Plasma Exchange for Acute Renal Failure of Myeloma—Logical, Yet Ineffective

It has been estimated that in the US multiple myeloma will be diagnosed in 15,980 persons and that 11,300 persons will die from it, constituting 2% of all deaths from cancer (1, 2). At the time of diagnosis only 52% of patients have normal serum creatinine concentrations (3) and in different series 12 to 20% were found to present with acute renal failure (4–6). Plasmapheresis has been shown to remove light chains, but the information on its effect on recovery of renal function had remained controversial. An uncontrolled observational series of 50 patients with myeloma and acute renal failure had initially shown that more patients who had chemotherapy plus plasma exchange recovered renal function (61% of cases) than patients who had been treated only with chemotherapy (27%) (7). A further report had stated that all 3 patients with multiple myeloma and severe acute renal failure recovered renal function after plasmapheresis (8). These encouraging early observations prompted Zucchelli (9) to perform a small, controlled, randomized trial on a sample of 29 patients with multiple myeloma and Bence-Jones proteinuria >1 g/d who presented with acute renal failure. Dialysis was required in 24 cases and 5 cases had serum creatinine concentrations in excess of 5 mg/dl. Patients were randomly allocated to plasma exchange plus corticosteroids, with cytotoxic drugs and hemodialysis when required (n=15), or to peritoneal dialysis together with corticosteroids and cytotoxic drugs (n=14). As anticipated, patients treated with plasmapheresis, but not the patients without, had a dramatic reduction of Bence-Jones proteinuria. More importantly, 13 of the 15 patients with plasmapheresis, but only 2 of the patients without, recovered renal function defined as serum creatinine <2.5 mg/dl. In this underpowered study with different renal replacement therapies in the two arms of the study, the 1-yr survival was significantly higher in patients with plasmapheresis (66%) than in patients without (28%). In contrast, however, a subsequent controlled randomized study by the Mayo clinic failed to show a benefit of plasmapheresis (10). A total of 21 patients with incident or prevalent active myeloma and progressive renal failure, a somewhat different and also small group, was randomized to either forced diuresis, chemotherapy, and plasmapheresis or to forced diuresis and chemotherapy: Of the 5 patients already on dialysis who received plasmapheresis, only 3 recovered, and of the polyuric patients the number of recoveries was similar in the two groups (i.e., 7 versus 5 patients). The 6-mo mortality was 20% in either group and rose to 60 to 80% at 1 yr. This study is somewhat difficult to interpret because it studies patients presenting with renal failure due to myeloma (incident patients) as well as patients with established renal failure due to myeloma (prevalent patients). Despite these less than convincing results, several authors (11–13) as well as guidelines (14, 15) recommended plasmapheresis in such patients. Such recommendations were based on uncontrolled clinical observations, which admittedly were occasionally impressive—however, plasmapheresis by necessity was never used as a treatment modality in isolation, but was always combined with cytotoxic therapy so that it was impossible to find out which did what. In this murky situation, controlled, randomized, prospective information was most welcome, particularly because it has been well documented that reversal of renal failure was a more important prognostic factor than response to chemotherapy (4, 16, 17). Renal dysfunction impacted on outcome even after bone marrow transplantation (2). Consequently, if effective, plasmapheresis might confer substantial clinical benefit. This randomized, open, controlled, Canadian study (18) was conducted between 1998 and 2004. Patients were stratified by chemotherapy and dialysis dependence and subsequently randomized to receive either 5 to 7 plasma exchanges in addition to conventional therapy (50 ml/kg body weight with acid citrate as anticoagulant and 5% human albumin and normal saline as replacement fluid) within the first 10 d of study entry or conventional treatment. The primary endpoint was a composite outcome comprising death, dialysis dependence, and GFR <30 ml/min per 1.73 m2 at 6 mo. What were the results? In the control group 69.2% and in the plasmapheresis group 57.9% of patients had an event (composite outcome). The unadjusted odds ratio (OR) for the primary composite outcome was even higher in the plasmapheresis group (OR 1.71), but it decreased to 1.20 when adjusted for baseline treatment with the VAD scheme (vincristin, doxorubicin, methylprednisolone), for staging of myeloma (according to Durie-Salmon), dialysis, age, serum albumin concentration, and urine protein excretion; the 95% confidence intervals were very large, however (0.73 to 4.01 and 0.42 to 3.44, respectively). The cumulative 6-mo survival was also similar in the two groups, 66.7% versus 67.2% in the control versus the plasmapheresis group, and the same was true for several other secondary outcomes. What can be concluded from the results? The authors are presumably correct in stating that it is unlikely that plasmapheresis yields a clinically meaningful outcome in patients with acute renal failure at the onset of myeloma. So the guidelines were dead wrong. It is highly unlikely that we shall ever get a better and more penetrating study, so in the future decisions on patient management must be based on the above study results. Nevertheless, it is useful to point to some of its shortcomings. First, the huge confidence intervals illustrate that the study was still underpowered, although it exceeded the previous studies by a wide margin. As a result it cannot be totally excluded that plasmapheresis might be beneficial in specific subgroups. Second, the authors failed to monitor urinary excretion, or better plasma concentrations, of light chains. This resembles antihypertensive treatment without measuring BP. Third, despite concealed central randomization there was a substantial imbalance in treatment allocation, i.e., 39 patients were evaluable in the control and 58 patients in the plamapheresis group. Fourth, the authors did not perform renal biopsies. It is not likely, but cannot be excluded, that the study included patients who failed to have cast nephropathy, particularly since renal biopsy studies showed considerable heterogeneity of renal findings in myeloma patients (19–21). What is remarkable is that the proportion of these incident patients who became independent of dialysis was large, i.e., 39.5% overall, 7/19 in the control and 10/24 in the plamapheresis group. By contrast, previous series had reported that no more than 3 to 12% recovered renal function (4, 22 to 25). The most plausible, but not the only, interpretation is that modern treatment has become much more effective than what had been available in the past (26). This may possibly also have contributed to the observation that today plasmapheresis on top of modern cytotoxic treatment no longer improves outcome.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.