Sacral Stress Fracture in an Endurance Runner: A Clinical Case Study

A 40 year old male presented in clinic with right sided buttock pain. The patient described pain in the region of the right superior SIJ line and right L5/S1 facet joint, radiating inferolaterally towards the greater trochanter. There was no pain referred distal to the gluteal crease. There was no tingling, pins-and-needles or numbness. There was no reported central back pain, or left sided back pain. Red flag screening questions for bladder and/or bowel changes, saddle anaesthesia, significant night pain or unexplained weight loss were all negative. The pain was intermittent, and described as a combination of both dull and aching, with sharper catches of pain. The patient scored his pain at 7/10.

The patient reported his symptoms were aggravated by running. The patient was a keen endurance runner (see below) and, on his last run 5 days ago, had only managed half a mile before he had to stop due to his pain. Prolonged walking/weight bearing also brought on some symptoms, as did a combination of flexion and trunk rotation to the left. His symptoms were temporarily eased with co-codamol. The patient was otherwise very fit and well, and taking no other medications. The patient reported he was aware of his symptoms on rising in the morning, but that they settled quickly (within a few minutes) once he was up and about. His sleep had been disturbed on the corresponding night when his symptoms had been aggravated by running; however since he had stopped running, his sleep was no longer disturbed.

As aforementioned, the patient was a keen endurance runner. He had been running 100 – 120 miles/week for the last 18-24 months. He reported no history of previous injury. He did no other forms of training (e.g. no weights/strength training). He considered himself to be a mid-stance toe-runner. The patient had recently signed up for the Dragon’s Back race in September 2021, an ultra running event through the Welsh mountains, running a distance of 380km over 6 days. In preparation for this, over the last few months, the patient had maintained his mileage but increased his pace from 8 minute/miles to 7:20 – 7:30 minute/miles and, on his shorter distance runs of 10 miles, to 7 minute/miles. On discussion of form/technique, he reported he suspected he had lengthened his stride pattern a little, as well as obviously increasing his cadence. Prior to his ultra running over the last 18-24 months, he had been a keen cyclist and participated in endurance cycling events.

Approximately 3 weeks before presentation in clinic, the patient had gone out for a 20 mile run. By the time he had completed this, he had developed his right sided buttock pain. There was no particular injury; just the gradual onset of increasing symptoms during the run. The following day he was aware of his pain when walking, but it was less intense. The day after that the patient went out for another run – 15 miles – and reported having to stop during the run due to the severity of pain. He then struggled with walking, particularly the right side stance phase, for a few days due to the severity of pain. The patient denied any alteration in posture – he reported he could stand fully erect and hadn’t noticed any postural “shift”. The patient then rested for one week, before working out on his turbo bike. This aggravated his symptoms, but not to the levels that he felt compelled to stop. 3 days later (5 days prior to his consultation) he tried to go out for a run again, and had to stop after ½ a mile, due to pain.

On objective examination the patient stood in good alignment, with no indication of scoliosis or shift. There was subtle anterior pelvic tilt. Single leg stand was well preserved bilaterally, with good control. Lumbar flexion did not aggravate the pain, and was limited by tight hamstrings (fingertips to mid shin). Lumbar extension was pain free. There was mild multi-segmental stiffness, accompanied by a tendency to flex the knees in order to accommodate this. Right and left lumbar side flexions were also pain free; although left side flexion was more restricted than right side flexion. Cluster tests of the SIJ were all negative/unremarkable. Full active range of movement was noted in all directions at the hip joints bilaterally, with the exception of hip extension, which was limited by psoas restriction/tightness. All hip movements were pain-free. Piriformis length was full and pain-free bilaterally. Neurological examination of the lower limbs was unremarkable for dermatomal, myotomal and reflex testing. Isometric muscle testing of the lower limb musculature revealed subtle weakness comparatively in the right gluteus medius, quadriceps and hamstrings. On palpation, there was no tenderness over the right SIJ line. There was mild tenderness over the right L5/S1 facet joint. A provision diagnosis was made of right sided L5/S1 facet joint referred pain, secondary to end range restriction into lumbar extension, tight psoas and subtle gluteal, quads and hamstring strength deficits on the right, resulting in relatively increased loading as the patient fatigued. A targeted strengthening programme was prescribed, together with some lumbar extension and psoas stretches, and advice to rest until symptoms were fully resolved when walking and then try a short run at a slower pace.

Two weeks later the patient returned with persistent symptoms. He was still unable to run as, almost as soon as he started, his pain returned and reached an intensity whereby he had to stop running. The patient reported his pain was very much linked to the stance phase of (running) gait on the right lower limb. He had tried running at a much slower pace (9 minute/miles) but this was too sore. The patient was re-examined. Findings were as previously noted. In addition though, the hop test produced immediate symptoms on the right. Whilst the hamstrings were tight, there was no tenderness over the ischial tuberosity or pain on isolated hamstring loading. FABER and FADIR tests were negative. Palpation of piriformis was unremarkable. However there was exquisite focal tenderness on palpation of the right sacral alar, just medial to the right SIJ line, in the mid zone of the sacrum. Following this assessment, the presentation appeared more consistent with a stress fracture of the right sacrum, secondary to the increased cyclical loading through running gait, caused by the increase in cadence over long distances. Possible differential diagnoses included L5/S1 facet joint referred, discogenic, right SIJ, piriformis/DGS (deep gluteal syndrome): however these all did not consistently fit with the presentation, nor did they explain the sacral pain on palpation. An MRI scan of the sacrum was therefore requested. This was returned with the following report: -

“All lumbar discs return normal hydration signal. No disc bulge or thecal sac compression is identified. The intervertebral foramina are well maintained at all levels.

There is pronounced abnormal bone marrow signal change involving the right sacral ala in its superior and mid-portions with high signal on STIR imaging and intermediate signal on T1-weighted images. There is the suggestion of a serpiginous low signal line running vertically through part of the sacral wing on the right suspicious for a fracture and appearances must be considered highly suggestive of a right-sided sacral stress fracture

There is no abnormal signal change within the left sacral ala.”

Discussion

Sacral stress fractures are uncommon skeletal injuries that are separated into fatigue and insufficiency-type injuries (Bednar and Almansoori, 2015). Their true prevalence is unknown due to their frequent misdiagnosis, lack of specific symptoms and successful resolution with time and observation (Bednar and Alminsoori, 2015). They are difficult to diagnose given the symptoms are representative of several injuries including lumbar referred pain, disc disease, nerve root involvement, SIJ pathology and deep gluteal syndrome/piriformis syndrome (Kahanov et al, 2015). Insufficiency fractures are much more common than fatigue fractures, but are almost exclusively reported in women over 55 years old (Bednar and Alminsoori, 2015). Fatigue or stress fractures in the lower extremity account for 80-90% of all stress fractures, representing between 0.7% and 20% of all sport medicine injuries. Specifically stress fracture incidence in runners approaches 16% of all injuries (Kahanov et al, 2015), although sacral stress fractures occur much less frequently in comparison to tibial, navicular, metatarsal, and femoral stress fractures (Kahanov et al, 2015).

A stress fracture represents the inability of the bone to withstand repetitive mechanical loading, which results in structural fatigue and local bone tenderness (Walden et al 2014). Bone stress injuries occur along a pathological continuum beginning with stress reactions, progressing to stress fractures and then complete bone fractures. The pathophysiology underlying bone stress fractures remains speculative; however there is growing acceptance that it involves an imbalance between load-induced microscopic damage formation and its removal (Walden et al, 2014)

The skeleton is exposed to mechanical loading during running, causing the bones to deform. The amount of deformation depends on the load magnitude and the ability of the bone to resist deformation. Whilst the amount of deformation (aka strain) may not be sufficient to cause a break in cortical bone, levels below that required for fracture are capable of causing microscopic damage (microdamage) (Walden et al, 2014).

Microdamage formation is threshold dependent, with the threshold for its formation depending on the interaction between the number of bone strain cycles, strain magnitude, and the speed at which strain is introduced. Once the threshold for microdamage has been surpassed, further increases in bone strain cycles, magnitude and/or rate result in additional damage. However, this is useful as it helps to dissipate energy and serves as a stimulus for targeted remodelling. Targeted remodelling involves activation of remodelling units, consisting initially of bone-resorbing osteoclasts, followed by bone-forming osteoblasts. The osteoclasts remove damaged tissue, whereas the osteoblasts deposit layers of new bone to create a new bone structural unit. Targeted remodelling maintains homeostasis between microdamage formation and its removal to preserve skeletal mechanical competence, as well as reduces tissue age and enables bone to adapt over time to meet changing demands. The adaptation effectively decreases bone strain for a given load, so that greater loads can be tolerated before surpassing the threshold for microdamage formation.

Remodelling usually occurs as fast as microdamage formation. In addition, there are additional remodelling units “in reserve”, which can be activated in response to an increase in microdamage formation. Therefore changes that result in increased loading, via magnitude and/or rate, can usually be tolerated. However, remodelling is time dependent, and the time required to reach a new equilibrium is approximately 3-4 months in cortical bone. If insufficient time is given to adapt to the new mechanical loading, damage may progress due to the fact that increased osteoclastic activity precedes osteoblastic activity, causing a temporary local reduction in bone mass. Therefore the bone is less able to absorb energy, which leads to further microdamage, thereby starting the bone stress reaction continuum (Walden et al, 2014)

With regards to this specific patient, there was no indication of underlying bone pathology, and the presentation appeared wholly consistent with a fatigue fracture, secondary to the increase in cadence, and therefore cyclical loading, over ultra-running distances, without sufficient time allowed to achieve a new equilibrium for bone stress/load toleration. The patient was not taking any medications that would affect bone health. However, as suggested by Vajapey et al (2019), over-the-counter Vitamin D supplements and calcium supplements were discussed.

Sacral stress fractures are uncommon injuries; therefore, perhaps unsurprisingly, the literature does not provide sufficiently detailed guidance on appropriate rehabilitation programmes. Kahanov et al (2015) perhaps provide the most detail, but there is an absence of targeted specificity (e.g. the recommendation is made to do resistance training with “heavier loads” but this is not quantified e.g. as a percentage of 1 repetitions maximum (RM)). The patient was advised to rest from running completely for 6 weeks (Kahanov et al, 2015), and maintain his aerobic fitness using the turbo bike (avoiding out-of-saddle position) as long as this remained relatively symptom free. The patient was advised to walk as tolerated. A strength based programme was commenced. Whilst ultimately endurance would be the focus for this particular patient, baseline strength needed improving first. During lockdown there was also some restriction to facilities available, so the programme was designed with those considerations too. A non-weight bearing loading programme was commenced using isometric exercises for the quadriceps, hamstrings, gluteus medius and gluteus maximus, working to a high intensity (80 – 85% or maximum) over a small number of repetitions (5 reps) (Champaign et al, 2014) whilst maintaining pain-free status. The patient was also advised to continue with lumbar extension mobilisation exercises and psoas and hamstring stretches. After 4 weeks post-cessation of running, and in the absence of pain, the patient was progressed to weight bearing loaded exercises such as squats, single leg squats, lunges, mountain climbers, elevated bridging, abdominal “bicycles” and balance and proprioceptive work. Initially the patient began with a lower load to monitor any response before the load intensity was increased to work at the intensity as aforementioned. The hop test was used as an objective marker by the patient and it was agreed that, once the hop test was pain-free and there was no tenderness over the right sacrum on palpation, running could be gradually introduced. This commenced with a 3 mile run on the flat (Kahanov et al, 2015). Recommendations were made with regards to footwear, with the advice to change footwear every 300 – 500 miles to preserve shock absorption (Kahanov et al, 2015) and to consider limiting an increase in intensity to no more than 10%/week. The patient continued with 3-4 strength based sessions/week, and gradually increased his running mileage in line with the aforementioned. The patient has made a successful return to long distance running.

Reflections

This was a really interesting patient to assess and treat. After the initial assessment, even though I diagnosed the right L5/S1 facet joint as the source of the problem, I wasn’t wholly convinced that that was the correct diagnosis; I sensed there was something a little more to it. The fact the pain developed so quickly on cyclical loading during stance phase when running, didn’t quite fit with my provisional diagnosis. If this was a focal facet joint sprain/OA, I wouldn’t have expected the onset to be quite so immediate; indeed, it would have been more likely for the symptoms to gradually improve with activity/exercise rather than intensify to such an extent that the patient was forced to stop. And this is a patient who was able to significantly increase his running pace, over long distances; to me this reflected an ability to tolerate pain/discomfort when running. If the patient reported the intensity of pain stopped him running, then this indicated to me it was a significant level of pain. Again, this didn’t quite fit with my initial diagnosis. From the subjective examination, and in my experience as a recreational runner, the sudden increase in pace was the most significant finding. From a ‘leg turnover’ perspective, it’s a substantial increase in cyclical loading to go from 8 min/mile to 7:20 min/mile. So after the initial assessment, I felt I really needed to go through everything in my head and pick apart the history and the clinical presentation. I did some on-line searching. As soon as I determined that you could indeed sustain a stress fracture of the sacrum, I felt this became the most likely diagnosis. However the literature indicates it is difficult to diagnose, and is an uncommon stress fracture. The literature indicated that focal palpation and the ‘hop‘ test were perhaps the best objective indicators, although there isn’t enough data to accurately quantify specificity and sensitivity. In the context of the history though, the combination of focal pain over the right sacral ala, the immediate onset of pain when running, a positive hop test and the history of the increase in cadence over ultra distances, meant that I felt the most likely clinical diagnosis was a stress fracture of the right sacral ala. Once I had determined this in my mind, the challenge of discussing this with the patient then presented itself. A diagnosis of a stress fracture would mean a cessation of running completely. Given this patient is a keen ultra-runner, with an event that he is working specifically towards, this would not necessarily be an easy conversation. However, I was fortunate in this regard. When I explained what I felt the diagnosis was, and highlighted the evidence from the subjective and objective assessment, the patient agreed. He too had done some on-line research and was aware it was a possibility. We discussed the pros and cons of imaging. I explained it would be unlikely to show up on x-ray and the most appropriate investigation would be an MRI scan; however it wouldn’t change the management plan if confirmed on scan. However the patient felt it was important for him personally to have it confirmed and he decided that he would pay for a private scan. He needed authorisation from his own GP; his own GP reported that it was so rare, that he was confident it wasn’t a sacral stress fracture, would not refer him for an MRI scan on the NHS but would consent to a private one, simply because the patient was funding it. The patient’s wife, also a GP, also disagreed with the diagnosis, again citing how rare it was. The privately funded MRI scan confirmed the diagnosis. On reflection, it was important both for the patient and for me that imaging confirmed the stress fracture. For the patient, it highlighted the importance of moderating his training and incorporating a strength component. For me, it validated my clinical reasoning. Even though, before I assessed this patient, I wasn’t even aware that a stress fracture could develop in the sacrum, I’m pleased I got to the diagnosis as quickly as I did. It also highlighted, again, that what a patient puts in front of you, may well be something that you’ve not seen before. Over the years, the number of patients I have assessed with unilateral back pain must run into the thousands; however one can then present with something I wasn’t even sure existed. It underlines the risks of complacency, the importance of really listening to the patient, the benefits of ‘patient mileage’ (it didn’t quite fit with many of the other unilateral back pain patients over the years) and the awareness that there are times when you just need to take time, outside of the assessment, to really reflect and run through the clinical reasoning. I would feel more confident diagnosing one ‘next time’; however the rareness of the condition means I may not see another one again.

References

Bednar D, Almansoori K; (2015); Sacral Stress Fracture Mimicking Lumbar Radiculopathy in a Mounted Police Officer: Case Report and Literature Review; Global Spine Journal; 5; e69 – 73

Kahanov L; Eberman L; Games K; Wasik M; (2015); Diagnosis, treatment, and rehabilitation of stress fractures in the lower extremity in runners; Open Access Journal of Sports Medicine; 6; 87 – 95

Vajapey S, Matic G, Hartz C, Miller T; (2019); Sacral Stress Fractures: A Rare but Curable Cause of Back Pain in Athletes; Sports Health; 11; 5; 446 - 452

Walden S, Davis I, Fredericson M; (2014); Management and Prevention of Bone Stress Injuries in Long-Distance Runners; Journal or Orthopaedic and Sports Physical Therapy; 44; 10; 749 - 765