The postpartum patient with a headache
A 29-year-old woman, 5 days postpartum after an uncomplicated vaginal delivery, presents with a severe fronto-occipital headache that is worse on standing and improves when lying flat. She has no fever, no neck stiffness, and no focal neurology. Blood pressure is 118/76 mmHg, and urinalysis is normal. A non-contrast CT head is normal. The obstetric team asks whether she needs a lumbar puncture or an MRI. You are asked to advise on the anatomical basis of her symptoms and the most appropriate imaging pathway.
1. What is the likely diagnosis, and what anatomical structures explain the postural nature of the headache?
Read the worked answer
Likely diagnosis
- Post-dural puncture headache (PDPH) or spontaneous intracranial hypotension related to delivery. Even without a documented dural puncture, obstetric patients can develop CSF leaks from dural tears during labour.
Anatomical basis
- The brain and spinal cord are suspended in CSF within the meninges. The choroid plexus produces CSF, which circulates through the ventricles and subarachnoid space and is absorbed by arachnoid granulations.
- When CSF volume is reduced, the brain 'sags' and traction is placed on pain-sensitive structures: the meninges, bridging veins, and cranial nerves (especially the tentorium and the trigeminal and vagus nerves).
- Traction on the tentorium cerebelli and the trigeminal nerve explains the fronto-occipital distribution. Traction on the vestibulocochlear nerve can cause tinnitus and hearing changes.
Postural nature
- In the upright position, gravity increases the downward displacement of the brain, worsening traction and pain. Lying flat reduces displacement and relieves symptoms.
Why not lumbar puncture first
- Lumbar puncture may worsen the leak and is not required for diagnosis if the clinical picture is typical.
2. What MRI findings would you expect, and how do they relate to the underlying anatomy?
Read the worked answer
MRI findings of intracranial hypotension
- Pachymeningeal enhancement - diffuse, smooth, non-nodular enhancement of the dura. This reflects venous engorgement as the brain compensates for reduced CSF volume.
- Subdural effusions - bilateral, thin, often asymptomatic.
- Sagging brain - downward displacement of the cerebellar tonsils, brainstem, and optic chiasm. The midbrain may appear 'pulled down'.
- Venous engorgement - distension of the dural venous sinuses and pituitary hyperaemia (the pituitary may appear enlarged).
- Subarachnoid space effacement - reduced CSF signal around the brainstem.
Anatomical correlation
- The Monro–Kellie doctrine states that the cranial vault is a closed box containing brain, blood, and CSF. When CSF volume falls, venous blood volume increases to maintain intracranial pressure, producing the pachymeningeal enhancement and venous engorgement.
- The sagging brain reflects loss of buoyancy.
Practical point
- MRI with contrast is the imaging study of choice. If a spinal CSF leak is suspected, a spinal MRI or CT myelogram can localise the leak.
3. If the MRI confirms intracranial hypotension, what management options would you discuss, and what are the anatomical considerations for an epidural blood patch?
Read the worked answer
Management options
- Conservative - hydration, caffeine, bed rest, and analgesia. Many cases resolve spontaneously.
- Epidural blood patch - injection of autologous blood into the epidural space at the level of the suspected leak. The blood clots and seals the dural defect.
- Epidural saline or fibrin glue - alternatives if blood patch is contraindicated.
- Surgical repair - for refractory leaks with a localised defect.
Anatomical considerations for blood patch
- The epidural space lies between the dura mater and the vertebral canal periosteum. It contains fat, venous plexuses, and connective tissue.
- The lumbar approach (L3–L4 or L4–L5) is safest, avoiding the conus medullaris (which ends at L1–L2 in adults).
- The needle must not enter the subarachnoid space; confirmation of epidural placement is by loss of resistance.
- The blood volume (typically 10–20 mL) should be enough to coat the dural defect but not so much as to cause compression.
Counselling
- Explain that the blood patch is effective in most patients but may need to be repeated.
- Discuss the small risk of infection, bleeding, and transient back pain.
- If symptoms persist despite a blood patch, further imaging (spinal MRI or CT myelogram) is warranted to localise the leak.