Incorporating PET imaging in a new risk stratification system for NDMM

Patients with newly diagnosed multiple myeloma (NDMM) are risk stratified according to the Revised International Staging System (R-ISS) designed by the International Myeloma Working Group (IMWG).¹ The R-ISS is based on parameters including the presence of beta-2 microglobulin, serum albumin levels, serum lactate dehydrogenase levels, and the presence of cytogenetic abnormalities. However, other high-risk factors such as the presence of multiple bone lesions and extramedullary disease have not been incorporated into the model. Recently, the IMWG recommend the use of imaging techniques, such as computed tomography (CT), magnetic resonance imaging, and ¹⁸F-fluorodeoxyglucose positron emission tomography/CT (¹⁸F-FDG PET/CT), to aid in the diagnosis of MM.¹

Cho et al. along with the Korean Multiple Myeloma Working Party,¹ recently designed a new risk stratification system that incorporates ¹⁸F-FDG PET/CT analysis with the R-ISS to effectively stratify patients with NDMM according to their survival outcomes. The development and validation of this new risk stratification model, the R-ISS/PET, was published in the Blood Cancer Journal and is summarized below. 

Study design and patient characteristics

• The initial cohort consisted of 380 patients with NDMM who had ¹⁸F-FDG PET/CT analysis upon diagnosis across ten hospitals of the Korean Multiple Myeloma Working Party from September 2009 to March 2020.
• Patient characteristics can be seen in Table 1.

Table 1. Patient characteristics*

Auto-HSCT, autologous hematopoietic stem cell transplantation; CA, cytogenetic abnormality; ECOG PS, Eastern Cooperative Oncology Group performance status; iFISH, interphase fluorescent in situ hybridization; LDH, lactate dehydrogenase; PET/CT, positron emission tomography/computed tomography; R-ISS, Revised International Staging System. *Adapted from Cho et al.1
Characteristic, % (unless otherwise stated)	n = 380
Median age (range), years	66 (34–86)
≥65 years	54.5
Male/Female	52.1/47.9
ECOG PS, %
0–1	77.4
2–3	20.0
Unknown	2.6
Increased LDH	27.9
Albumin ≥3.5 g/dL	61.6
Beta-2 microglobulin ≥ 5.5 mg/L	36.8
CA by iFISH
Standard risk	81.3
High risk	18.7
R-ISS
I	20.5
II	60.5
II	18.9
Extramedullary disease	13.4
No of focal lesions on PET/CT
≤3	47.6
>3	52.4
Frontline therapy
Proteasome inhibitors	80.8
Immunomodulatory agents	49.5
Auto-HSCT	34.5

• Regarding treatment response, 27.1% of patients had a complete response, 28.2% had a very good partial response, 29.7% had a partial response, 9.7% had stable disease, and 5.3% had progressive disease.
• The validation cohort consisted of 67 patients with NDMM from one hospital who had ¹⁸F-FDG PET/CT analysis upon diagnosis between June 2006 and February 2021. 
  • Baseline characteristics were similar to those in the original cohort.
• Median follow-up was 26 months (range, 0.1–153 months)

Results

Survival rates by focal lesions on ¹⁸F-FDG PET/CT evaluation or R-ISS

• Survival rates by the number of focal lesions on ¹⁸F-FDG PET/CT or R-ISS stage can be seen in Table 2.
  • The 2-year overall survival (OS) and 2-year progression-free survival (PFS) rates differed significantly between patients with ≤3 focal lesions and those with >3 (p = 0.094 and p < 0.001, respectively).
  • The 2-year OS and 2-year PFS rates differed significantly between the different R-ISS stages (both p < 0.001), with lower rates for higher stages.

Table 2. Survival rates by the number of focal lesions or R-ISS stage*

18F-FDG PET/CT, 18F-fluorodeoxyglucose positron emission tomography/computed tomography; FL, focal lesions; OS, overall survival; PFS, progression-free survival; R-ISS, Revised International Staging System. *Data from Cho et al.1
	2-year PFS rate, %	2-year OS rate, %
Number of FL on 18F-FDG PET/CT
≤3 FL	78.0	84.2
>3 FL	42.9	62.5
R-ISS stage
Stage I	71.8	95.3
Stage II	53.2	82.9
Stage III	26.8	61.2

Combining the number of focal lesions on ¹⁸F-FDG PET/CT with the R-ISS

A K-adaptive partitioning algorithm was undertaken to provide a statistically optimized combination of R-ISS with the number of focal lesions seen on ¹⁸F-FDG-PET/CT. Patients were defined into four groups, which successfully distinguished the patients with regard to 2-year PFS (p < 0.001) and 2-year OS (p < 0.001), as seen in Table 3. This was independent of transplant eligibility and treatment type.

Table 3. Survival rates by R-ISS/PET*

FL, focal lesions; OS, overall survival; PFS, progression-free survival; R-ISS/PET, Revised International Staging System with 18F-fluorodeoxyglucose positron emission tomography. *Data from Cho et al.1
R-ISS/PET stage	Combination	% of patients	2-year PFS	2-year OS
I	R-ISS I with FL ≤3	8.2	84.1	96.7
II	R-ISS I with FL >3 and R-ISS II with FL ≤3	41.1	64.7	89.8
III	R-ISS II with FL >3 and R-ISS III with FL ≤3	42.6	40.8	74.7
IV	R-ISS III with FL >3	8.2	17.1	50.3

The C-index values for OS and PFS were 0.668 (0.609–0.725) and 0.657 (0.615-0.698), respectively.

Multivariate Cox analysis (Table 4) also showed the R-ISS/PET was a significant factor in predicting long-term outcomes with regard to PFS and OS, with an increasing hazard ratio with each stage compared to Stage I. 

Table 4. Multivariate analysis of factors affecting PFS and OS with the R-ISS/PET*

CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; EMD, extramedullary disease; HR, hazard ratio; OS, overall survival; PFS, progression free survival; R-ISS/PET, Revised International Staging System with 18F-fluorodeoxyglucose positron emission tomography.  *Data from Cho et al.1
Factor	PFS		OS
	HR (95% CI)	p value	HR (95% CI)	p value
Age, ≥65 years vs <65 years	1.33 (0.98–1.80)	0.063	1.56 (0.99–2.45)	0.053
ECOG PS, 2–3 vs 0–1	1.01 (0.91–1.13)	0.848	1.14 (1.01–1.28)	0.034
EMD	1.22 (0.80–1.84)	0.353	1.68 (0.97–2.91)	0.062
R-ISS/PET stage
I	1.00		1.00
II	2.21 (1.00–4.90)	0.050	2.50 (0.59–10.7)	0.215
III	4.57 (2.09–10.0)	<0.001	5.11 (1.23–21.3)	0.035
IV	9.48 (3.88–12.2)	<0.001	10.3 (2.24–47.0)	0.003

Validation cohort

When assessed by the R-ISS stage, the external validation cohort showed similar 2-year PFS rates and 2-year OS rates to the initial cohort (Table 5). The classic stages showed a good degree of separation for survival outcomes (p = 0.268 for 2-year PFS and p = 0.037 for 2-year OS).

Table 5. Survival rates by R-ISS stage in the validation cohort*

OS, overall survival; PFS, progression-free survival; R-ISS, Revised International Staging System. *Data from Cho et al.1
R-ISS stage	2-year PFS rate, %	2-year OS rate, %
I	88.9	100
II	60.4	80.3
III	39.3	61.3

When the external validation cohort was assessed by the R-ISS/PET (Table 6), the stages also showed a significant separation for survival outcomes (p = 0.004 for 2-year PFS and p = 0.001 for 2-year OS).

Table 6. Survival rates by R-ISS/PET stage in the validation cohort*

OS, overall survival; PFS, progression-free survival; R-ISS/PET, Revised International Staging System with 18F-fluorodeoxyglucose positron emission tomography. *Data from Cho et al.1
R-ISS/PET stage	% of patients	2-year PFS, %	2-year OS, %
I	2.6	100	100
II	27.6	74.5	89.9
III	40.8	57.9	82.6
IV	17.1	25.6	42.0

Conclusion

The number of focal lesions assessed by ¹⁸F-FDG PET/CT was demonstrated to be a reliable imaging parameter for predicting survival outcomes and therefore a feasible factor to be incorporated with the R-ISS. This new risk stratification model, the R-ISS/PET, was validated by an external cohort and was able to define patients with NDMM into four risk groups according to survival differences. The R-ISS/PET was able to separate patients in the R-ISS Stage II group (which accounted for ~60% of the population) and allows for a more accurate prognosis than the R-ISS.

This study is limited by the retrospective nature of the analysis as there was no standardization of the interpretation of the imaging analysis. Therefore, the results may be subjected to inter-hospital variations between the ten centers where the data were collected. Furthermore, false negatives could be possible for patients with extensive bone involvement but low expression of hexokinase-2 (the enzyme involved in the glycolysis of FDG in malignant cells). Therefore, a prospective study using newer high-risk fluorescent in situ hybridization markers such as t(14;20), gain(1q21), and del(1p32) may be required to further validate the R-ISS/PET model.