Radiofrequency Ablation in Patients with PrimaryBreast CarcinomaA Pilot Study in 26 Patients

Francesco Izzo, M.D.1

Renato Thomas, M.D.1

Paolo Delrio, M.D.1

Massimo Rinaldo, M.D.1

Paolo Vallone, M.D.2

Anna DeChiara, M.D.3

Gerardo Botti, M.D.3

Giuseppe D’Aiuto, M.D.1

Pina Cortino, M.D.1

Steven A. Curley, M.D.4

1 Division of Surgical Oncology, The G. PascaleNational Cancer Institute, Naples, Italy.

2 Division of Diagnostic Radiology, The G. PascaleNational Cancer Institute, Naples, Italy.

3 Division of Pathology, The G. Pascale NationalCancer Institute, Naples, Italy.

4 Department of Surgical Oncology, The Universityof Texas M. D. Anderson Cancer Center, Houston,Texas.

RadioTherapeutics Corporation provided financialsupport for a research nurse.

Address for reprints: Steven A. Curley, M.D., De-partment of Surgical Oncology, The University ofTexas M. D. Anderson Cancer Center, Box 444,1515 Holcombe Boulevard, Houston, TX 77030-4009; Fax: (713) 792-4689; E-mail: scurley@mdanderson.org

Received March 27, 2001; revision received June15, 2001; accepted July 19, 2001.

BACKGROUND. The authors performed a pilot trial of ultrasound-guided percuta-

neous radiofrequency ablation (RFA) in patients with T1 and T2 breast tumors 1)

to confirm complete coagulative necrosis of tumor tissue and 2) to determine the

safety and complications related to this treatment.

METHODS. Twenty-six patients with biopsy-proven, invasive breast carcinoma un-

derwent RFA of their breast tumors followed by immediate resection. Treatment

was planned to ablate the tumor and a 5 mm margin of surrounding breast tissue.

Tumor viability after RFA was assessed by hematoxylin and eosin and nicotinamide

adenine dinucleotide vital staining.

RESULTS. Twenty patients (77%) had T1 tumors, and six patients (23%) had T2

tumors. The mean greatest dimension of tumors that were treated with RFA was 1.8

cm (range, 0.7–3.0 cm). The mean treatment time for two-phase RFA treatment was

15 minutes and 23 seconds (range, from 6 minutes and 25 seconds to 24 minutes

and 54 seconds). Coagulation necrosis of the tumor was complete in 25 of 26

patients (96%): One patient had a microscopic focus of viable tissue adjacent to the

needle shaft site. A single patient (1 of 26 patients; 4%) had a complication related

to RFA: a full thickness burn of the skin overlying a tumor that was immediately

beneath the skin.

CONCLUSIONS. This pilot experience with RFA in the treatment of patients with

early-stage, primary breast carcinoma revealed that 1) coagulative necrosis of the

entire tumor occurred in 96% of the patients, and 2) the treatment was safe, with

only a 4% complication rate. The authors have initiated a trial of RFA alone (no

resection) for patients with T1 and T2 breast tumors that will include sentinel

lymph node mapping and postablation irradiation. Cancer 2001;92:2036 – 44.

© 2001 American Cancer Society.

KEYWORDS: invasive breast carcinoma, radiofrequency ablation, thermal ablation.

Breast conservation surgery gained wide acceptance as a treatmentoption for patients with Stage I or II breast carcinoma after large,

randomized studies showed no difference in survival rates betweenpatients who underwent radical or modified radical mastectomy andbreast conservation therapy.1–3 The goal of breast conservation sur-gery (also referred to as lumpectomy, segmental mastectomy, or widelocal excision) is complete removal of the malignant breast tumor anda surrounding rim of normal breast tissue. The oncologic principle ofcomplete tumor resection including a surrounding rim of nonmalig-nant tissue to reduce local recurrence rates is followed, with a sec-ondary goal being the maintaining of an adequate amount of non-malignant breast tissue to produce a normal breast contour; thus theterm breast conservation.

2036

© 2001 American Cancer Society

Technologic advances over the last decade havefueled interest in even less invasive local treatmentsfor patients with breast carcinoma. So-called mini-mally invasive techniques that involve percutaneouseradication of breast tumors are being studied by anumber of investigators. Minimally invasive thermaldestruction, rather than surgical excision, of early-stage primary breast tumors may be appropriate incarefully selected patients. One treatment techniquethat allows in situ, percutaneous destruction of abreast tumor is radiofrequency ablation (RFA). Weperformed a pilot trial of ultrasound-guided, percuta-neous RFA in patients with T1 and T2 breast tumors 1)to confirm histologically complete coagulative necro-sis of tumor tissue and 2) to determine the safety andcomplications related to this treatment.

MATERIALS AND METHODSBetween July, 1999 and April, 2000, patients with StageI or Stage II invasive breast carcinoma were offered anopportunity to participate in an Institutional ReviewBoard-approved clinical trial involving percutaneousRFA followed by immediate resection of the breasttumor at the G. Pascale National Tumor Institute inNaples, Italy. To be considered for this trial, all pa-tients had mammographic and ultrasonographic evi-dence of a solitary breast mass measuring , 3.0 cm ingreatest dimension that was needle biopsy-proven in-vasive breast carcinoma. Patients were excluded ifthey were pregnant or lactating, if they had a history ofipsilateral breast surgery or irradiation, or if they hada platelet count , 50,000/mm3 or prothrombin activ-ity , 60% of normal. Tumors lying within 10 mm ofthe skin were excluded from RFA treatment. All par-ticipating patients signed written, informed consent

that was approved by the clinical research investiga-tion and Ethics Committee of the institute.

RFA was performed in the operating room undergeneral inhalation anesthesia. The breast, chest wall,and axilla was prepped and draped sterilely for thestandard surgical procedure to follow the RFA. Thebreast tumor was identified by intraoperative ultra-sonography using an Aloka 2000 system (Wallingford,CT) with a broadband, 5.0 –7.5 mHz, linear transducer.The skin overlying the tumor was punctured with an11 blade; then, a 15-gauge, insulated shaft needle elec-trode (LeVeen needle electrode; RadioTherapeuticsCorporation, Mountain View, CA) was inserted intothe tumor using real-time ultrasonographic guidance.After satisfactory placement of the needle electrodeinto the tumor, the 10-tine multiple array electrodeswere deployed from the needle tip (Fig. 1). The LeVeenneedle electrode is available with four diameters of thefully deployed array: 2.0 cm, 3.0 cm, 3.5 cm, and 4.0cm. The diameter of the array chosen for treatmentwas based on a treatment plan that included RFA ofthe entire tumor and at least a 5-mm zone of sur-rounding breast tissue. The monopolar multiple arrayneedle electrode was connected to an RF-2000 gener-ator (RadioTherapeutics Corporation) with twogrounding pads (Valley Laboratory, Boulder, CO)placed on each thigh to complete the current path.

RFA was performed following a predeterminedtwo-phase algorithm. Treatment was initiated at 10watts of power for 2 minutes, after which, power wasincreased in 5-watt increments every minute until tis-sue impedance rose rapidly and power dropped below10 watts, thus indicating complete coagulative necro-sis of the target lesion. After a 30-second pause, asecond phase of treatment was applied, again begin-

FIGURE 1. A photograph of LeVeen

multiple array radiofrequency ablation

(RFA) needle electrodes. The diameter of

the fully deployed array is either 2.0 cm,

3.0 cm, 3.5 cm, or 4.0 cm. The 10 tines

of the multiple array are retracted com-

pletely into the needle shaft for place-

ment of the needle. Using real-time ul-

trasonographic guidance to place the

needle tip, the tines are opened into the

tumor to be treated by RFA. Upon com-

pletion of the RFA treatment, the array is

retracted back into the needle shaft and

the needle is removed.

Thermal Ablation of Breast Carcinoma/Izzo et al. 2037

ning at 10 watts for 2 minutes followed by increases inpower of 5 watts per minute until tissue impedanceagain rose and power rolled off. Power (watts) andimpedance (ohms) were monitored continuously dur-ing treatment. The maximum power at the time ofincrease in the tissue impedance preceding power rolloff and the total time needed to complete two-phaseRFA were recorded.

After completion of the two-phase RFA treatment,the multiple array tines were retracted into the needleshaft, and the needle electrode was removed. Thepresence or absence of bleeding from the needle trackwas noted, and any evidence of thermal injury to theskin overlying the tumor was measured. Each patientthen underwent the planned resection of the breasttumor with axillary lymphadenectomy. The surgicalbreast specimen was sent immediately to the pathol-ogy laboratory, and margins were inked. The breasttissue and tumor were bisected centrally along theneedle track; the tumor and the zone of RF-inducedcoagulative tissue necrosis were measured. The tumorand surrounding tissue were then serially sectioned.Two representative areas of tumor and adjacent breasttissue were obtained from each 1–2 mm section cutthrough the breast tumor. One of these areas of tissuefrom each section was fixed in formalin, embedded inparaffin, sectioned, mounted on microscope slides,and then stained with hematoxylin and eosin (H&E).The second area from each section was immediatelyplaced in OCT compound embedding medium (TissueTek; Sakura Finetek, Inc., Torrance, CA), and then wassnap frozen in liquid nitrogen and stored at 2 70 °C fornicotinamide adenine dinucleotide-diaphorase (NADH-diaphorase) cell viability staining. This latter histo-chemical cell viability determination is based on thereduction of nitroblue tetrazolium chloride, a redoxindicator, by NADH-diaphorase. Viable cells constitu-tively express NADH-diaphorase, resulting in an in-tense blue cytoplasmic pigment using this assay.NADH-diaphorase activity ceases immediately uponcell death, leading to an absence of blue cytoplasmicpigment in nonviable cells.4 The snap-frozen tumorareas were cut on a cryostat into 8-mm, unfixed sec-tions and placed on glass slides. The tissue sectionslides were then covered with 100 mL of incubationmedia for 15 minutes under aerobic conditions atroom temperature. Incubation media consisted of 1mL of reduced [a]-NADH (Sigma-Aldrich Corporation,St. Louis, MO) at a concentration of 2.5 mg/mL; 2.5 mLof nitroblue tetrazolium chloride (Sigma-Aldrich Cor-poration) at a concentration of 2.0 mg/mL; 1 mL ofphosphate-buffered saline, pH 7.4; and 0.5 mL ofRinger solution. After 15 minutes of exposure to theincubation media, each slide was washed in distilled

water for 2 minutes. Glass coverslips were mountedover the tissue, and the slides were evaluated sepa-rately by two different pathologists (A.D. and G.B.) todetermine whether there were any viable tumor cellsin the area of RFA. A section of normal liver was usedas a positive control for NADH-diaphorase viabilitystaining, and a section of normal liver that had beenheated to 100 °C was used as a negative control.

RESULTSTwenty-six women participated in this study. Theirmedian age was 57 years (range, 37–78 years). Theclinical staging information based on mammography,ultrasonography, and palpation of the breast and ax-illa; the results of core needle biopsy prior to RFA; andthe breast surgery performed immediately after RFAare listed in Table 1.

The mean greatest dimension of breast tumorstreated with RFA was 1.8 cm (range, 0.7–3.0 cm). Basedon the ultrasonographic maximum tumor dimensionand the RFA treatment plan to coagulate the tumorand a surrounding rim of normal breast tissue, the2.0-cm, multiple-array needle electrode was used in 4patients (15%), the 3.0-cm electrode was used in 8patients (30%), and the 3.5-cm needle was used in 14patients (55%). The mean initial tissue impedance atthe onset of both phases of treatment was 95 ohms(range, 55–127 ohms). The mean time required tocomplete two-phase RFA treatment was 15 minutesand 23 seconds (range, from 6 minutes and 25 secondsto 24 minutes and 54 seconds). The mean maximum

TABLE 1Clinical Staging, Pretreatment Core Biopsy Results, and SurgicalProcedure in 26 Women After Undergoing RadiofrequencyAblation of Primary Breast Carcinoma

Characteristic No. of patients (%)

TNM classificationT1 20 (77)T2 6 (23)N0 19 (73)N1 7 (27)M0 26 (100)

Clinical stageI 19 (73)II 7 (27)

Core biopsy resultsInfiltrating ductal carcinoma 17 (66)Infiltrating lobular carcinoma 6 (23)Tubular carcinoma 3 (11)

Surgical procedure after RFAQuadrentectomy and axillary lymphadenectomy 22 (85)Modified radical mastectomy 4 (15)

RFA: radiofrequency ablation.

2038 CANCER October 15, 2001 / Volume 92 / Number 8

power at roll off during the first phase of RFA treat-ment was 45 watts (range, 25– 80 watts), and the meanmaximum power at roll off during the second phase oftreatment was 21 watts (range, 10 – 40 watts). In all butone RFA treatment, the first phase of treatment waslonger than the second phase, taking an average of70% of the total treatment time (range, 59 – 82%). In asingle instance, the first and second phases of RFAtreatment time were the same.

RFA of the breast tumor was monitored continu-ously with ultrasonography. In all patients, as tumorheating around the multiple array electrodes devel-oped, an ill-defined, hyperechoic zone developed. Sig-nificant ultrasonographic shadowing from the hypere-choic thermal lesion made it difficult to measure thesize of the area of RFA-induced coagulative necrosisduring treatment.

There was a single complication related to RFA inthe 26 patients (4% incidence rate): A full-thickness

burn measuring 2.5 cm in greatest dimension oc-curred in the skin immediately overlying a superficialtumor. The entire burn area was excised during thebreast tumor resection (modified radical mastectomywith immediate TRAM reconstruction), and the pa-tient had no further sequelae. None of the patientswho were treated first with RFA developed wound-healing or infectious complications. There was nobleeding from the needle track upon removal of theRFA needle electrode in any of the 26 patients.

On macroscopic examination of the surgical spec-imen, there was central charring of the tumor andneedle track with a surrounding area of yellow, coag-ulated, adipose and breast tissue (Fig. 2). The marginbetween coagulated and viable breast tissue was notsharply demarcated. On microscopic evaluation of theH&E-stained slides, thrombosis of blood vessels, pyk-notic nuclei, and eosinophic staining of tumor cellcytoplasm was noted. The NADH-diaphorase enzyme

FIGURE 2. A specimen photograph after wide local excision of a T1 breast tumor (arrows) treated with percutaneous radiofrequency ablation immediately prior

to resection. The skin is at the top of the photograph, and the central charring of the needle track through the center of the tumor is readily apparent. The light-yellow

color of the fat surrounding the tumor is consistent with radiofrequency ablation of the surrounding adipose and breast tissue and is different than the darker yellow

and red fat in the area outside the zone of thermal destruction (arrowhead).

Thermal Ablation of Breast Carcinoma/Izzo et al. 2039

histochemical analysis showed no viable tumor cellareas in the serially sectioned specimens from 25 pa-tients (96%). A single patient (4%) had a solitary mi-croscopic focus of blue staining near the needle track.The NADH-diaphorase staining also showed clear de-marcation between nonviable and viable breast tissuein the RFA zone around the tumor (Fig. 3).

DISCUSSIONThe local application of heat to treat patients withmalignant tumors is not a novel concept. The EdwinSmith papyrus describes topical application of hot oilor heated metallic implements that was used approx-imately 5000 years ago to treat patients with tumors.5

The use of an electrical current to produce thermaltissue necrosis in patients with breast carcinoma alsois not new: Metallic or clay-insulated electrodes wereinserted into locally advanced breast tumors in thelate 19th century to shrink the tumor and reduce painand bleeding.6

In general, thermal injury to cells begins at 42 °C,with the exposure times to low-level hyperthermianeeded to achieve cell death ranging from 3 hours to50 hours, depending on the tissue type and condi-tions.7 There is an exponential decrease in the expo-sure time necessary for a cytodestructive response asone increases the temperature above 42 °C. For exam-ple, only 8 minutes at 46 °C are needed to kill malig-nant cells, and 51 °C can be lethal after only 2 min-utes.8,9 At temperatures above 60 °C, intracellularproteins become denatured, lipid bilayers melt, andcell death is inevitable.10 It is interesting to note thatmalignant cells are more resistant to lethal damagefrom freezing compared with normal cells but aremore sensitive to hyperthermic damage comparedwith normal cells.11,12

The use of RF energy to produce thermal tissuedestruction has been the focus of increasing researchand practice for the past several years.13–16 During theapplication of RF energy, a high-frequency, alternat-ing current moves from the tip of an electrode into thetissue surrounding that electrode. Ion movement re-sults in frictional heating of the tissue as the ionswithin the tissue attempt to follow the change in thedirection of the alternating current. Cells begin to dieas the temperature within the tissue becomes elevatedbeyond 60 °C, resulting in a region of necrosis sur-rounding the electrode.17 A typical RFA treatment re-

sults in local tissue temperatures that exceed 100 °C,which produces coagulative necrosis of the tumor tis-sue and surrounding hepatic parenchyma. The tissuemicrovasculature is destroyed completely, and throm-bosis of hepatic arterial, portal venous, or hepaticvenous branches measuring , 3 mm in greatest di-mension occurs. Only tissue through which RF elec-trical current passes directly is heated above a cyto-toxic temperature. The geometry of the RF currentpathway around the ablation electrode creates a rela-tively uniform zone of radiant/conductive heat withinthe first few millimeters of electrode-tissue interface.The conductive heat emitted from the tissue radiatesout from the electrode; and, if the tissue impedance isrelatively low, then a dynamic expanding zone of ab-lated tissue is created. The final size of the region ofheat-ablated tissue is proportional to the square of theRF current, also known as the RF power density. TheRF power/current delivered through a monopolarelectrode decreases in proportion to the square of thedistance for the electrode. Therefore, the tissue tem-perature falls rapidly with increasing distance fromthe electrode. RF energy is applied initially at lowpower and is increased progressively to prevent rapidtissue heating and formation of desiccated tissue im-mediately adjacent to the multiple tines. High RF en-ergies applied at the onset of treatment producesrapid coagulation of tissue around the tines, thus re-stricting propagation of heat throughout the targetlesion, leaving microscopic bands of viable tumor cellsbetween the tines.

When considering any technique to achieve localtumor ablation, the primary goal is to produce com-plete in situ destruction of all malignant tissue, includ-ing a rim of surrounding nonmalignant tissue. Like atumor free surgical margin, an area of normal tissuearound the tumor is ablated to treat any microscopicextensions of malignant cells at the tumor periphery. Aconcordant secondary goal is to achieve effective localtumor control with minimal risks and side effects forthe patient. Our group has had significant experiencewith RFA in the treatment of patients with unresect-able primary and secondary hepatic malignan-cies.18 –21 We have now treated over 500 patients and800 tumors in the liver using open or percutaneousRFA with a mortality rate of 0.2%, a major complica-tion rate of less than 5%, and a local recurrence rate ofunder 8%. An optimistic current lying under the grow-

‹

FIGURE 3. (A) Nicotinamide adenine dinucleotide (NADH)-diaphorase staining of breast tumor cells treated with radiofrequency ablation. The complete absence

of blue staining in this photomicrograph indicates nonviable tumor after radiofrequency ablation (original magnification, 3125). (B) A photomicrograph of the zone

of demarcation between nonviable tissue (left) and viable tissue (right) stained blue by NADH-diaphorase.

2040 CANCER October 15, 2001 / Volume 92 / Number 8

FIGURE 3.

ing wave of popularity of RFA for the treatment ofpatients with hepatic malignancies holds that subsetsof patients may achieve long-term disease free sur-vival, but another 3– 4 years of follow-up are requiredbefore we can establish these survival rates.

RFA in patients with early-stage breast carcinomais an appealing treatment if it can produce local con-trol rates similar to the rates produced with wide localexcision with improved cosmetic results. The purposeof this study was to provide histopathologic confirma-tion of complete necrosis in patients with T1 or T2breast tumors who were treated with RFA. It is criticalto confirm the accuracy of ultrasound or other imag-ing techniques to guide placement of the RF needleelectrode, with the endpoint of coagulative necrosis ofthe entire tumor and a surrounding rim of normalbreast tissue. Thus, this study required resection of thetumor and surrounding breast tissue treated with RFA.Macroscopic changes were seen in the tumor tissueand included charring and a surrounding target lesionextending into the nonmalignant breast. The concen-tric circles of heat-treated tissue end with a red ring ofhemorrhagic tissue that is indicative of the borderbetween viable tissue and nonviable tissue.22 BothH&E and NADH-diaphorase staining showed com-pletely necrotic tissue in all but a single slide from onepatient. Repeat sectioning of the tissue blocks corre-sponding to this blue-stained area failed to confirmany viable tissue on repeat NADH-diaphorase stain-ing. It is possible that the positive staining on theinitial slide was an artifact or represented improperstaining. It is unusual to see viable cells near theneedle shaft, where tissue temperatures exceed 100 °C.In our experience with hepatic RFA, all local recur-rences have been at the periphery of a tumor, not atthe center near the electrode, indicating incompletetreatment at a distance away from the central needleshaft and tines. Nonetheless, we scored this area aspossible viable tumor, yielding a 4% incomplete treat-ment rate (1 of 26 patients).

The first report of RFA in the treatment of patientswith breast carcinoma described five patients withlocally advanced (Stage III) breast carcinoma or withprimary tumors measuring . 5 cm.23 Similar to thecurrent study, all five patients were treated with RFAand then immediately underwent resection of thebreast tumor and all surrounding breast tissue. Theauthors reported that coagulative necrosis of the pri-mary tumor was complete in four of five patients, witha single patient who had a microscopic focus of viablebreast tumor cells in the wall of a cyst at the peripheryof the RFA treatment zone. There were no complica-tions related to RFA in the five patients, although theauthors did not describe the location of the tumors

within the breast or their proximity to the skin.Clearly, our experience suggests that superficial tu-mors close to the skin probably should not be treatedwith RFA because of the risk of thermal injury to theskin. Five patients who were treated in our series withRFA had tumors that were deep near the chest wall.Excision of these lesions showed no injury to the mus-cle underlying the fascia, and there were no thermalinjuries to the chest wall, axilla, or adjacent organs inour patients. Clearly, the effect of RFA of a primarybreast tumor that is not resected will be associatedwith the formation of scar tissue within the breast.Whether this will have a negative impact on the cos-metic appearance of the breast or the accuracy offollow-up studies to assess for local recurrence ofbreast carcinoma has yet to be determined.

RFA is not the only local tissue-ablative techniqueto be studied for the treatment of patients with breastcarcinoma. Cryoablation has been used extensively totreat patients with carcinoma of the liver and prostategland; however, reports regarding the use of this tech-nology for the treatment of patients with breast carci-noma are anecdotal.24 In an experimental model ofbreast carcinoma, a single freeze-thaw cycle of murinetumors was associated with a local recurrence rate of80%.25 A total of five freeze-thaw cycles was necessaryto decrease recurrence rates to less than 5%. Ster-eotactically guided laser ablation of breast tumors hasbeen reported in 36 patients, with a complete tumornecrosis rate of 66%.26 High-intensity, focused ultra-sound also is undergoing evaluation, although bothlaser and focused ultrasound treatments result insmall areas of tissue ablation (5–10 mm), thus requir-ing complex treatment planning to produce multiple,overlapping zones of ablation to treat patients withclinically relevant breast tumors.26 –28 The multiple-tine electrodes used with current RFA equipment per-mits the treatment of patients with early-stage breasttumors, particularly those with Stage I or II diseasewho usually can be treated with a single placement ofthe needle electrode.

There are important limitations in using RFA totreat patients with early-stage breast carcinoma. First,ultrasonographic monitoring of the RFA treatmentdoes not provide accurate measurement of the his-topathologic zone of complete coagulative necrosis.The hyperechogenecity of the heated breast tissueproduces an indistinct leading edge of complete ne-crosis, and the ultrasonographic shadowing makes itimpossible to monitor the deeper zones of treatment.Because of the limitations of real-time assessment ofthe adequacy of RFA treatment, we currently are per-forming a study to compare magnetic resonance (MR)imaging and ultrasonography during and after RFA to

2042 CANCER October 15, 2001 / Volume 92 / Number 8

determine which technique is more accurate.29,30 Thatstudy will involve 30 patients who undergo RFA usingultrasonographic and MR monitoring. The treatedbreast tumor will be assessed by MR and ultrasound 1day and 7 days after RFA; then, the tumor will beresected for pathologic correlation of imaging accu-racy. The second problem associated with RFA is al-luded above: Specifically, there are no data at this timeon the long-term morphologic changes in the tumorand surrounding breast tissue after RFA. Although it ishoped that this treatment may produce a superiorcosmetic result while accomplishing all oncologic ob-jectives, this awaits confirmation by a prospectivestudy. Furthermore, many patients will be candidatesfor adjuvant breast irradiation, which will complicatefurther the local healing process after RFA. A thirdissue relates to marked heterogeneity in breast sizeand composition of fatty and stromal elements be-tween individual patients as well as the coexistence insome patients of nonmalignant disorders, such as fi-brocystic disease. RFA treatment time and efficacymay be affected by differences in breast tissue com-position, vascularity, inflammatory conditions, or bythe location of tumors near the chest wall or axilla.Commensurate with significant individual variationsin breast tissue composition is the wide range (55––127 ohms) we observed in the initial tissue impedanceprior to RFA. In contrast, the tissue impedance in livertumors and the surrounding parenchyma has lessvariability (45– 65 ohms).31 The importance of theseindividual patient variations also must be recordedcarefully and assessed in any studies that evaluate RFAor other local tissue-ablative techniques. Finally, re-gardless of cosmetic effect, the ultimate assessmentwill be oncologic effectiveness of RFA compared withcomplete margin negative resection of breast carci-noma. To begin accruing data to help answer thisquestion, we will initiate a randomized, Phase III trialof RFA with sentinel lymph node mapping and se-lected axillary lymphadenectomy compared with sur-gical resection with the same lymph node assessmentafter we complete the MR imaging study.

Our initial evaluation of RFA in the treatment ofpatients with early-stage breast carcinoma suggeststhat it should be studied further in patients with StageI disease. It remains unclear whether this treatmentshould be offered to patients with Stage II or largertumors. RFA appears to produce complete coagulativenecrosis in over 95% of small primary breast tumors,and studies to assess the oncologic effectiveness ofthis treatment, keeping in mind the treatment limita-tions and variations among patients noted above,should be performed.

REFERENCES1. Veronesi U, Saccozzi R, Del Vecchio M, Banfi A, Clemente C,

De Lena M, et al. Comparing radical mastectomy withquadrantectomy, axillary dissection, and radiotherapy inpatients with small cancers of the breast. N Engl J Med1981;305(1):6 –11.

2. Fisher B, Redmond C, Poisson R, Margolese R, Wolmark N,Wickerham L, et al. Eight-year results of a randomized clin-ical trial comparing total mastectomy and lumpectomy withor without irradiation in the treatment of breast cancer.N Engl J Med 1989;320(13):822– 8.

3. Fisher B, Anderson S, Redmond CK, Wolmark N, WickerhamDL, Cronin WM. Reanalysis and results after 12 years offollow-up in a randomized clinical trial comparing totalmastectomy with lumpectomy with or without irradiation inthe treatment of breast cancer. N Engl J Med 1995;333(22):1456 – 61.

4. Neumann RA, Knobler RM, Pieczkowski F, Gebhart W. En-zyme histochemical analysis of cell viability after argon la-ser-induced coagulation necrosis of the skin. J Am AcadDermatol 1991;25(6 Pt 1):991– 8.

5. Breasted JH. The Edwin Smith surgical papyrus, vol. 54.Chicago: Chicago University Press, 1930.

6. Rockwell AD. Electro-surgery: benign and malignant tu-mors. In: Rockwell AD, editor. The medical and surgical usesof electricity. New York: EB Treat & Company, 1903:565– 6.

7. Dickson JA, Calderwood SK. Temperature range and selec-tive sensitivity of tumors to hyperthermia: a critical review.Ann NY Acad Sci 1980;335:180 –205.

8. Hill RP, Hunt JW. Hyperthermia. In: Tannock IF, Hill RP,editors. The basic science of oncology. New York: PergamonPress, 1987:337–57.

9. Haines DE, Watson DD, Halperin C. Characteristics of heattransfer and determination of temperature gradient andviability threshold during radiofrequency fulguration of iso-lated perfused canine right ventricle. Circulation 1987;76:278 – 82.

10. Grundfest WS, Litvack FI, Doyle DL. Laser-tissue interac-tions: considerations for cardiovascular applications. In:White RA, Grundfest WS, editors. Lasers in cardiovasculardisease. Chicago: Yearbook Medical Publishers, 1987:32–43.

11. Bischof J, Christov K, Rubinsky B. A morphological study ofcooling rate response in normal and neoplastic human livertissue: cryosurgical implications. Cryobiology 1993;30(5):482–92.

12. Steeves RA. Hyperthermia in cancer therapy: where are wetoday and where are we going? Bull NY Acad Med 1992;68(2):341–50.

13. Siperstein AE, Rogers SJ, Hansen PD, Gitomirsky A. Laparo-scopic thermal ablation of hepatic neuroendocrine tumormetastases. Surgery 1997;122(6):1147–55.

14. Goldburg SN, Gazelle GS, Solbiati L. Ablation of liver tumorsusing percutaneous RF therapy. Am J Res 1998;170:1023– 8.

15. Lencioni R, Goletti O, Armillotta N, Paolicchi A, Moretti M,Cioni D, et al. Radio-frequency thermal ablation of livermetastases with a cooled-tip electrode needle: results of apilot clinical trial. Eur Radiol 1998;8(7):1205–11.

16. Nagata Y, Hiraoka M, Nishimura Y, Masunaga S, MitumoriM, Okuno Y, et al. Clinical results of radiofrequency hyper-thermia for malignant liver tumors. Int J Radiat Oncol BiolPhys 1997;38(2):359 – 65.

Thermal Ablation of Breast Carcinoma/Izzo et al. 2043

17. McGahan JP, Brock JM, Tesluk H, Gu WZ, Schneider P,Browning PD. Hepatic ablation with use of radio-frequencyelectrocautery in the animal model. J Vasc Interv Radiol1992;3(2):291–7.

18. Izzo F, Di Muria A, Ruffolo F, Parisi V, Curley SA. Intraop-erative use of radiofrequency interstitial tissue ablation totreat primary and metastatic liver tumors. Gastroenterol Int1997;10:515– 6.

19. Curley SA, Izzo F, Delrio P, Ellis LM, Granchi J, Vallone P, etal. Radiofrequency ablation of unresectable primary andmetastatic hepatic malignancies: results in 123 patients.Ann Surg 1999;230(1):1– 8.

20. Pearson AS, Izzo F, Fleming RY, Ellis LM, Delrio P, RohMS, et al. Intraoperative radiofrequency ablation orcryoablation for hepatic malignancies. Am J Surg 1999;178(6):592–9.

21. Curley SA, Izzo F, Ellis LM, Nicolas Vauthey J, Vallone P.Radiofrequency ablation of hepatocellular cancer in 110patients with cirrhosis. Ann Surg 2000;232(3):381–91.

22. Thomsen S. Mapping of thermal injury in biologic tissuesusing quantitative pathologic techniques. SPIE Conferenceon Thermal Treatment of Tissue with Image Guidance.1999;3594:82–95.

23. Jeffrey SS, Birdwell RL, Ikeda DM, Daniel BL, Nowels KW,Dirbas FM, et al. Radiofrequency ablation of breast cancer:

first report of an emerging technology. Arch Surg 1999;134(10):1064 – 8.

24. Staren ED, Sabel MS, Gianakakis LM, Wiener GA, Hart VM,Gorski M, et al. Cryosurgery of breast cancer. Arch Surg1997;132(1):28 –34.

25. Rand RW, Rand RP, Eggerding F, DenBesten L, King W.Cryolumpectomy for carcinoma of the breast. Surg GynecolObstet 1987;165(5):392– 6.

26. Mumtaz H, Hall-Craggs MA, Wotherspoon A, Paley M,Buonaccorsi G, Amin Z, et al. Laser therapy for breast can-cer: MR imaging and histopathologic correlation. Radiology1996;200(3):651– 8.

27. Harari PM, Hynynen KH, Roemer RB, Anhalt DP, Shimm DS,Stea B, et al. Development of scanned focused ultrasoundhyperthermia: clinical response evaluation. Int J Radiat On-col Biol Phys 1991;21(3):831– 40.

28. Hill CR, ter Haar GR. Review article: high intensity focusedultrasound—potential for cancer treatment. Br J Radiol1995;68(816):1296 –303.

29. Hall-Craggs MA. Interventional MRI of the breast: minimallyinvasive therapy. Eur Radiol 2000;10(1):59 – 62.

30. Fornage BD. Recent advances in breast sonography. Jbr-Btr:Organe Soc R Belge Radiol 2000;83(2):75– 80.

31. Curley SA. Radiofrequency ablation of malignant liver tu-mors. Oncologist 2001;6(1):14 –23.

2044 CANCER October 15, 2001 / Volume 92 / Number 8