Palliation of Soft Tissue Cancer Pain With Radiofrequency Ablation

Abstract

The purpose of this study was to analyze the feasibility, safety, and efficacy of radiofrequency ablation (RFA) to treat pain from soft tissue neoplasms. RFA was performed on 15 painful soft tissue tumors in 14 patients. Tumors varied in histology and location and ranged in size from 2 to 20 cm. Patient pain was assessed using the Brief Pain Inventory (BPI) at baseline and 1 day, 1 week, 1 month, and 3 months post RFA. All patients had unresectable tumors or were poor operative candidates whose pain was poorly controlled by conventional treatment methods. BPI scores were divided into two categories: pain severity and interference of pain. Although not all scores were statistically significant, all mean scores trended down with increased time post ablation. Based on these outcomes, RFA appears to be a low-risk and well-tolerated procedure for pain palliation in patients with unresectable, painful soft tissue neoplasms. RFA is effective for short-term local pain control and may provide another option for failed chemotherapy or radiation therapy in patients with cancer. However, pain may transiently worsen, and relief is often temporary.

Pain affects nearly every aspect of the life of a patient with cancer. Although a variety of palliative treatments are available, most have unpleasant side effects and/or limited effectiveness. Side effects from opioid and non-opioid analgesics may become intolerable. Additionally, palliative therapies to decrease tumor burden may be inadequate due to limitations of radiation doses, chemotherapy toxicities, or the patient's refusal or intolerance of additional surgical procedures. Cancer patients need other therapeutic options to maximize pain control and reduce suffering.

Multiple, minimally invasive, neurodestructive techniques have been safely applied for pain control, including radiofrequency lesioning, cryoanalgesia, and chemical neurolysis with agents such as phenol, alcohol, and hypertonic saline [1]. This same neurolytic principle may be the mechanism of pain relief in patients with bone pain from chordoma, osteoid osteoma, or osseous metastases treated with radiofrequency ablation (RFA). RFA of nerve ganglia has proven to be effective in the treatment of multiple pain syndromes, including trigeminal neuralgia, cluster headaches, chronic segmental thoracic pain, cervicobrachialgia, and plantar fasciitis. In addition, RFA has been used for inflammatory, idiopathic, and tumor-related pain [2-7].

RFA may provide a means for the palliation of cancer-related pain unresponsive to other therapies [8, 9]. RFA has been used effectively for over 10 years to treat benign bone tumors, such as osteoid osteomas [10]. Recently, RFA has been used successfully in the treatment of painful metastatic bone disease [11-15]. Although recent reports demonstrate safety and efficacy for bone cancer pain, similar results for soft tissue pain are not available. However, anecdotal evidence of dramatic palliation exists for certain patients with soft tissue cancer pain who receive RFA [16]. This rapid response could lessen the use of narcotics in palliative care and provide another non-sedating option for cancer patients. The goal of this study was to evaluate the short-term efficacy of RFA in relationship to pain and functional impact (as measured by the Brief Pain Inventory [BPI]) as a treatment for soft tissue tumor pain.

Materials and Methods

This study was a single-arm, retrospective, paired-comparison study with subjects serving as their own controls. The study group consisted of patients referred from two cancer center clinics. Patients who met the eligibility criteria for inclusion in this study were those who had unresectable, painful, soft tissue neoplasms recalcitrant to conventional pain therapies; those who were poor candidates for surgery; those who had maximized conventional methods of palliation; and/or those who had experienced dose-limiting side effects from pain medications. Pain was localized to the target tumor in all patients by physical assessment and imaging review by the interventional radiologist and the primary care physician, as well as the palliative care service physicians whenever possible. Patients excluded from this study were those who had uncorrectable coagulopathies or pain not localizable to dominant tumor sites, as well as patients who had changed the class of pain medications they had been taking or the cancer treatment regimen they were receiving within 2 weeks prior to planned RFA.

All patients underwent routine clinical laboratory and imaging evaluation prior to RFA. RFA of soft tissue is approved by the US Food and Drug Administration (FDA), and 11 of 14 patients underwent percutaneous RFA as standard treatment for tissue ablation, with the BPI used as a follow-up pain assessment tool. The remaining 3 patients were treated on a formal investigational review board-approved protocol.

From June 1999 to June 2003, percutaneous image-guided RFA was performed on 15 painful soft tissue tumors in 14 patients who returned a completed BPI. One patient had 2 tumors treated with RFA, of which only 1 was studied so that we could consider each treatment independently for a more accurate analysis. Tumor histology included breast carcinoma, colon carcinoma, chordoma, melanoma, lymphoma, prostate carcinoma, renal cell carcinoma, ovarian carcinoma, and adenocarcinoma of unknown origin. Tumor locations included chest wall, pelvic, breast, perirectal, renal, aorto/caval, retroperitoneal, and superficial soft tissues. Mean tumor size was calculated from the averaged maximum axial orthogonal distances for each tumor and was 6 cm, with a range of 2–20 cm.

RFA PROCEDURE

During the RFA procedure, standard guidelines for general anesthesia (4 patients) or conscious sedation (10 patients) were followed. One to four adhesive grounding pads (dispersive electrodes) were placed on the patients' thighs. The grounding pads were connected to the radiofrequency generator to complete the electrical circuit. The skin in the area of interest was cleansed with povidone iodine or chlorhexidine to provide a sterile site for probe placement. Lidocaine (1%) was then injected subcutaneously before placement of a monopolar RFA probe. Ultrasonography and/or computerized tomography (CT) was used to guide placement of the RFA probe through the skin and directly into the lesion. The probe was connected to the radiofrequency generator to complete the electrical circuit, and current was adjusted according to temperature or impedance, depending upon which RFA system was used.

Twelve tumors were treated with a Radionics Cool-tip™ 200-watt generator (Radionics, Inc., Burlington, Mass) in manual or automatic impedance control mode. Treatment typically involved a series of 12-minute ablation sphere sessions, with the choice of single or triple parallel “cluster” needle probes made according to desired treatment volumes and estimated risk to adjacent anatomy. Two tumors were treated with a RITA Model 500, 50-watt generator with a Model 70 probe (RITA Medical Systems, Inc., Mountain View, Calif).

Depending on the size of the tumor, the RFA probe was left in place for 10–12 minutes according to the manufacturer's guidelines. For larger lesions, the probe was repositioned subsequently to treat more of the tumor, avoiding adjacent heat-sensitive structures (eg, bowel and major nerves). The needle track was cauterized upon removal to decrease the risk of bleeding and tumor seeding.

PAIN INVENTORY DATA ANALYSIS

In the clinic or 1 day prior to RFA, patients were given self-addressed envelopes and copies of the BPI with dates filled out for completing the baseline pre-RFA questionnaire and again at 1 day, 1 week, 1 month, and 3 months post RFA.

The pre-RFA BPI was completed the day before the procedure. No coaching or training was given on how to interpret the BPI. Patients were not given copies of prior score sheets for reference. One patient had two tumors treated sequentially; only the second tumor was included in the analysis because complete 3-month follow-up data were not available for the first tumor.

The BPI has been shown to be a valid and reliable measure of cancer pain [17, 18]. The BPI scores were divided into two categories: pain severity and interference of pain with daily living activities and quality of life [17, 18]. Within each category, the ratings were combined to give a composite score [18]. Severity scores were calculated by averaging the scores from questions 3 through 6 on the BPI (worst pain, least pain, average pain, and pain right now) for each patient at each measurement interval. Interference scores were estimated by averaging the scores from questions 9a through 9g on the BPI (pain interference with general activity, mood, walking ability, normal work, relations with other people, sleep, and enjoyment of life) for each patient at each measurement interval.

The Wilcoxon signed-rank test was used to test for a statistically significant change in pain severity and/or interference from the pre-RFA baseline level at each measurement interval. A Pvalue < 0.05 was considered to be a statistically significant improvement from baseline. The mean, standard error of the mean (SEM) and median pain severity and interference scores were then calculated, as well as the mean and median score of each question on the BPI at each time point.

Results

The data showed statistically significant improvement in pain severity from the baseline, pre-RFA scores at 1 week post RFA (P = 0.0077, n = 13), and borderline significance at 3 months post RFA (P = 0.0568, n = 5) (Figure 1). For pain interference, a significant decrease from the pre-RFA scores was seen at 1 week post RFA (P = 0.0252, n = 13) and a borderline significant decrease at 1 day (P = 0.0546, n = 13) and 1 month post RFA (P = 0.0593, n = 10) (Figure 2). Although differences in scores for individual questions on the BPI were not significant, there was a clear numerical trend toward lower mean scores on all questions with increased time post procedure (Table 1).

Figure 1. Reduction in Pain Severity Score Post Procedure.

Mean and median BPI pain severity scores over time; error bars represent the standard error of the mean. P values are for Wilcoxon signed-rank tests of the null hypothesis that the difference in the current time period minus the baseline value is equal to zero.

Figure 2. Reduction in Pain Interference Post Procedure.

Mean and median BPI pain interference scores over time; error bars represent the standard error of the mean. P values are for Wilcoxon signed-rank tests of the null hypothesis that the difference in the current time period minus the baseline value is equal to zero.

Table 1.

Decline in Mean and Median (Range) BPI Pain Scores From Baseline to 3 Months Post Ablation

BPI QUESTION		BASELINE	1 DAY	1 WEEK	1 MONTH	3 MONTHS
Severity questions
3	Worst pain	5.7, 5.5 (2–10)	5.9, 7.0 (2–10)	3.8, 4.0 (1–7)	3.2, 3.0 (0–8)	2.5, 3.0 (0–4)
4	Least pain	2.6, 2.5 (0–5)	2.1, 2.0 (0–4)	2.1, 1.4 (0–5)	1.7, 1.5 (0–6)	0.6, 0.0 (0–2)
5	Average pain	4.8, 5.0 (2–10)	3.8, 4.0 (1–5)	2.7, 3.0 (1–5)	2.5, 2.5 (0–6)	1.2, 1.0 (0–3)
6	Pain right now	3.8, 3.5 (0–6)	2.5, 3.0 (0–4)	2.6, 3.0 (1–4)	2.7, 2.5 (0–6)	1.2, 1.0 (0–3)
Interference questions
9a	General activity	4.7, 5.0 (0–10)	4.2, 3.0 (2–9)	3.4, 3.0 (0–10)	2.3, 1.5 (0–6)	1.0, 1.0 (0–2)
9b	Mood	4.6, 5.0 (0–10)	2.6, 2.0 (0–6)	1.9, 1.0 (0–6)	2.3, 2.0 (0–8)	0.7, 0.5 (0–2)
9c	Walking ability	4.3, 5.0 (0–10)	2.9, 2.5 (0–6)	3.5, 2.0 (0–9)	2.1, 2.0 (0–7)	0.5, 0.0 (0–2)
9d	Normal work	5.4, 6.0 (0–10)	3.5, 3.0 (0–10)	3.7, 2.0 (0–10)	2.6, 1.5 (0–7)	0.8, 0.5 (0–2)
9e	Relations with other people	4.1, 2.0 (0–10)	2.6, 2.0 (0–10)	1.8, 0.0 (0–5)	1.6, 1.0 (0–5)	0.3, 0.0 (0–1)
9f	Sleep	4.7, 5.0 (0–10)	3.1, 2.5 (0–10)	1.8, 2.0 (0–7)	1.8, 1.0 (0–5)	1.5, 1.0 (0–1)
9g	Enjoyment of life	5.1, 5.0 (0–10)	3.4, 2.5 (0–9)	2.8, 2.0 (0–7)	2.0, 1.5 (0–6)	1.2, 1.0 (0–4)

The number of patients responding at each time interval declined as time after RFA passed. Not all patients answered all questions at each time point. A decrease in pain severity was reported by 69% (9/13) of patients at 1 day post RFA, by 77% (10/13) at 1 week, by 70% (7/10) at 1 month, and by 80% (4/5) at 3 months . A decrease in pain interference was reported by 85% (11/13) of patients at 1 day post RFA, by 77% (10/13) at 1 week, by 80% (8/10) at 1 month, and by 100% (5/5) at 3 months.

Some patients received additional benefits from RFA aside from pain control. For example, a 70-year-old man with melanoma that had metastasized to his kidneys, resulting in pain, hematuria, and severe renal insufficiency, was treated with RFA as a nephron-sparing, palliative maneuver (Figure 3). The procedure improved the patient's pain and completely eliminated hematuria without further compromising renal function until he died 7 months later. He required patient-controlled analgesia (morphine) prior to RFA, was rapidly tapered to oral opioids during the first week after RFA, and required no pain medication from several weeks post RFA until his death.

Figure 3. Palliation of Kidney Metastasis Post RFA.

Left: Coronal reconstruction from enhanced computed tomography (CT) scan pretreatment with a 9 cm × 7 cm metastatic melanoma lesion to the right kidney (arrow). Middle: Axial slice of non-contrast CT scan immediately following RFA shows gas within the treated area (arrow). Right: T-1 weighted contrast-enhanced coronal MRI scan 2 months post RFA. Area without enhancement (arrow) shows devascularized necrosis of the treated lesion.

As another example, a 52-year-old woman had painful metastatic ovarian carcinoma that had spread to her pelvis, causing dyspareunia, urinary frequency, and severe pain recalcitrant to conventional pharmacologic and radiation treatment (Figure 4). She received complete pain relief between 1 week and 1 month post RFA and was able to discontinue taking opioids, which were causing sedation and impairing her quality of life, until her death 8 months later.

Figure 4. Palliation of Pelvic Metastasis Post RFA.

Left: Contrast-enhanced computed tomography (CT) scan of painful metastatic ovarian carcinoma in the pelvis, causing dyspareunia, urinary frequency, and severe pain recalcitrant to conventional treatment. Middle: Unenhanced CT scan shows the RFA probes (arrow) in the tumor during treatment. Right: Enhanced CT scan 6 months post RFA shows interval shrinkage and loss of enhancement throughout tumor, except for enhancing nodule adjacent to the rectum (arrows).

Complications from RFA were minor or insignificant except in one patient who underwent RFA of a superficial subcutaneous abdominal tumor adjacent to a colostomy. Skin breakdown occurred, and the ablated tumor became infected 2 weeks post ablation and required surgical resection. No other major complications were observed.

Typically, patients were ready for discharge in less than 24 hours after the procedure; however, several patients required additional hospitalization unrelated to RFA. All patients were discharged home from 6 hours to 7 days later.

Discussion

Recent interest in minimally invasive, percutaneous therapies has resulted in less invasive and potentially safer methods of treating difficult to manage disease sites. RFA is approved for ablation of soft tissue, unresectable liver tumors, and painful bone metastases and has emerged as the safest and most predictable technology for thermal ablation in the bone, liver, heart, nerve ganglia, breast, lymph nodes, and soft tissue. Although the FDA cleared RFA for “soft tissue” ablation, it is not specifically approved for painful soft tissue tumors.

In our experience, procedure-related pain was better controlled with general anesthesia than with conscious sedation. The procedure and the immediate hours after the procedure may transiently increase the worst pain levels; therefore, aggressive post-procedural analgesia is important in the hours following RFA. Admittedly, this step could be a confounding factor, but short-term analgesia should not result in more lasting confounding effects. In this study, the most statistically significant decreases in pain severity and interference of pain were at 1 week post ablation. At 1 day post RFA, patients experience post-procedural discomfort at the probe insertion site or at the actual thermal lesion area. The manifestation of major pain relief from RFA 24 hours after treatment is not uncommon. In addition, based on anecdotal evidence, certain patients who receive RFA may experience dramatic palliation following the procedure [16].

Recent reports using RFA for palliation of bone neoplasms show promise for this indication. Callstrom et al [14] performed RFA on 12 adult patients with a single painful osteolytic metastasis, for whom conventional palliative methods were unsuccessful. Of these 12 patients, 10 (83%) reported that they had 100% relief of pain at some point after RFA. In addition, the results of a multicenter study of RFA to treat painful metastases involving bone showed that 95% (41/43) of patients experienced a statistically significant decrease in pain at 1 and 4 weeks post RFA [15].

In the largest multicenter study on complications of RFA of liver tumors, Livraghi et al [19] reported that less than 4.7% of patients experienced post-procedural discomfort, and much of the discomfort was associated with superficial lesions or lesions near the diaphragm. This figure underestimates procedural pain when RFA is used for extrahepatic palliation. Post-RFA discomfort may be managed with a combination of patient-controlled analgesia, use of a non-steroidal anti-inflammatory agent (eg, ketorolac), and other oral analgesics. Typically, these peri-procedural medications can be rapidly tapered in the hours following RFA.

Recent advancements in thermal ablation technology allow more effective treatments of larger volumes of soft tissue [20-24]. These treatments result in a larger volume of tissue death, or coagulation necrosis, and allow for potential clinical applications that were not previously feasible, including treatment of larger, painful soft tissue lesions. The technical advances most clinically useful are coaxial expandable electrodes and electrodes perfused with cooled saline. These simple advancements and larger volumes are bringing RFA into the minimally invasive oncologic arsenal, enabling treatment of large, painful tumors with minimal risk.

A multidisciplinary team approach should be employed when considering RFA for palliation. Consultation between the interventional radiologist, medical and/or surgical oncologist, and palliative care service members is ideal.

A general downward trend was observed over time in all responses on the BPI (Table 1). There may have been non-significant P values because fewer patients returned the questionnaire, resulting in a smaller sample size as time passed. The general trend toward fewer replies with the passage of time post RFA weakens the study and could potentially introduce confounding factors. However, it is likely that some patients replied early after RFA and were subsequently lost to BPI follow-up. The reasons for this could be due to lack of interest, other priorities, distracting end-of-life issues, avoidance, disease evolution, development of other medical problems, or death.

The major complication rate of RFA in the treatment of hepatic neoplasms has been estimated to be less than 3% [19, 25, 26]. Complications include bleeding, effusion, fever, fistula, bowel injury, portal vein thrombosis, needle track seeding, abscess, hematuria, transient numbness or weakness, and infection. Due to this low complication rate, many percutaneous RFAs are performed in the outpatient setting with the patient under conscious sedation. However, to prevent harm from patient movement while the probe is in place, general anesthesia has become the norm at our institution. Complications from hepatic RFA are usually managed non-operatively.

Outside the liver, complications may be more frequent, given the wide variety of locations that may present with cancer-related pain. The predictable nature of RFA contributes to its relatively low risk of collateral damage to nearby structures when appropriate safeguards are employed. In fact, the “heat sink” effect of adjacent blood vessels serves to protect vascular integrity from thermal damage. This protective effect, however, also preserves the viability of neighboring tumor cells and increases the likelihood of local tumor recurrence. The vast majority of complications from RFA should be apparent in the weeks after RFA.

The technique of RFA is dependent upon basic image guidance principles in practice, which translates into widespread availability. However, cautious use by experienced practitioners should be performed in the application of RFA to unconventional areas of the body and outside the liver. The general rules and protocols for RFA use in the liver or kidneys cannot be applied reliably to extrahepatic soft tissue metastases without risk to the patient. Careful attention should be paid to anatomy, with a detailed comprehensive pre-procedural clinic visit and imaging evaluation to complete the treatment plan. Bowel and major motor nerves should be avoided. If RFA is unsuccessful because of the creation of too small a thermal lesion, the procedure may be repeated. RFA may also be repeated if it is successful and pain subsequently recurs.

Successful palliation may not be affected by local tumor recurrence or incomplete tumor ablation. A tumor-free treatment margin may not be requisite to palliation. The mechanism of pain relief with RFA may be a decrease in interstitial or intratumoral pressure, a decrease in cytotoxin release, decreased pressure on adjacent anatomy, or destructive ablation of nerve fibers, which carry pain signals and, fortuitously, are the most sensitive to heat [27]. In some cases, merely putting a small thermal lesion hole in the tumor may control pain, perhaps by these speculative mechanisms. These mechanisms may also explain why painful bone metastases may respond to RFA directed at the periosteal-tumor junction.

One can envision the investigation of RFA combined with other treatment approaches. For example, RFA could be used in combination with conventional radiation therapy or bone cement to provide more rapid relief to a symptomatic lesion in a patient with multiple metastases. The addition of RFA should not preclude or change conventional treatment options. Unlike radiation therapy, RFA should not compromise bone marrow reserves. The procedure does not cause sedation during the precious moments of life in the end-stage patient. Collateral damage to nearby structures could be less with RFA than with other forms of local therapy, such as surgery or radiation therapy. Radiotherapy may be a more standard procedure for painful soft tissue tumor pain but may be slow to take effect and has limited durability. Radiopharmaceuticals used for painful bone cancer have similar weaknesses. Symptomatic relief for metastatic bone pain following antineoplastic therapies generally takes 4–12 weeks, which is a lengthy time interval given the potentially brief life expectancy of the patient [28]. Also, tumor recurrence following irradiation may not be amenable to repeated treatments.

Quality of life may be significantly improved following RFA in patients with cancer-related pain. The questions relating to interference of pain address only a small dimension of quality-of-life issues. Further studies with detailed quality-of-life questionnaires and medication diaries could clarify and better define the impact and durability of RFA for pain palliation.

Conclusion

Uncontrolled cancer-related pain is experienced by more than 70% of patients with metastatic disease [28]. This is a major healthcare problem without adequate solutions. RFA may be an effective method of pain control in the cancer patient with focal tumor pain. Preliminary results suggest that RFA is a feasible and safe option for patients with soft tissue cancer pain refractory to standard therapies. Future studies will address the durability of response and the strength of imaging as a surrogate marker for treatment outcome and pain recurrence. Additional study is also needed to refine the indications for RFA and to validate its efficacy before it can be broadly offered as a palliative measure.