Post‐Dural Puncture Headache: Part I Diagnosis, Epidemiology, Etiology, and Pathophysiology

Abstract

(Headache 2010;50:1144‐1152)

Post‐dural puncture headache (PDPH) is a frequent complication of dural puncture whether performed for diagnostic purposes or accidentally, as a complication of anesthesia. Because both procedures are common, clinicians interested in headache should be familiar with this entity. The differential diagnosis of PDPH is broad and includes other complications of dural puncture as well as headaches attributable to the condition which lead to the procedure. The patterns of development of PDPH depend on a number of procedure‐ and nonprocedure‐related risk factors. Knowledge of procedure‐related factors supports interventions designed to reduce the incidence of PDPH. Finally, despite best preventive efforts, PDPH may still occur and be associated with significant morbidity. Therefore, it is important to know the management and prognosis of this disorder. In this review, we will highlight diagnosis and clinical characteristics of PDPH, differential diagnosis, frequency, and risk factors as well as pathophysiology of PDPH.

Dural or lumbar puncture (LP) involves the passing of a needle through the dura mater into the cerebrospinal fluid (CSF)‐filled subarachnoid space.1 A common procedure, LP is performed to measure the CSF pressure and sample CSF for laboratory analysis. It is also used diagnostically to instill radio‐opaque material into the subarachnoid space in myelography. For therapeutic purposes, it may be used to remove CSF (in the treatment of idiopathic intracranial hypertension) and to administer intrathecal chemotherapy or spinal anesthesia. Finally, accidental dural puncture is a common complication of epidural anesthesia.1

Post‐dural puncture headache (PDPH) is a sometimes unavoidable complication. When severe or prolonged, PDPH may become incapacitating.2 Among diagnosed patients, 39% experience at least 1 week of impaired ability to perform activities of daily living.3 When severe, PDPH may necessitate hospital admissions that may be prolonged or recurrent.3 Because of its profound consequences, emphasis on prevention and optimal treatment of PDPH is important. The likelihood of developing a headache after dural puncture depends on a number of factors, which may lower or increase the probability of PDPH.4, 5 Neurologists and headache specialists should be familiar with techniques that reduce the risk of PDPH and with approaches to diagnosis and treatment. Herein, we review the diagnosis and clinical features, differential diagnosis, epidemiology, and risk factors, as well as the pathophysiology of PDPH. In a companion manuscript we will review prevention, treatment, and prognosis.

CLINICAL CHARACTERISTICS OF PDPH

Post‐dural puncture headache is frequently diagnosed on the basis of clinical features. History of a procedure, such as a diagnostic LP, myelogram or spinal anesthesia is almost invariably present.6 The cardinal feature of PDPH is orthostatic or postural headache.7 Headache typically appears or worsens when the patient moves from a supine to an upright position and is typically relieved when the patient lies down. Headache generally appears or worsens within 20 seconds of a postural change and reaches its maximum within a minute. Headache characteristically subsides within 20 seconds of recumbency.8

In PDPH, headache features are variable. The pain is usually severe, but may be mild or moderate.4 The quality may be burning, dull and/or throbbing.4, 9 The headache location is not diagnostic. Pain may be frontal or occipital with radiation to the neck and shoulders.4 Temporal areas and vertex pain are less common.4 Holocephalic pain involving the neck occurs. Headache can be aggravated by physical activity and movements of head may worsen the pain.6 Valsalva maneuver, coughing, sneezing, straining, or ocular compression may worsen the headache.6 PDPH may have associated features including low back pain, vertigo, tinnitus, hearing changes, cranial nerve palsies, diplopia, and even cortical blindness. In addition, the associated features of migraine such as nausea, photophobia, and phonophobia may occur.6, 9, 10

Post‐dural puncture headache may first develop from hours to months following dural puncture.11, 12 About 90% of the headaches start within the first 72 hours and 66% within the first 48 hours of LP.11, 12 Rarely, the headache develops between 5 and 14 days after the procedure.11, 12

DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS

Cause of Low Pressure Headache.— Post‐dural puncture headache is included in the International Classification of Headache Disorders, second edition (ICHD‐2, Table 1). The diagnostic criteria require a strong orthostatic component in the setting of a dural puncture. The differential diagnosis of PDPH is broad and includes many forms of orthostatic headaches (Table 2). Differential diagnosis is particularly important in the postpartum period; headache or neck/shoulder pain occurs in the first week postpartum in about 40% of women.13 These headaches are often attributed to dural leak because epidural anesthesia is very common. Among postpartum headaches, PDPH was the identified cause in a minority of women.14 In a retrospective review of 95 postpartum women with head and neck pain after delivery, Stella et al found that 81% received epidural blocks.15 Post‐dural puncture was the identified cause of headache in only 16%. Most headaches were considered tension or migraine headache (47%).15

Table 1.—. The International Classification of Headache Disorders Diagnostic Criteria for Post‐Dural Puncture Headache 55

(A) Headache that worsens within 15 minutes after sitting or standing and improves within 15 minutes after lying, with at least one of the following and fulfilling criteria C and D

• Neck stiffness

• Tinnitus

• Hypacusia

• Photophobia

• Nausea

(B) Dural puncture has been performed

(C) Headache develops within 5 days of dural puncture

(D) Headache resolves either

1. Spontaneously within 1 week

2. Within 48 hours after effective treatment of the spinal fluid leak (usually by epidural blood patch)

Table 2.—. Differential Diagnosis of Headache After Dural Puncture and a Differential Diagnosis of Orthostatic Headache 9, 19, 56-58

Differential diagnosis of orthostatic headache

1. Spontaneous low pressure headache

2. CSF shunt overdrainage

3. True hypovolemic state

4. Traumatic CSF leak (definite trauma, diagnostic dural puncture, accidental dural puncture during spinal anesthesia, spinal or cranial surgery)

5. Spontaneous CSF leak

Idiopathic

Weakness of dural sac (meningeal diverticula, abnormalities of connective tissue)

Dural tear from spondylosis or disc herniation

Trivial trauma

Differential diagnosis of headache after dural puncture

1. Nonspecific headache

2. Migraine

3. Caffeine‐withdrawal headache

4. Meningitis

5. Sinus headache

6. Pregnancy‐induced hypertension (pre‐eclampsia)

7. Drugs (cocaine, amphetamine)

8. Pneumocephalus‐related headache

9. Cerebral venous thrombosis

10. Subdural hematoma

11. Subarachnoid hematoma

12. Brain tumor

13. Lactation headache

14. Stroke (ischemic and hemmorhagic)

15. Posterior leukoencephalopathy

• CSF = cerebrospinal fluid.

Patients with postural headaches beginning after a dural puncture or epidural anesthesia should also be evaluated for secondary headache disorders other than PDPH.4 Postpartum, venous sinus thrombosis, and subdural hematoma are well known to cause positional headache.16 In a series of 22 postpartum women with headache who underwent cerebral imaging because of focal neurological deficits and/or failure to respond to initial therapy, 68% had abnormal radiographic findings; common findings included intracranial hemorrhage, cortical vein thrombosis (CVT) or sagittal sinus thrombosis (SST) and vasculopathy.15 Headache beginning immediately postpartum or more than 72 hours postdelivery may be attributable to secondary causes other than PDPH, particularly CVT and SST.17 Puerperal state is an important risk factor for CVT and changes in intracranial pressure because of inadvertent LP can lead to a subdural hematoma if shifting of the brain leads to tearing of dural bridging veins.4, 17, 18

Ischemic stroke can also present with a headache and is an important cause of maternal morbidity and mortality accounting for as many as 12% of maternal deaths.19

Other headache syndromes with postural features include migraine, cervical facet syndrome, and headaches associated with postural orthostatic tachycardia.4, 20 Migraine often remits during pregnancy and may relapse after childbirth; but it may also occur for the first time in the postpartum period.17

Clinical Maneuvers and Diagnostic Tests.— Several screening maneuvers may aid in the differential diagnosis of orthostatic headache. Firm continuous pressure on patient's abdomen by the examiner's hand may relieve headache by increasing CSF pressure.21 Relief usually develops within 30 seconds and the headache recurs once pressure is released in low pressure headache.21 Rozen et al placed patients in the Trendelenburg position to screen for low CSF pressure; a head‐down tilt theoretically causes a rapid increase in intracranial CSF pressure.20 Placement of an individual with decreased intracranial pressure in Trendelenburg position for 1‐2 minutes may lead to relief of PDPH.20 Continuation in Trendelenburg position for a longer period of time may lead to a pain‐free state, with return of pain when resuming a flat or supine position.20 The true sensitivity and specificity of the above mentioned maneuvers is unknown and their utility in diagnosis of PDPH is unproven.

Imaging is used both to confirm the diagnosis of PDPH and to identify or exclude other causes of headache.6, 20 In low pressure headache, magnetic resonance imaging (MRI) of the brain may show diffuse pachymeningeal enhancement with administration of gadolinium (Fig.). As the brain shifts downward, there may be cerebellar tonsillar descent, with crowding of the posterior fossa and obliteration of the basilar cisterns.6, 20 Enlargement of the pituitary gland and decreased ventricular size may be seen as well. In the setting of low pressure, meningeal thin‐walled vessels are thought to dilate, giving rise to pachymeningeal enhancement.4 However, the MRI is not perfectly sensitive; that is, even in the presence of a large CSF leak, the MRI may be normal.20

Figure Figure.—

Pachymeningeal enhancement on post‐gadolinium MRI in PDPH. From Mayo Clin Proc., 22 with permission. MRI = magnetic resonance imaging; PDPH = post‐dural puncture headache.

Finally, a diagnostic LP may demonstrate a low CSF opening pressure, the so‐called dry tap, a slightly raised CSF protein, and a rise in CSF lymphocyte count.4 However, low opening CSF pressure is not required for diagnosis of PDPH and cases of PDPH with normal opening pressure as well as cases of no PDPH and decreased opening pressure have been described.22

FREQUENCY AND RISK FACTORS FOR PDPH

The frequency of PDPH following LP varies with characteristics of individual patient (nonmodifiable risk factors), the type of needle and technique used (modifiable risk factors), as well as the diagnostic definition of PDPH and method of follow‐up. Most studies of PDPH were published prior to development of the ICHD‐2 criteria and so did not use them. Strupp et al was the only researcher who used original ICHD‐1 criteria for PDPH.23, 24 Most studies define PDPH as “headache exacerbated by sitting or standing up.”25 Although these criteria lack specificity, following Evans et al we have included all of the available studies in this review.25

Risk factors for PDPH are presented in Table 3. Nonmodifiable risk factors include age, female gender, low body mass index (BMI), history of prior PDPH, and history of chronic headache.4, 5 Modifiable risk factors are not discussed in this review, but will be presented in Part II of the review.

Table 3.—. Modifiable and Nonmodifiable Risk Factors for PDPH

Nonmodifiable risk factors

1. Age

2. Female gender

3. Low BMI

4. History of prior PDPH

5. History of chronic headache

Modifiable risk factors

1. Size of the spinal needle

2. Needle shape

3. Bevel orientation and angle of insertion

4. Stylet replacement

5. Operator experience (with spinal anesthesia)

• BMI = body mass index; PDPH = post‐dural puncture headache.

Age.— Age is a well‐studied risk factor for development of PDPH.26-32 Over the age of 60, PDPH is rare.28, 29 Risk is highest in 20 to 30‐year‐olds. This age group is 3‐5 times more likely to develop PDPH than those after 60.27-29 In a study of 9277 patients receiving spinal anesthetics, the incidence of PDPH in age group of 20‐29 (16%) was twice that of age group of 40‐49 (8%) and 4 times that of age group of 50‐59 (4%).28 Similarly, after 1021 spinal anesthesations, PDPH was most common between the ages of 20‐30 and the incidence of PDPH began to decrease after age of 40.29 Other smaller studies have shown consistent age‐dependent reduction in PDPH incidence. The discrepancy in incidence of PDPH may be due to the fact that older individuals have lower CSF pressures.33 In children, after LP, the incidence of PDPH ranges from 2% to 15%.34 The incidence of PDPH in adolescents may be close to the PDPH incidence in adults with Scher et al reporting PDPH incidence of close to 20% in 26 patients with ages 10‐18 (median 14) receiving a diagnostic LP.35 The incidence of PDPH in younger children may be very low with Scher et al reporting a 0% PDPH incidence in 29 children with ages 2‐9 (median 6).35 However, the notion of low PDPH incidence in children was challenged by Kokki et al, who found that PDPH incidence was not age‐related.36 The authors suggested that the incidence of PDPH in children may be underreported because of inability of infants to report presence of headache as well as a failure by the clinicians and parents to document physical and behavioral changes associated with positional headache.36 Hence, whether younger children have PDPH incidence comparable to adults is still an open question.

The mechanisms underlying age‐related difference in PDPH are likely multifactorial. Some of the proposed mechanisms include differences in pain perception, psychosocial factors (women are more likely to report pain than men), hormonally related differences in cerebral vessel reactivity (possibly because of estrogen‐induced increases in substance P receptor reactivity), differences in elasticity of the dura, and possibly gender‐based differences in cranial morphology.37, 38

Sex.— Women have almost twice the risk of developing a PDPH in comparison with men.26-28, 30, 31, 37, 39 Flaatten et al found that women in their 30s had a 3 times higher PDPH incidence than men from the same age group.27 Both Flaatten et al and Lynch et al found the difference in PDPH to be unrelated to age.27, 39

Other Risk Factors.— Also, previous history of PDPH is a risk factor for a recurrence of PDPH.38 Amorim et al found that 19% of patients with prior PDPH developed second PDPH vs 6.9% of patients without prior PDPH who developed it de novo.38 Meanwhile, Lybecker et al found that 2 of 3 patients with previous PDPH had developed it again compared with 3 of 114 patients who developed it for the first time. Also, Clark et al found history of chronic headaches in close to 40% of patients with PDPH, while Kuntz et al found that the group of patients with headache 1 week before the procedure had PDPH incidence of close to 70% compared with 30% PDPH incidence in nonheadache group. Increased incidence of PDPH with history of previous PDPH and chronic headache is likely to reflect the underlying predisposition of this group to developing headaches. Interestingly, both Kuntz and Levi et al found significantly lower BMI in patients who developed PDPH.30, 31, 39, 40 Furthermore, morbid obesity may decrease the incidence of PDPH.41 The mechanisms underlying the association between BMI and PDPH risk are unknown. Finally, race and pregnancy are not believed to be risk factors for development of PDPH.12, 42 It is plausible that leakage of CSF may be exaggerated by the rises in intra‐abdominal pressure during labor, but conclusive evidence is lacking.1

PATHOPHYSIOLOGY

The exact mechanism of headache in PDPH is uncertain, but several theories have been proposed. These mechanisms include downward pull on pain‐sensitive structures caused by CSF volume loss, compensatory vasodilation of intracranial vessels in an attempt to maintain intracranial volume constant (Monro‐Kellie doctrine) and hypersensitivity to substance P. PDPH is attributed to relative CSF hypovolemia resulting from persistent leakage of CSF.43 A large hole in the dura leads to a greater CSF volume loss, which increases the odds of developing PDPH.44 Orthostatic headache arises when approximately 10% of the estimated total CSF volume is lost.45 Individual characteristics are also thought to play an important role in the development of PDPH. Some develop headache with little CSF volume loss, while others do not develop PDPH even with large CSF volume losses.7, 43, 46, 47 Although CSF leak is a plausible cause for low CSF opening pressures, the relationship between opening pressure and presence of PDPH is uncertain.22 We will consider these theories one at a time.

Perhaps the most common theory for the origin of headache is that downward pull on pain‐sensitive structures when patient assumes an upright position generates pain.38 When CSF volume is low, in the upright position, gravity causes CSF to move into the spinal dural sac.7, 38 As a consequence, the brain loses buoyancy. As it sags, this creates tension on the meninges and other pain sensitive intracranial structures, like vessels and nerves.38, 48 Downward displacement of intracranial structures has been demonstrated radiologically in PDPH.20 Finally, the presence of downward pull on the brain with CSF hypotension is supported by the finding that in severe cases of CSF hypotension the sagging of the pons against the clivus can result in cranial nerve palsies and in extreme cases confusion, obtundation, or even coma.49

Another mechanism for PDPH emerges from the Monro‐Kellie doctrine. As total intracranial volume must remain constant, loss of intracranial CSF volume must be replaced, most importantly through an increase in intracranial blood volume.38 Cerebral arterial and venous dilation that occurs as part of this compensatory process may lead to PDPH.50 There is some direct evidence to support this hypothesis. Cerebral venodilation leading to decreased intracranial blood flow has been demonstrated by transcranial Doppler ultrasound.50 Venous and arterial dilation in a setting of CSF hypovolemia may be mediated by adenosine receptors.50 This may provide a basis for the therapeutic use of caffeine.51-53

A third mechanism involves hypersensitivity to substance P with an associated up‐regulation of neurokinin 1 receptors (NK1R). Low CSF levels of substance P are associated with a 3‐fold increased risk of PDPH.10 Pregnancy may increase the risk of PDPH because of decline in CSF density and increases in intra‐abdominal pressure; these factors may induce CSF leakage.1, 54 Finally, posterior dura varies in thickness greatly within individual and between individuals.4 It is thought that thicker areas of dura are less likely to leak CSF.4 Hence, PDPH incidence can arbitrarily depend on the location of dural perforation.4

CONCLUSION AND FUTURE PERSPECTIVES

In summary, PDPH is a commonly encountered problem in clinical practice with younger individuals and women being at a higher risk of developing this condition. It is important for general clinicians, neurologists, and headache specialists to be familiar with preventive measures for PDPH as these can significantly reduce PDPH incidence and associated morbidity. Furthermore, as a number of potentially life‐threatening conditions like CVT, SST, and subdural hematoma may mimic PDPH, knowledge of clinical characteristics of PDPH as well as diagnosis and differential diagnosis of PDPH is paramount. The pathophysiology of PDPH is not entirely clear, but CSF volume loss, intracranial blood vessel vasodilation, and low CSF levels of substance P may play a role. ICHD‐2 criteria for PDPH diagnosis should be implemented in future studies. Studies further exploring the pathophysiology of PDPH are needed.

STATEMENT OF AUTHORSHIP

Category 1

• (a)

  Conception and Design

  David Bezov, Richard B. Lipton, Sait Ashina

• (b)

  Acquisition of Data

  David Bezov, Richard B. Lipton, Sait Ashina

• (c)

  Analysis and Interpretation of Data

  David Bezov, Richard B. Lipton, Sait Ashina

Category 2

• (a)

  Drafting the Article

  David Bezov, Richard B. Lipton, Sait Ashina

• (b)

  Revising It for Intellectual Content

  David Bezov, Richard B. Lipton, Sait Ashina

Category 3

• (a)

  Final Approval of the Completed Article

  David Bezov, Richard B. Lipton, Sait Ashina