A selective history of the therapy of Hodgkin's disease.
This is, for the most part, the history of the development of curative combination chemotherapy for Hodgkin's disease, and its consequences, from 1963 up to the early 1990s [the MOPP (nitrogen mustard, oncovin, procarbazine, prednisone) programme]. Coupled with radiotherapy, it had a major impact on the management of the disease. It is also heavily weighted towards the description of the events that occurred at the Medicine Branch of the National Cancer Institute (NCI) in Bethesda, MD, USA, between 1963 and the early 1990s. To this is added a brief description of the follow-up development of the ABVD (adriamycin, bleomycin, vinblastine, dacarbazine) treatment drug programme by the group in Milan, Italy. The latter added significantly to the therapeutic armamentarium available to physicians in the 1970s and beyond. This selective history is perhaps justified because, although the focus of the studies discussed below was on Hodgkin's disease, there was a more important issue involved. A new breed of cancer investigators in the 1960s had been addressing the generic question of whether or not cytotoxic chemotherapy was ever capable of curing patients with any type of advanced malignancies. The answer would turn out to be 'yes', and this affirmative answer enabled chemotherapy to be used in many other tumours and, notably, in the adjuvant setting after treatment of apparently localized disease. The question was addressed in two important human models, childhood acute leukaemia (Frei & Freireich, 1965) and Hodgkin's disease, and used combination chemotherapy, a concept that deviated from the practice of the time. The response in the academic community to this radical departure from dogma was initially quite negative. The protected environment of The Medicine Branch, at the Clinical Center of the NCI, facilitated studies in both diseases, under the inspired leadership of Dr Emil Frei, who was Chief until 1965 (Fig 1). In recounting these events, the novelty of the approaches and study results do not seem so radical in retrospect, but they were. Dr Emil (Tom) Frei in Bethesda in the 1960s. Chief of Medicine at the NCI until 1965. Photograph from the author's collection. At the time the programmes began at the NCI, radiation therapy was already established for Hodgkin's disease patients presenting with apparently localized disease. We did not know the aetiology of Hodgkin's disease or the cellular origin of the Reed–Sternberg cell that characterizes the disease. Medical oncology, as a speciality, had not yet appeared on the scene. Physicians who would become the medical oncologists of the future worked with therapeutic radiologists and pathologists to integrate newer methods of diagnosis, staging and radiotherapy with the newly developed chemotherapy programmes and converted Hodgkin's disease from a largely fatal to a largely curable disease. For the integration of chemotherapy and radiotherapy in Hodgkin's disease, we owe a debt of gratitude to the late Henry Kaplan (Fig 2), who was unique in his field in recognizing the value of the new chemotherapy. He set in motion a brilliant series of studies of the use of MOPP and radiotherapy at Stanford University in the late 1960s and 1970s that still serve as a model of clinical investigations (Kaplan, 1962, 1966; Kaplan & Rosenberg, 1966; Kaplan et al, 1973). Curative chemotherapy had provided the missing link in the therapeutic equation. The late Harry Kaplan at Stanford in the early 1970s. Photograph from the author's collection. The group of senior faculty members from the Medicine Branch, shown in Fig 3, went on to apply these same principles to other cancers (Bagley et al, 1972; Canellos et al, 1974). The development of the CMF programme (cyclophosphamide, methotrexate and 5-fluorouracil) for breast cancer (Canellos et al, 1974) and the NCI's contract with the Istituto Tumori in Milan, Italy, to test it in the adjuvant situation helped to set in motion a similar series of events in breast cancer. The author, then Chief of Medicine at the NCI (second from right), with the senior faculty of the Branch who played a role in advancing the studies on all the lymphomas. Left to right: Drs George Canellos, Bruce Chabner, Phillip Schein, the author and Robert Young (about 1970). Photograph from the author's collection. The author (Fig 4) chose to end this tale in the early 1990s because, at that point, we entered a new era of treatment for Hodgkin's disease. Investigators began to focus on the problems of curing a minority of patients not cured by the initial programmes and reducing the morbidity of treatment revealed in long-term follow-up studies. A daunting task, but another story entirely. The author in the Medicine Branch of the NCI in 1968. Photograph from the author's collection. Pusey first used radiotherapy to treat Hodgkin's disease in 1902 (Pusey, 1902), only 8 years after Roentgen rays were discovered. The important observation that Hodgkin's disease spread by contiguity is owed to several investigators, most notably the Swiss radiotherapist Renee Gilbert in the 1920s (Gilbert, 1925) and the Canadian radiotherapist Vera Peters in the 1950s (Peters, 1950; Peters & Middlemiss, 1958). These observations had important implications for radiotherapy, as fields could now be shaped to encompass areas of presumed contiguous spread. By observing the natural history of the disease in large patient populations, Vera Peters (Peters, 1966) and Henry Kaplan (Kaplan et al, 1973) were able to map the patterns of spread and apply appropriate radiotherapy fields (Rosenberg & Kaplan, 1966). Although Vera Peters implied that the disease was curable (Peters & Middlemiss, 1958), it was Easson and Russell from the Christie Hospital, Manchester, UK, using her data, who first broached the subject of 'The Cure of Hodgkin's disease' (Easson & Russell, 1963). It was a sign of the pessimism of the times that the title was considered provocative and controversial. It was Kaplan who made the point that localized Hodgkin's disease could by cured by radiotherapy with finality and conducted the trials to show the doses and extent of radiotherapy necessary to cure early stages of the disease. At Kaplan's insistence, a conference of leading investigators was convened in Rye, New York, in 1966 that not only married the new histological classification developed by Lukes & Butler (1966) to treatment, but laid out the principles of staging necessary for an orderly approach to treatment. Combination chemotherapy was just in its seminal phase at this time but already the cure of localized stages with good radiotherapy was noted to be about 30%. In the early 1960s, the debate that dominated chemotherapy was not could it cure, but whether or not the palliation achievable was worth the toxicity associated with anticancer drugs. The first chemotherapy study to have an impact on the management of patients with Hodgkin's disease was published in 1963 (Scott, 1963). Eighty-nine patients with advanced Hodgkin's disease received a conventional induction course of nitrogen mustard (0·4 mg/kg), of whom 40 patients with 'satisfactory response' were randomized to receive either no further treatment or continuous treatment with the newly developed oral alkylating agent, chlorambucil. In the 16 patients who received chlorambucil, the time to relapse averaged 35 weeks compared with 11·7 weeks without further treatment. This highly significant difference in the duration of a 'satisfactory remission' provided the first useful information on alternatives in the day-to-day management of patients with Hodgkin's disease. No mention was made in this study, however, as to whether any survival benefit accrued to the patients maintained on chlorambucil. A report by Jacobs et al (1968) presented one of the first survival curves published in the modern chemotherapy era. Drug treatment in patients with advanced Hodgkin's disease was associated with a median survival of less than 2 years, with only 5% living beyond 4 years, all with evidence of disease, only slightly superior to patients left untreated for the entire course of their disease (Craft, 1940) (Fig 5). Survival curves of patients with Hodgkin's disease either untreated (Craft, 1940) or treated only with alkylating agents (Jacobs et al, 1968) before the advent of combination chemotherapy. The next major advance came with the identification of the antitumour activity of the plant-derived natural products, the vinca alkaloids. The availability of two non-cross-resistant classes of antitumour agents and the conceptual separation of induction and maintenance therapy gave impetus to a large study initiated through the combined effort of two clinical co-operative groups supported by the NCI (Acute Leukaemia Group B and the Eastern Solid Tumor Group). The principal investigator was the late Paul Carbone of the NCI's Medicine Branch (Carbone et al, 1968). Patients were randomized by disease and prior therapy to remission induction with cyclophosphamide or one of the vinca alkaloids. The objectives of this study were to compare the effectiveness of the vinca alkaloids with an alkylating agent in remission induction in lymphomas. Patients received either cyclophosphamide or placebo to maintain the remission, which enabled the effectiveness of continuous therapy to be compared with intermittent treatment with cyclophosphamide at the time of disease progression. The results of this study established the superiority of vinblastine over cyclophosphamide in patients with Hodgkin's disease. In all disease categories, the average duration of placebo-maintained remission was a remarkably short 4–6 weeks, irrespective of the drug used to induce remission. Daily oral cyclophosphamide significantly prolonged remission duration when compared with placebo, confirming the observation of Scott (1963). Remission duration with drug maintenance treatment was, however, only 32 weeks for Hodgkin's disease. Two features of this study influenced the studies designed later by NCI investigators. First, patients were separated by whether they had partial or complete remissions, a practice used to evaluate treatments in leukaemias but not generally used at that time in 'solid tumours'. Secondly, regardless of the approach to treatment, overall survival in the various subgroups in each disease category was almost identical. Those data enabled investigators at the NCI, using the first combination chemotherapy programmes, to set achieving a high complete remission rate as their major initial goal in the first combination chemotherapy programme, and to leave patients who attained a complete remission off therapy to evaluate the capacity of their new treatment programme to eradicate the tumour permanently. The appearance of vinca alkaloids was followed shortly by the discovery of the antitumour activity of the methylhydrazine derivative, procarbazine, then called ibenzmenthyzin, in Hodgkin's disease (Bollag & Grunberg, 1963; Mathe et al, 1963; DeVita et al, 1965a). The first intensive drug combination programme designed to exploit the new principles of chemotherapy in Hodgkin's disease began at the NCI in 1963 and used the combination of vincristine, methotrexate, cyclophosphamide and prednisone (MOMP) given for only 2·5 months (DeVita et al, 1965b; Moxley et al, 1967). The goal of this pilot protocol was to test the safety of combination chemotherapy in advanced Hodgkin's disease. The vinca alkaloid, vincristine, was selected over vinblastine because it has little associated marrow toxicity, and preclinical studies suggested therapeutic effects equal to vinblastine. In an era devoid of reliable methods to support patients during periods of bone marrow suppression, this was an important issue. Only 14 patients were treated with MOMP, and all were hospitalized and kept in reverse isolation. This small study showed that the approach was safe. In 1964, the MOMP programme was modified in several ways. At that time, most data on the doses and schedules of anticancer drugs in humans were derived and translated directly to the clinic from tumour-bearing mouse models. But, comparative cell kinetic data in mouse and human tumours had shown much-prolonged human tumour cell cycle times, compared with the mouse (Yankee et al, 1967; Young & DeVita, 1970; DeVita, 1971; DeVita & Schein, 1973). So, the duration of MOPP treatment was increased from 2·5 months, the standard of the time, to 6 months. And, procarbazine, by now an agent known to be active in Hodgkin's disease, was substituted for the antifol methotrexate. Each drug was used in its optimal dose and schedule. This new programme was named MOPP ('M' for nitrogen mustard, 'O' for oncovin, the brand name for vincristine, and 'PP' for procarbazine and prednisone). Cyclical administration every 28 d was introduced, and the interval selected between cycles was the narrowest possible to allow for recovery of the most sensitive normal target tissue, the bone marrow. Again, cell kinetic studies on the mouse bone marrow, compared with humans, showed that, like mouse and human tumours, the kinetics of human marrow were also quite prolonged compared with the mouse (Yankee et al, 1967; DeVita et al, 1969a; Young & DeVita, 1970; DeVita, 1971). Thus, for patients whose tumour, unlike leukaemia, infrequently involved bone marrow, adjustments in schedules were made to introduce wider intervals between treatments than previously allowed. In 1967, the results of the use of MOPP in the first 43 chemotherapy-naïve patients were reported in abstract form, and formal results were published in 1970 (DeVita et al, 1970). An 80% complete remission rate was noted. This was a fourfold increase over results achieved with the best use of single agents of the day, and those remissions proved durable and appeared to influence survival. The survival of treated patients was impressively different from those treated with single agents (Fig 6). British investigators, under the leadership of the late G. Hamilton Fairley, confirmed the effectiveness of combination chemotherapy with a modified version of MOPP (Nicholson et al, 1970). Others followed suit in either controlled (Huguley et al, 1975) or uncontrolled (Frei et al, 1972a; Moore et al, 1973) studies. While the Stanford group confirmed the NCI results, their concern with the neurotoxicity of vincristine led them to recommend capping the dose at 2 mg total dose. This widely practised dose adjustment negatively affected the dose intensity of vincristine in all future studies of MOPP (Moore et al, 1973). Life table analysis of survival for the entire group of 43 patients (solid line) and the 35 patients achieving a complete remission (CR; dotted line). Figures in parentheses next to each point indicate the number of patients at risk for that interval. From the original MOPP trial (DeVita et al, 1970). Published with permission. When the NCI investigators later reported on the long-term follow-up of the MOPP treatment of the first 198 patients, the initial results were maintained (DeVita et al, 1980; Longo et al, 1986) (Fig 7). Of 198 patients, 159 (80%), achieved a complete remission with a median of three treatment cycles. All 23 patients without symptoms attained complete remission, compared with 78% of the symptomatic patients. This difference was highly significant, and only one asymptomatic patient had relapsed in subsequent years. Plot of remission duration in 198 patients with advanced Hodgkin's disease treated with MOPP at 20 years follow-up from Longo et al. (1986). Reprinted with permission from the American Society of Clinical Oncology. Sixty-three per cent of patients achieving complete remission, who were at risk for longer than 10 years, remained disease free. Because 80% of the entire treated population achieved a complete remission, the proportion of all treated patients who remained free of relapse at 10 years and beyond was 54·6%. The vast majority of relapses occurred within 42 months after cessation of therapy. Two features had a powerful effect on outcome, the absence of symptoms and the dose delivery rate of the drug vincristine. The next question posed was whether maintenance drug therapy would be beneficial after achieving a complete remission, as 35–50% of patients relapsed by the fifth year of follow-up. The first such study was also conducted at the NCI (Young et al, 1973). After achieving complete remission, patients were randomly allocated to intermittent cycles of carmustine (the then new nitrosourea, BCNU) or no therapy or intermittent cycles of MOPP. Maintenance treatment was given for 15 months. When the results of this study were reported in 1973, and reanalysed in 1980, there were no significant advantages for either continued intermittent MOPP or intermittent carmustine therapy. Seven other studies designed to test the usefulness of maintenance therapy have been conducted, also with no long-term positive effects noted (Coltman et al, 1976; DeVita, 1981). An important biological principle, related to the issue of de novo resistance to chemotherapy, was first observed from the retreatment of relapsed patients. Only five out of 17 patients whose initial complete remission was less than 1 year in duration achieved a second complete remission (29%), compared with 14 out of 15 patients whose initial complete remission was 1 year or longer (93%), a statistically significant difference (P = 0·001) and, once again, remissions in this latter group were durable (Fisher et al, 1970). Overall, survival was significantly improved in those patients who achieved second remission over those who did not achieve a second remission (P = 0·005). Retained sensitivity to MOPP in patients who stayed in remission in excess of 1 year was a novel observation at the time and indicated that these patients were 'almost cured' by the induction programme, and might benefit from future intensification of treatment. This has proven to be the case in the modern era. The relative insensitivity of the tumour of patients who experience short remissions (< 12 months) suggested that the primary cause of treatment failure in this group was the presence and overgrowth of cells resistant to the drugs in the MOPP programme at the time of diagnosis. More importantly, the observation on retained sensitivity after a long drug-induced remission has carried over to virtually all advanced cancers where complete remissions in response to drug combinations are attainable, and illustrates an important and generally unexplained biological principle that is still under investigation (Fisher et al, 1970; Viviani et al, 1990; DeVita, 1991; Longo et al, 1992). The impressive survival curves in MOPP-treated patients published in 1970 indicated that it was possible to cure advanced Hodgkin's disease with combination chemotherapy (DeVita et al, 1970). As the capacity to cure early-stage disease with radiotherapy was limited, the cure of advanced disease made it ethically possible to ask the question whether the use of chemotherapy alone, in lieu of radiotherapy, would improve the cure rate when the tumour volume was low. A trial testing this hypothesis was initiated at the NCI in 1978, comparing MOPP therapy alone, in laparotomy-staged patients, with subtotal nodal irradiation therapy (Longo et al, 1991a). MOPP, indeed, proved superior to Only one other similar trial has been et al, 1992). Although the results are only they have made it possible to chemotherapy as an to radiotherapy in early-stage patients, an approach that is now under investigation in the modern era. a population of patients available for study, other important observations related to the use of chemotherapy were first made in these initial MOPP studies 1981). the first evidence of the long-term effects of cancer agents in cured humans et al, 1972; DeVita et al, the first evidence of from cytotoxic drugs & DeVita, 1973) and the impact of chemotherapy on et al, the first evidence of the recovery of the unique to Hodgkin's disease, with therapy that was in (Fisher et al, and the identification of unique to cancer patients at that time, the first and treatment of in an (DeVita et al, These conducted by a single group at one over 20 years to as by the time it to the results of all the studies designed in the 1960s and 1970s. It is a to the of a in clinical It also led to the first cure of the most of known as the large cell with a of the MOPP programme, another story in (DeVita et al, Although the MOPP studies established that an advanced cancer of a major in could be cured by combination chemotherapy, of patients did not achieve complete remission after MOPP, and of complete the availability of new in the late 1960s and investigators began to and test new The known as ABVD (adriamycin, bleomycin, vinblastine and dacarbazine) was developed at the Istituto Tumori in Milan, et al, on evidence of the of all drugs and their sensitivity with MOPP (Frei et al, et al, et al, A randomized trial was in to test whether ABVD chemotherapy could induce a complete remission rate with that of MOPP chemotherapy et al, Overall, cycles of either a of complete remissions, and this had an influence on the from and survival This study showed that ABVD chemotherapy was as as MOPP in durable remissions in advanced Hodgkin's disease. effects were similar in both drug but the later were therapy to the development of acute leukaemias in the alkylating MOPP the effect of and irradiation to the and proved to be the major with from both the NCI and the group showed that patients who did not achieve a complete remission after either MOPP or ABVD showed of tumour after the first cycles of chemotherapy, or within a months after an initial complete remission, and that a of these patients could be by the chemotherapy. To the of the availability of two non-cross-resistant drug investigators at the Cancer Institute designed was called the MOPP and ABVD that cycles of MOPP and ABVD to be given not for 6 months, the but over 12 months. By previously patients were randomized to receive either cycles of MOPP or 12 months of et al, Although there were no in two non-cross-resistant drug combinations appeared to a superior initial response to MOPP for 12 months, however, led to a in dose for vincristine, which had been given for that duration of time, and this have affected the response rate to MOPP. study of non-cross-resistant chemotherapy, at about the same time at the NCI, had a different It compared MOPP, this time in its with 6 months of MOPP with and a combination of drugs that had also been shown to be an to MOPP (Longo et al, The results showed no for chemotherapy when MOPP was given at & published a model to the failure of chemotherapy to cure more patients with advanced model that tumour cells towards drug resistance before to chemotherapy, and at cell that would be below the of of the Drug resistance was to their to be in all tumours to a or at the time of diagnosis. This was, combination chemotherapy was in by resistant with drugs. was to apply all available drugs this was not approaches were the next best the information in the two trials from the NCI and Milan, they the results of the study as clinical of their This set in motion a of clinical trials of delivery of the two non-cross-resistant the that dominated clinical trials in Hodgkin's disease & DeVita, The test of the hypothesis in Hodgkin's disease came from the Cancer and Leukaemia This co-operative group MOPP ABVD MOPP with ABVD in and patients (Canellos et al, 1992). As was the case in the NCI study, the study confirmed that a combination programme in this given in be as as therapy. To the early long-term follow-up data have shown any survival the three treatment and the practice is the use of a standard 1991; Canellos & 40 years after the development of combination chemotherapy, Hodgkin's disease patients are cured in 80% of and from the disease have in the At 15 years from the end of treatment, a patient has a risk of of a of treatment than of Hodgkin's disease patients receive combination chemotherapy and radiotherapy (Longo et al, and those that those with long first remissions, are with intensive treatment programmes, with cell as support for marrow The for newer combinations of more less drug combinations on with only one programme from the toxicity of both MOPP and ABVD et al, We we know that the Reed–Sternberg cell is a B although this has not helped and the aetiology of the disease still by the Reed–Sternberg cell are however, for either or derived from but of the latter methods has advanced in a clinical situation by a of treatment cell used in the of that is with toxicity with less have an in as a therapeutic approach of The set in motion by the cure of acute childhood leukaemia and Hodgkin's disease has maintained its A of advanced cancers be cured by combination chemotherapy, and the of chemotherapy as an to has the management of such as those of the breast and As to and drugs for them the there is for a more of treatment that a on the patient with cancer and, most cancers from a to a disease, like Hodgkin's disease & DeVita,
Read more
9