SciELO - Scientific Electronic Library Online

vol.54 issue2Dexmedetomidine as adjuvant drug for wake-up test during scoliosis correction surgery: case reportPrevention of pulmonary gastric contents aspiration author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Brasileira de Anestesiologia

Print version ISSN 0034-7094On-line version ISSN 1806-907X

Rev. Bras. Anestesiol. vol.54 no.2 Campinas Mar./Apr. 2004 



Management of pain in patient with inoperable sacral tumor. Case report*


Tratamiento de dolor en paciente con tumor sacral inoperable. Relato de caso



Juliano Rodrigues Gasparini, TSA, M.D.I; Renato Ângelo Saraiva, TSA, M.D.II

IAnestesiologista do Hospital Sarah Brasília
IICoordenador de Anestesiologia do Hospital Sarah Brasília





BACKGROUND AND OBJECTIVES: Different techniques may be used to control cancer pain. This report aimed at describing some therapeutic measures used to treat a patient with a severe and bad responsive pain.
CASE REPORT: A 70-year-old male patient a sacral chordoma of very difficult therapeutic. He referred major tumor-associated pain. Several techniques used to treat pain of this patient and results obtained are reported.
CONCLUSIONS: Neurolytic block is an efficient technique to control pain in cancer patients who have no possibilities with standard treatment, especially when properly indicated and following well established routines.

Key Words: PAIN, Chronic: cancer; TERAPHEUTIC: alchoolization, ethanol


JUSTIFICATIVA Y OBJETIVOS: Diversas técnicas pueden ser utilizadas en el control de dolor en pacientes oncológicos. El objetivo de este relato es mostrar el uso de algunas medidas terapéuticas utilizadas para tratar un paciente con cuadro doloroso importante de difícil control.
RELATO DE CASO: Paciente masculino, 70 años, con cordoma sacral y con posibilidad terapéutica muy difícil. Presentaba cuadro de dolor importante asociado al tumor. Se relatan diversas técnicas utilizadas en su tratamiento y los resultados obtenidos.
CONCLUSIONES: Los bloqueos neurolíticos, para el control del dolor en paciente con tumores cuja posibilidad terapéutica es difícil, constituyen técnica eficaz cuando bien indicados y realizados dentro de criterios establecidos.




Patients with inoperable tumors very often refer pain, which may be mild, but most frequent severe and impairs life quality.

In many situations, the anesthesiologist is called to help managing pain. An integrated team is of paramount importance for the best results.

Rescuing patients' life quality and dignity should be our priority, especially in cases where the prognosis is reserved. This report aimed at describing several therapeutic measures and their results in treating a patient with major pain, very difficult to control with pharmacological approaches.



A Caucasian, 70 year-old male patient, was referred to the Oncology service of Hospital Sarah, Brasilia, in February 2001 with pain in left leg and sacral region.

Left leg pain had started in 1999 and diagnostic tests evidenced a sacral chordoma (tumor originating from notochordal remnants). Patient was submitted to surgical tumor resection and permanent colostomy to allow surgical access and bypass intestinal traffic. He underwent 45 postoperative radiation therapy sessions and had satisfactory improvement in the next months.

Pain complaints have returned though, and patient was unable to sit and would eat standing up. Pain involved sacral region and left leg, was constant but presented paroxymal crisis. There are no data about pain scores of that period. Patient would use standard analgesics, with bad results. He had urinary incontinence and there were no other relevant historical data.

Tumor recurrence was evidenced by imaging screen. Small pulmonary nodes, single liver metastasis and insufflating osteolytic injury associated to soft tissue loss involving the whole sacrum, the fifth lumbar vertebra and the left iliac bone were observed.

The case was discussed during a clinical and radiological meeting by the general surgery, orthopedics, pathology, radiology, neurosurgery and oncology teams. A conservative approach was defined for clinical support and pain control. Surgical proposal would not be curative and would significantly increase morbidity and mortality (involving hemicor- porectemy with total sacral and last lumbar vertebrae resection).

Patient was put in an outpatient regimen. An association of codeine (30 mg), paracetamol (500 mg) and carbamazepine (200 mg/day) was started to control sacral pain and paroxymal crisis.

Patient has not come for follow-up visits in the next 5 months, when he returned with worsening of symptoms. Supine position was very painful. The association of codeine/paracetamol was withdrawn and oral morphine was introduced (initially 15 mg every 4 hours, if needed).

Doses were progressively increased and in November 2001 (4 months after his return for follow-up) hospitalization was needed for pain control. By this time he was under oral morphine (30 mg every 4 hours), carbamazepine (400 mg/day) and dipirone. Pain score would vary 5 to 8, according to a 0 - 10 analog visual scale.

We decided to start 2 mg.h-1 intravenous morphine continuous infusion in PCA with 1 mg demand doses and 15-minute blockade interval without the limit of 4 hours. Daily morphine consumption was approximately 70 mg.

After some days, transdermal fentanyl was introduced, starting with 25 µg.h-1 release patch. Fentanyl dose was increased and morphine dose was decreased. At 150 µg.h-1 fentanyl, morphine infusion was withdrawn. Patient had mean pain score of 3 to 5 in analog visual scale without paroxymal crisis and was sleeping well. He was discharged 9 days after admission under 150 µg.h-1 transdermal fentanyl, 300 mg/day carbamazepine, oral dipirone, 20 mg/day tenoxicam and oral morphine (15 mg every 4 hours, if needed).

Patient remained well for some months with outpatient follow-up. Drug doses were checked and adjusted as needed.

Pain kept worsening, and in May/2002 (2 years and 3 months after surgery) he was again admitted to control it. Left leg and sacral region pain persisted, however patient also referred right leg and abdominal pain. The score varied 7 to 8 and would reach 10 during crisis. At admission, patient was under 200 µg.h-1 transdermal fentanyl, 600 mg/day carbamazepine, 40 mg/day tenoxicam, 50 mg/day amitriptyline and 30 mg oral morphine every 6 hours.

Oral morphine was replaced by intravenous morphine during hospitalization, in a total dose of 60 mg/day divided in regular doses. Even so, pain control was not satisfactory. Morphine continuous infusion was started in doses varying 3 to 9 mg.h-1 and rescue doses of 10 mg. Simultaneously to morphine dose increase, transdermal fentanyl was withdrawn.

In face of the medication pain management failure, and the onset of opioid side-effects (nausea, vomiting and intestinal constipation), neurolytic spinal block was proposed.

Patient and relatives agreed with the procedure and blockade was scheduled to August/2002 (2 years and 6 months after surgery).

Due to patient's pain, blockade was performed under general anesthesia with the patient in the prone position and pads under the hip. Spinal puncture was performed at L2-L3 interspace and 0,3 mL of absolute sterile alcohol was administered in three doses at 3-minute intervals. Patient was maintained in this position for 1 hour. Procedure was carried out uneventfully, and no complications were observed in the PACU.

There has been no improvement 5 days after blockade and another blockade was performed again under general anesthesia. Position was changed to prone position with 30º head down. Spinal puncture was performed at the same L2-L3 interspace and 0.7 mL absolute alcohol was administered in 7 doses.

Patient showed satisfactory, though short-lasting improvement (2 days). Morphine infusion was increased to 10 mg.h-1 and, even so, patient needed rescue doses. Dexametasone was introduced also with brief improvement.

Due to uncontrollable pain in changing position, patient has evolved with infected sacral bed wound. A cordotomy was considered by that time, but neurosurgery would need resolution of the bed wound and control of infection. Since this would not be easy, procedure was postponed.

Transdermal fentanyl was reintroduced up to 250 µg.h-1 with partial pain relief. Morphine could be withdrawn and maintained only as rescue dose. Patient was discharged in October/2002, still referring pain (score 6) but with no paroxymal crisis.

Seven days later, patient returned with acute cholecystitis, which was clinically treated with good results. Pain was progressively increasing (score between 8 and 10) and fentanyl had to be increased to 300 µg.h-1. Patient was also under intravenous morphine with doses above 200 mg/day.

A new blockade was then scheduled for 03/09/2002. Patient could no longer walk due to pain and tumor evolution. He was virtually paraplegic, and this time a test was performed before blockade. With the patient awaken and sedated for comfort (fentanyl bolus) blockade was simulated. Patient was placed in the left lateral position with the chest bent forward approximately 45º. Two punctures were performed, at L2-L3 and at L4-L5, for the administration of 0.25% hypobaric bupivacaine (obtained by diluting 0.5% isobaric bupivacaine in sterile distilled water) in a volume of 1 mL (L2) and 2 mL (L4). Patient reported total right leg pain relief, which was his major complaint.

Considering the success of the test blockade, neurolytic blockade was performed with absolute alcohol 24 hours later. Procedure was performed with the patient awaken in the same interspaces, with 100% sterile alcohol in the volumes previously used as local anesthesia. During alcohol administration, patient complained of intense burning at the site where he referred pain. Patient was maintained in this position for 60 minutes and no intercurrences were observed.

Three days later the patient was virtually painless (score 2) in right leg. Initial proposal was a blockade for left leg and then for the sacral region, but he was depressed and did not agree with two more blockades, even after the good result of the most recent one. He referred severe pain during alcohol administration and that he would not stand another procedure.

The importance of a new blockade was discussed with him and his relatives: right leg improvement had been significant and similar results could be obtained. Patient agreed with just one more blockade, so the proposal was to induce a blockade able to reach sacral and lumbar roots.

Blockade was performed with the patient awaken and sedated with remifentanil continuous infusion. Patient was placed in the right lateral position with the chest bent approximately 45º forward and head down 40º aiming at alcohol spread to the sacral region. Puncture was performed at L4-L5 interspace and 4 mL absolute sterile alcohol were injected. Again, patient referred severe pain at neurolytic solution injection in the region where he referred pain. Patient was maintained in this position for 30 minutes because he could no longer tolerate the head down position.

Three days after blockade patient was feeling much better, with no pain at movements and could sit for 60 minutes, being discharged under 200 µg.h-1 transdermal fentanyl associated to 200 mg/day carbamazepine and 10 mg subcutaneous morphine in case of pain. At follow up, patient was evolving well, with just 1 or 2 morphine doses.



Pain management in patients with major tumors may be a challenging. Standard pharmacological measures involve anti-inflammatory drugs and opioids. These drugs are gradually and progressively started according to pain intensity. There are several and well-known recommendations in the literature on the use of such measures.

Coadjuvant drugs also are part of the pharmacological approach. Several drugs of different pharmacological classes are used to help managing chronic or acute pain, like antidepressants, anticonvulsants, adrenergic agonists, local anesthetics, sodium channel blockers, NMDA receptor blockers, GABA receptor agonists and many others.

The oral route is preferred for the administration of such drugs, but there are other options, especially transdermal, epidural and spinal.

But sometimes, all this therapeutic armamentarium is unable to achieve an acceptable result. The excess of side-effects or even failures in providing effective analgesia place the physician in a dilemma.

Even epidural or spinal catheters have their limitations. In addition to tachyphylaxis, infections and material cost, the tumor may evolve to nervous structures. This may cause permanent compression and impair the access of local anesthetics and analgesics.

Intervention techniques are of great value for selected cases. Several aspects should be taken into consideration when indicating such procedures. The type of pain (visceral, parietal or neuropathic), pain location (single site, multiple sites, well defined area, poorly defined area) and patient's life expectancy, as well as clinical and mental conditions, are extremely important in the decision process.

Neurolytic blockades for pain control are widely used and may be performed in sympathetic chain or by epidural or spinal route. There are several studies reporting the use of neurolysis to treat abdominal tumor-induced pain (e.g.: pancreas tumor with celiac plexus alcoholization).

Few cases report the use of epidural neurolytic blocks to treat tumor-induced pain 1. Results are satisfactory, but short-lasting. Very often several blockades are required to obtain adequate analgesia 1,2.

Even less cases are reported involving spinal neurolysis 3. Dogliotti has reported the first chemical spinal neurolysis with alcohol in 1931. With pharmaceutical advances to obtain increasingly potent drugs for pain control, this technique has been reserved for selected cases.

The most frequent and accepted indications for this procedure are cancer pain which tends to perpetuate, life expectancy between 6 and 12 months, failure of analgesic regimens or unacceptable side-effects, well localized somatic pain, pain affecting 2 to 3 dermatomes and evidence of response to spinal blocks with local anesthetics 4.

Counterindications are: mental status change unrelated to pain or analgesic drugs, neuropathic pain; sympathetic nervous system-maintained pain, poorly localized pain, widely distributed pain; tumor at procedure site and spinal space totally taken by tumor.

Most frequent complications are loss of muscle strength 5,6 by anterior roots destruction and sphincter dysfunction, which may be transient or permanent. The incidence of sus complications persisting for more than 4 months is approximately 12%.

Response to blockade is positive in 70% of cases, but there are no exact data for pain relief duration. A reliable evaluation of blockade response can only, be mode after 3 to 5 days. In general, it lasts from 2 to 4 months.

There are several solutions that can be employed for neurolysis, most commonly ethanol and phenol, yet cooled hypertonic or isotonic saline solutions may be used. Patients' position should place posterior roots toward neurolytic solution spread. Alcohol is hypobaric as compared to CSF and patient should be positioned according to figure 1. Volumes vary within a wide range: some authors recommend very low doses (0.1 to 0.3 mL), while others suggest greater volumes (1 to 4 mL). However, mean 2 mL doses seem to be the most widely used.

Our case is a good example of all that was said above. One can observe patient's evolution and his complaints, as well as the steps for his support treatment. There were 3 different moments in his clinical history. Initially the pain was easily controlled with simple pharmacological measures. A second moment came when those measures had to be intensified (dose increase, replacement of drugs by more potent ones), but still showing satisfactory results. Finally the drugs (even in high doses) were no longer effective, and/or had major side-effects.

Patient was treated in progressive stages. In the early stage of the disease, just a combination of non steroid anti-inflammatory drug (paracetamol), a weak opioid (codeine) and an anticonvulsant (carbamazepine) was enough to control pain for 5 months. During this stage, patient could remain at home, walk and perform his activities with minor restrictions.

But with disease and complaints progression, analgesic measures were intensified. Doses were increased and weaker drugs were replaced by more potent ones aiming at pain control. For a certain period, these measures were enough to maintain patient at home with slightly higher restrictions in his activities. Some hospitalization periods were necessary for faster dose adjustments. These periods were as short as possible, at patient's request. The disease, however, progressed and conventional measures became ineffective. Opioid doses were high, with poor results. Adjuvant drugs were added to regimens with brief responses and side-effects.

Neurolytic blockade for this patient seemed adequate. Tumor would increase in size and survival would not be much longer than one year. When the option was presented, patient had no longer good response to high opioid doses, could no longer walk, had colostomy and urinary incontinence. Most feared blockade complications were already part of his clinical presentation.

The first two attempts (with frustrated results) were not performed according to the recommended technique. Patient's position would not place dorsal roots in the proper position. Patient was anesthetized due to severe pain at blockade position) and a low volume was used. The combination of these factors may explain inadequate results.

At third attempt, we have followed the recommended technique. Patient was awaken in a good position on the table. A previous test blockade was performed with local anesthetics with complete response. Two punctures were performed to reach involved dermatomes, and a higher volume was used. The results was satisfactory and in line with the literature.

The fourth blockade has not followed the same technical rigor because it had to be more encompassing. Patient would not allow two more blockades. Scheduling included one alcohol blockade aiming at left lumbar roots and another one with phenol aiming at sacral roots with phenol. Considering an adequate solution for the patient, we have changed the technique trying to obtain both results with just one blockade. Position was basically maintained, but the table was tilted to head down (Figure 2), so as alcohol could reach sacral roots. Results were favorable, even with the change in technique.

After 4 blockades, patient had improved, could change position, sit (even if for a short period) and could sleep. Patient was discharged and regained quality of life.

Five months after discharge he is still painless and with no drug side-effects. He is under transdermal fentanyl (300 µg.h-1) and subcutaneous morphine for crisis (10 to 20 mg/day, occasionally).

Neurolytic blockades are an option for pain of difficult control in cancer patients. Criteria (indications) and adequate technique (including patient's position, roots to be blocked, neurolytic solution volumes) should be observed to achieve good results.



01. Hongo T, Tsunoda K, Egami Y et al - Efficacy of epidural neurolysis. Masui, 1995;44:1537-1541.        [ Links ]

02. Hayashi I, Odashiro M, Sasaki Y - Two cases of epidural neurolysis using ethyl alcohol and histopathologic chances in the spinal cord. Masui, 2000;49:877-880.        [ Links ]

03. Patt RB, Payne R, Farhat GA et al - Subarachnoid neurolytic block under general anesthesia in a 3-year-old with neuroblastoma. Clin J Pain, 1995;11:143-146.        [ Links ]

04. Patt RB, Reddy S - Spinal neurolysis for cancer pain: indications and recent results. Ann Acad Med Singapore, 1994;23:216-220.        [ Links ]

05. Garcia E, Patt RB, Ginsberg LE - Effective but near disastrous subarachnoid neurolysis for mechanical root pain in a paraplegic cancer patient. Clin J Pain, 1996;12:243-249.        [ Links ]

06. McGarvey ML, Ferrante FM, Patel RS et al - Irreversible spinal cord injury as a complication of subarachnoid ethanol neurolysis. Neurology, 2000;54:1522-1524.        [ Links ]



Correspondence to
Dr. Juliano Rodrigues Gasparini
SMHS Quadra 501 - Conjunto A - Hospital SARAH
70330-150 Brasilia, DF

Apresentado (Submitted) em 9 de abril de 2003
Aceito (Accepted) para publicação em 01 de julho de 2003



* Recebido do (Received from) Hospital Sarah Brasília, Brasilia, DF

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License