Spine Metastasis in Elderly

Encouraging Results for Better Survival

Quentin Beaufort, MD; Louis-Marie Terrier, MD; Arnaud Dubory, MD, PhD; Louis-Romée Le Nail, MD; Ann-Rose Cook, MD; Joseph Cristini, MD; Kévin Buffenoir, MD, PhD; Hugues Pascal-Moussellard, MD, PhD; Alexandre Carpentier, MD, PhD; Bertrand Mathon, MD; Aymeric Amelot, MD, PhD


Spine. 2021;46(11):751-759. 

In This Article


SpM is reported to be found at autopsy in up to 70% of patients that presented a primary tumor, although only around 10% of these patients were stated symptomatic.[16] Metastasis have a profound impact on patient prognosis, indeed according to the American Cancer Society, five-year survival rates in non-metastatic versus metastatic carcinomas drop from 80% to 100% to 16% according to the primary tumor (14.5% in our SpM series).[17] Moreover, it is described that SpM represent the terminal stage of the primary cancer, as previously studied a chronological metastatic process is described concerning the course of progression of cancer.[18,19]

With the increased prevalence of cancers as well as treatment advancements in controlling systemic disease and local recurrences, the incidence of patients developing SpM is also expected to increase. In result, patients who develop SpM are increasingly older, and this was the main postulate to study patients > 70years' old. Thereby, the rationale of this study was to identify accessible clinical prognostic factors of survival in elderly patients with SpM, to thereby improve disease anticipation and treatment.

Independently of the age of patients, several studies have shown that primary tumor histology is the most important prognostic factor of survival.[13,14,20] WHO status and epiduritis are also known to be strong predictor parameters of survival and for many patients a diagnosis of MSCC indicates the final stage of their disease. These parameters remain relevant in elderly patients. In our series, primary tumor in particular lung and visceral cancer, poor WHO status, no ambulatory function, and the presence of epiduritis were bad prognostic factors with a significant statistical and independence difference.

Moreover, neurological status (appreciated by Frankel score) remains a good main prognostic survival. SpM' impact on neurological function is included in some prognosis scores.[14,21,22] The ambulatory status (Frankel or ASIA) has already been shown as a prognostic factor in other spine pathologies such as degenerative,[23] traumatic,[24] even for SpM.[2,25] This item is essential for the prognosis of survival, especially for the elderly, since preserved ambulatory status protects against serious complications related to immobilization: infections, pressure ulcers, thromboses, denutrition… It therefore seems obvious for that criterion to be included in a future prognostic score.

Likewise, patients >80 years' old were associated with a significant difference in survival compared to other subgroups and the association remained significant in multivariate analysis. In an international multicenter series with elderly patients, the GSTSG found lower survival in the >80 years' old group than in the others.[9] It seems that this group of patients also presented a higher risk of postoperative complications (33.3%, P = 0.004), compared to the other groups and this was mainly due to an increase in wound complications or chest infections. However, despite higher complication rates, these patients still had significant improvements in their quality of life based on the EQ-5D. In a retrospective cohort study from a National Inpatient Database, Li et al[26] found a 12.17% complication rate and a 0.17% mortality rate after lumbar laminectomy, and for patients older than 85 years, these rates increased up to an 18.9% complication rate and 1.4% mortality rate, depending on co-morbidities. Recently, Yonezawa et al analyzed spine surgical resection such as en bloc spondylectomy for patients > 65 years' old in comparison to younger patients. They concluded that, even when they are selected, elderly patients undergoing curative surgical resection have a significantly higher rate of serious perioperative complications than nonelderly patients.[27]

In a similar study, Deyo et al[28] identified in the State of Washington hospital discharge registry an 18% complication rate and a 0.6% mortality rate for patients older than 75 years of age. Raffo and Lauerman reported a 35% major complication rate for patients older than 80 years who received instrumented surgery,[29] whereas Baladaud et al,[30] in a series of lumbar spine surgery in patients 80 years of age or older, reported only 13% major complication and 29.7% minor complication rates.

On the contrary, Reindl et al[31] compared the complication rates between lumbar spine surgery with or without instrumentation and total hip arthroplasty in the elderly group, and found no significant difference.

Therefore, all these data suggest that patients >80 years of age have more unpredictable outcomes and are more complicated than younger patients, even after straightforward surgery. Subsequently postoperative management should be carefully planned and anticipated.

According to the SAVES-V2 system,[32] 73 of 136 patients (53.2%) had at least minor grade 1 adverse event (AE did not require treatment and no adverse effect) and seven of 52 (13.4%) presented higher SAVES-V2 grade complications: three patients with AE grade 3 (required others emergency surgeries) and four patients with AE grade 5 (significant neural injury and Frankel grade deterioration). Grade 3 AEs were for evacuation of a postoperative hematoma or abscess. The ratios and the distribution of complications in our series according toSAVES-V2 were comparable with the series of younger oncologic patients.[33,34]

Four of 27 (14.8%) were in 70- to 75-years-old group, two of 14 (14.3%) in 75- to 80-year-old group and one of 11 (9.1%) in >80-year-old group, without significative differences (P = 0.147), implying that >80 years' old were not necessarily more fragile than other elderly subgroups. The majority of patients who received medical treatment without surgery or braces, presented SpM without associated pathological fractures.

Surprisingly, brace treatment appeared as a factor of poor prognosis with a statistically significant difference in OS compared to the two other treatment groups (symptomatic treatment and surgery). No differences were found between the three treatment groups concerning median age, initial neurological, and WHO status that could have explained this observation. However, nearly four of five of the patients in the surgery group presented an epiduritis, a complication evidently representing the surgical treatment group[16] but also synonymous to worse prognosis, which could falsely lower survival in this group. Brace treatment is often motivated by an unstable spine or unknown status and is therefore usually nursed in a supine position with bed rest.[35] In our series, this therapeutic choice was comparable for patients >80 years' old and patients 70 to 75 years' old (54%). The proportion was lower (43%) for 75 to 80 years' old (P = 0.356).

Therefore, lower survival in the brace treatment group might be explained by higher rates of medical complications such as deep vein thrombosis or pulmonary embolism, chest or urinary tract infection, and pressure ulcers, which were not analyzed in our series. Pease et al[36] also report that some patients experience a depressed mood when lying in a supine position for an extended period of time. Spine bracing with bed rest may perhaps be a starting point of sliding syndrome in this geriatric population. Further appropriate studies would be needed to confirm this data.

The role of bracing as a treatment for SpM is not well defined and the literature regarding its use is sparse.[37] In the past, braces were used, among other things, postoperatively to maintain the integrity of the spine, although this attitude is now criticized as being pointless, as this is the purpose of the surgery.[38] Some guidelines[37] suggest bed rest for threatening lesions to avoid painful movements and vertebral collapses with external stabilization by bracing when surgery is not feasible. If the spine is unstable, movements may cause an increase in pain as well as further spinal cord or nerve root damage. The idea of bracing is to achieve pain relief and prevent the risk of vertebral collapse by restricting gross body motion via stabilization of the back muscles. Until a recent publication of Gernot Lang et al, there was no evidence that braces effectively restricted spinal movements: in a study with healthy volunteers they demonstrated significantly reduced spinal motion compared with no bracing. Rigid orthosis permitted the greatest restriction of regional, intersegmental and segmental spinal motion among all ortheses.[39] In fact, we do not even know if bracing actually reduces pain or improves quality of live, and most references on spinal bracing relate to idiopathic adolescent scoliosis, osteoporotic fractures and thoracolombar traumatic fractures. Lee et al in a Cochrane review investigate the correct positioning of spinal braces and examine the effects of this therapeutic option on pain relief and the prevention of further vertebral collapse in patients with MSCC. They found no randomized RCTs on bracing in MSCC and no appropriate studies.[40]

In this prospective series, there were insufficient numbers of patients >90 years' old in order to reliably evaluate this age group, which remains a narrow group in oncology, although it is continuously growing. All the patients in our series received complementary radiation therapy.

In our study, there were 29 patients (11.9%) with hematological malignancies: (21, 72%) with multiple myeloma (MM). These etiologies of tumors remain very debatable in SpM' studies. We have already shown in a previous work that even if the survival of patients with MM and spine location was shorter than MM without SpM (59 vs. 85 months), OS was incomparably better than SpM of other primary tumors.[41] This observation was mainly due to the fact that in the MM (formerly named bone cancer), spine was not a "metastases target" in the pathological timeline, but an initial location or a starting point.[42] In addition, MM treatments have made important advances, improving significantly the OS. Furthermore, this discrepancy has already been raised for SpM survival prognosis scales which are not accurate and abilite to predict the survival of patients with spine myeloma location.[43]