COMPLICATIONS OF NON-ENDOSCOPIC DISCECTOMY : A RETROSPECTIVE STUDY OF TWENTY-ONE

Objective: To report and compare the number and grade of major complications presented with non-endoscopic thermal discectomy and nucleoplasty for the treatment of discogenic axial lumbar pain using laser and radiofrequency. Methods: A 21 years retrospective study was conducted of the clinical charts of patients whose reason for consultation was axial lumbar pain from degenerative disc disease, and who underwent surgery using non-endoscopic discectomy and nucleoplasty (NEDN). Two groups were established; the first, NEDN with laser, and second, NEDN with radiofrequency. The number and types of complications reported in the case-series were counted, and their statistical differences determined. Results: The inclusion criteria were fulfilled by 643 of the medical charts. 26 complications were reported, the most common being radiculitis (n=12). Statistically significant differences were found between the complications occurring in the two groups (p=0.01). Conclusion: The number of complications showed statistically significant difference. The severity of the complications and adverse outcomes provide an argument for choosing one technology over the other. Training and the learning curve stage are important factors to be taken into account, to avoid complications.


INTRODUCTION
Discogenic pain is defined as back pain caused by disc degeneration. 1Currently, there are different alternatives for the treatment of discogenic pain.Options range from conservative medical management to open arthrodesis. 24][5] One of the MISS procedures most widely used in the treatment of discopathy is non-endoscopic thermal discectomy and nucleoplasty (NEDN). 6,7This technique consists of the placement of non-endoscopic fibers to transmit thermal energy and increase temperature in the intradiscal zone, causing the ablation of pain generating nerves, 6 complemented by decompression with mechanical nucleoplasty. 7,8ince its description by Hijikata, 9 and its complementing with thermal energy by Choy et al., 10 this technique has been implemented in more than 500,000 patients around the world. 11The principle behind the treatment is that it produces sufficient intradisc temperatures to achieve annular collagen shrinkage, nucleus dehydration, and nociceptor ablation in the posterior annulus area, with resulting pain relief. 12Most fibers use either laser (LS) or radiofrequency (RF) as the heat source. 5][15][16][17][18][19][20][21] Nevertheless, their safety margin and degree of complications continue to create extensive controversies between different surgeons.The purpose of the study was to report and compare the number and grade of major complications present in non-endoscopic thermal discectomy and nucleoplasty, in the treatment of discogenic axial lumbar pain using LS and RF technologies.

MATERIALS AND METHODS
This is a retrospective study reviewing medical charts of patients treated with NEDN that reported any type of major complication associated with the procedure between 1993 and 2014.The study only included patients whose reason for consultation was axial lumbar pain, and whose final diagnosis was either degenerative disc disease (DDD), black disc, disc bulging, annular tear, or contained herniated disc.The disease was diagnosed by history, clinical examination, plane and dynamic X-ray and MRI.Also, for inclusion of the record in the study, the surgical notes had to report at least one positive discogenic test (≥5/10).Medical records were excluded belonging to patients with pathologies such as discopathy with more than 50% height loss, any degree of segmental instability, or positive discography not consistent with lumbar symptoms, radicular pain, and medical charts with incomplete data on follow-up and evaluation criteria.The study was not submitted to the ethics committee.

Setting and surgical procedure
Surgeries were performed at the Reina Sofia Clinic in Bogotá Colombia, by the same team of surgeons and using a standardized technique, with minor changes throughout the experience.The source of laser energy used was the Holmium YAG Laser (Trimedyne Inc.Irvine, CA).The RF electrode employed was Disc-FX Bipolar System (Elliquence, NYC, NY) and a high-frequencylow temperature radiofrequency energy source called Surgimax (Elliquence LLC, NYC), in bipolar mode.
The patient was placed in the prone position.An epidural needle was inserted at a 45° angle in the direction of the foraminal area, 8 cm to 12 cm from the midline.The aim was for the tip of the needle to enter the posterior third of the intervertebral space.After determining whether the level tested during the discography procedure reproduces the patient's familiar, concordant pain (positive result), we entered the disc through a system of dilator, cannulas and trephines.These elements was placed sequentially, always under the verification of the fluoroscope, to reach the annulotomy and enter the nucleus.In order to allow adequate access to the fiber and achieve neural decompression, a nuclear resection was performed.This mechanical discectomy was performed using the punch forceps.Lastly, an electrode was inserted, and the thermal discectomy and nucleoplasty was performed.

Clinical evaluation
"Major related complications" designates the presentation, during the intraoperative period and up to three months after NEDN, of phenomena such as transient or definitive motor deficit, dural tear, headache, radiculitis, chemical discitis, or vertebral end-plate burning, and infectious spondylodiscitis.
To minimize bias, including among observers, a company independent from the researchers conducted the review of medical records and gathered data from the patients with major complications.

Statistical analysis
Data analysis was performed with the statistical software R 3.1.1for Windows 8.The analysis determined certain aspects in the variables, including frequency and descriptive statistics.The tests selected for to compare the groups were X 2 and Wilcoxon for related pairs.Statistical significance was p<0.05.
LS was used from March 1993 to January 2008.The inclusion criteria were fulfilled by 171 medical charts.In turn, RF was applied between January 2008 and March 2014, and 472 medical charts met the inclusion criteria.In twenty-one years, a total of 26 major complications were reported.
For the 171 medical charts that fulfilled the inclusion criteria, complications occurred in 10.53% of the patients (n=18) when LS was used.The distribution of these complications was: motor deficit (n=2), one of them definitive with dorsiflexor paresis (2/5); headache caused by dural sac tear (n=2); post-operative radiculitis (n=8); chemical discitis or vertebral end-plate burning with the laser light (n=4), and infectious spondylodiscitis (n=2).
The differences in the presentation of complications were statistically significant (p=0.01).

DISCUSSION
The number and type of complications are important in determining the safety of a surgical procedure.Several factors can increase the possibility of complications: wrong choice of patient, poor technique, being in initial stages of the learning curve, and certain concomitant diseases. 22To the best of our knowledge, no publication has compared complications presented with percutaneous disc decompression between two analog technologies such as LS and RF, performed by the same surgical team, in the treatment of discogenic pain.Thus, it is important to analyze factors such as type, severity and solution of the complication, and the learning curve stage in which such complication occurred.
The percentage of complications in the global sample (LS+RF) was 4.0% (26/643), which could be deemed high if reports that fall below 1% are considered. 18,19,23We hypothesize that because of the complications, the medical charts may have been completed in greater detail, always complying with the "complete data on follow-ups and evaluation" inclusion criteria.Thus, the relation of the complications to the number of patients without complications led to a number of compliant records that was less than the total number of patients treated in the 21 years of the experience (n=1,357), therefore the resulting percentage was higher.Thus, if the study links complications to the total number of patients, this gives a global percentage of 1.9%, which is more like the percentage reported in general in the literature.
Another interesting data was that 69% of the complications (18/26) occurred in the initiation stage of the learning curve with LS.This result could be associated with factors such as the learning curve development and the high temperatures reached with this technology.Indeed, these factors caused the complication percentage with LS to become 10.4%, a figure that is considerably higher than cited in reports by authors who use the same technology, which is below 1%. 16][19] Regarding the type of complication, for laser, the literature reports spondylodiscitis as the most common, with an occurrence ranging from 0.24% to 1.2%. 15,16,24In contrast, this series shows radiculitis as the principal complication.This complication is related to the initial stages of the learning curve, incorrect (too deep) anesthesia protocol, and excessive heat from fibers with low irrigation.Concerning the use of RF with the same system used in this case-series, to our knowledge, this type of complication was not reported.With other RF available systems, minor complications have been reported, such as discomfort in the area of the incision, numbness, and leg weakness. 25,26  regard to types of complications, a sudden change is evident in the use and preference of RF over LS, due precisely to the number of complications and their nature and severity, specifically in the presentation of "thermal necrosis of vertebral endplate."This complication continues to occur even in the advanced stages of the learning curve. 16It is also important to mention that of the 26 complications presented in this sample, only one -a permanent motor deficit caused by thermal L5 root lesion with dorsiflexor paresis -did not respond to standard medical treatment, and the remaining 25 patients were treated and resolved.Chemical discitis only appeared upon using LS (n=4) and constituted with the definitive motor deficit, one of the reasons for assessing the effectiveness and safety of RF.

CONCLUSION
MISS has been reported as a group of effective and safe procedures.Nevertheless, there are several factors that could increase the rate of complications.In this series, we concluded that the training and learning curve stages are important factors to take into account.Also, it must be noted that whereas the two techniques have proven to be effective in the treatment of discogenic lumbar axial pain, in this sample, the occurrence of complications was different and statistically significant.Likewise, the severity of the complications, and their negative outcome, provide an argument for choosing one technology over the other.

Table 1 .
Sex and age distribution of the population.