To improve the complete response (CR) rate in advanced multiple myeloma (MM) without increasing the toxicity of high-dose therapy, we have used a new conditioning regimen. A combination of BE-8 [an anti-interleukin 6 (IL-6) murine monoclonal antibody] and dexamethasone followed by high-dose melphalan (220 mg/m2) and autologous stem cell transplantation was used to treat a series of 16 patients with advanced multiple myeloma. A strong inhibition of IL-6 activity evaluated by quantification of C-reactive protein was observed in all patients and was correlated with the high CR rate achieved with this combination therapy.
The outcome of patients autografted for newly diagnosed multiple myeloma (MM) is superior to that of those receiving conventional chemotherapy ( Attal et al, 1996 ). However most patients eventually relapse after autologous stem cell transplantation (ASCT) and the optimal conditioning regimen is still unknown. Taking into account the clear relationship between dose and response established in patients treated with melphalan for MM ( Samuels & Bitran, 1995), we have shown that an increased dose of 220 mg/m2 i.v. melphalan (HDM220) followed by ASCT was feasible in a pilot study including 27 patients with advanced MM ( Moreau et al, 1999 ). In this trial, we have observed an encouraging 85% response rate, including an 11·1% complete response (CR) in heavily pretreated patients who had relapsed after a prior intensive therapy.
Thus, we have conducted a phase II trial evaluating the combination of an anti-IL-6 murine mAb with DXM and HDM220 (anti-IL-6/DXM/HDM220) followed by ASCT in 16 patients with advanced MM. The aim of this study was to improve the CR rate achieved with HDM alone without increasing the toxicity of high-dose therapy (HDT).
PATIENTS and METHODS
Patients Sixteen patients with advanced MM were entered into this study (Table I) . At the time of anti-IL-6/DXM/HDM220 therapy, two patients had never responded to standard chemotherapy and were considered as primary refractory patients. The other 14 patients (81·2%) were in relapse after having responded to a prior high-dose therapy including HDM alone or in combination with total body irradiation (TBI) followed by ASCT. Relapsed patients were refractory to a salvage chemotherapy regimen in 10 cases and chemosensitive in four cases.
Conditioning regimen and ASCT Two hundred and fifty milligrams of BE-8, an anti-IL-6 murine mAb ( Klein et al, 1991 ) (provided by Dr Wijdenes, Diaclone Research, Besançon, France), was infused over 4 d (100 mg on day 1 by infusion over 2 h, then 50 mg/d on days 2–4) in combination with 40 mg/d DXM (days 1–4) followed by HDM220 (infusion over 30 min under forced diuresis) on day 5 and ASCT on day 7 [peripheral blood stem cells, median dose of 4·8 × 106 granulocyte–macrophage colony-forming units (GM-CFU)/kg and of 2·4 × 104 CD34/kg].
Capacity of BE-8 and DXM to inhibit IL-6 in vivo C-reactive protein (CRP) levels directly reflect IL-6 production in vivo. Thus, the combined activity of BE-8 and DXM to inhibit IL-6 in vivo was evaluated by their capacity to annihilate the hepatic production of CRP ( Klein et al, 1991 ). CRP serum levels were quantified at the beginning and at the end of BE-8 plus DXM therapy by highly sensitive laser nephelometry (sensitivity < 0·16 mg/l) ( Adam et al, 1985 ).
The tolerance of BE-8 was excellent and no adverse side-effects were documented.
One toxic death was observed. A 61-year-old woman in refractory relapse at the time of anti-IL-6/DXM/HDM220 died 40 d after ASCT from a streptococcus septicaemia. She was considered as a responding patient as the M-component had decreased by 60% at the time of death.
The median duration of hospitalization was 28 d (range 19–40). The median duration of neutropenia < 0·5 × 109/l and < 1 × 109/l was 10 d (range 7–13) and 11 d (range 7–17) respectively. The median time to an unsupported platelet count > 25 × 109/l was 15 d (range from 4 to not yet reached). Two patients never achieved a platelet count of 25 × 109/l without transfusion. They had been heavily pretreated before stem cell collection and HDM220 with regimens including TBI. Both patients had refractory disease at the time of anti-IL-6/DXM/HDM220 and did not respond to this therapy. Ten patients (62·5%) experienced severe grade 4 mucositis with a median duration of 7 d (range 5–13).
Inhibition of CRP production
The median value of CRP serum levels was 7·3 mg/l (range 0·42–56·4) before BE-8 and DXM therapy ( Fig 1) . In each case, the CRP value dropped at the end of BE-8 and DXM therapy. The final median value of CRP serum levels was 0·37 mg/l (range < 0·16–11·7) (P < 0·01, chi-squared test). Of note, CRP serum levels were always detectable at the beginning of the treatment but became completely undetectable at the end of the treatment in 7 out of 14 evaluable cases (P < 0·018, chi-squared test), suggesting a complete inhibition of IL-6 in vivo in these seven cases.
Three out of 16 patients did not respond to anti-IL/6-DXM/HDM220 and two of them died from progressive disease. Thirteen patients (81·2%) experienced a response. Six patients (37·5%) achieved CR that was confirmed by immunofixation and seven (43·7%) achieved partial response (PR). Seven out of 13 responding patients experienced a relapse and one of them died of progressive disease. With a median follow-up of 12 months (range 5–19), five patients are alive in PR (four) or CR (one) and seven patients are alive with progressive disease. One patient who had relapsed 8 months after HDM220 alone is still in PR 16 months after anti-IL-6/DXM/HDM220. The actuarial event-free survival (EFS) was 16·1% at 16 months (median duration 10 months) and the actuarial overall survival (OS) was 68·1% at 20 months.
It was possible to correlate the decrease of CRP levels during BE-8 and DXM therapy with the outcome and response after intensive therapy. Among the seven patients with undetectable serum CRP levels at the end of BE-8 therapy, five achieved CR that was confirmed by immunofixation compared with one CR in the group of seven patients with detectable serum CRP levels at the end of BE-8 therapy (P = 0·103). Of major interest, the median value of CRP levels after BE-8 treatment was undetectable in the six patients achieving CR after intensive therapy and significantly lower than that of the eight other patients without CR after intensive therapy (P < 0·05, Wilcoxon rank sum test).
Comparison with a series of patients treated with HDM220 alone without DXM and BE-8 therapy
We have previously reported a series of 27 patients with advanced MM treated with HDM220 followed by ASCT ( Moreau et al, 1999 ). The clinical and biological characteristics in these 27 cases were similar to those of the present series, although the number of patients with refractory relapse and the number of patients who received prior intensive therapy including double transplant programmes at the time of HDM220 were higher in the group treated with anti-IL-6 (Table I). The CR rate confirmed by immunofixation achieved after HDM220 alone was 11·1% compared with 37·5% in the present trial (P = 0·058). The OS and EFS survival rates were did not differ in the two series (Table I). Nevertheless, the EFS rate was 0% at 12 months in 10 patients with refractory relapse in the series treated with HDM alone, whereas 3 out of 10 patients with the same disease status at the time of anti-IL6/DXM/HDM220 therapy experienced sustained responses (one patient experienced a sustained response of 12, one a sustained response of 13 and one patient experienced a sustained response of 16 months).
To date, no randomized trial has shown that ASCT is the best therapeutic option in patients with MM relapsing after a prior autograft. Nevertheless, data from the Little Rock group ( Mehta et al, 1998 ) suggest that further intensive therapy after relapse after autotransplantation is beneficial to patients with MM. They have found that approximately 54% of those patients were projected to be still alive 3 years after a new ASCT. The conditioning regimen for these salvage ASCTs were either HDM200 alone (18 cases) or TBI plus HDM140 (14 cases), and the toxic death rate was 9·5%. The overall response rate was 81% with 33% CR, but the criteria for response were not clearly stated in this trial. In a similar population of patients, we have studied the impact of an increased dose of melphalan (220 mg/m2) followed by ASCT. This strategy was feasible with no toxic deaths in our series, and the 85% response rate was high among 27 patients who were heavily pretreated having a median overall survival of 31 months from the salvage ACST ( Moreau et al, 1999 ). Our findings therefore were very similar to those of the Little Rock group. However, the increased dose of melphalan of our study was associated with a significant extra-haematological toxicity, mainly grade 4 mucositis documented in 63% of cases with a median duration of 8 d. Thus, it seems difficult further to increase the doses of i.v. melphalan.
To improve the response rates and especially the CR rates which are correlated with survival ( Vesole et al, 1994; Samuels & Bitran, 1995) without increasing the toxicity, we have investigated a new approach, i.e. the combination of high-dose therapy HDM220 with anti-IL-6 murine mAb and DXM. We observed a strong inhibition of IL-6 activity in each patient, and the anti-IL-6 therapy did not add any significant toxicity. Although the overall response rate was similar when patients were treated with either HDM alone or anti-IL-6/DXM/HDM, the CR rate (established after immunofixation) seems to be superior in the trial using anti-IL-6 therapy. Moreover, these cases of CR were mainly observed among the group of patients with undetectable serum CRP levels at the end of BE-8 therapy, suggesting that the complete inhibition of IL-6 activity in vivo observed in this subgroup of patients could favour CR. One could argue that some favourable responses observed after such a therapy could only reflect a low tumour burden in patients with relatively low CRP levels before anti-IL-6 therapy. Figure 1 clearly shows that at least two patients who entered CR after BE-8 therapy had high CRP levels (one of 10 and one of 13·8 mg/l) before this treatment. In contrast, three other patients who had relatively low CRP levels below the median value of 7·3 mg/l before BE-8 therapy and always had CRP levels detectable after anti-IL-6 therapy did not achieve CR. These observations favour the hypothesis of a strong correlation between the CR rate and the ability of anti-IL-6 to inhibit CRP production that is not only due to the tumour burden itself. We cannot exclude the possibility that the improved response rate could also be partially attributed to the dexamethasone added to the anti-IL-6 therapy.
The combination of HDM with DXM and anti-IL-6 therapy that is tolerable and at least as effective as HDM220 alone is a new conditioning regimen that may induce a higher CR rate because of its increased ability to reduce the tumour burden after neutralization of IL-6 in vivo. Prospective and randomized trials of intensive therapy followed by ASCT comparing HDM alone with anti-IL-6/DXM/HDM are ongoing.