Clinical Profile and Treatment of Multiple Myeloma at a Tertiary Hospital in Kenya: A Five-Year Retrospective Review

Background Multiple myeloma (MM) is a chronic B-cell malignancy that involves proliferation of neoplastic clonal plasma cells in the bone marrow with circulating monoclonal immunoglobulins or constituent chains in serum or urine or both. It is a rare cancer with a lifetime risk of 0.76% and an age-adjusted incidence rate of 2.5–7.2 per 100,000 in high-income countries. There is a paucity of local data on the morbidity and treatment of MM. Methods This was a single-centre descriptive retrospective study at the Kenyatta National Hospital (KNH). The study population included inpatients and outpatients with a documented diagnosis of MM managed between 1st January 2014 and 31st December 2018. Demographic data, pathology reports, laboratory results, and clinical findings were transcribed and uploaded to a database, and data analysis was done using Stata 16® software. Results A total of 207 patient files were reviewed. The median age at presentation was 60 years with a slight male preponderance. Bone pain was the predominant complaint in 59% (139/207) of patients, with 17% of patients presenting with paraparesis or paraplegia. For patients who underwent imaging, osteolytic bone lesions were identified in 90.6% (126/139). Anaemia was present in 71% (147/207) patients, hypercalcemia in 55.4%, and renal dysfunction in 38.2%. There were 25 different treatment regimens prescribed, with 13 patients (7%) being on bortezomib-based triplet therapy. Conclusions MM in KNH is a disease of the middle aged, affecting men and women almost equally and presenting mainly with bone pain and anaemia. Although there seems to be a general improvement in diagnosis and care, access to novel and less toxic agents for treatment is still wanting.


Introduction
Multiple myeloma (MM) is a chronic B-cell malignancy that involves proliferation of neoplastic clonal plasma cells in the bone marrow with subsequent overproduction of monoclonal immunoglobulins or its constituent polypeptide chains (paraproteins) in serum and urine, which lead to characteristic end-organ damage including renal dysfunction, anaemia, extensive lytic lesions, and hypercalcemia.
MM is a rare cancer, with a lifetime risk of 0.8% in the United States [1] and an age-adjusted incidence rate of 2.5-7.2 per 100,000 in Western countries [1,2].Te reported incidence rate has been low in sub-Saharan Africa but is on the rise following improved diagnostic capabilities and increased life expectancy [3].Tere are no specifc populationlevel data available for MM in the Kenya cancer registry [4] and other East African registries [5].However, there are a number of hospital-based studies that have looked at the incidence and clinical characteristics of MM in Kenya [6][7][8].Following the updated diagnostic, pathological, and clinical criteria of MM internationally in 2014 and advances in new therapies including proteasome inhibitors and immunomodulators, there has been an increase in the overall survival of MM patients [9,10].
Tis study was designed to document the clinical profle and treatment of multiple myeloma at the Kenyatta National Hospital to shed light on the current status of MM in Kenya.

Materials and Methods
Tis was a retrospective study conducted at the Kenyatta National Hospital (KNH), the largest referral hospital in Kenya with a bed capacity of 1800, receiving referrals from county and subcounty hospitals from around the country.KNH has dedicated outpatient oncology clinics, a radiotherapy cancer treatment centre (CTC), and inpatient oncology wards managing the bulk of cancer patients in Kenya.On average, a total of 500 to 600 cancer patients are seen every week.A digital fling system based on the ICD-10 classifcation allowed for retrieval of all fle numbers corresponding to the ICD-10 C-90 diagnosis for MM in the hospital records.
Consent to use the patient records in the fle was sought from the Kenyatta National Hospital Ethics Review Committee and was granted.Any identifcation data were removed, and patients were identifed only by assigned study numbers.
Te study population included patients with multiple myeloma classifed under ICD-10 diagnosis C-90 (multiple myeloma excluding solitary plasmacytomas) managed at KNH between 1 st January 2014 and 31 st December 2018.All eligible cases were included.Data extraction was done, and study variables of interest were recorded into an online data extraction form and uploaded onto the Microsoft Excel database.
Categorical variables, e.g., sex, stage of disease, exposure to novel agent during treatment, and presence of supportive treatment modality, were reported as frequencies with percentages.Continuous data variables, e.g., age, were expressed as means and standard deviations if normally distributed or median and interquartile range if skewed.A multivariable Cox regression model was used to test the association between presence of anaemia, renal dysfunction, and hypercalcemia as independent variables and time to outcome (death).Censoring was done at date of the lastknown follow-up for those in whom mortality was not witnessed.In building the model, stepwise backward and forward elimination at the 10% level of signifcance was used to select variables to be retained in the fnal model.Te fnal model was ftted to determine the signifcant variables associated with mortality and reported as hazard rates.Signifcance was set at the 5% level of signifcance for a twosided test.After model ftting, a test of the proportional hazards assumption was performed and it revealed no evidence of violation of this assumption.
Te data were exported to the Microsoft Excel package and subsequently exported into a study STATA ® fle.Ex- ploratory data analysis was done to identify missing values and check the skewness and normality of the data as well as check for signifcant associations.

Results
A total of 384 fle numbers were retrieved from the registries under the ICD classifcation of C-90 for multiple myeloma (Figure 1).Tere were 58 duplicate fles (one patient with two fle numbers) and 41 fles missing from the registries.Four fles of solitary plasmacytoma were excluded.In total, 207 fles that met the study case defnition of MM were retrieved and included in the fnal analysis.
3.1.Sociodemographic Characteristics.Patient baseline characteristics are presented in Table 1.Te mean age was 58.5 years (SD 11.8), while the median age was 60 years (interquartile range, IQR: 50-66).Te majority of patients (59%) were aged between 51 and 70 years, and only 6.5% were under 40 years of age.Tere was a slight male preponderance (n � 113, 54.6%) with a male to female ratio of 1.2 :1.

Clinical Presentation.
Te prevalence of myeloma defning events including anaemia, renal dysfunction, hypercalcemia, and bone lysis was documented.In line with the IMWG diagnostic criteria, anaemia in this study was broadly defned as haemoglobin of less than 10 g/dl, which is determined as 2 g/dl (20 g/l) lower than the lower limit in the normal population of 12 g/dl as per the WHO criteria used in most epidemiological studies [11,12].Of the 147 patients who had anaemia, 88 (59.9%) had moderate anaemia (Table 2).Only 19 (8%) patients presented with symptoms documented by the clinician as suggestive of anaemia at diagnosis while 5 (2%) had bleeding tendencies.
Renal dysfunction was assessed based on laboratory fndings, defned by the IMWG as serum creatinine >177 μmol/l, and was found in 79/207 (38.2%) patients at diagnosis.Only 7 (3%) came in with overt fuid overload which may indicate renal failure.Other symptoms of renal dysfunction were not documented as a major presenting complaint at the point of diagnosis.Hypercalcemia was present in 127 (55.4%) of patients at diagnosis (n � 178).
Bone pain was the most common presenting complaint in 135 out of 207 patients (59%).Isolated lower back pain, without any other major complaint, was present in 76 (33%) patients, while lower back pain with paralysis and paresthesia of the lower limbs were recorded in 40 (17%) patients.Evidence of osteolytic bone lesions and/or compression fractures was seen in 126 patients (91.9%) out of 137 patients that had documented imaging.Magnetic resonance imaging (MRI) was the most common modality, done in 69 (50.4%) patients, 39 (28.5%) had a computer tomography (CT scan), and 29 (21.1%) had conventional radiography done.

Diagnostics and Staging.
MM was diagnosed based on a bone marrow aspiration cytology report or tissue histology with one or more myeloma defning event (MDE).M-protein was detected in 114 (82%) of the 139 SPEP tests performed.Te mean M-protein component at diagnosis was 35.52 g/l (SD 30.6 g/l) with a median of 33.8 g/l (IQR 5 -53 g/l).Other diagnostic tests included urine Bence-Jones protein testing, done in ffty-eight patients (25.3%), of which 42 (73%) were positive.Serum free light chain testing was performed in 16 (7.7%)patients.Ten patients (62.5%) had predominantly free kappa light chains, and six patients (37.5%) had predominantly free lambda light chains.

Treatment.
Treatment was prescribed in 184 out of 207 patients (88.9%).For the other 23 patients (11.1%), the treatment prescribed was not on fle; some patients died before treatment was prescribed, some discharged before treatment was initiated, while some had been referred for radiotherapy only, and their chemotherapy had not been documented.
Te gold standard for MM treatment is induction therapy followed by autologous stem cell transplant (ASCT) for eligible patients.Tere were no patients who received ASCT in the study population.
Tere were 25 diferent combination treatment regimens for MM used within the study period (Table 3).Treatment regimens were varied; the majority were doublets and triplets based on medication available at the time (Table 3).

Outcome.
Tere were 77 (33%) recorded deaths, from any cause, at date of the last-known follow-up on record.However, the outcome, dead or alive, in 130 patients (67%) at date of the last-known follow-up could not be ascertained from the records, and due to patient confdentiality, the patients or next of kin could not be contacted to provide information on outcome.

Discussion
Our study population was relatively young for MM with a median age of 60 years consistent with more recent studies in Kenya and Nigeria [8,14].An earlier study in Kenya by Othieno-Abinya et al. found a median age of 53 years [7].Te diference may be attributed to an increasing life expectancy in Kenya over the years, from 50 years in the year 2000 to 66.3 years in 2018 [15].Tere was a slight male preponderance, consistent with regional haematological cancer data from Uganda and Nigeria [16,17].
Tere was a heavy burden of MM-defning events (MDEs) in the study population which may be attributed to late presentation and delayed diagnosis.Te presence of anaemia was high at 71%, similar to a study conducted in South Africa with an anaemia rate of 94%, also with a majority African population [18].Te normal range of the haemoglobin level in Kenya may actually be lower than that stated by WHO as documented in several studies [19,20] and may contribute to the high prevalence of anaemia found in this study.
Over 90% of patients had evidence of osteolytic bone lesions and/or compression fractures on imaging at diagnosis.A large multinational systematic review conducted by Mohty et al. found the presence of lytic lesions in 67.5%-71.5% of patients [21], whereas this study had 91% of patients with multiple osteolytic lesions at diagnosis.Being a tertiary facility, most patients had prior imaging at presentation; hence, clinicians immediately ordered an MRI on the frst review with the patients based on indication.Te high uptake of MRI is attributed to the National Health Insurance Fund (NHIF) fully covering imaging costs for patients, making MRI more accessible.
A third of patients had high levels of plasma cell infltration (>60%) which indicates severe disease as described by Kastritis et al. [22].Tis would support that majority of patients presented in late-stage disease, likely ISS stage 3, as implied by the few who had a full staging work-up (9%).Resource-based adaptation of the international consensus staging guidelines for MM should be considered to better capture the epidemiology of the disease in LMIC with limited laboratory capabilities.
Despite the gold standard for MM treatment being induction therapy followed by autologous stem cell transplant (ASCT) for eligible patients [23,24], no patients received ASCT in the study population.Currently, few African countries ofer ASCT such as Algeria, Egypt, Morocco, Nigeria, South Africa, and Tunisia [25].In 2023, one private facility in Nairobi, Kenya, embarked on ofering ASCT; however, the cost is still prohibitive for most patients.Te limited access to ASCT speaks to a large disparity in the quality of care ofered to MM patients in Kenya and other sub-Saharan African countries as compared to high-income countries, where ASCT has led to a four-year overall survival rate of 81% compared to 65% and progression-free survival rate of 43 months compared to 22 months in patients receiving chemotherapy alone with melphalan, prednisone, and lenalidomide [26].
Te 2019 Kenya national cancer treatment protocol recommends bortezomib-based regimens as frst line for both transplant eligible and noneligible patients [27], in line with similar guidelines in Europe and North America [23,24].However, only 7% (13) of the patients at KNH were on triplet bortezomib-based regimen.Te low uptake of bortezomib-based regimens may be related to the relatively prohibitive cost of bortezomib in the country, as it was not widely available and was not covered under the National Health Insurance Scheme then.Te access is improving with coverage under the National Health Insurance Scheme; therefore, higher numbers are expected.Many diferent treatment combinations mirror the fndings of a large prospective multinational noninterventional study on MM treatment carried out in Africa, Europe, and the Middle East by Mohty et al. that revealed great diversity in current treatment regimens used in MM [21].Tey attributed this to the evolution of treatment regimens, as well as variable access to the increasing number of treatments.Clinician's prescriptions were likely to have been infuenced by the availability of drugs locally and their afordability, as at times, patients had to pay out-of-pocket for what was not covered under the national scheme.Te efect of this may be better explored in a prospective qualitative study looking at clinicians' decisions and the afordability and availability of 4 Advances in Hematology MM medical treatments in our setting.Tis can be expanded to assess the real-world implementation of MM national treatment protocols and the impediments to their adaptation, a key aspect being their integration into what is covered under the National Health Insurance Scheme.
Te study collected real-world data on myeloma management in a tertiary hospital in a low-resource setting.It was able to capture records of 207 patients over a period of fve years, a relatively higher number than previous local studies.However, the study is not generalizable as it was a single-centre study.Another limitation was the reliance on obtaining data exclusively from the patient records.

Conclusion
Multiple myeloma at Kenyatta National Hospital is a disease of the middle aged, with a slight male preponderance and a high burden of disease at presentation.Staging of disease was inadequate with few patients able to do the full panel of tests.Treatment protocols were varied, and there is a need to improve the availability of newer agents so as to ofer standardized protocols.A nationwide recommended standard-of-care protocol with basic minimum investigations would help in improving care and mobilizing policy-makers to facilitate in making these tests and medications accessible and afordable.Advances in Hematology

Table 1 :
Sociodemographic characteristics of the MM patients at KNH.

Table 2 :
Prevalence of myeloma defning events among multiple myeloma patients at KNH.

Table 3 :
Regimens used in the treatment of multiple myeloma at KNH.