Review Reports - Adult Patients with Philadelphia-Positive B-Cell Acute Lymphoblastic Leukemia Treated with a Pediatric-Inspired Multiagent Chemotherapy Regimen, in Combination with a TKI, Do Not Require Routine alloSCT

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

see the file

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The authors provide a very interesting and highly significant report of their own experience with a pediatric-inspired multiagent chemotherapy regimen in adult patients with Philadelphia-positive B-cell acute lymphoblastic leukemia. They elegantly provide evidence that supports the omission of asparaginase and the omission of alloSCT (that of course need to be validated through RCTs), two highly effective treatment modalities but also associated with serious treatment related adverse events .

Author Response

Reviewer 2:

Thank you for reviewing our paper and for your comments

Reviewer 3 Report

Comments and Suggestions for Authors

Although the retrospective descriptive study drew some interesting takeaways for the readers regarding the feasibility of omitting alloHSCT and asparaginase in certain individuals with Ph+ ALL, the timeline event description was not clearly described, it will be more clear to have a diagram to depict all schema modifications as well as comparisons of end points against historical data. Some other comments below:

Page 1 line 33: spell out acronym for OS and RFS when first mentioned.

Page 2 line 48: “Increased toxicity observed early on when imatinib was added led to 47 the subsequent omission of asparaginase in 2009.”

What happened to patients’ toxicity profile when TKI such as imatinib was not added to the pediatric-inspired chemotherapy backbone prior to 2009?

How was this toxicity profile historically compared to Ph-negative ALL also using pediatric-inspired chemotherapy backbone?

How long (range, and median) did patients receive imatinib or dasatinib prior to alloHSCT?

Page 2, line 52: “Further, we observed that many patients for whom we were unable to find an alloSCT donor, were nevertheless achieving good long-term outcomes with-53 out consolidative alloSCT”.

Why was a plausible biological reason or clinical rationale that patients achieved good long-term outcomes even without alloHSCT at CR1?

What were the cytogenetic risk group and molecular lesions of these patients?

Was MRD testing available prior to 2009?

Page 6, line 165-167: “ Patients who underwent induction in 2001-2009, whether 165 they were (22 patients) or were not (28 patients) transplanted, had similarly unfavorable outcomes – 4-year OS 50.0% (95% CI, 28.2-58.4) versus 39.4% (95% CI, 21.1-57.3), and 4- 167 year RFS 50.0% (95% CI, 28.2-58.4) versus 35.8% (95% CI, 18.3-53.7), respectively”

Define what are similarly unfavorable outcomes.

Any explanation on why did the alloHSCT did not make a difference in survival outcomes? Differences in post-transplant morbidities and mortalities? Differences in supportive care?

Page 9, lines 209-210 : “The evolution of our practice – omission of asparaginase and alloSCT in CR1 – was based on excess toxicity and inferior outcomes observed in earlier years. This study examined if our protocol changes addressed these concerns.”

Explain what were the causes of excess toxicity and inferior outcomes?

Were they strictly related to the use of alloSCT and asparaginase?

Would the use of pegasparginase or even detection of silent mutation help mitigate the toxicity?

What age groups experienced the most toxicities? What are the risk groups in terms of molecular lesions and/or cytogenetics of those cohorts that have inferior outcomes?

Page 9, line 247: “Consequently, pegylated-asparaginase was removed from the UKALL14 induction protocol for patients with Ph+ B-ALL (Patel et al, 2017)… “

The authors categorize toxicity of asparaginase/pegasparginase to be under one group, will pharmacologic difference result in different toxicity profile among these two agents? Was this difference also reported in the literature? How would they affect the interpretation of the efficacy of this CR1 cohort with or without transplant?

What are the advantages of using this PM-DFCI derived regimen vs. hyperCVAD regimen?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

Synopsis:

In their manuscript “Adult patients with Philadelphia-positive B-cell acute lymphoblastic leukemia treated with a pediatric-inspired multiagent chemotherapy regimen, in combination with a TKI, do not require routine alloSCT,” Eng and colleagues retrospectively review the Princess Margaret Hospital clinical experience treating adults with Philadelphia chromosome-positive B-cell acute lymphoblastic leukemia (Ph+ B-ALL) with pediatric-inspired regimens incorporating imatinib in three therapeutic eras: 2001-09 in which asparaginase was a component of the regimen and patients routinely underwent allogeneic hematopoietic stem cell transplant (HSCT) in first complete remission (CR1); 2010-15 in which patients no longer received asparaginase but continued to undergo allogeneic HSCT in CR1; and 2016-19 in which patients no longer routinely underwent allogeneic HSCT in CR1.

For top-line results, Eng and colleagues report that among 157 adult patients with Ph+ B-ALL with a median follow-up of 41 months, relapse-free survival (RFS) at 4 years was 47.8% and overall survival (OS) was 54.9%. Comparing individuals who underwent HSCT in CR1 with those who did not among the entire study cohort, there were non-significant trends towards improved RFS and OS in the latter (RFS 50.9% vs. 43.1%, p=0.5272; OS 62.6% vs. 42.7%, p=0.0661). These results roughly align with other published reports of the combination of tyrosine kinase inhibitors and chemotherapy, lending credibility to the authors’ analyses.

When patients were evaluated according to therapeutically distinctive temporal bins (2001-09, n=50; 2010-15, n=58, 2016-19, n=31), individuals treated in the 2016-19 period demonstrated a trend towards improved RFS and a significant improvement in OS compared to the earlier time periods (hazard ratios 0.571, p=0.1012; and 0.394, 95% CI 0.180-0.864, respectively). The authors proceeded to increasingly finer bins and report that patients who initiated treatment in 2016-19 and forewent HSCT in CR1 had superior outcomes (4-year RFS 69.3%, OS 87.0%) compared to other temporally defined HSCT groups. Patients who initiated treatment in earlier periods had inferior RFS and OS, without significant differences between the 2001-09 and the 2010-15 groups based on asparaginase use when analyzed independent of HSCT, though nominal improvements in RFS and OS were observed among those who initiated treatment in 2010-15 without asparaginase and who did not undergo consolidate HSCT compared to those treated earlier with asparaginase. Within the earliest time period (2001-09) nominal differences in RFS and OS were observed according to HSCT, with non-significant trends towards improved survival among those who underwent HSCT in CR1.

Multivariable analysis revealed that correlates of RFS and OS were achieving at least a three log(10) reduction in BCR::ABL1 mRNA levels by 9 and 12 months of therapy (HR 0.38 for RFS and 0.35 for OS) and undergoing HSCT in CR1 (HR 1.73 for RFS and 2.19 for OS). Other clinical factors including age, sex, WBC count, prior cancer extramedullary or CNS disease, asparaginase use, temporal period, and BCR::ABL1 mRNA levels at 3 and 6 months did not demonstrate significant survival correlation in univariable analyses with variable statistical power or in multivariable analysis.

Summary Comments

As a retrospective, single institution experience seeking to address the issue of whether allogeneic HSCT in CR1 is necessary to achieve optimal long-term survival outcomes, this manuscript is methodologically sound, clinically nuanced, and well-written. While such studies necessarily encounter unmeasured confounding variables, particularly when conducted over such a two-decade time period during which many concurrent clinical advances have occurred, the authors have made appropriate efforts to illuminate their impacts through post hoc analyses and acknowledge those that are beyond the scope of this work in the discussion.

My most substantive critique of this work is its novelty and potential for clinical impact, given that the field has been moving towards lower intensity induction therapy for Ph+ ALL for more than ten years by now, with comparative studies like GRALL (Chalendon et al. Blood, 2015) supporting the use of lower intensity, vincristine-based chemotherapy with imatinib compared to hyper-CVAD and imatinib for a population with a similar age range (18-59). Moreover, a long line of studies from the Italian GIMEMA group and others have generated compelling data from chemotherapy-free approaches based on progressively more potent tyrosine kinase inhibitors (beginning with the LAL1205 study by Foà et al. Blood, 2011 and more recently the LAL2116 trial incorporating early blinatumomab in Foà et al. N Engl J Med, 2020). The data provided by the authors here remain valid, and their analysis is sound. Critically, the authors report that many Canadian centers continue to rely on multi-agent chemotherapy and imatinib for adults with Ph+ B-ALL. This makes the data provided herein relevant and potentially impactful, but for many clinicians who have long since adopted less intensive treatment regimens even for fit adults with Ph+ ALL, the practical impact may be blunted.

With respect to the specific question of the necessity of HSCT in CR1, the authors’ data are compelling that this does not correlate with improved outcomes following the pediatric-inspired multi-agent chemotherapy induction regimen in use at Princess Margaret Hospital and other Canadian centers. This is a live question in adult leukemia that investigators have sought to address for a number of years, particularly for patients who achieve measurable residual disease (MRD) negativity based on molecular assays with varying levels of sensitivity, which have deepened over time. The authors provide a very thoughtful handling of previously published studies in the discussion, pointing out limitations to extrapolating outcomes from heterogeneous trials utilizing different tyrosine kinase inhibitors and, in some cases, different chemotherapy backbones. All of this is true. It also relates back to the point made in the previous paragraph about the applicability of the data from the current study derived from patients treated at a single institution with a regimen that has been progressively less commonly used in other regions.

Finally, the authors provide data that adds to the body of literature supporting the value of molecular measurements of MRD in Ph+ ALL through multivariable analysis. Monitoring of MRD via molecular assays has been a guideline-supported practice for some time now and has been adopted widely. The authors appropriately maintained consistency in their use of a three-log(10) reduction in BCR::ABL1 mRNA given the temporal span of this study. There is value in this report showing that for a lower potency first generation kinase inhibitor like imatinib, even when combined with more intensive chemotherapy, the predictive value of BCR::ABL1 RT-qPCR is greatest at months 9 and 12 rather than at months 3 and 6. One wonders whether even greater accuracy would be obtained in the most recent temporal bin with a lower level of molecular MRD (e.g., 10^-4 as per modern guidelines). The lack of genetic data to assign IKZF1^plus status resulted from the fact that these tests were not routinely available or performed for the vast majority of this study duration, although that is an element that must be considered, as the authors touch upon in the discussion.

None of these critiques detract from the fact that retrospective single-institution studies make a unique contribution to broader clinical understanding, but the potential impact on practice is a consideration in the review. In the end, the authors addressed their aims to retrospectively explore the value of allogeneic HSCT in CR1, the inclusion of asparaginase in induction, and molecular monitoring of minimal residual disease (MRD) in adults with Ph+ ALL and package their data and valid conclusions in a well-written manuscript.

I list specific constructive critiques below:

Major:

Please confirm the numbers provided in lines 167 and 168. The exact same numbers are given for 4-year OS and 4-year RFS. This would be highly unusual. Please verify accuracy.
It would facilitate interpretation if Table 1 had columns for patients in the three different periods of time (2001-2009, 2010-2015, 2016-2019) as well as the combined cohort. Additionally, but less critically, since this study seeks to evaluate the necessity of transplant, a supplemental table of patient characteristics similarly stratified by whether or not patients underwent allogeneic HSCT in CR1 might also be helpful.

Minor:

Please specify in the methods whether BCR:ABL1 mRNA was measured in the bone marrow or the peripheral blood. If this was stated, I missed it.
There may be a typo in line 133 where it says, “In 2001-2009, all 35 of patients received. After 2009, no patients received asparaginase.” The first sentence appears to have been truncated. Please correct.
In Table 1, it would help to have more granularity in the distribution of patients with previously reported prognostic factors. For example, for WBC count, the median value has limited utility by itself. It would be helpful to know how many patients had WBCs in different clinically relevant ranges (e.g., the percent of patients with WBC>30k).
Please correct the following typos in line 152: “…patients treated in 2010-2015 were no [sic] significant [sic] different to patients treated in 2001-2009.”
The Kaplan-Meier plots use red and green, which are difficult to distinguish for colorblind readers. Consider a more colorblind-friendly palette.
Even though this is a retrospective series, it would enhance the prospective relevance of this study if the authors could comment briefly in the discussion on the sensitivity of molecular MRD analysis in the clinical calculus of whether to recommend a patient for transplant.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf