Prospective Hounsfield Unit measurements of intercorporal bone grafts remodelling towards spinal fusion.

Low (lumbar) back pain and concomitant leg pain are common and a major cause of
global disability. Lumbar spinal fusion (lumbar spondylodesis) is a surgical
procedure is used for many diseases that cause lumbar back pain, and the use of
this treatment has increased in recent decades. Lumbar spondylodesis has been
shown to reduce pain and improve functional status. With spondylodesis, the
intervertebral disc is removed and bone graft is placed between the two
vertebrae. Over time, this becomes vital bone, creating a lasting fusion of the
two vertebrae. The vertebrae are also fixed with pedicle screws and metal rods
to provide immediate stability after surgery.

The success of the fusion is partly determined by the amount, and quality, of
bone formed between the two lumbar bodies. Radiological scoring is used to
monitor fusion. One method of quantifying bone formation in instrumented areas
is by using Hounsfield Units (HU) measured by Computed Tomography (CT) imaging,
where an increase in HU correlates with higher bone density. HU are measured in
a specific area (Region of Interest). Clinically, HU measurements are already
used to evaluate bone quality, for example prior to spinal surgery.

In the literature, only little is known about quantification of the biological
process of bone graft remodelling and the occurrence of fusion between
vertebral bodies. Only one prior study has used HU to evaluate the changing
bone density of bone graft after lumbar fusion, but they did not use the same
subjects for HU measurements at different time points in the follow-up.

By quantifying the bone density of individual patients' bone grafts at
different times during follow-up, clinicians can evaluate the quality and
progression of bone formation. In addition, it can help elucidate the
biological process of bone graft remodelling. This will help us understand how
long it may take for the bone graft to finish remodelling. In addition, and
perhaps more importantly - it may help to predict the failure of fusion with
early recognition of symptomatic pseudo-arthrosis. In doing so, we will compare
the trend in HU with the most commonly used fusion scoring criteria and
administer clinically related questionnaires.

Prior to this study, we have already conducted a retrospective pilot study
using patients with consecutive 1- and 2-year postoperative CT scans. In this
study, we developed a protocol for measuring bone graft on two follow-up CT
scans and evaluated whether this methodology could be reliably performed by a
single observer. The methodology was validated in this study. The results
showed that in 78% of the patients the bone graft was still remodelling to a
higher bone density, meaning that better quality bone is still being formed.

This study has three specific objectives:
1) to explore how the participants* intercorporal bone grafts HU develop over
time in the first two years after fusion surgery.
2) to explore the interobserver reliability of the HU measurements.
3) to explore if the participants* upward, downward or no trend in HU values
correlate with trends of scores on a commonly used interbody fusion
classification and the degree of back and/or leg pain.

This is a single-center prospective exploratory study. During this study a
convenience sample of participants, who have undergone lumbar spinal fusion in
Rijnstate Hospital and who have agreed to participate in the study, will have
(four) CT-scans of the lumbar spine taken in the first week after surgery, and
again after 6, 12 and 24 months.

Participants will be asked to undergo four low-dose CT-scans of the
instrumented segment over a two-year period. During standard care, fusion
patients undergo three conventional lumbar radiographs (after 6 weeks, 3
months, and 1 year) and one CT-scan (after 1 year). In our hospital, the
effective median exposure doses of standard lumbar radiographs and CT-scans are
0,2 mSv and 2,4 mSv, respectively. This exposes each patient to a total
radiation dose of (3 x 0,2 + 1 x 2,4) = 3,0 mSv in one year. Without taking any
additional measures, four *study* CT-scans would expose the participants to a
total radiation dose of (4 x 2,4 =) 9,6 mSv in two years. To mitigate this
increased radiation risk, and to lower the participants* radiation exposure
during the study, we will:
1) focus the CT-scan of the lumbar region on the instrumented segment only (2
instead of the regular 6-7 vertebral bodies). This measure will reduce the
radiation dose by up to 60% (approximately 1.4 mSv) per CT-scan, and hence
reduce the overall CT-based radiation dose by (4 x 1,4 =) 5,6 mSv.
2) replace the conventional lumbar radiographs by the study CT-scans. This
measure will prevent participants from also receiving a standard radiation dose
of 3 x 0,2 = 0,6 mSv mSv from the radiographs.

By taking the above two measures, the patient*s effective dose of radiation
exposure during the study will be around 7,2 mSV compared to the 3,0 mSv during
regular care. As such, compared to regular care, participants will be exposed
to higher dose of radiation by participating in this study. It is of note that
we will only recruit patients over the age of 45. In this age group the
lifetime risk of radiation induced carcinogenesis is known to decline rapidly.

Considering the above, and applying the risk classification as proposed by the
Dutch Commission of Radiation Dosimetry we estimate that the additional
radiation risk is moderate (*matig risico*). We disscused the above with the
clincal physicist and we believe that the benefits, of gaining insight in how
patients* intercorporal bone grafts HU develop over time in the first two years
after fusion surgery, and if the patients* upward, downward or no trend in HU
values correlate with trends of scores on a commonly used interbody fusion
classification and the degree of back and leg pain, outweighs this risk.