Post Dural Puncture Headache – We Can Prevent It

Although modern anesthesiology has made great progress in the last decades, neuraxial anesthesia (NA) is still the keynote of regional blockade [1]. NA is popular for its effectiveness in producing anesthesia for providing excellent intraoperative neuromuscular paralysis and in generating analgesia for relieving postoperative pain if continuously infused [2, 3]. As the NA techniques are used popularly in clinic, post dural puncture headache (PDPH), a common iatrogenic complication resulted from post-spinal taps or accidental dural puncture (ADP) subsequent to epidural block, is frequently reported [4] and becomes a challenge to health caregivers [5]. Although the incidence of PDPH in research volunteers is ~6% [6], in patients for whom the NA is for clinical purposes the prevalence of PDPH ranges from 10% to over 80% in different aged patients underwent either epidural or spinal or combined block [7].


Introduction
Although modern anesthesiology has made great progress in the last decades, neuraxial anesthesia (NA) is still the keynote of regional blockade [1]. NA is popular for its effectiveness in producing anesthesia for providing excellent intraoperative neuromuscular paralysis and in generating analgesia for relieving postoperative pain if continuously infused [2,3]. As the NA techniques are used popularly in clinic, post dural puncture headache (PDPH), a common iatrogenic complication resulted from post-spinal taps or accidental dural puncture (ADP) subsequent to epidural block, is frequently reported [4] and becomes a challenge to health caregivers [5]. Although the incidence of PDPH in research volunteers is ~6% [6], in patients for whom the NA is for clinical purposes the prevalence of PDPH ranges from 10% to over 80% in different aged patients underwent either epidural or spinal or combined block [7].
Investigations on the risk factors of PDPH revealed that female, age, perpendicular bevel orientation [8], previous history of PDPH [9], repeated dural puncture [10], needle gauge and design [11], and pregnancy [12] are factors substantially related with the occurrence of PDPH. The leakage of cerebrospinal fluid (CSF) was considered as the major cause of PDPH [13], whereas its real etiology is unknown. These procedure-and nonprocedure-related factors in combination determine the patterns of development of PDPH. Several procedures and methods were identified effective in treating and reducing the incidence of PDPH based on the knowledge of procedure-related factors, but whether could we prevent this morbid prior to its occurrence?
Techniques developed based on how to reduce CSF leakage are classified into either preventive or therapeutic ones. Although the results from the differently designed studies were inconsistent [4], one consensus on this topic reached is that we can prevent, at least in part, PDPH with currently available methods. Small size pencil point spinal needle [14,15], parallel bevel orientation [8], liquid use for the loss of resistance (LOR) in epidural puncture [16], and prophylactic epidural blood patch [17,18] are preventive considerations in reducing PDPH. Therapeutically, intrathecal saline injection [19,20], repeatable epidural blood patch [21], and compensatory intrathecal catheterization for drug or fluid administration [22] are means treating the on-going PDPH. Although the effectiveness of these methods is changing in different population at different ages under different clinical procedures, they are still promising for our patients.
Beside abovementioned procedure-related techniques, emerging pharmacological data support the use of analgesics. The most recent systematic review and meta-analysis [23] revealed that morphine, cosyntropin, and aminophylline are effective for reducing the incidence of PDPH with any severity, but dexamethasone on the contrary increases the risk of PDPH. For fentanyl, caffeine, and indomethacin, no conclusive evidence reached in the effectiveness and safety for preventing and treating PDPH due to the design quality and low power of the available studies. In consideration of the contribution of high body mass index (BMI) [24] and non-smoking [25] to PDPH, we thus cannot only attribute PDPH to CSF leakage. Caregivers need take careful consideration of the methods listed in this chapter to prevent and improve the clinical outcomes of this iatrogenic morbid because of its multifactorial originality.

History and epidemiology
First epidural blockade was reported by an American physician Dr. James Leonard Corning in 1885 [26]. The actual history of spinal anesthesia can be traced back to 1888 by German physician Dr. Heinrich Irenaeus Quincke and 1889 by Britain physician Dr. Walter Essex Wynter aspirated CSF from patients with meningitis for lowering intracranial pressure [27][28][29]. In 1898, Dr. Karl August Bier from Germany performed first elective spinal anesthesia for surgery [30,31]; at the same time, Dr. Bier and his assistant they themselves experienced spinal anesthesia, and reported plus another four patients (6/9) with PDPH symptoms [32]. Since that time, analgesics, hydration, and bed rest became the basic constitutes in treating PDPH [33]; however there were still some 40% cases showed no response to these therapeutics. From the early 1970s, anesthesiologists began to use epidural blood patch (EBP) to treat severe PDPH. Until 1990, EBP was first recommended by official guideline [34].
Over the past one century, the incidence of PDPH was sharply decreased from ~70% tõ 1% [35], whereas the recently reported occurrence of PDPH is still seeing a big difference in various clinical settings from different regions when diverse techniques were used in patients with different ages. A Nordic survey found the incidence of ADP in Obstetric setting is 1% [36], and 73% of the ADP patients developed PDPH [37]. The incidence of PDPH in Obstetric in Middle East is 2-4.6% [14,38], 22.7% in Western Africa [39], 16.9% in Southeast Asia [40], 16.6% in North Europe [41], and 6% in North America [42,43]. In non-Obstetric patients, about 18% patients developed PDPH after spinal anesthesia [44], however a lower incidence (4%) was then reported in the next year by the same group [45]. In the earlier time, another group from Denmark reported the occurrence of PDPH was 7.3% in patients underwent different types of surgeries below the diaphragm after spinal anesthe-sia [46]. In orthopedic patients, about 1.6% experienced PDPH after continuous spinal anesthesia (CSA) or combined spinal epidural anesthesia (CSE) in South America [47]. In patients who underwent placement of an intrathecal drug delivery system (IDDS), 23% developed PDPH [48]. .9% children with malignant disease attended for diagnostic or therapeutic lumbar puncture experienced PDPH [49,50].

Risk factors
Clinical and epidemiological studies support a connection between PDPH and certain demographic factors. For adult, the frequency of PDPH was less in older age patients (51-75 years) than younger age comparisons (30-50 years) [51]. Children younger than 13 years rarely get PDPH [49,52], but that does occur with increasing frequency in adolescents and are similar to those seen in adults [53]. To child younger than 13 years and adult older than 50 years, they have less PDPH incidence than their peers that largely may be related to the reduced CSF pressure [54,55]. While there are some inconsistencies upon gender as an independent risk factor for the development of PDPH, a recent meta-analysis confirmed the declaration that the odds of developing a PDPH were significantly lower for male than nonpregnant female subjects with an odds ratio (OR), 0.55 and 95% confidence interval (95% CI), 0.44-0.67 [56]. Lower weight is found to be strongly associated with the higher incidence of PDPH [24] and cumulating evidence showed an inverse relationship between BMI and PDPH [57,58] suggesting that heavier patients in general have higher intra-abdominal pressure, which in turn raises intra-epidural pressure and prevents cerebrospinal fluid from leaking when ADP occurs. New survey revealed that taller height, reduced pre-procedure intravenous hydration and lower systolic blood pressure (SBP) are novel risk factors that contribute to the pathogenesis of PDPH [59]. Although the incidence of PDPH from different countries, an indicator of racial difference, seems to be different [14,[38][39][40][41][42][43], the race itself looks unlike an independent risk factor for the PDPH that was observed in the same study [60]. Interesting findings showed that smokers had a considerably reduced rate of PDPH in comparison with non-smokers suggesting an inhibitory effect of tobacco smoking on PDPH that may be associated with the stimulation role of nicotine in dopamine neurotransmission [25].
In a more recent study, severe headache after lumbar puncture and sitting position were confirmed as predicting factors of the occurrence of PDPH, and in further sitting sampling position, history of depression, multiple effort of lumbar puncture, and high perceived stress during the procedure were found to be significantly associated with a longer duration of PDPH [61]. In the same study, migraineurs showed no change at the risk of developing PDPH compared to the non-headache subjects, and epidural puncture does not trigger migraine attacks [61]. However, there was report showing that patients had a history of chronic or recurrent headache has more chance in nearly 60% to develop PDPH than those without such a history [62]. For the multiple effort of lumbar puncture that indicates the inexperience in such clinical procedures increases the possibility of PDPH [6], but in contrast, other studies found no different between experienced and inexperienced practitioners, nor does between multiple and single dural puncture [63].
Although the leakage of CSF is regarded as the major cause of PDPH, the volume of CSF removed and its role in causing PDPH is unclear. Davignon and Dennehy reported that removal of 15-20 ml of CSF reliably caused headaches [64], but Kuntz et al. did not find such a causal relationship [65]. So it is hard to draw a conclusion from the available data that volume change in CSF causes PDPH. In clinical practice, the volume usually removed during diagnostic lumbar punctures or spinal anesthesia is less than 5ml that means it is not likely to be a significant factor for the PDPH. However, we cannot exclude the possibility that chronic leakage of CSF over more than 15 ml after ADP or spinal anesthesia is causative for the PDPH (see detailed pathophysiology of CSF leakage below).
Prophylactic treatment with 8 mg of dexamethasone not only increases the severity and incidence of PDPH, but is also ineffective in decreasing the prevalence of intra-operative nausea and vomiting during cesarean section indicating that dexamethasone treatment is a significant risk factor for the development of PDPH [66]. Nonetheless, hydrocortisone i.v. (100 mg in 2 ml 8 hourly for 48 h) was found effective in reducing PDPH following spinal anesthesia [67] suggesting that glucocorticoid with different potency and half life of action may possess different function in PDPH prevention and therapy.
Pregnancy is considered as a particular factor that relates to PDPH due to the young age, female, sometimes sitting position, pregnancy-associated depression and anxiety, and the special popularity of regional anesthesia in this population [68][69][70], but a meta-analysis showed that pregnancy itself does not increase the risk of PDPH [71]. For some cases of PDPH, we cannot exclude some other co-founding factors including fatigue, sleep deprivation, and night work that lead to higher incidence of ADP in clinical personnel when performing epidural analgesia. Table 1 summarizes the risk factors of PDPH.

Anatomy of meninges
There three membranes, known as that cover the spinal cord lying within the vertebral canal. The outermost layer is the dura mater, a non-adherent, dense, and tough fibrous sheath closely applied to the inner layer of bone surrounding the spinal canal. Between the dura and the walls of the spinal canal is a potential imaginary space, the epidural space or cavum epidurale, which normally occupied by a small amount of by loose areolar tissue, fat, and the anterior and posterior plexuses of the vertebral veins. Dura mater is attached above to the margin of the foramen magnum, to the axis, and to the third cervical vertebra, and below to the level of the second sacral vertebra. In normal, a potential space known as the subdural space exists between dura mater and arachnoid mater, a thin and delicate membrane lies closely beneath the dura mater. Beneath the arachnoid mater is the pia mater that intimately applied to the spinal cord. Both the arachnoid and pia mater are continuous with the arachnoid and pia surrounding the brain. There is a space between the arachnoid mater and pia mater: the subarachnoid space, which normally is filled with CSF.
The conventional conception for the structure of spinal dura mater is that it is of elastic and collagen fibers running in the longitudinal direction. Based on this, clinical studies found the PDPH incidence is less in patients who underwent spinal anesthesia during which the dura mater was cut with a perpendicular orientation to the direction of the spinal dura fibers than those with parallel bevel orientation [8,72]. But a more extensive electron microscopic study challenged this traditional conception of the of the anatomy of the spinal dura mater, they found that the dura mater is consisted of collages fibers that are arranged in several layers parallel to the surface, and each layer does not arranged in any specific orientation [73]. Moreover, the thickness of the posterior dura mater demonstrates big difference within and between individuals suggesting that perforation at thicker dura is less likely to result in CSF leakage than the thin dura because the thicker the dura mater, the easier the retraction after perforation, and this inter-or intra-individual variation in dural mater thickness may be an unpredictable variable affecting the management of dural puncture [74,75].

Physiology of CSF
CSF secretion in adults varies between 400 to 600 ml per day, i.e. 0.28-0.42 ml/min, and about 60%-75% of CSF is produced by the choroid plexuses of the lateral ventricles and the tela choroidea of the third and fourth ventricles. The total volume of CSF in the adult is ~150 ml, of which 125ml distributes in cranial and spinal subarachnoid spaces and 25 ml in the ventricles. Therefore, CSF is renewed four to five times every 24 hours in young adults. CSF circulation is a dynamic phenomenon that pulses in response to the systolic pulse wave in choroidal arteries. Ageing-related cerebral atrophy and reduction in CSF turnover enlarge the CSF compartment markedly, and aging-associated slowing down of the CSF renewal all may explain the reason why aged population had a lower incidence of PDPH [51].
CSF pressure, one part of the intracranial pressure, is the result of a dynamic equilibrium between CSF secretion, absorption and resistance to flow. Physiological values of CSF pressure vary according to individuals and study methods between 13 and 20 cmH 2 O in adults and 4 and 6 cmH 2 O in infants [76]. In the lumbar region in the supine position, CSF pressure ranges between 5 and 15 cmH 2 O, and this pressure can increase to over 40 cmH 2 O when on the vertical position [77]. In the prone position, CSF pressure changes from 8 to 21 cmH 2 O, but in lateral decubitus position it reduces to 7-17 cmH 2 O [78]. Besides, the normal range for lumbar CSF pressure in children is 10 to 28 cmH 2 O when measured in a flexed lateral decubitus position [79]. CSF pressure is determined by parenchymal and venous pressures. Increased SBP exerts negative feedback on choroidal secretion by decreasing the pressure gradient across the blood-CSF barrier and by reducing cerebral perfusion pressure.

Pathophysiology of CSF leakage and pathological mechanisms
CSF leak is defined as the escape of CSF from any tear or hole in the meninges. The direct consequence of CSF leak is the drop of CSF volume and then the pressure. It is not that clear for the causal relation between CSF leak and PDPH pathogenesis, however the explanatory theory on the role of CSF leak in PDPH onset is still widely accepted. Leakage of the CSF is the most common cause of spontaneous intracranial hypotension [77]. Theoretically, lumbar puncture-induced CSF leak is consisted of two phases: acute and chronic phases. The acute phase is largely resulted from the abrupt outflow of CSF from the broken within minutes to several hours, during which the CSF pressure dives down to a lower level (~3-4 cmH 2 O) that eventually leads to shifts of intracranial contents and gravitational traction on pain sensitive structures, which worsens when the patient is upright and is relieved on lying down [80]. The chronic phase refers to a stage of which started from the formation of the new CSF pressure balance (several hours to one day after dural puncture) to the complete resolution from the puncture (1-6 weeks). This phasic alteration in CSF leakage can explain, at least in part, why the onset of PDPH in some patients occur 1 to 7 days after ADP, but some appeared 12 days post puncture [81]. In addition, the loss of CSF may activate adenosine receptors that subsequently dilate intracranial arteries and veins and then clinical manifestations of PDPH [82]. Based on the findings that pregnancy and the immediate postpartum period are associated with the lowest CSF density [83] and the particular high incidence of PDPH in Obstetric setting [68][69][70], the CSF density change during the chronic leakage of CSF was also considered as a potential reason of the PDPH.
For some patients they showed "crushing" postural headache without abnormal diagnostic lumbar puncture and computed tomography (CT) angiogram suggesting that the conventional understanding on the pathogenesis of PDPH based on the over-rigorous pooled analyses needs to be reconsidered [80]. In fact, the occurrence of PDPH has its own pathological bases. For the cranial innervations, studies showed that the dura mater is heavily innervated and most likely cause intense pain [84], and abnormal distention of intracranial nerve and extracerebral blood vessels all can consequently activate the trigeminal nervous system that were thought to be the origins of headache [85]. Functional immunohistochemistry found that neuropeptides and nitric oxide synthase (NOS) are expressed in the nerve fibers of the supratentorial dura mater and the structural alterations of nitroxidergic axons innervating blood vessels of the dura mater support the idea that nitric oxide (NO) is involved in the induction of headache [86]. Therefore further studies are needed on the relationship among dura mater innervations, expression of neuropeptides and NOS, and PDPH.
Artemin, a member of the glial cell line-derived neurotrophic factor family, is a vasculaturederived growth factor shown to regulate migration of sympathetic neuroblasts and targeting of sympathetic innervation [87]. Recent evidence supports the role of artemin in cold pain [88] and inflammatory pain [89]. Moreover, the expression of artemin was detected in the smooth muscle of dural vasculature, and its receptor GFRα3 was found present in nerve fibers that closely associated with tyrosine hydroxylase (TH) or calcitonin gene-related peptide (CGRP) [90] suggesting that artemin is involved in dural afferent activity through modulating both primary afferent and sympathetic systems. In further, catecholaminergic nerve fibers innervate human cranial dura mater in density, and these nerve fibers are more abundant in the perivascular dural zone than in the intervascular zone at the basal region [91]. In collection, given TH functions as the precursor of catecholamine (norepinephrine and epinephrine) [92], a potential interaction exists among artemin, sympathetic regulation, and catecholaminergic transmission in nerves located in cranial dura mater, and this interaction may underlie the occurrence of PDPH.
Dural innervations are of importance as, like its cranial counterpart, the spinal dura mater and its nerve root sleeves may be a source of primary pain. Different types of nervous branches are given off to the spinal dura mater within the vertebral canal. The nerves in the spinal dura mater have already been described as nociceptive sensory fibers [93][94][95][96], and they also belong to sympathetic vasomotor [97,98]. We hereby proposed that a tonic inhibition of the spinal and cranial dural nerves exists under normal CSF pressure, but this inhibition would be reduced or reversely activated by chronic leakage of CSF after lumbar puncture. However, such reduction in tonic inhibition or/and reverse activation of the dural nerves does not determine the occurrence of PDPH, which depends on the alteration extent of the tone that can evoke nociceptive activation, i.e. it should at least reach the activation threshold. Based on this hypothesis, it can partially explain why not all patients after ADP will develop PDPH [37]. Figure 1 depicts the potential pathological mechanisms of PDPH.

Needle gauge and tip configuration
A huge body of evidence and systematic review support the view that the diameters of the needles that pierce the spinal dura mater and the tip design, cutting-edge or pencil-point, are two key facets that determine the eventual incidence of PDPH [99]. Of the same type needles, 29G (19%) compared with 25G Quincke needle (17%) led to no reduction of PDPH [44]. Similarly, both 25G and 26G Quincke needle had same incidence of PDPH (8-9%), but in comparison, 24G Sprotte non-cutting tip needle results in a significantly lower incidence (1.5%) [100]. Patients receiving spinal anesthesia with a 27G Quincke needle suffered significantly more frequently from PDPH (6.6%) than the 27G pencil-point needle controls (1.7%) [45]. In Obstetric women, 25G Quincke, 27G Quincke and 27G Whitacre spinal needles produce 8.3%, 3.8% and 2.0% of PDPH, respectively [14], and 10% in the 25G Quincke and none in the 24G Gertie Marx spinal needle [101]. For the non-Obstetric patients, the incidences of PDPH for 27G Quincke and 27G Whitacre spinal needles were 2.7% and 0.37%, respectively [102].

Therapeutic strategy
The occurrence of PDPH resulted from ADP or spinal anesthesia is completely unavoidable, although we can reduce its incidence via various preventive means. Therefore, health caregivers need familiar with all potential therapeutic strategies, and treat them following different treatment protocols that are divided into four steps: conservative treatment (1 st step), aggressive medical treatment (2 nd step), conventional invasive management (3 rd step), and aggressive invasive management (4 th step) [4]. Table 3 summarizes the 4-step therapeutic maneuvers for PDPH.

First step: Conservative treatment
Initially, conservative methods are recommended for the treatment of PDPH largely because of its self-limiting property. The use of abdominal binder for PDPH patients is mainly based on its pressure transmission from the increased pressure of abdominal cavity to CSF pressure [106]. Although no powerful evidence supports this hypothesis, and the CSF pressure can change along with the intra-abdominal pressure [107], it is uncertain the consequently increased CSF pressure at early period of lumbar puncture would push more CSF exit from the broken. Conventionally, it has been suggested that PDPH would be less common if patients routinely have a period of bed rest after dural puncture because about 1-70% patients after the puncture experienced postural headache. In addition, giving supplementary fluids additional to the normal dietary intake can restore the loss of CSF. Although the degree of CSF leak does not correlate with the severity of the symptoms in a PDPH [40], it is assumed that improvements in the ratio of CSF production to CSF leak will improve the clinical picture. Dehydration can result in a decrease in CSF production [108]. However, if a patient is appropriately hydrated, and the rate of CSF production is normal, there is no evidence that overhydration will increase the rate of CSF production any further. For both bed rest and hydration, the most recent systematic review did not find convincing evidence supporting the routine bed rest after dural puncture is beneficial for the prevention of PDPH onset, and also it is still unclear whether vigorous intravenous fluid supplementation has any prophylactic or therapeutic benefit in alleviating PDPH [109].
Due to the ethical consideration, pre-operative communication will let the patients know that PDPH is a common iatrogenic complication, and the subsequent problems include the inability to perform daily activities, an extended length of stay (LOS) at hospital, and a higher visiting rate to the emergency room after discharge. All these will raise patients' anxiety to their possibly miserable experiences after regional anesthesia, and the psychological stress will be exacerbated if PDPH eventually occurred. Therefore, psychological support will help PDPH patients more precisely understand: 1) PDPH is a self-terminating process; 2) many medical procedures can alleviate and treat PDPH; 3) active cooperation with clinicians will promote resolution; 4) keep normal diet; and 5) think solutions with faith but not with fear. So try to comfort or reassure PDPH patients psychologically will enhance their confidence to the treatments, and improve the outcomes [110].
Symptomatic analgesia was also used as the conservative management of PDPH. Oral acetaminophen (1000 mg) along with fluid administration was suggested [4,111], whereas the actual effect is unknown. Prophylactic administration of acetaminophen (500 mg)-caffeine combination did not prevent PDPH [112]. Non-steroidal anti-inflammatory drugs (NSAIDs), the most popular over-the-counter drugs for analgesia also can be used for PDPH treatment [32]. Antiemetics combined with other analgesics were suggested for headache or migraine, but whether such medication performs effective function in PDPH is not elucidated. Dexamethasone, a traditional anti-inflammatory glucocorticoid, is found possessing antiemetic effect for postoperative patients [113], and also was used in migraine treatment [114], but for PDPH no convincing evidence was found [23]. Erol reported that gabapentin, a gammaaminobutryic acid (GABA) analog, significantly reduced pain, nausea and vomiting compared to ergotamine/caffeine combinations in patients with PDPH [115] suggesting that gabapentin exerts function in PDPH patients through both analgesic and antiemetic effect.

Second step: Aggressive medical treatment
Aggressive medical treatments that include subarachnoid catheter left in situ, occipital nerve block, intravenous methylxanthines, and symptomatic therapies are recommended once the conservative management was not that effective in treating PDPH. Leaving a subarachnoid catheter in situ after spinal anesthesia or ADP has at least three benefits: 1) mechanical blockade of the CSF leak: intrathecal (i.t.) catheter left in situ for 24 h closes the hole in the arachnoid dura preventing the leakage of CSF; 2) indirect inflammation: catheter-evoked inflammatory responses helping closing the broken; 3) therapeutic administration: convenient drug or fluid infusion or injection through the emplaced catheter for postoperative analgesia or artificial CSF supplementation. This is particularly effective in reducing the incidence of PDPH (~14%) after ADP with large gauge epidural needles than those without catheter placement (70-85%) [22]. Of this technique, Kuczkowski and Benumof composed a 5-step protocol for PDPH prevention and treatment following ADP [22], and their subsequent investigation on it proved more effective than ever in conquering the incidence of PDPH to 6.6% [116]. Similarly, subsequent catheter placement into the epidural space after ADP in cesarean delivery and leaving the catheter for postoperative analgesia for 36-72 h reduced the incidence of PDPH significantly (7.1% compared to 58% in non-catheterized patients) [117]. However, attention needs to be paid on this procedure due to the potential risk of catheter-associated infection [118] and cauda equina syndrome [119,120]. In addition, placement of the catheter needs informed consent and should be reconsidered for it causes discomfort especially when prolonged retention is intended [32].
Since the first case has been reported on the successful treatment of PDPH with occipital nerve block (ONB) [121], several other PDPH patients from different institutes were presented after treatment with ONB [122,123]. The sensory fibers of the greater occipital nerve (GON) originate from the C2 and C3 segments of the spinal cord, and its cutaneous sensory distribution extends over the posterior part of the head, spreading anteriorly to the vertex towards the area supplied by the ophthalmic division of the trigeminal nerve [124]. The lesser occipital nerve (LON) arises from the lateral branch of the ventral ramus of the second cervical nerve. Near the cranium it perforates the deep fascia, and is continued upward along the side of the head behind the auricula, supplying the skin and communicating with the GON [125]. In migraine patients, a unilateral greater ONB can initiate an inhibitory process that shuts down several symptom generators which alleviates allodynia first then headache [126]). Moreover, physical compression of LON causes migraine [127]. In PDPH patients, nerve stimulatorguided bilateral blockade of the GON and LON were found effective in controlling PDPH symptoms [128].
Methylxanthines, a group of derivatives of xanthine, act on adenosine receptors nonselectively as antagonists that in turn lead to vasoconstriction, which exactly negate the compensatory cerebral vasodilation that occurs in response to loss of CSF volume, a theoretical cause that has been implicated in the pathogenesis of PDPH [82]. In addition, methylxanthines can activate sodium-potassium pumps [129] that are involved in the regulation of CSF production [130], which may finally lead to headache relief. There are three major methylxanthines include aminophylline, caffeine, and theophylline that are widely used in patients who suffered PDPH. In a randomized trial, 1 mg/kg aminophylline i.v. given immediately after umbilical cord clamping significantly reduced the incidence of PDPH from 23.3% (no aminophylline injection) to 5% in Cesarean patients underwent spinal anesthesia, and the severe headache after 48 hours was also markedly lower (3%) than the control (11%) [131]. Although caffeine was considered as a potential drug in relieving PDPH, but currently available evidence does not endorse its therapeutic value for PDPH because no valid pharmacological rationale for this drug exists, which majorly due to the clinical trials are few in number, small in sample size, weak or flawed in methodology, and non-effectiveness, contradictory, and even confliction in results [132]. Furthermore, prophylactic oral multidose caffeine-paracetamol also cannot prevent PDPH [112] suggesting that the use of caffeine in PDPH treatment needs to be evaluated carefully and herein I do not recommend it. However, oral theophylline was found effective in treating PDPH painful symptom [133], and studies compared the analgesic efficacy of theophylline infused i.v. to placebo found the therapy significantly decreased the painfulness of PDPH and it was suggested as an easy, rapid, minimally invasive, an effective treatment for PDPH [134]. If methylxanthines are ready to use in managing PDPH, the side effects for these types of drugs should be bear in mind like the central nervous system (CNS) stimulation, seizures, gastric irritation, and cardiac dysrhythmia when patients had psychiatric history, gastroenterological and cardiovascular problems [135,136].
Beside abovementioned analgesics in the conservative treatment, more potent pharmacological analgesics are recommended in this aggressive medical treatment phase. Cosyntropin, an adrenocorticotropic hormone (ACTH) analog with less antigenic than the naturally occurring hormone, was used successfully to treat PDPH. Cosyntropin functions through: 1) stimulating the adrenal cortex to secrete glucocorticoids, mineralocorticoids, and androgens; 2) activating adenyl cyclase, which increases intracellular cAMP that can promote CSF production; and 3) increasing β-endorphins in the CNS which subsequently leads to an increase in the pain threshold. The doses of cosyntropin changed from 0.25 mg to 1.0 mg i.v. or intramuscularly (i.m.), and the headache was controlled by 80%-100%, and incidence of PDPH and the need of EBP were reduced, and the time from ADP to occurrence of PDPH was prolonged [137][138][139][140][141]. The side effects of ACTH administration include infection, mood elevation, and intracranial hemorrhage [142].
Although epinephrine was initially regarded as a vasoconstrictor in neuraxial analgesia, its analgesic effect in the spinal cord is rarely investigated in human being. Accumulating data support the analgesic role of spinal noradrenergic transmission in various pain conditions [143]. Controversial results were reported for the possible effect of epinephrine added into local anesthetics and opioids on the incidence of PDPH. Using continuous intrathecal patientcontrolled analgesia (PCA) consisting fentanyl, bupivacaine and epinephrine completely prevented the occurrence of PDPH in Cesarean patients after ADP [144], but other studies did not find correlation between occurrence of PDPH and type of local anesthetics or additives including epinephrine and opioids [145]. Opioids are still the mainstay of pain management, and intrethecal morphine was also found to be effective in controlling PDPH [23]. While dexamethasone was found no beneficial to PDPH patients, and even is a risk factor for the development of PDPH [66], hydrocortisone i.v. (100 mg in 2 ml 8 hourly for 48 h) on the contrary was found effective in reducing PDPH following spinal anesthesia [67].
Gabapentin and pregabalin are two GABA analogs that have been reported to be useful in the management of epilepsy and neuropathic pain. For PDPH patients, increasing number of cases were reported after successful treatment with gabapentin or pregabalin [146][147][148][149][150]. Sumatriptan, a serotonin type 1-D receptor agonist, was reported effective in the treatment of PDPH, with complete resolution of symptoms [151], but Connelly et al. did not find significant relief in headache [152]. Methergine is used widely in the management of refractory headache and migraine [153,154], and also it was found to be an effective drug in alleviate PDPH [155], although the actual efficacy of its single use in this context is not sure [156]. The antidepressant mirtazapine has a net positive effect on noradrenergic neurotransmission, and was reported effective in relieving PDPH [157]. In sum, although these satellite cases treated with abovementioned medications showed effective in managing PDPH, their clinical application needs to be evaluated substantially before prescribed for PDPH.

Third step: Conventional invasive management
When medical therapies in the step 2 fail for relief of PDPH and when the symptoms of PDPH is debilitating or severe, management should move on to procedural invasive therapies. The most widely used conventional invasive procedure is blood patch, and some alternative materials like hydroxyethyl starch and saline will be discussed if EBP is contraindicated. EBP is a treatment of choice for PDPH with high success rate and low incidence of complications [158].

Theoretical basis of EBP
As early 1960s, Gormley found that the incidence of PDPH in patients with "bloody taps" were comparably lower than those with only saline taps [159], and then the thought of EBP began to develop. Until 1970, the use of EBP became popular, and in 1990, first official guideline recommended EBP for PDPH treatment [34]. Theoretically, EBP is assumed to work by increasing CSF pressure and stimulating fibrin and platelet formation, and secondly, the introduced blood into the epidural space will clot and exerts tamponade effect by occluding the perforation that subsequently will prevent further leakage of CSF.

EBP technique
EBP has the same technique as the epidural anesthesia. Modern epidural kits are disposable and are sterilized package that includes all equipments and drugs without preservative. The epidural needle (usually Tuohy needle) is typically 16-18G, 8cm long with surface markings at 1cm intervals, and has a blunt bevel with a 15-30 degree curve at the tip (the Huber tip). Wings attached at the junction of the needle shaft with the hub, which allow better control of the needle as it is advanced. In general, a traditional glass syringe with an easily slide plunger is used to identify the epidural space. The newer commercially available disposable epidural packs contain a plastic syringe with a plunger that has very low resistance. For EBP, there is no need to use epidural catheter for continuous medication.
To identify the epidural space, several methods can be used like loss of resistance (LOR) and hanging drop technique. Of the hanging drop technique, it has been abandoned by modern anesthesia. Given the reports on the better anesthesia quality and the possible complications of large amounts of air injected into the epidural space and surrounding structures [160], therefore the LOR to saline is preferred in EBP. Even though the EBP technique can be performed with the patient either in the sitting or lateral decubitus position, the latter is the preference due to the sitting position causes more incidence of PDPH [78]. On this position, the patient should be encouraged to adopt a curled up position, as this tends to open the spaces between the spinous processes and facilitates the identification of the intervertebral spaces.
After the back has been prepared with sterile solution and draped in sterile fashion, the desired level is selected.
There are two approaches for the epidural needle insertion: midline or paramedian approach. The midline approach needs insert the epidural needle through the supraspinous ligament, and then advance the needle into the interspinous ligament, until distinct sensation of increased resistance is felt as the needle passes into the ligamentum flavum. For the paramedian approach, the inserting point of the epidural needle is 1-2 cm lateral to the spinous processes, and then insert and advance the needle perpendicularly to the skin until the lamina or pedicle is encountered, and then redirect it approximately 30° cephalad and 15° medially in an attempt to "walk the needle" off the lamina, at which point the needle should be in close proximity to the ligamentum flavum. After felt the resistance from the ligamentum flavum, the needle is then advanced further using LOR to saline.

Contraindications of EBP
Although the performance of EBP has the same contraindications that normally apply to epidural anesthesia, it also has some particular concerns. General absolute contraindications include: 1) patient refusal; 2) coagulopathy; 3) therapeutic anticoagulation; 4) skin infection at injection site; 5) localized sepsis in lumbar area; 6) raised intracranial pressure; 7) hypovolemia; 8) unexplained neurological symptoms; 9) active neurological disease; and 10) generalized sepsis. General relative contraindications include: 1) uncooperation; 2) pre-existing neurological disorders; 3) fixed cardiac output states; 4) anatomical abnormalities of vertebral column; and 5) prophylactic low dose heparin. Particular contraindications of EBP: 1) raised white cell count and pyrexia; 2) human immunodeficiency virus (HIV)-positive patients with other active bacterial or viral illnesses; 3) oncology (EBP in these patients may raises the potential for seeding the neuraxis with neoplastic cells).
Special attention should be paid when anticoagulants are used: 1) full anticoagulation with warfarin or standard heparin is absolute contraindication to EBP; 2) partial anticoagulation with low molecular weight heparin (LMWH) or low dose warfarin (International Normalized Ratio, INR < 1.5) is relative contraindication; 3) for low dose standard heparin (Minihep), wait for 4 h after a dose before performing EBP, and it should not be given until 1 h following blood patch; 4) allow a 12 h interval between LMWH administration and EBP; 5) NSAIDs including aspirin do not increase the risk of epidural hematoma; 6) avoid EBP for 24 h when fibrinolytic and thrombolytic drugs are used, and check clotting prior to needle insertion; 7) EBP needs avoid when patients were diagnosed thrombocytopenia especially when the platelet count is less than 100 000/mm 3 .

EBP procedures
Once the EBP was determined to be applied, following procedures should be scheduled.

1.
Give a full explanation of the cause of the headache, the reasons for performing an EBP, the technique, potential hazards and anticipated success rate;

Obtain informed consent;
3. Move patient to fully equipped work area;

4.
Undertake under the direct supervision of a consultant or senior physician in the operating room with an assistant;

5.
Lie flat for an hour before the EBP procedure ( that may improve its efficacy by reducing the volume of CSF in the extradural space); 6. Secure i.v. access for fluid titration;

7.
Two operators are required, both scrubbed, gowned and masked; 8. Position patient in lateral position;

9.
Operator 1: sterilize skin over back, drape and perform epidural puncture at the same level as previous puncture or one level below; 10. Operator 2: simultaneously sterilize skin over antecubital fossa, drape and perform venepuncture withdrawing 20 ml of blood; 11. Blood is handed to operator 1 who injects blood via epidural needle until either the patient complains of a tightness in the buttocks or lower back, or until 20 ml is injected;

12.
If back or leg pain (due to arachnoid irritation) occurs, stop injecting and wait a few seconds. If pain persists, abandon procedure. If a catheter has been used, remove it immediately after injection is complete;

13.
Inject remaining blood into blood culture bottles for culture and sensitivity;

15.
Keep the patient under close review. If symptoms have not completely resolved, refer to consultant, and a repeat blood patch may be required;

Record procedures in medical book;
17. Refer the patient visit anesthetic clinic 2-4 weeks.

Distribution of the blood patch
No consensus reached as to the precise volume of blood required for EBP, but it is now recognized that the 2-3 ml of blood originally described by Gormley [159] is inadequate, and that 20-30 ml of blood is more likely to ensure success. However there was successful case treated with larger volumes of blood up to 60 ml in patients with spontaneous intracranial hypotension [161].
Several studies reported the distribution of the blood patch in the epidural space using radiolabelled red blood cells [162] or magnetic resonance imaging (MRI) scanning [163]. After injection, the blood spread caudally and cephalad regardless of the direction of the bevel of the needle, and also the blood can reach to the anterior epidural space circumferentially, and also can pass into the paravertebral space. When injecting 14 ml of blood, the highest level it can reach is six spinal segments, and caudally three segments. It is presumed that the compression of the thecal space that elevates the subarachnoid pressure for the first 3 h contributes to the rapid resolution of the headache. About 7-13 h after the EBP, there will be a thick layer of mature clot over the dorsal part of the thecal sac formed due to the procoagulant effect of CSF [164,165].

PDPH outcomes of EBP
The reported success rate of the EBP technique is 70-98% if carried out more than 24 h after the dural puncture [36,166]. Kokki et al. reported that EBP performed later than 48h following lumbar puncture or ADP is effective in parturients with postdural puncture symptoms [167].
If the first EBP failed to relieve the headache, repeated EBP can be used with the same success rate each. In a randomized controlled trial (RCT) in which a "sham" procedure was assigned as the comparison to the EBP, 11 out of 12 patients (92%) reported successful relief for the first EBP application, and the twelfth being relieved by a second procedure, whereas the shamtreated patients reported no benefit from the procedure (168). In another RCT, the respective successful rates of EBP and conservative treatments were 42% and 10% at day 1, and 84% and 14% at day 7 in PDPH patients. For those without recovery, the severity of headache was mild in all EDBP patients, but moderate or severe in conservatively treated patients (169). If the headache persists or debilitates after several attempts, other invasive maneuvers should be considered as appeared in the step 4.

Complications of EBP
While EBP is an effective treatment with a low complication rate, it is also an invasive method that can cause permanent neurological sequelae such as early and late back pain, radiculopathy, spinal-subdural hematoma, spinal-epiarachnoid hematoma, intrathecal hematoma, arachnoiditis, and infection. In consideration of headache as a common symptom of PDPH and cerebral venous thrombosis (CVT), therefore it is hard to distinguish them especially after EBP [170][171][172][173][174][175] suggesting that it should be carefully evaluated before EBP was planned in patients with altered coagulation state (see above Contraindications of EBP), and when possible treatments that would affect the coagulation were ready to be given. Besides, rare cases were reported that EBP may cause epidural scarring that eventually results in slow spread of epidural local anesthetics, unilateral block and low efficacy if later epidural block was performed [176].

Prophylactic EBP
There were studies suggesting the use of prophylactic EBP in preventing PDPH, but the data were conflicting. Reported cases including patients underwent post-myelogram [177], and spinal anesthesia and ADP with an epidural needle [178,179], have confirmed the benefit of prophylactic patching. Nonetheless, other studies found prophylactic EBP cannot decrease the incidence of PDPH or reduce the need for criteria-directed therapeutic epidural patch for parturients after ADP, but it can shorten the duration of PDPH symptoms [18]. One possible explanation for the failure of EBP is that the pressure gradient between the thecal and epidural space may be high enough immediately after blood injection which leads to patch separation from the site of the perforation. Therefore a greater volume of blood may be needed to produce a successful patch [32].

EBP for Jehovah's Witness
Due to Biblical interpretation principles, Jehovah's Witness patients do not authorize even autologous transfusions because blood removed from the body lost the continuity [180]. In the early time, alternative methods like epidural saline or Dextron were suggested [181]. Until 2003, Silva Lde et al. reported a closed system, through which two Jehovah's Witness patients with PDPH were treated successfully with autologous EBP [182]. Since then several cases were reported treated with the closed system successfully [183][184][185]. For this system, in brief, it includes two serum catheters cut in 60 cm segments, one two-way connection, one three-way tap and one 20 ml syringe. The system was assembled to allow one connection to venepuncture needle (20G), one connection to the three-way tap, and the remaining two ways were connected to a 20 ml syringe and to the other serum catheter segment, which would be connected to the epidural needle. After approval by the ethical committee, and informed consent by the patient, the system was filled with saline (6 ml). After epidural puncture and intravenous puncture, the already described system was connected to the epidural needle. Occluding the epidural needle way by moving three-way tap, 20 ml venous blood was aspirated to the syringe and venous catheter way was occluded. Then, epidural needle way was opened and syringe's content was injected in the epidural space. After removal of the epidural needle and system, venous access was maintained for fluid infusion. Patient remained at rest for 2 h and discharged with the recommendation to visit the clinic in case of recurrence.

Alternative maneuvers to EBP
For PDPH patients with absolute contraindications, alternative methods are suggested. In theory, epidural injection of other materials like saline or hydroxyethyl starch (HES) would produce the same mass effect, and restore normal CSF dynamics. Advocates of the epidural saline patch [32] include: 1) a single 30 ml bolus of epidural saline after development of headache; 2) 10-120 ml of saline injected as a bolus via the caudal epidural space; 3) 1.0-1.5 l of  [187]. Epidural HES patch was also found effective in treating PDPH when patients contraindicated to EBP like bacteremia and leukemia [188,189]. Although there were successful cases managed with other epidural materials, it is still not conclusive for their clinical use due to the lack of high quality evidence.

Forth step: Aggressive and invasive management
When epidural patch with blood, saline or HES failed to resolve the headache, the diagnosis needs to be reevaluated and more aggressive methods can be considered. These invasive managements only apply to those with persistent, severe and debilitating headache after treatment using above means in the step 3.
Fibrin glue, also known as fibrin sealant, is a biological adhesive made up of fibrinogen and thrombin that are applied to the tissue sites to glue them together or block bleed by creating a fibrin clot [190]. Fibrin glue is used frequently in repairing cranial dural perforations to block CSF leak after intradural procedures [191]. There are successful PDPH cases treated with epidural fibrin glue injected through percutaneous CT guidance or blindly in patients [192,193] and animal models [194]. Moreover, the effectiveness of fibrin glue in sealing the hole and stopping the leakage of CSF has been studies using in vitro model of postdural puncture leakage and got supportive results for its application under this condition [195]. However, conflicting cases reported that such artificial formulation had a risk of the development of aseptic meningitis [196], and in further it has been warned against the application of fibrin glue when they were used in CNS because fibrin glue contains tranexamic acid (t-AMCA) which may cause severe nervous complications [197].
Surgery was considered as the last option for the treatment of PDPH if all abovementioned methods failed to resolve it. Neurosurgical procedure can be performed to identify and suture the hole in dura mater under the operating microscope. In one refractory PDPH patient in whom the headache lasted over two year, surgical repair successfully resolved the headache immediately, and the patient was rapidly mobilized from bed without orders for bed rest or any further precautions [198]. Before performing surgeries, the exact location of the CSF leak should be identified. Several medical diagnostic techniques are currently available to help detect the CSF leakage: dynamic CT myelography for fast-flow CSF leak, delayed CT myelography or magnetic resonance myelography for slow-flow leak.

Preventive strategy
Although the effectiveness of the prophylactic EBP is controversial [18,176,177], there are studies found advocating role for other strategies in reducing the incidence of PDPH. Deliberate intrathecal saline injection (5 ml) before spinal administration of hyperbaric bupivacaine as a preventive approach is an effective and simple way to minimize PDPH in patients undergoing cesarean section (the incidence is 2% versus 16% without prophylactic saline) [20]. Subsequent spinal catheterization with epidural catheter following ADP can be used to prevent extra leakage of CSF, and then prevents PDPH [22,116,199]. Preventive epidural morphine 3mg given after the end of anesthesia and another 3 mg given on the following day reduced the incidence of PDPH from 48% to 12% [200]. Preventive administration of cosyntropin after ADP in parturients was associated with significant reduction in the incidence of PDPH and need for EBP and significant prolongation of the time from ADP to occurrence of PDPH [141]. Orally used prophylactic frovatriptan 2.5 mg/diet for 5 days markedly decreased the occurrence of PDPH [201]. Other methods like prophylactic administration of caffeine, magnesium, aminophylline, dexamethasone, or intravenous fluid infusion all cannot reduce the incidence of PDPH [202,203].

Recommendations for clinicians
The occurrence of PDPH is determined by multiple factors including patient's demographic variables, caregiver's aspects, procedure-related factors, and post-accidental strategies. It is unclear which of them weighs over the others and what the accurate weight for each factor is in contributing to the onset of the headache. An arbitrary predictive curve of the incidence of PDPH to its risk factors herein is modeled and depicted in Figure 2 to show the association between the headache and different risk factors, and also give a potential prediction of the PDPH occurrence. In this model, each risk factor is scored "1", and all currently identified risk factors are summed up and in total get a scoring "10". Of the PDPH incidence changes from "0" to "100". For example, a 24-year non-smoking full-term pregnant woman with depression history was assigned to spinal anesthesia for Cesarean section by a third-year resident under the supervision by his consultant, so this woman had a risk scoring of "6", and the probability for the PDPH occurrence is ~32%.

Risk factor Scoring
Young Beside the risk predictive model, we need bear in mind following recommendations before and during performing spinal anesthesia.

1.
Soothe patients psychologically to release stress;

2.
Advance corresponding knowledge of the health care personnel; 3. Be energetic and active when performing procedures;

4.
Hydrate patients prior procedures at least 500 ml;

5.
Stabilize blood pressure at individual's physiological level; 6. Use lateral decubitus position for procedures;

7.
Avoid repeat attempts of lumbar puncture;

No dexamethasone any time;
9. Reduce the volume of CSF withdrawn.
For the strategies after spinal anesthesia or ADP, refer to Table 3 for the therapeutic means and section 9 for the preventive maneuvers.

Concluding remarks
As a well known iatrogenic complication, PDPH has its special morbidity that affects patient's daily life, even though it carries self-terminating characteristic. Based on its particular risk factors and special pathophysiological concerns, many preventive and therapeutic methods are developed, although they need to be verified by further quality studies. While we cannot find a once-for-all method for the perplexing headache, we can individually assess the patient and predict the risk of PDPH using the arbitrary predictive model in combination with the prior-procedure preventive strategies. The key to avoid such an annoying morbid is to bear all associated concerns in mind and keep alert when facing a patient ready to spinal anesthesia or epidural puncture, and actively seek and carry out effective remediation once PDPH occurred. Yes, PDPH, we can prevent it only if we paid our attention carefully on this issue.