Recently, several studies have demonstrated that AIS could be successfully corrected using minimally invasive techniques [11, 14, 15, 19, 32]. Unlike the percutaneous technique and lateral interbody fusion commonly applied in adult degenerative spinal deformity, MISS for AIS used one to three small skin incisions from a posterior approach, which allowed surgeons to perform facetectomy, instrumentation, and fusion [15, 33]. Compared to standard PSF, dissection of soft tissue and the spinal posterior structure was less common, and the cosmetic concerns of adolescent patients could be relieved [19, 32]. However, as a novel technique, the effectiveness of MISS for the management of AIS still needs to be clarified. In this study, we found that MISS was associated with adequate deformity correction, better restoration of TK, less EBL, lower blood transfusion rate, shorter LOS, and less pain than PSF.
Correction in the coronal and sagittal planes
Due to the limited available instrumentation for MISS, the application of compression, distraction, and in situ contouring are still challenges for this technique [31]. However, MISS yielded correction of the main curve Cobb angle, LL, CB, and SVA that was comparable to standard PSF in this meta-analysis. The better restoration of TK in the MISS group, especially for patients with > 10 fusion levels, was an unexpected finding, although the difference was minor, and whether it had any clinical significance remained debatable. Recently, Sarwahi et al. reported that MISS was associated with significantly greater restoration of TK and a lower postoperative hypokyphosis rate than standard PSF [21]. One plausible explanation was that the preservation of paraspinal muscles and ligaments in the MISS procedure played an important role in maintaining sagittal alignment after surgery [35]. Overall, MISS was adequate to achieve the goal of deformity correction in both the coronal and sagittal planes.
Estimated blood loss
Minimizing blood loss is of paramount importance for the surgical management of spinal deformity, as the use of allogenic blood transfusion is a risk factor for the development of surgical site infection after spinal fusion [36]. The impact of bleeding-related and transfusion-related morbidity is even more pronounced in pediatric and adolescent populations. The smaller body size of these patients could accelerate the proportional loss of blood volume during deformity correction surgery for AIS [37]. In a cohort of 188 patients with AIS, Ialenti et al. reported a mean EBL of 907 ± 775 mL with standard PSF [34]. In another large cohort of 43,983 patients who underwent PSF for AIS, Yoshihara and Yoneoka reported a blood transfusion rate of 30.4% [38]. Chiu et al. showed that screw insertion and derotation maneuvers were the main contributors to EBL during standard PSF [39]. These two steps were not modified during the MISS procedure; nonetheless, the EBL and the corresponding need for blood transfusion were significantly decreased in the current study, with an EBL of 288.25 mL and a blood transfusion rate of 8.0%. We believe that the decreased EBL in the MISS group was attributable to the smaller incisions, reduced tissue dissection, and smaller area of subperiosteal exposure [14]. The ORT was significantly longer in the MISS group than in the PSF group in this study, which reflected the learning curve when applying the new technique [32]. As the longer ORT could also contribute to the increased EBL, we considered that the EBL in the MISS group might continue to decrease with increasing surgeon experience [34, 40].
Length of hospital stay, postoperative pain, and SRS-22 score
This meta-analysis found that the LOS was shorter by 1.48 days in the MISS group than in the PSF group. We considered that this positive outcome was related to the reduced dissection of muscle and EBL in the MISS procedure, which enhanced patient recovery and allowed early discharge. Sultan et al. reported that intraoperative blood loss was an independent risk factor for prolonged LOS in AIS patients undergoing PSF because patients with increased EBL were more likely to develop postoperative complications and require additional management [41]. The shorter LOS may also be tied to patients’ postoperative pain and satisfaction [42, 43]. Martin et al. reported that achieving adequate pain control could decrease the opioid requirement and minimize opioid-induced side effects, allowing AIS patients to mobilize early [44]. Additionally, patients with higher satisfaction tend to return home faster [44]. Consistent with previous studies, we found that patients in the MISS group reported significantly less pain, according to the SRS-22 pain score (4.31 vs. 3.80) and the VAS score (5.94 vs. 6.28), which was attributed to the reduced dissection of tissue with this new technique. Although the results were not statistically significant, the SRS-22 self-image/appearance score was also higher in the MISS group than in the PSF group (4.05 vs. 3.95). Better pain management and scar appearance could improve overall patient satisfaction, which was reflected by the higher overall SRS-22 score in the MISS group (4.22 vs. 4.11) and led to the shorter LOS in this study.
Complications
The overall complication profile is an essential aspect of the safety evaluation of a novel surgical technique, particularly in the early stage of application. In this study, the MISS yielded a complication spectrum and complication rate similar to those of standard PSF (15.0% vs. 18.0%), suggesting that this new procedure would not cause additional harm and was a safe alternative to conventional open surgery. The most commonly reported complication in both groups was surgical site infection. In this study, the surgical site infection rate was 2.0% in the MISS group, similar to the PSF group. In a previous study of 540 AIS patients who underwent standard PSF, 2.8% of cases developed postoperative surgical site infection, which was consistent with our results [45]. Additionally, surgical site infection is associated with ORT [14]. We believe that the decreased ORT of the MISS procedure with increasing surgeon experience would lead to a decrease in the surgical site infection rate.
Concerns regarding the MISS procedure
Despite MISS’s advantages of better restoration of TK, less EBL, fewer transfusions, shorter LOS, and less postoperative pain, several important concerns must be acknowledged. The ORT was significantly longer with MISS than with PSF (376.40 min vs. 287.60 min). The difference in ORT may be the effect of the steep learning curve, and de Bodman et al. and Zhu et al. reported that the ORT of the MISS procedure could be shortened by nearly 60 min with increased surgeon experience [32, 40]. Nonetheless, even when MISS is performed by a proficient surgeon, the ORT might still be slightly longer than that of standard PSF, and this should be emphasized as a potential limitation of MISS for AIS [33]. Prolonged surgical times keep patients under extended anesthesia, which may have deleterious effects on perioperative morbidity [18].
Concerns have been raised that more fluoroscopic imaging may be needed during the MISS procedure, which could increase the radiation exposure of both patients and surgeons [32, 33]. However, Si et al. reported that a satisfactory screw placement rate of 93.6% could be obtained using the freehand technique, and no significant difference in radiation exposure was found between the MISS group and the PSF group [19]. Additionally, Zhu et al. achieved a satisfactory screw placement rate of 93.8% in the MISS group through O-arm navigation, and patients and surgeons could be exposed to less radiation from an O-arm than with traditional fluoroscopy [32].
Appropriate patient selection is important for a new technique, although little is known about the indications of the MISS procedure. To date, most studies have focused on moderate (< 80°) and flexible (> 50%) AIS [14, 15, 19, 20, 31, 32]. Whether the indication can be extended to patients with more severe and rigid curves remains unclear.
Limitations
This study had several limitations. First, the heterogeneity of the included patients should be acknowledged, as they had different curve types and fusion levels. Two studies focused on Lenke types 1–4, two studies focused on Lenke type 5, the other three studies included mixed curve types (Lenke type 1–6), and the fusion levels in these studies ranged from 5 to 12. Heterogeneity may have impacted meaningful pooling. Although we performed subgroup analysis, heterogeneity was still present. Other sources of heterogeneity may have been differences in surgeons, management protocols of spinal centers, and study designs. Second, it is important to address the inherent potential for bias underlying the use of retrospective studies of MISS versus PSF. It is clear that in the absence of randomization, patient selection for MISS or PSF is likely to be a key source of bias in all of the included studies. The patients who are most appropriate for MISS are likely to be selected to undergo MISS, and vice versa. Third, because MISS is a new technique, the results reported by the included studies reflected the early experience of surgeons. Regarding the learning curve of the MISS, de Bodman et al. indicated that the outcomes could be improved after the first 25 cases [40]. However, five of the included studies reported data for their first 25 cases [15, 20, 31,32,33]. Therefore, the results of this study may be limited when surgeons gain more experience.