Nutshell Lumbar Spine: Back to Basics

“Low Back Pain” (LBP) will be experienced by ~80% of adults; leading to employment disability, lost work days, and contributing to other health issues secondary to sedentary lifestyle, reduced activity, or loss of agility. In 2010, it ranked #3, behind only ischemic heart disease and COPD, as a leading cause of US mortality, morbidity, lost work days, or poor health.

LBP is an overly-inclusive term; it would be helpful if clinicians provided additional information such as acute vs chronic, precipitating circumstances, localized vs radiating, abrupt activity-related onset vs age-related evolution, or potentially related co-morbidities such as renal disease, chronic steroid use, IVDA, MRSA sepsis, recent instrumentation, new or unusual activity (weekend warriors, training for marathon, new employment requiring lifting, wearing 5” high heels), chronic or acute trauma, associated fevers, etc.—but sadly the majority will tend to state only ‘Pain”. Spelunking the patient’s records is therefore invaluable.

Common risk factors for back pain include age, obesity, pregnancy, occupational strain, suboptimal fitness levels, sporadic athletes (weekend warriors interspersing intense effort with long spells of non-activity), risk factors such as HLA-B27, stress, depression, scoliosis/structural or muscle imbalance, and rising fast—the child with the over-loaded backpack pre=teen? Keep it to 15% of child’s body weight PLEASE!).

As with any radiographic study, adequate localizing and temporal history, important co-morbidities or specific details, and search for comparison images are PRICELESS. Remember the comparison may not be a prior L spine exam—abdominal or pelvic radiographs, CT,MR, barium studies, even a hysterosalpingogram, may show an area of current concern at an earlier point of time allowing a ‘worry/don’t worry’ or ‘changed/unchanged’ assessment.

ACUTE TRAUMA moves faster and is higher-risk for morbidity/mortality. The ACR AC suggestions change accordingly.

REMEMBER the Hopkins Emergency Department cannot always adhere to community-medicine criteria or practices.  Our patients are far more complex, often quite ill or unstable, quite likely to have inadequately documented histories and/or inadequately treated conditions, often carry long lists of comorbidities, may have suboptimal diets, personal habits, and living situations; and may be on many prescription, OTC, and self-administered medications and substances.


Initial consultation for ‘uncomplicated’ back pain does not mandate immediate imaging; ALL modalities rate only a ‘2’. Up to 6 weeks medical management and physical therapy are advised in the absence of obvious co-morbidity or historical  red flags. Add a history of trauma, osteopenia, steroid use or advanced age and the initial entrance exam is RADIOGRAPHY (7). 

Where history includes neoplasm, infection, or immunocompromised, MR without and with IV contrast ranks 8, without contrast 7; although most clinicians would start with Radiography first, it is insufficient to r/o these entities.

If a patient in general fails 6 weeks conservative management, MR without contrast ranks 8 although again, most would also screen with radiography. History of prior lumbar instrumentation, progressing/new signs or symptoms? MR.

Lumbar spine emergencies—rapid progression of neurologic signs/symptoms, sudden loss of bowel/bladder control, etc. —> MR without contrast, STAT.

TRAUMA?? Most clinicians would screen lower-grade suspected trauma with radiographs first especially if there are distinct localizing signs in adults, and XR gets a 9 for screening under-14 yr. olds; but non-contrast CT is definitive (9) where indicated in adults by severity of trauma, neurologic findings, etc.

Mechanism of trauma counts—flexion over seatbelt when car struck tree at 60 mph? Hyperextension falling backwards from stance, vs off a speeding horse?  Overextended or awkward tennis, baseball, or golf swing? Twisting injury in soccer game or gymnastics class?  Osteoporotic, bent over and sneezed and got acute pain? Over-loaded straining to achieve your Cobra or Camel yoga postures?  Axial loaded falling 15 feet off roof while cleaning gutters on ladder? All more helpful than “Pain”, or “Trauma”.

The variables are complex and highly individualized; and like all good recipes, subject to tweaking. The ACR AC provides a useful guide but individual situations, modality availability, etc, may lead to on-site variations in choices.

START with the AP view

Air and feces, osteopenia, foreign matter, marked scoliosis, hardware, rotation, may all limit detail—state limits of exam. Bilateral symmetry is your friend

Count apparent number vertebral bodies . Five is expected, but a surprising number of patients have either 4 or 6. Scoliosis images going from distal c spine to sacrum are ONLY way to ascertain true number of thoracic and lumbar bodies, since T12 ribs can be hypoplastic and the lumbo-sacral junction is prone to anatomic variation.

'Sacralization’ and ‘lumbarization’ are being replaced by ‘lumbosacral transitional vertebrae’, LSTV; pseudarthrosis implies there is apposed marginal sclerosis implying biomechanical imteraction/impingement of the transitional processes ---estimated at anywhere from 4-20% of the population). LSTV and altered biomechanics are believed to contribute to LBP, but cannot be assumed to be the sole cause.

Please dictate minor posterior fusion defects as ‘anatomic variant posterior fusion defect spinous process”, NOT ‘spina bifida occulta’. These are common, usually incidental findings; some think 10-20% of the population may have it.  And if one can SEE it; it is NOT ‘occulta’.  And that minor variation CAN be associated with other significant spinal dysraphisms, cord tetherings, diastematomyelias, dermal sinus tracts or other anomalies

The midline spinous processes (inverted ‘V’) should be co-linear and centered between the ‘owl-eyes’ of the pedicles (tubes running in sagittal plane, therefore cancellous circles with cortical peripheries; blastic and lytic pedicle lesions can modify or distort these ‘eyes’). MISSING spinous process implies surgical resection as rule; not all laminectomies are stabilized with hardware.

SACRUM, SI joints, and sacral foramina—see PELVIS handout.  If you do not have this early in your check list, you will forget—air/feces/lordosis often obscure detail.

The vertebral bodies are curving in the sagittal plane; heights/widths/radiodensity should be similar but definition of end plates and body integrity is better on lateral.

Transverse processes—rare to see anything significant; normal variant 2dary ossification center FAR more likely than isolated trauma to the process.  More severe trauma—all bets off.

PSOAS margin—Bilaterally symmetric inverted ‘V’ of tissue interfaces as psoas travels inferolaterally from the thoracolumbar junction towards the medial iliac crests. Mention if seen (visible in +80% of normal; air and feces may limit).  This is a radiographic guide to the retroperitoneum; a patient with psoas abscess or hemorrhage may report ‘back pain’; and it is not uncommon for retroperitoneal, pelvic, or renal abscess or neoplasm to spread to the thoracolumbar junction via the valveless Batson venous plexus. We have many anticoagulated, immunocompromised patients, and patients living longer with neoplastic disease, all increasing the incidence of such hemorrhage, infection, or neoplasm.

Part of the pelvic ring, hips, abdomen are on the AP lumbar image—check for osteoarthrosis, calculi, surgical clips, abnormal renal silhouettes, old trauma…if it is there we are responsible for it.  This may all be at the end of the report:  “Again seen are…”  “Again noted are…”. “Surgical clips, partial visualization of known iliac crest harvest, incompletely seen left hip cartilage loss, unchanged…” (Remember to qualify when only part of something is visible.)


POSITIONING—vertebral bodies posterior margins should be superimposed, as are iliac crests, end plates, posterior elements—rotation or lateral flexion will offset these bilaterally symmetric margins. T12 through most of the sacrum should be visible.

ALIGNMENT—posterior vertebral bodies should form flowing lordosis. Use cephalo-caudal center of posterior body margins, as with C spine; osteophytes distort posterior endplate landmarks.  Flattening may be positional, spasm, or pain; comment on it, compare to old images if possible.

If there is anterior or posterior subluxation (SPONDYLOLISTHESIS) specify in mm or quadrants (Grade 1 = up to 25%, Grade 2 to 50%, Grade 3 to 75%, Grade 4 close to 100% subluxation of one end plate over/under another) and redundantly state “Anterior 3 mm subluxation of L3 under L2”… or ” … over L4”, for precision and clarity.

END PLATES = thin radiodense lines, flat or slightly concave, similar in density, thickness, and shape to levels above/below; not angled, nor blurred, nor thickened/sclerotic, zig-zagged, effaced.

DISC SPACES-- widest by 1-2 mm at L4-5 since this is the apex of lordosis; if 4-5 equals/less than level above/below ‘minimal narrowing’. (Caveat:  straightened lordosis may also subtly diminish this disc space to equal level above/below). NARROWING common with age and spondylosis, seen with osteophytes and sometimes discogenic sclerosis; WIDENING may imply vertebral body compression or rarely, facet subluxation distracting bodies. ‘Disc disease and osteophytes’ can be called just that, or ‘degenerative disc disease’, or ‘spondylosis, but NOT ‘arthritis’ or ‘osteoarthrosis’ since NOT synovial joints.

VERTEBRAL BODY – roughly rectangular with either straight or slightly concave margins all 4 visualized edges.  AP diameter should exceed cephalocaudal height and height should be similar all levels. This is cancellous bone; look for focal radiodensities or lucencies, smudged trabeculae, anterior wedging, biconcave endplates, other asymmetries or focal changes, comparing to level proximal and distal.

POSTERIOR ELEMENTS – facet joints form oblique flowing ‘shingles’ as inferior articular facet of the more cephalad body overlaps posteriorly on the superior articular facet of the more caudal body. The radiolucent space defined between facets, the linear posterior facet joint, is a true synovial joint and CAN be said to have ‘arthritis’ or ‘osteoarthrosis’.

The bony junction between each level’s superior/inferior articular facet is the PARS INTERARTICULARIS, literally ‘part in between’.  Seeing lucency here usually implies bilateral pars defect; unilateral pars defect may be visualized but not as reliably. If there is anterior/posterior subluxation of vertebral bodies look for pars defect; and v.v.

SPINOUS PROCESS – check to see if present; posterior laminectomy or laminotomy would be embarrassing to miss. LUMBAR HARDWARE/INSTRUMENTATION—need not be very precise naming/describing. “Instrumentation”, ‘fusion’, “posterior (or anterior)  hardware”, are correct if vague; most posterior surgical fixation are ‘rod and pedicle screws’, anterior instrumentation often ‘plate and screw and implants” but look at old reports for suggestions. ‘Anterior implants’ covers autologous and allograft (cadaver) bone, bone-simulating materials, metal, wire cages, etc—don’t exceed your actual grasp or knowledge. And ‘Surgery/fixation unchanged from last exam” is perfectly adequate. Look for changes in hardware position, hardware- bone interfaces (<1 mm is fine), fractured or separating rods, recurrence of infection/neoplasm.

AORTA, IVC FILTERS, HARDWARE, CALCULI, PILLS IN COLON—If it is visible on image one is responsible. Patients with aortic dissection or aneurysm may come in complaining of ‘back pain”.  Measure diameter, compare to old images (L spine, abdo, GI studies, CT….) if possible, convey if urgent/significant or if CHANGED from prior exams.

PILLS IN COLON of interest to clinician. Patient may have denied ‘taking meds’, not realizing OTC or health food store supplements count; or may complain ‘meds don’t work’ because enteric coating overly-protective; or doctor may not believe patient who claims to take meds if therapeutic windows suggest noncompliance.  Iff the dose is upped or pt. switches to non-enteric version, negative effects (gastritis, GI bleed, overdose or toxicity) could follow.


Evolved because most disc disease, ie the most common change with age, is at L4-5-S1.  Look at the better detail on this view but personal opinion:  Does not add much except dose.


Assess dynamic stability.  Look for anterior/posterior subluxation; while lordosis may flatten or reverse during range of motion, posterior vertebral body line and spinous process articulations should ‘flow’ and continue to be contiguous/smooth.

One must assess both (1) whether or not alignment maintained (ie, is there subluxation or change in visualized subluxation on the neutral view?) AND (2) adequacy of range of motion—‘Alignment unchanged and unremarkable between flexion and extension, with unremarkable range of motion”, ‘Unchanged…but with minimally limited range of motion’, ‘…markedly limited in extension, adequate in flexion’, ‘…but virtually no motion between flexion and extension”; “No evidence of motion fused levels between flexion and extension”. It is not a dynamic test if pt. DID NOT actually manage to move, due to pain, spasm, hardware, or fear.  If patient has fusion instrumentation, NOT moving between flexion and extension is a good thing.


Look for the ‘Scotty Dog’—the pars forms an eye, the transverse process the nose and snout, the ear the superior articular facet and the front paw the inferior articular facet.  A radiolucency creating a ‘collar’ implies a pars defect. Do not try to understand the positioning (there are different ways to obtain this view), and use the R/L marker ONLY to see which way the ‘dog’ is ‘looking’—it LOOKS to the side of interest, ie R vs L. 

The ‘eye’ should be radiodense circle with cancellous center; the articular facets (ear-meets-leg-above, leg-meets-ear-below) may show sclerosis, narrowing, osteophytes—ie osteoarthrosis or arthritis.

The pars is most likely to have a defect due to chronic stress/overuse/acquired attenuation but true acute trauma, neoplasm, infection, always possible.