Magnetic-Induction Hyperthermia

Results of a 5-Year Multi-Institutional National Cooperative Trial in Advanced Cancer Patients

F. KRISTIAN STORM,-MD, HARVlEY W. BAKER, MD, EE)WARD F. SCANLON, ~D, HENRY P. PLENK, MD, P~AUL M. MEADOWS, ME), STEPHEIN C. COHEN, MD, CARL E. OLSON, MD, JOHN W. THOMSON, MD, JANARDAN, D. KHANDEKAR, MD, DENISE ROE, MS, ANNE NlZZE, PAC, AND DONALD L. MORTON, MD

Nine US institutions performed 14,807 Phase 1-1I treatments of magnetic-induction: (Magnetr0de [Henry Medical Electronics, Inc., Los Angeles, CA]) hyperthermia in 1170 adults. All had advanced tumors: 20% had untreated inoperable cancer or disease progression despite surgery (10%), iadiation therapy (XRT) (3%), chemotherapy (27%), or combinations (40%); 67% had pain; and 73% had reduced activity. Eighteen percent were advanced primaries, 26% were recurrent, and 56% metastatic tumors in the head and neck (7%), body wall (7%), extremity (4%), abdominal cavity (17%), pelvis (17~, lung (15%), or liver (30%); 36% were <5 cm and 64% >_5 cm. Treatments were to safe tolerance for 30 to 60 minutes for five or more treatments. Results in 960 evaluable patients were c.omplete~.....~r~sp_o_n~e_~ (1-34 months; median, 7 months), partial response 18% (1-39 months; median, 4"months), minimal response 10% (1-15 months; median, 3 months), and no change 33% (1-32 months; median, 3 months), with decreased pain in 30% and improved activity in 21%, in-d"ependent of histologic type or site. Regression was dependent on treatment type and minimum temperature: heat only, 23%; heat + XRT, 60%; heat + less-than-standard XRT because of prior XRT failure, 39%, heat + intravenous (IV) chemotherapy, 28%; heat + same previously failed IV chemotherapy, 20%; heat + intraarterial (IA) chemotherapy, 28%; heat + same previously failed IA chemotherapy, 15%; heat + standard XRT + chemotherapy, 58%; heat + less-than-standard XRT + chemotherapy, 47%; <40C, 31%; 40 to 40.9C, 45%; 41 to 41.9C, 54%; 42 to 42.9C, 47%; 43 to 43.9C, 40%; 44 to 44.9C, 33%; 45 to 45.9C, 55%; 46 to 46.9C, 63%; >47C, 100%. There were 49 (0.33%) skin burns and 2 systemic injuries (stomach ulcer at 1 month; lung fibrosis at 9 months). This trial indicates that localized hyperthermia has a significant role in palliation of human advanced solid cancer. Cancer 55:2677-2687, 1985.

From the Division of Surgical Oncology, John Wayne Clinic, Jonsson Comprehensive Cancer Center, UCLA School of Medicine, Los Angeles, California. Supported by the Department of Health and Human Services, National Cancer Institute, grants no. P01-CA29605 (D.L.M.), N01CM1 17523 (H.P.P.), R01-CA 24883 (F.K.S.); Radiation Therapy Oncology Group grant no. 2UI0-CA 17906 (H.P.P.); American Cancer Society Clinical Fellowship (F.K.S.); the Fraternal Order of the Eagles (P.M.M.); the Cohen Foundation for Oncologic Research (S.C.C.); Nancy and Carroll OConnor (F.K.S.); Michael A. Wayne and the John Wayne family (D.L.M. and F.K.S.); the Committee to Cure Cancer through Immunization (D.L.M); the Norman Sprague Jr., M.D., Foundation (F.K.S.)~ the Charles Al~right Foundation (F.K.S.); the Charles Fabrikant Foundation (F.K.S.); and a Phi Beta Psi Sorority grant (F.K.S.). Address for reprints: F. Kristian Storm, MD, Division of Surgical Oncology, 9th Floor Louis Factor Building, UCLA School of Medicine, Los Angeles, CA 90024. The authors thank the following: oncologic surgeons--Harvey W. Baker, MD, Richard Berk, MD, Constantine Hatzitheofilou, MD, C. Edwin Irish, MD, Larry R. Kaiser, MD, David H. Krag, MD, Donald L. Morton, MD, Kenneth P. Ramming, MD, Edward F. Scanlon, MD, Allan W. Silberman, MD, Philip A. Snedecor, MD, F. Kristian

Storm, MD; medical oncologists--Elmer Brestan, MD, Stephen C. Cohen, MD, Philip B. Dreisbach, MD, William P. Galen, MD, Robert Goslin, MD, Charles M. Haskell, MD, Jamardan D. Khandekar, MD, Allen Patton, MD, Gregory Sarna, MD; radiation therapists--Nell Bowie, MD, Donald Eads, MD, John J. Gallucci, MD, Raphael M. Garces, MD, Irving J. Horowitz, MD, Paul M. Meadows, MD, Carl E. Olson, MD, Henry P. Plenk, MD, Marcia J. S. Richards, MD, Robert Ring III, MD, William T. Sause, MD, John W. Thomson, MD; engineers and physicists--Doug Brewer, YuKong Chan, PhD, Charles Deal, Robert S. Elliott, Phl)~ William H. Harrison, BA, David C. Matthes, MS; RNs/technicians--Lisa A1-Hashimi, RN, Milzi Benz, RN, Mary Ann Bright, RN, Dinah Celmin, CRT, Beverly Drury, RN, Lupe Ettinger, RN, Caron Finn, RN, Jane Foard, RN, Robin Foley, RN, Kathy Duggan Holmes, RN, Patricia Leprich, RN, Sonja Maxwell, RTT, Connie Mueller, RN, Jan McNitt, RN, Claudia Ongley, RN, Ann Packard, RN, Linda Peham, RN, Kim Raja, RN, Carla Rich, RTT, Helen Rowell, RTT, Karen Stickel, RN, Barbara Sosaya, RN; statisticians--Ed Korn, PhD, Denise Roe, MS, Kenneth Pringle, Senior Programmer;, data-entry personnel--Stephen Conti, AA, Dan Greene, MSPH, Mindy lllingsworth, J. Anne Nizze, PAC, and Jamie Quintero, MPH. Accepted for publication October 17, 1984.

2677

2678

CANCER June 1 1985

Vol. 55

BODY of in vitro, in vivo, and clinical A SUBSTANTIAL evidence indicates that hyperthermia has significant anti-cancer activity, particularly when combined with radiation therapy or chemotherapy.~ Hyperthermia has been produced by water baths, perfusion, enshroudment in water-circulating suits, extracorporeal heat exchange, and electromagnetic or.acoustic waves, namely, shortwaves, microwaves, or ultrasound. We performed a comprehensive Phase I-II investigation of localized hyperthermia in patients with advanced solid cancer, particularly deep-seated visceral malignancy. Localized hyperthermia is not focused, and treats the region of the body that contains the tumor; this approach assumes that tumors heat as well as normal tissue and that in many cases tumors selectively retain higher levels of heat due to relatively poor blood flow and inadequate vasoregulation.2 Patients were also treated with localized hyperthermia combined with standard doses of radiation therapy or chemotherapy, based upon the premise of an additive or synergistic response.3-6 Patients whose disease progressed after radiation therapy were subsequently treated with additional low doses of radiation combined with heat. Similarly, patients who had failed the best known chemotherapy for their disease were treated with the same ineffective agent(s) plus hyperthermia, in the hopes of bringing about a secondary response. This cooperative study was desig.ned to answer several important, clinically relevant questions, ts hyperthermia as a single agent benefidial, or is it best employed as adjuvant treatment with radiation therapy or chemo- therapy? Woullll standard doses of these agents become prohibitively toxic when combined with heat? Does response depend upofi tumor histologic type, size~ or location, within .the body, or upon the absolute tumor temperfiture achieved? Finally, could localized hyperthermia be l~r(orme~lshfely by Well-trained physicians and nurses, given a standard clinical setting, without extensive on-site engineering expertise7
Conduct and Methods of Study Institutions and Principal Investigators

versity of Texas affiliate)--Stephen C. Cohen, MD; St. Marys Hospital, Milwaukee, Wisconsin (University of Wisconsin affiliate)--Carl E. Olson, MD; Desert Hospital, Palm Springs, California (University of Southern California afflliate)--Philip B. Dreisbach, MD; and Pomona Valley Corrgnunity Hospital, Pomona, California (Loma Linda University affiliate)--Robert Ring, MD. Study Dates Patients were entered into the cooperative study over a 5-year period from 1977 through 1983.
Subject Selection Criteria

AII patients had histologic proof of malignancy, which was locally advanced primary, recurrent, or metastatic disease for which standard therapy offered little or no reasonable chance of cure. Candidates included those with advanced cancer for which no standard or potentially beneficial therapy exists. Hyperthermia and standard therapy were used together during initial treatment of advanced tumors where an additive or synergistic beneficial effect was expected.3-6 No proven.method of cancer therapy was withheld because of this investigation. Patients had a projected life expectancy of at least 1 month. All treated tumors were measurable in at least one plane. Any histologic type was accepted, as long as the lesion fulfilled size and accessibility requirements for evaluation. Patient accrual was by nonsolicited intramural and extramural referral.
Pretreatment Evaluation Tumors were carefully measured using calipers, and the measurements were recorded in millimeters. When an accurate depth measurement was impossible, an estinaate of tumor depth was made from the skin surface. -For inaccessible lesions, special studies, such as radiographs, angiograms, ultrasound, or compuierized tomography (CT) scans, assessed thq pretreatrncnt size, depth, and "location of the. tumor: Volume calculation was made whe~re feasible. Adequate evaluation of the patients general health, extent of disease, and capacity to undergo treatment included: (1) history, physical examination, and .appropriate laboratory studies; (2) activity status---~rank 0 = fully active (Karnofsky 90-100), rank 1 = restricted but ambulatory and able to carry out light work (Karnofsky 70-80), rank 2 = out of bed more than half of working hours and capable of self care but ufiable to work (Karnoffsky 50-60), rank 3 = limited to self care with more than half of waking hours confined to bed or chair (Karnofsky 30-40), and rank 4 = completely bedridden (Karnofsky (10-20): and (3) pain status--rank

Nine institutions participated in this study: UCLA School of Medicine, Los Angeles, California--F. Kristian Storm, MD, Sponsor; Good Samaritan Hospital, Portland, Oregon (University of Oregon affiliate)--Harvey W. Baker, MD; Evanston Hospital, Evanston, Illinois (Northwestern University affiliate)--Edward F. Scanlon, MD; LDS Hospital, Salt Lake City, Utah (University of Utah affiliate)--Henry P. Plenk, MD, and John W. Thomson, MD; Baptist Hospital, Pensacola, Florida (University of South Alabama affiliate)--Paul M. Meadows, MD: Baptist Hospital, San Antonio, Texas (Uni-

No. 11

MAGNETIC-INDUCTION HYPERTHERMIA

Storm el a].

2679

0 = pain-free, rank 1 = nonnarcotic analgesics required, were fully trained at the UCLA Medical Center in the rank 2 = oral narcotics required, rank 3 = oral and/or principles and application of this device, ensuring uniintramuscular narcotics required, and rank 4 - intra- form hyperthermia delivery according to the methods venous narcotics required. of the study. None could deviate from or modify these methods, or the standard equipment, without exclusion. All were required to attest to the conformity of treatments Hyperthermia Equiprnent and Training of at their facility and to abide by Department of Health Principl Jnvestigators and Human Services (DHHS) regulations for human We used unaltered, nonmodified original generators subject protection.

and standard electrodes supplied by Henry Medical Electronics. Inc., Los Angeles, California (Magnetrode Thermometry Equipment System Hyperthermia~ Devic~l. All investigators used c~]ibrated thermocouples or This ~ystem creates electromagnetic energy in the thermistors reliable to _+0.1C- read serially throughout form of shortwave-band radiowaves at the ISM (industnal-scientific-amedical) frequency of 13.56 MHz, and treatment during brief periods of radiofrequency (RF)employs either capacitive or inductive applicators to wave interruption, or nonperturbing fiberoptic sensors provide noninvasive, localized hyperthermia.7 The ca- reliable to _+0.1C read continuous~l~ d...u~riffg-ireatment. Acceptable protocol probes included B.ajley Microtherpacitor electrode heating method uses a pair of cooled mocouples, Model Nos. MT 23-29 and IT 14z23"(Bailey contact conductive plates suitably arranged on either side of the volume being treated. On tissues with a high Instrument Co., Saddlebrook, N J); YSI needle thermiswater content, such as muscle, internal organs, or tumors, tors, Series No. 500 (Yellow Springs-Ilastyument Company, Yellow Springs, OH); gallium-arse~de by Clinieffectiveness is reduced by excessive heating of overlaying ,therm, Inc. (Dallas, TX); and phosphor Fluoroptic (Luxsubcutaneous fat. For tumors located subdermally with less than 1 cm of overlaying normal skin and subcuta- tron, Inc., Mountain View, CA-). neous tissue, specific tumor heating was accomplished by incident wave contouring using various-shaped contact Treatment Temperature Monitoring electrodes, as well as surface cooling (3% of cases). Normal tissues: Skin and/or subcutaneous tissue, as Magnetic-loop applicators are self-resonant, noncon- well as oral and/or rectal core temperatures, were meatact circular structures with built-in impedance matching ~fured in all patients to assure that normal tissues recircuitry. The coil is a single turn of a rolled conducting mained within their physiologic thermal tolerance range. sheet that overlaps in a nonconductive manner. The ~tlt~lOrS. Tumor temperatures were assessed whenever area of the overlap and the gap distance provide the safely possible. In each case, individual risk was weighed proper amount of capacitance. The element parameters against potential knowledge gained. Ideal tumor-temwere selected to produce very large circulating currents perature monitoring included the normal tissue-tumor in the loop structure. These currents create a strong interface, as well as radial depths within the tumor, electromagnetic field into which the body or limb is including its peripheral and central regions. Sterile needle, immersed; this provides localized hyperthermia withrut wire, or fiberoptic thermometers, with or without sterile excessive heating of surface tissues. With deep internal locating tubes, were placed aseptically using local anestumors and normal overlaying skin and subcutaneous thetic (1% xylocaine without epinephrine), or surgically tissues of 1 cm or greater thickness, magnetic-loop in and about accessible tumors to be treated. electrodes were used. Since different body parts (thorax, The highest recorded temperature throughout the abdomen, neck) were treated by this method, electrodes tumor that was sustained throughout therapy (i.e., the of different sizes were employed (97% of cases). minimum tumor time-at-temperature) was entered into Patients were treated in a standard clinical setting, our data as the tumor heating capacity. without a screen room, Faraday cage, or other shielding. The hyperthermia system meets or exceeds Federal Dose and Treatment Schedule Communications Commission (FCC~ and American Nation~.l Standards Institute (ANSI) standards for perHyperthermia application (all categories): Patients were sonnel safety. Pursuant to Medical Device Amendments treated within the physiologic normal tissue-temperature of 1976 and 1981, this device was classified as "inves- range, to comfortable tolerance. Treatments were for 30 tigational" and was used for human subjects under a to 60 minutes at tumor-treatment temperature per treatClinical Investigations Exemption approved by the Food ment application, no more than once per day, at 24- to and Drug Administration (FDA). 72-hour intervals. Only awake, responsive patients were All principal investigators (or their representatives) treated, with or without mild sedation.

2680

CANCER June 1 1985

Radiation therapy application: Hyperthermia was given within 4 hours of radiation therapy (XRT), at 24- to 72-hour intervals, throughout the course of XRT. Standard applicable doses and fractions of XRT were employed in accordance with acceptable and common community practice, and were given only to patients who had not received larior XRT to the field of treatment. The total XRT dose varied according to the organ being treated, but did not exceed the standard recommended dosage for that organ. Less-than-standard-dose XRT was employed in accordance with abceptable and common community practice in patients who had failed prior standard-dose XRT for their disease, and where additional XRT was safely feasible. Additional XRT doses varied according to the organ being treated, but did not exceed .standard recommended dosage for that organ. All radiation therapy was administered by a qualified radiation therapist. Chemotherapy application: Drugs were administered immediately before or simultaneously with hyperthermia. Only FDA-approved chemotherapeutic agents, doses, and schedules were used, in accordance with acceptable and common community practice. The chemotherapy in each case was an indicated agent(s) applicable to the cancer being treated, lnvestigational drugs and doses were prohibited. Chemotherapy was administered under the supervision of qualified physicians experienced in the use of antineoplastic agents.. Evaluation During Hyperthermia During h.~erthermia, all patients had their vital signs (pulse, respiration, and blood pressure) and core temperatures serially monitored. Any adverse symptom during ltceatment was noted and investigated. Clinical evaluation of~tiormal tissue in the treatment field was made se~ial13)-~lufihg~ treatment. If the 45C thermal tolerance limit of any normal tissue occurred, or if any adverse reaction was observed, treatment was immediately terminated, and appropriate supportive care was instituted as required. Criteria Jbr Response Patients evaluableJbr therapeutic response: Heat alone: patients completed five or more serial heat treatments, defined as "one course of hyperthermia." Heat plus radiation therapy: patients completed the initial prescribed course of radiation therapy combined with hyperthermia, defined as "one course of thermoradiotherapy." Heat plus chemotherapy: patients completed five or more serial treatments of simultaneous heat and drug administration, defined as "one course of thermochemotherapy."

Tumor response. Complete response (CR) was defir~ed as the disappearance of all evidence of the treated tumor. Partial response (PR) was defined as a decrease of 50% or more in the size of the treated tumor. Minimal response (MR) was defined as a decrease of 25% or more in the sizo~of the treated tumor. Stable disease (S) was defined as no change, i.e., _+25%, in the size of the treated tumor. Since nonviable tumor matrix may be replaced by scar after effective hyperthermia with little or no reduction in size,z growth arrest of a previousl growing tumor was categorized as disease "stabilization." Progression (P) was defined as greater than 25% increase in the size of the treated tumor. Pain response." Pain response was defined as improvement by at least one pain rank. Activity response: Activity response was defined as improvement by at least one activity rank. Duration of response." Duration of response was calculated from the time of initiation of hyperthermia therapy. Treatment response exclusions: Patients unable to complete one course of therapy, for whatever reason, were considered nonevaluable (NE) for therapeutic response.
Assessment of Toxicity

Toxicity was evaluated during and after the completion of a course of hyperthermia at serial intervals (2-4 weeks). Evaluation included physical examination and appropriate blood chemistries for the organ(s) treated. Patients undergoing chemotherapy also had serial complete blood counts (CBC), differential prothrombin time (PT), partial prothrombin time (PTT), and platelet counts. Data Collection and Analysis

All patients were entered into the comphterized database. No patients or data were e~xcluded" because of "inadequote" treatment or "porr" results.-tData-entrY personnel were trained.by the ~hief statistician (D.R.) and data-entry coordinator (A.N.). Additionally, statisticians independently verified the accuracy of data entry. for 5% of the data forms. All analyses were conducted by statisticians not directly connected with~.the trial. Those patients who had one or more cours~ of therapy were considered evaluable for treatment response, and all patients were considered evaluable for toxicity.
Informed Consent Patients understood the investigational nature of this study and gave fully informed consent to participate, in

No. 11

MAGNETIC-INDUCTION HYPERTttERMIA

Storm et al.

2681

accordance with all applicable institutional, state, and federal guidelines. They could withdraw from the study without prejudice.

TABLE 1. Production of Regional Hyperthermia in Tumors

Minimum tumor time-al-temperature in C (30-60 min) <40
40-40.9 41-41.9 42-42.9

No. (%)

Of the 1170 patients entered on study, 960,completed at least one course of therapy and we]~evaluable for tumor response. Table 5 shows the overall response of the entird group. There were. 353 (37%) patients with objective tumor regression (CR + PR + MR) for 1 to 39 months, 313 (33%) with disease stabilization for 1 to 32 months, and 294 (30%) with disease progression. Pain improved overall in 30% of patients, and activity in 21% (Table 6). Interestingly, not only did 228 of 651 (35%) patients with tumor response have pain improvement but 49 of 287 (17%) with disease progression also had improvement in pain status. Overall response did not correlate with tumor histologic type (Table 7), location of the tumor within the body (Table 8), or depth of the treated lesion (Table 9). Overall response did vary by the type of treatment employed (Table 10), as well as by minimum tumor temperature (Table 11). The incidence of regression was higher in those patients treated with combination therProduction of Localized Hyperthermia in Tumors moradiotherapy or thermochemotherapy than in those Heating capacity was determined in 591 tumors; 409 treated with hyperthermia alone. The rare occurrences (69%) had one site monitored and 182 (31%) had two of sustained minimum tumor temperatures of 46C or or more sites monitored. No injuries or deaths were associated with temperature monitoring. In nearly all higher produced a high response rate. When these two variables (i.e., treatment type and minimum tumor cases monitoring occurred during the initial treatment. Table 1 shows the overall result of localized hyper- temperature) were combined (Table 12), combination thermia within tumors. It is emphasized that the data therapy and higher temperaturgs seemed most beneficial. Table 13 shows the overall toxicity in 14,807 treatrepresent the highest temperature mg, asured throughout the tumor, throughout treatment (i.e., the minimum ments of localized hyperthermia and hyperthermia combination therapy. There were 49 (0.3%) surface tissue tumor time-at-temperature for 30--60 minutes). At these injuries, of which only 6 resulted in focal areas of skin tumor temperatures, only, 2 of 924 (0.2%) patients had loss, and 2 (0.01%) systemic injuries. After 1 month of skin temperatures ever recorded at 45C or higher, and thermoradiotherapy to the upper abdomen, one patient only 24 of 932 (3%) had subcutis at or above this developed a peptic ulcer, and 9 months after receiving temperature. Tumor heating capacity did not appear to

Patient and Tumor Profile One thousand, one hundred seventy adults, 55% of whom weye male, ranging in age from younger than 20 years t0~older than 80 years (70% 50-90 years), underwent 14,807 regional hyperthermia treatments. Patients received a median of 10 treatments (average, 12.7: range, 1-160). All patients had a~lvanced" disease a] the time of treatment~ Two hundred thirty (2~) had gad no prior therapy; the others~had had disease progression despite surgery in tit (10%), radiation therapy in 40 (3%), chemotherapy in 306 (27%), or combination therapy in 459 (40%). Seven hundred sixty-three patients (67q) had pretreatment pain, of Whom 568 (74%) required narcotics and 911 (79%) had restricted activity, of whom 550 (60%) had a Karnofsky rating of 50 to 60 or less. Two hundred fifteen (18%) patients had advanced primary, tumors, 303 (26%) had locally recurrent tumors, and 648 (56%) had metastatic disease. There were 707 (61%) adenocarcinomas, 135 (12%) epidermoid carcinomas, 130 (11%) sarcomas, 137 (12%) melanomas, 13 (1%) lymphomas, and 45 (4%) other less common solid tumors. Five hundred twenty (63%) of the tumors were 5 cm or larger in least dimension, and 302 {37%) were less than 5 cm. -." Treated tumors were lo{ated in the head and neck in 77 (7%) patients, the body wall in 79 (7%), an extremity in 42 (4%), a lung in 177 (15%), the liver in 351 (30%), the pelvis in 202 (17%), another intra-abdominal location in 201 (17%), or another body site in 32 (3%). Nine hundred twenty-four (86%) of the tumors were considered deep internal or visceral (below the body wall), and 152 (14%) were deemed superficial (on or in the body wall).

116 (20%)
95 (16%) 142 (24%) 104 (18%)

43-43.9
44-44.9

55 (9%)
23 (4%)

45-45.9 46-46.9
47-47.9

27 (5%) 8 (1%)
0

48-48.9
49-49.9 "

2
1

>_50

9 (2%)

be correlated with tumor histologic.type (Ta .ble2), location within the body (Table 3), depth, dr- siid(Table 4.)~ Results

2682

CANCER June 1 1985

TABLE 2. Tumor Heating Capacity by Histologic Type

Vol. 55

Minimum temp C <40 40-41.9 42-44.9 >-45

Adenoca n = 332 48 (14%) 146 (44%) 117 (36%) 21 (6%)

Epidermoid Ca n = 70

Sarcoma n = 90

Melanoma n = 71

Lymphoma n=9 3 2 4 0

Other n=12 0 9 2 1

22 (31%) 28 (40%) 18 (26%) 2 (3%)

24 (27%) 29 (32%) 25 (28%) 12 (13%)

18 (25%) 25 (35%) 17 (24%) 11 (16%)

~ ~(~p: temperature; adenoca: adenocarcinoma.

TABLE 3. Tumor Heating Capacity by Location

Minimum temp C (:40 40-41.9 42-44.9 >_45,

Head, neck n = 56 ~ 2b (36%) 24 (43%) ~ 8 (14%) 4 (7%)

Body wall n = 70~ 21 (30%) 26 (38%) 15 (21%) 8

Extremity n = 36 9(25/,,) 10 (28%) 10 (28%) 7 (19%)

lntrabdominal n = 93 11 (12%) 41 (44%) 32 (34%) 9 (10%)

Pelvis n = 119 25(21%) 46 (39%) 44 (37%) 4 (3%)

Lung n = 34 6(18%) "17 (50%) 9 (26%) 2 (6%)

Liver n = 161 17(11%) 73 (45%) 58 (36%) 13 (8%)

Other n=13 7 0 6 0

temp: temperature.

thermoradiotherapy, one patient had areas of lung fibrosis exceeding that expected after radiation therapy alone. There was no increase in chemotherapy toxicity reported with simultaneously administered localized hyperttkermia. There were no treatment-related deaths.
Discussion

Hyperthermia has a long and interesting history dating to before the year 1900, and several excellent review texts are currently available.~,s,9 Over the last decade, convincing evidence for the tumoricidal effect of heat has come from basic science laboratories throughout the world. Unfortunately, few trials have evaluated the clinical effectiveness and toxicity of localized hypertherTABLE 4. Tumor Heating Capacity by Depth and Size Depth Minimum temp C <40 40-41.9 42-44.9 ~45 Superficial n = 135 46 (34%) 43 (32%) 28 (21%) 18 (13%) Deep n = 410 67 (16%) 178 (43%) 139 (34%) 26 (6%) <5 cm n = 122 32 (26%) 52 (43%) 29 (24%) 9 (7%) Size >-5 cm n = 395 76 (19%) 153 (39%) 136 (34%) 30 (8%)

temp: temperature. TABLE 5. Overall Response and Response Duration (n = 960) Response Complete regression Partial regression Minimum regression Stabilization Progression N/A: not applicable. No. (%) ~ 85 (9%) 173 (18%) 95 (10%) 313 (33%) 294 (30%) Response duration 1-34 mo (7 mo median) 1-39 mo (4 mo median) 1-15 mo (3 mo median) 1-32 mo (3 mo median) N/A

mia and hyperthermia combination therapy using standard doses of radiation theralyy, or chemotherapy for human tumors throughout the ~dy; none has approached the magnitude of this national cooperative study. This multi-institutional Phase 1-II trial was undertaken to determine the potential benefit and safety of localized hyPerthermia as administered by physicians in a standard clinical setting for patients with advanced solid cancer. Evaluation of deep-seated internal tumors was possible due to the de.v._elopment of effective magnetic-induction shortwave- :applicators.7 However, regardless of the method of heating employed, the knowledge of tumor response derived from this study can be applied universally to use localized hyperthermia. Nine institutions participated in this study. Each principal investigator was trained similarly in hyperthermia application by the sponsor. At most facilities, a "hyperthermia tearrr~ was formed, and included representatives from each of the major oncologic specialities: surgery, radiation therapy, medical oncology, and nursing. Treatments were generally given by a trained registered nurse under the direct supervision of a qualified medical doctor. Early in the investigation, patients were often observed in the hospital for 24 to 28 hours after treatment; however, it soon became clear that this was usually unnecessary, and during the last several years

TABLE 6. Response of Patients with Pain and Reduced Activity Status Status Pain Activity No. 938 952 Improved/duration of ihaprovement 277 (30%) 1-34 mo (2 mo median) 201 (21%) 1-39 mo (3 mo median) Not improved 661 (70%) 751 (79%)

No. 11

MAGNETIC-INDUCTION HYPERTHERMIA

Storm et aL

2683

TABLE 7. Response by Tumor Histologic Type Histologic type Adenoca Epidermoid Ca Sarcoma Melanoma Lymphoma Other CR 44 (6%) 22 (19%) 2 (2%) 15 (13%) 0 2 ~6%) PR 96 (16%) 31 (27%) 21 (21%) 16 (14%) 1 8 (25%) MR 62 (11%) 14 (12%) 8 (8%) 5 (4%) 1 3 (9%) S 207 (35%) 29 (26%) 31 (31%) 27(23%) 5 14 (44%) P 178 (30%) 17 (15%) 37 (37%) 54 (46%) 3 5 (16%)

CR: complete regression; PR: partial regression; MR: minimum

regression; S: stabilization; P: progression; Adenoca: adenocarcinoma; Ca: carcinoma.

TAm-E 8. Response by Tumor Location Location Head and neck Body wall Extremity Lung Liver Intraabdominal Pelvis Other CR 19 (29%) I1 (17%) 7 (21%) 9 (6%) 8 (3%) 9 (6%) 17 (11%) 3 (t2%) PR 19 (29%) 15 (23%) 10 (30%) 15 (10%) 32 (11%) 30 (19%) 44 (28%) 7 (28%) MR 4 (6%) 5 (8%) 4 (!2%) 23 (16%) 19 (6%) 21 (13%) 16 (10%) 2 (8%) S 13 (20%) 13 (20%) 4 (12%) 57 (39%) 117 (38%) 44 (28%) 56 (36%) 6 (24%) P 11 (17%) 20 (31%) 8 (24%) 43 (29%) 128 (42%) 52 (33%) 23 (15%) 7 (28%)

CR: complete regression; PR: partial regression; MR: minimum regression; S: stabilization; P: progression.

ve
~ty ns ~rs

the majority of treatments were administered on an outpatient basis. Since this trial was designed to test a new treatment modality, only patients with locally advanced primary, recurrent, metastatic, or refractory disease were included. This was with full knowledge that such a population of patients would unfavorably bias response against the

treatment. Although, principal investigators could choose whether to use hyperthermia alone or in combination with radiation therapy and/or chemotherapy based upon their best clinical judgement, no proven method of
cancer therapy was withheld because of this study. Patients understood the investigational nature of the trial, and gave fully informed consent to participate in

TABLE 9. Response by Tumor Depth and Size "CR Depth ..-~ "[ " ~" , ~ , ~ ~2.~. (!9%i Superficial -Deep -49 (6%) Size <5 cm 36 (12%) >-5 cm 36 (7%) PR 28 (23%) 125 (17%) 38 (13%) 95 (18%) MR 11 (9%) 70 (9%) 16 (5%) 64 (12%) S 26 (21%) 274 (36%) 101 (33%) 177 (34%) P 35 (28%) 238 (31%) 111 (37%), 146, (28%)

CR: complete regression: PR: partial reg~ression: MR: minimum regression: S: stabilization; P: progression. TABLE 10. Response by Treatment Type Treatment type Heat alone Heat + STD XRT Heat + <STD XRT Heat + IV chemo Heat + IV chemo-PF Heat + IA chemo Heat + IA chemo-PF Heat + STD XRT + chemo Heat + <STD XRT + chemo n = 142 n = 227 n = 90 n = 95 n = 165 n = 99 n = 46 n = 79 n = 17 CR 4 (3%) 48 (21%) 5 (6%) 4 (4%) 3 (2%) 5 (5%) 2 (4%) 14 (18%) 0 PR 16 (11%) 53 (23%) 20 (22%) 17 (18%) 17 (10%) 18 (18%) 2 (4%) 24 (30%) 6 (35%) MR 12 (9%) 36 (16%) 10 (11%) 6 (6%) 13 (8%) 5 (5%) 3 (7%) 8 (10%) 2 (12%) S 27 (19%) 71 (31%) 42 (47%) 43 (45%) 43 (26%) 40 (40%) 21 (46%) 19 (24%) 7 (41%) P 83 (58%) 19 13 (14%) 25 (26%) 89 (54%) 31 (3!%) 18 (39%) 14 (18%) 2 (12%)

CR: complete regression: PR: partial regression: MR: minimum regression; S: stabilization; P: progression: STD XRT: standard-dose radiation therapy: <STD XRT: less-than-standard,lose ratiation therapy

was combined with hyperthermia in patients who had received prior radiation: IV: intravenous: IA: intra-arterial, Chemo-PF: prior failure on that same chemotherapy alone.

2684
TABLE I 1. Minimum temp C <40 40-40.9 41-41.9 42-42.9 43-43.9 44-44.9 45-45.9 46-50 CR 5 (5%) 14 (18%) 12 (10%) 6 (6%) 4 (8%) 0 3 (14%) 4 (21%)

CANCER June 1 1985

Response by Minimum Tumor Temperature PR 12 (13%) 17 (22%) 36 (31%) 26 (28%) 10 (21%) 4 (27%) 7 (32%) 10 (52%) MR 11 (12%) 5 (6%) 15 (13%) 12 (13%) 5 (10%) 1 (7%) 2 (9%) 2 (11%) S 31 (34%) 16 (20%) 26 (22%) 23 (24%) 11 (23%) 8 (53%) 5 (23%) 3 (16%) P 32 (35%) 27 (34%) 28 (24%) 27 (29%) 18 (38%) 2 (13%) 5 (23%) 0

Vok 55

temp: temperature: CR: complete regression; PR: partial regression; MR: minimum regression: S: stabilization; P: progression.

acaordance whh all applicable institutional, state, and federal requirements. The heat delivery system was tested finder a clinical investigational device exemption approved by the FDA. The Magnetrode System Hyperthermia Device proved suitable as an effective heat delivery system for both surface and internal solid tumors] As shown in Table 1, potentially significant minimum tumor temperatures could be sustained throughout the tumor during .treatment. Unfortunately, great variation in tumor heat retention occurred, probably due to preserved vasoregulatory mechanisms present in many tumorsm that could not be overcome by this method of heating. However, treatments continued, regardless of the absolute tumor temperature achieved, in hopes that even low doses of heat might prove beneficial and that otherwise undesirable increases in blood flow might impro,~ drug delivery and oxygenation to the tumor, thereby enhancing the effectiveness of chemotherapy or radiation therapy. Interestingly, there were no significant differences in tumor heating capacity based upon tumor histologic

type, location, or size (.Tables 2, 3, and 4). When the data were analyzed by primary tumor origin (e.g., breast, thyroid, colon, bladder, etc.), we determined the number of cases to be too small to warrant specific conclusions. Table 5 summarizes ifi~ ~ov~ei~a~i response in 960 evaluable patients who complete~:at lea~i@ne course of treatment. The tumor regressions in 353 (37%) of these patients with advanced disease .were gratifying, particularly considering that 85 (24%) oiRche regressions were complete responses lasting from 1 to 34 months (median, 7 months). Disease stabilization was included as a potential response parameter since it appears that many effectively heated tumors may not decrease in size.27~ Even with substantial tumor cell kill, the background stroma may be preserved and subsequently become fibrotic. Also, tumor vascular thrombosis, which frequently accompanies hyperthermia, may prevent rapid absorption of necrotic tumor debris (similar to a healing infarction), such that no appreciable change in volume occurs. Documented growth arrest of enlarging tumors that is associated with stabilization or improvement in consti-

TABLE 12. Tumor Regression* by Treatment Type and Minimum Tumor Temperature (Number of Regressions
per Number of Tumors at That Temperature)

Minimum temp C <40 40-40.9 41-41.9 42-42.9 43-43.9 44-44.9 45-45.9 46-50

Heat alone 2/16 0/15 6/25 3/15 2/13 0/1 2/9 12/12

Heat + STD XRT 14/27 16/18 23/26 15/18 4/7 3/5 1/2 ~ 1/1

Heat + <STD XRT 5/19 9/15 5/8 5/12 1/2 1/3 1/1

Heat + IV chemo 2/7 2/4 10/15 4/8 0/3 0/3 1/1 0/2

Heat + IA chemo 0/1 0/4 6/15 5/9 2/3 1/2 0/1

Heat + STD XRT + chemo 3/9 5/7 3/5 5/6 5/5 2/2

Heat + <STD XRT + chemo 1/3 2/2 1/2 1/1 0/1 1/1 1/1

* Tumor regression: CR + PR + MR. CR: complete regression; PR: partial regression; MR:. hainimum regression; S: stabilization; P: progression; STD XRT: standard-dose

radiation therapy; <STD XRT: less-than-standard-dose radiation therapy was combined with hyperlhermia in patients who had received prior radiation; IV: intravenous; IA: intra-arterial.

No. 11

MAGNETIC-INDUCTION HYPERTHERM1A

Storm el al.

2685

tutional status (pain and activity), with no new lesions in the treated field, may be an important criteria of response to hyperthermia. It is recognized that some patients with advanced cancer treated with radiation therapy or chemotherapy do not respond, even though their tumor growth seems to stabilize, either because of decreased growth as the tumor increases in volume alonK.~v~e Gompertzian curve or for other unknown reasons. However, in contrast, the patients reported in this study who had temporary disease stabilization subsequently did demonstrate tumor growth and]or a significant decline in constitutional status, indicating that -their tumors were still in th~)ogarithmic phase of tumor growth) Thus, the313 (33%) incidences of disease stabilization, for l~ to 32 months (median, 3 months) probably elobs not reflect a placebo effect, and may be clinically important when using hyperthermia. The 30% incidence of improvement in pain and 21% improvement in activity (Table 6) were also gratifying, particularly considering that 74% of the patients with pain required narcotics and 60% of the patients with restricted activity had a Karnofsky status of 50 to 60 or less. Interestingly, despite continued tumor progression and lack of an objective tumor response to treatment, 49 of 287 (17%) such patients with pain and 23 of 290 (8%) such patients with restricted activity also had improvement in symptoms. The reason for this surprising subjective response in the absence of objective tumor regression or stabilization is unknown. Table 7 shows the response to treatment by tumor histologic type. Tumor regressions were observed in all major tumor types: adenocarcinoma (35%), epidermoid carcinoma (59%), sarcoma (31%), melanoma (31%), lymphoma (20%), and other solid tumors (40%). Epidermoid carcinoma seemed to be particularly responsive to treatment. The overall response to treatment was independent of the location of the tumor within the body, as well as the depth or size of the lesion (Tables 8 and 9). This was probably due more to the heat delivery system employed than to the specific characteristics of the tumor site or to the tumor dimension. Two of the authors (J.D.K. and E.F.S.) recently reported an average temperature of 43.9C using this heating system for tumors of the pancreas,~2 the most centrally located organ evaluated to date. Overall response varied by the type of treatment employed (Table 10). With heat alone, there were 32 of 142 (.23%) tumor regressions; the response to combined standard-dose radiation therapy or chemotherapy was 137 of 227 (60%) and 55 of 194 (28%), respectively. Even more significantly, subsequent tumor regression occurred in 35 of 90 (39%) patients treated with heat

TABLE 13. Treatment Toxicity Toxicity Local First-degree burn 17 Second-degree burn 26 Third-degree burn 6 Systemic Peptic ulcer 1 Lung fibrosis 1 Deaths (treatment-related) Incidence

49/14,807 (0.3%)

2/14,807 (0.01%) 0/14,807 (0%)

combined with less-than-standard-dose radiation therapy after they failed standard-dose radiation alone, and in 40 of 211 (19%) patients who were secondarily treated with combination thermochemotherap~-after documented disease progression, o~ "th~tt S~rfi~ chemotherapeutic agent(s) alone. These results i~icate ibmt" hyperthermia may be a useful clinicgl adjuvant to radiation therapy and chemotherapy where a syuergistic or additive response has been predicted.34 Unfortu~l~ely, the results of this Phase 1-11 study do not shed further light on the mechanism(s) of interaction of these treatment modalities. Moreover, since patients received a variety of the commonly employed chemotherapeutic agents most appropriate for their disease, including single drug and combinations of doxorubicin, cisplatin, bleomycin, nitrosoureas, 5-fluorouracil, cytoxan, low- and high-dose methotrexate, mitomycin C, velban, vincristine, imidizo,lp: carboximide, phenylalanine mustard, and others, administered by various oral (PO), intravenous (IV), and intra-arterial (IA) routes in many different types of tumors, no specific conclusions could be drawn regarding the efficacy of any specific thermochemotherapy regimen. As previously predicted in a large number of in vitro, in vivo, and pilot clinical investigations,~~9 tumor response correlated with the minimum tumor temperature
achieved (Table 11). At 40 to 45C, the overall incidence

of tumor regression ranged from about 40% to 50%. There seemed to be a break in response at 46C and above, with more than 80% tumor regression at such high temperatures. Most interestingly, an unexpected and significant incidence of tumor regression was observed at minimum tumor temperatures of less than 40C (Tables 11 and 12). Analyzing this group separately, less than 40C alone resulted in 2 of 16 (13%) tumor regressions, less than 40C heat plus radiation therapy resulted in 19 of 46 (41%) regressions, less than 40C heat plus chemotherapy resulted in 3 of 17 (18%) regressions, and less than 40C heat plus radiation therapy and chemotherapy resulted in 4 of 12 (33%)

2686

CANCER June 1 1985

Vol. 55 ~

regressions. Thus, hyperthermia treatment should not be abandoned based upon some preconceived desirable tumor temperature, since clinical responses may occur at quite low temperatures. Tumor response was evaluated by both the type of treatment employed and the minimum tumor temperature achieved (Table, 12). These results show that although hyperthermia alone may be beneficial (particularly in the rare incidences of superheating), most favorable responses occurred when localized hyperthermia was combined with radiation therapy or chemotherapy. Of special importance is the fact that combination localized hyperthermia treatment was so well tolerated by these sick patients with such advanced disease. Table 13 shows the rare treatment-related complications. Significant surface-tissue injury occurred in only six (0.04%) patients, and systemic injuries in two (0.01%). Both systemic injuries occurred in patients receiving combined thermoradiotherapy: After l month of treatment, one patient developed a peptic ulcer that healed without sequellae after discontinuing treatment; 9 months after treatment, one patient with a lung tumor manifested areas of lung fibrosis considered more severe than that expected from radiation therapy alone. In the 14,807 treatments administered at nine institutions, there were no treatment-related deaths. However, it should be emphasized that although all of the patients in this trial had advanced or refractory cancer, this study group was a selected population based upon knownpotentially predictable side effects of heat therapy in certain situa- tions. For ex~t~ple, patients with any of the following conditions werenot treated: (l) circulatory compromise in the treatment field with impaired or damaged normal tissue ,~va~regulafion, and a reduced capacity to augment blood~10w and~dissiI~ate hbat that could result in normal tissue injury (incl~dihg extensive surgical scarring, brawny edema, vascular disease, radiation fibrosis, and split-thickness or pedicle skin grafts); (2) significant tumor fistulae communicating with the aerodigestive tract, where massive tumor necrosis could potentially result in fistula enlargement or increased infection; (3) impaired tissue sensation from disease, heavy sedation, or anesthesia, since burns could occur without patient awareness; and (4) cardiac pacemakers or large ferrous (iron-containing) metal prostheses (i.e., total joints, femoral rods, etc.). This study, however, confirmed that nonferrous metals in the field of treatment, including commonly employed small internal surgical vascular clips and implantable chemotherapy infusion pumps (e.g., Infusaid, Infusaid Corp., Norwood, MA) could be safely treated.

It should also be understood that the principal investigators participating in this study were highly trained in the application of hyperthermia using thi~ heat delivery system. Whether the low toxicity observed in this study can occur with other hyperthermia delivery systems is unknown. ~, This Phase I-II trial clearly indicates that localized hyperthermia has a role in the palliative treatment of human advanced solid cancer, particularly when combined with radiation therapy or chemotherapy. In such patients, hyperthermia treatment with this technique should no longer be considered investigational. However, since this treatment modality is not as yet known to be curative, and as responses have generally been temporary (range, 1-39 months; median, 3-7 months), localized hyperthermia should not be used in place of established potentially curative methods of cancer therai~y. Moreover, although implied by these results, it remains to be determined in a prospective randomized Phase III trial whether localized hyperthermia in combination with radiation therapy or chemotherapy is more effective than these modalities alone. The outcome of this multiinstitutional national cooperative Phase I-II study now warrants such comparative treatment!.ffials, as the tumoricidal effects of localized hyperth~rmia appear real and potentially important in the palliative treatment of human solid cancer.
Conclusions

Localized hyperthermia alone or in combination with radiation therapy or chemotherapy was administered to 1170 patients during 14,807 treatments at nine institutions in a national cooperative trial. In the study population, comprised of patients with advanced primary, recurrent, metastatic, or refractory solid cancer, the vast majority of whom had failed one or more prior therapies, of 960 evaluable patients, the overall response was complete regression in 85 (9%), partia.I regreksion in 173 (18%), minimal regression in 9,5~ (10%), arid growth stabilization in 313 (33~o.). Tumor response occurred from 1 to 39 months (median, 3-7 months) and was associated with improvement in pain and abtivity in a significant number of patients. Tumor regressions were more frequent with combination thermoradiotherapy or thermochemotherapy than with hyperth~rnrf~i alone. Moreover, a significant number of tumor regressions were observed at low temperatures (<40C), as well as when heat was combined with less-than-standard doses of radiation therapy in patients who failed radiation alone, and when heat was added to previously ineffective chemotherapy alone.

No. 11

MAGNETIC-INDUCTION HYPERTHERMIA

Storm el al.

2687

This Phase I-II trial shows that localized hyperthermia using the Magnetrode System Hyperthermia Device is safe and palliative in a substantial number of patients with advanced cancer. Whether this form of treatment is equivalent to or better than other forms of palliative therapy was not the object of the study. However, the overall result of this trial indicates that localized hyperthermia using-this heat delivery system should no longer be regarded as investigational, and that Phase 111 trials are now warranted.
REFERENCgS 1. Storm FK, ed. Hyperthcrmia in Cancer Therapy. Bbston: GK Hall, ~983. 2. Storm FK,Harrison ~WH, Elliott RS, Morton DL. Normal tissue and solid tumor effects of hyperthermia in animal models and clinical trials. Cancer Res 1979; 39:2245-2251. 3. Kim JH, Antich P, Ahmed S, Hahn EW. Clinical experience with radiofrequency hyperthermia. J Microwave Power 1981; 16:193204.

4. Overgaard J. Influence of sequence and interval on the biologic response to combined hyperthermia and radiation. Natl Cancer lnst Monogr 1982; 61:325-332. 5. Hahn G. Potential for therapy for drugs and hyperthermia. Cancer Res 1979; 39:2264-2268. 6. Marmor JB. Interaetions of hyperthermia and chemotherapy in animals. Cancer Res 1979; 39:2269-2276. 7. Storm FK, Harrison WH, Ellion RS, Morton DL. Clinical RF hyperthermia: A new approach to cancer therapy. IEEE Trans M77" 1982; 30:1149-1158. 8. Hahn GM, ed. Hyperthermia and Cancer. New York: Plenum Press, 1982. 9. Hornbeck N, ed. Hyperthermia and Cancer: Human Clinical Trial Experiences. Boca Raton, Horida: CRC Press, 1983. 10. Olch AJ, Kaiser LR, Silberman AW, Storm FK, Graham LS, Morton DL. Blood flow in human tumors during hyperthermia thera v: Demonstration of vasoregulatio.~ and an applicable physiological ~:lel. J Surg Oncol 1983; 23:125-" 132. 11. Parks LC, Smith GV. Systemic hyl~erthermia by extracorporial induction: Techniques and results. In: Storm FK, ed. Hyper~hermia in .Cancer Therapy. Boston: GK Hall, 1983; 407-446. 12. Khanderkar JD, Scanlon EF, Garces RM, Prasad G, Eawrence GA. A Phase II study with radiofrequency hyperthermia,.radiation, and chemotherapy in carcinomas of the pafidr~as and- stomach (Abslr 615). Proc Am Assoc Cancer Res 1983:24:155.