Improved survival among older acute myeloid leukemia patients – a population-based study

Abstract

Background. Survival in acute myeloid leukemia (AML) has improved in younger patients over the last decade. This study was conducted to evaluate the relative survival rates in older AML patients over two decades in the US.

Material and methods. We analyzed Surveillance, Epidemiology, and End Results (SEER) registry database to evaluate relative survival rate in older (≥ 75 years) AML population diagnosed during 1992–2009. We selected AML patients from 13 registries of SEER 18 database to compare RS during 1992–2000 and 2001–2009.

Results. The relative survival rates improved significantly during 2001–2009 compared to 1992–2000 for all age groups and sex. For young elderly patients (75–84 years) RS increased from 13.1 ± 0.8% to 17.4 ± 0.9% at one year Z-value = 3.98, p < 0.0001 and from 2.0 ± 0.4 to 2.6 ± 0.5%, Z-value = 3.61, p < 0.0005 at five years. Similarly, for very elderly (≥ 85 years) patients RS increased from 5.3 ± 1.0% to 8.0 ± 1.0%, Z-value = 3.03, p < 0.005 at one year, but no improvement seen at five years.

Conclusion. The relative survival in elderly AML has increased significantly during 2001–2009 compared to 1992–2000.

Acute myeloid leukemia (AML) is a heterogeneous hematologic stem cell malignancy in adults with incidence rate of 3–5% per 100 000 populations [1]. Incidence rate peaks in late 1970s to early 1980s, with around 23% of AML being diagnosed in this age group. Median age at the time of diagnosis is 65–69 years [2–5]. It is a very aggressive disease and is fatal without anti-leukemic treatment.

There have been significant advances in the understanding of the pathogenesis and prognostic factors of AML. Due to advances in treatment utilizing all-trans retinoic acid and arsenic trioxide, cure rate of acute promyelocytic leukemia, a rare subtype of AML has improved dramatically. For other patients of AML, treatment depends on prognostic factor assessment and age of patients. As per NCCN guidelines, treatment options for older patients include standard dose cytarabine with anthracycline, clofarabine, decitabine and five azacytidine [6]. None of these chemoregimens are highly effective. Previous population-based studies [7,8] have shown improvement in survival of AML patients during recent periods. Improvement in survival has been attributed to the advances in supportive care. Unfortunately, these studies did not show any improvement in survival of elderly patients. We conducted this study to analyze survival of older AML patients over last two decades.

Methods

The Surveillance, Epidemiology, and End Results (SEER) program from the National Cancer Institute is a population-based cancer registry that covers 26% of the US population. SEER program has high quality incidence and survival data since 1973. All SEER registries have an annual contractual completeness of at least 98% [9]. Periodic audits are done by regional registries as well as national cancer institute SEER. We analyzed 13 SEER registry to include patients from 1992 to 2009. These 13 SEER registries include: San Francisco-Oakland SMSA, Connecticut, Detroit (Metropolitan), Hawaii, Iowa, New Mexico, Seattle (Puget Sound), Utah, Atlanta (Metropolitan), San Jose-Monterey, Los Angeles, Alaska Natives and Rural Georgia.

Newly diagnosed, first primary only AML patients, aged ≥ 60 years were selected from the Surveillance, Epidemiology, and End Results (SEER) database. Ederer II was added as the default method for calculating cumulative expected survival and actuarial method was used to calculate survival using SEER*Stat Version 8.0.4 (released 15 April 2013). All patients were microscopically confirmed AML cases based on World Health Organization (WHO) classification system, including International Classification of Diseases for Oncology (3rd edition, ICD-O-3) histology codes in SEER data. AML cases which were actively followed, known malignant behavior, known age and cases in research database were included for analysis. In total 31 AML patients with no survival data were excluded. Patients with death certificate only or autopsy cases and alive with no survival time were also excluded from analysis. These cases were followed from date of diagnosis of AML to date of end of study December 2010 or death whichever came first.

There were 14 741 cases of AML who met initial inclusion criteria of age ≥ 60 years at the time of diagnosis, year of diagnosis 2001–2009, race and sex in the defined geographical area. We excluded 247 death certificate only or autopsy only cases. We also excluded 536 microscopically unconfirmed cases, 3358 cases who were not diagnosed as first primary AML and 12 living patients with no survival times.

The relative survival (RS) rates were analyzed for various cohorts categorized by race [White, African-Americans (AA) & other], gender & age (≥ 60, 60–74, 75–84 & ≥ 85 years) to determine survival differences from 1992–2000 and 2001–2009. Since the study cut-off was December 2010, all the AML patients diagnosed during 2001–2009 do not have three- and five-year survival data. Hence, AML (4182) diagnosed during January 2001–December 2007 were included in three-year RS analysis and AML (3024) patients diagnosed during January 2001–December 2005 were included in five-year RS analysis. We compared age-adjusted incidence rate of AML during 1992–2000 and 2001–2009.

RS rate measures the survival of the cancer patients in comparison to the general population of similar age and sex to estimate the effect of cancer. The RS rates accompany standard error (SE). We used SEER-Stat, statistical software provided by NCI for statistical analysis.

Results

There were 10 588 AML patients (age ≥ 60 years) during 1992–2009, 5816 (55%) were male and 4772 (45%) were female, 8887 were Caucasian, 675 were Black and 992 were other races. During 1992–2000 and 2001–2009 there were 5122 and 5466 patients, respectively. During 1992–2000, median age of patients at diagnosis of AML was 74 years (range 60–102 years). The quantile distribution by age at diagnosis was: 25% of patients were below 68 years, 50% below 74 years and 75% below 80 years of age. During 2001–2009 median age of AML patients upon diagnosis was 75 years (range 60–101years). Age distribution was: 25% of patients were below 68 years, 50% below 75 years and 75% below 81 years of age at the time of diagnosis of AML. A baseline characteristic is shown in Table I.

Table I. Baseline characteristics.

	1992–2000 (N)	2001–2009(N)
Age
60–74	2692	2645
75–84	1818	1998
85+	612	823
60+	5122	5466
Sex
Male	2824	2992
Female	2298	2474
Race Caucasian	4389	4498
African American	302	373
Other races	424	568
Median age at diagnosis	74 years (60–102)	75 years (60–101)
Median follow-up time	3 months (< 1–226)	4 months (< 1–226)

The RS rates of older AML patients (≥ 60 years) during 1992–2009 were 23.6 ± 0.4% at one year and 6.4 ± 0.3% at five years. The RS decreased from 33.7 ± 0.7% to 15.4 ± 0.6% to 6.9 ± 0.7% at one year and 10.7 ± 0.5% to 2.3 ± 0.3% to 0.5 ± 0.3% at five years for patients’ age 60–74 years, 75–84 years and ≥ 85 years, respectively. The median RS was six months, 2.6 months and 1.76 months for patients’ age 60–74 years, 75–84 years and ≥ 85 years, respectively.

The RS rates improved significantly during 2001–2009 compared to 1992–2000 for all age groups, sex, Caucasian and African Americans. For patients aged 60–74 years, RS rates increased from 29.3% to 38.3%, p < 0.0001 at one year; from 11.7% to 17.0%, p < 0.0001 at three years and from 7.8% to 12.7%, p < 0.0001 at five years. The median RS improved from 4.77 to 7.29 months.

For patients aged 75–84 years, RS increased from 13.1% to 17.4% at one year, p < 0.0001; from 3.6% to 4.6%, p < 0.001 at three years and 2.0 ± 0.4% to 2.7 ± 0.6%, p = 0.01 at five years with improvement in median RS from 2.41 to 2.75 months. Similarly, for patients aged ≥ 85 years, RS increased from 5.3% to 8.0%, p < 0.005 at one year; from 0.7% to 1.3% at three years, p = 0.004, but RS decreased to zero at five years.

There was significant improvement in RS from 1992–2000 to 2001–2009 in both sexes and all races examined (Table II).

Table II. One-, three- and five-year relative survival rates.

Cohort 1992–2000, n = 5122 2001–2009, n = 5466	1-year relative survival % (± SE)		Z-value/p-value	3-year relative survival % (± SE)		Z-value/p-value	5-year relative survival % (± SE)		Z-value/p-value
	1992–2000	2001–2009		1992–2000	2001–2007		1992–2000	2001–2005
Age (years)
≥ 60	20.8 (± 0.6)	26.2 (± 0.6)	6.93/< 0.0001	7.6 (± 0.4)	10.1 (± 0.5)	5.9/< 0.0001	5.0 (± 0.3)	7.3 (± 0.5)	4.73/< 0.0001
60–74	29.3 (± 0.9)	38.3 (± 1.0)	7.28/< 0.0001	11.7 (± 0.6)	17.0 (± 0.9)	6.5/< 0.0001	7.8 (± 0.6)	12.7 (± 0.9)	5.24/< 0.0001
75–84	13.1 (± 0.8)	17.4 (± 0.9)	3.98/< 0.0001	3.6 (± 0.5)	4.6 (± 0.6)	3.38/< 0.001	2.0 (± 0.4)	2.7 (± 0.6)	2.50/0.012
≥ 85	5.3 (± 1.0)	8.0 (± 1.0)	3.03/0.002	0.7 (± 0.4)	1.1 (± 0.5)	2.9/0.004	0.6 (± 0.4)	0.0 (± 0.0)	–
Sex
Male	20.6 (± 0.8)	27.2 (± 0.8)	6.25/< 0.0001	7.0 (± 0.5)	9.3 (± 0.6)	4.8/< 0.0001	4.0 (± 0.4)	6.4 (± 0.7)	4.26/< 0.0001
Female	21.0 (± 0.9)	25.1 (± 0.9)	3.42/< 0.001	8.4 (± 0.6)	11.1 (± 0.8)	3.5/0.0005	6.1 (± 0.5)	8.4 (± 0.8)	2.33/0.02
Race
Caucasian	20.8 (± 0.6)	25.8 (± 0.7)	5.80/< 0.0001	7.6 (± 0.4)	10.4 (± 0.6)	5.3/< 0.0001	5.1 (± 0.4)	7.3 (± 0.6)	3.88/0.0001
African American	18.9 (± 2.3)	27.8 (± 2.4)	2.21/0.027	5.2 (± 1.4)	9.2 (± 1.8)	2.4/0.016	2.5 (± 1.0)	8.0 (± 2.1)	2.74/0.006
Other	21.2 (± 2.0)	28.7 (± 1.9)	3.14/0.002	8.8 (± 1.4)	9.0 (± 1.5)	1.7/0.09	5.1 (± 1.1)	7.0 (± 1.6)	1.93/0.054

n, the number of individuals.

Discussion

AML is a fatal disease if left untreated. It is primarily a disease of the elderly. In the Swedish Acute Leukemia Registry, 75% of patients diagnosed with AML during 1997–2005 were aged 60 years or older [10]. Old age is a poor prognostic factor for complete remission, overall survival, remission duration and relapse-free survival [11]. Despite significant advances in understanding of pathogenesis and prognosis factors, prognosis of AML remains poor. Up to 70–80% of younger patients (< 60 years of age) achieve complete remission after treatment [12,13]. Unfortunately, they relapse over time and only 40–45% patients survive at five years. Among older patients (age > 60 years), 40–50% achieve complete remission, but cure rates are less than 10% and median survival is less one year. Many patients die from complications of treatment of leukemia [6].

Even though, there has not been significant change in chemotherapy in recent years, survival rates for young AML patients have improved. A population-based study by Derolf et al. [7] included 9729 AML patients diagnosed between 1973 and 2005. This study showed five- and 10-year survival rate improved in all age groups except for patients > 80 years of age. Another population-based study by Pulte et al. [8] showed improvement in five- and 10-year RS of AML patients from 1980–1984 and 2000–2004 for all patients except those aged over 75 years old. Highest improvements were seen for patients aged 15–34 with five- and 10-year RS around 52.3% and 47.9% during 2000–2004. Improvements in RS were also seen in the 35–54 and 55–64 age groups. There was no improvement in survival of aged 75 years or older. A recent population-based study by Thien et al. [14] evaluated survival of AML patients from nine SEER registries over three decades (1977–1986, 1987–1996, 1997–2006). There was no improvement in survival rates in patients aged ≥ 75 years. RS rates were lowest for those aged ≥ 85 years. In contrast, our study included AML patients from 13 SEER registries during 1992–2009. Hence, a direct comparison of our findings cannot be made.

Our study is the first study to report a modest but significant improvement in survival of elderly AML patients. This finding is encouraging but highlights the important point that survival in elderly patients remains poor. More research is needed to develop more effective and less toxic treatment regimens for this patient population. SEER database is the largest population base registry. There are some limitations of this study. This is a retrospective study. Since our study is based on SEER database, underreporting of AML is a possibility. SEER does not collect specific risk factors of malignancies.

In conclusion, survival rates of elderly AML patients improved significantly during 2001–2009 compared to 1992–2000.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

This paper was presented at the 55th ASH Annual Meeting and Exposition (December 7–10, 2013) in New Orleans, LA.