Keywords
spinal surgery - surgical results - transthyretin
Introduction
Spinal surgery is extremely important part of neurosurgery practice. Along with technological
progress, surgical success in spinal pathologies has been increasing with increased
diagnostic possibilities, the development of surgical techniques, and particularly
the development of microsurgery. However, despite all these developments, poor results
and advanced morbidity can occur in a considerable number of patients after spinal
surgery. Transthyretin (TTR) is a homotetrameric protein weighing 54.98 kDa with four
identical subunits. This protein was originally named “prealbumin” because it shows
more anodal migration in electrophoresis compared with albumin.[1] After nerve damage, in cases where TTR is low, recovery of locomotor activity is
delayed and nerve conduction velocity slows down.[2] Moreover, TTR has an accelerating effect on nerve regeneration.[3] The mechanisms underlying this neurogenic effect have not been clearly understood.
In this study, we aimed to investigate the effects of preoperative serum TTR levels
on surgical success, pain scores, and postoperative morbidity.
Methods
Note that, in this study, 188 patients who underwent spinal surgery in our clinic
between 2010 and 2011 were included. Written consent was obtained from the patient
(or legal guardian) that his medical data could be published; patients who refused
to sign the informed consent form were excluded from the study. A certificate of conformity
was obtained for the study from the ethics evaluation commission of Ankara Diskapi
Yildirim Beyazit Training and Research Hospital (26.08.2010, Decision No: 07). All
spinal surgery cases were included in the study group:
-
Patients who underwent cervical discectomy + cage.
-
Spinal tumor cases.
-
Patients who underwent lumbar discectomy.
-
Patients who underwent posterior lumbar stabilization.
-
Patients who were operated for lumbar stenosis.
Patients using steroids, those with chronic inflammatory disease, those with chronic
renal failure, those with high liver function tests, those patients who had myocardial
infarction in the last month, and patients with symptoms of preoperative acute infection
were excluded from this study. The demographic data (age, sex, height, weight, weight
6 months ago, and body mass index [BMI]) of the patients included in the study and
length of hospital stay were recorded. The activity levels of patients were evaluated
between 1 and 4 points: sedentary (1), mild activity (2), moderate activity (3), and
heavy activity (4). Comorbidities accompanying the disease, including the presence
of pulmonary diseases, diabetes mellitus, hypertension, hypercholesterolemia/hyperlipidemia,
gastroesophageal diseases, smoking, coronary artery disease, and hyperthyroidism were
recorded. Venous blood samples were obtained from the patients on the morning of the
surgery, just before the operation and their complete blood count, protein, albumin,
TTR, C-reactive protein (CRP), and ferritin levels were measured. Preoperative and
postoperative neurological examinations of patients were performed, and muscle strength
examinations in relation with the patients' primary spinal pathology were recorded:
0, no contraction; 1, slight muscle contraction, no movement; 2, motion without gravity;
3, motion with gravity; 4, some resistance against the examiner, but not fully; and
5, full muscle strength. Preoperative and postoperative Karnofsky scores of patients
were evaluated. Preoperative and postoperative Oswestry Disability Index (ODI) and
visual analog scale (VAS) and pain levels of patients were evaluated in the preoperative,
early postoperative (second day), and late postoperative (first month) periods. It
was recorded whether the patients had wound site infection in the postoperative periods.
A postoperative wound site infection diagnosis was made in patients with purulent
drainage from the wound site and the causative microorganism could be grown in culture
in the postoperative period. The wound sites of the patients were evaluated on the
7th postoperative day. During this period, wound site healing problem was evaluated
as positive in patients whose wound site did not close and were macerated. Preoperative
complications (dural tears, nerve damage, and massive hemorrhage) were recorded. Postoperative
morbidity was defined as permanent damage after surgery and/or the requirement for
serious additional investigation and treatment. Deep wound infection, prosthetic malposition,
myocardial infarction, pulmonary thromboembolism, newly developed neurological deficit,
incorrect surgical distance, pseudomeningocele, and vascular injuries were recorded
as morbidity. In this study, patients' preoperative variables (particularly blood
TTR level, CRP/TTR, albumin/TTR, ferritin/TTR ratios) and postoperative results were
compared.
Statistical Analysis
Data analysis was performed using SPSS for Windows 11.5 package program (SPSS Inc.,
Chicago, Illinois, United States). Whether the distribution of continuous variables
was close to normal was investigated using the Shapiro–Wilk test. The homogeneity
of variances was examined using the Levene's test. Descriptive statistics were observed
as mean ± standard deviation or median (smallest–largest). The significance of difference
between the groups in terms of means was determined using Student's t-test. When the number of independent groups was two, the significance of the difference
between the groups in terms of median values was determined using the Mann–Whitney
U test. When the number of independent groups was more than two, the significance of
the difference between groups was investigated using the Kruskal–Wallis test. If the
Kruskal–Wallis test statistics were reported to be significant, the conditions causing
the difference were determined using the Conover's test of multiple comparisons. Whether
there was a statistically significant change between preop and postop measurements
was examined using the Wilcoxon signed-rank test. Categorical variables were evaluated
using the Pearson's chi-square test. The Spearman's correlation test was used to investigate
whether there was a significant relationship between continuous variables. For p < 0.05, the results were considered to be statistically significant.
Results
Note that 188 patients were included in this study. The age range of the patients
was 18 to 85, and the mean age was 49.3 ± 12.5 years. Of the patients participating
in the study, 113 were female (60.1%) and 75 were male (39.9%). The patients were
grouped as per the spinal surgery they underwent. Group I included 150 patients (79.8%)
who were operated for lumbar discectomy, posterior lumbar stabilization, and lumbar
stenosis. Group II included 26 patients (13.8%) who underwent anterior cervical discectomy
and cage application. Group III included 12 patients (6.4%) who were operated for
spinal tumors. Of the patients, 41.5% (n = 78) had at least one comorbid disease. The mean preoperative motor strength of
the patients was 4 (0–5). In the postoperative period, the mean motor strength was
evaluated as 5 (0–5). This improvement in postoperative examination was considered
to be statistically significant (p < 0.001). The mean preoperative Karnofsky score of the patients was 80 (50–100),
while the postoperative Karnofsky scores were reported to be 100 (40–100). This postoperative
improvement in the Karnofsky scores was statistically significant (p < 0.001). The preoperative ODI scores of the patients were 43 (0–100), while the
postoperative ODI scores were reported as 10 (0–88). This improvement in the ODI scores
was statistically significant (p < 0.001). The mean preoperative VAS values of the patients were reported to be 8
(0–10). The mean VAS values in the early postoperative period were 4 (0–10) and the
mean VAS values in the late postoperative period were found as 1 (0–10). Compared
with the preoperative VAS values, the decrease in both early postoperative and late
postoperative VAS values was statistically significant (p < 0.001). Postoperative wound site infection was observed in 14 patients (7.4%),
whereas postoperative morbidity was observed in 25 patients (13.3%). Peroperative
complications occurred in 13 patients (6.9%). When the correlation analysis was examined,
the following results were reported to be statistically significant:
As the age of patients increases, hospital stay is prolonged (p = 0.009, correlation coefficient [kk] = 0.189).
As the BMI of the patients increases, Karnofsky scores decrease (p = 0.022, kk = −0.167), ODI scores increase (p < 0.001, kk = 0.261), early postoperative VAS (p = 0.08, kk = 0.193) and late postoperative VAS (p = 0.04, kk = 0.193) levels increase, and also length of hospital stay is prolonged
(p = 0.011, kk = 0.186).
As the patient's activity level increases, early postoperative VAS values decrease
(p = 0.017, kk = −0.174), and length of hospital stay is shortened (p = 0.005, kk = −0.206).
As the patient's preoperative hemoglobin level increases, ODI scores decrease (p = 0.002, kk = −0.222), early postoperative VAS (p < 0.001, kk = −0.247) and late postoperative VAS values (p = 0.022, kk = −0.166) decrease, and also length of hospital stay is shortened (p < 0.001, kk = −0.248).
As the patient's preoperative total protein level decreases, Karnofsky scores decrease
(p = 0.004, kk = 0.207) and ODI scores increase (p = 0.024, kk = –0.165).
As the patient's preoperative albumin level decreases, Karnofsky scores decrease (p < 0.001, kk = 0.295), ODI scores increase (p < 0.001, kk = −0.275), early postoperative VAS (p = 0.021, kk = −0.169) and late postoperative VAS values (p = 0.01, kk = −0.189) increase, and also length of hospital stay is prolonged (p = 0.017, kk = −0.174).
As the patient's preoperative TTR level decreases, Karnofsky scores decrease (p < 0.001, kk = 0.309), ODI scores increase (p < 0.001, kk = −0.344), early postoperative VAS (p < 0.001, kk = −0.259) and late postoperative VAS values (p < 0.001, kk = −0.323) increase, and also length of hospital stay is prolonged (p < 0.001, kk = −0.355).
As the patient's preoperative CRP level increases, Karnofsky scores decrease (p = 0.016, kk = −0.175), early postoperative VAS (p = 0.046, kk = 0.146) and late postoperative VAS values (p = 0.034, kk = 0.154) increase, and also length of hospital stay is prolonged (p < 0.001, kk = 0.236).
As the patient's preoperative ferritin level decreases, Karnofsky scores decrease
(p = 0.017, kk = 0.174) and ODI scores increase (p = 0.012, kk = −0.183).
As the patient's preoperative CRP/TTR ratio increases, Karnofsky scores decrease (p < 0.001, kk = −0.344), ODI scores increase (p < 0.001, kk = 0.291), early postoperative VAS (p < 0.001, kk = 0.282) and late postoperative VAS values (p < 0.001, kk = 0.341) increase, and also length of hospital stay is prolonged (p < 0.001, kk = 0.390).
As the patient's preoperative albumin/TTR ratio increases, Karnofsky scores decrease
(p < 0.001, kk = −0.248), ODI scores increase (p < 0.001, kk = 0.297), early postoperative VAS (p < 0.001, kk = 0.230) and late postoperative VAS values (p < 0.001, kk = 0.288) increase, and also length of hospital stay is prolonged (p < 0.001, kk = 0.333).
When the data were compared between genders in term of scores, postoperative early
VAS scores decreased less in women compared with preoperative VAS scores (p = 0.027). No significant correlation could be established between other variables
and genders (p > 0.05).
When a comparison was made between diagnostic groups, group II had the highest improvement
in motor strength scores while group III had the least improvement. The change in
motor strength scores between all groups showed a statistically significant difference.
When the change in Karnofsky scores was compared, the difference between group I and
group II was not statistically significant; however, the differences between group
I and III and group II and III were statistically significant. Similarly, when the
change in ODI scores was compared, the difference between group I and group II was
not statistically significant; however, the differences between group I and III and
group II and III were statistically significant. When the change in early and late
postoperative VAS scores was compared, the difference between group I and group II
was not statistically significant; however, the differences between group I and III
and group II and III were statistically significant. When the length of hospital stay
was compared, the patients in group I and group II showed similar characteristics;
however, patients in group III had longer hospital stays.
Low levels of TTR are predictive for postoperative wound site infection. TTR levels
were 209.4 ± 81.9 g/dL in the group with wound site infection and 286.6 ± 70.8 g/dL
in the group without infection. When statistical analysis was performed, the rate
of wound site infection increased in the patient group with low TTR levels (p < 0.001). Similarly, increased CRP/TTR ratio and albumin/TTR ratio were reported
to be correlated with a high risk of wound site infection (p = 0.007 and p = 0.005, respectively).
TTR levels of patients with delayed wound site healing (182.4 ± 68.5) were reported
to be lower than the group without delay in wound site healing (292.6 ± 65.9) (p < 0.001). Similarly, increased CRP/TTR ratio and albumin/TTR ratio were reported
to be correlated with delay in wound site healing (p < 0.001). Furthermore, lower hemoglobulin (p = 0.016), albumin (p = 0.015), and ferritin (p = 0.005) levels were reported in the group with delayed wound site healing compared
with the group without delay. TTR levels (202.9 ± 86.4) were lower in the group with
postoperative morbidity compared with the group without morbidity (292.8 ± 64.7) (p < 0.001). Moreover, high CRP/TTR and albumin/TTR ratios were found to be statistically
correlated with increased morbidity (p < 0.001). In addition, lower hemoglobulin (p = 0.002), protein (p = 0.034), albumin (p = 0.041), and ferritin levels (p = 0.003) were determined in the group with morbidity compared with the group without
morbidity; CRP levels were observed to be higher (p = 0.018).
To summarize the results in terms of TTR, in patients with low TTR: postoperative
Karnofsky scores were lower, postoperative ODI levels were higher, postoperative early
and late VAS scores were higher, length of hospital stay was prolonged, peroperative
complication rates were higher, wound site infection rates were higher, delay in wound
site healing increased, and morbidity rates were higher.
Discussion
Spinal interventions form an important part of daily neurosurgery practice. Although
we use developed microneurosurgical techniques, advanced radiological examinations,
and modern stabilization and fusion materials, spinal surgery still can lead to unsuccessful
results, causing more serious morbidity and mortality.[4] It is extremely important for the surgeon and the patient to predict which patient
is at risk before surgery, which patient will develop postoperative complications,
which patient will benefit from surgery, and which patient will be exposed to morbidity.
However, because there are multiple factors that affect this evaluation, it is not
possible to predict postoperative results based on a single factor. Possible complications
after spinal surgery have been the subject of many retrospective studies.[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13] In these studies, many variables had an effect on surgical results and contributed
to morbidity and mortality; however, the predictive role of a single factor in the
occurrence of these adverse events has not been studied. Moreover, there is no clear
definition of the terms “complication,” “adverse event,” and “morbidity” in the literature.
Because of this ambiguity, a 7% complication rate was reported in one series[13]; however, this rate was 42% in another series.[4]
In this prospective study, the effects of preoperative TTR levels on postoperative
surgical outcomes (Karnofsky score, ODI, VAS) of patients who underwent spinal surgery,
length of hospital stay, perioperative complication rates, postoperative morbidity,
wound site infection development, and wound site healing were examined. The relationship
between malnutrition and diseases has been well known. Changes in nutritional status
cause suppression in organ functions and immune system.[14] In a study, up to 50% of hospitalized patients were susceptible to protein energy
malnutrition.[15]
Wound site healing was delayed in patients with malnutrition, and the rates of infection
and morbidity increased.[14] Moreover, surgical morbidity and mortality increased in patients with impaired nutritional
status.[16]
[17] Because of the importance of malnutrition, it is recommended to perform a nutritional
evaluation in all hospitalized patients, especially in patients undergoing major surgery.[18] However, there is no consensus in the literature on which parameters the nutritional
assessment should be based on previous studies.[19]
[20]
[21] To perform nutritional evaluation, clinicians require a fast, easily available,
inexpensive, and effective screening test. For this purpose, visceral proteins such
as albumin, transferrin, and TTR have been used. In previous studies, TTR was defined
as a good indicator of nutritional status in patients with malnutrition.[22]
[23]
[24]
[25] Moreover, a significant correlation has been identified between protein energy nutrition
and TTR levels in large clinical studies.[19]
[20]
[21] However, in cases where acute phase reactants[26] and inflammatory cytokines[27] are high, low TTR levels have been shown to be an independent indicator of acute
phase reactants in evaluating malnutrition. Bernstain stated that TTR is the best
parameter for evaluating malnutrition.[28] In conclusion, TTR is accepted as an inexpensive, applicable, and reliable tool
for evaluating malnutrition in patients.[24]
[29]
Abnormal TTR levels correlate with the increasing number of complications in patients
undergoing elective surgery.[30] A close relationship was reported between low TTR levels and the risk of infection,
infection-related mortality rates, and infection-related complications.[31] Moreover, low TTR levels were correlated with increased complications and delayed
wound site healing in patients operated for ovarian cancer.[32] Jewell et al reported that TTR levels are the best indicator of perfect wound healing.[33] Beck and Rosenthal reported that low TTR levels correlated with prolonged hospital
stay, delayed wound healing, and prolonged sepsis duration.[14] Furthermore, higher mortality levels were reported in patients with low TTR levels
compared with patients with normal TTR levels. Similarly, low serum TTR levels were
reported to be correlated with delayed wound healing in other studies.[32]
[34]
[35]
[36] High TTR levels are a strong indicator of complete wound site healing in burn patients
and that the TTR level correlates with wound healing.[37]
[38] Salvetti et al reported that preoperative low TTR levels are a marker of increased
surgical site infections for elective spinal surgery cases.[39]
Furthermore, young stroke patients with high serum TTR levels have a better prognosis;
low TTR levels were reported to be seriously compatible with a poor prognosis.[40] Under physiological conditions, TTR passes to the peripheral nerve either by crossing
the blood–nerve barrier or through cerebrospinal fluid (CSF). Similarly, it passes
to the central nervous system through both CSF and blood. TTR circulating in CSF has
neuroprotective properties.[41] It has been shown in a similar study that TTR accumulates in damaged neural tissues
and increases neural growth.[40] In a study conducted with mice that were genetically unable to produce TTR, TTR
had an enhancing effect on regeneration in the damaged nerve.[3] Moreover, TTR has a neuroprotective effect through polyphenols.[42] In another study, low TTR levels detected in CSF were compatible with increased
dementia in Alzheimer's patients.[43]
In conclusion, TTR is an indicator of nutritional status in studies previously published
in the literature, and low TTR levels have been shown to be an indicator of impaired
nutrition. Impaired nutrition is known to cause increased morbidity, mortality, delayed
wound site healing, increased susceptibility to infection, and prolonged hospital
stay in surgical patients. Low TTR levels have been shown to delay wound site healing,
increase mortality and morbidity, prolong hospital stay, and cause susceptibility
to infection.
Moreover, TTR's neuroprotective effects have been shown to accelerate nerve regeneration.
Furthermore, TTR, found in human serum and CSF, is an easy to measure and inexpensive
protein with multiple functions.[44] For these reasons, in this study, the effects of preoperatively measured serum TTR
levels on surgical outcomes, complication rates, wound site healing, postoperative
infection, and hospital stay were studied in patients operated for spinal pathologies.
In our study, the preoperatively measured low TTR levels were reported to be consistent
with low Karnofsky score, increased ODI scores, and high early and late postoperative
VAS values. Therefore, low TTR levels were reported to correlate with poor clinical
outcomes after spinal surgery. The length of hospital stay increased in patients with
low TTR levels. Moreover, patients with low TTR levels have more wound site infection
and delayed wound site healing. Similarly, patients with low TTR levels were exposed
to increased peroperative complications and postoperative morbidity. However, a high
CRP/TTR ratio was reported to be compatible with low Karnofsky score, increased ODI
scores, and high early and late postoperative VAS values.
Based on these results, similar to TTR, CRP/TTR ratio was evaluated as a data that
can be used to predict postoperative clinical outcomes in spinal surgery. Furthermore,
high CRP/TTR ratio was reported to be correlated with increased wound site infection
and delayed wound site healing. It was found to be correlated with perioperative complications
and increased morbidity in patients with a high CRP/TTR ratio. Based on these results,
serum TTR levels preoperatively measured in patients undergoing spinal surgery emerge
as a parameter that can be used to predict postoperative surgical results, wound site
healing status and wound site infection, peroperative complications, and morbidity
risks.
Conclusion
In patients with low preoperative serum TTR level measured before spinal surgery,
the following were found
-
Postoperative Karnofsky scores were lower.
-
Postoperative ODI levels were higher.
-
Postoperative early and late VAS scores were higher.
-
Length of hospital stay was prolonged.
-
Peroperative complication rates were higher.
-
Wound site infection rates were higher.
-
Delay in wound site healing increased.
-
Morbidity rates were higher.
In conclusion, preoperative low TTR levels were reported to be an effective parameter
that can be used to predict surgical outcomes, wound site infection and wound site
healing status, peroperative complications, and morbidity in spinal surgery.
