This study possesses distinct characteristics. 1) Case selection: we exclusively included cases of NDMM to mitigate potential influences of treatment and other factors on reticular fiber overgrowth. 2) Assessment of reticular fiber overgrowth (RFO): RFH in NDMM was evaluated using reticular fiber staining. Krzyzaniak et al. [12] initially classified myelofibrosis in 297 MM cases. Their findings indicated that BMF was suspected in H&E-stained sections in 48 (28.28%) of these cases and confirmed by reticulin fiber staining in 26 (8.8%) cases, which is lower than our observed rate. This discrepancy is primarily attributed to instances where RFH was inconspicuous in HE-stained sections and was not subsequently verified by reticular fiber staining. 3) RFH evaluation: the extent of RFH was assessed solely in hematopoietic areas, with the final grade determined by the most severe fibrosis area exceeding 30% [19, 20]. However, previous studies have not reported on this topic. Moreover, the grading of BMF remains an ongoing discussion. Currently, there are no unified diagnostic criteria for BMF in NDMM, resulting in variability in the reported incidence in prior literature [10,11,12,13,14]. Recently, Paul et al. [22] evaluated the BMF of 253 patients with NDMM. The results showed that 122 (48.2%) patients had detectable BMF, including MF-1, MF-2, and MF-3, while 131 (51.8%) had no BMF, namely MF-0. This finding is somewhat different from the results of our study. We suspect the reasons are as follows: First, fiber density was assessed only in hematopoietic areas, and the final score was determined by the highest-grade present in at least 30% of the marrow area in our research. Second, in our study, patients with stages I, II and III accounted for 17.9%, 24.1%, and 57.9%, respectively, while in Paul et al. ‘s study, patients with stages I, II and III accounted for 31.2%, 30.6%, and 38.2%, respectively. The proportion of stage III patients in our study was significantly higher than that of stage I patients. Third, a difference existed between the observers, especially in the case of MF-1. Previous studies reported high inter-observer differences, with a kappa value of only 0.373 (95% CI 0.174–0.573) [19].
Currently, the semi-quantitative grading of RFH, categorized into MF-0, MF-1, MF-2, or MF-3 based on the extent of overgrowth, is widely acknowledged. A significant debate centers on whether the MF-1 group should be classified as BMF. The diagnostic criterion for primary BMF requires an MF grade of 2 or higher [20]. In our clinical practice, reticular fiber staining is routinely conducted on all BM biopsy specimens, and a mild increase in reticular fibers (MF-1) is frequently observed in non-neoplastic lesions. In our study, statistical analysis was conducted on 11 patients in the MF-0 group and 51 patients in the MF-1 group, revealing no statistically significant differences between the MF-0 and the MF-1 groups (data was shown in Supplement Table 1). Consequently, in our research, a grade of MF ≥ 2 was identified as indicative of BMF, leading to the categorization of MF-2 and MF-3 into the BMF+ group, whereas MF-0 and MF-1 were classified under the BMF- group.
This study further investigated the association between clinicopathological features and BMF in patients with NDMM. The results revealed a connection between BMF and a diffuse infiltration pattern, corroborating the findings of Bartl et al. [10] Notably, since reticulin fibrosis is often reversible, earlier studies have documented that the presence of tumor cells in BM diminishes following MM treatment, with a reduction in fibrosis observed in 71% of cases [23]. This also supports the argument that BMF is related to the infiltration of tumor cells. Furthermore, our study revealed that characteristics indicative of a high tumor burden, such as a notable increase in plasma cell load and advanced D-S stage, were significantly more prevalent in the BMF+ group than in the BMF– group. These factors are recognized prognostic indicators for survival in MM. In a related study, Subramanian et al. examined 44 patients with MM and found that 9 cases (20.5%) exhibited BMF, predominantly of the plasmablastic type (8 cases), a proportion significantly higher than that of other types [14]. Contrastingly, our study did not reach a similar conclusion. This was potentially attributable to the exclusive inclusion of patients with NDMM and the minimal incidence of the plasmablastic type (merely one case).
Combined extramedullary invasion or plasma cell leukemia are poor prognostic factors for NDMM. Koshiishi et al. studied 91 cases of NDMM and found that 5.49% (5/91) had extramedullary disease, and the patients with BMF tended to exhibit extramedullary disease [23]. In our study, 13.0% (19/146) of the patients with MM presented with extramedullary invasion, and 5.5% (8/146) were diagnosed with plasma cell leukemia. This discrepancy was not observed in our study, and two potential reasons may account for this: 1) the incidence of extramedullary disease in our study was 13.0%, which is higher than that reported in the study by Koshiishi et al., and 2) in our study, a diagnosis of BMF was made with an MF ≥ 2, whereas Koshiishi et al. classified BMF at a threshold of MF ≥ 1.
The relationship between BMF and the prognosis of patients with MM remains unclear. Previous case series studies have indicated that patients with MM and BMF have poorer survival [14, 24, 25]. Sailer et al. stated that patients with BMF had a survival time of only 18 months [26]. Barry Paul et al. conducted reticular fiber staining in 393 patients with NDMM and conducted a follow-up, with a median duration of 83 months (range: 3.9–212 months). The findings revealed that the presence of BMF correlated with reduced OS and PFS [22]. However, our study did not provide conclusive evidence. The ISS staging, in conjunction with β2-MG and albumin levels, was employed to assess the patient’s prognosis. Our study found no marked differences between BMF and ISS staging. Our results indicated that patients with BMF had a shorter survival than those without BMF, albeit the differences were not statistically significant. Moreover, further univariate and multivariate survival analyses revealed that MF did not impact OS. A probable explanation is that the patients in our study received a standardized regimen comprising IMiD, PIs, and ASCT, contributing to their prolonged survival in comparison to the previously mentioned studies (45.4 months in the BMF– group, 39.2 months in the BMF+ group). Megumi Koshiishi [23] and colleagues analyzed 91 cases of NDMM treated with PIs and IMiDs. Their findings indicated that BMF did not significantly affect treatment response, OS, or PFS. Therefore, it can be inferred that the occurrence of BMF may not influence the prognosis of patients receiving modern treatment. Besides MM, BMF characteristics associated with other neoplastic diseases have been reported. An increase in reticulin fibers in BM has been observed in various hematological malignancies, including MPN, myelodysplastic syndrome (MDS), and acute myeloid leukemia [27, 28]. Notably, in chronic myelogenous leukemia (CML), myelofibrosis has been identified as a significant predictor of therapeutic efficacy and patient outcomes [29]. In MDS, patients with BMF exhibit significantly worse survival compared with patients with no BMF [30].
The significance of reticular fiber staining in diagnosing parathyroid tumors, adrenocortical adenomas, and pituitary adenomas is well-established [31, 32]. Reticular fibers, challenging to discern under H&E staining, become readily observable post-immersion in a silver ammonia solution and subsequent reduction with formaldehyde, yielding a black color. This method simplifies the examination of the characteristics of reticular fibers using an optical microscope. Characterized by its speed, cost-effectiveness, stability, and straightforward interpretation, reticular fiber staining effectively reveals variations in quantity. All specimens in this study were decalcified using a strongly acidic solution. The decalcified specimens were effectively stained with reticular fibers, demonstrating consistency and underscoring the maturity and stability of this widely used technique in clinical diagnosis.
The BM microenvironment is complex and comprises extracellular matrix proteins, cytokines, bone marrow stromal cells, mesenchymal stem cells, osteoblasts, osteoclasts, inflammatory cells, megakaryocytes, and microvessels [33]. Notably, substantial evidence now supports the notion that crosstalk and interactions among these components are crucial for myeloma cell proliferation and disease progression [25]. Plasma cells and BM stromal cells are known to secrete high levels of transforming growth factor-β1, and megakaryocytes produce cytokines that mediate tumor plasma cell proliferation. These factors can stimulate myofibroblasts to synthesize collagen, promote the synthesis of BM extracellular matrix proteins, and inhibit the degradation of extracellular matrix components, eventually leading to the occurrence of BMF [34, 35].
This study has some limitations. First, this study was a retrospective study, and the clinical follow-up time was not long enough, so there are inevitably some unnoticed bias factors and confounding factors. Second, the inherent heterogeneity of myeloma may have impacted the results. Third, sampling variability was a concern due to the focal distribution of the disease. Fourth, the specimens were treated with strong acid decalcification, precluding further next-generation sequencing and FISH detection. Finally, the treatment regimens varied among patients, potentially influencing prognosis outcomes. Therefore, future studies with larger samples and rigorous methodologies are essential to provide more comprehensive information and reveal prognostic effects.
Add Comment