Abstract
Introduction
A key dogma in the management of extensive-stage SCLC (ES-SCLC) is rapid diagnosis and treatment. However, the impact of treatment timeliness in a real-world setting is incompletely understood.
Methods
Patients with ES-SCLC were identified from the Victorian Lung Cancer Registry (VLCR). Referral-to-diagnosis, referral-to-treatment, and diagnosis-to-treatment intervals were calculated and defined as timely if they were less than 28 days, 42 days, and 14 days, respectively. The impact of timeliness on overall survival was determined using univariable and multivariable Cox proportional hazards regression.
Results
From January 2012 to April 2023, the VLCR enrolled 1195 patients with ES-SCLC, including 995 who were treated with chemotherapy. The median referral-to-diagnosis, referral-to-treatment, and diagnosis-to-treatment intervals were 8 days (interquartile range [IQR] 4–17 days), 18 days (IQR 10–31 days), and 7 days (IQR 4–13 days), respectively. Mortality risk was greater with a shorter diagnosis-to-treatment time (hazard ratio [HR], 1.31 for <14 days versus ≥14 days; 95% confidence interval [CI]: 1.12–1.54; p = 0.001), which was corroborated by multivariable regression (adjusted HR, 1.32; 95% CI: 1.13–1.56; p = 0.001). Poorer survival was also observed with a shorter referral-to-diagnosis interval (adjusted HR, 1.37 for <28 days versus ≥28 days; 95% CI: 1.10–1.71; p = 0.005) and shorter referral-to-treatment interval (adjusted HR, 1.58 for <42 days versus ≥42 days; 95% CI: 1.28–1.95; p <0.0001).
Conclusion
Most patients captured in the VLCR received timely care. Although earlier treatment predicted worse overall survival, the effects of timely care on symptom control and quality of life remain important considerations that require further characterization.
Keywords: Small cell lung cancer, Timeliness, Registry, Real-world data, Survival
Introduction
SCLC is an aggressive malignancy with a poor prognosis.1 Approximately 70% of individuals diagnosed with SCLC have extensive-stage SCLC (ES-SCLC) at the time of presentation.2 Treatment advances for ES-SCLC in the modern era have been limited, such that average overall survival is approximately 1 year after standard treatment using a combination of platinum-doublet chemotherapy and anti–programmed death ligand 1 immunotherapy.3 Because of its tendency for rapid growth and high likelihood of quick response to platinum chemotherapy,1 prompt commencement of systemic therapy is widely regarded to be a key principle in the management of ES-SCLC.4 However, despite this biologic rationale, observational studies have failed to show that a shorter time interval from diagnosis to chemotherapy commencement results in improved survival in ES-SCLC.5,6
Although tumor biology and treatment responsiveness are important factors providing impetus for expedited treatment, patient experience, and patient-centered outcomes, such as relief of symptoms, improvement in performance status, quality of life, and lessening of anxiety, are other major objectives driving early treatment.7 To this end, the Australian Optimal Care Pathway for lung cancer provides management guidelines and recommended time frames for the workup and treatment of patients with new presentations of lung cancer.8 These guidelines recommend completion of diagnostic tests within 2 weeks of the first specialist appointment and commencement of treatment within 6 weeks of the initial specialist referral. The Australian Optimal Care Pathway is consistent with the National Optimal Lung Cancer Pathway employed in the United Kingdom, which recommends a maximum referral-to-diagnosis interval of 28 days and a maximum referral-to-treatment interval of 49 days.9 In addition, it advises a 2-week target between diagnosis and commencement of systemic therapy for SCLC.
In view of the limited available evidence linking timeliness of care to patient outcomes in ES-SCLC, we conducted an analysis of the Victorian Lung Cancer Registry (VLCR) to gain further insights into this matter. Using registry data, we sought to determine the proportion of patients with ES-SCLC in Victoria, Australia, who received timely diagnosis and treatment, and to evaluate the effect of timely care on survival outcomes.
Materials and methods
Study Design and Population
The VLCR is a clinical quality registry designed to capture all new cases of lung cancer diagnosed in Victoria, Australia, and encompasses 51 Victorian hospitals from 19 health care networks.10 Patients with newly-diagnosed lung cancer receiving treatment at VLCR-participating institutions are identified from diagnosis coding (C34.0 to C34.9) according to the International Classification of Diseases (10th edition) system.11 Once identified, potential participants are notified and given 2 weeks to opt out of the registry.
Baseline characteristics recorded in the VLCR database include age, sex, country of birth, postcode, Eastern Cooperative Oncology Group (ECOG) performance status,12 Indigenous status, smoking history, and medical comorbidities. Recorded comorbid conditions include diabetes mellitus, renal insufficiency (defined as a serum creatinine level greater than 200 μmol/L), coronary artery disease (including history of myocardial infarction or receipt of coronary artery stenting, angioplasty or coronary artery bypass grafting), respiratory comorbidity (defined as documented chronic obstructive pulmonary disease or forced expiratory volume in 1 second of less than 66%) and neoplastic comorbidity (defined as a history of cancer other than lung cancer). Information regarding cancer diagnosis, stage (according to the TNM Classification for Lung Cancer13), treatment, and survival is also collected. Death dates are sourced from hospital site medical or administrative records and through linkage with The Registry of Births, Deaths and Marriages Victoria.
In this study, we evaluated all registered patients with ES-SCLC captured in the VLCR database from January 2012 to April 2023, referred to hereafter as the total population. From the total population, patients who received chemotherapy and had recorded diagnosis and treatment initiation dates were classified as the treatment timeliness population. For each patient in the treatment timeliness population, the following three timeliness intervals were measured: (1) referral-to-diagnosis, defined as the time from the initial referral to a lung cancer specialist service to the confirmation of a lung cancer diagnosis; (2) referral-to-treatment, defined as the time from the initial referral to the commencement of lung cancer treatment (of any modality); and (3) diagnosis-to-treatment, defined as the time from the diagnosis of lung cancer to the initiation of lung cancer treatment.
Statistical Methods
Descriptive data were summarized for the total population. Patients in the treatment timeliness population were further categorized into the following four groups according to the length of their diagnosis-to-treatment interval: less than 7 days, 7 to 13 days, 14 to 27 days, or greater than or equal 28 days. These groups were compared using chi-square or Fisher’s exact test for categorical variables and analysis of variance for continuous variables.
In the treatment timeliness population, Cox proportional hazards regression was used to analyze the effect of timeliness of care on overall survival, in which referral-to-diagnosis, referral-to-treatment, and diagnosis-to-treatment were explanatory variables. The referral-to-diagnosis, referral-to-treatment, and diagnosis-to-treatment intervals were defined as timely if they were less than 28 days, 42 days, and 14 days, respectively, on the basis of the targets outlined in the Australian Optimal Care Pathway for lung cancer.8 Overall survival was defined as the time from lung cancer diagnosis to death from any cause. Mortality risk was reported using hazard ratios (HRs) with 95% confidence intervals (CIs). The Kaplan–Meier product-limit method was used to plot survival as a function of time, and comparisons between survival curves were made using the log-rank test.
Multivariable Cox regression was used to assess the independent association between timeliness of care and overall survival, with adjustment for age, sex, ECOG performance status, comorbidities (diabetes mellitus, renal insufficiency, coronary artery disease, respiratory comorbidity and neoplastic comorbidity), use of immunotherapy, preferred first language (English or other), geographic remoteness and funding source of the treating hospital (private or public). Geographic remoteness was measured by the Modified Monash Model (MMM), which categorizes locations as MM 1 (metropolitan areas), MM 2 (regional centers), MM 3 (large rural towns), MM 4 (medium rural towns), MM 5 (small rural towns), MM 6 (remote communities) or MM 7 (very remote communities).14 Multivariable regression models for specific subgroup analyses (age <70 versus ≥70 y and immunotherapy treatment versus no immunotherapy treatment) were developed using these same covariates, except for the covariate that defined the subgroups of interest.
Sensitivity analyses using propensity score matching were performed to account for potential confounding. Propensity scores were generated using a multivariable logistic regression model with timely care as the dependent variable and the same patient characteristics described above as independent variables. The association between timely care and overall survival was then assessed using a Cox regression model with the propensity score included as an additional covariate. Landmark analysis was also performed, wherein patients who died within 30 days of diagnosis were excluded from the multivariable Cox regression models. This was done in view of the potential confounding effect of these patients on overall survival, recognizing that such patients may already be approaching end-of-life, with a trajectory that would unlikely be altered by treatment.
In the multivariable Cox models, missing data for ECOG performance status were handled by including “missing” as its own category of the ECOG variable. In addition, sensitivity analyses using best-case and worst-case scenarios were conducted, in which missing ECOG scores were assigned as 0 or 4, respectively. This was further complemented by complete case analyses restricted only to participants with available ECOG scores.
All calculated p values were two-sided, and p less than 0.05 indicated statistical significance. Analyses were performed with Stata version 18 (StataCorp, College Station, TX) or SAS version 9.4 (SAS Institute, Cary, NC).
Ethics
This research was approved by the Human Research Ethics Committees of Monash University (project number 47950) and the individual hospitals that participated in the VLCR.
Results
Population and Treatment
From January 2012 to April 2023, the VLCR database enrolled 1195 patients with ES-SCLC (Table 1). The median age was 69 years (interquartile range [IQR] 62–76 y), 59.9% (n = 716) were male, and 46.0% (n = 550) had an ECOG performance status of 0 or 1. Out of the total population, 1037 (86.8%) patients received anticancer treatment in the form of chemotherapy or radiotherapy, of whom 438 (42.2%) were treated with both chemotherapy and radiotherapy, 558 (53.8%) were treated with chemotherapy only, and 41 patients (4.0%) were treated with radiotherapy only. Altogether, 996 patients (83.3%) received chemotherapy, and 479 (40.1%) patients received radiotherapy.
Table 1.
Characteristics of the Study Population
| Characteristic | Total Population (n = 1195) | Treatment Timeliness Population, Stratified by Diagnosis-to-Treatment Interval (n = 995) |
|||||
|---|---|---|---|---|---|---|---|
| <7 days (n = 439) | 7–13 days (n = 314) | 14–27 days (n = 153) | ≥28 days (n = 89) | Total (n = 995) | P value | ||
| Age – median (IQR) | 69 (62–76) | 68 (61–74) | 69 (62–75) | 69 (61–75) | 68 (61–75) | 68 (61–74) | 0.38 |
| Sex – n (%) | 0.99 | ||||||
| Female | 479 (40.1) | 175 (39.9) | 126 (40.1) | 63 (41.2) | 36 (40.5) | 400 (40.2) | |
| Male | 716 (59.9) | 264 (60.1) | 188 (59.9) | 90 (58.8) | 53 (59.6) | 595 (59.8) | |
| ECOG score – n (%) | 0.67 | ||||||
| 0 | 173 (14.5) | 73 (16.6) | 58 (18.5) | 21 (13.7) | 17 (19.1) | 169 (17.0) | |
| 1 | 377 (31.6) | 139 (31.7) | 121 (38.5) | 58 (37.9) | 29 (32.6) | 347 (34.9) | |
| 2 | 165 (13.8) | 59 (13.4) | 42 (13.4) | 23 (15.0) | 10 (11.2) | 134 (13.5) | |
| 3 | 79 (6.6) | 20 (4.6) | 10 (3.2) | 7 (4.6) | 3 (3.4) | 40 (4.0) | |
| 4 | 11 (0.9) | 2 (0.5) | 0 (0) | 0 (0) | 0 (0) | 2 (0.2) | |
| Unknown | 390 (32.6) | 146 (33.3) | 83 (26.4) | 44 (28.8) | 30 (33.7) | 303 (30.5) | |
| Diabetes mellitus | 260 (21.8) | 87 (19.8) | 69 (22.0) | 38 (24.8) | 16 (18.0) | 210 (21.1) | 0.50 |
| Renal insufficiency | 31 (2.6) | 6 (1.4) | 6 (1.9) | 2 (1.3) | 1 (1.1) | 15 (1.5) | 0.96 |
| Coronary artery disease | 192 (16.1) | 61 (13.9) | 48 (15.3) | 34 (22.2) | 15 (16.9) | 158 (15.9) | 0.11 |
| Respiratory comorbidity | 265 (22.2) | 81 (18.5) | 71 (22.6) | 41 (26.8) | 19 (21.3) | 212 (21.3) | 0.16 |
| Neoplastic comorbidity | 198 (16.6) | 66 (15.0) | 50 (15.9) | 31 (20.3) | 15 (16.9) | 162 (16.3) | 0.51 |
| Country of birth – n (%) | 0.03 | ||||||
| Australia | 770 (64.4) | 290 (66.1) | 202 (64.3) | 95 (62.1) | 71 (79.8) | 658 (66.1) | |
| Other | 425 (35.6) | 149 (33.9) | 112 (35.7) | 58 (37.9) | 18 (20.2) | 337 (33.9) | |
| Preferred language – n (%) | 0.49 | ||||||
| English | 1094 (93.1) | 408 (92.9) | 289 (92.0) | 137 (89.5) | 84 (94.4) | 918 (92.3) | |
| Other | 101 (6.9) | 31 (7.1) | 25 (8.0) | 16 (10.5) | 5 (5.6) | 77 (7.7) | |
| Current smoker or ex-smoker – n (%) | 1139 (95.3) | 427 (97.3) | 294 (93.6) | 147 (96.1) | 83 (93.3) | 951 (95.6) | 0.07 |
| Indigenous – n (%) | 14 (1.2) | 6 (1.4) | 2 (0.6) | 3 (2.0) | 1 (1.1) | 12 (1.2) | 0.55 |
| Treating institution – n (%) | 0.51 | ||||||
| Private | 71 (5.9) | 35 (8.0) | 17 (5.4) | 13 (8.5) | 6 (6.7) | 71 (7.1) | |
| Public | 1124 (94.1) | 404 (92.0) | 297 (95.6) | 140 (91.5) | 83 (93.3) | 924 (92.9) | |
| MMM category | 0.05 | ||||||
| MM 1 (metropolitan) | 815 (68.2) | 307 (69.9) | 205 (65.3) | 108 (70.6) | 52 (58.3) | 672 (67.5) | |
| MM 2 (regional) | 87 (7.3) | 23 (5.2) | 33 (10.5) | 8 (5.2) | 6 (6.7) | 70 (7.0) | |
| MM 3-5 (rural) | 291 (24.4) | 109 (24.8) | 75 (23.9) | 37 (24.2) | 31 (34.8) | 252 (25.3) | |
| MM 6-7 (remote and very remote) | 2 (0.2) | 0 (0) | 1 (0.3) | 0 (0) | 0 (0) | 1 (0.1) | |
| Chemotherapy treatment – n (%) | 996 (83.4) | 439 (100) | 314 (100) | 153 (100) | 89 (100) | 995 (100) | - |
| Radiotherapy treatment – n (%) | 479 (40.1) | 188 (42.8) | 140 (44.6) | 73 (47.7) | 37 (41.6) | 438 (44.0) | 0.72 |
| First treatment administered – n (%) | 0.79 | ||||||
| Chemotherapy | 918 (76.8) | 404 (92.0) | 290 (92.4) | 141 (92.2) | 82 (92.1) | 917 (92.2) | |
| Radiotherapy | 102 (8.5) | 27 (6.2) | 19 (6.1) | 11 (7.2) | 4 (4.5) | 61 (6.1) | |
| Both | 17 (1.4) | 8 (1.8) | 5 (1.6) | 1 (0.7) | 3 (3.4) | 17 (1.7) | |
| Neither | 158 (13.2) | - | - | - | - | - | |
| Immunotherapy treatment – n (%) | 339 (28.4) | 151 (34.4) | 110 (35.0) | 53 (34.6) | 25 (28.1) | 339 (34.1) | 0.66 |
ECOG, Eastern Cooperative Oncology Group; ES-SCLC, extensive-stage SCLC; IQR, interquartile range; MMM, Modified Monash Model.
Of the 996 chemotherapy-treated patients, one did not have diagnosis and treatment commencement dates documented in the registry database. Therefore, the treatment timeliness population consisted of 995 patients. Apart from country of birth, all other patient characteristics were evenly distributed between the four diagnosis-to-treatment interval categories (Table 1, Supplementary Table 1). In the 438 patients who received both chemotherapy and radiotherapy, chemotherapy was administered first in 82.2% of cases (n = 360), radiotherapy was administered first in 13.9% of cases (n = 61), and chemotherapy and radiotherapy were initiated on the same day in 3.9% of patients (n = 17). Immunotherapy was administered to 34.1% of the treatment-timeliness population.
Timeliness of Diagnosis and Treatment
In the treatment timeliness population (n = 995), the median diagnosis-to-treatment interval was 7 days (IQR 4–13 days), corresponding to 75.7% (n = 753) commencing treatment within the 14-day benchmark (Table 2). The referral-to-diagnosis and referral-to-treatment intervals were measurable for 926 patients because of the lack of a definitive referral date in the remaining 69 patients. The median referral-to-diagnosis time was 8 days (IQR 4–17 days), with 87.4% (n = 809) meeting the target of diagnosis within 28 days of specialist referral. The median referral-to-treatment interval was 18 days (IQR 10–31 days), with 85.2% (n = 789) achieving the referral-to-treatment timeliness target of 42 days.
Table 2.
Timeliness of Diagnosis and Treatment Commencement in Treatment Timeliness Population
| Time Interval | Median (IQR) (d) | n (%) |
|||||
|---|---|---|---|---|---|---|---|
| <7 d | 7–13 d | 14–20 d | 21–27 d | 28–41 d | ≥42 d | ||
| Referral-to-diagnosis (n = 926) | 8 (4–17) | 377 (40.7)a | 249 (26.9)a | 117 (12.6)a | 66 (7.1)a | 70 (7.6) | 47 (5.1) |
| Referral-to-treatment (n = 926) | 18 (10–31) | 99 (10.7)a | 238 (25.7)a | 179 (19.3)a | 132 (14.3)a | 141 (15.2)a | 137 (14.8) |
| Diagnosis-to-treatment (n = 995) | 7 (4–13) | 439 (44.1)a | 314 (31.6)a | 93 (9.3) | 60 (6.0) | 58 (5.8) | 31 (3.1) |
IQR, interquartile range.
Fulfills definition for timely care.
Impact of Timing of Diagnosis and Treatment on Overall Survival
Diagnosis-to-Treatment Interval
The median overall survival was 8.0 months (95% CI: 7.6–8.5 mo) for patients with a diagnosis-to-treatment interval of less than 14 days, and 10.1 months (95% CI: 8.6–11.5 mo) for patients with a diagnosis-to-treatment interval of greater than or equal to 14 days. This corresponded to an unadjusted mortality HR of 1.31 (95% CI: 1.12–1.54; p = 0.001), favoring delayed treatment. Kaplan–Meier curves illustrating overall survival according to diagnosis-to-treatment interval are presented in Figure 1. When the diagnosis-to-treatment interval was analyzed as a continuous variable, the unadjusted HR for overall survival was 0.99 (95% CI: 0.99–1.00, p = 0.005), indicating a 1% reduction in mortality risk for every 1-day increase in the diagnosis-to-treatment interval.
Figure 1.
Kaplan–Meier curves for overall survival according to diagnosis-to-treatment interval.
The multivariable model also indicated an increase in mortality risk when the diagnosis-to-treatment interval was less than 14 days, with an adjusted HR of 1.32 (95% CI: 1.13–1.56; p = 0.001) (Table 3). In this model, additional variables associated with significantly poorer survival included older age (HR, 1.89 for age ≥80 y compared with age <50 y; 95% CI: 1.23–2.92; p = 0.004), poorer performance status (HR, 13.14 for ECOG 4 compared with ECOG 0; 95% CI: 3.20–54.00; p = 0.0004) and public hospital-based treatment (HR, 1.32 relative to private hospital–based treatment; 95% CI: 1.01–1.72; p = 0.04). Residence in rural or remote areas (HR, 0.82 for MM 3–7 compared with MM 1; 95% CI: 0.70–0.97; p = 0.02) and use of immunotherapy (HR, 0.78; 95% CI: 0.67–0.91; p = 0.001) were associated with improved survival. When patients who died within 30 days of diagnosis were omitted, a similar adjusted HR of 1.26 (95% CI: 1.07–1.49; p = 0.005) was observed. In the propensity matched analysis, the adjusted HR for overall survival was 1.30 (95% CI: 1.11–1.53; p = 0.001), indicating significantly poorer survival for individuals who started treatment within 14 days of diagnosis.
Table 3.
Multivariable Cox Proportional Hazards Model for Overall Survival According to Diagnosis-to-Treatment Interval
| Covariate | HR (95% CI) | p Value |
|---|---|---|
| Diagnosis-to-treatment interval | ||
| <14 days | 1.00 | - |
| ≥14 days | 1.32 (1.13–1.56) | 0.001 |
| Age | ||
| <50 y | 1.00 | - |
| 50–59 y | 1.14 (0.75–1.71) | 0.54 |
| 60–69 y | 1.26 (0.85–1.87) | 0.25 |
| 70–79 y | 1.52 (1.02–2.26) | 0.04 |
| ≥80 y | 1.89 (1.23–2.92) | 0.004 |
| Sex | ||
| Female | 1.00 | - |
| Male | 1.13 (0.98–1.30) | 0.09 |
| ECOG score | ||
| 0 | 1.00 | - |
| 1 | 1.20 (0.98–1.48) | 0.08 |
| 2 | 1.42 (1.10–1.83) | 0.007 |
| 3 | 3.08 (2.11–4.49) | <0.0001 |
| 4 | 13.14 (3.20–54.00) | 0.0004 |
| Missing | 1.51 (1.23–1.86) | 0.0001 |
| Diabetes mellitus | 1.05 (0.89–1.24) | 0.55 |
| Renal insufficiency | 1.47 (0.86–2.52) | 0.16 |
| Coronary artery disease | 0.95 (0.79–1.16) | 0.63 |
| Respiratory comorbidity | 0.88 (0.74–1.04) | 0.13 |
| Neoplastic comorbidity | 0.94 (0.78–1.14) | 0.54 |
| Preferred language | ||
| Language other than English | 1.00 | - |
| English | 1.20 (0.93–1.56) | 0.16 |
| Treatment institution | ||
| Private | 1.00 | - |
| Public | 1.32 (1.01–1.72) | 0.04 |
| MMM classification | ||
| MM 1 | 1.00 | - |
| MM 2 | 0.94 (0.72–1.22) | 0.63 |
| MM 3-7 | 0.82 (0.70–0.97) | 0.02 |
| Immunotherapy treatment | 0.78 (0.67–0.91) | 0.001 |
CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; MMM, Modified Monash Model.
Referral-to-Diagnosis Interval
The median overall survival was 8.2 months (95% CI: 7.8–9.0 mo) in patients who were diagnosed within 28 days of referral and 10.3 months (95% CI: 8.7–12.1 mo) in those who were diagnosed 28 days or more after referral. The unadjusted HR for overall survival did not differ significantly between these groups (unadjusted HR, 1.22; 95% CI: 0.99–1.52; p = 0.07) (Fig. 2). Similarly, the referral-to-diagnosis interval exhibited no association with survival when evaluated as a continuous variable (unadjusted HR, 1.00; 95% CI, 0.99–1.00; p = 0.18).
Figure 2.
Kaplan–Meier curves for overall survival according to referral-to-diagnosis interval.
In the multivariable analysis (Supplementary Table 2), the risk of death was 37% higher for those diagnosed within 28 days of referral (adjusted HR, 1.37; 95% CI: 1.10–1.71; p = 0.005). Older age (HR, 1.82 for age ≥80 y compared with age <50 y; 95% CI: 1.16–2.86; p = 0.009), poorer ECOG performance status (HR, 16.98 for ECOG 4 compared with ECOG 0; 95% CI: 4.10–70.40; p < 0.0001) and public hospital-based treatment (HR, 1.37 compared with private hospital–based treatment; 95% CI: 1.03–1.83; p = 0.03) were significantly associated with increased risk of mortality. Conversely, residence in rural or remote areas (HR, 0.79 for MM 3–7 relative to MM 1; 95% CI: 0.66–0.94; p = 0.009) and treatment with immunotherapy (HR, 0.78; 95% CI: 0.67–0.91; p = 0.002) were associated with improved overall survival. The landmark analysis yielded consistent results, with an adjusted HR of 1.31 (95% CI: 1.05–1.64; p = 0.02) when patients who experienced death within 30 days were excluded. The propensity score-adjusted HR for early diagnosis (<28 days from referral) relative to later diagnosis (≥28 days from referral) was 1.29 (95% CI: 1.03–1.60; p = 0.03).
Referral-to-Treatment Interval
Patients with a referral-to-treatment interval of shorter than 42 days had a median overall survival of 8.1 months (95% CI: 7.7–8.7 mo), whereas those with a referral-to-treatment interval of 42 days or longer had a median overall survival of 11.6 months (95% CI: 10.0–13.1 mo). Patients who began treatment within 42 days of referral had a higher risk of death (unadjusted HR, 1.48; 95% CI: 1.20–1.82; p = 0.0002) compared with those who started treatment later (Fig. 3). When evaluated as a continuous variable, every 1-day increase in the referral-to-treatment interval was associated with a 1% reduction in the risk of death (unadjusted HR, 0.99 for referral-to-treatment as a continuous variable; 95% CI: 0.99–1.00; p = 0.002).
Figure 3.
Kaplan–Meier curves for overall survival according to referral-to-treatment interval.
Using multivariable regression (Supplementary Table 3), the adjusted mortality HR was 1.58 (95% CI: 1.28–1.95; p <0.0001), indicating a greater risk of death with earlier treatment. In this model, older age (HR, 2.02 for age ≥80 y compared with age <50 y; 95% CI: 1.29–3.17; p = 0.002), poorer performance status (HR, 13.01 for ECOG 4 compared with ECOG 0; 95% CI: 3.16–53.54; p = 0.0004) and treatment at a public institution (HR, 1.38; 95% CI: 1.04–1.84; p = 0.03) placed patients at a higher risk of death. Rural or remote residence (HR, 0.79 for MM 3-7 compared with MM 1; 95% CI: 0.66–0.94; p = 0.007) and use of immunotherapy (HR, 0.78; 95% CI: 0.67–0.90; p = 0.001) were associated with lower mortality risk. Exclusion of patients who died within 30 days of diagnosis produced an adjusted HR of 1.51 (95% CI: 1.22–1.87; p = 0.0001). Propensity score matching also corroborated a significant increase in mortality hazard for those treated within 42 days of referral, with an adjusted HR of 1.54 (95% CI: 1.25–1.90; p = 0.0001).
Subgroup Analyses
ECOG Performance Status
ECOG scores were captured for 69.5% of participants (n = 692) in the treatment-timeliness population. There was no significant difference between patients with missing and nonmissing ECOG data in terms of age, sex, use of immunotherapy, or comorbidities (Supplementary Table 4).
In the subgroup of patients with available ECOG data, a diagnosis-to-treatment interval of less than 14 days was associated with worsened overall survival (adjusted HR, 1.39; 95% CI: 1.14–1.70; p = 0.001) (Supplementary Table 5), consistent with the findings from the primary treatment timeliness population. Poorer overall survival was also observed with a shorter referral-to-diagnosis interval (adjusted HR, 1.31 for <28 d versus ≥28 d; 95% CI: 1.00–1.71; p = 0.047) and shorter referral-to-treatment interval (adjusted HR, 1.62 for <42 d versus ≥42 d; 95% CI: 1.26–2.08; p = 0.0002). Analyses of both best-case (missing ECOG scores designated as 0) and worst-case scenarios (missing ECOG scores designated as 4) also revealed inferior overall survival in patients with shorter timeliness intervals (Supplementary Table 6), yielding adjusted HRs of 1.32 (95% CI: 1.13–1.56; p = 0.001) and 1.33 (95% CI: 1.33–1.56; p = 0.001), respectively, in patients with diagnosis-to-treatment intervals of less than 14 days.
Immunotherapy
In the subgroup of patients who received immunotherapy (n = 339), survival was generally more favorable with delayed treatment. Among immunotherapy recipients (Supplementary Table 7), the adjusted HR for overall survival was 1.73 (95% CI: 1.25–2.41; p = 0.001) for patients with a diagnosis-to-treatment interval of less than 14 days, relative to those whose treatment was initiated later. In immunotherapy-treated patients, a referral-to-treatment interval of less than 42 days coincided with inferior overall survival compared with a referral-to-treatment interval of greater than or equal 42 days (adjusted HR, 1.87; 95% CI: 1.22–2.86; p = 0.004), but the time from referral to diagnosis did not significantly impact mortality (adjusted HR, 1.04; 95% CI: 0.70–1.54; p = 0.85).
There was evidence of a significant interaction between timeliness and immunotherapy in the diagnosis-to-treatment model for overall survival (p = 0.03), but not in the referral-to-diagnosis (p = 0.63) or referral-to-treatment (p = 0.14) models.
Age
The adjusted HRs for overall survival for timely versus nontimely treatment, as quantified by diagnosis-to-treatment interval, were 1.25 (95% CI: 0.99–1.45; p = 0.06) in patients aged under 70 years and 1.39 (95% CI: 1.10–1.76; p = 0.006) in those aged 70 years or above. There was no evidence of any interaction effect (p = 0.45) between age (<70 y versus ≥70 y) and diagnosis-to-treatment interval (<14 d versus ≥14 d).
Discussion
Earlier commencement of anticancer treatment has been associated with longer survival in a range of cancer types,15, 16, 17 but this association has not been seen in ES-SCLC. Data from the VLCR revealed that most patients with ES-SCLC received care that conformed with timeliness targets set out in national guidelines; 85% of patients began treatment within 42 days of being referred for specialist review, and 76% began treatment within 14 days of diagnosis. However, contrary to expectation, survival was poorer when referral-to-diagnosis, referral-to-treatment, and diagnosis-to-treatment intervals were shorter. These findings were consistent across the different statistical methodologies and sensitivity analyses used in our study.
Our results are also consistent with the limited evidence in the existing literature pertaining to SCLC, timeliness of care, and survival.5,6,18,19 An analysis of data from the National Cancer Database in the United States found that among 41,116 patients with ES-SCLC, a shorter interval between diagnosis and chemotherapy initiation was associated with worse survival; relative to patients who initiated treatment within 7 days of diagnosis, overall survival HRs were 0.92 (p < 0.001) for a diagnosis-to-treatment interval of 8 to 14 days, 0.82 (p < 0.001) for 15 to 28 days and 0.77 (p < 0.001) for greater than 28 days.6 Similarly, a Polish study of 3479 patients with SCLC reported that individuals who commenced treatment in the first 42 days after their initial medical visit had worse survival (HR 1.20, p = 0.002) than those who commenced treatment later.18 In addition, a smaller, single-center study of 323 patients with SCLC, 62% of whom had ES-SCLC, detected a 1-year overall survival rate of 40% in early chemotherapy initiators (defined as a diagnosis-to-treatment interval less than or equal to the population median of 18 days) compared with 53% in late chemotherapy initiators, though this difference was not statistically significant (p = 0.08).19
The presence of unmeasured confounding variables may account for the observed link between earlier treatment and poorer survival in the VLCR data. Certain clinical factors, such as volume of disease, brain metastases, emergency hospitalization at presentation, frailty, and individual patient preferences, may influence survival but are challenging to record as covariates in clinical quality registries, including the VLCR. More specifically, a higher prevalence of unfavorable prognostic features in the early treatment cohort may have attenuated any apparent benefit of early treatment; however, the available data did not allow for more detailed exploration of this possibility. Similarly, hospital-related factors, such as coordination of medical care services and availability of clinicians, diagnostic investigations, and multidisciplinary team meetings, are not easily captured in large population-based studies and were not specifically accounted for in our analyses.
ECOG performance status, which is an established independent predictor of survival in lung cancer,20 was not recorded for more than 30% of patients in the total population of our study, representing a potential source of bias. Reassuringly, examination of participants with and without recorded ECOG data revealed no significant intergroup differences with respect to age, immunotherapy treatment, and comorbidities. We also conducted analyses using best-case and worst-case assumptions, along with complete case analyses for patients with missing ECOG scores. These methods supported a link between earlier management and inferior overall survival, strengthening the validity of our primary findings.
The “sicker quicker” phenomenon, which has been well described in lung cancer population-based studies,21,22 provides a possible explanation for the observed inverse relationship between treatment timeliness and overall survival. This phenomenon posits that patients who present with higher symptom burden, more extensive disease, or declining performance status are more likely to be treated with greater urgency but are destined to have poorer outcomes because of the severity of their illness. This may result in selection bias because of an overrepresentation of patients with inherently poorer prognoses in the cohorts that receive earlier diagnosis and treatment. This phenomenon may be especially pronounced in patients who present late in the course of their cancer. In these patients, treatment that is considered “early” relative to their presentation may in fact fall late in the context of their overall disease course, and may be initiated at a point that is too late to meaningfully alter the natural history of their cancer. Within the VLCR database, the lack of more detailed prognostic information at presentation, including tumor burden, brain metastases, and emergency hospitalization, restricted our ability to verify the sicker quicker phenomenon in our patient cohort. Availability of such data would enable clarification of any associations between adverse clinical features and speed of treatment onset, thus helping to ascertain whether early and late treatment initiators are fundamentally different in terms of prognosis.
Consistent with expectation, other factors independently associated with improved survival were younger age and immunotherapy treatment. Notably, most of the study period took place before immunotherapy was established as a standard of care in the first-line treatment of ES-SCLC by the IMpower133 study,3 which reported improved survival with the addition of atezolizumab to platinum-based chemotherapy. Hence, immunotherapy was administered to only 34.1% of the treatment timeliness population in the VLCR. We considered the possibility that more widespread use of immunotherapy might alter the identified relationship between timeliness and survival. However, in the subgroup of patients treated with immunotherapy, delayed initiation of treatment was still in keeping with improved overall survival, supporting the relevance of our results in the immunotherapy era.
Surprisingly, residences in rural and remote areas were associated with improved overall survival compared with residence in metropolitan centers. This observation is at odds with Australian nationwide data suggesting shorter life expectancy and higher prevalence of chronic disease burden with increasing remoteness.23 The reason for this is unclear, but it is plausible that in rural and remote areas, complex and seriously unwell patients are more likely to decline chemotherapy and instead accept palliative management than metropolitan patients, thereby selecting for a healthier chemotherapy-treated population in rural and remote settings.
Irrespective of the effect on survival, early diagnosis and treatment have other potential advantages in cancer management. In individuals presenting with symptomatic tumors, rapid commencement of treatment may provide benefit through early relief of cancer-related symptoms, enhancement of performance status, and improvement in quality of life, even if this does not translate to a delay in death. Furthermore, timely care may alleviate the patient’s distress that often accompanies perceived delays in cancer diagnosis and treatment initiation.24,25 We hypothesize that health-related quality of life would be improved by earlier initiation of treatment, but this question could not be specifically addressed by the data available in the VLCR, and thus requires verification in future studies incorporating longitudinal assessment of quality of life indices and patient-reported outcome measures.
This study has numerous strengths. The VLCR is a large, multicenter clinical quality registry and evidence repository containing data that spans more than a decade and provides unique insights into lung cancer management and patient outcomes in Australia.26 To our knowledge, there are no other published studies describing treatment timeliness and its impacts in ES-SCLC in Australia, and only a limited number of published works on this topic worldwide. Our study provides added nuance regarding the impact of diagnostic and treatment timeliness by examining the entire journey from initial referral-to-treatment initiation. Unlike previously published studies, which primarily focus on the interval between cancer diagnosis and chemotherapy commencement, our work also investigated the effect of the referral-to-treatment and referral-to-diagnosis intervals, given that a “timely” diagnosis-to-treatment interval may not represent timely overall care if there are delays between presentation and diagnosis.
Limitations of this work include the potential consequences of confounding variables not measured in the registry data. As described above, certain patient- and hospital-related variables, which are known to influence survival but are difficult to capture in registry data sets, are not represented in the VLCR. Furthermore, the VLCR does not include longitudinal metrics relating to symptom burden and cancer-related quality of life, which would have provided additional insights into the benefits of early diagnosis and treatment besides prolongation of survival.
In conclusion, in this study of ES-SCLC, most patients received timely care according to guideline-defined timeliness targets. However, earlier diagnosis and treatment initiation were associated with poorer overall survival, de-emphasizing prolongation of survival as the primary objective of timely management. Possible explanations for this include unmeasured confounding, the sicker, quicker phenomenon, and the notion that accelerating treatment alone is insufficient to overcome the natural history of more aggressive presentations of ES-SCLC. The potential—but yet unconfirmed—advantages of timely care on quality of life, symptom control, and performance status demand additional exploration in further studies.
CRediT Authorship Contribution Statement
Alison Hiong: Conceptualization, Methodology, Formal analysis, Writing – original draft, Writing – review & editing.
Julien Robinson: Conceptualization, Methodology, Writing – review & editing.
Eldho Paul: Methodology, Formal analysis, Writing – review & editing, Visualization.
Sanuki Tissera: Project administration, Writing – review & editing.
Jessie Zeng: Project administration, Writing – review & editing.
Craig Underhill: Writing – review & editing.
Sagun Parakh: Writing – review & editing.
Louis Irving: Writing – review & editing.
Wasek Faisal: Writing – review & editing.
Rob Blum: Writing – review & editing.
Phillip Parente: Writing – review & editing.
Michelle Caldecott: Writing – review & editing.
Javier Torres: Writing – review & editing.
Evangeline Samuel: Writing – review & editing.
Barton Jennings: Writing – review & editing.
Katharine See: Writing – review & editing.
David Langton: Conceptualization, Writing – review & editing, Project administration.
James Bartlett: Writing – review & editing.
Maggie Moore: Writing – review & editing.
Gary Richardson: Conceptualization, Writing – review & editing, Project administration.
Inger Olesen: Conceptualization, Writing – review & editing, Project administration.
Tom John: Conceptualization, Writing – review & editing, Project administration.
Marliese Alexander: Conceptualization, Writing – review & editing, Project administration.
Matthew Conron: Conceptualization, Writing – review & editing, Project administration.
Lisa Briggs: Conceptualization, Writing – review & editing, Project administration.
Tom Wood: Conceptualization, Writing – review & editing, Project administration.
Olivia Cassar Wong: Conceptualization, Writing – review & editing, Project administration.
Nicola Atkin: Conceptualization, Writing – review & editing, Project administration.
Susan Harden: Conceptualization, Writing – review & editing, Project administration.
Zoe McQuilten: Conceptualization, Writing – review & editing, Project administration.
John Zalcberg: Conceptualization, Writing – review & editing, Project administration.
Rob G Stirling: Conceptualization, Methodology, Writing – review & editing, Supervision, Project administration.
Disclosure
The authors declare no conflict of interest.
Footnotes
Cite this article as: Hiong A, Robinson J, Paul E, et al. Treatment timeliness in extensive-stage SCLC and impact on survival: a registry-based observational study. JTO Clin Res Rep 2026;7:100978
Note: To access the supplementary material accompanying this article, visit the online version of the JTO Clinical and Research Reports at www.jtocrr.org and at https://doi.org/10.1016/j.jtocrr.2026.100978.
Supplementary Data
References
- 1.Rossi A., Di Maio M., Chiodini P., et al. Carboplatin- or cisplatin-based chemotherapy in first-line treatment of small-cell lung cancer: the COCIS meta-analysis of individual patient data. J Clin Oncol. 2012;30:1692–1698. doi: 10.1200/JCO.2011.40.4905. [DOI] [PubMed] [Google Scholar]
- 2.Huang J., Faisal W., Brand M., et al. Patterns of care for people with small cell lung cancer in Victoria, 2011-19: a retrospective, population-based registry data study. Med J Aust. 2023;219:120–126. doi: 10.5694/mja2.52017. [DOI] [PubMed] [Google Scholar]
- 3.Horn L., Mansfield A.S., Szczęsna A., et al. First-line atezolizumab plus chemotherapy in extensive-stage small-cell lung cancer. N Engl J Med. 2018;379:2220–2229. doi: 10.1056/NEJMoa1809064. [DOI] [PubMed] [Google Scholar]
- 4.National Comprehensive Cancer Network NCCN clinical practice guidelines in oncology small cell lung cancer. version 4.2025. 2025. https://www.nccn.org/professionals/physician_gls/pdf/sclc.pdf
- 5.Bhandari S., Pham D., Pinkston C., Oechsli M., Kloecker G. Timing of treatment in small-cell lung cancer. Med Oncol. 2019;36:47. doi: 10.1007/s12032-019-1271-3. [DOI] [PubMed] [Google Scholar]
- 6.Bhandari S., Kumar R., Pham D., Gaskins J., Kloecker G. Treatment timing in small cell lung cancer, a national cancer database analysis. Am J Clin Oncol. 2020;43:362–365. doi: 10.1097/COC.0000000000000676. [DOI] [PubMed] [Google Scholar]
- 7.Altman D.E., Zhang X., Fu A.C., et al. Development of a conceptual model of the patient experience in small cell lung cancer: a qualitative interview study. Oncol Ther. 2023;11:231–244. doi: 10.1007/s40487-023-00223-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cancer Council Optimal care pathways. https://www.cancer.org.au/health-professionals/optimal-cancer-care-pathways
- 9.Partners R.M. National optimal lung cancer pathway (NOLCP) https://rmpartners.nhs.uk/our-work/improving-diagnostic-treatment-pathways/tumour-specific-cancer-pathways/lung-cancer/national-optimal-lung-cancer-pathway-nolcp/ [DOI] [PMC free article] [PubMed]
- 10.Stirling R.G., Evans S.M., McLaughlin P., et al. The Victorian Lung Cancer Registry pilot: improving the quality of lung cancer care through the use of a disease quality registry. Lung. 2014;192:749–758. doi: 10.1007/s00408-014-9603-8. [DOI] [PubMed] [Google Scholar]
- 11.Steindel S.J. International classification of diseases, 10th edition, clinical modification and procedure coding system: descriptive overview of the next generation HIPAA code sets. J Am Med Inform Assoc. 2010;17:274–282. doi: 10.1136/jamia.2009.001230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Oken M.M., Creech R.H., Tormey D.C., et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol. 1982;5:649–655. [PubMed] [Google Scholar]
- 13.Goldstraw P., Chansky K., Crowley J., et al. The IASLC lung cancer staging project: proposals for revision of the TNM stage groupings in the forthcoming (eighth) edition of the TNM classification for lung cancer. J Thorac Oncol. 2016;11:39–51. doi: 10.1016/j.jtho.2015.09.009. [DOI] [PubMed] [Google Scholar]
- 14.Australian Government Department of Health Disability and Ageing. Modified monash model. https://www.health.gov.au/topics/rural-health-workforce/classifications/mmm
- 15.Cushman T.R., Jones B., Akhavan D., et al. The effects of time to treatment initiation for patients with non-small-cell lung cancer in the United States. Clin Lung Cancer. 2021;22:e84–e97. doi: 10.1016/j.cllc.2020.09.004. [DOI] [PubMed] [Google Scholar]
- 16.Tsai C.H., Kung P.T., Kuo W.Y., Tsai W.C. Effect of time interval from diagnosis to treatment for non-small cell lung cancer on survival: a national cohort study in Taiwan. BMJ Open. 2020;10 doi: 10.1136/bmjopen-2019-034351. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Khorana A.A., Tullio K., Elson P., et al. Time to initial cancer treatment in the United States and association with survival over time: an observational study. PLoS One. 2019;14 doi: 10.1371/journal.pone.0213209. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Radzikowska E., Roszkowski-Sliz K., Chabowski M., Glaz P. Influence of delays in diagnosis and treatment on survival in small cell lung cancer patients. Adv Exp Med Biol. 2013;788:355–362. doi: 10.1007/978-94-007-6627-3_48. [DOI] [PubMed] [Google Scholar]
- 19.Tas F., Ozturk A., Erturk K. Timing of chemotherapy after diagnosis of small cell lung cancer. J Chemother. 2024;36:607–612. doi: 10.1080/1120009X.2024.2305062. [DOI] [PubMed] [Google Scholar]
- 20.Simmons C.P., Koinis F., Fallon M.T., et al. Prognosis in advanced lung cancer--a prospective study examining key clinicopathological factors. Lung Cancer. 2015;88:304–309. doi: 10.1016/j.lungcan.2015.03.020. [DOI] [PubMed] [Google Scholar]
- 21.Forrest L.F., Adams J., White M., Rubin G. Factors associated with timeliness of post-primary care referral, diagnosis and treatment for lung cancer: population-based, data-linkage study. Br J Cancer. 2014;111:1843–1851. doi: 10.1038/bjc.2014.472. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Klarenbeek S.E., Aarts M.J., van den Heuvel M.M., Prokop M., Tummers M., Schuurbiers O.C.J. Impact of time-to-treatment on survival for advanced non-small cell lung cancer patients in the Netherlands: a nationwide observational cohort study. Thorax. 2023;78:467–475. doi: 10.1136/thoraxjnl-2021-218059. [DOI] [PubMed] [Google Scholar]
- 23.Australian Government Australian Institute of Health and Welfare Rural and remote health. https://www.aihw.gov.au/reports/rural-remote-australians/rural-and-remote-health
- 24.Frosch Z.A.K., Jacobs L.M., O’Brien C.S., et al. “Cancer’s a demon”: a qualitative study of fear and multilevel factors contributing to cancer treatment delays. Support Care Cancer. 2023;32:13. doi: 10.1007/s00520-023-08200-9. [DOI] [PubMed] [Google Scholar]
- 25.Miles A., McClements P.L., Steele R.J., Redeker C., Sevdalis N., Wardle J. Perceived diagnostic delay and cancer-related distress: a cross-sectional study of patients with colorectal cancer. Psychooncology. 2017;26:29–36. doi: 10.1002/pon.4093. [DOI] [PubMed] [Google Scholar]
- 26.Stirling R.G., Samankula U., Lloyd M., et al. Impacts of a clinical quality registry on lung cancer quality measures: a retrospective observational study of the Victorian lung cancer registry. Clin Oncol (R Coll Radiol) 2025;44 doi: 10.1016/j.clon.2025.103878. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.