Anatomy

The anatomic properties of the sacroiliac joints have been well studied, both at the gross and histologic scales. Grossly, the sacroiliac joint is an articulation between the ilium and the sacrum – usually S1–S4 but including the fifth lumbar vertebra up to 5% of the time.^2,3 The joint’s shape is commonly described as auricular, or L-shaped, with two arms of usually unequal length: the longer arm is dorsocaudally directed, and the shorter arm is dorsocranially oriented.⁴ The embryologic development of the sacroiliac joint begins in the tenth week and obtains a definitive form by the four month.⁵ Flat until at least the time of puberty, the articular surfaces eventually become roughened, with numerous and highly variable grooves and protuberances on each opposing face.⁶ In the adult, there is commonly a longitudinal sacral groove at S2 which admits an iliac ridge, although this arrangement may be reversed.⁴

Gross supporting structures exist anteriorly and posteriorly. Stabilization against anterior motion of the sacral promontory may be aided by the anterior sacroiliac ligament^7,8 and the sacrotuberous and sacrospinous ligaments.^9,10 Resistance against downward translation of the sacrum may be aided by a posterior structure, the posterior sacroiliac ligament.⁹ The interosseous ligament, another posterior structure, is thought to be a primary joint stabilizer.¹⁰ Anterior and posterior fibrous joint capsule fibers mesh with these structures.

In keeping with the somewhat controversial nature of the sacroiliac joint, the histologic characteristics of this structure render strict definitions difficult. The cartilaginous architecture of the sacral and iliac aspects of this joint differ, the sacral face being hyaline with an overall thickness of 1–3 mm, the iliac face composed of columns of fibrocartilage oriented perpendicular to the joint surface and interposed with islands of hyaline cartilage, with a thickness of less than 1 mm.⁴ The overall histologic architecture of the joint has been postulated to change throughout life, beginning as a diarthrodial joint and progressively losing mobility.⁶ Further variation is provided by gender: a cadaveric study of 47 specimens by Vleeming found that articular interdigitation was greater in males than in females.¹¹ Thus, the sacroiliac joints have been described as diarthrotic, synarthrotic, and amphiarthrotic.³

Physiology

Of physiologic concern is the innervation of the sacroiliac joint, although definitive understanding of this has been elusive. The anterior aspect of the joint is likely innervated by the posterior rami of the L2 through S2 roots, while the posterior portion is likely innervated by the posterior rami of L4 through S3. Innervation is thought to be highly variable, even between two joints in a given individual.^12–14 Anterior joint innervation may be further subserved by the obturator nerve, superior gluteal nerve, or the lumbosacral trunk.^15,16

Physiologic characterization of the sacroiliac joint, as with all joints, falls under two broad categories: kinetic and kinematic evaluation. The primary kinetic and kinematic considerations of the sacroiliac joints, by definition, involve sacroiliac force transmission and subsequent sacral motion relative to the ilia. In the standing position, superincumbent body weight is associated with an inferiorly directed translatory force acting upon the sacrum.¹⁷ An equally important induction of an anteriorly directed rotary force upon the sacral prominence relative to the ilia has been understood since the nineteenth century;¹⁸ the axis of this rotary force has been theorized to coincide with the horizontal line connecting the paired iliac tuberosities,² although other axes of rotation have been postulated, and more than one may exist.^19,20 In the standing position these two forces – one linear and the other rotatory and thus properly termed a torque – should by Newton’s first law induce a respective inferior translation and anterior rotation (or flexion, termed nutation) of the sacrum relative to the ilia, unless sacroiliac motion were completely restrained. Likewise, changing positions from standing is accompanied by changes in translatory forces and torques acting on the sacrum, and should therefore induce differing sacral translations and rotations relative to the ilia.

The physical parameter that couples the above forces acting on the sacroiliac joint with resultant sacroiliac motion is none other than joint mobility. As sacroiliac mobility has not been well quantified, the effect of sacroiliac kinetics on sacroiliac kinematics has not been well established. The presence of sacroiliac mobility in the pregnant female was suspected by Hippocrates and later by other researchers.^2,19 In pregnant women, sacroiliac joint laxity has been quantified using Doppler ultrasound; side-to-side asymmetries in such laxity have been postulated to be associated with pregnancy-related pelvic pain and predictive of postpartum pelvic pain.²¹ In the nonpregnant patient, however, reported joint range of motion in both the normal and pathologic states varies widely. An early analysis was performed by Mennell, who noted that with a change in position from prone to sitting, the posterior superior iliac spines were displaced apart from each other by 0.5 inch.²² Attempting to verify such sacroiliac motion in normal subjects, Colachis et al. used Kirschner wires to mark bilateral iliac landmarks, and interwire distances were noted in nine different subject positions. Maximal iliac motion was found with trunk flexion from the standing position. The interiliac relationship previously described by Mennell was reversed, however, with greater posterior superior iliac spine approximation found with sitting versus the prone position.¹⁹ Subsequent research, in order to maximize sensitivity of detecting sacroiliac motion, has evaluated the extremes of motion about the sacroiliac joint, such as the reciprocal straddle position and other extremes of hip motion. Smidt et al.²³ found interinnominate rotation to average 9° around an oblique sagittal axis and 5° around the transverse axis. Similar analyses by Barakatt et al.²⁴ recorded interinnominate motions of up to 36°. Increased detail has been allowed with the use of radiologic studies. In a follow-up cadaveric study utilizing computed tomography, Smidt²⁵ found sagittal sacroiliac motion of up to 17° with hips placed at end-ranges of motion. In contrast to such large values, Sturesson et al., in a roentgen stereophotogrammetric study of 25 individuals diagnosed with sacroiliac disorders, simulated typical motions about the joints and found mean sacral translation of 0.5 mm and usual sacral rotation of less than 3.6°.²⁶ Such minimal amounts of rotation – less than 1.6° – were confirmed by this author in a second study involving six women, five of whom experienced postpregnancy posterior pelvic pain.²⁷ This order of magnitude of rotation echoed similar radiostereometric results obtained by Egund et al.²⁰ The minimal amount of sacroiliac joint mobility described in these latter studies not only contrasts the results of previously noted investigations but also tends to cast doubt on the clinician’s ability to detect any related change in joint alignment by physical examination alone. Yet the clinical detectability of sacroiliac joint misalignment has been proposed.^28,29

Joint pain

Unresolved issues in the anatomic and physiologic characterization of the sacroiliac joint foreshadow the controversy surrounding clinical features associated with putative derangement of this structure. In both historical/epidemiological and clinically observable data, high degrees of variability confound the clinician’s ability to reliably diagnose, much less qualify, dysfunction of the sacroiliac joint. The very presence of pain emanating from the sacroiliac joint, the pathologic factors that cause such pain, the subjective location of such pain, as well as aggravating and alleviating factors, have all been studied to a limited degree, and with limited agreement as to outcome.

Pain emanating from the sacroiliac joint may stem from multiple primary and relatively well-defined causes: a list drawing from familiar pathologic categories, such as trauma, infection, tumor, and systemic illness. In the particular case of the sacroiliac joint, these broad categories have been reported specifically as traumatic pelvic ring fracture, intrapartum diastasis, pyarthrosis, metastatic adenocarcinoma, and spondyloarthropathy.^12,30,31 Even well-understood pathologies within the sacroiliac joint present with cryptic clinical findings: Bohay reported that in sacroiliac joint pyarthroses the most common symptom and sign was fever. Other common symptoms were also non-specific, including ipsilateral hip, leg, or buttock pain or low back pain. Physical examination focused on the joint is thought to be helpful in timely diagnosis of a suppurative process, and serial Patrick’s and Gaenslen’s tests and sacroiliac joint compression (described below) are recommended.³¹

A separate cause of sacroiliac joint pain, intrinsic to and emanating from the joint itself, has been termed sacroiliac joint dysfunction,^12,32 and is postulated to stem from an anatomic derangement of the joint. Attempts have been made to characterize this entity,^{28,29,33–35} although its existence has not been universally accepted in the allopathic medical literature.

Within this same body of literature, further clinical analyses have been undertaken, with tacit understanding of the uncertain nature of the underlying pathology of sacroiliac joint dysfunction. These analyses, which are few in number, range from the direct testing of the most basic clinical hypotheses to more advanced, multivariate evaluations.

Borrowing from the model set forth by Dwyer and Aprill,^36,37 Fortin et al. attempted to derive sacroiliac joint pain referral maps from asymptomatic volunteers. Ten subjects with no history of prior back pain underwent unilateral contrast injection into the sacroiliac joint. This was followed by analyses of the distribution of subsequently produced pain and hypoesthesia. Delineation of area of hypoesthesia was repeated after infusion of Xylocaine into the joint. Hypoesthesia after instillation of an average of 1.6 mL of contrast was noted to comprise one of three patterns: medial buttock inferior to the posterosuperior iliac spine, this area plus the superior greater trochanter, and this last area plus the superior lateral thigh. Areas of provoked pain were found to coincide with these areas of hypoesthesia. Given that the approach used was believed to minimize needle penetration of adjacent potential pain generators, including ligaments and muscle, it was inferred that these referral patterns were derived solely from the sacroiliac joint itself.³⁸

In a follow-up study, Fortin et al. tested the clinical significance of one of the previously derived pain referral zones. Of 54 consecutive patients referred for treatment of lumbar discogenic or facet pain, 16 were tentatively diagnosed with sacroiliac joint-mediated pain based on pain diagrams that indicated maximal discomfort within a region inferior to the posterosuperior iliac spine. All of these patients subsequently demonstrated concordant pain and sensory changes after provocative instillation of contrast into the sacroiliac joint, confirming the diagnosis. Ten of these patients underwent diagnostic evaluation of the lowest two intervertebral discs and lowest two facet joints ipsilateral to their initial pain, and in all cases these structures were deemed not to be pain generators.³⁹

However, subsequent investigation of 100 subjects with and without sacroiliac joint pain, by Schwarzer et al.,⁴⁰ called into question the clinical validity of sacroiliac pain reproduction by intra-articular contrast injection. Failure to reproduce pain was noted to have some negative predictive value, while pain reproduction was noted to have little positive predictive value. The gold standard here was pain reduction with subsequent anesthetic injection into the joint. Comparison of historical features of sacroiliac joint versus nonsacroiliac joint-mediated pain yielded only a single statistically significant distinguishing factor: the presence of groin pain with sacroiliac joint-mediated pain. No potentially exacerbating or mitigating factors, such as sitting, standing, walking, flexion, or extension, were found to correlate with the presence or absence of sacroiliac joint pain.

Following Schwarzer’s use of anesthetic instillation into the joint as the diagnostic gold standard for sacroiliac pain, Dreyfuss et al.¹² used this injection technique to compare historical features and physical examination findings in patients with and without sacroiliac joint pain. In a group of 85 patients who had been referred for sacroiliac injection, composite preinjection pain diagrams revealed a solitary factor that distinguished sacroiliac joint mediated pain from nonsacroiliac joint pain: the presence of pain above the L5 dermatome in patients without sacroiliac joint pain. Schwarzer’s analysis of potentially exacerbating or mitigating factors was modified slightly and extended greatly to include sitting, standing, walking, lying down, coughing/sneezing, defecation, use of heeled footwear, and job activities. Symptom relief with standing was found to be possibly specific for sacroiliac joint pain, although the statistical significance of this finding was not certain. None of the other factors was found to correlate with the presence or absence of sacroiliac joint pain. A similar lack of predictive value was found on analysis of previous response to a variety of therapeutic modalities, including certain classes of oral medication, physical therapy, manipulation, and certain modalities.

In a more focused analysis of pain referral zones, Slipman et al.⁴¹ studied 50 patients with sacroiliac pain whose diagnosis was tentatively made by means of physical examination maneuvers and confirmed by fluoroscopically guided sacroiliac joint block. Pretreatment interviews were utilized to localize the patients’ pain: buttock pain was found to be the most prevalent area of referred pain, occurring in 94% of patients, followed by lower lumbar (72%), thigh (48%), and lower leg pain (28%). Other painful areas included the foot and ankle as well as the groin, upper lumbar region, and abdomen. The greatest prevalence of buttock pain, followed by a descending prevalence of progressively distal lower extremity pain referral, is reminiscent of the previous results of Fortin et al.^38,39 Slipman et al. further noted a statistically significant inverse relationship between patient age and the presence of referred pain distal to the knee.

The literature thus far, although rather scant, has already yielded some clinical insight into the nature of sacroiliac joint-mediated pain. Firstly, Schwarzer was able to confirm the very presence of such pain by alleviating this symptom with injection of anesthetic into that joint in patients whose symptoms were not relieved by similar injections into neighboring structures.⁴⁰ The works of Schwarzer, Fortin, and Slipman served to establish and later validate the use of fluoroscopically guided anesthetic injection as the diagnostic gold standard for sacroiliac joint-mediated pain. Although by no means a gold standard, the pain referral zones presented first by Fortin et al. and later by Slipman et al. are at least foundations on which further clinical analyses may be conducted. Assessing possible reasons for the limited clinical usefulness of pain referral zones in the diagnosis of sacroiliac joint pain, Slipman has noted the complexity and variability of this joint’s innervation, possible sclerotomal referral patterns, and the proximity of other potential secondary pain generators activated by primary sacroiliac joint pathology.

The sacroiliac joint’s complex and incompletely understood innervation, described above, may lead to referred pain in the L2–S3 distribution. Sclerotomal referral indicates pain generated in a structure originating from a given embryonic somite being referred to another structure originating from that same somite. As the osseous structures of the spinal column originate from the ventromedial portion of their respective somites, and the muscles of the trunk and extremity originate from the corresponding posterolateral portions, sclerotomal, or somatic, referral may include a broad swath of the low back, buttock, and lower extremity. A further confounding factor is the proximity of other potential pain generators that may be directly irritated by a putative derangement in the anatomy or mechanics of the sacroiliac joint. An early study by Yeoman, involving sciatica, hints at such a pathologic mechanism,⁴² which has been further reported more recently.⁴³

In spite of these inherent obstacles, the current literature points undeniably towards continued refinements in the ability to diagnose sacroiliac joint-mediated symptomatology based on historical and epidemiological factors, even if such refinements include further insight into which factors are not clinically useful.