Autologous Hematopoietic Stem Cell Transplantation in Light Chain Amyloidosis (AL) With Renal Involvement

Letter to the Editor

Approximately 70% of patients with primary light chain amyloidosis (AL) present with kidney involvement. Achievement of renal response is associated with improved overall survival (OS). OS for patients with AL treated with conventional chemotherapy ranges from 13–18 months (1) with many progressing to end stage renal disease (ESRD), and a poor long-term outcome. High dose (HD) melphalan followed by autologous hematopoietic stem cell transplantation (auto-HCT) has been shown to improve organ function and OS by inducing high rates of both hematologic and organ responses (2). Organ response rates range from 31%–46% and are time-dependent, with a median time to best response of up to 1 year (3).

In a large 8-year longitudinal study of patients undergoing HD melphalan and auto-HCT, Skinner et al demonstrated an improved median OS of 4.6 years (4). Historically, the treatment related mortality (TRM) of this approach has been as high as 25–40% (5). However, with better patient selection and improvements in supportive care, TRM has now decreased to less than 10% (3).

We studied 74 consecutive patients with AL treated with HD melphalan and auto-HCT at our center between the years 1999 and 2011. Fifty five had renal involvement and were included in this study.

Granulocyte colony-stimulating factor-primed peripheral blood progenitor cells were collected using standard mobilization protocols and apheresis techniques in all patients.

All patients received HD melphalan (n=53) or melphalan-based (n=2) preparative regimens. Thirty-eight (69%) patients received melphalan 200 mg/m², while 17 (31%) received melphalan at lower doses at the treating physician’s discretion. Filgrastim 5 μg/kg was administered daily from day +1 for those transplanted prior to 2010, or day +5 for those transplanted 2010 onwards, until neutrophil engraftment. All patients received infection prophylaxis with sulfamethoxazole-trimethoprim or pentamidine and acyclovir or valacyclovir for 6 months.

Renal involvement and responses were defined according to the International Consensus Criteria, which defines renal involvement as >0.5 grams of proteinuria per 24-hour period (6). Renal response was defined as a 50% reduction in proteinuria in 24 hours without worsening of creatinine or creatinine clearance by ≥25% over baseline, and renal progression as a 50% increase in proteinuria in 24 hours (at least 1g/24 hours) or >25% increase in creatinine or creatinine clearance (7). Patients were staged according to Palladini et al (8) and were stratified according to proteinuria >5g/24 hours and estimated glomerular filtration rate (eGFR) <50 mL/min. Patients with none, one, or two of these parameters were stratified as stages 1, 2 or 3, respectively.

Neutrophil engraftment was defined as the first of 3 consecutive days with an ANC ≥0.5 ×10⁹/L. Platelet engraftment was defined as the first of 7 consecutive days with a platelet count ≥20 ×10⁹/L without transfusion support. Adverse events were graded according to the National Cancer Institute Common Toxicity Criteria, version 4.

Renal responses were assessed at 6, 12, and 24 months. Hematologic responses (HR) were determined by the International Myeloma Working Group Criteria (IMWG) or revised International Society of Amyloidosis criteria, when available (9, 10) and were assessed at 1, 3, 6, and 12 months, and approximately every 3–6 months thereafter. FLC assays were performed at diagnosis, prior to auto-HCT, and at each subsequent evaluation.

Primary endpoints were OS and PFS. Secondary endpoints were TRM, toxicity rates, hematologic and renal response rates, and the correlation between hematologic and renal response. Predictors of OS and PFS were evaluated using Cox proportional hazards regression analysis.

Patient characteristics are summarized in Table I. Median bone marrow plasma cell infiltration at diagnosis was 6% (range, 1–40%). Forty-one (74.5%) patients had AL with <10% bone marrow plasma cells at diagnosis (AL only). Fourteen (25.5%) patients had AL with >10% bone marrow plasma cells at diagnosis, (AL + PC>10%).

Table I.

Patient Characteristics

Parameter	N (%)/median [range]
Time from Dx to auto-HCT (mos)	6.6 [2–120]
Age (years)	58 [41–74]
Male	31 (56)
λ light chain subtype	41 (75)
Absolute difference between involved and uninvolved serum free light chains (dFLC: mg/L)	27 [0–2159]
Median serum albumin before auto-HCT (g/dl)	2.6 [1.5–4.3]
Other visceral organs involved	14 (25)
Heart	5 (9)
Liver	5 (9)
GI	3 (5)
Peripheral nerves	1 (2)
Induction Rx before auto-HCT*	42 (76)
Proteinuria (g/24 hrs)	5 [0.17–28]
Serum creatinine (mg/dL)	1.20 [0.53–9.28]
≤2.0	44 (80)
>2.0	11 (20)
eGFR (mL/min/1.73 m2)	54.6 [10.3–155.4]
Renal Stage
1	16 (29)
2	27 (49)
3	11 (20)
Hemodialysis-dependent	2 (4)

Forty-two (76%) patients received induction therapy prior to auto-HCT, thirty-three (60%) with novel agents: IMiD-based with lenalidomide in 7 (13%) patients and PI-based with bortezomib in 26 (47%) patients. Nine (16%) received induction with conventional agents: dexamethasone in 5 (9%), melphalan-based in 3 (5%), and cladribine in 1 (2%) patient.

Median time to neutrophil and platelet engraftment was 10 (range, 7–15) and 12 (range, 6–23) days, respectively. Grade 2–4 nonhematologic toxicities included nausea in 23 (42%), infection in 18 (33%), diarrhea in 15 (26%), mucositis in 12 (22%), and cardiovascular events in 11 (20%). Renal toxicity was seen in 7 (13%) patients. Six patients died within 100 days and 7 died within 1 year, with 100-day and 1-year TRM of 11% (95% CI 5–23) and 13% (95% CI 6–26), respectively. Causes of death at 1 year included infection/sepsis/multiorgan failure in 6 patients, and sudden cardiac death in 1 patient. There was a significant decrease in the TRM when comparing auto-HCT performed after 2008 with those performed before 2008 (9% vs. 20%).

Overall rates HR was observed in 43 (86%) of 50 evaluable patients after auto-HCT, with 8 (16%) complete responses (CR), 12 (24%) very good partial responses (VGPR), and 23 (46%) partial responses (PR). This compares to 21 of 51 patients (41%) who achieved a HR prior to auto-HCT. Median time to HR was 3.1 (range,1–24) months. Fifty, 42 and 35 patients were evaluable for renal response at 6, 12 and 24 months, respectively. Renal responses were seen in 8 (16%), 13 (31%) and 17 (49%) patients at 6, 12, and 24-months, respectively. Seventeen of 35 evaluable patients achieved a renal response at 2 years. There was no difference in renal response between patients achieving >VGPR (50%), PR (50%) and <PR (33%, p=0.67). Two patients required hemodialysis prior to auto-HCT and remained hemodialysis-dependent afterward.

With a median follow up of 37 (range, 4–126) months, the estimated 3-year PFS and OS rates were 54% and 73%, respectively (Figure 1). At last follow up, 31 (56%) patients were alive and in remission.

Figure 1:

Progression-free and overall survival for all 55 patients who received auto-HCT.

Median follow up from auto-HCT 37 [4–126]

At last follow up, 31 pts (56%) were alive and in remission

On multivariate analysis, patients not treated with IMid or PI-based induction and dFLC > 26 mg/dL had significantly shorter 3-year PFS (15% vs. 83%, HR=7.6, p=0.002) and significantly shorter 3-year OS (49% vs. 95%, HR=9.5, p=0.004).

Our results support the role of auto-HCT by demonstrating >80% HR rate and almost 50% renal response rate at 2 years. Use of induction therapy and a lower disease burden at the time of auto-HCT, as defined by dFLC ≤26 mg/dL, were associated with longer PFS and OS in a multivariate model.

The association of organ response with HR was not unequivocal in our study due to small numbers and almost 30% patients being inevaluable for a response. We saw a steady improvement in renal response over 2 years, with 16%, 31% and 49% renal responses at 6, 12 and 24 months, respectively. In a similar study from Mayo Clinic, Leung et al (11) reported a renal response rate of 61% at 2 years. Although the patients were comparable, a lower response in our cohort may be due to the inclusion of dialysis-dependent patients, use of a lower dose of melphalan in a subset of patients, a lower rate of hematologic CR, and a smaller number of patients evaluable for renal response.

Huang et al (12) conducted a randomized clinical trial where patients either received induction with bortezomib + dexamethasone or no induction before HD melphalan and auto-HCT. Renal response at 2 years was 75% in the induction arm and 53% in the non-induction arm. The better renal response rate may be due to a lower baseline serum creatinine (0.8 mg/dl vs. 1.2 mg/dl in our study), a higher hematologic CR in his study (68%) and the uniform use of bortezomib in patients treated on the bortezomib arm.

The role of induction therapy prior to auto-HCT in AL is evolving. However, several studies have failed to show improvement in outcomes. Perz et al. did not show any improvement in HR with the use of VAD induction before auto-HCT and the induction treatment was associated with 7% mortality (13). Some more recent studies, however, support the role of PI or IMiD-based induction in improving outcomes after auto-HCT (12). The significance of reduction of FLC as a predictor of survival was first reported by Dispenzieri et al. (14). A recent report from the National Amyloidosis Center in the UK identified the magnitude of the FLC response at 6 months from baseline as a strong predictor of PFS and OS (15).

HD melphalan is nephrotoxic, and some studies have reported a high risk of renal toxicity. In our study, grade 2–3 renal toxicity was seen in 7 patients (13%), and was reversible. All 7 patients had prior renal insufficiency. Forty two (76%) patients were transplanted between 2006 and 2011, which may be one of the reasons for a relatively short median follow up of 37 months. We showed a decrease in TRM after 2008, likely due to better patient selection, use of induction therapy, and improvements in supportive care.

HD melphalan followed by auto-HCT is a safe and effective treatment option for select patients with AL with renal involvement, with high rates of hematologic and renal responses, long PFS and OS, and acceptable rates of toxicity.