient stands or sits. When hair is propelled tip first into this tube, the barbs on the hair shaft prevent it from being expelled in a reverse direction2’. Hair and skin debris then stimulate a foreign body reactionz2. Rarely the hair may not come from the patientI8, and a case is reported23 in which a pilonidal sinus contained a small bird feather of the variety used in feather bedding.
could happen in other places subjected to rubbing by hair - for example on barber’s hands36. Raffman3’ reviewed the evidence for hair follicles within pilonidal tracks and could find no convincing evidence of skin appendages - hair follicles or sweat glands - within pilonidal tracks. Thus the histological findings are compatible with the view that presacral pilonidal disease originates as a folliculitis and then progresses because of formation of a foreign body granuloma. Although rarely congenital sacrococcygeal sinus may resemble pilonidal sinus, the available evidence strongly supports an acquired aetiology for the great majority of adult pilonidal infections. Thus radical removal of all tissue overlying the sacrum in order to remove all embryological remnants should not be regarded as essential to treatment.
Pathology When considering how to treat sacrococcygeal pilonidal disease, it is important to remember that the condition is merely a particular form of foreign body granuloma2’. Hair is found within the track in roughly two-thirds of male cases and one-third of cases in womenL6and hair fragments - surrounded by foreign-body giant cells - are frequently seen within the granulation tissue. The majority of the track and hair-containing cavity is lined with granulation tissue. Only the midline epithelial pit through which the hair enters is lined by epithelium except in rare cases of long-standing recurrence where the epithelial lining can extend further along the track32. The term pilonidal ‘cyst’ is therefore misleading since it implies that there is a cyst lining which must be excised. In pilonidal cysts, the granulation-lined cavity communicates with one or more epithelialized natal cleft pits but the cavity does not open laterally on to the skin to form a pilonidal sinus. A better term for this condition is ‘pilonidal granuloma’.
Congenital theory Other hair-induced granulomas The identification of postcoccygeal epidermal cell rests24, Foreign body granulomas, resulting from penetration of the skin dysrhaphic sacrococcygeal cysts2’ and the observation of a familial tendency to pilonidal sinus26 led to the s ~ g g e s t i o n ~ ~ , by ~ ~ hair, have also been described at other sites of skin f r i ~ t i o n ~Interdigital ~ - ~ ~ . pilonidal sinuses produced by human that the epithelialized tracks found in adult pilonidal sinus hair and by sheep’s wool have be-n reported respectively in the originate as congenital abnormalities. If this is the case, it hands of barbers36 and sheep shearers47.Unlike sacrococcygeal follows that removal of all epithelialized tracks down to the pilonidal disease, the pathogenesis of these conditions usually presacral fascia is invariably necessary to achieve a cure. involves direct penetration of healthy skin by hair. However, the majority of congenital neural canal sinuses are lumbar rather than sacral29 and most extend to the dura. Risk of malignant change Infection does occur within these congenital tracks and as might Very rarely, squamous carcinoma can arise in a pilonidal sinus be expected this tends to occur in children3’ and can lead to usually of long-standing4’. Giant condyloma a c ~ m i n a t u m ~ ~ , meningitis3’. Vestiges of these congenital tracks have been basal cell carcinoma5’, and adenocarcinoma’ involving detected but they are lined by cuboidal epithelium - unlike pilonidal sinus have also occasionally been reported. The pilonidal sinus tracks which are lined by granulation tissuez2. incidence of any of these neoplasms is far too small for pilonidal Congenital tracks are usually situated more cranially over the sinus to be considered as carrying a significant malignant sacrum than pilonidal sinuses, do not contain hair, and can often potential. be shown to communicate with the spinal canal32. It has also been suggested that pilonidal sinuses could originate from a vestigial scent gland33 similar to the preen gland found in some Clinical course birds, or that it might develop following infection in a n inclusion Pilonidal abscess d e r m ~ i dThe ~ ~ absence . of histological evidence of intermediate conditions demonstrating either of these structures prior to A pilonidal abscess is formed if an infected hair follicle expands development of pilonidal sinuses invzlidates both these theories. into the subcutaneous tissues or a pre-existing pilonidal granuloma becomes acutely inflamed. Roughly half of all cases Pilonidal sinus has been reported as arising in skin derived of sacrococcygeal pilonidal disease first present with an abscess52 away from the midline and subsequently brought by rotation which is usually situated cephalad to the site of hair follicle flap to overlie the sacrum’4 after previous removal of all tissue i n f e ~ t i o n ’ ~It. may not be possible to identify the infected hair overlying the sacrum. It was similar experience of recurrence follicle at the time of acute abscess formation because of after complete excision of all tissue overlying the sacrum which surrounding oedemaI4. Incision, drainage and curettage of the originally led Patey and Scarff3’ to suggest that the condition abscess cavity to remove hairs and d e b r i ~ results ~ ~ . ~in~ was acquired. Advocates of a congenital origin of pilonidal sinus complete healing within about 1 month in over 90 per cent of have suggested that sacral skin could be indrawn to form a sinus cases. Simple incision and drainage results in healing in tube by tethering to the coccyx during growth. The hair follicles 58 per cent of patients within 10 weeks55. Once healing has in this skin then produce the hair which is found within pilonidal occurred, 40 per cent of patients will have no further symptoms, sinuses. Patey and Scarff3’ observed that while pilonidal tracks and a further 20 per cent will experience only minor sympcontained hair, the track lining did not contain hair follicles. t o m ~ ~However, ~ * ~ 40 ~ .per cent of patients treated by incision They argued that the hair, which stimulated a foreign-body and drainage for acute pilonidal abscess will develop a pilonidal reaction, must have entered the sinus from the skin surface as
’
124
Br. J. Surg., Vol. 77, No. 2, February 1990
Pilonidal sinus: T. G . Allen-Mersh sinus requiring further treatment'". Bascom has suggested that if the epithelialized pilonidal pit is excised via a small (7 mm) incision at about 5 days after abscess drainage14 when the pit can usually be identified, the early recurrence rate is reduced to approximately 15 per cent.
penetrates other normal skin to produce a foreign body Treatment follows the usual principles of anal fistula surgery.
Pilonidal sinus A pilonidal sinus is formed when the granulation-lined pilonidal cavity drains via a sinus track which opens away from the midline as an area of proud granulation onto the skin. During an interval between episodes of inflammation, the diagnosis can be confirmed by identifying the epithelialized follicle opening within the natal cleft 4-8 cm cephalad from the anus. The sinus track then runs cephalad from this midline opening in roughly 93 per cent of cases1", and can usually be palpated as an area of induration deep to the sacral skiz. Not infrequently there may be more than one epithelialized opening within the natal cleft but the laterally situated granulation-lined opening is single in cases that have not previously undergone surgery. Pain (84 per cent of patients) and discharge (78 per cent) are the two most frequent symptoms af pilonidal sinus', while fever, chills and bleeding are rare'. Symptoms fluctuate with recurrent bouts of infection, but rarely subside completely between bouts.
There is no shortage of reports recommending the best treatment for pilonidal sinus. Virtually without exception these studies are flawed because of one or more of the following defects in study design: retrospective analysis, patient selection and exclusion not stated, lack of randomization or lack of a control group, short or incomplete follow-up, and failure to assess treatment costs and patient inconvenience. While it may not be possible to decide about what should be done for pilonidal sinus, some conclusions can be drawn from these studies about what should not be done, and about the direction of future treatment studies.
Results of treatment of sacrococcygeal pilonidal sinus
Method of analysis of published studies The principal reports of treatment of pilonidal disease in the last 30 years have been reviewed (Tables 1-6). Average healing time for each treatment method given in the tables has been derived as the quotient of the total number of patient days in all studies of each method which were spent healing divided by the total number of patients treated by that method. Recurrence (Table 7) was divided into early (c1 year) or late (> 1 year). In every evaluable study, minimum follow-up (in contrast to average follow-up) was used to assign recurrence rate to either the early or late follow-up group. Only patients actually followed up, either by clinic review or telephone, were included. Patients not followed up may fare worse63so results are likely to contain an optimistic bias in all cases since inadequate follow-up was not restricted to any one treatment method. The relative cost of treatment by open and primary closure methods was calculated as shown in Table 8 . Figures assume that: 90 per cent of patients heal by first intention after primary closure and 90 per cent of laid open wounds heal within 45 days; laying open does not require hospital admission; district nurse visits open wounds every second day for one month and third day thereafter; ~ ; patient patient's income is the UK 1988 national a ~ e r a g e "and returns to work 21 days after primary closure and 30 days after laying open unless healing is delayed when patients are off work for 60 days. Cost estimates of medical and nursing attention, and hospital facilities are in keeping with current UK provincial rates for private medical care.
Recurrent pilonidal sinus Late recurrence of pilonidal sinus after complete healing can be produced by further hair follicle infection, but early recurrence after treatment is usually associated with granulation-lined tracks which are a consequence of chronic infection and foreign-body reaction in an incompletely healed surgical wound. The pathogenesis in these cases is no longer the precipitating inflamed hair follicle since this has usually been excised. The unhealed surgical wound is a focus for infection and skin debris, including hair, which continues the granulomatous process. Occasionally, after many years of chronic infection, these tracks can develop a squamous epithelial lining3*. It follows that in treatment of recurrent disease, more emphasis should be given to healing or excision of tracks than to identification and excision of the midline pit. Perianal pilonidal sinus In roughly 7 per cent of cases16 of pilonidal sinus the track runs caudally where it may present as perianal sepsis. Four causes of perianal pilonidal disease have been recognized: (1 ) a sacrococcygeal pilonidal sinus may extend into the perianal tissue^'^*'^ and rarely may communicate with the anal canals8; (2) following laying-open of a fistula in ano, hair may enter either the healing wound or the scars9; (3) hair may enter through an anal fissure"'; and (4) hair may penetrate normal anal skin as it
Closed techniques Lord and Millar have de~cribed"~ an outpatient treatment where affected midline epithelial follicles are cored out under
Table 1 Mean healing time and combined recurrence/failure rate following track debridement and excision or phenolization of epithelial pit ~~~
Reference Track curettage/brushing with excision of Lord and Millar65 Dorton66 EdwardP Ba~com~~ Kobe1 and Ma&*
Year follicle opening 1965 1970 1977 1980 1988
Number of cases 33 31 120 50 366
21-
1. G . Allen-Mersh Westrninstef Hospital, Dean R yle Street, London W l , UK Correspondence to: Mr T. G . Allen-Mersh
Pilonidal sinus: finding the right track for treatment Management of pilonidal sinus is frequently unsatisfactory. N o method satisfies all requirements for the ideal treatment - quick healing, no hospital admission, minimal patient inconvenience, and low recurrence but greater awareness of the strengths and weaknesses of existing methods would lead to improved management. Early excision of the pilonidal pit at the time of treatment of pilonidal abscess reduces the high (40per cent) risk of subsequent sinus. Treatments for pilonidal sinus that flatten the natal cleft halve the risk of recurrence. En bloc excision of pilonidal sinus with secondary healing should be abandoned and emphasis given to development of treatments, such as primary asymmetric closure, which have more potential. Some treatments are operator-dependent and, to achieve the best results, junior surgeons must be correctly trained and supervised. Future treatment studies must be prospective and randomized, and should compare healing time, recurrence rates beyond 3 years, nurse and hospital visits, patient inconvenience and loss of income. Keywords:
Pilonidal sinus, pathogenesis, treatment
The ideal treatment of pilonidal sinus should provide a high chance of cure with a low recurrence rate, and should avoid hospital admission and general anaesthetic while involving minimalinconvenience and time off work for the patient. In 1985 over 7000 patients were admitted, for an average stay of 5 days, to hospitals in England for treatment of sacrococcygeal pilonidal disease'. Most pilonidal sinuses resolve with treatment by the age of 40years irrespective of the method of treatmentzs3.This encourages surgeons to believe that their method of treatment is adequate and that there is no compelling need to change. However, the average duration of symptoms from pilonidal sinus exceeds 3 years2 and many patients endure discomfort, inconvenience, and loss of income during prolonged or multiple courses of treatment which are often left to relatively inexperienced surgeons. Although most patients are eventually cured, treatment falls short of the ideal. Over the last 25 years many treatments have been advocated for pilonidal sinus but no consensus has emerged, and there have been few studies comparing relative treatment costs, discomfort and time off work. In this paper the rationale for, and results of, various treatments are reviewed in order to determine those that are promising and require further evaluation and those that should be discarded.
Incidence Sacrococcygeal pilonidal disease affiicts young adults after puberty4. In a population study of Minnesota college students, pilonidal sinus was noted at routine physical examination in 365 (1.1 per cent) of 31497 males and 24 (0.11 per cent) of 21 367 females', but proportionately more females undergo treatment and so the ratio of patients treated is closer - roughly 4:14. The ratio of ma1es:female.s admitted to English hospitals in 1985' for treatment of pilonidal disease was 15:l. The abscess:sinus ratio (0.7:l) was lower for males (05:l) than for females 1.1:l). This may be because pilonidal sinus is more likely to be complicated, requiring multiple hospital admissions, in males. The observation that the onset of pilonidal disease coincides
0007-1323/90/0201234953.00 Q 1990 Butterworth & Co (Publishers) Ltd
with puberty and that de nouo pilonidal disease is rare after age 40 is compatible with an association with sex hormones which are known to affect pilosebaceous glands6. This may explain the earlier onset of the condition in women7 since puberty occurs earlier in females than in males. Differences in incidence between races are also found. The incidence is highest among Caucasians and less among Africans and Asians'. This probably results from differences in hair characteristics' such as kinking, medullation, cuticular scale count and average curvature, and from different daily rates of hair growth' between races. The incidence is also affectedby factors that are not related to hair characteristics. During the 193945 war the increase in sweaty activity associated with buttock friction" - particularly jeep and lorry driving" -and reduced opportunity for personal hygiene produced an epidemic of pilonidal disease. Seventyseven thousand American servicemen were admitted to US Army hospitals'2 where average inpatient stay for treatment was 55 days. Karydakis has noted" that between 1960and 1971 the prevalence of pilonidal sinus at the time of enrollment into the Greek army rose from 4.9 per cent to 14.8 per cent. He suggests that this increase may be due to a 3.2 kg rise in average body weight among Greek army recruits. An association between incidence and obesity has also been observed among American college students'.
Pathogenesis There are two conflicting theories of pathogenesis. Both have implications for the extent of surgical resection thought necessary to achieve a cure. Acquired theory It is most likely that sacrococcygeal pilonidal infection originates within a natal cleft hair follicle which becomes distended with keratinI4. Eccrine and apocrine sweat glands and their ducts are normally sterile, but anaerobic organisms are found in pilosebaceous follicles' '. The distended follicle becomes inflamed and the resulting folliculitis produces oedema which occludes the follicle opening. The obstructed follicle then
123
Pilonidal sinus: T. G. Allen-Mersh
expands and eventually ruptures into the underlying subcutaneous fat to produce a pilonidal abscess. I n 93 per cent of cases the direction of this abscess and subsequent secondary tracks is cephalad’6. MillarI7, who studied the direction of hair growth in the postanal region of 112 patients, observed that where there was cephalad tracking the inflamed follicle was situated at a mean of 8 cm from the anus, but where there was caudal tracking the follicle was a mean of 4.5 cm from the anus. The direction of the sinus track always coincided with the orientation of hair follicles at the site of the inflamed follicle. Normal perianal hair distribution is in a ring around the anus pointing away from the anus towards the natal cleft. Cephalad to this there is a change in the direction of hair which then points caudally from above into the natal cleft. There were no cases in which the direction of the pilonidal track ran in the opposite direction to the orientation of hair follicles in that area. The subcutaneous pilonidal abscess then drains on to the skin surface - usually via a laterally situated track overlying the sacrum18. The natal cleft follicle remnant now communicates via a track extending from the hair follicle through the abscess cavity to a laterally situated granulating sinus track opening. An epithelialized tube, extending into the sinus track from the natal cleft opening, is then formed by the residue of epithelial cells ~. present at the site of rupture of the original f ~ l l i c l e ’ Movement of the buttocks exerts both a suctionlg and a drilling or ‘cigarette rolling’20 effect which encourages loose natal cleft hair and debris to enter the track whenever the patient stands or sits. When hair is propelled tip first into this tube, the barbs on the hair shaft prevent it from being expelled in a reverse direction2’. Hair and skin debris then stimulate a foreign body reactionz2. Rarely the hair may not come from the patientI8, and a case is reported23 in which a pilonidal sinus contained a small bird feather of the variety used in feather bedding.
could happen in other places subjected to rubbing by hair - for example on barber’s hands36. Raffman3’ reviewed the evidence for hair follicles within pilonidal tracks and could find no convincing evidence of skin appendages - hair follicles or sweat glands - within pilonidal tracks. Thus the histological findings are compatible with the view that presacral pilonidal disease originates as a folliculitis and then progresses because of formation of a foreign body granuloma. Although rarely congenital sacrococcygeal sinus may resemble pilonidal sinus, the available evidence strongly supports an acquired aetiology for the great majority of adult pilonidal infections. Thus radical removal of all tissue overlying the sacrum in order to remove all embryological remnants should not be regarded as essential to treatment.
Pathology When considering how to treat sacrococcygeal pilonidal disease, it is important to remember that the condition is merely a particular form of foreign body granuloma2’. Hair is found within the track in roughly two-thirds of male cases and one-third of cases in womenL6and hair fragments - surrounded by foreign-body giant cells - are frequently seen within the granulation tissue. The majority of the track and hair-containing cavity is lined with granulation tissue. Only the midline epithelial pit through which the hair enters is lined by epithelium except in rare cases of long-standing recurrence where the epithelial lining can extend further along the track32. The term pilonidal ‘cyst’ is therefore misleading since it implies that there is a cyst lining which must be excised. In pilonidal cysts, the granulation-lined cavity communicates with one or more epithelialized natal cleft pits but the cavity does not open laterally on to the skin to form a pilonidal sinus. A better term for this condition is ‘pilonidal granuloma’.
Congenital theory Other hair-induced granulomas The identification of postcoccygeal epidermal cell rests24, Foreign body granulomas, resulting from penetration of the skin dysrhaphic sacrococcygeal cysts2’ and the observation of a familial tendency to pilonidal sinus26 led to the s ~ g g e s t i o n ~ ~ , by ~ ~ hair, have also been described at other sites of skin f r i ~ t i o n ~Interdigital ~ - ~ ~ . pilonidal sinuses produced by human that the epithelialized tracks found in adult pilonidal sinus hair and by sheep’s wool have be-n reported respectively in the originate as congenital abnormalities. If this is the case, it hands of barbers36 and sheep shearers47.Unlike sacrococcygeal follows that removal of all epithelialized tracks down to the pilonidal disease, the pathogenesis of these conditions usually presacral fascia is invariably necessary to achieve a cure. involves direct penetration of healthy skin by hair. However, the majority of congenital neural canal sinuses are lumbar rather than sacral29 and most extend to the dura. Risk of malignant change Infection does occur within these congenital tracks and as might Very rarely, squamous carcinoma can arise in a pilonidal sinus be expected this tends to occur in children3’ and can lead to usually of long-standing4’. Giant condyloma a c ~ m i n a t u m ~ ~ , meningitis3’. Vestiges of these congenital tracks have been basal cell carcinoma5’, and adenocarcinoma’ involving detected but they are lined by cuboidal epithelium - unlike pilonidal sinus have also occasionally been reported. The pilonidal sinus tracks which are lined by granulation tissuez2. incidence of any of these neoplasms is far too small for pilonidal Congenital tracks are usually situated more cranially over the sinus to be considered as carrying a significant malignant sacrum than pilonidal sinuses, do not contain hair, and can often potential. be shown to communicate with the spinal canal32. It has also been suggested that pilonidal sinuses could originate from a vestigial scent gland33 similar to the preen gland found in some Clinical course birds, or that it might develop following infection in a n inclusion Pilonidal abscess d e r m ~ i dThe ~ ~ absence . of histological evidence of intermediate conditions demonstrating either of these structures prior to A pilonidal abscess is formed if an infected hair follicle expands development of pilonidal sinuses invzlidates both these theories. into the subcutaneous tissues or a pre-existing pilonidal granuloma becomes acutely inflamed. Roughly half of all cases Pilonidal sinus has been reported as arising in skin derived of sacrococcygeal pilonidal disease first present with an abscess52 away from the midline and subsequently brought by rotation which is usually situated cephalad to the site of hair follicle flap to overlie the sacrum’4 after previous removal of all tissue i n f e ~ t i o n ’ ~It. may not be possible to identify the infected hair overlying the sacrum. It was similar experience of recurrence follicle at the time of acute abscess formation because of after complete excision of all tissue overlying the sacrum which surrounding oedemaI4. Incision, drainage and curettage of the originally led Patey and Scarff3’ to suggest that the condition abscess cavity to remove hairs and d e b r i ~ results ~ ~ . ~in~ was acquired. Advocates of a congenital origin of pilonidal sinus complete healing within about 1 month in over 90 per cent of have suggested that sacral skin could be indrawn to form a sinus cases. Simple incision and drainage results in healing in tube by tethering to the coccyx during growth. The hair follicles 58 per cent of patients within 10 weeks55. Once healing has in this skin then produce the hair which is found within pilonidal occurred, 40 per cent of patients will have no further symptoms, sinuses. Patey and Scarff3’ observed that while pilonidal tracks and a further 20 per cent will experience only minor sympcontained hair, the track lining did not contain hair follicles. t o m ~ ~However, ~ * ~ 40 ~ .per cent of patients treated by incision They argued that the hair, which stimulated a foreign-body and drainage for acute pilonidal abscess will develop a pilonidal reaction, must have entered the sinus from the skin surface as
’
124
Br. J. Surg., Vol. 77, No. 2, February 1990
Pilonidal sinus: T. G . Allen-Mersh sinus requiring further treatment'". Bascom has suggested that if the epithelialized pilonidal pit is excised via a small (7 mm) incision at about 5 days after abscess drainage14 when the pit can usually be identified, the early recurrence rate is reduced to approximately 15 per cent.
penetrates other normal skin to produce a foreign body Treatment follows the usual principles of anal fistula surgery.
Pilonidal sinus A pilonidal sinus is formed when the granulation-lined pilonidal cavity drains via a sinus track which opens away from the midline as an area of proud granulation onto the skin. During an interval between episodes of inflammation, the diagnosis can be confirmed by identifying the epithelialized follicle opening within the natal cleft 4-8 cm cephalad from the anus. The sinus track then runs cephalad from this midline opening in roughly 93 per cent of cases1", and can usually be palpated as an area of induration deep to the sacral skiz. Not infrequently there may be more than one epithelialized opening within the natal cleft but the laterally situated granulation-lined opening is single in cases that have not previously undergone surgery. Pain (84 per cent of patients) and discharge (78 per cent) are the two most frequent symptoms af pilonidal sinus', while fever, chills and bleeding are rare'. Symptoms fluctuate with recurrent bouts of infection, but rarely subside completely between bouts.
There is no shortage of reports recommending the best treatment for pilonidal sinus. Virtually without exception these studies are flawed because of one or more of the following defects in study design: retrospective analysis, patient selection and exclusion not stated, lack of randomization or lack of a control group, short or incomplete follow-up, and failure to assess treatment costs and patient inconvenience. While it may not be possible to decide about what should be done for pilonidal sinus, some conclusions can be drawn from these studies about what should not be done, and about the direction of future treatment studies.
Results of treatment of sacrococcygeal pilonidal sinus
Method of analysis of published studies The principal reports of treatment of pilonidal disease in the last 30 years have been reviewed (Tables 1-6). Average healing time for each treatment method given in the tables has been derived as the quotient of the total number of patient days in all studies of each method which were spent healing divided by the total number of patients treated by that method. Recurrence (Table 7) was divided into early (c1 year) or late (> 1 year). In every evaluable study, minimum follow-up (in contrast to average follow-up) was used to assign recurrence rate to either the early or late follow-up group. Only patients actually followed up, either by clinic review or telephone, were included. Patients not followed up may fare worse63so results are likely to contain an optimistic bias in all cases since inadequate follow-up was not restricted to any one treatment method. The relative cost of treatment by open and primary closure methods was calculated as shown in Table 8 . Figures assume that: 90 per cent of patients heal by first intention after primary closure and 90 per cent of laid open wounds heal within 45 days; laying open does not require hospital admission; district nurse visits open wounds every second day for one month and third day thereafter; ~ ; patient patient's income is the UK 1988 national a ~ e r a g e "and returns to work 21 days after primary closure and 30 days after laying open unless healing is delayed when patients are off work for 60 days. Cost estimates of medical and nursing attention, and hospital facilities are in keeping with current UK provincial rates for private medical care.
Recurrent pilonidal sinus Late recurrence of pilonidal sinus after complete healing can be produced by further hair follicle infection, but early recurrence after treatment is usually associated with granulation-lined tracks which are a consequence of chronic infection and foreign-body reaction in an incompletely healed surgical wound. The pathogenesis in these cases is no longer the precipitating inflamed hair follicle since this has usually been excised. The unhealed surgical wound is a focus for infection and skin debris, including hair, which continues the granulomatous process. Occasionally, after many years of chronic infection, these tracks can develop a squamous epithelial lining3*. It follows that in treatment of recurrent disease, more emphasis should be given to healing or excision of tracks than to identification and excision of the midline pit. Perianal pilonidal sinus In roughly 7 per cent of cases16 of pilonidal sinus the track runs caudally where it may present as perianal sepsis. Four causes of perianal pilonidal disease have been recognized: (1 ) a sacrococcygeal pilonidal sinus may extend into the perianal tissue^'^*'^ and rarely may communicate with the anal canals8; (2) following laying-open of a fistula in ano, hair may enter either the healing wound or the scars9; (3) hair may enter through an anal fissure"'; and (4) hair may penetrate normal anal skin as it
Closed techniques Lord and Millar have de~cribed"~ an outpatient treatment where affected midline epithelial follicles are cored out under
Table 1 Mean healing time and combined recurrence/failure rate following track debridement and excision or phenolization of epithelial pit ~~~
Reference Track curettage/brushing with excision of Lord and Millar65 Dorton66 EdwardP Ba~com~~ Kobe1 and Ma&*
Year follicle opening 1965 1970 1977 1980 1988
Number of cases 33 31 120 50 366
21-
