7.0: Headache associated with nonvascular intracranial disorder

7.1: High CSF Pressure

Raskin (1988) points out that the frequency and severity of headaches that accompany intracranial lesions correlate poorly with the presence or degree of intracranial hypertension (Northfield, 1938; Pickering, 1939); the latter therefore, is unlikely to be an important mechanism of headaches in these settings. Whether intracranial hypertension per se can result in headache at all is controversial. Kunkle et al (1943) found, in four subjects, that elevation of the cerebrospinal fluid pressure to between 680 and 850 mm H2O for 2 minutes did not result in headache, and concluded that intracranial hypertension does not produce headache. However, Fay (1940) performed a similar experiment and found thresholds for headache production that varied greatly among subjects. Headache is often the sole symptom of the acute increase in intracranial pressure that occurs with shunt malfunction in treated hydrocephalic patients (Dahlerup et al, 1985). It may be that the rate at which the CSF pressure increases is critical to production of headache.

The syndrome of benign intracranial hypertension (pseudotumor cerebri) is associated with headache in nearly all patients (Ahlskog and O’Neill, 1982; Corbett, 1983). The lateral ventricles are not enlarged in this disorder, so traction and displacement of periventricular pain-sensitive structures are not likely to be a mechanism of headache, as they might be in obstructive hydrocephalus. Headache is sometimes, but inconsistently, relieved by lumbar puncture in this disorder, despite the restoration of CSF pressure to normal levels. Intracranial hypertension in animal results in compensatory dilatation of pial arteries (Forbes and Wolff, 1928), vasomotor instability, and loss of cerebral vasomotor regulation (Langfitt et al, 1965). The magnitude and persistence of these vasomotor responses may be important to the mechanism of headache. Studies of cerebral hemodynamics in patients with benign intracranial hypertension have shown that cerebral blood volume is increased but have led to differing conclusions regarding the impairment of autoregulation in this condition (Raichle et al, 1978; Mathew et al, 1975). The mechanism underlying the increase intracranial pressure appears to be an increased flow resistance in the arachnoid villi and/or an increase in dural sinus pressure (Black and Conner, 1986; Aisenberg and Rottenberg, 1980). How this occurs remains unknown. Many patients who have a pseudotumor cerebri may have undiscovered venous sinus thrombosis. Many of the patients do not have arteriography and the etiology is not discovered unless arteriography or contrast enhanced MRI is used. I have recently seen two patients in whom original MRI were misinterpreted as normal in patients with pseudotumor. Both these patients had thrombosis of the transverse sinus. Both were young overweight women, but no coagulopathy was discovered. They were successfully treated with anticoagulation as well as measures, such as Diamox, to reduce intracranial pressure.

In pseudotumor cerebri additional symptoms may include spinal and radicular pain (Bortoluzzi et al, 1982) or facial pain (Hart and Carter, 1982), and they may occasionally be the presenting complaints. Many patients with pseudotumor continue to complain of headache long after papilledema and increase intracranial pressure have subsided, which suggests that the mechanism of headache in this condition depends on other factors. Propranolol and ergonovine treatment has helped to quell headaches for many of these patients (Raskin, unpublished observations).

7.2: Low CSF pressure

Lumbar puncture headache

This is a frequent complication that curiously occurs more often when the cerebrospinal fluid is normal than when it shows evidence of disease. According to Davenport (1964), up to 30% of patients develop headache following lumbar puncture. It apparently occurs from a fall in intracranial pressure and continues during the period when the pressure remains low as a result of leakage of fluid through the puncture wound in the dura. The size of the needle used, the amount of fluid withdrawn, and multiple punctures are factors that influence the development of headache.

The headache, which is best described as ``bursting,'' is often extremely severe. The pain is usually worse in the occipital region and forehead, but it may extend into the neck and shoulders. It is completely relieved when the patient lies down with the feet at a higher level than the head. The headache usually lasts for only about 2 or 3 days but may persist for as long as 2, or even 3, weeks, and rarely for many months.

Cisternal and ventricular punctures are rarely followed by severe headaches, but when air, anesthetic, or radiographic contrast material (pantopaque or metrizamide) are injected into the ventricles or particularly when they are injected into the subarachnoid space by the lumbar route, severe and prostrating headaches may occur.

The frequency and severity of headaches following lumbar puncture may be reduced by using a fine gauge needle, having the patient lie prone for 4 hours after lumbar puncture, and maintaining bed rest for 24 hours after the procedure. Lying flat in bed and drinking large quantities of fluids, particularly those with caffeine (tea, coffee, some soft drinks) often give significant relief (Mathew, 1978).
Lumbar puncture headache, ordinarily disabling only for a few days, has no serious sequelae. Occasionally, however, lumbar puncture may produce subarachnoid bleeding from laceration of vessels of an unexpected spinal cord arteriovenous malformation. Subarachnoid hemorrhage may also occur from rupture of a congenital aneurysm. In other cases where there is an intracranial mass lesion, particularly when it is situated in the posterior fossa, lumbar puncture may result in herniation of the medulla into the foramen magnum and thus cause sudden death of the patient.

Dr. Frank B. Walsh (Walsh and Hoyt, 1969) observed an individual who suffered from severe vascular hypertension but did not complain of headache except following lumbar puncture. On three successive taps, the spinal fluid pressure was 575, 405, and 600 mm of water.

As emphasized by Raskin (1988), headache following lumbar puncture usually begins within 48 hours and is dramatically positional: it begins when the patient stands or sits upright; there is relief upon reclining or with abdominal compression. It is believed that the mechanism is the loss of the CSF position to the brain so that when a patient is upright there is probably dilatation and tension based on the brain’s anchoring structures, the pain-sensitive dural sinuses, resulting in pain.

Headache due to cerebrospinal fluid fistula

Low CSF or lumbar puncture headaches usually subside spontaneously in a day or two and do not require aggressive therapy. However, some patients are debilitated by persisting headache and these patients are candidates for injection of saline or blood in the epidural space. Based on the observation that headache seems to follow traumatic, bloody lumbar punctures less often than atraumatic spinal taps, the epidural injection of autologous blood was begun (Gormley, 1960). It has now become the most predictable and rapidly effective measure for stopping post-lumbar puncture headache (Olsen, 1987).

7.3: Intracranial infection

Such headaches are usually severe. They occur in any of the various types of meningitis or as a result of intracranial bleeding where the blood is in contact with the meninges. There is associated cervical rigidity, tenderness of the scalp, and almost always a positive Kernig's sign. The headache itself is continuous and nonlocalized. Examination of the spinal fluid reveals an excess of white blood cells in inflammatory processes and either blood or xanthochromia in cases of subarachnoid hemorrhage.
The pain of injury or inflammation, as well as the pain provoked by extravascular blood in the subarachnoid space, appear to be caused by the concerted action of substances that are contained in whole blood and are formed from inactive plasma precursors. Serotonin and the kinins are probably the most important of these agents.

Individuals who suffer from fever that is associated with septicemia or infectious diseases often complain of dull, deeply situated headaches and aching of their eyeballs. The mechanism of such pain is not understood. Sutherland and Wolff (1940) have compared it to headaches produced by injections of histamine. Wolff (1955) noted cerebral vasodilation in experimental animals that had been given typhoid vaccine. He concluded that headaches associated with acute infections, sepsis, and bacteremia are primarily caused by the distention of intracranial arteries.

7.4: Intracranial sarcoidosis and other noninfectious inflammatory diseases

A diffuse sometimes severe headache may accompany chronic glanulomatous disorders involving the intracranial structures. One of the most frequent is sarcoidosis usually producing its effect from invasion of the basal structures and cranial nerves. Sarcoid involvement of the cranial nerves and visual pathways is discussed elsewhere, however, one of the presenting features may be a diffuse headache not associated with increased intracranial pressure.

7.5: Headache related to intrathecal injections
Local pain at the sight of injection, if it is through the lumbar base, is not uncommon, however, more frequently a diffuse headache may follow injection of agents such as methotrexate used in the treatment of meningeal carcinomatosis. There is usually an associated nausea and it is uncertain whether this is part of a type of vascular migraine-like phenomenon or due to the chemotherapeutic agent.

In the IHS classification category 7.5.1 is a direct effect of the agent with the following diagnostic criteria:

A. Headache follows intrathecal injection within 4 hours.

B. Headache is diffuse and present also in the recumbent position.

C. Headache clears completely within 14 days. (If it persists consider post lumbar puncture)

If it is due to a presumed chemical meningitis, the category is 7.52 with the following diagnostic criteria:

1) Headache follows intrathecal injection within 5 to 72 hours

2) Headache is diffuse and present also in the recumbent position

3) Cerebrospinal fluid pleocytosis with negative culture

7.6: Intracranial neoplasm (Brain tumor headache)

It is important that increased intracranial pressure from tumors be considered as a cause of headache. About 75% of patients with intracranial tumors complain of headaches, and increased intracranial pressure is frequently associated with such lesions. The headaches that are associated with increased intracranial pressure are variously described as ``bursting,'' ``boring,'' etc. Often they are paroxysmal at the onset and rather frequently occur at night when they may last only for an hour or two. The patient may be able to point to the area of maximum headache. As time passes, the headaches last longer and become associated with vomiting that may be of the projectile type. It is interesting that the nature of the vomiting has received more attention than it merits, since true projectile vomiting is rare. Sneezing, vomiting, straining at stool, bending over, etc., are likely to increase the pain which may also be exacerbated by changes in position.

Usually headaches have little localizing value in cases of cerebral tumor, although there are certainly exceptions to this generalization. In cases of dural tumor, pain is often described as being present in the skull at a place overlying the tumor that has impinged on the bone. In some such cases, it is possible to palpate a protrusion of the skull, and we have observed several patients in whom a meningioma was suspected from palpation. In malignant tumors invading the bone or the meninges, there may be localized pain.

Intermittent headaches are occasionally caused by intraventricular tumors, primarily cysts. According to Dandy (1932), the best examples of such tumors are cysts obstructing the ventricular drainage posteriorly, but similar obstructions occur in the third and lateral ventricles. Dandy remarked upon the proof of the ball valve action of such tumors by the high pressure that is present in the ventricles at the time of the height of the headache and by the absence of increased pressure when headache is absent.

Further proof that headache is associated with increased intracranial pressure is seen in individuals who have been subjected to subtemporal or suboccipital decompression since, in most cases, the decompression bulges when the headache is severe, and it collapses when the headache has disappeared (Dandy, 1932). However, we have observed one patient in whom, after subtemporal decompression for presumed pseudotumor cerebri, the headaches continued. When they were most severe, the decompression site was quite flat or even depressed. The patient was more comfortable when the decompression was bulging. This observation correlates with the findings of Kunkle et al. (1943) who found that in four subjects, elevation of intracranial pressure to between 680 and 850 mm of water for 2 minutes did not result in headache. Fay (1940) performed a similar experiment and found that cerebro-spinal fluid thresholds for headache production varied greatly among individuals. Headache arising as a result of intracranial abscess is indistinguishable from that caused by neoplasm except that it tends to be more constant.

Northfield (1938) suggested that it is not the increased intracranial pressure that is responsible for the associated headache, but rather the sudden alteration of pressure, whether it be an elevation or a lowering. Northfield cited clinical and experimental observations to support this concept. Wolff (1955) concluded that the headache associated with either increased or decreased intracranial pressure results from traction upon, or displacement of, pain-sensitive intracranial structures and is independent of generalized intracranial pressure changes per se. The tumor or abscess may press upon the dura overlying the brain, on the falx, or on the tentorium. It may result in traction or pressure on blood vessels that are pain-sensitive or the lesion may put direct pressure upon the sensory nerves. Hydrocephalus associated with cerebral tumor accounts for diffuse pressure that may influence widely separated pain-sensitive structures.

Northfield's paper concerning headache in cases of cerebral tumor is instructive, and his findings conform fairly well to the following observations of Wolff (1955).

(1) Although the headache of brain tumor is often referred from a distant intracranial source, the area of pain approximately overlies the tumor in about one-third of all patients. (2) The presence of headache in the absence of papilledema in patients with brain tumor is of great localizing value. In two-thirds of the patients with such headaches, the area of pain immediately overlies, or is near, the tumor and, when it is unilateral, it is always on the same side as the tumor. (3) Headache is almost always present with tumors of the posterior fossa. (4) Headache may be absent with any of the common types of supratentorial tumor. (5) The headache associated with tumors of the posterior fossa is almost always over the back of the head, although it may occur elsewhere. (6) Headache is usually the first symptom of tumors of the posterior fossa except for growths in the cerebellopontine angle. (7) Headache is the first symptom in one-third of supratentorial tumors. (8) The headache of tumors of the cerebellopontine angle is frequently, and sometimes solely, postauricular. (9) Headache due to supratentorial tumors is rarely in the back of the head unless it is associated with papilledema. (10) When supratentorial tumor causes headache in the back of the head, headache in the front of the head is usually also present. (11) Headache which is both frontal and occipital indicates extensive displacement of the brain and has little localizing value. (12) The headache of brain tumor is usually intermittent, but when it is continuous, its value in localization is greatly enhanced.

 

 

Return to Other Headaches