HD9

BEACOPP _escalated reflects the introduction of granulocyte-colonystimulating factor (G-CSF) in hematology. In the 1980s it was hypothesized that dose density might increase the response rate of chemotherapy-sensitive tumors. Dirk Hasenclever developed a statistical model of the association of tumor growth kinetics and the effects of chemotherapy. The model was used to simulate the effect of dose escalation, dose density, and schedule changes in the COPP/ABVD regimen. This simulation demonstrated that the most potent effect would be achieved through dose escalation of cytostatics. This effect was estimated at an increase of 10% to 15% in PFS after 5 years. From this model, the BEACOPP regimen with G-CSF support was developed and tested versus the former standard COPP/ABVD in the HD9 study. The HD9 study randomized 1195 patients in 3 treatment groups (COPP/ABVD, BEACOPP _baseline , and BEACOPP _escalated ) and clearly demonstrated the superiority of BEACOPP _escalated . The 10-year data for this study have confirmed the initial results: the freedom from treatment failure and OS rates for COPP/ABVD were 74% and 75%, respectively, after 120 months. The corresponding results for BEACOPP _escalated are 82% and 86%, respectively. OS is thus 11% better than with standard COPP/ABVD. This effect is particularly pronounced in the group of patients with an intermediate risk profile according to the international prognostic score (IPS 2-3), which also forms the largest subgroup of patients (IPS 0-1: 28%, IPS 2-3: 38%, IPS 4-7: 13%).

HD12

However, the toxicity of 8 cycles of BEACOPP _escalated is high. In addition to the considerable acute toxicity, the development of secondary acute leukemia prompted concern. In the HD9 study, the incidence of secondary acute leukemia was 3% compared with only 0.4% in the COPP/ABVD arm. The follow-up study HD12 therefore looked at reducing the chemotherapy to 4 cycles of BEACOPP _escalated followed by 4 cycles of BEACOPP _baseline (“4+4” regimen). At 5 years, there were no significant differences regarding OS or PFS, although there was a decrease in absolute numbers with the 4+4 regimen. Importantly, the incidence of severe toxicities could not be reduced by the reduction of chemotherapy, so 8 cycles of BEACOPP _escalated remained standard of care in the GHSG.

HD15

In the subsequent study, de-escalation of chemotherapy was investigated with a reduction in the number of escalated cycles from 8 to 6 and with the introduction of a dose dense BEACOPP _baseline regimen (BEACOPP-14). The study was designed to show noninferiority of the experimental treatment groups. In addition, positron emission tomography (PET)-guided radiotherapy of residual disease ≥2.5 cm was investigated. Only PET-positive patients received consolidating radiotherapy. A total of 2182 patients were randomized among the 3 study arms. Surprisingly, when comparing 6 cycles of BEACOPP _escalated with 8 cycles, both PFS (90.3% vs 85.6%) and OS (95.3% vs 91.9%) were significantly superior with the reduced number of cycles. With regard to radiotherapy, the negative predictive value for PET at 12 months was 94.1% (95% CI 92.1%–96.1%) and only 11% of all patients received additional radiotherapy without compromising the tumor control. It is thus safe not to irradiate patients with residual tumor masses ≥2.5 cm, if they are PET negative. In summary, the HD15 established 6 cycles of BEACOPP _escalated as a new standard of care based on a significantly improved PFS and OS.

HD18

The idea of this trial is to reduce the number of chemotherapy cycles further from 6 to 4 in patients with a PET-negative response after 2 cycles of BEACOPP _escalated (early interim PET). If this reduction could be done without a loss of efficacy, it would be a major improvement for the patients. The cumulative dose of alkylating agents after 4 cycles of BEACOPP _escalated should be low enough to improve the toxicity profile of the early intensification approach significantly. The study is still enrolling patients and results for this part of the study will not be available before 2018. Another question to be answered by this study is related to patients with PET-positive disease at early interim staging. These patients have received additional treatment with the anti-CD20 antibody rituximab to improve their outcome. Results for this part of the HD18 trial will be available at the end of 2013. It might be noted though that the response adapted de-escalation of therapy using an early interim PET has never been shown to be safe in a randomized trial so far. Thus, without evidence from randomized trials, early interim PET must currently not be used to guide therapy outside clinical studies in advanced-stage disease.

Acute Toxicity

Thousands of patients have been treated with BEACOPP _escalated within controlled international multicenter trials. This review refers to the experience derived from the GHSG studies. These trials include large tertiary cancer centers, but more than half of the patients within the GHSG trials are enrolled by small centers and private practitioners enrolling about one patient per year. In this real-life setting, 94% of all patients experience at least one World Health Organization toxicity grade III/IV.

Leukopenia

Leukopenia is observed in almost 90% of patients; anemia and thrombocytopenia are observed in about 50% of patients. Consequently, infections are documented in around 20% to 25% of patients. Acute treatment-related mortality (TRM) of 6 cycles of BEACOPP _escalated varied from 0.8% up to 2.5%. TRM highly depends on 2 easily assessable patient-related risk factors: age and performance status. Patients ≤40 years old show a TRM of only 0.7%; however, this risk increases to 2.5% in patients between 40 and 49 years, 3.8% in those between 50 and 59, and 14.3% in patients aged older than 60 years. Furthermore, a poor performance status increases the risk of TRM. Thus, the TRM risk can be evaluated easily and patients at risk can be monitored adequately. In the ongoing HD18 study prophylactic measures have been implemented including obligatory antibiotic treatment during neutropenia and TRM is less than 1% with more than 1000 patients enrolled so far.

Organ toxicity is observed in almost all organ systems but with very low frequencies. An exception to this rule is severe peripheral neuropathy due to the use of vincristine, occurring in up to 10% of patients.

Gonadal Toxicity

Gonadal toxicity must be mentioned. Karolin Behringer has investigated this important question in the G5 (ie, GHSG studies HD13, HD14, and HD15). A total of 1323 (55%) of 2412 contacted female and male survivors were evaluable for analysis (mean follow-up 46 and 48 months, respectively). After 6 to 8 cycles of BEACOPP, menstrual activity was strongly related to age (<30 years: 82% vs >30 years: 45%; P <.001; prolonged recovery time up to 3 years). Thirty-four percent of women older than 30 years suffered severe menopausal symptoms (3- to 4-fold more frequently than expected). In contrast, male survivors had mean levels of testosterone within the normal range and reported no increased symptoms of hypogonadism. However, in men follicle-stimulating hormone and inhibin-B levels indicated infertility for most survivors regardless of age. Importantly, for both men and women, the sexual quality of life was shown to be more closely related to patient characteristics and sexual functions before start of treatment than to the intensity of treatment. Sexual functions recover in most patients after the end of treatment.

Thus, there is well-grounded information on gonadal toxicity of BEACOPP _escalated regarding sexual functions, hypogonadism, and infertility. It is important to note that all of these side effects depend on age and gender. Therefore, the individual who wishes to avoid gonadal damage can be balanced against the individual who wishes to be cured with the primary treatment and not to experience the psychological stress of relapsing from the lymphoma and being in need of a very demanding second-line treatment. Publications suggest that physicians might overestimate the meaning of infertility for patients suffering from HL and, accordingly, underestimate the meaning of relapse-free survival for these patients.

Long-Term Toxicity

Little is known about the long-term toxicities of BEACOPP _escalated . Second malignancies including non-HL, solid tumors, and acute myeloid leukemia (sAML/myelodysplastic syndrome [MDS]) were the focus when BEACOPP was introduced. At 5-year median follow-up, there were 0.4% cases of sAML/MDS in the COPP/ABVD group, 0.8% in the BEACOPP baseline group, and 1.9% in the BEACOPP escalated group. At 10-years median follow-up, there were 0.4%, 1.5%, and 3.0% of patients with sAML/MDS. The total number of second malignancies was not different between the 3 treatment arms. Interestingly, the risk of sAML/MDS after BEACOPP escalated was only 0.9% at 4 years in the HD12 study. It is known that radiotherapy combined with chemotherapy add to the risk for second malignancies. In HD9, 70% of patients received additional radiation, whereas in HD12, only 38% of 1502 patients enrolled were irradiated. The reduction of radiotherapy very likely contributes to an improved safety profile in HL patients, which is certainly also true for the cumulative dose of the alkylating agents cyclophosphamide and procarbazine. With 6 cycles of BEACOPP _escalated instead of 8 cycles in the GHSG HD15 trial (and only 11% of patients receiving consolidating radiotherapy), the risk of sAML/MDS decreased to 0.3%. So far, no increased risk for solid tumors has been detected. The median observation time of 10 years for BEACOPP _escalated might be too short; however, an even longer follow-up is needed. It can be assumed that the reduction of radiotherapy to only every tenth patient might result in a markedly decreased risk of second solid tumors, which is also true for cardiovascular disease.