We strongly agree with Rajkumar (Haematological cancer: Lenalidomide maintenance—perils of a premature denouement. Nat. Rev. Clin. Oncol. 9, 372–374; 2012)1 that it is premature to recommend lenalidomide maintenance to all patients with myeloma, but we would also like to underline other aspects of the problem not discussed in the recent commentary article.

It is our assertion that not every patient with myeloma needs maintenance therapy. In addition to considering the risk–benefit ratio for a patient during maintenance therapy, we are convinced that the well-known concept of disease plateau—which was established decades before the novel-agent era in myeloma2—is also applicable in the setting of maintenance therapy. Furthermore, a small percentage of patients (15%, but this could improve with novel agents used in the pre-transplant setting) after autologous stem-cell transplantation (ASCT) achieve disease-free survival or progression-free survival of longer than 10 years.3 For those patients, any maintenance therapy is useless or, even worse, harmful. When considering maintenance therapy, especially after ASCT, the primary aim should be a survival benefit for the patients. Other considerations should include patient quality of life and cost effectiveness.

To assess the value of maintenance therapy we should define the meaning of maintenance therapy as compared with consolidation therapy—if measurable disease is present, we should define a given therapy as 'consolidation'. This aspect is very important, because the advantages in overall survival and progression-free survival seen in patients receiving maintenance therapy in the trial published by McCarthy et al.4 was in patients who had not achieved a complete response on their previous therapy. The presence of measurable disease in these patients was not commented on by Rajkumar1 in his commentary article, and compares with the patients with complete response in the trials conducted by Palumbo et al.5 and Attal et al.6 In both of these trials, an overall survival benefit was not associated with lenalidomide maintenance therapy, although a progression-free survival benefit was seen.5,6

It seems that there is no doubt that lenalidomide is useful in patients with residual disease7 (that is, as consolidation therapy); however, we have some concerns that lenalidomide treatment can be useful in patients who have already achieved complete response. With the advent of new agents, achieving a complete response after ASCT is now more likely than in the past,8 and we should better understand in future trials those patients who will benefit from lenalidomide or bortezomib maintenance therapy and those who will not. For these reasons we believe that lenalidomide use should be rationally planned.