SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Objective

To characterize symptoms and signs of AL amyloidosis that may bring patients to the attention of rheumatologists, evaluate Ig VL gene usage in this subgroup of patients, and assess the impact of soft tissue and bone involvement and VL gene usage on survival.

Methods

Clinical features of soft tissue and bone involvement were assessed in 191 patients with AL amyloidosis. VL gene sequencing was carried out to determine light-chain family, rate of somatic mutation, and evidence of antigen selection. The association of soft tissue and bone involvement with VL gene usage was assessed by logistic regression analysis, and survival time was analyzed using log rank tests and Cox regression models.

Results

Soft tissue and bone involvement occurred in 42.9% of the patients, and 9.4% had dominant soft tissue and bone involvement. The most common manifestations were submandibular gland enlargement, macroglossia, and carpal tunnel syndrome. Dominant soft tissue and bone involvement was significantly associated with VLκI gene usage. Mutation rate and evidence of antigen selection in the VL genes were not found to be confounding factors, providing evidence against a contribution of autoimmunity in this type of AL amyloidosis. Survival time was initially longer in patients with dominant soft tissue and bone involvement than in patients with other dominant organ involvement; however, this difference diminished over time.

Conclusion

Amyloid infiltration into soft tissue, joints, periarticular structures, and bones can bring patients with AL amyloidosis to the attention of rheumatologists. Recognition of the presenting symptoms is essential for accurate diagnosis and appropriate treatment, since the long-term outlook for untreated patients with dominant soft tissue and bone involvement is not better than that for patients with other dominant features of AL amyloidosis.

Primary or AL amyloidosis is a plasma cell disorder characterized by the overproduction and tissue deposition of a monoclonal Ig light chain or fragments containing the light-chain variable region (VL). The deposits form amyloid fibrils that can be identified by Congo red staining. Light chain deposition produces tissue damage and eventually organ failure, leading to death in untreated patients within 1–1.5 years (1). A variety of effective treatments are now available for patients with AL amyloidosis, including dose-intensive intravenous melphalan followed by autologous stem cell support (2) and dexamethasone-based regimens (3, 4). The immunomodulatory drugs thalidomide (5, 6) and lenalidomide (7) have also been found to be effective.

One of the most important determinants of outcome is early diagnosis, since severe amyloid organ disease may preclude the use of potentially effective treatment regimens. Amyloid cardiomyopathy can rapidly lead to diastolic dysfunction, heart failure, and arrhythmias that can complicate use of chemotherapy and steroids. Amyloid kidney disease can alter drug clearance and make the kidney highly susceptible to nephrotoxic insults. In the liver, amyloid involvement can produce cholestatic liver disease, and alter drug metabolism. Involvement of the gastrointestinal tract itself alters oral absorption and can lead to malnutrition.

AL amyloidosis can also present with symptoms and signs that mimic a variety of rheumatic conditions. Skin thickening due to amyloid deposits may simulate scleroderma (8). Amyloid infiltration into periarticular and synovial tissue can produce stiffness and multiple joint swelling resembling rheumatoid arthritis (9, 10). Xerostomia caused by amyloid deposition in salivary glands can be mistaken for Sjögren's syndrome (11, 12). Such symptoms and signs of a systemic disorder may bring patients to the attention of rheumatologists, and must be distinguished from authentic rheumatic diseases. This study was undertaken to characterize features of such involvement in a large group of patients with AL amyloidosis, analyze the VL gene sequences associated with this type of disease, and assess the impact of these manifestations on survival.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Patient clinical and laboratory assessment.

One hundred ninety-one patients with biopsy-proven systemic AL amyloidosis who were evaluated in the Amyloid Treatment and Research Program at Boston University School of Medicine from September 1999 to September 2005 were included in the study. Fifteen patients with AL amyloidosis had multiple myeloma. Informed consent for data and sample collection was obtained with permission of the Boston University Medical Center Institutional Review Board, in accordance with the Declaration of Helsinki.

At initial evaluation, patients were categorized based on clinical and laboratory manifestations of renal, cardiac, hepatic, gastrointestinal, neuropathic, pulmonary, or soft tissue involvement according to recently published consensus criteria (13). The organ system producing the most significant clinical symptoms was determined to be the dominant organ system by the treating physicians.

Soft tissue, joint, and bone involvement was diagnosed by symptoms and physical findings of macroglossia, submandibular gland enlargement, cervical adenopathy, carpal tunnel syndrome (CTS), muscle pseudohypertrophy, amyloid arthropathy (13), and vertebral compression fractures. Arthropathy was identified by firm thickening of synovial membranes and periarticular tissue. Although biopsies of such tissue were rarely available, all patients had biopsy-proven amyloid deposits either in the abdominal fat, in the bone marrow microvasculature, or in an involved organ. In cases of vertebral compression fractures, the presence of amyloid was confirmed by bone biopsies; other causes of these processes were excluded. CTS was defined by the examining neurologist on the basis of history, presence of Tinel's sign or Phalen's sign, electrophysiologic changes, or based upon pathologic finding of Congo red–positive material in tissue obtained during surgical release. Of note, macroglossia and the “shoulder pad” sign caused by periarticular soft tissue deposition of amyloid fibrils were pathognomonic for AL amyloidosis (14); the other manifestations were less specific.

Patients with any soft tissue, joint, or bone manifestation were recorded as having soft tissue and bone involvement, while patients in whom soft tissue, joint, or bone symptoms were the major manifestation of disease were considered to have dominant soft tissue and bone involvement. Skeletal symptoms of multiple myeloma, such as bone pain, osteolytic lesions, or generalized osteoporosis, were not considered to be evidence of soft tissue and bone involvement due to AL amyloidosis.

The presence of amyloid was visualized by Congo red staining, producing apple-green birefringence under polarized light. The plasma cell disorder was assessed by immunohistochemical staining of the bone marrow for κ and λ light chains and for CD138+ plasma cells, by serum and urine protein immunofixation electrophoresis, and by quantitative Ig and free Ig light chain testing.

VL gene sequencing.

Bone marrow aspirates from the 191 patients being evaluated for AL amyloidosis were used for cloning and sequencing of Ig VL genes as previously described (15). Briefly, the aspirate cells were treated with ammonium chloride to lyse red blood cells (16). Total RNA was extracted, and complementary DNA was synthesized, amplified by multiplex polymerase chain reaction (PCR) with a set of 5′ primers specific for framework region 1 of 7 VLλ (λI, λII/V, λIII, λIVa, λIVb, λVI) and 4 VLκ (κI/IV, κII, κIII) families and 3′ λ or κ C-region primers (Integrated DNA Technologies, Coralville, IA) (17), cloned, and sequenced. In each case, the clonal sequence was determined by the identity of at least 50% of 6–9 independently cloned and sequenced products. Sequences were compared using the Jellyfish gene analysis software package (Field Scientific, Lewisburg, PA). To correct for nucleotide sequence errors introduced by the framework region 1 primers, once the light chain was identified, additional PCR amplification with 5′ primers for the appropriate VL leader region and 3′ primers for the appropriate C region was performed, and resequencing was carried out. In every case in which the monoclonal sequence had been previously obtained, a single VL region sequence was confirmed and submitted to the GenBank database (accession nos. EF589382–EF589571).

Analysis of VL gene usage and somatic hypermutation.

VL genes with the corrected framework region 1 sequence were evaluated for their homology with the germline donor sequences using a database of rearranged immunoglobulin genes, V BASE (http://www.mrc-cpe.cam.ac.uk), and the International Immunogenetics Information system (IMGT/V-QUEST [http://imgt.cines.fr]) (18). The assignment of the germline gene counterpart was made based on maximum homology of the nucleotide sequences. Homology with germline sequence was determined for complete VL genes, with the exception of the codons associated with the V–J junction and framework region 4. All sequences were functional without stop codons, frameshift mutations, or pseudogenes. The number of somatic mutations in the individual VL gene sequences was assessed by comparing each sequence with its closest germline donor from the beginning of framework region 1 to the end of the VL segment, excluding the V–J junction and framework region 4. The mutation rate was calculated by dividing the number of nucleotide substitutions by the total number of nucleotides in the individual monoclonal VL gene sequence.

Statistical analysis.

VL sequences derived from post–germinal center B cells that have undergone somatic hypermutation in response to antigen selection and affinity maturation display a relative increase in replacement mutations within the complementarity-determining regions (CDRs), while replacement mutations are suppressed in the framework regions in order to maintain light chain structure. The probability that an excess or scarcity of observed replacement mutations in the framework regions and CDRs occurred by chance was calculated using the multinomial distribution model described by Lossos et al (19, 20). The model was used to predict the number of replacement mutations, the predicted number was then compared with the number of observed replacement mutations, and P values for both framework regions and CDRs were calculated.

To determine the association of soft tissue and bone involvement and dominant soft tissue and bone involvement with VL gene usage, baseline variables were compared between patients with and those without soft tissue and bone involvement or dominant soft tissue and bone involvement, using the t-test for continuous symmetrically distributed variables, the Mann-Whitney test for continuous asymmetrically distributed variables, and the chi-square test and Fisher's exact test for categorical values. To calculate the odds ratios (ORs) associated with the presence or absence of different VL families in patients with AL amyloidosis, logistic regression analysis was performed. Soft tissue and bone involvement and dominant soft tissue and bone involvement were evaluated as the dependent variables, and κ and λ VL families as well as 4 κ (κI, κII, κIII, κIV) and 6 λ (λI, λII, λIII, λIV, λVI, λX) VL subfamilies were used as the independent variables. To control for possible confounding variables, age, sex, history of multiple myeloma, mutation rate per monoclonal sequence, and evidence of antigen selection in VL sequences were forced into the model. For all analyses, P values less than 0.05 were considered significant. In all post hoc tests, the significance level was adjusted using the Bonferroni correction.

To assess the impact of soft tissue and bone involvement on survival, survival of patients with dominant soft tissue and bone involvement was compared with that of patients with other dominant organ involvement. For patients who survived through the end of the study period, data were considered censored at that time point. Survival curves were estimated using the Kaplan-Meier method, and compared using the Wilcoxon test and log rank test. The 95% confidence intervals (95% CIs) for differences in survival medians were estimated as previously described (21). The assumption of proportional hazards was assessed prior to attempting Cox modeling, and a modified Cox model was used in place of the standard model to account for nonproportional hazards. A dummy variable for the effect of dominant soft tissue and bone involvement on survival beyond a threshold (t0) was used, and assumed that proportional hazards before and after this threshold remained valid. Akaike's information criterion was used to determine the threshold that achieved the best fit for the model.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Incidence of soft tissue and bone involvement and dominant soft tissue and bone involvement.

The demographic, clinical, and laboratory characteristics of the 191 patients with AL amyloidosis with or without soft tissue and bone involvement are shown in Table 1. Soft tissue and bone involvement was found in 82 of the 191 patients (42.9% [95% CI 35.9–50.3%]). Patients with soft tissue and bone involvement and those without soft tissue and bone involvement were similar with regard to many features, including age, sex, history of multiple myeloma, percentage of bone marrow clonal plasma cells, number of patients with monoclonal protein in serum and urine, number of patients with VLκ and VLλ gene families, and proportion of patients with different dominant organ systems involved. However, patients with soft tissue and bone involvement had significantly higher numbers of involved organs (median 4 versus 2; P < 0.001). Eighteen of the 82 patients with soft tissue and bone involvement (22%) had dominant soft tissue and bone involvement; thus, dominant soft tissue and bone involvement occurred in 9.4% of all of the patients with AL amyloidosis.

Table 1. Demographic, clinical, and laboratory features of the 191 patients with AL amyloidosis according to the presence or absence of soft tissue and bone involvement
FeaturePatients with soft tissue and bone involvement (n = 82)Patients without soft tissue and bone involvement (n = 109)
  • *

    P < 0.001 versus patients with soft tissue and bone involvement.

  • Dominant soft tissue and bone involvement was excluded from the analysis of dominant organ system involvement since it was present only in the soft tissue and bone involvement group; therefore, for the soft tissue and bone involvement group, n = 64 patients.

  • P values less than 0.01 were considered significant for distribution in the patients with soft tissue and bone involvement versus the patients without soft tissue and bone involvement, with Bonferroni adjustment for multiple comparisons.

  • §

    P = 0.011 versus patients with soft tissue and bone involvement.

Age, median (range) years60.8 (36.5–79.8)60.9 (32.6–80.4)
Sex, no. (%) male/female54 (65.9)/28 (34.1)57 (52.3)/52 (47.7)
Plasma cell disorder, no. (%)  
 AL amyloidosis73 (89.0)103 (94.5)
 AL amyloidosis plus multiple myeloma9 (11.0)6 (5.5)
% clonal plasma cells in the bone marrow, median (range)10 (5–90)10 (5–40)
Positive findings on serum protein immunofixation electrophoresis, no. (%)68 (82.9)85 (78.0)
Positive findings on urine protein immunofixation electrophoresis, no. (%)74 (90.2)94 (86.2)
VL gene family, no. (%)  
 κ22 (26.8)29 (26.6)
 λ60 (73.2)80 (73.4)
No. of organ systems involved, median (range)4 (1–6)2 (1–6)*
Dominant organ system involved, no. (%)  
 Cardiac30 (46.9)29 (26.6)§
 Renal24 (37.5)58 (53.2)
 Neuropathic5 (7.8)7 (6.4)
 Hepatic/gastrointestinal3 (4.7)13 (11.9)
 Pulmonary2 (3.1)2 (1.9)

Clinical presentation of soft tissue and bone involvement.

Submandibular gland enlargement was the most common finding in the group of patients with soft tissue and bone involvement; this feature was found in 61 of the 191 patients with AL amyloidosis (31.9%). Typically, the enlarged submandibular glands (Figure 1A) were reported to be firm and nontender. Macroglossia was the second most common finding and was present in 44 of 191 patients (23.0%). The degree of macroglossia varied from slight tongue thickening with tooth indentation to massive enlargement (Figure 1B) and interference with eating, swallowing, speaking, and breathing. CTS, temporally related to the onset of AL amyloidosis and presumed to be due to amyloid deposition in carpal tunnel ligaments, was seen in 25 patients (13.1%). Amyloid deposition in subcutaneous tissue, including breast, eyelid, lower lip, and skin nodules, was seen in 5 patients (2.6%). Pseudohypertrophy of skeletal muscles was found in 3 patients (1.6%). Only 2 patients (1.0%) in this series had cervical lymphadenopathy (Figure 1C) due to amyloid deposition.

thumbnail image

Figure 1. Soft tissue manifestations of AL amyloidosis. A, Submandibular gland enlargement. B, Macroglossia, a pathognomonic feature of AL amyloidosis. C, Symmetric cervical lymphadenopathy due to amyloid deposition. D, Amyloid arthropathy, with firm periarticular thickening of synovial membranes around proximal interphalangeal and metacarpal joints, resembling rheumatoid arthritis.

Download figure to PowerPoint

Amyloid arthropathy, involving various joints, occurred in 7 patients (3.7%). All patients had firm symmetric periarticular thickening of synovial membranes (Figure 1D) associated with pain, stiffness, and limited range of motion without erosive changes. In the majority of cases, multiple joint swelling with contractures was observed. Shoulders (the “shoulder pad” sign) were the most commonly affected joints, followed by the proximal interphalangeal joints, hips, knees, metacarpal joints, wrists, and elbows. Amyloid infiltration into thoracic vertebrae leading to spondylolisthesis of multiple vertebrae occurred in 2 patients (1.0%).

To assess whether a clustering of individual rheumatologic features occurred in patients with AL amyloidosis with dominant soft tissue and bone involvement, or if these features were equally likely to be seen in patients with other organ involvement, the incidence of soft tissue and bone involvement in patients presenting with dominant soft tissue and bone involvement was compared with that in patients presenting with other dominant organ involvement. Macroglossia, arthropathy, amyloid deposits in other soft tissue, muscle pseudohypertrophy, cervical adenopathy, and amyloid infiltration into the vertebrae were observed significantly more frequently in the group of patients with dominant soft tissue and bone involvement, while submandibular gland enlargement and CTS were seen with similar frequency in both groups (Table 2). Thus, submandibular gland enlargement and CTS could occur in patients with AL amyloidosis with any type of dominant organ involvement. The spectrum of soft tissue and bone involvement in the 18 patients who presented with dominant soft tissue and bone involvement is detailed in Table 3.

Table 2. Frequency of various manifestations of soft tissue and bone involvement according to dominant organ involvement*
FeaturePatients with dominant soft tissue and bone involvement (n = 18)Patients with other dominant organ involvement (n = 64)P
  • *

    Values are the number (%) of patients. Other biopsy-proven soft tissue sites of involvement included breast, eyelid, lower lip, and skin nodules.

Submandibular gland enlargement15 (83.3)46 (71.9)0.380
Macroglossia14 (77.8)30 (46.9)0.040
Carpal tunnel syndrome7 (38.9)18 (28.1)0.381
Arthropathy6 (33.3)1 (1.6)<0.001
Other soft tissue sites4 (22.2)1 (1.6)0.001
Muscle pseudohypertrophy3 (16.7)00.009
Cervical adenopathy2 (11.1)00.046
Vertebral lesion2 (11.1)00.046
Table 3. Spectrum of soft tissue and bone involvement in the 18 patients with dominant soft tissue and bone involvement
Patient/age/sexResult of serum protein immunofixation electrophoresisResult of urine protein immunofixation electrophoresisVL gene familyInvolvement*
  • *

    CTS = carpal tunnel syndrome; PIP = proximal interphalangeal; MCP = metacarpophalangeal.

  • Patient had AL amyloidosis with coexisting multiple myeloma.

  • Joint contractures.

1/69/MκκκIMacroglossia, submandibular gland enlargement, CTS, arthropathy (shoulder pad, hips)
2/59/FκκκIMacroglossia, submandibular gland enlargement, CTS, arthropathy (shoulder pad, hips, PIP joints)
3/71/FIgGκκκIMacroglossia, submandibular gland enlargement, arthropathy (shoulder pad, hips, wrists, PIP joints), muscle pseudohypertrophy, other soft tissue (skin nodules)
4/50/MκκκIMacroglossia, submandibular gland enlargement, CTS, arthropathy (shoulder pad, knees, MCP joints, PIP joints)
5/37/MλλVIMacroglossia, submandibular gland enlargement, arthropathy (shoulder pad, elbows, knees), muscle pseudohypertrophy, cervical adenopathy
6/60/MIgGλλIIIMacroglossia, submandibular gland enlargement, CTS, arthropathy (MCP joints, PIP joints), muscle pseudohypertrophy, other soft tissue (eyelid, lip)
7/63/MκκIBone (T4, T6, T7, and T12 compression fractures)
8/59/MIgGλλISubmandibular gland enlargement, bone (T10 and T11 compression fractures)
9/67/MIgAλIgAλλIIMacroglossia, submandibular gland enlargement, CTS
10/74/FλλλIICTS, other soft tissue (skin nodules)
11/68/MκκκIMacroglossia, submandibular gland enlargement, CTS
12/56/MλλλVIMacroglossia, submandibular gland enlargement
13/68/MλλλIMacroglossia, submandibular gland enlargement
14/58/FλλIIIMacroglossia, submandibular gland enlargement
15/70/MκκκIVMacroglossia, submandibular gland enlargement, cervical adenopathy
16/67/FIgGκIgGκκIOther soft tissue (breast)
17/75/FIgAλIgAλ; λλIIIMacroglossia, submandibular gland enlargement
18/57/MκκIMacroglossia, submandibular gland enlargement

In 5 patients, diagnoses of various rheumatic conditions were considered before the diagnosis of AL amyloidosis was made. Four patients presenting with soft tissue and articular symptoms were initially diagnosed as having seronegative rheumatoid arthritis, polymyalgia rheumatica, connective tissue disease, and Sjögren's syndrome, respectively. In 1 patient, restrictive lung disease, muscle weakness, and dysphagia were initially attributed to scleroderma. None of the immunologic markers of these diseases were found in any of the cases.

VL gene sequence characteristics and the presence of soft tissue and bone involvement.

An analysis of VL gene family usage was performed to determine whether specific light-chain sequences were associated with soft tissue and bone involvement in AL amyloidosis. Monoclonal VL sequences, identified in multiple independently amplified and cloned sequences, were successfully characterized in all 191 bone marrow specimens, including specimens from 15 patients who were found to have AL amyloidosis with multiple myeloma. In all cases, the monoclonal light chain type identified in the bone marrow was consistent with the results obtained from immunohistochemistry or immunofixation electrophoresis. Of the monoclonal sequences obtained for each patient, 85–100% were identical (median 98.3%).

The distribution of VL gene family usage did not differ significantly depending on the presence or absence of soft tissue and bone involvement (P = 0.40) (Figure 2). In the group of 18 patients with dominant soft tissue and bone involvement, VLκ and VLλ were equally represented. The most frequently observed light-chain sequence was κI, found in 8 patients, followed by λIII, found in 3 patients, λI, λII, and λVI, each found in 2 patients, and κIV, found in 1 patient.

thumbnail image

Figure 2. Frequency of VL gene family usage in monoclonal plasma cells from 191 patients with AL amyloidosis with and without soft tissue and bone involvement (STBI). The distribution of VL gene family usage in the group of patients with soft tissue and bone involvement did not differ significantly from that in the group of patients without soft tissue and bone involvement (P = 0.40).

Download figure to PowerPoint

As expected, somatic mutations were found in all sequences. The mutation rate varied between 1% and 14.6%, with a median of 5.1% per sequence. The overall frequency of somatic mutations was similar in sequences obtained from patients with soft tissue and bone involvement (median 5.3%; range 2.1–13.3%) and from those without soft tissue and bone involvement (median 5.1%; range 1–14.6%) (P = 0.13). In patients with dominant soft tissue and bone involvement, the median rate of mutations per sequence was 4.8% (range 3.8–13.3%).

Evidence of antigen selection was observed in the majority of sequences from both groups of patients. The frequency of antigen selection was comparable in patients with soft tissue and bone involvement and those without soft tissue and bone involvement (68.3% versus 58.7%; P = 0.23). In the group with dominant soft tissue and bone involvement, evidence of antigen selection was exhibited in 66.7% of the sequences.

Risk factors associated with soft tissue and bone involvement and dominant soft tissue and bone involvement.

Factors associated with the risk of soft tissue and bone involvement and dominant soft tissue and bone involvement were identified using logistic regression. Soft tissue and bone involvement was not associated with the VL family or subfamily in either unadjusted models (OR 1.01 [95% CI 0.53–1.93], P = 0.97 for the VL family and OR 0.95 [95% CI 0.48–1.88], P = 0.87 for the VL subfamily) or adjusted models (OR 0.80 [95% CI 0.40–1.56], P = 0.59 for the VL family and OR 0.79 [95% CI 0.38–1.62], P = 0.51 for the VL subfamily). However, in the adjusted model, male sex was a risk factor for soft tissue and bone involvement in VLκ (OR 2.05 [95% CI 1.09–3.85], P = 0.03) and VLκI (OR 2.04, [95% CI 1.09–3.82], P = 0.03) models.

Dominant soft tissue and bone involvement appeared to be significantly associated with VLκ (OR 3.12 [95% CI 1.16–8.37], P = 0.024) and VLκI (OR 3.16 [95% CI 1.16–8.58], P = 0.025) families in unadjusted models. However, after adjustment for possible confounders, the significance of this association was reduced (OR 2.57 [95% CI 0.92–7.20], P = 0.07 for VLκ; OR 2.80 [95% CI 0.99–7.88], P = 0.052 for VLκI). The major confounding variable appeared to be a history of multiple myeloma, which was a significant risk factor for dominant soft tissue and bone involvement in both the VLκ (OR 3.89 [95% CI 1.02–14.86], P = 0.046) and VLκI (OR 4.26 [95% CI 1.12–16.18], P = 0.03) models. Other variables such as age, sex, mutation rate, and evidence of antigen selection did not demonstrate a confounding effect on the presence or absence of dominant soft tissue and bone involvement.

Soft tissue and bone involvement in patients with AL amyloidosis and multiple myeloma.

Fifteen patients (7.9% of the cohort) presented with AL amyloidosis and multiple myeloma. The overall frequency of soft tissue and bone involvement was comparable in patients with multiple myeloma (60%) and patients without multiple myeloma (41.5%; n = 176) (P = 0.26). Evaluation of various types of soft tissue and bone involvement showed a significantly higher frequency of CTS in multiple myeloma patients than in patients without multiple myeloma (40% versus 10.8%; P = 0.006). The frequencies of other clinical features did not differ significantly between groups. However, macroglossia (33.3% versus 22.2%), submandibular gland enlargement (40% versus 31.3%), joint and bone involvement (13.3% versus 3.4%), and involvement of other soft tissue (6.7% versus 2.3%) were more frequent in patients with multiple myeloma. Cervical adenopathy and muscle pseudohypertrophy were seen only in patients without multiple myeloma.

Dominant soft tissue and bone involvement was the dominant clinical finding in 4 of the 15 patients with multiple myeloma (26.6%), compared with 14 of the 176 patients without multiple myeloma (8%) (P = 0.04). However, the difference between the 2 groups failed to reach significance after adjustment for multiple comparisons. The clinical features of dominant soft tissue and bone involvement included CTS in all 4 patients with multiple myeloma, submandibular gland enlargement and macroglossia each in 3 patients, and joint involvement in 2 patients. Other dominant clinical findings in patients with multiple myeloma included cardiac involvement in 5 patients (33.3%), renal involvement in 3 patients (20%), and gastrointestinal, neuropathic, and pulmonary involvement, each in 1 patient (6.7%). No statistically significant difference was observed between patients with multiple myeloma and patients without multiple myeloma with regard to these dominant organ categories.

In the group of 4 patients with multiple myeloma and dominant soft tissue and bone involvement, VLκ was found in 3 (75%; all κI) and VLλ in 1 (25%; λII). In the 11 multiple myeloma patients with other dominant organ involvement, VLκ was found in 4 patients (36.4%; 2 κI, 1 κIII, and 1 κIV) and VLλ in 7 (63.6%; 1 λI, 2 λII, 3 λIII, and 1 λVI). The frequencies of VLκ and VLλ did not differ significantly between the 2 groups (P = 0.28).

Survival of patients with dominant soft tissue and bone involvement.

The mean followup time in the group with dominant soft tissue and bone involvement was 49 months, compared with 33 months in the group without dominant soft tissue and bone involvement. The median survival time in the entire group was 35 months. The median survival time among patients with dominant soft tissue and bone involvement was 70 months, compared with 30 months among patients without dominant soft tissue and bone involvement. The 95% CI for the difference in median survival ranged from −14.7 to 80, indicating better survival for patients with dominant soft tissue and bone involvement. However, the difference between the 2 groups did not reach statistical significance (P = 0.21), most likely due to the fact that few deaths were observed in the dominant soft tissue and bone involvement group.

A positive effect of dominant soft tissue and bone involvement on early survival was confirmed when Kaplan-Meier survival curves were plotted separately for patients with and those without dominant soft tissue and bone involvement (Figure 3). The 2 survival curves were found to be statistically significantly different by Wilcoxon test (P = 0.035), although no difference was observed when the log rank test was used (P = 0.120), and, as shown in Figure 3, the survival curves converged over time.

thumbnail image

Figure 3. Kaplan-Meier plots of the probability of survival among AL amyloidosis patients with and without dominant soft tissue and bone involvement (DSTBI). The vertical line shows a cut point at 52 months, after which the difference in survival time was no longer statistically significantly different.

Download figure to PowerPoint

To analyze this further, we sought a cut point, t0, after which survival was no longer statistically significantly different. Cut points between 52 and 54 months were found to provide the best fit. With a cut point of 52 months, the early effect of dominant soft tissue and bone involvement on survival was significant when either VLκ or VLκI was considered a secondary predictor (P = 0.015 and P = 0.012, respectively, with estimated hazard ratios [HRs] of 0.33 and 0.31). After 52 months, the beneficial effect of dominant soft tissue and bone involvement was no longer significant for either VLκ or VLκI as a secondary predictor (P = 0.09 and P = 0.12, respectively, with estimated HRs of 3.47 and 3.24).

Other factors associated with the increased risk of death with both secondary predictors (VLκ and VLκI) were male sex (P = 0.012 and P = 0.022, respectively, with estimated HRs of 1.71 and 1.63) and history of multiple myeloma (P = 0.018 and P = 0.024, respectively, with estimated HRs of 2.22 and 2.07). Factors that did not make an independent contribution to survival were age, light-chain family, mutation rate, and evidence of antigen selection.

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Soft tissue is 1 of 7 sites of organ involvement in AL amyloidosis that have been identified by international consensus. Soft tissue involvement has been described as including macroglossia, submandibular swelling, amyloid lymphadenopathy, vascular amyloid manifested by claudication or bruising, muscle involvement, and painful periarticular amyloid deposition (13). In the present study, we focused on those features that might bring a patient with AL amyloidosis to the attention of a rheumatologist. For this reason, bone lesions were also included in the analysis.

In our group of 191 patients with AL amyloidosis, 82 (42.9%) had evidence of soft tissue and bone involvement. Soft tissue and bone involvement tended to occur in the setting of multiple organ involvement; the median number of involved organ systems was 4, compared with 2 in the group of patients without soft tissue and bone involvement. Such involvement consisted of submandibular gland enlargement in almost one-third of the patients and macroglossia in almost one-quarter of the group. CTS, common in patients with rheumatic diseases (22, 23), occurred in 13.1% of the patients with AL amyloidosis.

The frequency of amyloid arthropathy, which can mimic seronegative rheumatoid arthritis (9), was relatively low in our series (3.7%), as was amyloid deposition in unusual subcutaneous sites including breast, eyelid, lower lip, and skin nodules (2.6%). Muscle involvement was even rarer, with a frequency of 1.6%, comparable with the frequency reported in other studies (24, 25). Peripheral lymphadenopathy due to amyloid deposition was observed in 1% of patients. Only 6 case reports of peripheral lymphadenopathy in AL amyloidosis have been published (26–31). Bony involvement and vertebral collapse due to amyloid infiltration is an unusual manifestation of AL amyloidosis and should be distinguished from the lytic bone lesions seen in multiple myeloma (24). In the present study, we observed this in 1% of the patients with AL amyloidosis without evidence of multiple myeloma.

The incidence of soft tissue and bone involvement identified in this study was higher than that shown in our previous studies (32, 33), primarily because those studies did not include submandibular gland involvement in the soft tissue and bone involvement category. In this analysis, submandibular gland enlargement and CTS occurred with equal frequency in patients with dominant soft tissue and bone involvement and patients with dominant involvement of other organs. Thus, these manifestations were not exclusively part of the rheumatic syndrome. Dominant soft tissue and bone involvement occurred in 9.4% of patients, similar to the frequencies found in 2 other cohort series (7.3% and 9%) (34, 35).

The majority of soft tissue, joint, and bone manifestations of AL amyloidosis are nonspecific and may frequently be attributed to various rheumatic diseases. Similar to previously published case reports, soft tissue and bone involvement in our group of patients with AL amyloidosis was misdiagnosed as seronegative rheumatoid arthritis (9, 10), scleroderma (8), Sjögren's syndrome (11, 12), and polymyalgia rheumatica (36). Although AL amyloidosis is a rare condition, it should be considered in the differential diagnosis of atypical symptoms in patients believed to have rheumatic diseases.

There was no significant difference in light-chain families or subfamilies associated with soft tissue and bone involvement. In contrast, dominant soft tissue and bone involvement was found to be associated with VLκ and VLκI families in unadjusted models; however, the significance of this association was reduced after adjusting for possible confounders. A trend toward an association of VLκI gene usage and primary soft tissue disease has previously been reported (37). In another study, VLκIII protein was isolated from the shoulder tissue of a patient with dominant soft tissue and bone involvement (14). None of the VL sequences in our group of patients with dominant soft tissue and bone involvement and articular involvement were of the VLκIII type (Table 3), and the predicted protein sequences did not encode the amino acids at positions 32, 34, 59, and 95 that were present in the VLκIII protein.

In this study, the major confounding variable was a history of multiple myeloma, which was found to be a significant risk factor for dominant soft tissue and bone involvement in both the VLκ and VLκI models. The risk of having dominant soft tissue and bone involvement was 3.9–4.3 times higher in patients with AL amyloidosis with multiple myeloma than in patients with AL amyloidosis without multiple myeloma. Interestingly, the risk of developing soft tissue and bone involvement was not associated with the presence of multiple myeloma. AL amyloidosis may occur in 7–15% of patients with multiple myeloma (38), and a variety of soft tissue and bone manifestations in patients with AL amyloidosis with multiple myeloma have previously been reported. While CTS (24), macroglossia (24, 39, 40), arthropathy (41–43), and myopathy (25, 44, 45) were common soft tissue findings, lymphadenopathy (24, 46), submandibular gland enlargement (40), and amyloid bone lesions (47, 48) were rarely reported in multiple myeloma–associated AL amyloidosis.

In the analysis of soft tissue and bone involvement, the major confounding variable was male sex in both the VLκ and VLκI models. In our cohort, the risk of developing soft tissue and bone involvement was 2-fold higher in men than in women. Male predominance has been reported for various amyloid musculoskeletal and soft tissue manifestations, such as arthropathy (41, 49), myopathy (25, 44, 45), bone lesions (50), and lymphadenopathy (31).

Dominant soft tissue and bone involvement has been thought to be a relatively indolent presentation of AL amyloidosis (14). Our survival data did demonstrate an improved median survival time and higher rate of early survival among patients with dominant soft tissue and bone involvement compared with those with other dominant organ involvement. However, over time, the Kaplan-Meier survival curve for patients with dominant soft tissue and bone involvement converged with the curve for patients with other dominant organ involvement. This result suggests that, eventually, patients with dominant soft tissue and bone involvement either developed significant morbidity and mortality from their soft tissue and bone involvement or developed life-threatening involvement of other organ systems. Indeed, both of these outcomes were observed in patients in the present study. Two other variables associated with an increased risk of death in this analysis were male sex and history of multiple myeloma, consistent with the findings of a previous study (37).

In conclusion, AL amyloidosis can present with symptoms mimicking a variety of rheumatic syndromes. Patients with symptoms or signs of an infiltrative process involving soft tissue, joints, periarticular structures, or bones, without a clear rheumatic disease diagnosis, should be screened for AL amyloidosis, so that effective treatment can be instituted in a timely manner.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

Dr. Prokaeva had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. Prokaeva, Seldin.

Acquisition of data. Prokaeva, Spencer, Kaut, Skinner, Seldin.

Analysis and interpretation of data. Prokaeva, Connors, Skinner, Seldin.

Manuscript preparation. Prokaeva, Seldin.

Statistical analysis. Prokaeva, Ozonoff, Doros, Seldin.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES

The authors would like to acknowledge Jennifer Ellis, PhD, Elizabeth Fingar, BA, and Patrick Smith, BA for their assistance in light chain cloning and Kip Bodi, MS for upgrading the ALBase of amyloidogenic light chains (http://pulm.bumc.bu.edu/aldb/home).

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. AUTHOR CONTRIBUTIONS
  7. Acknowledgements
  8. REFERENCES