1903 COMMENTARY See referenced commentary on pages 1900 –2, editorial on pages 1887–9, and original article on pages 2143–9, this issue. Details on the design and results of the Phase II trials have already been published.1,5,6 The Unconventional Di Bella Cancer Treatment Supported by a specific act of the Italian Parliament. The committees and members of the Italian Study Group for the Di Bella Multitherapy Trials are as follows. Research Coordinating Center: D. Greco, R. Raschetti, B. Caffari, F. Chiarotti, R. Da Cas, B. De Mei, G. Di Giovambattista, M. Maggini, F. Menniti-Ippolito, S. Modigliani, P. Popoli, P. Ruggeri, S. Spila-Alegiani, C. Tomino, G. Traversa, P. Bruzzi, T. Gamucci. Steering Committee: G. Benagiano, D. Amadori, P. Bruzzi, E. Buiatti, E. Ciranni, F. Cognetti, G. Colucci, P.F. Conte, G. Di Bella*, D. Greco, S. Iacobelli, F. Mandelli, E. Marubini, M. Massotti, S. Monfardini, F. Oleari, R. Raschetti, G. L. Sannazzari, L. Tomatis, U. Veronesi. (*Dr. G. Di Bella, the son of Dr. L. Di Bella, only participated in the first two meetings of the Steering Committee. On many occasions, he strongly disagreed with the performance and results of the trials. No part of the current article has been discussed with or endorsed by Dr. G. Di Bella.) International Review Board: P. Calabresi (Rhode Island), F. Cavalli (Bellinzona, Switzerland), P. Kleihues (Lyon Cedex, France), J.G. Mc Vie (London, UK), H. Pinedo (Amsterdam, The Netherlands), K. Sikora (Lyon Cedex, France), T. Tursz (Villejuif Cedex, France). Independent Endpoint Evaluation Committee: A. R. Bianco, R. Labianca, P. L. Rossi Ferrini, G. Simonetti, A. Sobrero. Investigators: Ancona—R. Cellerino, S. Antognoli, R. Berardi, R. Bracci, P. Lippe, F. Pulita; Aosta—F. Di Vito, D. Martinod, M. Musi; Aviano—A. Veronesi, R. Lazzarini; Bari—G. Colucci (Principal Investigator [PI]), V. Attolico, F. Giotta, M. Longo, E. Maiello; Bologna—S. Tura, F. Gherlinzoni, M. Tani; Bolzano—H. Amor, P. Borona, F. Girardi, C. Graiff; Cagliari—G. Broccia, M. G. Corona, A. Desogus, V. Mascia, S. Pasqualucci; Catania—G. Failla, M. R. Aiello, S. Cordio, P. Finocchiaro. Chieti—S. Iacobelli (PI), E. Melena, M. De Tursi, M. T. Scognamiglio, N. Tinari; Forlı̀—D. Amadori (PI), M. Maltoni, O. Nanni, F. Scardovi, P. Serra; Genoa—R. Rosso, F. Boccardo, A. Barp, T. Coli, M. Grimaldi; Milan—N. Cascinelli, L. Ascani, E. Bajetta, P. Bidoli, A. Cassata, D. De Candis, W. Giannessi, G. Procopio; Milan—S. Di Donato, A. Boiardi, D. Cerri, M. Eoli, A. Silvani; Milan—U. Veronesi (PI), M. Colleoni, N. Rotmensz; Naples—S. Monfardini (PI), G. Comella, A. De Matteis, A. Gravina, C. Gridelli, F. Perrone, M. R. Salzano, C. Sandomenico; Naples—C. Battista, G. Iodice, R. Fico, A. Nicchia; Padua—M. Lise, G. L. De Salvo, E. Di Lenardo, G. Iadicicco; Palermo—B. Agostara, I. Cafarelli; Perugia—M. Tonato, F. Radicchi, S. Sorbolini, O. Taschini; Pisa—P. F. Conte (PI), A. Carmignani, L. Del Mastro, A. Gen© 1999 American Cancer Society A Reflection on the Italian Experience Giuseppe Traversa, M.D.1 Marina Maggini, D.Biol.1 Francesca Menniti-Ippolito, M.Sc.1 Paolo Bruzzi, M.D.2 Flavia Chiarotti, M.Sc.1 Donato Greco, M.D.1 Stefania Spila-Alegiani, D.Biol.1 Roberto Raschetti, M.Sc.1 Giuseppe Benagiano, M.D.1 for the Italian Study Group for the Di Bella Multitherapy Trials 1 Istituto Superiore di Sanità, Rome, Italy. 2 Istituto Nazionale Ricerca sul Cancro, Genoa, Italy. O n November 13, 1998, the great hope raised among Italians by the multidrug anticancer treatment scheme developed by a Modena physiologist, commonly known as Multitrattamento Di Bella (MDB, Di Bella Multitherapy), received a substantial blow. On that date, the government-sponsored clinical trials were concluded, with completely negative results.1 It was almost 1 year from the time when the issue blasted to public attention, although the MDB was not a new discovery. Indeed, Luigi Di Bella had begun to treat cancer patients more than 20 years before, using a combination of different substances; his multitherapy evolved over the years and the full list of substances that he used throughout this period comprised more than 100 drugs. Nonetheless, the compounds most frequently used in the last 10 years, constituting the backbone of what became known as MDB, were only a handful: somatostatin (or its synthetic equivalent, nari, C. Orlandini; Potenza—L. Manzione, A. Dinota, G. Lombardi; Reggio Calabria—G. Gasparini, M. G. Arena, M. Fanelli, F. Modafferi, A. Morabito; Rome—F. Mandelli (PI), P. Fazi, F. Mauro, M. Martelli, C. Ricci, M. Vignetti; Rome—F. Cognetti (PI), A. D’Alessio, S. De Marco, V. Ferraresi, T. Gamucci, D. Giannarelli, E. Magnani, P. Rellecati, E. M. Ruggeri, M. Zeuli; Turin—G. L. Sannazzari (PI), M. Borgognone, R. Chiò, A. Gasco, R. Ragona, M. Sacco, R. Soffietti, R. Verna, M. Tessa; Turin—A. Pileri; Trento—E. Galligioni, A. Lucenti, F. Moltrer, F. Paolazzi. Muzzi, Director), for the production of the melatonin tablets and retinoid solution and the Group for the supervision and control of custom-made drugs of the Istituto Superiore di Sanità, Roma (E. Ciranni, Director; R. Alimenti, S. Alimonti, G. Cavazzutti, L. Gagliardi, B. Gallinella, G. Incarnato, F. La Torre, A. Mosca, G. Multari, L. Turchetto, L. Valvo). The authors thank the Stabilimento Chimico Farmaceutico Militare, Florence, Italy (Colonel G. Received April 26, 1999; revision received July 22, 1999; accepted July 29, 1999. Address for reprints: Giuseppe Traversa, M.D., Department of Epidemiology and Biostatistics, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy. 1904 CANCER November 15, 1999 / Volume 86 / Number 10 octreotide), melatonin, bromocriptine, and a “syrup” containing vitamins and retinoic acid; in most cases, Dr. Di Bella added a cytotoxic agent, usually cyclophosphamide or hydroxyurea. Today, several months after the Italian public was informed that MDB was ineffective, it seems appropriate to reflect on the entire experience; the exercise should be of interest to the scientific community at large and the oncologist in particular. Specifically, in this commentary we propose to analyze what transformed a nonconventional therapy, recommended by a physician operating on a solo basis in a private office, into a national priority; to revisit the discussions that took place in Italy among oncologists, politicians, prosecutors, and magistrates; and to analyze the role of the media in promoting and diffusing this alternative anticancer treatment. We will also briefly describe how the trials were conducted under almost unbearable public pressure and which were the principal findings; what controversies arose while the trials were being carried out; and what issues were debated after their conclusion. The Background of the Di Bella Case Many different arguments have been put forward to explain why, more than 20 years after Dr. Di Bella started proposing his cure for all cancers, at the end of 1997 the public and the mass media suddenly and unexpectedly started paying sustained attention to this physiologist and his treatment (Table 1). In the background is the chronic difficulty that oncologic centers have in providing adequate supportive care to critically ill patients: too often, patients and their relatives are left alone to face the disease. This situation, however, cannot alone explain what happened. A first important issue was the cost of Dr. Di Bella’s treatment for patients. MDB, unlike all other approved anticancer treatments, could not be paid for by the National Health Service (NHS) and, because of the price of somatostatin, the financial burden on patients and families was often unaffordable: 1 month of treatment amounted to some 7500 Euros. This led to the creation of associations of patients and their relatives, which, by requesting the Government to eliminate this “clear discrimination” between the poor and the rich, galvanized public opinion. At this stage the role of media became more and more critical and started a self-feeding chain reaction. The leading anchorman of a national commercial television networks conducted, on November 16, 1997, the first program entirely dedicated to the Di Bella case. According to statistics kept by the Osservatorio della comunicazione radiotelevisiva di Pavia (Obser- vatory of Radio and Television Communications) during the period between December 16, 1997, and January 17, 1998, the daily proportion of prime time news devoted to the Di Bella case reached 36% (the average share was 8.4%). In most instances, journalists praised the MDB (between 47% and 83% of the total time allotted to this topic on the 6 most popular television networks).2 The opinion of all those who appeared as guests on television programs was also strongly polarized. By and large, Dr. Di Bella was praised for his humanity and the quality of his contacts with patients. The percentage of those interviewed who expressed a positive opinion ranged between 91% and 100%, depending on the network.2 During the period mid-December 1997 through mid-January 1998, the number of published newspaper articles about MDB amounted to more than 100 per newspaper.2 With this coverage, no one should be surprised at the tremendous impact the MDB treatment had on cancer patients. The number of those who requested the treatment rose exponentially; patients who were denied the costly treatment appealed to local magistrates to force the NHS to pay for it. In December 1997, for the first time, a local magistrate ordered a local health authority to provide MDB to a patient free of charge. In the following months, several hundred patients (no exact number is available) received the MDB on the basis of courts’ actions. The mounting tide of demand for the right to MDB treatment contributed to the creation of a larger movement, that for “a generalized freedom of treatment,” and this immediately became a political issue. Dr. Di Bella claimed he could cure (and that he had indeed cured) up to 100% of his patients with a variety of cancers, and that even terminal patients could improve with his treatment. In interview after interview, he expressed the opinion that the absence of any published evidence was not a weakness,3 because, he claimed, he had been ostracized by his own university years previously as well as by members of the “oncology clique,” who were eager to protect their links with the pharmaceutical industry. In this scenario, the profound skepticism expressed from the beginning by the scientific community completely backfired. Far from being considered a rational point of view, the position held by the top scientists and oncologists was often considered by many to be further proof of the impermeability of the establishment to anything from the outside. As the popularity and credibility of Dr. Di Bella grew faster and faster, that of official medicine reached its nadir. During the summer and autumn of 1997, the Italian National Committee for Drugs (Commissione Unica del Farmaco) rejected a request to release so- The Di Bella Case/Traversa et al. 1905 TABLE 1 Chronology of the Di Bella Case 1963 1969 1971 1977 1981 1997 1997 1997 1997 1997 1998 1998 1998 1998 Dr. Di Bella starts his experiments in his private laboratory, mainly using melatonin to treat patients affected by various diseases. A communication on the effect of melatonin on platelet turnover in rats is presented at the 45th Congresso della Società Italiana di Biologia Sperimentale, Sassari, Italy. This is the year for which the first patient record is present in Dr. Di Bella’s archive. Dr. Di Bella adds somatostatin to his protocol of therapy. Dr. Di Bella claims to have obtained good results in treating patients with various cancer types using somatostatin, bromocriptine, melatonin, and cyclophosphamide (Abstract book, 2nd International symposium of somatostatin, Athens, Greece, June 1–3). July 17—Two associations of patients organize and publicize a conference of Dr. Di Bella (Hotel Excelsior, Rome). August 5—The Italian National Committee for Drugs refuses to include somatostatin in their list of drugs for compassionate use (for all cancer patients) paid for by the National Health Service (NHS). September 18—Cancer patients demonstrate against the Minister of Health, demanding freedom of treatment and MBD paid for by the NHS. November 16—The first television program entirely dedicated to Dr. Di Bella is broadcast. December 16—A local magistrate orders a local health authority to treat a cancer patient with MDB without charge. The fact was widely reported by Italian national newspapers. January 9, 10—Two Italian regions (Puglia and Lombardia) decide on free administration of MDB. January 14—The Italian National Cancer Advisory Committee and Dr. Di Bella reach an agreement on the decision to conduct trials to test MDB anticancer activity. February 15—Cancer patients demonstrate in St. Peter’s Square during the Pope’s Sunday prayer. February 17—The Italian Parliament approves specific legislation authorizing the conduct of the trials. matostatin and octreotide for the compassionate treatment of patients with all cancer types and, thus, to provide them free of charge. At the moment of the study, the two drugs were provided free of charge only for selected indications (for which scientific evidence was available): somatostatin for severe hemorrhage of the upper gastrointestinal tract and octreotide for gastrointestinal and pancreatic endocrine tumor syndromes. In December 1997, Dr. Di Bella and his followers were asked to submit to the National Cancer Advisory Committee clinical records of patients who were considered to have been successfully treated. After an initial refusal, about 70 clinical records were submitted. When carefully examined, these records provided little evidence to support the effectiveness of MDB. This finding was confirmed, several months later, when the results of a study conducted with the archive of patients treated by Dr. Di Bella was presented. By January 1998, hundreds of patients were being treated with the MDB and an even greater number was pressuring the Minister of Health to receive the treatment free of charge. Then, under the combined pressure of patients, media, local magistrates, and politicians, the negative initial official position became untenable and had to be gradually reconsidered. At that stage, the Minister of Health stated publicly that she would consider the possibility of a free administration of the MDB, but only after the positive conclusion of appropriate clinical trials.4 At the end of January, after several meetings between Dr. Di Bella and the National Cancer Advisory Committee, a consensus was reached to recommend that a series of Phase II clinical trials be conducted to test the anticancer activity of MDB. A written agreement was signed that described the type of cancers to be included in the trials, the procedures to be performed, and the standardized regimen to be utilized. In spite of the fact that Dr. Di Bella himself signed the document, consensus on the decision to conduct the trials was far from unanimous. On the one hand, the majority of Di Bella supporters were against the trials, fearing that they were a diversion intended to postpone the time of the free availability of the combination treatment for all patients. They went public in stating that because of bias or conflicts of interest, oncologists would not give MDB the opportunity of a fair evaluation. On the other hand, a number of physicians and scientists expressed the opinion that the decision to conduct 10 clinical trials in the absence of any published evidence was unethical. In the end, the decision to carry out the clinical trials received great support from the Italian oncology community and the National Cancer Advisory Committee; the participation of the most prominent oncologic centers in Italy was therefore assured. Preparations for the trials continued, and on February 17, 1998, the Italian Parliament approved a specific legislation authorizing the conduct of the studies. The Istituto Superiore di Sanità (Italian National Institute of Health) was given the responsibility of organizing and coordinating the conduct of the trials.5 The Trials A total of 11 Phase II clinical trials were planned. Ten addressed eight different types of cancers; in addition, 1906 CANCER November 15, 1999 / Volume 86 / Number 10 one study was specifically designed to evaluate the MDB in patients (affected by different tumors) who were critically ill. The following cancers were studied: non-Hodgkin lymphoma; chronic lymphoid leukemia; breast carcinoma in women age $70 years who were eligible for surgery; breast carcinoma with Eastern Cooperative Oncology Group performance status [ECOG PS] 0 –2); breast carcinoma with ECOG PS 3– 4; nonsmall cell lung carcinoma; colorectal carcinoma; pancreatic carcinoma; head-neck and esophageal carcinoma; and glioblastoma. The critically ill patients suffered from metastatic tumors originating from the lungs, esophagus, stomach, pancreas, liver, colon, rectum, bladder, endometrium, or cervix uteri and ovary. Special attention in designing the study was devoted to ethical aspects. For each study a specific informed consent form was prepared. After receiving the form, patients had to wait 48 hours to confirm their participation in the study. The aim of the study was to determine whether the MDB regimen had antitumor activity worthy of further controlled clinical evaluation. The main endpoint of each study was measurable clinical activity. Details of the design and conduct of the trials have already been reported.1,6,7 The protocols were approved by the National Cancer Advisory Committee and by an ad hoc Ethical Committee. A Steering Committee, which was composed of the principal investigators of each trial, the son of Dr. Di Bella, a group of statisticians and epidemiologists, and researchers of the National Institute of Health, was put in charge of the practical conduct of the study. A research coordination unit, created within the Italian National Institute of Health, coordinated all phases of the study. To ensure independent confirmation of all diagnoses, a committee of five members (listed in the footnotes at the beginning of this article), chosen from among oncologists and radiologists not involved in the trials, was charged with the review of all iconographic materials utilized in the evaluation of objective responses to the treatment. Finally, a panel of seven scientists (listed in the footnotes), selected from among leading international experts in oncology, acted as supervisors of the entire study. Each part of the study was reviewed by and discussed with this international review board. METHODS The enrollment of patients started in the 26 oncology centers identified throughout Italy as suitable for the study, in the middle of March 1998. More than 200 people were involved in all phases of the study. The main eligibility criteria, common for all protocols, were as follows: age $18 years, histologically or cytologically confirmed diagnosis of neoplastic disease, measurable and/or assessable lesions, advanced stage of disease, no prior MDB treatment, and signed informed consent. In the trial of nonsmall cell lung carcinoma, both patients with no previous chemotherapy and patients in progression after first-line chemotherapy were enrolled, although results were analyzed separately in the two strata. Patients with advanced pancreatic carcinoma had not previously been treated with chemotherapy. All other patients had already undergone chemotherapy and/or hormonal therapy; they were admitted only if the treatment course had been completed at least 4 weeks before the beginning of the trial. For all patients ECOG PS was evaluated. The treatment consisted of daily administration of melatonin (by mouth [os], 20 mg), bromocriptine (os, 2.5 mg), solution of retinoids (os, 7 g), somatostatin (slow subcutaneous injection, 3 mg over 8 hours), or octreotide (subcutaneous injection, 1 mg). Hydroxyurea (os, 1 mg) was added to the treatment of patients with glioblastoma, and cyclophosphamide (os, 50 mg) was added for patients who were not terminally ill, with time of administration varying according to the specific protocol. Patients were treated until progression occurred or unacceptable toxicity developed. For the evaluation of the anticancer activity of MDB, all patients were observed until one of the following events occurred: complete or partial response (confirmed after 4 weeks), disease progression, death, discontinuation of treatment because of toxicity, or voluntary abandonment. The MDB anticancer activity was expressed as the proportion of complete or partial objective responses, according to the definition accepted by the World Health Organization (WHO).8 The evaluation of objective response was performed utilizing different schedules, depending on individual protocols: every month for chronic lymphoid leukemia and lung carcinoma; every 2 months for non-Hodgkin lymphoma, breast carcinoma, colorectal carcinoma, head-neck and esophageal carcinoma, and advanced solid neoplasms; and at 3 months for pancreatic carcinoma and glioblastoma. The sample size for each of the studies was calculated on the basis of the minimum number of patients required to demonstrate a predefined level of anticancer activity.9 For the calculations, a one-tail test was applied, taking into account the proportion of partial or complete responses expected under the hypothesis of inactivity (p0) or sufficient anticancer activity (p1) of The Di Bella Case/Traversa et al. 1907 TABLE 2 Selected Characteristics of the Patients Included in the MDB Trials (n 5 386)a Breast carcinoma Lung carcinoma Lymphoid Colorectal Pancreatic Head and neck Advanced Lymphoma leukemia 0–2 3–4 Treated Untreated carcinoma carcinoma cancer Glioblastoma neoplasms (n 5 32) (n 5 22) (n 5 33) (n 5 34) (n 5 65) (n 5 51) (n 5 34) (n 5 29) (n 5 32) (n 5 20) (n 5 34) % females Median age (yrs) Median mos from diagnosis % previous surgery % previous chemotherapy % previous hormone therapy % previous radiotherapy ECOG PS (%) 0 1 2 3 4 40 58 34 25 100 30 62 62 14 100 100 54 78 91 94 100 59 58 91 94 20 62 12 35 100 10 67 3 33 — 62 63 21 100 100 41 69 3 66 — 9 62 34 69 100 50 61 10 100 — 71 70 14 62 68 — 41 — 4 82 70 85 76 — 60 — 23 — 29 — 7 — 94 — 100 6 15 22 25 22 31 — 68 18 9 4 — 39 39 21 — — — 3 3 74 21 12 52 35 — — 23 59 18 — — 32 35 32 — — 34 48 17 — — 3 44 53 — — — 30 70 — — — — 9 29 62 MDB: Di Bella Multitherapy; ECOG PS: Eastern Cooperative Oncology Group performance status. a This table was first published in BMJ (Italian Study Group for the Di Bella Multitherapy Trials. Evaluation of an unconventional cancer treatment [the Di Bella Multitherapy]: results of Phase II trials in Italy. BMJ 1999;318:224–8) and is reproduced by permission. the MDB, with an alpha error equal to 5% and a power of 95%. The p0 ranged from 5% to 10% and p1 from 20% to 30%; consequently, the number of patients needed in each protocol ranged from 24 and 69, and the minimum number of responses necessary to recommend further investigation varied between 2 and 12. Only in the glioblastoma trial was Simon two-stage optimal design followed.10 Kaplan–Meier survival analysis was applied to estimate the cumulative incidence and the median time to progression.11 RESULTS All patients were enrolled between March and July. One of the trials on breast carcinoma (preoperative treatment with MDB and tamoxifen) was discontinued in May because, over a period of more than 2 months, only 2 patients had opted to participate in this study. The final analysis considered a total of 386 patients who fulfilled the eligibility criteria. The main characteristics of these patients are summarized in Table 2. The detailed results of trials have already been reported and published.1,6,7 In summary, these results indicate that, at the beginning of the trials, a good majority of all patients were in fair-to-good condition; in addition, 21% of them had not received any previous chemotherapy. Although no patient experienced complete remission, in 3 cases a partial response could be documented: a man age 52 years with a non-Hodgkin lym- phoma, a woman age 50 years with breast carcinoma, and a man age 50 years with a nonendocrine pancreatic carcinoma. All three of these patients had disease progression after 350 days, 182 days, and 355 days, respectively. At the time of the scheduled evaluation of response, 47 patients (12%) had stable disease; for these subjects treatment was therefore continued. In June 1999, only 7 of these patients were still receiving MDB. The time to disease progression is illustrated in Figure 1 for all patients. The median time to progression was about 1 month for patients with chronic lymphoid leukemia, patients with breast carcinoma in poor condition (ECOG PS 3– 4), and critically ill patients. It was less than or approximately 2 months for those with non-Hodgkin lymphoma, breast carcinoma (ECOG PS 0 –2), previously treated lung carcinoma, colorectal carcinoma, pancreatic carcinoma, headneck or esophageal carcinoma, and glioblastoma. Previously untreated lung carcinoma patients had the longest median time to progression (almost 3 months). With only 3 partial responses among 386 patients, the results of the trials were considered incompatible with the initial claim that the MDB might represent a cure for any type of cancer. The possibility that, for a specific cancer among those included in the trials, there might be a therapeutic efficacy of clinical interest is also remote. 1908 CANCER November 15, 1999 / Volume 86 / Number 10 FIGURE 1. (A–D) Time to disease progression is shown for patients included in the trials of Di Bella Multitherapy (Kaplan–Meier survival analysis). (E–H) Time to disease progression is shown for patients included in the trials of Di Bella Multitherapy (Kaplan–Meier survival analysis). The Di Bella Case/Traversa et al. The Debate in Italy Conducting the trials was not an easy experience for the investigators at the 26 centers or for those responsible for the overall coordination, because there was an absolute need to produce convincing evidence quickly. It is fair to say that, in our opinion, both objectives were attained. Within approximately 1 year from the beginning of the Di Bella case, a scientifically valid response was provided to patients, physicians, scientists, and politicians. Putting aside the technical, logistic, and managerial difficulties, two major general problems were encountered: the progressively increasing critical attitude regarding the conduct of the trials by Dr. Di Bella and his followers, and the roles played by public prosecutors and local magistrates. Dr. Di Bella’s son, who is a physician, was a member of the Steering Committee but participated only in the first two meetings (in February and March 1998). As soon as the enrollment of patients started, there was a gradual increase in the criticism of how recruitment was being carried out; this was followed by a series of diverse critical remarks (which changed in nature as the study progressed). Initially, it was argued that Dr. Di Bella had not actually agreed on the details of the study protocols and the therapeutic regimen. Then it was argued that, though he had signed the protocols, the therapeutic regimen was different from the one he has currently adopted.12 Later the criticism shifted to focus on a bias in the selection of patients. It was argued that a vast majority of the patients enrolled were terminally ill. The fact of the matter is that only two protocols included terminal cancer patients. Whereas one could question, on theoretic grounds, the choice to extend the study to terminally ill patients, the fact remains that many such patients received compulsory treatment by the NHS with the MDB, under orders by local courts. Not to have tested MDB in these cases would have been irresponsible under the circumstances. Indeed, these were the patients (with their relatives) who were mostly seeking the MDB. Another criticism was the alleged possibility that previous chemotherapy cycles might have had a “negative” influence on the efficacy of the MDB. On the face of it, however, this statement was in sharp contrast to the initial repeated declarations by Dr. Di Bella that the MDB was effective even when other anticancer treatments had failed. The inclusion of a number of patients with lung carcinoma previously untreated with other regimens served to dispel this claim, because these patients also had a complete absence of objective response. 1909 It is worth noting that most of criticisms started as soon as it became clear that the trials were actually going to be carried out; they increased throughout the study as information on patients who had died in spite of the MDB was spread, and became strongest when the first final results (for four protocols) were made public by the end of July 1998.13,14 This situation is of some interest because it illustrates the peculiarity of many trials conducted to assess unconventional treatments. Usually, the clinical researcher launches a trial to demonstrate the validity of his or her own therapeutic hypothesis against a null hypothesis of no effectiveness. He or she knows that any flaw discovered by independent critics in the design or in the conduct of the trial will hinder the rejection of the null hypothesis, that is, demonstration of the validity of the treatment. In the MDB trials, those who supported the effectiveness of the treatment, and in particular Dr. Di Bella himself, were by their own choice only marginally involved in the design and conduct of the trials. As a consequence, their critical review of the trials started from a reverted null hypothesis, based on the assumption that the treatment was indeed effective: accordingly, any flaw they perceived in the design or conduct of the trials was interpreted to be supportive of its effectiveness. It is this reversion of the roles of the scientific debate (researcher vs. independent scientific reviewers) that contributes to a climate of reciprocal distrust. This can be avoided only if the supporters of an unconventional treatment follow the rules that are mandatory worldwide to assess new treatments. Unfortunately, this does not happen very frequently. It is still a matter of debate whether, in similar conditions, and given the large number of patients willing to receive the MDB, it might have been preferable to carry out Phase III comparative trials.15–17 Obviously, there are arguments for and against this. On the one hand, if the results of Phase II trials were encouraging, Phase III trials would be carried out anyway. On the other hand, the decision to initiate with a Phase III trial may turn out to be neither feasible nor ethical. Is it possible to ask patients who seek a specific unconventional treatment to agree to random assignment to a conventional one as well? Another issue that deserves specific mention relates to the inquiries made by public prosecutors in several Italian cities regarding the correctness of the trials. Nearly all stages and elements of the trials became objects of close scrutiny. In the future, this may represent a recurrent scenario in the evaluation of unconventional treatment efficacy, in which the proponents of a treatment may decide to take legal action, after an initial agreement, to stop the process of 1910 CANCER November 15, 1999 / Volume 86 / Number 10 objective evaluation. What we argue is the concern that if a “trial gets under trial during the trial” a potential conflict of interest may arise. If the investigators become the objects of scrutiny during the course of the study, which are the guarantees for the patients involved that every decision made by the investigators is adopted exclusively in their best interest? Moreover, why should the best interests of patients be better served by prosecutors who often ignore even the basis of clinical testing, rather than by steering committees, independent review boards, and ethical committees? The broader issue of “freedom of treatment” raised by Dr. Di Bella and his supporters deserves discussion. There are two major aspects to consider, namely, the duty of health authorities to regulate treatments that can be offered to patients, and the selection of treatments paid for by the NHS. Regulatory authorities base their evaluations solely on the available evidence regarding the efficacy and toxicity of new treatments, whereas the decision to reimburse for a given treatment takes into account other factors as well, such as cost-effectiveness. Prescribing the Di Bella treatment was perfectly legal, because all the individual drugs were marketed in Italy (with different indications), with the exception of the retinoid solution that had to be prepared in pharmacies. The law in Italy allows a physician to prescribe any registered drug, under his own responsibility, for indications different from the approved ones. This, however, was not the main issue. It was the high cost per month of MDB that became the critical issue because of the perceived injustice between wealthy patients who could afford the treatment and poor patients who could not. This was the basis for the decisions of a number of local magistrates throughout Italy to force the NHS to provide MDB at no cost.18 The ensuing national debate revolved around the right to decide which treatments can be undertaken by willing individuals and who has to decide about their reimbursement. The answer to both questions is definitely not straightforward. Central to the issue are two conflicting ethical values: the need to protect patients and citizens from ineffective or even harmful treatments, and the freedom of an individual to choose treatments even in the absence of scientific evidence of their effectiveness. Can this right be dependent on whether or not a patient can financially afford the treatment? The fact is that this freedom represents a right of every citizen, though no financial support can be claimed when evidence of efficacy is lacking. All of this exists against a background of limited financial resources on the part of the NHS, which is financially unable to satisfy even all the needs for treatment “of proven efficacy.” The conclusion that no clear-cut solution exists seems indisputable; therefore, the only possible way to cope with these problems is to couple regulatory action with adequate and credible information for the public at large. In the Di Bella case a majority of Italians initially felt that the information provided by health authorities was not sufficient to label MDB “ineffective.” As a consequence, the refusal by the NHS to provide it was regarded as a proof of prejudice on the part of the scientific establishment. We must conclude that those who labeled as “irrational” the desire of thousands of desperate patients to “try” MDB failed to recognize that for many of these patients the choice might have been a rational one. If, in fact, conventional treatments offer no chance of cure, a treatment that promises cure (i.e., a very large benefit) could indeed be the best option, as long as the probability that it will fulfill this promise is greater than zero. In fact, the Decision Theory holds that the expected benefit is equal to the size of the benefit multiplied by the probability of achieving it. The repeated anecdotal media reports of “incurable” patients who were “cured” by the Di Bella treatment, or at the very least benefited from it, definitely made this probability greater than zero for the vast majority of individuals lacking the scientific information necessary to discard these reports, unless they had been verified. Not surprisingly, many physicians shared this view, even though they were skeptical about the potential effectiveness of MDB. To provide up-to-date information on MDB, from the planning stage it was decided to make public the results of the trials as soon as they became available. In this respect, it is worth noting that the presentation of the results of 4 protocols on July 28, 1998, had almost no effect on the number of requests for MDB. The Di Bella group continued to present patients who claimed to have been cured with the MDB as proof of the lack of validity of the trial findings. To desperate people who had lost all hope, a single patient who claimed to have been cured may have been far more convincing than the results for hundreds of other subjects included in the trials. Only after the communication of the final trial findings on November 13, 1998, a change in the public perception started to materialize. The complete results were received by the public without any widespread negative reaction. Even the politicians and journalists who had spoken in favor of Dr. Di Bella seemed to have accepted the idea that the case was about to be concluded. CONCLUSIONS It is our opinion that the Di Bella case was created by different concomitant factors. An important one was The Di Bella Case/Traversa et al. the high cost of the treatment; another critical factor was the reaction of the media, which is more often than not guided by the imperative “first of all, do not kill the news (in pursuit of the truth).” This cynical attitude played a major role in diffusing, among hopeless patients, great expectations for an untested treatment. If we have to learn from this experience and avoid future repetition of such sagas, it is necessary that the public, and therefore the government, rely on credible scientific institutions for guidance. However, this will occur only if the medical community is perceived as dedicated to the patients’ well-being, rather than to personal and corporate interests. Today, almost 8 months after the presentation of the final data, the situation seems to have stabilized to the point that the publication of further negative findings concerning a survival study of patients treated with MDB outside the Phase II trials19 did not fuel a new, appreciable controversy. The rapid performance and the completion, within 1 year, of the trials disproving the Di Bella claims provided solid ground for a more rational discussion and gave a basis to the government’s decision not to approve MDB as an anticancer treatment. Finally, the need to rely on evidence provided by clinical trials to accept a new treatment has been reaffirmed; the wide public discussion that took place throughout the country contributed to the spread of this principle outside the narrow confines of the scientific community. REFERENCES 1. 2. 3. 4. Italian Study Group for the Di Bella Multitherapy Trials. Evaluation of an unconventional cancer treatment (the Di Bella Multitherapy): results of phase II trials in Italy. BMJ 1999;318:224 – 8. Osservatorio della comunicazione radiotelevisiva di Pavia. Il caso Di Bella nella televisione e nella stampa italiana. Roma, RAI-ERI, 1999. Abbasi K. Di Bella’s miracle method. BMJ 1998;316:1617. Support for a pragmatic health minister [editorial]. Nature 1998;392:421. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 1911 Government of Italy. DL n. 23 del 17 febbraio 1998. Disposizioni urgenti in materia di sperimentazioni cliniche in campo oncologico e altre misure in materia sanitaria. G.U. n. 39 del 17 febbraio 1998. Gruppo di Coordinamento Centrale per la Sperimentazione del Multitrattamento Di Bella. Risultati della Sperimentazione del Multitrattamento Di Bella (MDB): protocolli 4, 6, 8, 10. Roma: Istituto Superiore di Sanità. 1998. Report No: Rapporti ISTISAN;98/17. Gruppo di Coordinamento Centrale per la Sperimentazione del Multitrattamento Di Bella. Risultati della Sperimentazione del Multitrattamento Di Bella (MDB): protocolli 1, 3, 5, 7, 9. Roma: Istituto Superiore di Sanità. Servizio per le Attività Editoriali, 1998. Report No. Rapporti ISTISAN; 98/24. Miller AB, Hoogstraten B, Staquet M, Winkler A. Reporting results of cancer treatment. Cancer 1981;47:207–14. Levental BG, Wittes RE. Research methods in clinical oncology. New York: Raven Press, 1988. Simon R. Optimal two-stage designs for phase II clinical trials. Control Clin Trials 1989;10:1–10. Bland JM, Altman DG. Survival probabilities (the Kaplan– Meier method). BMJ 1998;317:1572. Pellegrini R. Di Bella’s method of curing cancer is becoming popular in Italy. BMJ 1998;317:52. Brera GR, Violato C. Methodological and ethical difficulties in clinical oncology studies. Lancet 1998;352:1391. Raschetti R, Bruzzi P, for the Di Bella Multitreatment Italian Trial Coordinating Group. Methodological and ethical difficulties in clinical oncology trials. Lancet 1999;353:153– 4. Müllner M. Di Bella’s therapy: the last word? BMJ 1999;318: 208 –9. The Editorial Committee of the Italian Study group for the Di Bella Multitherapy Trials. Raschetti R, Bruzzi P, Greco D, Menniti-Ippolito F, Spila-Alegiani S, Traversa G, et al. We totally disagree with Müllner’s view. BMJ 1999;318:1074. Liberati A, Magrini N, Patoia L, Pagliaro L. Missing the forest while looking at the tree. BMJ 1999;318:1073– 4. More clinical judgement, fewer “clinical” judges [editorial]. Lancet 1998;351:303. Gruppo di Coordinamento Centrale per la Sperimentazione del Multitrattamento Di Bella. Risultati della Sperimentazione del Multitrattamento Di Bella (MDB): Studio osservazionale. Roma: Istituto Superiore di Sanità. Servizio per le Attività Editoriali, 1999. Report No. Rapporti ISTISAN; 99/12.