Both surgeons used a navigation-assisted technique for all patients who presented with a localized bone sarcoma of the pelvis or sacrum, indicating there was no selection process that determined which patients were treated with this technique.
In all patients, the diagnosis was made by biopsy and review of the pathology by a musculoskeletal pathologist (CF, JH, WJ, EV-S, DR, ME). Treatment plans for all patients were made by a multidisciplinary sarcoma tumor board, and decisions regarding chemotherapy and/or radiation therapy were made by the relevant specialists with input from the surgeon as required. Diagnosis and margins were confirmed by the final resection specimen and rereviewed at the multidisciplinary sarcoma conference to determine the need for any additional postoperative treatment.
All patients underwent CT scanning of the pelvis postoperatively, and the postoperative CT was compared with the resection plan. Surgical margins, both bone and soft tissue, were all measured on the resection specimen, and again any variation from the planned intended margin was noted.
Patients underwent surgical treatment with the intent of achieving a negative margin resection in all patients. Two navigation systems were used: the BrainLab Kolibri™ system (BrainLab, Munich, Germany) was used for nine patients and the Medtronic O-arm navigation system (Medtronic, Minneapolis, MN, USA) for the remaining patients. This was because the Medtronic O-arm system only became institutionally available to one surgeon after nine patients had already been treated. Registration arrays were placed into a portion of the bone not being resected but readily visible to the camera. Once placed, they were not moved for the entirety of the procedure. After preoperative planning of osteotomies was performed, the navigation system was used to guide all osteotomies being made in whatever manner the surgeon felt was most appropriate for that particular patient. For instance, in some patients, a navigated saw or osteotome was used, whereas in others, a navigated drill was used to drill a series of pilot holes with the osteotomy and then completed with a traditional osteotome. Soft tissue dissection and closure were carried out in a standard fashion (Fig. 1).
Patients are generally kept toe-touch weightbearing for 8 weeks if reconstruction was performed. In procedures not involving reconstruction, patients were allowed to immediately bear weight. Physical therapy was initiated once the wound was healed and the patient was allowed to fully weightbear.
We studied surgical outcomes including (1) preoperative factors: age, diagnosis, grade, and tumor location; (2) intraoperative factors: surgical margins and intraoperative complications; and (3) postoperative factors: postoperative complications. We also studied oncologic factors related to local control: recurrence and disease status at last followup. Additionally, although potentially not influenced by the use of navigation-aided resections, we tallied the development of metastases and survival in the study population, which are more closely related to the patient’s diagnosis. Patients were followed postoperatively according to the standard National Comprehensive Cancer Network guidelines for bone sarcoma followup. This includes imaging of the local site and chest every 3 months postoperatively for 2 years followed by the same routine every 6 months until Year 5.
In our series, 21 of 23 patients had negative margin resection. In all patients, the bone margin was negative, but two patients (both of whom underwent sacral resections) were found to have positive soft tissue margins (Table 2). The width of the margin was up to the discretion of the surgeon and the treating institution and we did not record specific widths of bony or soft tissue margins. They were deemed either positive or negative. In addition, we performed postoperative CT scans on all patients and manually compared the final resection CT scan with the planned resection and found no significant variations.
Six of 23 (26%) patients experienced local recurrence within the study period. Two of the six recurrences occurred in patients who had positive margins at the initial resection. One patient with sacral chordoma undergoing a margin negative resection had locoregional recurrence in an inguinal lymph node; this was considered metastasis rather than a local recurrence because it was outside of the operative field.
Although there may not be a relationship to the use of navigation-assisted resection, we tallied the development of metastases and survival in this cohort. It is likely that these parameters are more closely related to histology and grade of the tumor than type of resection, but for completeness, the data are being reported. Two patients developed lung metastases during followup; one patient had a high-grade unclassified round cell sarcoma and one had a Ewing sarcoma. Three patients died during the study period; two patients died of disease at 11 and 27 months and a third died from sepsis 3 months after the procedure. Seventeen patients demonstrated no evidence of disease at last recorded followup (Table 2).
There were four intraoperative complications: one dural tear during chordoma resection in a patient with prior lumbar decompression surgery for spinal stenosis treated with repair; one iliac vein laceration treated with ligation with an initial period of leg swelling, which resolved, and no long-term sequelae; one bladder injury in a patient with a large chondrosarcoma extending across the midline anteriorly treated with repair; and one patient had unexpected foot drop after surgery, which did recover within 2 months. Eight patients had wound complications resulting in operative débridement. None of these 13 complications were believed to be the result of the use of a navigation system for resection nor were they thought to be problems that could have been avoided with the use of navigation, because they were all directly related to dissection of the soft tissue portion of the tumor margin. Two patients in the series, however, developed postoperative femoral nerve palsy, both of which recovered within 6 months. In both of these patients, the tumor was not in the immediate vicinity of the femoral nerve, and the navigation registration array was placed in the pubic ramus in the vicinity of the femoral nerve. It is likely that this may have led to intraoperative stretch of the nerve leading to the observed femoral palsies. As such, these were considered navigation-related complications.
Pelvis and sacral sarcomas represent some of the most challenging problems in orthopaedic oncology. The anatomic complexity of the region makes wide margin resections difficult to achieve, potentially contributing to a high risk of local recurrence. In the last decade, several case reports have demonstrated the usefulness of computer-assisted surgery in resection and reconstruction of pelvis and sacral malignancies [2, 8, 9, 12]. However, only three reports include the outcomes of a series of patients treated with a computer-assisted technique for pelvis and sacral sarcoma [4, 11, 17]. We found in our patients that computer-assisted osteotomies were associated with achieving negative bony margins in all patients and had a relatively low incidence of complications.
This study has a number of major limitations. First, a major limitation is that we did have some loss to followup: four patients have no available data after their 2-year postoperative visit. However, we do have complete data available for a minimum of 2 years on every patient who survived for that period, which is the period of highest risk for local recurrence. However, it is possible that these patients may have experienced poor outcomes after the last followup, making this a notable limitation. Second, there were two surgeons and two navigation systems that were used in this study, introducing the possibility that some of the data presented reflects a learning curve rather than true outcomes, particularly in light of the complexity of the surgical procedure. We tried to minimize this risk by only including patients after a period when both surgeons had significant familiarity with their respective systems, and we chose a study period that began after each surgeon had performed a minimum of 10 resections using navigation in addition to sawbones and cadaver sessions with both systems.
Third, the navigation systems used do not have formal oncology software packages, and so modifications were made to existing spinal software to allow the resections. This introduced the major limitation that there was no software available at the time of the study that could be used to directly overlay and compare the planned resection with the actual resection. This was therefore done manually, which introduces some potential bias. For this reason, we did not focus on accuracy of the resection as a parameter in this study, but we feel that this would have strengthened the case for the benefit of navigation in this type of resection. At the time of this writing, however, several manufacturers have developed oncology-specific software, which allows this type of direct comparison, so we will plan to report these data in future publications involving navigation-related resection.
Fourth, the small number of patients makes meaningful interpretation and analysis difficult. Because this is a retrospective cohort study, we do not have a control group. This prevents a direct comparison to patients treated in the same framework without a navigation-assisted technique. Additionally, because the cohort is small, we performed no statistical analysis and only proportions are presented. However, we attempted to avoid any selection bias by using this procedure on every patient who presented with the correct indication, and we applied reasonably consistent methods and data collection. In addition, the followup data are relatively complete, thereby maximizing the information that we can glean from a study of this cohort. Fifth, the patient population is heterogeneous. The cohort of “pelvis and sacral sarcomas,” although narrower than other navigation-related cohort studies to date, still includes a wide variety of patient diagnoses and ages, which limits the strength of the conclusions and makes the metastasis and survival data less relevant to the use of the navigation technique. However, the primary goal of this study was to determine the likelihood of achieving negative margins utilizing a navigation technique, which we felt could be adequately studied even in a heterogeneous population. Sixth, we recognize that our followup is relatively short at an average of 27 months. However, because bone sarcoma local recurrences generally occur within the first 2 years after initial treatment [7, 13], we believe that our followup was long enough to capture the majority of local recurrences. The focus of this study was not to evaluate metastases or survival, which would require substantially longer followup and may not be directly related to the use of a navigation-assisted technique; however, we felt it was important to include this information and plan to report further on detailed survival data at longer followup points for this cohort in future publications.
A total of 21 of our 23 patients had negative margin resection using this computer-assisted technique. In our study, the bone resection margins were all clear of tumor, but in two of the patients, soft tissue margins were positive. Only one other study has studied resection margins in pelvic malignancies utilizing a navigation technique . In this study, however, several pelvic metastases and colorectal malignancies extending into the sacrum were included, making this a much more heterogeneous sample, and had a short followup of 13 months. That study demonstrated a similar proportion of negative margin resection with 29 of 31 negative margins and, similar to our study, noted that the actual bone resection margins in all patients were negative with two of the patients having positive soft tissue margins. Wong and Kumta  looked at the proportion of negative margins in resections including all anatomic locations and included seven pelvis and five sacral primary tumors. In this study, all patients achieved negative bone and soft tissue margins, although the margins were noted to be marginal in five patients. Another study included 10 patients with pelvis and sacral tumors as part of a larger cohort and reported negative margin resection in all 10 patients . No other studies to date have reported outcomes on this specific cohort of patients. Given the paucity of available data, it is difficult to draw firm conclusions, but these results do suggest that navigation may play a role in allowing the surgeon to consistently achieve negative bone margins during difficult pelvis and sacral resections.
Historical studies on pelvic sarcoma have noted recurrence rates of 60% or higher [6, 14]. In our study, we noted six of 23 local recurrences. When breaking these patients down by anatomic site, four were in patients treated for pelvic sarcoma (four of 15) and two in patients treated for sacral tumors (two of eight). Two of the recurrences developed in patients who had positive margins at the initial resection. Both of these patients had tumors located in the sacrum. Similarly, in the Wong and Kumta study , three of the four recurrences seen were in patients with sacral tumors, suggesting tumors in this location may be more prone to recurrence. Because navigation-assisted resection may influence the probability of achieving a negative margin resection, there may be an impact on the local recurrence, but this study and other studies like it to date have had too few patients to be able to demonstrate a clear difference.
We noted several complications in this study, the majority of which were wound complications, occurring in eight (eight of 23) patients. This represents a similar rate as other published studies on pelvic and sacral tumors [4, 11, 17]. Other intraoperative complications (iliac vein laceration, bladder laceration, and dural tear) also did not have any relationship to the navigation technique. We did note, however, two transient femoral nerve palsies in patients who had the reference trackers placed in the pubic ramus because that was the only portion of the pelvis not being resected. It is possible these palsies may have been related to the placement of the reference tracker by forcing the nerve to stretch around the tracker pin, leading to neurapraxia. Because this is not a complication typically seen with pelvic resection, we considered it to be navigation technique-related. This type of complication could potentially be avoided by careful placement of the arrays away from neurovascular structures in patients in whom there is limited bone available.
However, it is important to recognize that tracker-related problems may occur and careful planning is needed to avoid such issues.
In addition to the high proportion of negative margin resection, we identified several patients in whom the navigation assistance allowed for subjective improvement in the resection. We felt that navigation assistance was particularly useful in pelvic procedures in which the majority of the tumor was intraosseous without substantial soft tissue mass. In these patients, the navigation allowed “visualization” of the tumor inside the bone allowing a measure of safety against an inadvertent violation of the tumor with an osteotomy (Fig. 2). Again, this is a purely subjective evaluation based on the authors’ own experience. In one patient, the navigation allowed for partial acetabular resection of only the anterior and inferior portions of the acetabulum, allowing for negative margin resection while sparing the hip (ie, not necessitating reconstruction), which we feel would otherwise have been difficult or impossible to achieve without navigation assistance (Fig. 3). In three patients with sacral tumors, we noted that additional nerve roots could be spared utilizing the guidance of the navigation system. In one patient we were able to perform a posterior-only resection in which the treating surgeon would have otherwise utilized a combined AP approach, sparing the patient the substantial additional morbidity of the entire anterior approach. This is in agreement with findings of both Jeys et al.  and Wong and Kumta , both of which demonstrated similar findings in their patients with sacral tumors. We also subjectively felt that navigation facilitated more precisely matching osteotomies to a custom-made prosthesis than could have been done without navigation assistance. Another previously reported benefit  that we realized in this series was the ability to perform complex multiplanar osteotomy around the acetabulum, allowing for a joint-sparing resection while still achieving a margin negative resection.
We found that navigation-assisted resection of pelvic and sacral tumors is associated with a high proportion of patients achieving a negative margin resection of these difficult tumors. Although we cannot compare this with a similar group of patients resected without navigation, it is possible that the ability to more reliably achieve negative bony margins may be associated with a lower risk of local recurrence. Navigation may add risk of additional complications like nerve injury to the procedure, although it may have additional benefits such as sparing sacral nerve roots or allowing complex joint-sparing resection. Further study will be needed to improve computer-assisted techniques for application in tumor resection, but our initial results demonstrated potential value in the utilization of this technique, and we recommend consideration of its use for complex pelvic or sacral resections.
We thank the following pathologists who evaluated the resection specimens in this series: Christopher Fletcher, Jason Hornick, Wei Jiang, Esperanza Villaneuva-Siles, Danielle Rocchio, and Michelle Ewart.
1. Cartiaux O, Docquier PL, Paul L, Francq BG, Cornu OH, Delloye C, Roucent B, Dehez B, Banse X. Surgical inaccuracy of tumor resection and reconstruction within the pelvis: an experimental study. Acta Orthop. 2008;79:695–702.
2. Cho HS, Kang HG, Kim HS, Han I. Computer-assisted sacral tumor resection. A case report. J Bone Joint Surg Am. 2008;90:1561–1566.
3. Cho HS, Oh JH, Han I, Kim HS. Joint-preserving limb salvage surgery under navigation guidance. J Surg Oncol. 2009;100:227–232.
4. Cho HS, Oh JH, Han I, Kim HS. The outcomes of navigation-assisted bone tumour surgery: minimum three-year follow-up. J Bone Joint Surg Br. 2012;94:1414–1420.
5. Enneking WE, Dunham WK. Resection and reconstruction for primary neoplasms involving the innominate bone. J Bone Joint Surg Am. 1978;60:731–746.
6. Fuchs B, Hoekzema N, Larson DR, Inwards CY, Sim FH. Osteosarcoma of the pelvis: outcome analysis of surgical treatment. Clin Orthop Relat Res. 2009;467:510–518.
7. Hawkins DS, Arndt CA. Pattern of disease recurrence and prognostic factors in patients with osteosarcoma treated with contemporary chemotherapy. Cancer. 2003;98:2447–2456.
8. Hüfner T, Kfuri M Jr, Galanski M, Bastian L, Loss M, Pohlemann T, Krettek C. New indications for computer-assisted surgery: tumor resection in the pelvis. Clin Orthop Relat Res. 2004;426:219–225.
9. Ieguchi M, Hoshi M, Takada J, Hidaka N, Nakamura H. Navigation-assisted surgery for bone and soft tissue tumors with bony extension. Clin Orthop Relat Res. 2012;470:275–283.
10. Jeys L, Grimer R, Carter S, Tillman R, Abudu S. Outcomes of primary bone tumours of the pelvis: the ROH experience. J Bone Joint Surg Br. 2012;94(Suppl):39.
11. Jeys L, Matharu GS, Nandra RS, Grimer RJ. Can computer navigation-assisted surgery reduce the risk of an intralesional margin and reduce the rate of local recurrence in patients with a tumour of the pelvis or sacrum? Bone Joint J. 2013;95:1417–1424.
12. Krettek C, Geerling J, Bastian L, Citak M, Rücker F, Kendoff D, Hüfner T. Computer aided tumor resection in the pelvis. Injury. 2004;35(Suppl 1):S–A79–83.
13. Nathan SS, Gorlick R, Bukata S, Chou A, Morris CD, Boland PJ, Huvos AG, Meyers PA, Healey JH. Treatment algorithm for locally recurrent osteosarcoma based on local disease-free interval and the presence of lung metastasis. Cancer. 2006;107:1607–1616.
14. Ozaki T, Flege S, Kevric M, Lindner N, Maas R, Delling G, Schwarz R, von Hochstetter AR, Salzer-Kuntschik M, Berdel WE, Jürgens H, Exner GU, Reichardt P, Mayer-Steinacker R, Ewerbeck V, Kotz R, Winkelmann W, Bielack SS. Osteosarcoma of the pelvis: experience of the Cooperative Osteosarcoma Study Group. J Clin Oncol. 2003;21:334–341.
15. So TY, Lam YL, Mak KL. Computer-assisted navigation in bone tumor surgery: seamless workflow model and evolution of technique. Clin Orthop Relat Res. 2010;468:2985–2991.
16. Wirbel RJ, Schulte M, Mutschler WE. Surgical treatment of pelvic sarcomas: oncologic and functional outcome. Clin Orthop Relat Res. 2001;390:190–205.
17. Wong KC, Kumta SM. Computer-assisted tumor surgery in malignant bone tumors. Clin Orthop Relat Res. 2013;471:750–761.
18. Wong KC, Kumta SM. Joint-preserving tumor resection and reconstruction using image-guided computer navigation. Clin Orthop Relat Res. 2013;471:762–773.
© 2018 Lippincott Williams & Wilkins LWW
19. Zeifang F, Buchner M, Zahlten-Hinguranage A, Bernd L, Sabo D. Complications following operative treatment of primary malignant bone tumours in the pelvis. Eur J Surg Oncol. 2004;30:893–899.