Introduction

The management of cancer pain is complex. While pain is present at the beginning of the disease in only 30% of cases, it affects more than 80% of patients in a terminal phase. Pain is particularly frequent in connection with bone metastases, which represent the third most common metastatic site after the liver and lungs. Approximately 20% of cancer patients will present with bone metastases during their evolution, most of the time painful and with variable clinical expressions [1]. These bone metastases can be responsible not only for pain but also for fractures, in particular the long bones, and medullary compressions during vertebral involvements affecting the functional prognosis of the patient. For several years, the capabilities of antalgic drug therapy have increased with the arrival of new opioids, but these treatments can have major side effects detrimental to the quality of life. Progress in the systemic treatment of metastatic disease in general, as well as radiotherapy (external radiotherapy and metabolic radiotherapy [2]), is contributing to advances in the management of bone metastases [3]. Thus, the metastatic disease can become a “chronic” disease, while the important issue is to provide an improvement of survival, with a good quality of life. With this main objective in view, interventional radiology has developed over the last few years, occupying an increasingly important role in the management of bone metastases [4]. We report here a trial of bone cementoplasty in a series of 42 patients treated over a period of 30 months at the Center.

Patients and method

We report here a retrospective study including patients at the Eugene Marquis Comprehensive Cancer Center of Rennes, who benefited from cementoplasty for bone metastasis between January 2007 and July 2009. Over this period, 42 cementoplasty procedures were carried out. In the majority of the cases (35/42 cases, corresponding to 83%), the indication of cementoplasty was submitted for consideration by a multi-disciplinary team (MDT) meeting on “bone cancer pain and complex cancer pain,” bringing together an interventional radiologist, anesthetist, nuclear medicine specialist, radiotherapist, neurosurgeon, and an algologist-cancer specialist. This MDT was set up in March 2007 at the CRLCC Eugene Marquis.

The cases were analyzed retrospectively, but a routine treatment assessment sheet was recorded (in 12 cases, that is 28%) since October 2008, with a regular follow-up over at least 1 month, which comprises a telephone follow-up by a nurse specialized in algology and a consultation at 1 month by the referral doctor or the algologist.

The median age of the population was 59 years (ranging from 21 to 81 years); there were 24 women (57%) and 18 men.

The patients all had an advanced disease. The performance status (PS) was very impaired (PS 3 or 4) for 11 patients, that is, in two out of 16 patients (38%), and in 15 the general state was preserved (PS = 0 for 2 patients and PS = 1 for 13). While 16 patients (38%) were at the stage of palliative care, 26 were still receiving a “palliative” chemotherapy at the time of the cementoplasty treatment.

At the date of analysis of the study (31 August 2009), 17 patients (40.5%) had died. The maximum follow-up was 27 months, with respect to the first operation carried out in April 2007.

The primary sites of the cancer were: central (15 patients, that is 36%), colon and kidney (5 cases of each localization, or 12%), liver (3 patients, or 7%), then lungs and uterus (2 of each or 5%), and finally 1 melanoma, 1 bilio-pancreatic cancer, 1 urothelial carcinoma, 1 prostate cancer, 1 rectal cancer, 1 soft-tissue sarcoma, 1 Ewing sarcoma, 1 thyroid cancer, 1 cancer of the bladder, and 1 cancer of the esophagus.

The cementoplasty sites were as follows: long bone in ten cases, vertebrae (dorsal or lumbar, never cervical) in 13 cases and metastases of the pelvis in 20 cases. For 22 patients, walking was prohibited or impossible (pain).

Pain is always the principal indication for cementoplasty, with the aim of consolidation when metastases affect the spinal column or long bones.

There were 41 osteolytic metastases (or mixed osteolytic-osteosclerotic, with predominance of lytic lesions), and one osteosclerotic metastasis (prostate cancer).

Method of performing the cementoplasty

Most of the patients (30, corresponding to 71%) received the treatment under general anesthesia, while the others underwent a local epidural anesthesia or epidural block (1 patient with hypnosis). No patient had cementoplasty without anesthesia. All the treatment acts were performed by the same interventional radiologist.

For lesions of the spinal column (except the sacrum) and long bones, the guidance and injection were carried out under scopic monitoring, while, for lesions of the sacrum and wings of ilium, we made use of a guidance scanner with injection under scopic monitoring.

The needles used had a gage of 11 G, and a bilateral transpedicular approach was adopted for involvements of the spinal column. For the long and flat bones, one or two needles were used according to the size of the lesion. Cement was prepared at the time of application, and injected systematically using a high-pressure injector (radio-opaque bone cement OBC100, Osteo-Firm (COOK)). The duration of injection did not exceed 10 min, while the volume injected varied from 2 to 10 cc and the duration of the treatment act from 45 to 90 min.

The majority of patients (36 cases, i.e., 86%) received a preliminary radiotherapy; 11 times at more than 3 months before the treatment act and 25 times in the weeks preceding or following the cementoplasty. This irradiation was delivered briefly (8 Gy in 1 session for 7 patients; 20 Gy in 5 sessions for 8 patients; and 30 Gy in 10 sessions for 10 patients).

The total tolerance of the treatment act was assessed peri-operatively, on regaining consciousness and then again at D1.

The efficacy was assessed in terms of three classes: “failure,” when the pain was not alleviated, “partial” if the patient was relieved but only moderately or temporarily, and “good” if the following conditions were met:

  • Reduction of more than 10% of the dose of opioids in the days immediately after the treatment act or during the following month;

  • Resumption of walking in the month following the treatment act, if this had been previously impossible (risk of fracture) or difficult (intense pain);

  • Reduction in pain at rest and during effort of more than one point on the VAS (Analogical Visual Analog Scale);

  • Concordant opinion of the patient and the doctor (or of the ancillary medical team).

Statistics

Statistical analysis of the data was carried out using version 9.1 of the program SAS (Statistical Analysis System). The quantitative results are expressed in median and extreme values. Student’s t test, the Fisher-Yates test [exact chi-squared test], and Wilcoxon’s test were applied. We used Student’s t test for paired series to follow the post-treatment evolution of certain quantitative variables (pain scored on VAS, dose of morphine). A value of P = 0.05 was taken as the threshold for statistical significance.

Results

The median duration of hospitalization is 2 days (from 0.5 to 43 days; reflecting more the context of palliative care that the treatment act itself).

The intensity of the pain decreases significantly between the maximum observed the day before treatment and the maximum observed during the month following treatment (P = 0.04) in the 25 patients included in the survey; it is probable that those patients whose pain intensity had not been reassessed systematically presented with little or even no pain at all after the treatment. The morphine oral equivalent dose is on average almost identical before the treatment and at the end (256 and 254 mg, respectively), and a little weaker (217 mg) at 1 month. Using paired data for 35 patients yielding the same morphine equivalent doses before treatment and 1 month after, we find that there is a non significant (P = 0.11) tendency to a reduction in the dose of analgesics.

In terms of efficacy, we note 57% good results (24 patients), 9.5% failures (4 patients), and 31% partial results (13 patients), corresponding to a total of 88% “partial or good” results. For one patient, the lack of data in the records prevents us from drawing any conclusions. In the days following the treatment act (generally the next day and over a period ranging up to 1 month), 16 patients out of 22 (73%) could resume walking (with or without assistance) (Fig. 1a, b), whereas, prior to treatment, they were unable to walk because of pain or the risk of fracture.

Fig. 1
figure 1

a Lytic metastasis of upper 1/3 of left humerus head, with fracture. b Same patient as in a after cementoplasty: Disappearance of pain

Full size image

The tolerance to treatment is excellent in peri-operative and immediate post-operative stages. At D1, we observe a temporary increase in pain for 6 patients and one case of iatrogenic radiculalgia by direct contusion of the root during the transpedicle puncture carried out under general anesthesia. We note moderate cement leakages into the soft tissues (8 cases) at the level of the intervertebral disks (2 cases) and into the epidural region in one case (Fig. 2a, b).

Fig. 2
figure 2

a Lytic metastasis of an internal malleolus preventing walking. b Same patient as in Fig. 1a after cementoplasty: resumption of walking possible

Full size image

By comparing the various pre-therapeutic data with the overall assessment of the treatment, we do not find any statistically significant correlation, except as regards the anesthesia: cementoplasty of the flat bones was less often carried out under general anesthesia (P = 0.0016) than for long bones or vertebrae. Full efficacy appears less frequent (P = 0.07) when cementoplasty is applied to flat bones. We do not observe any differences in results between the cementoplasty treatments carried out according to bone localization.

Table 1 summarizes the main data for each group of patients separated into: cementoplasty for metastases of long bones, spinal column, and flat bones (pelvis and scapula).

Table 1 Distribution of patient cementoplasty by type of bone
Full size table

Discussion

All the cementoplasty procedures carried out in this study were in compliance with the indications recently summarized by Kastler et al. [4]: zone of very wide and/or major bone destruction for which no surgical treatment is possible; osteolytic lesions of bone parts for which surgery would be too damaging and, in particular, lesions of the pelvis and sacrum; zones of bone destruction with high risk of fracture, especially if located in a weight-bearing zone likely to compromise the solidity and stability of part of the skeleton, or upright station; lesions of the long bones with important functional consequences, such as the tips of the lower limbs compromising upright station and walking; osteolytic lesions unattainable by radio or chemotherapy.

Concerning the improvement of symptoms, we find 57% of cases with full efficacy and 88% with good or excellent results.

The results for the pain come very shortly as we can see for the case described by Wong [5]: the patient was able to stand and walk just after the cementoplasty. Sometimes, the result can come later [6] but always improve quality of life.

A review of the literature [4] reveals the disappearance or very important reduction of pain in 70% of the patients affected by bone metastases or myeloma. This early antalgic effect, between the 16th and the 72nd hour, leads in the majority of cases to a resumption of the upright station and contributes to limiting the complications of decubitus. These results are confirmed in this study, since we find 16 of the 22 patients (73%) who previously could no longer walk were able to resume walking a few hours after the treatment act, with much simpler sequels than with surgery in this population of tired immunodepressed patients. The benefit of cementoplasty is particularly clear (without the difference reaching the level of statistical significance) for involvements of the long bones (70% full and 30% partial efficacy, that is, 100% overall efficacy) and of the spinal column (75% full and 17% partial efficacy, that is, 92% overall efficacy) compared to flat bones (42% full and 42% partial efficacy, that is, 84% overall efficacy). In 2008, an Italian team [7] reported an experiment concerning 70 cementoplasty procedures performed on 50 patients with lytic metastases of the flat bones and long bones (except spinal metastases): the excellent antalgic result was not improved by the radiotherapy carried out in addition to the cementoplasty in 7 patients.

Two patients presented with fractures 1 month after their cementoplasty for femoral metastasis. This raises the question of the role of the cementoplasty of long bones such as the femur compared to surgery among patients in excellent condition.

Finally, a recent large series [8] included a population of 136 patients that had benefited from cementoplasty, but only 19 of these patients received this treatment for a metastasis (the others were treated for myeloma, osteoporosis, or post-traumatic lesions). Sixteen of these 19 patients describe an immediate beneficial effect in the 24–48 h following the treatment act, with the effect being prolonged in time with a median follow-up of 12 months (3–38 month). These authors (op. cit.) consider that, for optimum results, vertebroplasty should ideally be carried out before the radiotherapy. They suggest that it could be regarded as the first line of treatment of pain in vertebral metastases, before invasive surgery. This contrasts with other teams who propose this treatment for patients with short life expectancy [9].

In the study, the majority of the metastases were osteolytic or of mixed osteosclerotic–osteolytic nature, while only one case was osteosclerotic with a partial result. While vertebroplasty is also effective in osteosclerotic metastases (or of mixed type), the technique is more problematic and local complications are more frequent than when dealing with osteolytic lesions [10]. This also explains why we find cementoplasty has a good efficacy in myeloma, with a complete interruption of opioid medication for 59.3% of patients in a series of 28 [11].

The main complication of vertebroplasty is represented by cement leakage outside the vertebral body [4]. In a series of 50 patients, Anselmetti et al. [7] report 10% of leakage into the soft tissue, which in general is perfectly well tolerated except when it occurs at the level of the foramen, inducing a radicular compression sometimes requiring a surgical correction. Other complications are possible [10], such as the reflux of cement into the segmentary arteries and even the aorta, or pulmonary embolism. On the whole, these complications are rare (1–5% of cases, according to the series) [12]. The frequency of complications decreases with the experience of the operator. The 11 local leakages of cement found in the study were all without clinical consequence.

In the study, a temporary increase in pain was noted for six patients having received a large injection of cement (>5 cc), probably due to intense cytotoxic effect linked with the marked local heating during polymerization.

Radiotherapy was carried out for 36 patients (86%) in the months or days preceding or following the cementoplasty. The antalgic effect of the radiotherapy is often delayed in time (2–6 weeks), while its bone reconstruction capacity is more limited and occurs even later (approximately 3 months) [1]. For a patient receiving cementoplasty, it is recommended to associate the treatment act with radiotherapy [12]. It is evidently difficult to weigh up matters regarding antalgic efficacy when the two methods are associated together (initially cementoplasty and then radiotherapy). There are no studies in the literature comparing the two techniques. A French randomized test (EVAR) was launched in 2005 [13] to assess the contribution of vertebroplasty associated with radiotherapy compared to radiotherapy alone, but was unfortunately closed because of the low rate of inclusion.

The study suffers from its retrospective nature and the lack of assessment data available in the case records.

To conclude, in the retrospective series of 42 patients, cementoplasty appears to offer an interesting contribution to improving the quality of life of certain patients affected by bone metastases. Indeed, it allows a partial or complete improvement of symptomatology in 88% of cases, with a resumption of walking in nearly three-quarters of the patients who previously could no longer walk. The complications are limited to a temporary increase in pain following the injection of more than 5 cc of cement and some asymptomatic cement leakage into soft tissues.

The complexity of implementing the technique is a factor limiting its wider application, explaining the small number of patients who can benefit from the treatment [8]. This makes it even more difficult to assess its efficacy and develop the recommendations, especially since no controlled comparative studies are available. The setting up of this treatment technique will require human and financial resources, but a small number of teams organized around a “multidisciplinary team meeting dedicated to palliation of bone metastasis” forming a network over the territory should nevertheless be sufficient as long as the indications are not very frequent.