To Delay or Not to Delay, That Is the Question in ALL Patients Who Do Not Receive Prophylactic Cranial Irradiation

Condensed abstract:

To determine the impact of delaying diagnostic lumbar puncture and initial intrathecal therapy on treatment outcome of children with newly diagnosed acute lymphoblastic leukemia (ALL) treated without prophylactic cranial irradiation, Yeh and colleagues updated their TPOG-ALL-2002 clinical trial. They found that this treatment strategy reduced the rate of traumatic lumbar puncture with blasts to zero, and tended to improve outcome of patients with a CNS-2 or CNS-3 status based on the event-free survival, the overall survival, and the cumulative risk of CNS relapse, as compared to historical controls.

With 5-year survival rates of childhood acute lymphoblastic leukemia (ALL) approaching and even exceeding 90% in many developed countries,¹ current efforts are focused on improving not only the cure rate but also the quality of life of the patients by reducing treatment-related toxicities. Since prophylactic cranial irradiation can lead to many serious long-term sequelae, including secondary cancers, neurocognitive impairment and endocrinopathy,² eliminating the use of this treatment modality has been a priority for leukemia therapists worldwide. Two clinical trials in the previous decade achieved excellent 5-year event-free survival rates of 86% and 81% after totally omitting prophylactic cranial irradiation and relying on intensive dexamethasone, vincristine and asparaginase, high-dose methotrexate, and, for patients with blasts in cerebrospinal fluid, extra doses of intrathecal therapy.^3,4 Importantly, isolated and combined central-nervous-system (CNS) relapse rates in these studies were only 2.7% and 1.2%, and 2.6% and 0.8%, respectively.

Encouraged by these early findings, most study groups now restrict prophylactic cranial irradiation to patients at high-risk of CNS relapse, and some, such as the Taiwan Pediatric Oncology Group (TPOG), as reported in this issue of the Journal,⁵ have eliminated this treatment modality altogether.^5–12 Comparison of contemporary trials of limited use versus complete omission of prophylactic cranial irradiation shows similar rates of event-free survival and CNS control (Table 1). The likelihood of 5-year event-free survival and the cumulative risk of an isolated CNS relapse ranged from 72% to 89% and 1% to 4.1%, respectively, in the 4 studies that omitted irradiation,^5,10,13,14 compared with 85% to 87.3% and 1.4% to 2% in the 3 trials in which up to approximately 10% of the patients were irradiated,^7,8,12 and with 82.5% to 85% and 0.9% to 1.3% when more than 10% were irradiated.^6,9,11 Indeed, a 2016 meta-analysis of more than 16,000 patients enrolled in clinical trials of 10 cooperative childhood ALL study groups indicated that prophylactic cranial irradiation has no impact on the overall survival or cumulative risk of death, any adverse event, isolated CNS relapse, or combined CNS relapse in any subgroup of high-risk patients, with the exception of those with overt CNS leukemia (CNS-3 status, ≥ 5 leukocytes/µL with blasts).¹⁵ Although cranial irradiation was associated with a reduced risk of isolated CNS relapse in patients with CNS-3 status, it failed to reduce overall adverse events, hematologic relapses, or mortality rates in these patients.¹⁵

Table 1.

Characteristics and treatment outcome of selected contemporary clinical trials

Group Trial (years of enrollment)	Total No Patients	Age at diagnosis (years)	No. (%) of Patients receiving prophylactic cranial irradiation; Indications	% 5-year event-free survival (±SE)	% Cumulative risk of isolated and combined CNS relapse
AIEOP-BFM ALL 2000 (2000–2006)6	3720	1–17	(18%); 12 −18 Gy for prednisone-poor- response, no complete remission on day 33, presence of BCR-ABL1, KMT2A-AFF1, or MRD ≥5 × 10−4 on day 78, or age 1–2 years with CNS-3; 24 Gy for age >2 years with CNS-3	83.9% (0.9%) for dexamethasone- treated patients; 80.8% (0.9%) for prednisone- treated patients	0.9% isolated CNS relapse and 0.7% combined CNS relapse for dexamethasone-treated patients; 1.9% isolated CNS relapse and 1.5% combined CNS relapse for prednisone- treated patients
COG AALLO331 for standard-risk and AALL0232 or NCI high-risk B-ALL (2003- 2011)7	8379	1–30	645 (7.7%); 18 Gy for 127 patients with CNS-3 status, 12 Gy for 518 with NCI high-risk ALL with slow early response; no modification of therapy for CNS-2 status	85% (0.6%) for 7214 CNS-1, 76% (2%) for 836 CNS-2, and 76% (5%) for 124 CNS-3 patients	2%, 5.6% and 5.1% isolated CNS relapse for CNS-1, CNS-2 and CNS-3 patients, respectively (P<0.001); 2.8%, 7.7% and 5.1% of combined CNS relapse for CNS-1, CNS-2, and CNS-3 patients, respectively (P<0.001)
DCOG-10 (2004–2012)8	778	1–18	5 (0.6%); 12 Gy for age >3 years with high-risk features such as no completion remission after induction, MRD ≥5 x 10−4 at day 78, presence of KMT2A- AFF1	87.0% (1.2%)	1.4 % isolated CNS relapse and 0.9% combined CNS relapse
DFCI 05–001 (2005–2010)9	551	1–18	~127 (~23.5%); 12 Gy for T-ALL or B-ALL with leukocyte count ≥100 x 109/L; 24 Gy for CNS-3 status	85% (1.5%)	~1.3% isolated CNS relapse and ~1.1% combined CNS relapse
EORTC CLG 58951 (1998- 2008)10	1947	0–17	None	82.6% (0.9%)	1.7% isolated CNS relapse and 1.8 % combined CNS relapse
JCCLSG ALL 2000 MRD study (2000- 2004)11	229	1–15	98 (42.8%); 18 Gy for patients with leukocyte count >100 × 109/L, age ≥10 years, or age 6–9 years with leukocyte count >5 × 109/L	82.5% (2.6%)	0.9% isolated CNS relapse and 1.5% for combined CNS relapse
MRC UKALL 2003 (2003- 2011)12	3126	1–24	49 (2%); 24 Gy for patients with CNS 3 status	87.3% (1.4%)	1.9% isolated CNS relapse
NOPHO ALL 2008 (2008- 2014)13	1288	1–17	None	89% (1%) for age 1–9 years; 80% (3%) for age 10–17 years	1% isolated CNS relapse and ~1% combined CNS relapse for age 1–9 years; 3% isolated CNS relapse and ~0.5% combined CNS relapse for age 10–17 years
SJCRH XV (2000–2007)14	498	1–18	None	87.3% (2.9%)	2.7% isolated CNS relapse and 1.2% combined CNS relapse
TPOG-ALL- 2002 Era 2 (2009–2012)5	444	0–18	None	72.0% (2.5%)	4.1% isolated CNS relapse and 1.3% combined CNS relapse

Abbreviations: ALL, acute lymphoblastic leukemia; AIEOP, Associazione Italiana di Ematologia Pediatrica Group; BFM, Berlin-Frankfurt-Münster; CNS, central nervous system; COG, Children’s Oncology Group; DCOG, Dutch Children’s Oncology Group; DFCI, Dana-Farber Cancer Institute consortium; EORTC – CLG, European Organisation for Research and Treatment of Cancer - Children Leukemia Group;JCCLSG, Japanese Children’s Cancer and Leukemia Study Group; MRC UKALL, Medical Research Council United Kingdom acute lymphoblastic leukemia; NOPHO, Nordic Society of Pediatric Hematology and Oncology; SJCRH, St. Jude Children’s Research Hospital; TPOG, Taiwan Pediatric Oncology Group; NCI, National Cancer Institute; MRD, minimal residual disease.

Although intensive systemic and intrathecal therapy can replace prophylactic cranial irradiation even in patients at high risk of CNS relapse, it is associated with acute and late sequelae (although less serious than those induced by irradiation), and could poise difficult challenges to therapists working in developing countries with limited resources. It should also be stressed that even with regimens of intensive systemic and intrathecal chemotherapy, findings of a CNS-3 or CNS-2 status (< 5 leukocytes/µL with blasts), or traumatic lumbar puncture with blasts (≥ 10 erythrocytes/µL with blasts; TLP+) were still linked to poor event-free survival and overall survival as well as an increased risk of CNS relapse.^3,7

Recognizing that TLP+ is associated with an inferior treatment outcome,¹⁶ and that delaying the diagnostic lumbar puncture with intrathecal chemotherapy to day 8 of remission induction as tested in the Tokyo Children’s Cancer Study Group L89–12 study can almost completely eliminate the risk of TLP+ (0.6%),¹⁷ investigators at Mackay Memorial Hospital in Taiwan pioneered an innovative study of CNS-directed treatment, in which diagnostic lumbar puncture and triple intrathecal therapy were delayed until the clearance of circulating blasts but no later than day 10 of remission induction, with elimination of prophylactic cranial irradiation.¹⁸ The success of this single-institution pilot study of 156 patients, which yielded a 5-year event-free survival of 84.2%±3.0% with only 2 patients developing a combined CNS relapse and none an isolated CNS relapse prompted the revision of their ongoing TPOG-ALL-2002 multi-institutional study in 2009 to adopt this method of CNS-directed treatment.¹⁹ The revised protocol (n= 456 patients treated) yielded a 5-year event-free survival of 71.4%±2.6%, a 5-year overall survival of 77.8%±2.4%, an isolated CNS relapse rate of 4.3%±1.1%, and an any-CNS relapse rate of 5.8%±1.3%, results that were not significantly different from those of the first 913 patients treated on the same protocol with prophylactic cranial irradiation, before the revision.¹⁹

Although delaying initial intrathecal therapy in this study reduced the risk of TLP+, there was concern that systemic chemotherapy given prior to diagnostic lumbar puncture might have eradicated or reduced leukemia cells in the cerebrospinal fluid, masking the diagnosis of CNS-2 or CNS-3 in patients who otherwise would have received intensified CNS-directed treatment, and that the delay in intrathecal therapy might have also compromised the outcome of patients diagnosed to have a CNS-2 or CNS-3 status. To address these issues, Yeh and colleagues⁵ updated their TPOG-ALL-2002 results in this issue of Cancer, limiting them to the 903 and 444 patients with known CNS status before and after the protocol revision, respectively. They found that the rate of TLP+ was indeed reduced, from 1.8% (16 of 903) before the revision to zero (0 of 444) afterwards; however, there was no significant difference in the frequency of patients with CNS-2 or CNS-3 status before and after the revision (6.6% vs. 6.3%, P=0.81). Moreover, the presenting clinical and biologic features were not significantly different between patients with a CNS-2 or CNS-3 status diagnosed before or after revision, suggesting that delaying intrathecal therapy did not select for higher-risk patients. More importantly, despite the elimination of prophylactic cranial irradiation, the clinical outcome in patients with a CNS-2 or CNS-3 status who were treated after the revision tended to fare better than those treated earlier: 5-year event-free survival (62.9%±9.4% vs. 52.3%±5.8%, P=0.199), 5-year overall survival (77.9%±8.0% vs. 61.8%±5.7%, P=0.079), 5-year cumulative risk of isolated CNS relapse (3.6%±3.5% vs. 6.3%±3.1%, P=0.639), and 5-year cumulative risk of any CNS relapse (3.6%±3.5% vs.17.3%±4.8%, P=0.096). Finally, there were no significant differences in outcome for the entire cohort before and after revision: 5-year event-free survival (75.7%±1.4% vs. 72.0%±2.4%, P=0.260), 5-year overall survival (83.0±1.3% vs. 79.4%±2.1%, P=0.069), 5-year cumulative risk of isolated CNS relapse (4.0%±0.7% vs. 4.1%±1.0%, P=0.960), and 5-year cumulative risk of any CNS relapse (6.3%±0.8% vs. 5.4%±1.2%, P=0.563). Investigators of TPOG concluded from their findings that delaying diagnostic lumbar puncture and initial intrathecal therapy with total omission of prophylactic cranial irradiation did not compromise treatment outcome in children with ALL.

Although Yeh et al.⁵ are to be commended for their update of the TPOG-ALL-2002 study, the extent of the impact of this modified CNS-directed treatment on clinical outcome remains in question. First, the trial was not randomized and the proportion of patients in the later cohort who had their diagnostic lumbar puncture and initial intrathecal therapy delayed because of the presence of circulating blasts at diagnosis was not known. Hence, even though the presenting features were not significantly different between the two cohorts in the TPOG study, outcome should also have been compared between patients who had the procedures delayed and matched controls of the earlier cohort. Second, the overall results of the revised TPOG protocol were inferior to those of the other 3 studies in which prophylactic cranial irradiation was entirely omitted (Table 1); however, MacKay Memorial Hospital, one of the participating centers of the TPOG trial, achieved a superior outcome among the participating centers: with a 5-year event-free survival, 84.5%±2.7%; 5-year overall survival, 91.8%±2.0%; isolated CNS relapse rate, 0.6%±0.6%; and isolated plus combined CNS relapse, 2.3%±2.1%. Yeh and colleagues⁵ attributed this excellent outcome to the implementation of a number of supportive care measures and improved compliance, a factor that could not be adjusted for the combined study. Finally, TLP+ status was reported in only 1.8% of the patients (n=16) in the earlier cohort of the TPOG study but in 4.7% and 5.6% of patients in other reported studies.^3,7

These caveats aside, TPOG investigators should be congratulated on their pioneering study of delaying diagnostic lumbar puncture until the clearance of blasts together with the elimination of prophylactic cranial irradiation. By abolishing the TLP+ status (and hence the need for extra intrathecal therapy) and by omitting cranial irradiation, they should be able to secure an improved quality of life with less treatment-induced neurocognitive impairment and other toxicities in their patients.²⁰ Until additional studies can establish the positive impact of this approach on outcome, it would seem reasonable to delay diagnostic lumbar puncture and initial intrathecal therapy until blasts are cleared in patients presenting with hyperleukocytosis. This strategy has the additional benefit of (i) avoiding methotrexate toxicity due to poor clearance in patients presenting with acute kidney injury and (ii) circumventing the sedation of patients with a mediastinal mass or other acute illness at diagnosis of ALL. Most important, perhaps, it will overcome several obstacles facing newly diagnosed ALL patients in developing countries with limited resources.